Toxin/antitoxin systems induce persistence and work in concert with restriction/modification systems to inhibit phage

ABSTRACT Myriad bacterial anti-phage systems have been described and often the mechanism of programmed cell death is invoked for phage inhibition. However, there is little evidence of “suicide” under physiological conditions for these systems. Instead of death to stop phage propagation, we show here that persister cells, i.e., transiently-tolerant, dormant, antibiotic-insensitive cells, are formed and survive using the Escherichia coli C496_10 tripartite toxin/antitoxin system MqsR/MqsA/MqsC to inhibit T2 phage. Specifically, MqsR/MqsA/MqsC inhibited T2 phage by 105-fold and reduced T2 titers by 3,000-fold. During T2 phage attack, in the presence of MqsR/MqsA/MqsC, evidence of persistence includes the single-cell physiological change of reduced metabolism (via flow cytometry), increased spherical morphology (via transmission electron microscopy), and heterogeneous resuscitation. Critically, we found restriction-modification systems (primarily EcoK McrBC) work in concert with the toxin/antitoxin system to inactivate phage, likely while the cells are in the persister state. Hence, a phage attack invokes a stress response similar to antibiotics, starvation, and oxidation, which leads to persistence, and this dormant state likely allows restriction/modification systems to clear phage DNA. IMPORTANCE To date, there are no reports of phage infection-inducing persistence. Therefore, our results are important since we show for the first time that a phage-defense system, the MqsRAC toxin/antitoxin system, allows the host to survive infection by forming persister cells, rather than inducing cell suicide. Moreover, we demonstrate that the MqsRAC system works in concert with restriction/modification systems. These results imply that if phage therapy is to be successful, anti-persister compounds need to be administered along with phages.

before the phage can complete its life cycle, a mechanism analogous to apoptosis.This paper and others like it are important as the argument that bacteria persist and go dormant to survive phage infection rather than committing suicide is well reasoned and supported by data.I only have a few minor comments in no particular order.Line 130: is MqsC not involved here?Line 133: "...and produced MqssR/A/C..." This sentence is confusing.Do the authors mean T2 phage virions are produced?Line 143: Describe briefly for your non-phage expert readers what the EOP and ECOI assays are.Line 187: This sentence seems like it should be introduced at the beginning of the paragraph to help readers understand the timing of the phage lifecycle as it relates to the timing of persister formation.Line 194: The inactivation of McrBC and its impact on phage replication is important and should be included as a main figure rather than a supplement.Line 246: Inclusion of literature related to (p)ppGpp and lambda phage would be a welcome addition to the discussion: This manuscript reports the results of experiments designed to shed light on the roles that bacteriophage T2 and the toxin antitoxin (TA) module MqsRAC play in tolerance to phage and antibiotic challenges.TA modules are ubiquitous in bacteria and evidence suggests that they play roles in stabilizing extra-genomic elements, inducing latency in response to stress and thwarting the development of phage.Some of the data supporting roles for TA systems remain controversial including the formation of so-called persisters.The presence of TA modules throughout the bacterial kingdom and their potential role in pathogenesis and tolerance to antibiotic treatment make their study highly significant.Here, the authors present confirmatory evidence that the MqsRAC module plays an important role in suppressing infection of E. coli by phage T2.The authors also explore the effects of the antibiotics ampicillin and ciprofloxacin on the antiphage activity of MqsRAC.The results are interesting, but the authors' conclusions are unclear and not supported by the data.The presentation is muddled as the authors claim that T2 causes the formation of persisters that protect against the drugs, or that the drugs cause persisters that are resistant to the phage.Moreover, the distinction between persister cells and those that are merely tolerant to stress is notoriously difficult to assess, and not done correctly in this case.The manuscript might be improved by attention to the clarity of the conclusions as well as the following issues.1. Experiments are only done with T2, so the generality of the phenomena is unclear.The Laub lab showed that the Mqs system does not protect against T4, which is closely related to T2.If persistence is a general anti-phage protectant, it should work against another phage besides T2.This should be tested.2. The authors state that they employed 10X the MIC of ampicillin to test for the ability of Mqs to induce tolerance to phage.But the results shown in Figure 1 and S3 reveal that treatment of MG1655 with 10 X MIC of Amp has no effect on cell growth, whether Mqs is expressed or not.The standard for identification of persisters is treatment that eliminates 99% of the population (Balaban, et al, 2019).3. When challenged before Amp (or Cipro) treatment with T2, the Mqs-population is eradicated, suggesting a synergistic effect of the T2 and the drug (Fig. 1).This inhibition is significantly blocked by MqsRAC.The major question is why does T2 infection sensitize the cells to Amp (or Cipro) treatment?This is not consistent with the authors' conclusion that the formation of persisters by T2 and/or Msq is the explanation for the findings.4. The manuscript consistently contends that persisters are the result of T2 infection, but the effects measured require the presence of the Msq plasmid.This is confusing.5. Is it really a surprise that survival of cells to T2 is decreased by loss of one RM system?Same for MMC; of course, it will kill the cells whether or not they are non-growing.6.The y-axis in Figure 4 has a misspelling (hcange).

Preparing Revision Guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex.Go to Author Tasks and click the appropriate manuscript title to begin the revision process.The information that you entered when you first submitted the paper will be displayed.Please update the information as necessary.Here are a few examples of required updates that authors must address: • Point-by-point responses to the issues raised by the reviewers in a file named "Response to Reviewers," NOT IN YOUR COVER LETTER.
• Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file.
• Each figure must be uploaded as a separate file, and any multipanel figures must be assembled into one file.• Manuscript: A .DOC version of the revised manuscript • Figures: Editable, high-resolution, individual figure files are required at revision, TIFF or EPS files are preferred For complete guidelines on revision requirements, please see the journal Submission and Review Process requirements at https://journals.asm.org/journal/Spectrum/submission-review-process.Submissions of a paper that does not conform to Microbiology Spectrum guidelines will delay acceptance of your manuscript." Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me.If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail.Arrangements for payment must be made before your article is published.For a complete list of Publication Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees.Need to upgrade your membership level?Please contact Customer Service at Service@asmusa.org.
Thank you for submitting your paper to Microbiology Spectrum.

Response for Spectrum 03388-23 "Toxin/Antitoxin Systems Induce Persistence and Work in Concert with Restriction/Modification Systems to Inhibit Phage"
We wish to thank the editor and two reviewers for their help with this manuscript and feel the manuscript is much-improved due to the careful review.We have addressed all the issues as indicated below (the Reviewers' comments are underlined below and the changed text is highlighted yellow in the revised manuscript).Note the line numbers below refer to the revised text.
Specifically, we have conducted three sets of new experiments to better demonstrate that E. coli forms persister cells upon phage T2 attack: we show (i) ampicillin creates persister cells via a kill curve with MG1655/pCV1 (Fig. 1B), (ii) T2 phage makes persisters via a kill curve (6 th line of evidence of persistence with phage) for MG1655/pCV1 + T2 (Fig. 1B), and (iii) confirmed amp has reduced killing based on a pCV1 plasmid feature (but kills the plasmid-free host as expected) (Table R1 below).We also improved the figures (specifically Fig. 1A, Fig. 2, Fig. 4, Fig. 5, and Fig. S1).

Reviewer #1
The manuscript by Fernandez-Garcia provides evidence that toxin-antitoxin systems induce bacterial persistence and this persistence promotes cell survival after phage infection.The authors go on to show that in persister cells, restriction modification systems operate to protect 'sleeping' cells, presumably by destroying phage DNA.The authors frame their results in a way that contradicts the popular idea that phage defense systems protect cell populations from phage infection by killing the infected cell before the phage can complete its life cycle, a mechanism analogous to apoptosis.This paper and others like it are important as the argument that bacteria persist and go dormant to survive phage infection rather than committing suicide is well reasoned and supported by data.I only have a few minor comments in no particular order.
Thank you for your encouragement! 1. Line 130: is MqsC not involved here?
We agree and now list MqsC here (line 112) and apologize for the confusion.We discovered the binary (MqsR/MqsA) type II toxin/antitoxin system in E. coli K-12 in 2004 and characterized in subsequent years, but the current paper deals with a tripartite system (MqsR/MqsA/MqsA) from E. coli C496_10, identified by the Laub lab.To avoid confusion, we have added more detail to the Methods (line 57, where we now indicate the source), (line 112, where we indicate the function of each protein).
As suggested, we have made our points more clearly by updating this sentence in the first paragraph of the Results.First, we broke it into two sentences.Then, we added detail to explain what the 'same expression system' means (natural promoter, previously indicated on line 48 of the Intro) and explained better why we focused on T2 phage (since it was previously shown to be inhibited by MqsRAC) (line 115).As suggested, we added text to explain the efficiency of plating assay (EOP) and the efficiency of center of infection (ECOI) (line 125): "… we performed both an efficiency of plating (EOP) assay (to determine the number of phages present (1)) and an efficiency of the center of infection (ECOI) assay (to determine the number of pre-adsorbed phages (2))." 4. Line 187: This sentence seems like it should be introduced at the beginning of the paragraph to help readers understand the timing of the phage lifecycle as it relates to the timing of persister formation.
As suggested, we moved this sentence (line 167).Thank you for this suggestion.

Line 194:
The inactivation of McrBC and its impact on phage replication is important and should be included as a main figure rather than a supplement.
As suggested, we have moved Table S7 to the main text (now Table 1)."Additional evidence of the role of the inhibitory role (p)ppGpp during phage infection includes that lambda phage is not able to replicate with high (p)ppGpp concentrations due to (p)ppGpp-mediated inhibition of transcriptional activation of the lambda origin of replication (3) and that (p)ppGpp inhibits lambda P R by inhibiting formation of the first phosphodiester bond (4).Other (p)ppGpp-relevant phage results include that ppGpp stimulates lambda paQ promoter due to an increased rate of productive open complex formation (5) and that (p)ppGpp controls the lambda lysis-versus-lysogenization decision based on its differential influence of the activities of the lambda pL, pR, pE, pI, and paQ promoters (6)."

Reviewer #2
This manuscript reports the results of experiments designed to shed light on the roles that bacteriophage T2 and the toxin antitoxin (TA) module MqsRAC play in tolerance to phage and antibiotic challenges.TA modules are ubiquitous in bacteria and evidence suggests that they play roles in stabilizing extra-genomic elements, inducing latency in response to stress and thwarting the development of phage.Some of the data supporting roles for TA systems remain controversial including the formation of so-called persisters.The presence of TA modules throughout the bacterial kingdom and their potential role in pathogenesis and tolerance to antibiotic treatment make their study highly significant.Here, the authors present confirmatory evidence that the MqsRAC module plays an important role in suppressing infection of E. coli by phage T2.The authors also explore the effects of the antibiotics ampicillin and ciprofloxacin on the antiphage activity of MqsRAC.The results are interesting, but the authors' conclusions are unclear and not supported by the data.The presentation is muddled as the authors claim that T2 causes the formation of persisters that protect against the drugs, or that the drugs cause persisters that are resistant to the phage.Moreover, the distinction between persister cells and those that are merely tolerant to stress is notoriously difficult to assess, and not done correctly in this case.The manuscript might be improved by attention to the clarity of the conclusions as well as the following issues.
1. Experiments are only done with T2, so the generality of the phenomena is unclear.The Laub lab showed that the Mqs system does not protect against T4, which is closely related to T2.If persistence is a general anti-phage protectant, it should work against another phage besides T2.This should be tested.
The Laub group has already tested two phages (T2 and T4) and only found activity with T2, as indicated by the reviewer.However, we do not agree with the logic of the reviewer since if our hypothesis and data show persister cells are only made by MqsRAC, with no effect with the empty plasmid system, then we will only see effects with MqsRAC when it is actively inhibiting phage (and there is no dispute that MqsRAC inhibits T2 phage as the Laub group discovered this and we confirm it here).Hence, there will only be persisters formed when MqsRAC is active (i.e., RNase MqsR cleaves most mRNA and the cell becomes dormant), so there is no need to test other phages (e.g., T4) as MqsRAC is not active with T4 so no persisters should be formed and we will not glean much from this experiment (i.e., no different that the empty plasmid system).So, the requested experiment serves only as a negative control and we have already shown, with two antibiotics, that without MqsRAC but with T2 phage, there are no persisters formed (see T2 + Amp and T2 + cipro results in Table S3).
2. The authors state that they employed 10X the MIC of ampicillin to test for the ability of Mqs to induce tolerance to phage.But the results shown in Figure 1 and S3 reveal that treatment of MG1655 with 10 X MIC of Amp has no effect on cell growth, whether Mqs is expressed or not.The standard for identification of persisters is treatment that eliminates 99% of the population (Balaban, et al, 2019).
The reviewer makes an insightful observation, that 10X MIC kills less than expected for pCV1-bearing cells.First, 10X MIC kills ~90% of the pCV1-bearing cells, as shown in Table 1/Table S3 in the text, rather than the expected survival of ~0.1% (so we disagree that ampicillin has 'no effect').This is one reason we confirmed our results with a second class of antibiotics, ciprofloxacin, which has expected killing with pCV1-based plasmids (0.2%, Table S3).
Second, as suggested, we re-examined the results of Table S3 since the results with cells containing pCV1 have higher than expected ampicillin survival.As indicated previously in our results, we found that the presence of both plasmid pCV1 and pCV1-mqsRAC increase ampicillin resistance consistently by over 100X compared to the host MG1655 alone which has the expected 99.99% killing (100 µg/mL, 10 MIC ampicillin, for 3 h, OD = 0.5 so exponentially-growing cells, i.e., same conditions as in the paper):

Strain
Exponential (OD600 = 0. We added this information to the caption of Table 1/Table S3 and clarified the text (line 136).
We had already sequenced both pCV1 and pCV1-mqsRAC (to know what we were using and ensure our results were dependent on MqsR/MqsA/MqsC) but unfortunately, we cannot determine why the plasmid increases Ap tolerance.There is nothing wrong with our approach as the host without the plasmid is killed in the expected manner with only 0.01% survival.To prevent confusion, we have explained this in the captions of Tables S3/S7.Note that pCV1 and pCV1-mqsRAC also consistently increased Amp tolerance in the stationary phase but stationary phase results with Amp are not as meaningful since Amp is less effective with slow-growing stationary-phase cells, of course.Furthermore, it is good we confirmed our results with ciprofloxacin to show phages induce persistence.
The plasmid map for our sequenced pCV1 is shown below (Figure R1).Therefore, our ampicillin has the expected killing for MG1655 (0.01% survival or 99.99% killing) and the only anomaly is that there is some pCV1-based factor, present for both pCV1 and pCV1-mqsRAC, that makes this system somewhat more tolerant to ampicillin.But again, we have confirmed all of our results with a separate antibiotic (cipro).And remember there is a 10 8 -fold phenotype in regard to survival with and without T2 phage with ampicillin treatment (Fig. 1A).
We also have performed an additional experiment with ampicillin, generating a kill curve, as indicated below.
3. When challenged before Amp (or Cipro) treatment with T2, the Mqs-population is eradicated, suggesting a synergistic effect of the T2 and the drug (Fig. 1).This inhibition is significantly blocked by MqsRAC.The major question is why does T2 infection sensitize the cells to Amp (or Cipro) treatment?This is not consistent with the authors' conclusion that the formation of persisters by T2 and/or Msq is the explanation for the findings.
We disagree.The whole point is MqsRAC inhibits T2 phage by making persister cells.We previously provided not one but instead 5 lines of evidence showing persister cells are formed: (i) multiple antibiotic tolerance (ciprofloxacin and ampicillin), (ii) heterogeneous resuscitation, (iii) metabolic inactivity via flow cytometry, (iv) morphology of persister cells via TEM, and (v) MMC kills the MqsRAC-induced persister cells.Hence, our conclusion is sound that T2 fails to kill the MqsRAC-containing cells due to the formation of persister cells.
To provide a 6 th line of evidence that persister cells are formed, we now show kill curves for MG1655/pCV1 + T2 (0.1 MOI) and for MG1655/pCV1 + 10 MIC Amp for more evidence (also addresses point #2 showing persisters form amp with pCV1) (Fig. R2 and in Fig. 1B in the text); hence, both T2 phage and ampicillin show the gold standard for persistence: a plateau in killing.
4. The manuscript consistently contends that persisters are the result of T2 infection, but the effects measured require the presence of the Msq plasmid.This is confusing.
We apologize for the confusion.The main point of the manuscript is to demonstrate that a toxin/antitoxin system does not induce cell suicide (as claimed previously) but instead induces persistence.We have provided 6 lines of evidence to demonstrate our claim: (i) multiple antibiotic tolerance (ciprofloxacin and ampicillin), (ii) heterogeneous resuscitation, (iii) metabolic inactivity via flow cytometry, (iv) morphology of persister cells via TEM, (v) MMC kills the MqsRAC-induced persister cells, and (vi) kill curves show both ampicillin and T2 phage form persister cells.In addition, and perhaps the cause of the confusion, the MqsRAC system is not native to E. coli K-12 (the host we used for all the experiments) but instead is from E. coli C496_10; we now have tried to make this more clear (see lines 57 & 112 & 116).
5. Is it really a surprise that survival of cells to T2 is decreased by loss of one RM system?Yes, especially since it is a completely novel result; i.e., it has never been shown before that TAs work with any other phage defense method.Moreover, of all phage defense systems, only CRISPR-Cas has been shown to work with restriction enzymes.So we find that not only does MqsRAC not work alone, we identified the other phage defense method with which it functions.
Same for MMC; of course, it will kill the cells whether or not they are non-growing.
We agree this result is straightforward, but its utility is that it provides additional proof for persistence as MMC is known to kill persister cells (7).Furthermore, we wish to demonstrate that phages alone will lead to persister cells and should be combined with anti-persister measures such as MMC (and we have shown here for the first time that the persister cells generated my TAs can be killed successfully with MMC).We note this is also novel and has not been shown previously.
Thank you very much for catching this!The y-axis has been corrected!While the reviews of the revised version are overall positive, a few important points by referee #3 are made that should be addressed prior to further consideration.Some of these can be addressed through text-only modifications, but others may require additional analysis and interpretation of the current data set.I will be more than happy to consider a revised manuscript that addresses these concerns.
Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me.If you do not wish to modify the manuscript and prefer to submit it to another journal, notify me immediately so that the manuscript may be formally withdrawn from consideration by Spectrum.

Revision Guidelines
To submit your modified manuscript, log into the submission site at https://spectrum.msubmit.net/cgi-bin/main.plex.Go to Author Tasks and click the appropriate manuscript title to begin.The information you entered when you first submitted the paper will be displayed; update this as necessary.Note the following requirements: • Any supplemental material intended for posting by ASM should be uploaded separate from the main manuscript; you can combine all supplemental material into one file (preferred) or split it into a maximum of 10 files, with all associated legends included For complete guidelines on revision requirements, see our Submission and Review Process webpage.Submission of a paper that does not conform to guidelines may delay acceptance of your manuscript.
Data availability: ASM policy requires that data be available to the public upon online posting of the article, so please verify all links to sequence records, if present, and make sure that each number retrieves the full record of the data.If a new accession number is not linked or a link is broken, provide Spectrum production staff with the correct URL for the record.If the accession numbers for new data are not publicly accessible before the expected online posting of the article, publication may be delayed; please contact production staff (Spectrum@asmusa.org)immediately with the expected release date.
Publication Fees: For information on publication fees and which article types are subject to charges, visit our website.If your manuscript is accepted for publication and any fees apply, you will be contacted separately about payment during the production process; please follow the instructions in that e-mail.Arrangements for payment must be made before your article is published.The authors have responded adequately to this reviewer's concerns.

ASM Membership
Reviewer #3 (Comments for the Author): In the recent years we saw a dramatic development in the field of antiphage immune systems.Many of these are operating at toxin-antitoxin (TA) modules and are commonly believed to act by promoting altruistic suicide of the infected cell, executing the so-called 'abortive infection'.However, the hard evidence for this mechanism is often not as hard as one would like it to be.The reason being that it is an experimentally challenging thing to directly observe.For one thing, when the phage infects the cell, it executes a hostile takeover program that disrupts normal cellular function, e.g.commonly the host chromosome is degraded (e.g. in the case of T2).Therefore, it is hard to say who killed the cell -was it the defence system?Or was the cell already killed by the phage and the defence system merely 'aborted the infection' in a sense that is commonly used by virologists studying the viruses of eukaryotes: aborted the efficient production of new viral particles, without implying the cell death.
Beta-lactam antibiotics such as ampicillin kill bacteria interfering with their cell wall synthesis.The efficiency of killing is proportional to the growth rate -the faster one grows, the more sensitive one is.However, bactericidal antibiotics (such as betalactams) do not sterilise bacterial cultures, some cells survive.When only a small fraction survives, these are dubbed as persisters.When the whole population is killed slowly, this is referred to as tolerance.The distinction between the two can be fuzzy.Expression of any toxic protein that slows down the growth leads to tolerance / persistence.TA effectors are toxic, and many slow down the growth engaged -which has led to active research on TAs role in persistence.It is a very controversial research topic muddled with botched up experimental systems, overinterpretation and hype.
In this study Fernández-García characterise a recently discovered antiphage toxin-antitoxin-chaperone (TAC) system MqsRAC that protects E. coli from T2.They characterise the interplay between MqsRAC and the classical antiphage systems: restriction systems/DNA processing enzymes and probe the connections between antibiotic persistence, MqsRAC and phage attack/defence.
As MqsRAC is triggered upon sensing the phage, it would compromise the bacterial growth, thus driving bacterial antibiotic tolerance.This is exactly what the authors see.I am less sure about the biological meaningfulness, really.
Note that for bacteria to 'persist' after the T2 attack, the chromosome needs to survive the T2-excercised chromosomal digestion.Does this mean MqsRAC TAC is triggered early during infection...I strongly recommend doing DNA staining.
Importantly the authors see that MqsRAC and restriction modification act together to defeat the phage attack.To my mind this is the most interesting result of the paper.However, these experiments (correct me if I am wrong) are only presented in the presence of ampicillin.I would really like to see these assays done in the absence of the antibiotic, with focus on PFUs, not CFUs.As it stands, the most biologically meaningful result is not fully exploited / documented.
Overall, while I think the data is interesting and thought-provoking, I am not sure it is a fully biologically watertight.

3 .
Line 143: Describe briefly for your non-phage expert readers what the EOP and ECOI assays are.

•
Upload point-by-point responses to the issues raised by the reviewers in a file named "Response to Reviewers," NOT IN YOUR COVER LETTER • Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file • Upload a clean .DOC/.DOCX version of the revised manuscript and remove the previous version • Each figure must be uploaded as a separate, editable, high-resolution file (TIFF or EPS preferred), and any multipanel figures must be assembled into one file

:
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees.Need to upgrade your membership level?Please contact Customer Service at Service@asmusa.org.The ASM Journals program strives for constant improvement in our submission and publication process.Please tell us how we can improve your experience by taking this quick Author Survey.