Nuclear RPSA senses viral nucleic acids to promote the innate inflammatory response

Innate sensors initiate the production of type I interferons (IFN-I) and proinflammatory cytokines to protect host from viral infection. Several innate nuclear sensors that mainly induce IFN-I production have been identified. Whether there exist innate nuclear sensors that mainly induce proinflammatory cytokine production remains to be determined. By functional screening, we identify 40 S ribosomal protein SA (RPSA) as a nuclear protein that recognizes viral nucleic acids and predominantly promotes proinflammatory cytokine gene expression in antiviral innate immunity. Myeloid-specific Rpsa-deficient mice exhibit less innate inflammatory response against infection with Herpes simplex virus-1 (HSV-1) and Influenza A virus (IAV), the viruses replicating in nucleus. Mechanistically, nucleus-localized RPSA is phosphorylated at Tyr204 upon infection, then recruits ISWI complex catalytic subunit SMARCA5 to increase chromatin accessibility of NF-κB to target gene promotors without affecting innate signaling. Our results add mechanistic insights to an intra-nuclear way of initiating proinflammatory cytokine expression in antiviral innate defense.

13.) Please include cytokine levels for mock infections / stimulations where possible, e.g. in Fig. 1a, 3a, 5b, 6l etc. 14.)The lay-out of some of the graphs could be more consistent, e.g.uninfected -infected, DMSO -inhibitor (e.g. in Fig. 3).The labelling in Fig. 6 is very small and unclear.15.) Fig. 5 which shows an involvement of p65 could be moved to earlier (after Fig. 3), as many of the genes shown are well known to be NF-kB dependent.
Reviewer #3: Remarks to the Author: Jiang et al. report on the role of RPSA during innate immune responses, finding the protein acts as a nuclear sensor of viral nucleic acids that in turn modulates gene regulatory responses as part of the innate immune response.More specifically, they identify that virally-mediated phosphorylation of residue Tyr204 on RPSA causes recruitment of the SMARCA5/ISWI complex to NFkB proinflammatory promoters independent of p65 activation to increase chromatin accessibility and gene transcription.This is a compelling paper with a lot of data, and potentially identifies an important, alternative host defense sensor and pathway that appears independent of traditional pattern recognition receptor pathways but nonetheless modulates their activity by integrating regulatory information at the site of chromatin.While the potential significance of the work is high, and many of the experiments show promising and interesting results, some of the work and its interpretation are stretched thin at times leading to unsubstantiated conclusions.Suggestions to improve the manuscript: One of the main issues with the manuscript is that it mixes and matches data types, experiments, and interpretation.For example, several NGS profiling experiments are performed in genome-wide fashion (e.g.ATAC-seq, H3K4me3 ChIP-seq, p65 ChIP-seq) while several others are not (ChIP q-PCR for RPSA, SMARCA5, RNAPII).Given the claims of the authors, I would highly recommend performing ChIP-seq with RPSA and SMARCA5 to fully understand the extent to which these proteins co-occupy chromatin and at which targets they localize to -for example, how specific are RPSA/SMARCA5 for NFkB targets?A more integrated view of how NFkB, RPSA, SMARCA5, and chromatin accessibility relate to one another would help solidify the mechanistic conclusions put forth by the authors.
With respect to the text near the call-out to Fig. 3c/d, the authors claim that Rpsa-iKO have less DNA accessibility near TSS regions, but Fig. 3c does not seem to convincingly indicate this -at least it is not clear looking at the figure.Also, it is very odd that DNA promoter accessibility is shown to decrease in both WT and Rpsa-iKO samples during HSV infection even though the promoters of the genes shown are dramatically up-regulated (Fig. 3d).These observations are extra confusing because it seems that the Rpsa-iKO samples show MORE DNase accessibility on pro-inflammatory promoters, where the model proposed would suggest there should be less.I would highly recommend expanding the genome-wide profiling to include ATAC-seq at both control and infected conditions to include a more comprehensive analysis to make this data more robust instead of relying on the candidate promoter DNase hypersensitivity data.
A more detailed description of the RNA-seq data is needed.For example, Supp.Fig 3b is hard to process as it contains a mix of up-and down-regulated genes relative to each of the experimental groups mentioned.I might include a global clustering/heatmap of the experiment (like shown in Fig. 2e) that provides a birds-eye-view of the gene expression changes that occur in the experiment to provide additional context.For example, it seems clear that many proinflammatory genes have reduced induction when Rpsa is depleted, but are there also a significant number of genes that are perhaps hyper-induced too?
In Figs. 1 & 2 the authors make the claim that RPSA is a nuclear sensor for viral RNA/DNA.While several insightful experiments were performed (i.e.nuclear transfection of dsDNA), additional controls should also be considered.For example, is RPSA functional during the innate response to non-nuclear viruses (e.g.flaviviruses) or other pattern recognition receptor signals/NFkB signaling pathways (e.g.Tlr4 pathway activation/TNF)?These would appear to be important controls.Fig. 2g-h help address this, but additional stimuli would help make this more convincing.In addition, given the model that RPSA is needed to help increase pro-inflammatory gene expression during these responses, what is compensating its role in driving inflammatory gene expression in the case of cytoplasmic activation by e.g.polyIC where RPSA is presumably not activated?
The following statement is not well supported in the literature: "Chromatin remodeling complexes increase chromatin accessibility to maintain transcriptional activity, which are marked by trimethylation of histone H3 at lysine 4 (H3K4me3) around TSS of target genes".Certainly, H3K4me3 is associated with gene TSS and active transcription, but H3K4me3 is not in and of itself proof of ISWI chromatin remodeling, as many locations where chromatin remodelers are recruited (i.e.enhancers) do not necessarily have this chromatin mark.Does Fig. 3f show all genes in the genome?Would this imply there is a global decrease in H3K4me3?How does that impact the interpretation of the data?
The ChIP-seq profile shown in Fig. 5d is a bit concerning -most ChIP-seq data for transcription factors, including previously published p65 ChIP-seq experiments, show maximal enrichment for the TF just upstream of the TSS in the open chromatin region of the promoter.These experiments seem to show a distribution of ChIP-seq signal just downstream of the TSS, more similar to that of a ChIP-seq experiment like H3K4me3 or H3K27ac.
The authors mention a multitude of targets from their proteomics data, but do not report the data itself.Not only does this limit the interpretability of their results, but it is also unclear exactly how the interaction with SMARCA5 fits into those results -was it by far one of the strongest interactions, or are there a multitude of other proteins that interact with RPSA with similar strength?
The association of RPSA with SMARCA5 is intriguing and presents the exciting possibility that RPSA is critically involved in DNA accessibility.However, one problem is that DNA accessibility ( ATACseq), RNAPII and TF recruitment, and chromatin marks like H3K4me3 are often highly correlated with one another and transcription.As a result, it can be hard to definitely attribute specific mechanisms when looking at changes in these data types, particularly when they change in unison as e.g. a loss of ATAC-seq signal could simply mean that the transcription of the gene is down.In addition to thinking about how the nuance that this might introduce into their interpretation of the results, I might recommend the authors consider a more targeted in vitro chromatin remodeling assay to assess RPSA's role in modulating ISWI activity.
Have the authors considered looking at these mechanisms regulate gene expression on the viral genome?Does RPSA/SMARCA5 etc. get recruited the HSV genome?Does the RPSA-iKO show strong modulation of viral gene expression for HSV or IAV?
Another suggestion for a relatively complicated manuscript such as this is to include a schematic/model that helps summarize the findings as a final figure panel or supplementary figure .Minor comments: I would consider rewording the sentence in the abstract: "nuclear short way" does not seem very concise -I recommend replacing the term with a better description.

Many mentions of "viral nuclear acids" -maybe the authors mean viral nucleic acids?
Consider including MDA/IFIH1 in the 3rd sentence for more completeness.
Although the writing in the manuscript is relatively easy to understand, there are lots of minor language problems -needs to be carefully edited before publication.
Might be nice to comment on RPSA's previously known roles and alternative functions (and maybe how the protein was originally named).

Point-by-point Response
To Reviewer #1: In the present study, Jiang and colleagues suggest RPSA as a nuclear sensor of viral nucleic acids and propose a mechanism of RPSA-induced virus-infection triggered proinflammatory cytokine activation in cells and in vivo in mice.The authors demonstrated association of RPSA with viral nucleic acids upon infection and showed direct binding of RPSA to HSV-1 DNA by various methods supporting the idea that RPSA is a direct sensor of nuclear replicating viruses.Virus-infection induced phosphorylation at residue Tyr204 promotes RPSA interaction with the chromatin remodeling complex SMARCA5, therefore increasing chromatin accessibility and subsequent proinflammatory cytokine expression.In summary, the authors identified a novel nuclear innate sensor and proposed a unique and interesting short-cut mechanism enhancing the inflammatory response through epigenetic modification by RPSA.This is a carefully executed study with conclusions well supported by the data.
The findings are novel and of high interest to a wider audience.The manuscript provides sound evidence for the identified sensor and propose a unique mechanism for immune defense.To my point of view, there is no additional evidence needed.The following points could be considered to make an excellent study even more convincing.
Response: Thanks for the positive comments.Following the comments and insightful suggestions, we provided additional data to address the concerns and further improve the quality of this work.

1) The authors show convincing results on RPSA binding to nucleic acids.
Furthermore, infection induced Tyr204 phosphorylation followed by binding to

SMARCA5. It would be interesting to know whether binding of viral nucleic acids induce the phosphorylation event? This would connect the two major findings.
Response: Thanks for the insightful suggestions.Our previous results in Fig. 4f and g 2 / 19 showed that the nucleus-located RPSA was only phosphorylated upon HSV-1 or IAV infection, suggesting the presence of viral nucleic acids is especially important for the phosphorylation event of RPSA.To further address this critical concern, we established the cytoplasm replicated virus VACV (Vaccinia virus) infection model, and our results showed that VACV infection did not enhance the phosphorylation of nuclear RPSA (Revised Supplementary Fig. 6g).Moreover, we also stimulated RAW264.7 cells with cGAMP to initiate innate signals from the cytoplasm then tested the phosphorylation level of nuclear RPSA, and found cGAMP stimulation could not induce such a PTM as HSV-1 or IAV virus infection induced (Revised Supplementary Fig. 6e).Taken together, our results show that the nuclear RPSA binding of viral nucleic acids mediates RPSA phosphorylation and subsequently boosts transcription events.

2) RPSA promotes proinflammatory cytokine gene transcription upon infection of nuclear replicating viruses, Influenza and HSV. So, both, viral DNA and RNA seem to
be recognized by RPSA.The authors should consider proving binding of RPSA to RNA/dsRNA, not only to dsDNA or could discuss their hypothesis on how RPSA could sense both RNA and DNA.
Response: Our previous RNA immunoprecipitation analysis in the nuclear fraction of IAV-infected BMDM indicated the interaction between IAV genome RNA (gRNA) and nuclear RPSA (Fig. 1e).To further clarify the direct interaction between IAV genomic RNA and RPSA, we labeled IAV gRNA with biotin, then the RNA pulldown assay was carried out in Revised Fig. 1l.Our results showed the direct interaction between RPSA and viral gRNA.Strikingly, the immunofluorescence assay exhibited that RPSA co-localized with EU-labeled IAV gRNA within A549 cell nucleus (Revised Fig. 1g).So, we conclude in the Results of our revised manuscript that nuclear RPSA is a viral nucleic acids-binding molecule.
Minor point: Nucleic acid is misspelled in the title and most part of the manuscript.

To Reviewer #2:
In this manuscript, the author investigate the role of the ribosomal protein RPSA in the activation of inflammatory gene expression after virus infection.The authors provide evidence that RPSA enhances the expression of pro-inflammatory cytokine genes driven by NF-kB after infection with HSV-1 and influenza A virus, while not affecting the interferon response.It is shown that the expression of pro-inflammatory cytokines is enhanced via RPSA interacting with SMARCA5 and enhancing chromatin accessibility for NF-kB p65.The authors propose that RSPA acts as a direct sensor of HSV-1 DNA (and possibly influenza virus RNA) as well as an effector which promotes chromatin accessibility.However, as it stands it is unclear how the proposed sensing activity integrates with well-established nucleic acid sensors during viral infection.
Overall, the authors provide convincing evidence that RPSA promotes virus-induced cytokine expression by enhancing chromatin accessibility.While the authors provide evidence that RPSA can bind DNA, it is currently not entirely clear that DNA binding can be equated with sensing per se.However, even if further experiments were to show that RPSA transduces the signal from STING or MAVS for instance to promote NF-kB-mediated transcription, rather than acting as a sensor in its own right, this would still be an important finding in the field.
Response: Many thanks for the insightful suggestions, which are very helpful to us for further improving the quality of this work.To clarify the uncertainties, we performed additional experiments to address the concerns and tried efforts to make this work more compelling.We provided a set of experimental evidence to elucidate that RPSA senses nucleus-located pathogenic DNA and RNA and then boosts proinflammatory cytokine gene expression at epigenetic levels (Revised Fig. 1b, 1c 3g, Supplementary Fig. 5a, 5c, 5d, Supplementary Fig. 6e, 6g and Supplementary Fig. 7b).RPSA enhances proinflammatory cytokine gene expression depending on NF-κB activation, thus compensating the established cytoplasm innate signal cascade.

Specific points:
1.) Evidence should be provided at the start that the absence of RPSA does not affect HSV-1 and IAV-1 entry and replication etc.For instance, viral mRNAs could be quantified alongside the qPCR analysis in Fig. 1 a-c

Response:
We quantified viral gDNA levels in peritoneal macrophages upon HSV-1 infection and demonstrated that the knocking down RPSA did not affect HSV-1 entry (Revised Supplementary Fig. 1c).We also quantified both HSV-1 and IAV mRNA levels in the RPSA deleted-and control A549 cells upon infection.These results showed that RPSA deficiency did not affect intracellular viral entry and replications (Revised Fig. 1b-c).

2.) It is shown that RPSA affects the expression of pro-inflammatory cytokines during both HSV-1 and IAV infection. Does RPSA also bind IAV RNA, and co-localise with IAV viral factories, as would be expected if it indeed acts as a sensor?
Response: Our previous RNA immunoprecipitation analysis in the nuclear fraction of IAV-infected BMDM indicated the interaction between IAV gRNA and nuclear RPSA (Fig. 1e).To further clarify the direct interaction between IAV genomic RNA (gRNA) and RPSA, we labeled IAV gRNA with biotin.Then the RNA pulldown assay was carried out, as shown in Revised Fig. 1l.Our result showed the direct interaction between RPSA and IAV gRNA.Strikingly, the immunofluorescence assay exhibited that RPSA co-localized with EU-labeled IAV gRNA within A549 cell nucleus (Revised Fig. 1g).
3.) Currently there is not much information on the DNA (and/or RNA) binding activity in the RSPA protein, which would allow to dissect its role in the sensing pathway.
Could structural predictions be used to identify a binding site and predict

/ 19
residues/fragments for mutagenesis analysis?Response: Our previous data have shown that both the N-terminal and C-terminal of RPSA were necessary for its binding to HSV-1 gDNA (Fig. 1m).We tried to predict the interface between RPSA and HSV-60 dsDNA using HDOCK server  4.) In Fig. 2a-b, it is shown that loss of RPSA decreases cytokine mRNA and protein levels, but does not abolish the induction completely.It would be interesting to see whether the cytokine response to HSV-1 in these is completely or partially cGAS-and STING-dependent (using deletion cells, RNAi or inhibitors).This would provide some clues as to whether there is redundancy between two sensors, or whether RPSA boost responses that are mainly STING-dependent.RPSA depends on and made up for the cytoplasm innate signal.Strikingly, RPSA overexpression no longer rescued the proinflammatory gene expression when we used the combined inhibitors.Moreover, inhibition of STING in peritoneal macrophages partially decreased cytokine mRNA level, and the knocking down of RPSA further reduced the mRNA level (Revised Supplementary Fig. 5a).Collectively, our data indicated that RPSA is partially dependent on the innate signal cascade from the cGAS-STING pathway.Importantly, these two sensor pathways play nonredundancy roles in defending against viral infection.6.)The experiments with nuclear and cytoplasmic transfection in Fig. 2f-h are misleading.For "nuclear" transfection, only dsDNA is used, while "cytoplasmic" transfection the stimulus is CpG DNA and poly (I:C).The same stimuli should be used with the two different transfection methods to show whether there are any differences in terms of localisation, rather than nucleic acid type.
In consistent with our previous conclusions, the presence of viral nucleic acids in the host nucleus is necessary for activating RPSA.

7.) In this context it would also be interesting to examine whether cytokine production in response to a STING agonist is affected. If it is, this would support a pathway
where RPSA is activated as another effector mechanism of STING signalling, rather than a sensor (which would also be interesting).

Response:
We stimulated RPSA-stable overexpressed RAW264.7 and the wild-type cells with cGAMP and then determined proinflammatory cytokine gene expression.
The results showed that the loss of RPSA did not affect cGAMP-induced cytokine mRNA expressions (Revised Supplementary Fig. 3g).

8.) The fractionation in S4b is not convincing -maybe immunofluorescence could be used to quantify p65 translocation.
Response: We checked P65 translocation upon HSV-1 stimulation using the immunofluorescence method, as shown in Revised Supplementary Fig. 4c.

9.) The phosphorylation of RPSA is intriguing, and points towards the involvement of additional proteins in the activation/regulation of this response. To integrate RPSA into known signalling cascades phosphorylation could be tested in dependence of STING signalling (for HSV-1) or MAVS (for IAV), using inhibitors, RNAi or deletion
infection neither enhanced the interactions among RPSA, SMARCA5 and P65 (Revised Supplementary Fig. 7a), demonstrating that sensing of pathogenic nucleic acids in the host nucleus is essential for activating RPSA.Together with the direct interaction between RPSA and viral nucleic acids, our results demonstrate that nuclear RPSA acts as a sensor that accelerates proinflammatory cytokine gene expression under an epigenetic mechanism.

11.) Additional experiments with STINg inhibitor (or deletion) and Takinib separately would be more informative than the current "pan-inhibitor" experiments.
Response: Many thanks for the suggestion.Initially, we wondered whether RPSA boosted proinflammatory factor mRNA expression depending on the innate signals from the cytoplasm and, if so, which one would be the critical pathway.We checked the contribution of each path by adding the individual inhibitor and the combined inhibitors (pan-inhibitors).As the results shown in Revised Fig. 3a, RPSA restored the proinflammatory cytokine gene expression when blocking one of the pathways, suggesting RPSA did depend on and make up for the cytoplasm-derived innate signal.However, we need to figure out whether and how RPSA is selective based on current results.It would be interesting to systematically figure out the relationship between RPSA and these signaling pathways, as well as the contribution of each signaling pathway in the future.

12.) Very little information is provided about an involvement of RPSA in human cells -this would enhance the translatability of the findings.
Response: To provide solid evidence of the involvement of RPSA in human infection diseases, we constructed the RPSA-deleted human A549 epithelial cells.By testing the expression of pro-inflammatory cytokine genes in both HSV-1 and IAV infection models, we demonstrated that RPSA also enhanced the pro-inflammatory and some NFkB-regulated chemokine genes expression without disturbing the Ifnb gene expression in human cells (Revised Fig. 1b-c).

This is a compelling paper with a lot of data, and potentially identifies an important, alternative host defense sensor and pathway that appears independent of traditional pattern recognition receptor pathways but nonetheless modulates their activity by
integrating regulatory information at the site of chromatin.While the potential significance of the work is high, and many of the experiments show promising and interesting results, some of the work and its interpretation are stretched thin at times leading to unsubstantiated conclusions.Suggestions to improve the manuscript: 1) One of the main issues with the manuscript is that it mixes and matches data types, experiments, and interpretation.For example, several NGS profiling experiments are performed in genome-wide fashion (e.g.ATAC-seq, H3K4me3 ChIP-seq, p65 ChIP-seq) while several others are not (ChIP q-PCR for RPSA, SMARCA5, RNAPII).

Given the claims of the authors, I would highly recommend performing ChIP-seq with
RPSA and SMARCA5 to fully understand the extent to which these proteins co-occupy chromatin and at which targets they localize to -for example, how specific are RPSA/SMARCA5 for NFkB targets?A more integrated view of how NFkB, RPSA, SMARCA5, and chromatin accessibility relate to one another would help solidify the mechanistic conclusions put forth by the authors.

Response:
We tried to perform the ChIP-seq experiment in flag-RPSA over-expressiing RAW264.7 cells infected with or without HSV-1 using the anti-DYKDDDDK antibody (Cat.No. 14793, CST) in two independent experiments.However, we found the NGS data displayed a high background in this cell line using this anti-DYKDDDDK antibody.It is barely to analyze the data in high quality; even a weak signal of differential peaks could be enriched in the NFkB and inflammation-related pathways (Supplementary Fig. 4e).We then considered the conditions in two aspects: (1) RPSA was not so "directly" binding to the packages genome.It only interacted with the histones or chromosome remodeling complex.
During the ChIP procedure, especially when enlarged the experiment system, the majority RPSA complex might break down; thus we could not get the data in high quality.(2) We also seriously considered the antibody not good enough for performing the ChIP-seq experiment in this cell line.According to the manufacturer's instructions and the reported lectures, this antibody is good enough for and mainly used in ChIP-qPCR assays.Very few works showed high specific enrichment in NGS assay with this antibody.However, this does not affect the accuracy of the main conclusions.Based on our ChIP-qPCR results in Fig. 3b, the RPSA binds to the promotors of proinflammatory cytokine genes like Il6 and Il1b.In this system, we used the promoter of the Ifnb gene as the rigorous negative control to indicate the selectivity.In consisted with our molecular biology experiments, such as the co-IP experiment shown in Fig. 4a, HSV-1 infection enhanced the interaction between the RPSA, SMARCA and P65.We also showed IRF3 was excluded from the complex, indicating the selectivity of RPSA/SMARCA5/P65 in regulation gene expression.Notably, the co-IP experiment in Fig. 5d also displayed that the loss of SMARCA 5 robustly impaired the interaction between RPSA and P65.The SMARCA5 complex was reported binding of distinct transcriptional factors and tended to locate at promotors in mouse embryonic stem cells or hematopoietic stem cells (Barisic, D. et al.  odd that DNA promoter accessibility is shown to decrease in both WT and Rpsa-iKO samples during HSV infection even though the promoters of the genes shown are dramatically up-regulated (Fig. 3d).These observations are extra confusing because it seems that the Rpsa-iKO samples show MORE DNase accessibility on pro-inflammatory promoters, where the model proposed would suggest there should be less.I would highly recommend expanding the genome-wide profiling to include ATAC-seq at both control and infected conditions to include a more comprehensive analysis to make this data more robust instead of relying on the candidate promoter DNase hypersensitivity data.

Mammalian
Response: Thanks for the concerns and helpful suggestions.To make a clear view of the results, we showed all of the ATAC-seq signals and the global cluster results in both HSV-1 infected and uninfected RAW264.7 cells in Revised Fig. 3c and d.These results showed HSV-1 infection did enhance gene transcription and loss of RPSA significantly reduced the transcriptional activity.The global clustering analysis results showed that deletion of RPSA robustly reduced a set of classical pro-inflammatory cytokine genes expression (Revised Fig. 3d), and the KEGG pathway analysis displayed the differential peaks from ATAC-seq mainly enriched in inflammatory responses, especially on NF-kappa B signaling pathway (Revised Supplementary Fig. 5b).
In Fig. 3e in the revised manuscript, we determined the chromosome openness using DNase I sensitivity assay and calculated the results with 2 -dCT method relative to those uninfected samples.Thus the decreased fold change means these chromatin regions were sensitive to DNase I digestion and had high transcription activity.Our results in Revised Fig. 3e indicated that the deletion of RPSA selectively reduced the transcriptional activity near pro-inflammatory genes promoters without affecting the Ifnb gene.Sorry for confusing the reviewer.We described the method and experimental procedures of DNase I sensitivity assay in more detail in the Revised Materials and Methods.

/ 19
3) A more detailed description of the RNA-seq data is needed.For example, Supp.Fig 3b is hard to process as it contains a mix of up-and down-regulated genes relative to each of the experimental groups mentioned.I might include a global clustering/heatmap of the experiment (like shown in Fig. 2e) that provides a birds-eye-view of the gene expression changes that occur in the experiment to provide additional context.For example, it seems clear that many proinflammatory genes have reduced induction when Rpsa is depleted, but are there also a significant number of genes that are perhaps hyper-induced too?
Response: Sorry for the description not clear.We revised the descriptions of RNA-seq data rigorously in the new version.Our previous data showed the statistics of differentially expressed gene number (Revised Supplementary Fig. 3c) and KEGG analysis (Revised Supplementary Fig. 3d) of downregulated genes, but not a mixture of up-and down-regulated genes.The global cluster analysis is shown in the Revised Supplementary Fig. 3b.Indeed, the loss of RPSA also induced a set of genes expression.As a multifaceted protein, the RPSA has functions across various biological processes, including being the receptor of laminin and a subunit of the ribosome.Thus, the loss of RPSA may aspect functions besides interrupting host sensing pathogenic nucleic acids.cGAS or RNA sensor RIG-I, the nucleic acid sensors are reported to recognize viral nucleic acids, not through binding the specific sequence.That is to say, there is nosequence specific manner of recognizing viral nucleic acids.cGAS also directly recognizes genome DNA if they meet (so there is a hypothesis of physiological separation between cGAS translocated and naked genome DNA in the nucleus).We think that RPSA, like other nucleic acid sensors, also recognizes self-nucleic acids under pathogenic conditions such as SLE.Indeed, we found that RPSA binds naked genomic DNA but not the packaged genomic DNA (Fig. 1K), confirming our hypothesis.Furthermore, RPSA showed low affinity to some types of DNA, such as G3-YSD and E.coli ssDNA, implying the structure or modifications of DNA or RNAs might also be essential for RPSA sensing the nucleic acids.In summary, it remains unanswered how these innate nucleic acid sensors are precisely prevented from recognizing self-nucleic acids in normal conditions, although several potential ways such as physical separation proposed.We discussed these issues at Page 19, lines 384-

387.
2) Based on the reviewers' response in the rebuttal to performing ChIP-seq for RPSA/SMARCA5, if the antibody/assay isn't good enough for ChIP-seq, it probably isn't very trustworthy for qPCR either.The results from the ChIP for RPSA in Fig. 3b aren't terribly convincing anyway -according to the data in the figure, RPSA is recruited to each of the promoters tested (including the Ifnb promoter), which is at odds with the description in the text.I might remove this data and back off any claims about RPSA's recruitment to the genome given this data is probably unreliable, and at best shows weak support for the proposed conclusions.I would say the lack of genome-wide , 1g, 1l; Fig. 2g, 2h; Fig. 3a; Fig. 4c; Supplementary Fig. 1c, Supplementary Fig.
to show that RPSA deletion affects cytokine expression, but not virus levels in the cells.Showing the (unchanged) levels of IFN-b in Fig. 1 already would support this notion.

(
Figure for reviewer below Fig. a).In the top1 predicted model, the RPSA inserts into the DNA groove with two critical amino acids arms (See Figure for reviewer below Fig. b), located separately with the N-or C-terminal of RPSA.It is interesting to investigate and prove the functions of these two arms in sensing nucleic acids in the future.

Figure
Figure for reviewer.Structural prediction of RPSA and HSV-60 complex.a docking scores of RPSA, cGAS and SAFA.b binding model between RPSA and HSV-60 DNA.
6 / 19expression by activating NF-κB and MAPK signaling cascades, mainly depending on TLR-TAK1 and cGAS-STING axis.Consistently, as shown in Revised Fig.3a, both pathways contribute to initiating proinflammatory gene expression.The overexpressed RPSA restored the proinflammatory cytokine gene expression when the inhibitors blocked one of the classically recognized signaling cascades, indicating
2).With respect to the text near the call-out to Fig.3c/d, the authors claim that Rpsa-iKO have less DNA accessibility near TSS regions, but Fig.3c does not seem to convincingly indicate this -at least it is not clear looking at the figure.Also, it is very 13 / 19 4)In Figs. 1 & 2 the authors make the claim that RPSA is a nuclear sensor for viral RNA/DNA.While several insightful experiments were performed (i.e.nuclear transfection of dsDNA), additional controls should also be considered.For example, is RPSA functional during the innate response to non-nuclear viruses (e.g.flaviviruses) or other pattern recognition receptor signals/NFkB signaling pathways (e.g.Tlr4 pathway activation/TNF)?These would appear to be important controls.Fig. 2g-h help address this, but additional stimuli would help make this more convincing.In addition, given the model that RPSA is needed to help increase pro-inflammatory gene expression during these responses, what is compensating its role in driving inflammatory gene expression in the case of cytoplasmic activation by e.g.polyIC and Il12b without affecting the Ifnb gene site.6) The ChIP-seq profile shown in Fig. 5d is a bit concerning -most ChIP-seq data for transcription factors, including previously published p65 ChIP-seq experiments, show maximal enrichment for the TF just upstream of the TSS in the open chromatin region of the promoter.These experiments seem to show a distribution of ChIP-seq signal just downstream of the TSS, more similar to that of a ChIP-seq experiment like H3K4me3 or H3K27ac.Response: We prudently re-checked the raw data of P65 ChIP-seq and confirmed the results in Fig. 5d are correct.As shown in the Figure for reviewer bellow, the ChIP-seq results robustly enriched NF-κB-p65 motif, as well as the well-known co-activator motifs, such as cJun and ATF3, highly suggested our data was generated from P65 ChIP-seq.Also, as compared to the results in the revised Fig. 3f, the signal of P65 ChIP-seq mainly distributed around the TSS, which is different from the distribution features of H3K4me3 ChIP-seq signals.

Figure
Figure for reviewer.Motif enrichment of P65 ChIP-seq.

Fundamentally, both cGAS
and RIG-I can recognize the naked self-genome DNA and self RNA (G.R.Pathar, et al.Structural mechanism of cGAS by the nucleosome.Nature 587, 668-672 (2020); M. Jiang, et al.Self-recognition of an inducible host lncRNA by RIG-I feedback restricts innate immune response.Cell 173, 906-919.e13(2018)).How to preferentially sense the viral nucleic acids by these innate sensors remains unclear.