Bacteriophages avoid autoimmunity from cognate immune systems as an intrinsic part of their life cycles

Dormant prophages protect lysogenic cells by expressing diverse immune systems, which must avoid targeting their cognate prophages upon activation. Here we report that multiple Staphylococcus aureus prophages encode Tha (tail-activated, HEPN (higher eukaryotes and prokaryotes nucleotide-binding) domain-containing anti-phage system), a defence system activated by structural tail proteins of incoming phages. We demonstrate the function of two Tha systems, Tha-1 and Tha-2, activated by distinct tail proteins. Interestingly, Tha systems can also block reproduction of the induced tha-positive prophages. To prevent autoimmunity after prophage induction, these systems are inhibited by the product of a small overlapping antisense gene previously believed to encode an excisionase. This genetic organization, conserved in S. aureus prophages, allows Tha systems to protect prophages and their bacterial hosts against phage predation and to be turned off during prophage induction, balancing immunity and autoimmunity. Our results show that the fine regulation of these processes is essential for the correct development of prophages’ life cycle.

b) It is not as surprising or novel as the authors communicate to the reader that there are transcripts on the opposite strand of operons, especially given the compaction of bacteriophage genomes.The notion that they have functions to the phage is also not as novel or surprising as the authors state.If they didn't, why would they be transcribed?
c) The authors do not thoroughly demonstrate experimentally (as they do in the referenced PNAS paper) that this is what they call a "noncontiguous operon".They do not concretely show biochemically that either the opposing transcript or an unconfirmed Xis gene product regulate the function of Tha-1.Though they provide several genetic experiments, by making deletions, etc. to the antisense region, they are inherently changing the sense transcript.To tease apart these functions I feel I need more biochemical evidence of what is taking place.
For one example, although an ORF is annotated (xis) the authors never provide concrete evidence that a Xis protein is translated.The authors mention in passing that they tried to express the protein and were unable to, but give no detail beyond that.Without such evidence, I am left wondering if this is just a case of anti-sense transcription and dsRNA degradation, downregulation of the cI/int promoter, some role for the Xis protein (direct protein-protein inhibition of Tha-1, stabilizing dsRNA transcripts for degradation, further down-regulating the cI/int promoter, activating degradation of Tha-1), a combination, or something else entirely.Without knowing if a protein is even produced, the model becomes unclear.Since this is a key message of the paper, I would ask for more biochemical data supporting the model, and more experiments to investigate the function of Xis.
Major comments: 1) As stated above, the authors do not convincingly demonstrate that the ORF xis is translated.First, the protein should be epitope tagged and shown to be expressed on a Western blot.Alternatively, they could fuse GFP or another reporter to xis or the N-terminus of xis.Second, coexpressing Tha-1 and ORF61 is sufficient to activate and block growth -so the authors should test whether over-expression of Xis in this context can block toxicity.Preferably this would be with xis synonymously recoded so there is minimal effect from any antisense transcripts.A control with a mutated start codon could be tested as well.Similarly, if you engineer 7206 to overexpress Pxis+xis (+/-rest of antisense transcript), does this overcome the protective effect of Tha-1 on 7206?They do replace the predicted Xis start codon with a stop codon and this strain fails to lyse.However, this alone is not thoroughly convincing.Although minor, a cleaner experiment would be to mutate the start codon, as opposed to adding a stop codon in its place.In figure 5b, ∆Pxis, xisreversed, it would be helpful to see what a start codon mutation looks like in this assay.Also, to be thorough, it would be nice to see that Tha-1 is still expressed in this context so that the result is due to expression of Xis rather than interfering with expression of Tha-1 in some way.This is a running issue throughout as the authors make many artificial genetic constructs but are unable to directly see what this is doing to expression of gene products.This leads to many assumptions being made about how the genetic constructs should probably operate but not how they do operate.
2) As the authors show, at least some of the effect of the xis transcript on Tha-1 is likely due to antisense transcripts and degradation of dsRNA by RNase III.If RNase III is not essential in Staph, the authors could test regulation of Tha-1 in an RNase III KO or knock-down.This may allow them to ascertain how much regulation of Tha-1 is through RNA degradation and how much is due to a potential Xis protein.Somewhat related, it may also be informative to knockdown the xis transcript during lytic growth and see if the 80α can still form particles/plaques.That would tease apart how much regulation of Tha-1 is due to lowering expression and would be a nice companion to the deletion mutants.
3) Though I agree that Tha-1 by all available evidence is likely a toxic RNase, it would be simple to co-express Tha-1 and ORF61 and look at what is happening to RNA on a gel.I understand that finding the specificity of the cleavage might be beyond the scope, but to see RNA being cleaved is important.4) ORF61 activates the defense system, but according to the phage panel data, pTha-1 blocks many phages.Do all of these phages have a homologous gene to ORF61?If not, please explain.5) In the phage panel data, as the authors note, Tha-1 naturally expressed within the genome offers little protection compared to the pTha-1 plasmid.As such, this is an artifact of overexpression and the broad anti-phage language should be toned down.6) Regarding the phage plaquing data -since these phages are lysogenic, is it possible that rather than providing protection the phages are integrating?Can lytic variants of these phages be tested?
7) The authors mention that this system could provide abortive infection.Typically, this is tested by infecting a strain containing the system with phages at MOIs both greater than and less than one.The experiments shown here are at MOI of 1 and 10.These MOIs should elicit the same general growth pattern if the system works by abortive infection as all bacterial cells would be infected.Please clarify.8) Perhaps this is an artifact of the methodology or an issue with the figure (Fig. 4c), but the Nanopore RNA reads during lytic growth do not appear to contain the Xis start codon (and definitely not an RBS) as shown in figure 4C.9) Typically, phage defense systems are conserved in many contexts, and those found in prophages are also present in "defense islands" to some extent.Tha-1 appears to be quite rare and only found in a few S. aureus prophages.Is it possible that this is an 80α-specific superinfection exclusion mechanism, or perhaps a mechanism that blocks spurious lytic induction of 80α?The authors do not discuss this possibility.Relatedly, I also find it curious that the system targets itself, rather than recognizing some conserved region of foreign phages.Often, prophage encoded anti-phage systems do not target themselves.
10) It is unclear how diverse the phages in the tested phage panel are.Are they related to 80α? Something to indicate how related they are would be nice, e.g. a phylogenetic tree, whole genome alignment, etc.For all the reader knows these are all very similar phages.
Minor issues: Lines 62-64: "However, no concrete examples have been described where the noncontiguous operon is advantageous for the regulation of a specific biological process."-The referenced PNAS paper shows that this organization regulates menaquinone biosynthesis.
Line 75: "labelled" should be "annotated as" Line 83: "lamboid" should be "lambdoid" Line 199: "data now shown" should be "data not shown"?Line 480: "decreased fitness cost" should be either "decreased fitness" or "increased fitness cost"  In this study, Rostøl et al. identify a new prophage-encoded anti-phage defence system that they call THA (Tail-activated, HEPN domain-containing Anti-phage system).They show that this system is encoded together with a cognate "immunity" protein that allows phages expressing it to bypass the defence and replicate normally.The immunity protein for this system was previously mistakenly annotated as an excisionase protein.The authors provide strong evidence that it is not an excisionase and does in fact act as an immunity protein that inhibits the activity of the THA system.The THA systems encoded in S. aureus phages are predicted to function as RNAses due to the presence of the HEPN domain, and the activity of the characterized system is hypothesized to induce growth arrest, preventing phage replication and protecting the bacterial population.The study of anti-phage defence mechanisms is of high interest in the microbiology community at the present time, so this work will be of quite broad interest.
Overall, this is a well-presented story; importantly the authors examined the activity of the system in the context of the prophage and have provided convincing data about how the system functions with respect to the inhibition provided by THA and the reversal of this inhibition by the Xis protein.However, there is a lack of insight into how this system is actually functioning to block phage replication.At which stage of the phage life cycle is the inhibition occurring?Does phage genome replication occur normally?Are phage virion intermediates produced?It appears that the cells are not lysed, so is there is some block to phage late gene expression (maybe only lysins)?It appears that a trigger for activity is located in the phage minor tail proteins, but it is not clear if this is a protein trigger, or an RNA sequence trigger.Do the authors have evidence that there is a cleavage site in this region of the phage genome (i.e.do the mutations provide evasion of the system at the protein or RNA level)?In addition, it's not clear if the growth arrest noted is reversible or irreversible, as might be expected for an Abi system.I would expect that more detailed insight into the specific mechanism of anti-phage activity be provided for publication in Nature Micro.

Major concerns:
I found the term "regulator" used in reference to the activity of the Xis protein (e.g., lines 29, 74-76) confusing as this brings to mind transcription factor type regulatory proteins.At the protein interaction level these types of biological interactors are usually referred to as inhibitors.The authors should consider changing this terminology.
The MOI of 10 used in Fig. 2b did not result in complete cell lysis or premature cell lysis.Did the authors determine if the CFUs/ml are changing throughout infection; are the cells lysing or is there just bacteriostatic activity that is relieved at some point?Of the cells that regrow, what are they?Resistant mutants, lysogens?Quantification of changes in mRNA during phage infection of 7206 in a wild-type vs.Tha-1 prophage mutant should be done to determine if there is an arrest in all cellular activities, or if there is mainly a loss of transcription of phage products.
The authors must show that the H270 mutant is expressed and soluble to the same levels as wild-type THA.While the His residue may be involved in nucleic acid cleavage, mutation of this histidine residue may also lead to decreased protein expression/stability or solubility.A Western blot would provide evidence that it is the specific loss of enzymatic activity rather than instability/insolubility that causes the loss of activity.Do phages that bypass THA defence encode a similar Xis, or is it strictly differences in the minor tail protein target sequence?Some additional alignments would help paint a clearer picture of the relationship between the Xis and Tha proteins.The escape mutants were done in a plasmidbased system and don't provide biological context if Xis is able to allow phage infection in the presence of Tha from a prophage.Can the repressor of another phage be exchanged with 80a and see if 80a(new repressor) can plate on a wild type 80a lysogen?Or can the Xis from 80a be put into 7206?Are the Xis proteins broadly active or are they Tha sequence specific?Co-expressing the 80a and NM2 Xis/Tha-1/2 in the presence of bypassing phage would add knowledge for Tha functionality.Alternatively, can a chimera or truncated form of Tha be generated to determine if the Nterminal domain is essential for recognition of the phage tail proteins?
Minor concerns: Figure 2a -The blank spaces make this heatmap difficult to read.Do the clear/white boxes represent no inhibition at all and the clear boxes with * mean infection to wild-type levels but the plaques were smaller?It should be clearly stated if these data are PFU/mL or fold-inhibition.Figure 3 -It would be easier for the reader if this figure was inverted; as presented these look like plaques, not colonies.Figure S1b -There are two orf3's noted.Line 32 -Noncontiguous is spelled incorrectly.Line 155-162.This could be reworded to improve clarity.The movement between prophages and plasmid-expressed protein (I assume that is what Tha-1 alone means) is confusing.It would be useful to state how many phages were inhibited only by Tha-1 from the prophage.Line 390 -There is an empty pair of brackets.Supplemental Figure 1g -The labels say Φ11 but the figure legend says 80a.
Reviewer #3 (Remarks to the Author): I have reviewed the manuscript.I find the science presented in the paper fascinating.Although the authors emphasized that their findings represent a new mechanism of phage exclusion, e.g.Crispr-CAS, I believe the data reveals a fascinating mechanism of temperate phage biology.The experiments are performed logically with clear results.I recommend publication of the manuscript in Nature.But in my opinion the manuscript as presented is too long and wordy.I suggest it should be thoroughly edited for brevity and crispy reading.I also have some comments that the authors may respond to before the manuscript is accepted.My comments are not in order of chronology or in terms of importance although some are more important than others: 1.Of course, as I mentioned, it should be thoroughly edited.2. I don't know why the gene referred to as xis was called so -may be because of homology or gene location -but since the authors have shown that it does not function as the 'classical' excisionase, I suggest to change it to one that more reflects its demonstrated function.3. The triangular symbols at the top of, for example, figure 3 is misleading the way they are presented.If they reflect Dilution as noted, then it should be horizontally inverted reflecting more and more dilutions; as presented they reflect phage titers, not dilutions.4. The manuscript misleadingly and frequently uses the word Tha1 when they really mean Tha+ wild type situations.Please carefully edit the entire manuscript about this point.Was the mutant Tha1 gene cloned in the plasmid?In line 166, the authors certainly mean, "….In the absence of WT Tha gene".Also, in line 143 , same problem.5. What kind of mutant is Tha1?Base-substitution?Deletion?6.I strongly disagree with the use of the word non-contiguous operon.It is very misleading word to describe the case it refers to.I Insist that the authors use a different word to describe overlapping antisense operons.7. Line 121."….We sequenced five 80a mutant phage" 8. Line 422.Shouldn't the statement "Tha targets tail proteins" be " tail protein targets Tha? 9. Are the authors using the word "moron" in the sense that Roger Hendrix defined it?Whatever the authors are talking about, please clarify and give a reference.10.The Authors may point out that ORF61 has a dual role: tail protein as well as an activator of Tha gene.11.Lines 317-322.These lines really belong to Experimental Procedures, not in the results section.This is one example.There are other situations in the manuscript where some of the statements belong to Experimental Procedures.12. Line 417.The paragraph doesn't add much; may be deleted.13.I also believe some of the discussions talk about evolutionary aspects of the current results.I feel that those are not quite important points in the context of the current findings; may be left out.Sankr Adhya ** Although we cannot publish your paper, it may be appropriate for another journal in the Nature Portfolio.If you wish to explore the journals and transfer your manuscript please use our manuscript transfer portal.You will not have to re-supply manuscript metadata and files, but please note that this link can only be used once and remains active until used.For more information, please see our manuscript transfer FAQ page.
Note that any decision to opt in to In Review at the original journal is not sent to the receiving journal on transfer.You can opt in to In Review at receiving journals that support this service by choosing to modify your manuscript on transfer.In Review is available for primary research manuscript types only.
We would like to express our sincere appreciation to the reviewers for their valuable comments and suggestions.We believe that the new experiments conducted in response to their questions will provide fresh insights into the functioning of the system.We would also like to extend our gratitude to the reviewers, with special recognition for Reviewer 3, as their comments have highlighted a potential limitation in our previous presentation of the relevance of our findings.In Staphylococcus aureus, it is noteworthy that the majority of strains harbor at least one prophage, with an average of four prophages per cell.In this context, where a significant proportion of the prophages encode the immune systems characterised in our study and their corresponding anti-autoimmune mechanisms, it becomes evident that these immunity mechanisms (which protect cells against infection by other phages) and measures to prevent autoimmunity (when the prophages become activated) should be integral components in describing the various stages of the prophage life cycle.Now we have evidence that this is not just happening in S. aureus but also in other species, such E. coli and the prototypical phage 80 (paper in revision at Nature; reference 2023-08-13773).
Traditionally, life cycles have been analysed based on the impact of different processes (such as excision, replication, or packaging) on a phage's ability to reproduce.As demonstrated in our research, alterations in either the immune systems we have characterised or their corresponding mechanisms that regulate their function can profoundly affect the biology of prophages encoding these systems.Therefore, we propose that immunity and counter-immunity should be incorporated into the definition of prophage life cycles.We are grateful to the reviewers for guiding us toward this important perspective and we have now incorporated this idea into the manuscript.

Reviewer #1
Rostøl et al. examine two uncharacterized genes within the cI/int operon of a lysogenic lambdoid phage in S. aureus, 80α.Tha-1 is a HEPN-containing protein that the authors claim is a generalized phage-defense system.The other gene, transcribed in the other direction within this region, does not encode an excisionase as annotated, but instead the region is involved in regulating Tha-1, possibly through an open-reading frame, xis.Growth arrest via Tha-1 can be activated by expression of ORF61, suggesting this is what is detected to activate the defense system in response to foreign phage infection.Surprisingly, this putative defense system targets itself during lytic growth and the overlapping antisense transcript allows the phage to avoid auto-immunity during its own lytic cycle.
Overall, the authors use many clever genetic experiments to test hypotheses, but there is a significant lack of biochemistry to support their model.Additionally, though the paper describes an interesting potential defense system/prophage exclusion mechanism, I am unclear how this paper has a broad impact.
We thank the reviewer for constructively assessing our work.We have now performed additional experiments to confirm the proposed model.
In accordance with the reviewers' feedback, we have also made adjustments to highlight the importance of our work.As suggested by this reviewer, it may indeed seem surprising that the phage encodes a defense system that targets itself.However, our research has unveiled that this is not an isolated occurrence; rather, it is a prevalent characteristic observed in many Staphylococcal phages, as demonstrated here.Furthermore, we have obtained preliminary evidence of similar phenomena in other species, such as E. coli.Consequently, in alignment with reviewer 3's suggestion, we concur with the idea that the phage's immunity mechanisms, along with their mechanisms to counteract this immunity, should be integral components when defining the phage life cycle.We believe that this perspective is both unique and of significant importance, with the potential for broad-reaching impact.
The authors advance the concept of "noncontiguous operon" as the major impact of the paper.Regarding this: a) I am wary of how widely accepted the term "non-contiguous operon" is in the scientific community.For that reason, I personally don't want to endorse this terminology because it is antithetical to the definition of an operon as a single transcript.
The term "noncontiguous operon" was initially employed in the original paper to describe a distinctive genetic organisation wherein within a classical operon, there exists a gene transcribed in the opposite direction.The authors asserted in that publication that this genetic arrangement would somehow enable the unique control of these genes.In our manuscript, we have opted to utilise the established terminology for this genetic configuration and, for the first time, have substantiated the claim regarding the regulatory advantages inherent to this genetic organisation.We appreciate, however, the concern the reviewer has with the heavy emphasis on the term noncontiguous operon in our initial submission.We understand that it is not yet a commonly used term in the scientific literature, and that it might distract from the underlying regulatory concepts we tried to convey with the use of the phrase.As a result, we have removed it from the title, and have used it less throughout the text and figures.Where it is used, it is often used together with "complex transcripts", the term the eukaryotic virus field uses, or defined as overlapping, antisense transcripts.Yet, we still feel like the noncontiguous operon term has value in the context of our study, so we decided to leave it into the extent just described.
b) It is not as surprising or novel as the authors communicate to the reader that there are transcripts on the opposite strand of operons, especially given the compaction of bacteriophage genomes.The notion that they have functions to the phage is also not as novel or surprising as the authors state.If they didn't, why would they be transcribed?
We concur with the reviewer that the presence of overlapping/antisense genes being transcribed is not unexpected; indeed, why wouldn't they exist?Historically, when genomes were annotated and operons identified in silico, it was assumed that when a gene was oriented in the opposite direction to others, the flanking genes were associated with different operons.However, the original paper demonstrated that this assumption does not always hold true.In certain instances, the genes neighboring the gene in the opposite direction were transcribed from the same promoter, resembling the structure of a classical operon.This operon-like structure was disrupted by the presence of the antisense gene, thus aligning with the authors' definition of a "noncontiguous operon."This definition was also predicated on the notion, as mentioned earlier, that this genetic arrangement offered additional regulatory possibilities.
While we acknowledge that antisense transcripts are common in compact viral/phage genomes, to the best of our knowledge, there has not been a clearly defined case where a noncontiguous operon has been shown to regulate a significant in vivo phenotype.It could be argued that antisense genes often serve primarily to compact the genome and decouple the simultaneous expression of adjacent genes, rather than regulating a shared function.We have now provided a very clear and relevant example of their important regulatory function.We have made revisions in the text to elucidate this point.
c) The authors do not thoroughly demonstrate experimentally (as they do in the referenced PNAS paper) that this is what they call a "noncontiguous operon".They do not concretely show biochemically that either the opposing transcript or an unconfirmed Xis gene product regulate the function of Tha-1.Though they provide several genetic experiments, by making deletions, etc. to the antisense region, they are inherently changing the sense transcript.To tease apart these functions I feel I need more biochemical evidence of what is taking place.
Regarding the opposing transcripts, they clearly exist based on the transcriptomics (Fig. 5b-c), and nuance is added with our characterisation of the ith-1 (xis renamed) promoter with a reporter assay (Fig. 5a) combined with previous knowledge of the cI promoter.Moreover, we now have strong evidence for the translation of the ith-1 (xis) gene.As for also changing the sense transcript when modulating the antisense transcript, this is the only way to study the antisense operonic structure and would also need to be done if characterising the region with biochemistry.Through our various genetic prophage mutants, we dissect the structure by removing one possible contributing factor at the time, assessing the contribution of the Ith-1 (Xis) protein, the ith-1 (xis) promoter, and the localisation of these elements in an antisense orientation to the cI (and Tha-1) operon.We have performed additional experiments, detailed below, in an RNase III knockout, and using Tha-1 and Ith-1 (Xis) expressed from different plasmids, to bolster our previous results.
For one example, although an ORF is annotated (xis) the authors never provide concrete evidence that a Xis protein is translated.The authors mention in passing that they tried to express the protein and were unable to, but give no detail beyond that.Without such evidence, I am left wondering if this is just a case of anti-sense transcription and dsRNA degradation, downregulation of the cI/int promoter, some role for the Xis protein (direct protein-protein inhibition of Tha-1, stabilizing dsRNA transcripts for degradation, further down-regulating the cI/int promoter, activating degradation of Tha-1), a combination, or something else entirely.Without knowing if a protein is even produced, the model becomes unclear.Since this is a key message of the paper, I would ask for more biochemical data supporting the model, and more experiments to investigate the function of Xis.
As mentioned, we were unable to express and purify the proteins under study.To address the reviewer's comments, we have established a system where the anti-phage protein (referred to as Tha-1) is expressed from a plasmid.Cells carrying this plasmid were subjected to a challenge with a phage that is sensitive to Tha-1.As expected, the expression of Tha-1 effectively blocked phage infection.We conducted a subsequent experiment by introducing an additional plasmid that expressed either the wild-type xis gene (now denoted as ith-1, signifying the inhibitor of Tha-1), the ith-1 gene with a stop codon in the first position, or a recoded version of the Ith-1 protein (with an identical sequence but different codons).When we exposed the various strains to the phage, we observed that protection against phage infection was abolished in the presence of the WT Ith-1 protein, as well as the recoded Ith-1.However, Ith-1 M*1 did not inhibit Tha-1, resulting in the phage being unable to infect the cells.This confirms that Ith-1 (formerly Xis) mainly functions as a protein and effectively inhibits its counterpart Tha-1 protein (see new Figure 4 for details).
Major comments: 1) As stated above, the authors do not convincingly demonstrate that the ORF xis is translated.First, the protein should be epitope tagged and shown to be expressed on a Western blot.Alternatively, they could fuse GFP or another reporter to xis or the N-terminus of xis.
We strongly agree that providing evidence for Ith-1 (Xis) being translated is essential for our work.Our previous results, summarised in the new Fig. 4, demonstrate that Ith (Xis) perform its role as a protein.Importantly, the manuscript also contains other data supporting ith-1 being translated, including the ith-1 M1* mutation in Fig. 3e, and the ith-1 M1* mutation in Fig. 6a-c.Also, during our preliminary experiments (not included in the manuscript), we used an ith-1 N6* stop codon mutation, which behaved similarly to ith-1 M1* .In sum, we believe the in vivo and genetic data in our revised manuscript confirms that ith is translated into Ith protein, and that the protein is the main regulatory of Tha activity.The balance of Ith expression during lysogeny and phage induction is what the antisense operonic structure contributes to, and why we have emphasised the antisense transcripts (noncontiguous operon) in our work.
Second, co-expressing Tha-1 and ORF61 is sufficient to activate and block growthso the authors should test whether over-expression of Xis in this context can block toxicity.Preferably this would be with xis synonymously recoded so there is minimal effect from any antisense transcripts.A control with a mutated start codon could be tested as well.
We thank the reviewer for this suggestion, especially for the recoded ith-1 (xis), which we think has been of great help proving this gene is translated without dsRNA interference.In Fig. 4a described above, we performed a very similar experiment, but instead of using the ORF61 overexpression plasmid, we used phage infection to trigger Tha-1 activity.We believe this is better and more natural since ORF61 is highly expressed with the plasmid system, and might be able to outcompete Ith-1 inhibition (even if this does not happen at physiological levels).This approach was also more convenient experimentally, since the Tha-1/ORF61 co-expression already uses two plasmids, and we do not have a reliable third plasmid in S. aureus compatible with these two.
Similarly, if you engineer 7206 to overexpress Pxis+xis (+/-rest of antisense transcript), does this overcome the protective effect of Tha-1 on 7206?Again, a great suggestion.To avoid finicky phage engineering while still answering the reviewer's suggestion, we took advantage of a pre-existing 80α mutant that lacks ith-1 (and a non-functional version of tha-1 to allow the phage to obtain sufficient titres) (80α Δith-1, tha-1 F32L ).We know that 80α is largely insensitive to plasmid-expressed Tha-1, producing similar numbers of smaller plaques (Fig. 2a).However, with the mutant 80α, infection is blocked by Tha-1.Therefore, ith-1 of the wild-type 80α can normally block the Tha-1 of the host.This data is included as Fig. 4b.

Fig 4b
In addition, we hypothesised that the ith of an infecting phage can circumvent Tha immunity.From previous transcriptomics data, we know that the ith-1 of 80α is expressed early upon infection (explaining Fig. 4b, above).We think immune evasion with Ith proteins is quite common for our phages, and helps explain the infection patterns observed in Fig. 2a.An analysis of the ith genes of the phages in the panel (Fig. S6d) shows that many phages have ith-2 genes, and all these phages are largely insensitive to Tha-2 inhibition (Fig. 2a).Only 80α and Φ11 have ith-1 genes, and these phages are not very sensitive to pTha-1.
To conclude, the ith gene of an incoming phage can help circumvent Tha immunity, and this has added an extra layer to our work.
They do replace the predicted Xis start codon with a stop codon and this strain fails to lyse.However, this alone is not thoroughly convincing.Although minor, a cleaner experiment would be to mutate the start codon, as opposed to adding a stop codon in its place.
We are a bit confused by this comment.We do not add an additional codon, we change the ATG start codon to a TAA stop codon.Changing the first codon to a stop codon is the same as mutating the start codon, though it is slightly less likely to allow non-standard start codons upstream to be used and produce a longer, though functioning, Ith protein.In preliminary experiments, we also used an N6* stop codon in ith-1, with the same results as the M1* version (not in the manuscript).We also observe similar, incomplete lysis when doing inductions for Fig. 6b, where the ith stop codon and ith deletion behave similarly.We have therefore left the figure as it is (old Fig. 3d, new Fig.3e) In figure 5b, ∆Pxis, xisreversed, it would be helpful to see what a start codon mutation looks like in this assay.Also, to be thorough, it would be nice to see that Tha-1 is still expressed in this context so that the result is due to expression of Xis rather than interfering with expression of Tha-1 in some way.This is a running issue throughout as the authors make many artificial genetic constructs but are unable to directly see what this is doing to expression of gene products.This leads to many assumptions being made about how the genetic constructs should probably operate but not how they do operate.
With the aforementioned results confirming the expression of Ith-1 and its ability to counteract Tha-1 activity, all the experiments now possess a clear rationale, and their results are easily explainable.We have used multiple approaches to validate the proposed model, including the expression of various proteins, either individually or in conjunction with their respective partners, from plasmids, as well as modifications to the phage's genetic structure.These experiments have effectively dissected the roles of the different proteins involved in the process.We believe that there is no alternative hypothesis that could account for the results we have obtained.Therefore, introducing more mutants or constructs would only complicate the narrative of this study.For instance, the reviewer suggested that Tha-1 expression might be negatively controlled by Ith-1.However, in the experiments involving plasmids expressing Tha-1 and Ith-1, the defense system Tha-1 is expressed from a distinct promoter, clearly independent of any control by Ith-1.In these experiments, Ith-1 alleviates the function of Tha-1.
Another crucial point to consider is that tha-1 is part of an operon that also includes the master repressor (cI) and the phage integrase (int).If Ith-1 controlled the expression of all these genes, one would expect that the evolved ith-1 mutant phages, whether carrying mutations in tha-1 or in the tail gene, would have impacted their functionality.However, when tha-1 is removed (in the ith-1 mutant), this mutant works like wild-type, with these mutations having no impact on the biology of the phage, suggesting that Ith-1 does not regulate other phage functions aside from its effect on controlling Tha-1 activity.
2) As the authors show, at least some of the effect of the xis transcript on Tha-1 is likely due to antisense transcripts and degradation of dsRNA by RNase III.If RNase III is not essential in Staph, the authors could test regulation of Tha-1 in an RNase III KO or knock-down.This may allow them to ascertain how much regulation of Tha-1 is through RNA degradation and how much is due to a potential Xis protein.
Following the reviewer's comment, we have conducted the suggested experiment.Our hypothesis was that genes transcribed in opposite directions might be reciprocally regulated, either through RNA polymerase collisions or double-stranded RNA (dsRNA) degradation by RNase III.Upon comparing the phenotypes of the various phage mutants, we observed that they were similar in both the wildtype and RNase III mutant strains (refer to Fig. S10a).Therefore, we can conclude that the observed regulatory effect most likely occurs through head-on RNA polymerase collisions.
Somewhat related, it may also be informative to knock-down the xis transcript during lytic growth and see if the 80α can still form particles/plaques.That would tease apart how much regulation of Tha-1 is due to lowering expression and would be a nice companion to the deletion mutants.
If straight-forward, this would be a nice experiment to do.We feel that it is unfeasible, however.Assuming the reviewer means knockdown of RNA transcripts, we are not aware of efficient RNA knockdown platforms in S. aureus, and we feel like developing this for the current study would be excessive.If the reviewer refers to transcriptional knockdown by e.g. a dCas9 binding at the beginning of the ith-1 (xis) gene, we think this would affect transcription of the downstream integrase gene (from the cI promoter), and possibly also of the tha-1 transcript due to RNA polymerase pileup at the DNA-bound dCas9, preventing full expression of the tha-1 gene to be translated.Also, knockdown technologies are often somewhat inefficient, and it would be hard to ensure sufficient ith-1 transcript knockdown to prevent all translation of Ith-1 proteins.
Instead, we think the prophage induction of the 80α variants in Fig. 6b addresses the concern raised.For ith-1 M1* and Δith-1, transcription from the ith-1 promoter still occurs despite of the Ith-1 protein not being made.To address the antisense transcript possibly leading to dsRNA degradation of the tha-1 transcript, we performed the induction experiment in a Δrnc background described above, ruling out that free ith-1 transcripts significantly reduce tha-1 transcripts through dsRNA degradation by RNase III.
3) Though I agree that Tha-1 by all available evidence is likely a toxic RNase, it would be simple to coexpress Tha-1 and ORF61 and look at what is happening to RNA on a gel.I understand that finding the specificity of the cleavage might be beyond the scope, but to see RNA being cleaved is important.
Thanks to the reviewer's comment, we ran total RNA on an Agilent Bioanalyzer, visualising rRNA.This shows some RNA degradation of rRNAs upon Tha-1 activation by ORF61, and this data is now in Fig. 3b Fig 3b HEPN domains typically target ssRNA, and we suspect structural rRNAs are difficult substrates, most likely being cleaved at protruding rRNA loops or incompletely folded rRNA during ribosome biosynthesis.We think the main Tha-1 substrate is mRNA and possibly tRNA loops, leading to transcriptional shut-off.4) ORF61 activates the defense system, but according to the phage panel data, pTha-1 blocks many phages.Do all of these phages have a homologous gene to ORF61?If not, please explain.
However, we believe the presence of the immune evasion proteins ith-1/2 are more informative of how sensitive a phage is to Tha-1/2 inhibition.Many tested phages encode ith-2, explaining why they are resistant to Tha-2, and more phages are targeted by Tha-1.This is discussed in lines 350-353: "Of the phages in the panel, only two phages encode tha-1/ith-1, while six phages contain tha-2/ith-2 (Fig. S7c-d).Interestingly, none of the ith-1 phages are strongly targeted by Tha-1, and none of the ith-2 phages are strongly targeted by Tha-2 (Fig. 2a)." We have made a new supplemental figure (Fig. S7) showing the phylogeny of ORF61, ORF54, Tha-1/2, and Ith-1/2.This in large part helps explain the sensitivity to a given phage to Tha inhibition.In addition, the presence of Ith proteins in phages adds a new layer to the ith/tha operon, where during lysogeny, it acts an immune system, but during phage infection, it can help an incoming phage evade Tha immunity.

5)
In the phage panel data, as the authors note, Tha-1 naturally expressed within the genome offers little protection compared to the pTha-1 plasmid.As such, this is an artifact of overexpression and the broad anti-phage language should be toned down.
We notice two phages more sensitive to pTha-1 than prophage-encoded Tha-1, and one for pTha-2.In response, we have changed the wording in this part of the text removing the word broad, and broad no longer refers to Tha immunity in this context.We suspect overexpression of Tha from a plasmid can either allow recognition of a poorly recognised minor tail protein, or overcome ith inhibition by the incoming phage.

6)
Regarding the phage plaquing data -since these phages are lysogenic, is it possible that rather than providing protection the phages are integrating?Can lytic variants of these phages be tested?
We do not think that Tha immunity promotes phage integration of the following reasons: i) In Fig. S5 (old Fig. S4), we show(ed) that the Tha-2 of ΦNM2 does not impact lysogenisation of an incoming 80α phage. ii) The trigger of Tha, one of two minor tail proteins (ORF61/ORF54 homologues) are only expressed late in the lytic cycle, and should not be expressed when establishing lysogeny.
iii) Fig. 2a shows that many phages are blocked by the prophage Tha-1/2 of 80α/ΦNM2, and many of the incoming phages share integration sites with one of these two phages, making it difficult for them to integrate.
Yet, to be sure, we performed an infection of wild-type 80α or 80α-vir, a lytic version, into cells expressing Tha-2 ΦNM2 , and immunity was observed in both cases.This is now Fig. S2d: Fig S2d 7) The authors mention that this system could provide abortive infection.Typically, this is tested by infecting a strain containing the system with phages at MOIs both greater than and less than one.The experiments shown here are at MOI of 1 and 10.These MOIs should elicit the same general growth pattern if the system works by abortive infection as all bacterial cells would be infected.Please clarify.
Following the reviewer's comment, we have repeated the experiment at MOIs 0.1 and 5, fulfilling the criteria of higher and lower MOIs than 1.The results remain very similar, and have now replaced the old data in Fig. 2b-c.We believe the previous results at MOI of 1 still reflect most cells not being infected, due to i) the poor adsorption of our S. aureus phages (many phage particles do not infect, or take a long time to do so), and ii) general random error with measuring phage titres and CFUs needed to calculate MOIs.8) Perhaps this is an artifact of the methodology or an issue with the figure (Fig. 4c), but the Nanopore RNA reads during lytic growth do not appear to contain the Xis start codon (and definitely not an RBS) as shown in figure 4C.
The reviewer is correct.Fig. 5c (previously 4c) has been adjusted to accurately reflect where the transcripts begin.Below is a zoomed-in picture (not part of the manuscript) of the reads in the relevant region, showing that the RBS and full ith-1 gene are part of the transcripts (transcription starting 21 nucleotides upstream of the ATG).9) Typically, phage defense systems are conserved in many contexts, and those found in prophages are also present in "defense islands" to some extent.Tha-1 appears to be quite rare and only found in a few S. aureus prophages.Is it possible that this is an 80α-specific super-infection exclusion mechanism, or perhaps a mechanism that blocks spurious lytic induction of 80α?The authors do not discuss this possibility.
Relatedly, I also find it curious that the system targets itself, rather than recognizing some conserved region of foreign phages.Often, prophage encoded anti-phage systems do not target themselves.
Firstly, Tha-1/2 seem common in S. aureus.A pBLAST search of Tha-1 80α returns about 90 strong hits, almost all in S. aureus, while Tha-2 ΦNM2 returns more than 100 strong hits.This is reflected by the phages we used in the phage panel in Fig. 2a, where out of 17 temperate phages, two have Tha-1 and six have Tha-2 (Fig. S6).In addition, two phages (ROSA and Φ7206) encode ith-2, presumably to avoid immunity by Tha-2, without themselves encoding tha-2 genes.This means that they have evolved to circumvent Tha immunity of lysogens they regularly infect.This is referred to in lines 354-356: "Interestingly, two phages (ROSA and Φ53) only encode ith-2, and no tha-2, making ith-2 a strict immune evasion factor with no role in regulating a cognate tha-2 gene for these phages, used only to infect Tha-2-containing lysogens." We therefore anticipate that the Tha system is common in S. aureus phages, with possibly similar, more distantly related genes in other organisms.Therefore, we do state that the system seems quite common, e.g. in lines 505-507: "It is commonly encoded by S. aureus phages under the lysogenic cI repressor promoter, and exemplified by Tha-1 (found in e.g.80α) and Tha-2 (found in e.g.ΦNM2)." The idea that the antisense transcription can prevent spurious prophage induction or change lysogenisation dynamics upon phage infection is a very intriguing one.We performed several experiments to investigate this before our initial submission.It would make sense that the ith-1 promoter might prevent transcription of the integrase after induction, helping to avoid reintegration (thus affecting induction rates).However, we found no evidence for this.For example, in the experiment below (not in the manuscript), we looked at spontaneous induction rates in wild-type 80α and in an 80α mutant lacking the ith-1 promoter, the ith-1 gene, and the tha-1 gene ("80a delta ORF2-3").There was no significant change in spontaneous induction, suggesting the antisense transcription does not strongly regulate cI or int expression.

Not in the manuscript
Due to this negative data, and not wishing to complicate the paper further, we do not mention this possibility.
Thirdly, we do not find it that strange that the system would recognise a conserved protein shared by the cognate phage.As a lysogen, the prophage will be infected by both closely related and distantly related phages.We suspect that it would be hard to impossible to find a highly conserved gene that is shared between most closely related phages and not by the cognate phage (e.g.80α).This does not mean that the prophage should not evolve to also protect against closely related phages.Indeed, this is one of the reasons why such a diversity of anti-phage systems exists, and why there is often redundancy (e.g.80α also has an anti-lytic phage system (PMID: 36779718) and another system found by us).
Finally, we would like to add that the reviewer finding it curious to target a conserved gene shared with the cognate prophage adds to the novelty of our research, and shows it would be of interest to the scientific community.10) It is unclear how diverse the phages in the tested phage panel are.Are they related to 80α? Something to indicate how related they are would be nice, e.g. a phylogenetic tree, whole genome alignment, etc.For all the reader knows these are all very similar phages.This is a great point, and convinced us to perform phylogenetic analysis on the phages.However, the resulting figure below (not in the manuscript) does not tell the reader much about why phages are targeted.This is because the presence of ith genes in the phage, and small variations in minor tail proteins, are likely the only factors determining sensitivity to Tha.Also, in general, due to S. aureus temperate phages being quite closely related and being mosaic, phylogenetic comparisons often do not reflect the function of the phages (very similar phages can have different repressors or integrases, while distantly related phages can have many similar genes).
While we are happy to include a refined version of the above figure in the manuscript if requested, we think the phylogenetic analysis of ORF61/54 homologues, and of Ith and Tha, are much more informative (new Fig. S7).From this, we can in large part explain why some phages are targeted by Tha-1/2 and not others (Fig. 2a).

Minor issues:
Lines 62-64: "However, no concrete examples have been described where the noncontiguous operon is advantageous for the regulation of a specific biological process."-The referenced PNAS paper shows that this organization regulates menaquinone biosynthesis.
The referenced paper, by one of our co-authors, does describe the noncontiguous operon structure of the menaquinone biosynthesis operon.However, it does not provide a physiological explanation of why a noncontiguous operon is beneficial for the regulatory process, or whether a different configuration would also allow the regulation of this process.This is why we state that the ith/tha noncontiguous operon is the first known case of where the genetic organisation is essential to the biological process in question.Line 75: "labelled" should be "annotated as" Changed Line 83: "lamboid" should be "lambdoid" Fixed Line 199: "data now shown" should be "data not shown"?Changed to "data not shown" Line 480: "decreased fitness cost" should be either "decreased fitness" or "increased fitness cost" Changed to "decreased fitness" Horizontal (y axis) grid lines added to Fig. 5a (old 4a) Fig. 4b and c: It would be nice to see total coverage in both directions rather than aligned reads (are these all the reads that mapped?).
For Fig. 5b-c (old Fig. 4b-c, read depth has been added on top of each panel: We agree.This section contains a lot of information, and we have added a schematic (Fig. 6a) to help the reader visualise the different constructs.
Perhaps the authors should rename xis since they show it is likely not an excisionase.
We have renamed Xis Ith (Inhibitor of Tha), with the 80α version being Ith-1, and the ΦNM2 variant being Ith-2.

Reviewer #2
In this study, Rostøl et al. identify a new prophage-encoded anti-phage defence system that they call THA (Tail-activated, HEPN domain-containing Anti-phage system).They show that this system is encoded together with a cognate "immunity" protein that allows phages expressing it to bypass the defence and replicate normally.The immunity protein for this system was previously mistakenly annotated as an excisionase protein.The authors provide strong evidence that it is not an excisionase and does in fact act as an immunity protein that inhibits the activity of the THA system.The THA systems encoded in S. aureus phages are predicted to function as RNAses due to the presence of the HEPN domain, and the activity of the characterized system is hypothesized to induce growth arrest, preventing phage replication and protecting the bacterial population.The study of antiphage defence mechanisms is of high interest in the microbiology community at the present time, so this work will be of quite broad interest.
We thank the reviewer for their appreciation of our work and for their kind comments.
Overall, this is a well-presented story; importantly the authors examined the activity of the system in the context of the prophage and have provided convincing data about how the system functions with respect to the inhibition provided by THA and the reversal of this inhibition by the Xis protein.However, there is a lack of insight into how this system is actually functioning to block phage replication.At which stage of the phage life cycle is the inhibition occurring?Does phage genome replication occur normally?Are phage virion intermediates produced?It appears that the cells are not lysed, so is there is some block to phage late gene expression (maybe only lysins)?
We acknowledge the need for additional mechanistic insights into the Tha mechanism, and as a result, we have incorporated supplementary data to offer a more comprehensive understanding of the system.In particular, our new Southern blot analysis of samples collected after prophage induction reveals that the ith-1 (xis) mutant prophage is induced similarly to the wild-type and undergoes normal replication for the initial 60 minutes post-induction, after which replication is halted (see Fig. 3f).This timing corresponds to the expression of ORF61 (as indicated by our transcriptomic analyses, Fig. S6), which activates Tha-1.Furthermore, our new experiments demonstrate that when Tha-1 is active, it exhibits RNase activity (Fig. 3b), which is largely irreversible, as demonstrated in Fig. 3c.This explains why the prophage ceases replication.In sum, we believe that these recent findings, when combined with our earlier data, provide a detailed depiction of the Tha immunity mechanism.
It appears that a trigger for activity is located in the phage minor tail proteins, but it is not clear if this is a protein trigger, or an RNA sequence trigger.
We thank the reviewer for pointing this out.While we believed the Tha trigger was a minor tail protein, and not the RNA, we did not provide conclusive evidence of this, and the ORF61 F13C escaper mutation could also be an RNA escaper.Indeed, a recent report showed that the trigger for CBASS immunity in staphylococci is phage RNA (https://www.biorxiv.org/content/10.1101/2023.03.07.531596v1).While we have been unable to purify Tha-1 and ORF61 proteins to perform clean in vitro experiments, we performed an in vivo twoplasmid toxicity experiment with two stop codons in ORF61 (D2* and H50*), now in Fig. S3d: With these two mutations being in distant locations, we think it is highly unlikely that they could both fully disrupt the RNA from being recognised if RNA was the trigger.We also do not detect any predicted domains in the N-terminal of Tha, RNA-binding or otherwise.Instead, the ORF61 protein is the likely trigger.Do the authors have evidence that there is a cleavage site in this region of the phage genome (i.e.do the mutations provide evasion of the system at the protein or RNA level)?
We do not think that Tha immunity is effected through specific cleavage of the orf61 RNA, which is what a specific cleavage site orf61 RNA would imply.The RNA could hypothetically be the Tha activator (discussed above), but not the Tha substrate.Firstly, we observe growth arrest in the Tha-1/ORF61 expression system, in the absence of any phage, suggesting host RNAs being targeted (Fig. 3a).In fact, using this system, phages that are insensitive to the system can be targeted if we activate Tha-1 with ORF61 (Fig. 3d).We also see non-specific cleavage of rRNA (Fig. 3b).Thirdly, we see a lack of lysis in the prophage induction where Ith is not translated in Fig. 3e (80α ith-1 M1* ).If the specific orf61 (polycistronic) transcript was specifically cleaved, other late genes would be expressed, including the holin and lysin, which would result in normal lysis timing.
In addition, it's not clear if the growth arrest noted is reversible or irreversible, as might be expected for an Abi system.I would expect that more detailed insight into the specific mechanism of antiphage activity be provided for publication in Nature Micro.
We have now performed a time course experiment with Tha-1/ORF61 expression, where we remove aliquots over time and see if the number of colony forming units (CFUs) decreases over time, in Fig. 3c:

Fig 3c
If Tha-1 toxicity was fully reversible, we would expect the number of CFUs to remain stable over time in the Tha-1-ORF61 scenario.However, we see an early drop in CFUs already after 10 minutes, where only 10% of the original cells are still viable, and the number gradually decreases further.The number of escapers is negligible (the number of cells that can grow on +aTc plates).We were a bit surprised by the high toxicity of Tha activation, but it likely means that Tha is very potent once activated.We must interpret this data carefully, since the expression levels of the proteins from the plasmids are likely higher than the physiological levels during immunity against a phage.However, we are unable to perform this experiment in a wild-type setting since i) we do not know exactly when Tha is triggered during phage infection, and ii) the phage will express a lot of toxic genes, which can lead to cell death regardless of Tha activity.Still, since ORF61 is expressed late in the phage life cycle, we think that at this point, the cell is doomed from the combination of Tha activity and toxic phage genes.So regardless of whether Tha causes cell death directly, the cell likely dies after Tha has been triggered under physiological conditions.This is mentioned in the discussion, which reads (lines 525-528): "We observed rapid cell death upon activation of Tha-1 at non-physiological protein levels (Fig. 3c).Yet, we suspect that during phage infection, the minor tail protein trigger is expressed so late that the host cell is unlikely to survive, either through Tha-mediated cell death or toxic phage-expressed gene products." Major concerns I found the term "regulator" used in reference to the activity of the Xis protein (e.g., lines 29, 74-76) confusing as this brings to mind transcription factor type regulatory proteins.At the protein interaction level these types of biological interactors are usually referred to as inhibitors.The authors should consider changing this terminology.
We thank the reviewer for pointing this out.We have changed the terminology in the lines mentioned.Lines 30-32: "To avoid autoimmunity and allow the phage life cycle to complete, these systems are inhibited by a small overlapping gene previously thought to encode the phage excisionase."Lines 83-84: "a gene previously annotated as excisionase (xis) and renamed ith-1 (Inhibitor of Tha-1), inhibits Tha both through protein-protein interactions and antisense transcription." We have also refrained from the regulator terminology throughout the rest of the manuscript.
The MOI of 10 used in Fig. 2b did not result in complete cell lysis or premature cell lysis.Did the authors determine if the CFUs/ml are changing throughout infection; are the cells lysing or is there just bacteriostatic activity that is relieved at some point?Of the cells that regrow, what are they?Resistant mutants, lysogens?
In the infection at MOI 5 in Fig. 2b, and at MOI 10 in the old Fig. 2b, there is indeed no lysis observed, premature or otherwise.This is consistent with the lack of lysis during induction of the 80α prophage with an ith-1 stop codon (Fig. 3e), which does not lead to a delayed lysis (as it would for an eventual relief of the bacteriostatic state).To us it is also intuitive, since a non-specific RNase-induced growth arrest should not directly lyse the cell, and the phage will likely not produce sufficient holins and lysins in a growth arrested environment.We think that the cells that do grow at higher MOIs are the uninfected cells, and not survivors following phage clearance and resumption of growth.This is also supported by the toxicity observed in Fig. 3c (see above).Even with the majority of cells being infected and arrested, there will be many cells that are uninfected, and grow rapidly to a high OD.We therefore do not think the regrown cells during protection are resistant mutants or lysogens.
In response to the reviewer's suggestion, we attempted an experiment where we infect cells and measure CFUs and PFUs over time.In this context, we want to see a drop in CFUs shortly after infection, proportional to the drop in PFUs, and indicating that the infected cells are unviable (due to Tha immunity or phage toxicity).If, unexpectedly, there was a drop in PFUs but not in CFUs, this would show that the bacteriostatic state could be relieved.However, this experiment did not work since we do not see a marked drop of CFUs or PFUs in up to 20 minutes post-infection (not shown here, and not included in the manuscript).The lack of PFUs going down shows that the adsorption rates for our phage are low during a short period of time, and it makes impossible to detect CFU drops and help prove/disprove an Abi mechanism.We decided not to attempt incubating for longer than 20 minutes, since then the cells would divide so much that the CFU drop from infection will be eclipsed by cell division and with an increase in CFUs.
Of the cells that regrow in the condition without protection (80α Δ(ith-1+tha-1 ), we suspect that these survive due to lysogenisation and superinfection exclusion.However, we are unable to show this directly, since Φ7206 integrates in the same chromosomal location as 80α, and we could not detect Φ7206 in cells after the experiment.We think that even without integrating, Φ7206 can enter an episomal, transient prophage state that protect against external Φ7206 infection during the duration of the experiment, or at least until reaching stationary phase.We still believe the lysogenisation is the reason for this, since we also infected RN4220 cells (without an 80α prophage) in the same experiment at MOI 5 (not in the manuscript, but shown below), and these recover and grow at exactly the same rate as the 80α Δ(ith-1+tha-1) cells.In these RN4220 cells ("RN4220 5"), we do detect integrated Φ7206 at the end of the experiment.Thus, overall, we believe that the 80α Δ(ith-1+tha-1) cells that recover, despite of no Tha-1 protection, are due to transient superinfection exclusion from Φ7206 cells.

Figure not in the manuscript
So, in short, we suspect that infected cells with Tha-1 stop to grow and cannot be rescued from the bacteriostatic state, and lyse very slowly (with fewer phages released) or not at all.We are unable to prove this directly, and would need microscopy, which we think is beyond the scope of this paper.
Recovering cells in the non-protected cells lyse, and the cells that recover do so due to superinfection exclusion by the invading phage.
Quantification of changes in mRNA during phage infection of 7206 in a wild-type vs.Tha-1 prophage mutant should be done to determine if there is an arrest in all cellular activities, or if there is mainly a loss of transcription of phage products.
We did not perform this experiment because the results would be difficult to interpret.Quantification of phage mRNAs would not prove that mRNAs have been cleaved by Tha-1, since changes could also come from a decrease in transcription of phage genes due to the general growth arrest.To add to this, the growth arrest from Tha-1/ORF61 co-expression (Figs.3a, c-d) shows that Tha-1 does cause general growth arrest also in the absence of phage, confirming that host RNA is targeted.It is highly likely that phage transcripts are also targeted, but even in the absence of this, Tha-1 would still effect immunity by shutting the host down.
The authors must show that the H270 mutant is expressed and soluble to the same levels as wildtype THA.While the His residue may be involved in nucleic acid cleavage, mutation of this histidine OD (600 nm) residue may also lead to decreased protein expression/stability or solubility.A Western blot would provide evidence that it is the specific loss of enzymatic activity rather than instability/insolubility that causes the loss of activity.
Following the reviewer's suggestion, we performed a western blot of wild-type Tha-1 and the H270A mutant Tha-1, which shows similar expression levels.This data is included in Fig. S3a: Do phages that bypass THA defence encode a similar Xis, or is it strictly differences in the minor tail protein target sequence?Some additional alignments would help paint a clearer picture of the relationship between the Xis and Tha proteins.
We thank the reviewer for this question, and we have performed phylogenetic analyses on ORF61/ORF54, Ith-1/2 (previously Xis), and Tha-1/2, which is Fig.S7.This shows that the sensitivity of phages to Tha immunity can mostly be explained by whether they encode their own ith-1/2.More phages have ith-2, which we think helps them to avoid immunity from Tha-2 from ΦNM2.The minor tail protein sequences can in part help explain Tha sensitivity as well (for example the middle cluster of phages for ORF61 does not activate Tha-1 immunity, and Φ12/ΦSLT and the lytic phages lack homologues to ORF61/ORF54, which is why they are not targeted).We have added lines to explain this in the figure legends for Fig. S7: "Phages Φ12, ΦSLT, Stab21, and K, which are not targeted by Tha-1/2 in Fig. 2a, lack ORF61/ORF54 homologues.
We are not including additional alignments, since we think the phylogenetic trees convey the message in a clearer way.
The escape mutants were done in a plasmid-based system and don't provide biological context if Xis is able to allow phage infection in the presence of Tha from a prophage.Can the repressor of another phage be exchanged with 80a and see if 80a (new repressor) can plate on wild type 80a lysogen?Or can the Xis from 80a be put into 7206?
We think the ith-1/2 presence on the phages (discussed above) combined with the interference data in Fig. 2a helps answer the question, and shows that ith genes can be used as immune evasion factors for the incoming phage.Indeed, two phages (ROSA and Φ7206) encode only ith-2, not tha-2.These two phages are insensitive to Tha-2 interference in Fig. 2a, which means that their ith-2 genes are only used to circumvent the Tha-2 system in lysogens, and not for regulating their own cognate Tha-2 system.From transcriptomic data (Fig. S6), we know that the ith-1 gene of 80α is expressed early during phage infection, presumably to avoid targeting by Tha-1, and we think this is also true for other phages.
In addition, from Fig. 2a, we know that 80α is largely insensitive to plasmid-expressed Tha-1.However, with a mutant 80α (which lacks ith-1 and has a non-functional version of tha-1 to allow the phage to obtain sufficient titres) (80α Δith-1, tha-1 F32L ), infection is blocked by Tha-1.Therefore, ith-1 of the wild-type 80α can normally block the Tha-1 of the host.This data is included in Fig. 4b: While the reviewer is questioning whether this would also hold true when Tha is expressed from a lysogen instead of a plasmid, we think this assumption is sensible to make.Fig. 2a shows that if anything, protection by Tha is sometimes stronger when from a plasmid (presumably due to higher expression), so if the ith-1 of 80α can block plasmid-encoded Tha-1 immunity, it will even more easily be able to block prophage-encoded Tha-1.While the suggestions of the reviewer (80α with a different repressor, or moving the 80α ith-1 to Φ7206) could work and would be the most natural scenario, we think that the ith-1/2 correlation with phage sensitivity in combination with the data in Fig. 4b clearly shows that the ith gene of an incoming phage can circumvent Tha inhibition.Also, the phage engineering required would be time-consuming, and it might be hard to make 80α function with a new repressor.
Are the Xis proteins broadly active or are they Tha sequence specific?Co-expressing the 80a and NM2 Xis/Tha-1/2 in the presence of bypassing phage would add knowledge for Tha functionality.
We have now addressed this question, and we have several lines of evidence for the specificity of Tha-1/2 and Ith-1/2.Firstly, as discussed above, the phylogenetics of Tha and Ith shows they are classified into two groups, Tha-1 and Ith-1, exemplified by 80α, and Tha-2 and Ith-2, exemplified by ΦNM2.In the interference data in Fig. 2a, no phage with Ith-1 is strongly targeted by Tha-1, and no phage with Ith-2 is strongly targeted by Tha-2.
Secondly, we performed an experiment to demonstrate that there is no cross-talk between the two types of Ith/Tha.We used a two-plasmid system, where one plasmid expressed Tha-1 or Tha-2, and the second plasmid encoding Ith-1 or Ith-2.Using Φ7094, a phage sensitive to both Tha-1 and Tha-2, we showed that plaqueing could only occur when Tha-1 and Ith-1 were together, and when Tha-2 and Ith-2 were together.This is because the cognate Ith inhibited the Tha immunity.The Ith-2 of ΦNM2 could not inhibit the Tha-1 of 80α, or vice versa.This is included in Fig. 4c:

Fig 4c
Based on this, we believe that there are two types of Ith/Tha systems, which do not cross-react.
Alternatively, can a chimera or truncated form of Tha be generated to determine if the N-terminal domain is essential for recognition of the phage tail proteins?
While this would be a nice proof of concept, we think this experiment is beyond the scope of the current manuscript.We do indeed believe that the N-terminal domain of Tha-1/2 is responsible for sensing ORF61/ORF54, while the C-terminal HEPN domain is the effector.We think it would be quite hard to get the chimera to work, since we would need to get the exact domain boundaries correct to allow the binding of the minor tail protein to relay the correct conformational change to the C-terminal domain.We think this would require a lot of trial and error, and it might not be possible.Instead, we hope that the Ith/Tha non-cross-reactivity described above and the remaining phage sensitivity data is sufficient to understand that activates and inactivates Tha immunity.

Minor concerns
Figure 1a -Reference to the significance of the * and ** should be made in the figure legend.
Correct.However, with the introduction of the schematic showing the different 80α mutants in Fig. 6a, the * and ** are no longer needed, and have been removed.
Figure 2a -The blank spaces make this heatmap difficult to read.Do the clear/white boxes represent no inhibition at all and the clear boxes with * mean infection to wild-type levels but the plaques were smaller?It should be clearly stated if these data are PFU/mL or fold-inhibition.
White spaces have been added, improving the figure.White means no inhibition, and * means small plaque phenotype, which gives similar numbers of plaques that are smaller than on the control strain.This is now stated in the Fig. 2a legend: "White squares signify no immunity.Prophage repressor indicates repression of the incoming phage by its cognate phage repressor.SP, small (but similar number of) plaques." With regards to the type of inhibition, the legend states: "The 10-fold reduction in plaqueing efficiency is shown from at least three independent experiments" This means fold-change, and not PFU/ml.Addressed, the first orf3 has been changed to orf2.

Fixed
Line 155-162.This could be reworded to improve clarity.The movement between prophages and plasmid-expressed protein (I assume that is what Tha-1 alone means) is confusing.It would be useful to state how many phages were inhibited only by Tha-1 from the prophage.
Reworded.The relevant section has more details for clarity.
Line 390 -There is an empty pair of brackets.
A missing reference has been added Supplemental Figure 1g -The labels say Φ11 but the figure legend says 80a.
The legends of S1f-g now say Φ11.

Reviewer #3
I have reviewed the manuscript.I find the science presented in the paper fascinating.Although the authors emphasized that their findings represent a new mechanism of phage exclusion, e.g.Crispr-CAS, I believe the data reveals a fascinating mechanism of temperate phage biology.The experiments are performed logically with clear results.I recommend publication of the manuscript in Nature.
We extend our gratitude to the reviewer for their constructive criticism and their appreciation of our research.Furthermore, upon careful consideration of their comments, we have recognised the significance of incorporating immunity and counter-immunity as integral components of the temperate phage life cycle.As a result, we have made the necessary modifications to our manuscript to introduce and substantiate this innovative concept.
But in my opinion the manuscript as presented is too long and wordy.I suggest it should be thoroughly edited for brevity and crispy reading.I also have some comments that the authors may respond to before the manuscript is accepted.My comments are not in order of chronology or in terms of importance although some are more important than others: 1. Of course, as I mentioned, it should be thoroughly edited.
We have thoroughly edited the manuscript for brevity and readability, and we hope this makes it easier to read and digest.This includes changes suggested by the reviewer below.However, with the inclusion of additional data due to concerns raised by the reviewers, we have also added new sections.Depending on the editor's point of view, we might also be able to shorten it further, but we wonder if this might also decrease readability for the broader readership of Nature Microbiology, since it would assume more prior knowledge of the field to understand without a detailed explanation.
2. I don't know why the gene referred to as xis was called so -may be because of homology or gene location -but since the authors have shown that it does not function as the 'classical' excisionase, I suggest to change it to one that more reflects its demonstrated function.
The gene was originally labelled xis due to a publication where they conclude a similar gene in phage L54a functioned as an excisionase (PMID: 2526804).We think this is untrue, also in the case of L54a.
We agree that it is sensible to rename the gene, and we have chosen Ith, standing for Inhibitor of Tha.Both Tha and Ith come in two flavours, Tha-1 and Ith-1, exemplified by 80α, and Tha-2 and Ith-2, exemplified by ΦNM2 (Fig. S7).This naming convention makes it intuitive that type 1 and 2 recognise different minor tail proteins (ORF61 and ORF54, respectively), and that there is no crossreaction (inhibition) between the components of each Tha/Ith pair, e.g.Tha-1 is inhibited by Ith-1, but not by Ith-2 (see Fig. 4c) 3. The triangular symbols at the top of, for example, figure 3 is misleading the way they are presented.If they reflect Dilution as noted, then it should be horizontally inverted reflecting more and more dilutions; as presented they reflect phage titers, not dilutions.
This is true, and we have changed the figure text and figure legends to accurately describe the triangles.
4. The manuscript misleadingly and frequently uses the word Tha1 when they really mean Tha+ wild type situations.Please carefully edit the entire manuscript about this point.Was the mutant Tha1 gene cloned in the plasmid?In line 166, the authors certainly mean, "….In the absence of WT Tha gene".Also, in line 143 , same problem.
As explained above, we categorised Tha into two different types, Tha-1, exemplified by 80α, and Tha-2, exemplified by ΦNM2.Tha-1 and Tha-2 recognise different minor tail proteins, and are inhibited by different Ith proteins.Both Tha-1/2 are Tha systems, however.We therefore believe the distinction is important, especially with the inclusion of the phylogenetic data in Fig. S6, which helps explain the interference data seen in Fig. 2a.From this, it seems that phages carrying an ith-2 gene cannot be targeted by Tha-2, giving the ith-1/2 genes an additional role, namely immune evasion during infection, as well as the previous role in regulating the cognate Tha system.
The above was not clear in our previous version of the manuscript, both because we did not explain thoroughly, and because we did not elaborate on the prevalence of the two types of Ith/Tha.We have included wording to explicitly explain the classification in the new version.Lines 181-187 read: "We observed that multiple phages have an identical genetic architecture and composition to 80α in their lysogeny regions, including ith-1/tha-1 pairs.Phages like Φ11 and Φ69 have almost identical ith-1-tha-1 genes to 80α.A second group of phages, including ΦNM2 (Fig. S2a) and ΦNM1, have similar genetic structures but the genes vary slightly in their sequences from the first group (Fig. S2b), containing ith-2 and tha-2, which is even more abundant in the databases." It is also mentioned in several other places, e.g. in lines 333-335: "Since 80α and ΦNM2 encode tha-1 and tha-2, respectively, activated by different minor tail proteins, we wondered if their respective ith-1 and ith-2 genes have different immune suppression specificities." Finally, to simplify, we have abandoned the use of THA to describe the system, referring to it only as Tha.
See above.Tha-1 and Tha-2 refer to different wild-type versions of the immune protein.
6.I strongly disagree with the use of the word non-contiguous operon.It is very misleading word to describe the case it refers to.I Insist that the authors use a different word to describe overlapping antisense operons.
We appreciate that the use of the term noncontiguous operon has proven controversial.We have therefore toned down the emphasis of this term, removing it from the title, and discussing it less in the introduction and the discussion sections.Where we do use it, we usually define it in each case, and use it along with alternative phrases (like antisense operon), as well as often mentioning it along with the eukaryotic counterpart phrase, "complex transcripts".The reviewers find that the paper has improved in revision, and therefore we'll be happy in principle to publish it in Nature Microbiology, pending minor revisions to satisfy the referees' final requests and to comply with our editorial and formatting guidelines.

Decision
If the current version of your manuscript is in a PDF format, please email us a copy of the file in an editable format (Microsoft Word or LaTex)--we can not proceed with PDFs at this stage.
We are now performing detailed checks on your paper and will send you a checklist detailing our editorial and formatting requirements in about a week.Please do not upload the final materials and make any revisions until you receive this additional information from us.
Thank you again for your interest in Nature Microbiology Please do not hesitate to contact me if you have any questions. Sincerely,

Final Decision Letter:
Message : 4th March 2024 Dear Professor Penadés, I am pleased to accept your Article "Bacteriophages avoid autoimmunity from cognate immune systems as an intrinsic part of their life cycles" for publication in Nature Microbiology.Thank you for having chosen to submit your work to us and many congratulations.
Over the next few weeks, your paper will be copyedited to ensure that it conforms to Nature Microbiology style.We look particularly carefully at the titles of all papers to ensure that they are relatively brief and understandable.
Once your paper is typeset, you will receive an email with a link to choose the appropriate publishing options for your paper and our Author Services team will be in touch regarding any additional information that may be required.Once your paper has been scheduled for online publication, the Nature press office will be in touch to confirm the details.
You may wish to make your media relations office aware of your accepted publication, in case they consider it appropriate to organize some internal or external publicity.Once your paper has been scheduled you will receive an email confirming the publication details.This is normally 3-4 working days in advance of publication.If you need additional notice of the date and time of publication, please let the production team know when you receive the proof of your article to ensure there is sufficient time to coordinate.Further information on our embargo policies can be found here: https://www.nature.com/authors/policies/embargo.html After the grant of rights is completed, you will receive a link to your electronic proof via email with a request to make any corrections within 48 hours.If, when you receive your proof, you cannot meet this deadline, please inform us at rjsproduction@springernature.com immediately.You will not receive your proofs until the publishing agreement has been received through our system Due to the importance of these deadlines, we ask you please us know now whether you will be difficult to contact over the next month.If this is the case, we ask you provide us with the contact information (email, phone and fax) of someone who will be able to check the proofs on your behalf, and who will be available to address any last-minute problems.
Acceptance of your manuscript is conditional on all authors' agreement with our publication policies (see https://www.nature.com/nmicrobiol/editorial-policies).In particular your manuscript must not be published elsewhere.
Please note that Nature Microbiology is a Transformative Journal (TJ).Authors may publish their research with us through the traditional subscription access route or make their paper immediately open access through payment of an article-processing charge (APC).Authors will not be required to make a final decision about access to their article until it has been accepted.Find out more about Transformative Journals Authors may need to take specific actions to achieve compliance with funder and institutional open access mandates.If your research is supported by a funder that requires immediate open access (e.g. according to Plan S principles) then you should select the gold OA route, and we will direct you to the compliant route where possible.For authors selecting the subscription publication route, the journal's standard licensing terms will need to be accepted, including self-archiving policies.Those licensing terms will supersede any other terms that the author or any third party may assert apply to any version of the manuscript.
If you have any questions about our publishing options, costs, Open Access requirements, or our legal forms, please contact ASJournals@springernature.com An online order form for reprints of your paper is available at https://www.nature.com/reprints/author-reprints.html.All co-authors, authors' institutions and authors' funding agencies can order reprints using the form appropriate to their geographical region.
We welcome the submission of potential cover material (including a short caption of around 40 words) related to your manuscript; suggestions should be sent to Nature Microbiology as electronic files (the image should be 300 dpi at 210 x 297 mm in either TIFF or JPEG format).Please note that such pictures should be selected more for their aesthetic appeal than for their scientific content, and that colour images work better than black and white or grayscale images.Please do not try to design a cover with the Nature Microbiology logo etc., and please do not submit composites of images related to your work.I am sure you will understand that we cannot make any promise as to whether any of your suggestions might be selected for the cover of the journal.
You can now use a single sign-on for all your accounts, view the status of all your manuscript submissions and reviews, access usage statistics for your published articles and download a record of your refereeing activity for the Nature journals.
To assist our authors in disseminating their research to the broader community, our SharedIt initiative provides you with a unique shareable link that will allow anyone (with or without a subscription) to read the published article.Recipients of the link with a subscription will also be able to download and print the PDF.
As soon as your article is published, you will receive an automated email with your shareable link.
With kind regards,

Fig 3a :
Fig 3a: Please invert the colors to distinguish this from a plaque assay.Fig 4a: Grid lines would be nice.Fig.4b and c: It would be nice to see total coverage in both directions rather than aligned reads (are these all the reads that mapped?).

Fig 5c :
Fig 5c: The caption says green and blue, I'm seeing green and orange.Fig 5: A schematic similar to Fig. 1a, but with the position of Pxis, xis would be helpful to follow the several different constructs.Perhaps the authors should rename xis since they show it is likely not an excisionase.

Figure 1a -
Figure 1a -Reference to the significance of the * and ** should be made in the figure legend.Figure2a-The blank spaces make this heatmap difficult to read.Do the clear/white boxes represent no inhibition at all and the clear boxes with * mean infection to wild-type levels but the plaques were smaller?It should be clearly stated if these data are PFU/mL or fold-inhibition.Figure3-It would be easier for the reader if this figure was inverted; as presented these look like plaques, not colonies.FigureS1b-There are two orf3's noted.Line 32 -Noncontiguous is spelled incorrectly.Line 155-162.This could be reworded to improve clarity.The movement between prophages and plasmid-expressed protein (I assume that is what Tha-1 alone means) is confusing.It would be useful to state how many phages were inhibited only by Tha-1 from the prophage.Line 390 -There is an empty pair of brackets.Supplemental Figure1g-The labels say Φ11 but the figure legend says 80a.

Fig 3a :
Fig 3a: Please invert the colors to distinguish this from a plaque assay.

Fig 5c :
Fig 5c: The caption says green and blue, I'm seeing green and orange.

Figure 3 -
Figure 3 -It would be easier for the reader if this figure was inverted; as presented these look like plaques, not colonies.
submitting your revised manuscript "Bacteriophages avoid autoimmunity from cognate immune systems as an intrinsic part of their life cycles" (NMICROBIOL-23030723C).It has now been seen by the original referees and their comments are below.