Oxidation of hemoproteins by Streptococcus pneumoniae collapses the cell cytoskeleton and disrupts mitochondrial respiration leading to the cytotoxicity of human lung cells

ABSTRACT Streptococcus pneumoniae (Spn) causes pneumonia that kills millions through acute toxicity and invasion of the lung parenchyma. During aerobic respiration, Spn releases hydrogen peroxide (Spn-H2O2) as a by-product of the enzymes pyruvate oxidase and lactate oxidase and causes cell death with signs of both apoptosis and pyroptosis by oxidizing unknown cell targets. Hemoproteins are molecules essential for life and prone to oxidation by H2O2. We recently demonstrated that during infection-mimicking conditions, Spn-H2O2 oxidizes the hemoprotein hemoglobin (Hb), releasing toxic heme. In this study, we investigated details of the molecular mechanism(s) by which the oxidation of hemoproteins by Spn-H2O2 causes human lung cell death. Spn strains, but not H2O2-deficient SpnΔspxBΔlctO strains, caused time-dependent cell cytotoxicity characterized by the rearrangement of the actin cytoskeleton, the loss of the microtubule cytoskeleton, and nuclear contraction. Disruption of the cell cytoskeleton is correlated with the presence of invasive pneumococci and an increase in intracellular reactive oxygen species. In cell culture, the oxidation of Hb or cytochrome c (Cytc) caused DNA degradation and mitochondrial dysfunction from inhibition of complex I-driven respiration, which was cytotoxic to human alveolar cells. Oxidation of hemoproteins resulted in the creation of a radical, which was identified as a protein-derived side chain tyrosyl radical by using electron paramagnetic resonance. Thus, we demonstrate that Spn invades lung cells, releasing H2O2 that oxidizes hemoproteins, including Cytc, catalyzing the formation of a tyrosyl side chain radical on Hb and causing mitochondrial disruption that ultimately leads to the collapse of the cell cytoskeleton. IMPORTANCE Streptococcus pneumoniae (Spn) colonizes the lungs, killing millions every year. During its metabolism, Spn produces abundant amounts of hydrogen peroxide. When produced in the lung parenchyma, Spn-hydrogen peroxide (H2O2) causes the death of lung cells, and details of the mechanism are studied here. We found that Spn-H2O2 targets intracellular proteins, resulting in the contraction of the cell cytoskeleton and disruption of mitochondrial function, ultimately contributing to cell death. Intracellular proteins targeted by Spn-H2O2 included cytochrome c and, surprisingly, a protein of the cell cytoskeleton, beta-tubulin. To study the details of oxidative reactions, we used, as a surrogate model, the oxidation of another hemoprotein, hemoglobin. Using the surrogate model, we specifically identified a highly reactive radical whose creation was catalyzed by Spn-H2O2. In sum, we demonstrated that the oxidation of intracellular targets by Spn-H2O2 plays an important role in the cytotoxicity caused by Spn, thus providing new targets for interventions.

1.The length of the manuscript is very long and suggested to shorten the introduction and materials and methods sections.2. The authors have used already published methods to validate their hypothesis and it is suggested to use the references instead of providing detailed methodology.3. Figure numbers are missing in each page in the bottom of the manuscript.It was difficult to correlate/interpret the figures with the text.
Reviewer #2 (Comments for the Author): Comments for the authors: The authors have interestingly revealed possible links between H2O2 produced by Streptococcus pneumoniae and the associated cytotoxicity of it.Pneumococcal WT lab strains and mutants producing little H2O2 were used to compare differences in cytotoxicity, and cytotoxicity was linked to oxidation of Hb-Fe3+.The topic is relevant, as it could contribute to understanding the progression of invasive disease, and resulting in new therapeutics options.There are some major concerns which will be addressed below.

Major comment
There is no data shown or mentioned regarding the differences in growth between the strains, nor is the amount of bacteria present at most measurements known to the reader (and the authors?).Since the authors mention that the both SpxB and LctO are part of a pathway that is responsible for the bulk of the ATP produced.The loss of these genes is likely to affect growth to a considerable degree, resulting in different bacterial loads between the WT and mutant conditions.Whether the results observed are majorly affected by these differences is presently impossible to determine, while it could be a very important factor.Different growth dynamics, might also result in a different ratio of live:dead bacteria at the time of measurement, which could release all sorts of molecules into the growth medium which could also affect the results.CFU/DNA copy numbers, or the addition of growth curves for the used strains (as supplementary material) would greatly improve the interpretation of the shown data.Intracellular growth might be quite different and difficult to measure, but in the data as presented, it is unclear what the dynamics of bacterial growth are.The observed difference in cytotoxicity could be simply because of differences in the bacterial load.

Minor comments
Many experiments are performed with 10 µM Hb-Fe3+, which is shown to oxidize and result in cytotoxicity.However, most Hb in vivo would be Hb-Fe2+ to start with, which would first need to be oxidized to Hb-Fe3+ (which has been observed and reported previously).Hb-Fe3+ could then be oxidized to Hb-Fe4+.In reference 19, most Hb-Fe2+ is stated to be oxidized into Hb-Fe3+ in experimental conditions.It is unclear to me how much of the Hb-Fe2+ would be expected to be oxidized twice.So whether this oxidation of Hb-Fe3+ that is shown, would occur in physiological conditions.On what do the authors base the idea that Hb-Fe3+ oxidation could be expected in invasive disease and would it not be more logical to add Hb-Fe2+, which, if it also occurs in in vivo conditions, could be oxidized twice and also form the radical as observed?Line 78: The authors mention Spx and LctO as the only enzymes that produce H2O2, with percentages shown in line 82-83.There is at least one more enzyme, and possibly more, that also produce H2O2: GlpO, involved in the glycerol metabolic process.GlpO catalyzes the following reaction: O2 + sn-glycerol 3-phosphate = dihydroxyacetone phosphate + H2O2.As the mutants show, this likely does not contribute much to the overall H2O2 production by Spn, so it has little consequence for the content for the paper, but the statement that these two enzyme are solely responsible for H2O2 production is incorrect.Line 158: Choose either using or with Line 182: Washed once and then collected?Or just a single addition of PBS?Probably the latter, but the word washed confused me.Line 191: Why was this concentration of Hb used?Does this have any physiological basis?It is used throughout the paper.Although it is possibly difficult to determine the level of Hb the pneumococcus is exposed to, it is important to understand the relevance of this amount of Hb in the in vivo situation.Line 224-226: where the nuclei counted manually or using software algorithms?Line 239: Why was the 100 µM H2O2 concentration used?Especially since you mention in line 379-380 that the TIGR4 strain used produces 700µM within 6h of incubation.Line 383-384: What was the conclusion?Line 418: 8h for TIGR4 but 6h for EF3030 Line 434-435: H2O2 itself is already a ROS.It increases other intracellular ROS? Line 439-440: Does it not just visualize an increase of intracellular H2O2?That the mutants do not produce?Have you tested the same experiment with just extracellular H2O2 in similar concentration to see whether the observed ROS is not just the increased H2O2?H2O2 alone did not affect tubulin, suggesting another actor, but a control for H2O2 would have been more correct.Line 442-444: Have you quantified this observation?Line 446-447: If H2O2 is the cause of cytotoxicity, why did the 100 µM H2O2 not affect cytoskeleton, why were no other concentrations tested to see if this is the result of just H2O2 or if other factors are involved?Should the oxidation of cell molecules also not be induced by the extracellularly added 100µM H2O2.Line 459-460: This is worth showing Line 481: Or that the amount of H2O2 produced (+/-200µM, fig.3E) was sufficient for the cytotoxicity seen, especially since the cytotoxicity dropped from ~25% in figure 1B in 6 hours to 7% in 14 hours (3A)?Perhaps this experiment is only showing that Hb-Fe3+ can reduce H2O2.Line 485: It is also possible the cells reduced the produced H2O2, although increased catalase would have been expected in the TIGR4 16h condition in fig 1G.Line 516: You show that growth is significantly affected in biofilm in 5D, and although not significant, over a 100-fold lower CFU in the HU condition is quite a lot less.Line 535: Typo of S. pneumoniae Line 543-544: Would you not expect more of a signal in the mutants than as well?There is some in the TIGR4 double mutant, but nothing in the EF30303 mutant.Line 586-587: It does not necessarily show that the DNA damage that is observed is part of the cause of the cytotoxicity.It would also be interesting to add a bacteria but no Hb-Fe3+, as well as a double mutant, as a control to see whether the damage is caused by oxidized components, H2O2 or the bacteria itself.Was the H2O2 concentration measured in the growth of R6 with Hb-Fe3+?Otherwise the DNA damage could still be a result of H2O2.Line 593: Again, what is this concentration based off? Figure 1: Showing the results of statistical comparison between two conditions in a graph seems to be chosen somewhat randomly, as in 1I for example.1C: Were CFU counts/ODs measured to determine if the bacterial load was similar?Otherwise the effect could be due to the bacterial growth being affected by the catalase.If it is cytotoxic for A549 cells in those concentrations, perhaps it is the same for the pneumococci.1D/1H: Why were two different timepoints chosen for the different strains?Additionally, there seem to be no bacteria in the TIGR4 DAPI stain?1G: Would an effect of H2O2 on tubulin not be expected if that is what is the main difference between WT and the double mutant?The expected concentration is different though than the ~700µM that was mentioned previously.The authors do not make note of this observation, or try to explain what is seen.

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.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.

Comments for the editor:
Thank you for the opportunity to review "Oxidation of hemoproteins by Streptococcus pneumoniae collapses the cell cytoskeleton and disrupts mitochondrial respiration leading to cytotoxicity of human lung cells" for the journal I am not experienced in EPR and therefore I am limited in the ability to comment on the methods and materials used for this part of the paper.

Comments for the authors:
The authors have interestingly revealed possible links between H 2 O 2 produced by Streptococcus pneumoniae and the associated cytotoxicity of it.Pneumococcal WT lab strains and mutants producing little H 2 O 2 were used to compare differences in cytotoxicity, and cytotoxicity was linked to oxidation of Hb-Fe 3+ .The topic is relevant, as it could contribute to understanding the progression of invasive disease, and resulting in new therapeutics options.There are some major concerns which will be addressed below.

Major comment
There is no data shown or mentioned regarding the differences in growth between the strains, nor is the amount of bacteria present at most measurements known to the reader (and the authors?).Since the authors mention that the both SpxB and LctO are part of a pathway that is responsible for the bulk of the ATP produced.The loss of these genes is likely to affect growth to a considerable degree, resulting in different bacterial loads between the WT and mutant conditions.Whether the results observed are majorly affected by these differences is presently impossible to determine, while it could be a very important factor.Different growth dynamics, might also result in a different ratio of live:dead bacteria at the time of measurement, which could release all sorts of molecules into the growth medium which could also affect the results.CFU/DNA copy numbers, or the addition of growth curves for the used strains (as supplementary material) would greatly improve the interpretation of the shown data.Intracellular growth might be quite different and difficult to measure, but in the data as presented, it is unclear what the dynamics of bacterial growth are.The observed difference in cytotoxicity could be simply because of differences in the bacterial load.

Minor comments
Many experiments are performed with 10 µM Hb-Fe 3+ , which is shown to oxidize and result in cytotoxicity.However, most Hb in vivo would be Hb-Fe 2+ to start with, which would first need to be oxidized to Hb-Fe 3+ (which has been observed and reported previously).Hb-Fe 3+ could then be oxidized to Hb-Fe 4+ .In reference 19, most Hb-Fe 2+ is stated to be oxidized into Hb-Fe 3+ in experimental conditions.It is unclear to me how much of the Hb-Fe 2+ would be expected to be oxidized twice.So whether this oxidation of Hb-Fe 3+ that is shown, would occur in physiological conditions.On what do the authors base the idea that Hb-Fe 3+ oxidation could be expected in invasive disease and would it not be more logical to add Hb-Fe 2+ , which, if it also occurs in in vivo conditions, could be oxidized twice and also form the radical as observed?Line 78: The authors mention Spx and LctO as the only enzymes that produce H 2 O 2 , with percentages shown in line 82-83.There is at least one more enzyme, and possibly more, that also produce H 2 O 2 : GlpO, involved in the glycerol metabolic process.GlpO catalyzes the following reaction: O2 + snglycerol 3-phosphate = dihydroxyacetone phosphate + H2O2.As the mutants show, this likely does not contribute much to the overall H 2 O 2 production by Spn, so it has little consequence for the content for the paper, but the statement that these two enzyme are solely responsible for H 2 O 2 production is incorrect.Line 158: Choose either using or with Line 182: Washed once and then collected?Or just a single addition of PBS?Probably the latter, but the word washed confused me.3E) was sufficient for the cytotoxicity seen, especially since the cytotoxicity dropped from ~25% in figure 1B in 6 hours to 7% in 14 hours (3A)?Perhaps this experiment is only showing that Hb-Fe3+ can reduce H2O2.Line 485: It is also possible the cells reduced the produced H 2 O 2 , although increased catalase would have been expected in the TIGR4 16h condition in fig 1G.Line 516: You show that growth is significantly affected in biofilm in 5D, and although not significant, over a 100-fold lower CFU in the HU condition is quite a lot less.Line 535: Typo of S. pneumoniae Line 543-544: Would you not expect more of a signal in the mutants than as well?There is some in the TIGR4 double mutant, but nothing in the EF30303 mutant.Line 586-587: It does not necessarily show that the DNA damage that is observed is part of the cause of the cytotoxicity.It would also be interesting to add a bacteria but no Hb-Fe 3+ , as well as a double mutant, as a control to see whether the damage is caused by oxidized components, H 2 O 2 or the bacteria itself.Was the H 2 O 2 concentration measured in the growth of R6 with Hb-Fe3+?Otherwise the DNA damage could still be a result of H 2 O 2 .Line 593: Again, what is this concentration based off? Figure 1: Showing the results of statistical comparison between two conditions in a graph seems to be chosen somewhat randomly, as in 1I for example.1C: Were CFU counts/ODs measured to determine if the bacterial load was similar?Otherwise the effect could be due to the bacterial growth being affected by the catalase.If it is cytotoxic for A549 cells in those concentrations, perhaps it is the same for the pneumococci.1D/1H: Why were two different timepoints chosen for the different strains?Additionally, there seem to be no bacteria in the TIGR4 DAPI stain?1G: Would an effect of H2O2 on tubulin not be expected if that is what is the main difference between WT and the double mutant?The expected concentration is different though than the ~700µM that was mentioned previously.The authors do not make note of this observation, or try to explain what is seen.

Response to Reviewers
Dear Dr. Navarro-Garcia, Thank you for the providing the opportunity for us to submit a revised version of my manuscript titled "Oxidation of hemoproteins by Streptococcus pneumoniae collapses the cell cytoskeleton and disrupts mitochondrial respiration leading to cytotoxicity of human lung cells".We have greatly appreciated the feedback and are thankful for the reviewer's valuable comments on the manuscript.Based on their suggestions, the manuscript has undergone revision and the changes within the revised submission have been highlighted.
Below are the point-by-point responses to each reviewers' comments.

Comments from Reviewer 1 1. The length of the manuscript is very long and suggested to shorten the introduction and materials and methods sections.
Response: Thank you for the suggestion, the introduction and materials and methods sections have been shortened where applicable.Due to comments from Reviewer 2, the manuscript has lengthened slightly with suggested experiments that have now been included.We hope that these additions are amenable despite the minimal increase in length.

The authors have used already published methods to validate their hypothesis and it is suggested to use the references instead of providing detailed methodology.
Response: We agree, certain methods in this section have now been referenced and certain detailed descriptions removed.

Figure numbers are missing in each page in the bottom of the manuscript. It was difficult to correlate/determine the figures in the text.
Response: We apologize for the confusion; after checking the manuscript as well as the figures alone, figure numbers can now be found in the resubmission documents.

Comments from Reviewer 2 1. (Major Comment)
There is no data shown or mentioned regarding the differences in growth between strains, nor is the amount of bacteria present at most measurements known to the reader….The observed difference in cytotoxicity could be simply because of differences in the bacterial load.
Response: Thank you so much for this response, we agree that that the addition of growth curves/CFU counts would improve interpretation of the data.We have performed these growth curves and have included them in the revised manuscript.As seen in the revised Figure 1A, growth of both TIGR4 spxBlctO and EF3030 spxBlctO is significantly higher than their WT counterparts.This signifies that despite the increase in growth of the spxBlctO strains, there is still minimal cytotoxicity indicating cytotoxicity seen in manuscript Fig. 1B and 1C is due to Spn-H 2 O 2 alone and not an effect of the strain's growth or bacterial load.As well, the measure of cytotoxicity is done through the quantification of lactate dehydrogenase (LDH) and any molecules released during bacterial growth are presumed to not have an effect on measurements.While Spn does produce a homolog of LDH, it is below the level of detection in the kit used so the only measurable LDH seen is from the human alveolar A549 cells (1).In reference to CFU/mL, those experiments for catalase have been performed and response to that can be seen in comment #24.

Many experiments are performed with 10M Hb-Fe 3+ , which is shown to oxidize and results in
cytotoxicity…On what do the authors base the idea that Hb-Fe 3+ oxidation could be expected in invasive disease and would it not be more logical to add Hb-Fe 2+ , which, if it also occurs in in vivo conditions, could be oxidized twice and also from the radical as observed.
Response: The general oxidation of hemoglobin with H 2 O 2 in blood begins with Hb-Fe 2+ to Hb-Fe 4+ and then auto reduction of Hb-Fe 4+ to Hb-Fe 3+ (2).It is not uncommon for Hb-Fe 3+ to be oxidized again to create Hb-Fe 4+ .These oxidations would also include the toxic radical intermediates between each state.We appreciate the suggestion of beginning the experiments with Hb-Fe 2+ as there would be a higher presence of Hb-Fe 2+ in blood during infection but there would still be initial oxidation of Hb-Fe 2+ to Hb-Fe 3+ and further to Hb-Fe 4+ .We choose to start these reactions at Hb-Fe 3+ , downstream of initial oxidation as we have previously observed that three is no differences in downstream oxidation when starting with Hb-Fe 3+ rather than Hb-Fe 2+ (3).As well, the process of introducing hemoglobin into experiments is an easier process with Hb-Fe 3+ than Hb-Fe 2+ .If Hb-Fe 2+ were used, oxidation to Hb-Fe 4+ would still be seen and similar radial formation be observed.

Line 78: The authors mention Spx[B] and LctO as the enzymes that produce H 2 O 2 , with percentages
shown in line 82-83… As the mutants show, this likely does not contribute much to the overall H 2 O 2 production by Spn, so it has little consequence for the content in the paper, but the statement that these two enzymes are solely responsible for H 2 O 2 production is incorrect.
Response: Thank you for pointing this out, is has been edited so SpxB and LctO aren't regarded as the sole producers of H 2 O 2 for Spn (Lines76-80).

Line 158: Choose either using or with.
Response: Have changed to "using".

Line 182: Washed once and then collected?
Or just a single addition of PBS?Probably the latter, but the word washed confused me.
Response: We apologize for the confusion.In this method of collecting bacteria for inoculum, sterile PBS is pipetted on an incubated Blood Agar Plate (BAP) containing bacterial growth.The PBS is then repeatedly pipetted up and down over the BAP to collect bacteria in a suspension of PBS (inoculum), this is what we term "washing".This inoculum is them used for subsequent experiments.For less confusion to other readers, "washing" has been changed to "collected".

Line 191: Why was this concentration of Hb used? Does this have any physiological basis? It is used throughout the paper. Although it is possibly difficult to determine the level of Hb the pneumococcus is exposed to, it is important to understand the relevance of this amount of Hb in the in vivo situation
Response: While this concentration does not match the exact physiological hemoglobin concentration, where levels can vary, hemoglobin Soret peaks at ~415 nm and alpha and beta chain peaks at ~540 and ~570 respectively have been found within our lab to be measurable at a 10M concentration during a reasonable time course study.Other reported and standardized concentrations have previously been 20M (4, 5), though full oxidation of this concentration by Spn-H 2 O 2 would be longer than the 6 h that is observed with the 10M concentration.This is the in vitro proof of concept of what would occur in vivo.

Line 224-226: [Were] the nuclei counted manually or using software algorithms?
Response: These nuclei were counted using software algorithms with the NIS-Elements Basic Research Software, version 4.30.01 build 1021.

Line 239: Why was the 100M H 2 O 2 concentration used?
Response: We agree with your point here and thank you for pointing this out.This experiment has been performed again with a similar concentration to what Spn WT strains would produce and the updated figure and results which can be found within the manuscript.A549 cells that were treated with 750 M of H 2 O 2 still show no change in tubulin signal.While it would be hypothesized that the addition of a similar level of H 2 O 2 to that of Spn-H 2 O 2 production would cause loss, it is important to note that over a 16 h time course that is presented here, Spn produces H 2 O 2 continuously.The H 2 O 2 addition at a similar level is done once at the beginning of the time course and acts as a limiting reagent to be oxidized by the cells where cells can then recover after the H 2 O 2 is no longer present.

Line 383-384: What was the conclusion?
Response: The conclusion for cytotoxicity results can be found in the following paragraph.This line here is to introduce the reason for using a spxBlctO mutant rather than a single mutant in either enzyme, which is that both single mutants are cytotoxic to other bacteria when incubated alongside Spn.

Line 418: 8h for TIGR4 but 6h for EF3030
Response: While EF3030 exhibits lower density and growth than TIGR4 as shown in the newly added growth curves, EF3030 produces higher amounts of H 2 O 2 than TIGR4 within the same time period.Due to EF3030's increased production of H 2 O 2 , and to show a similar level of cytoskeleton and tubulin loss as TIGR4, this occurs at 6h rather than 8h.This phenomenon has not yet been addressed but would be an interesting future study to perform.Response: We agree with your statement here.While the data does appear to be visually significant, due to a n=2, any statistical analysis being performed is reporting no significance.

Line 535: Typo of S. pneumoniae
Response: Thank you for pointing this out, this typo has been edited to the correct spelling of S. pneumoniae.

Line 543-544: Would you not expect more of a signal in the mutants [then] as well?
There is some in the TIGR4 double mutant, but nothing in the EF3030 mutant.
Response: We agree, this would be expected given the control.As the control in this figure is THY + Hb-Fe 3+ , the THY media (Todd-Hewitt Broth that contains 0.5% yeast extract) alone contains redox active compounds/transition metal ions able to catalyze minimal radical formation in aerobic conditions.These compounds that may catalyze this reaction with DEPMPO are the beef heart infusion, which contains undigested proteins and metal ions, that is found in THY and combination of proteins and peptides in yeast extract.The reasoning why this minimal signal would not be seen in the spxBlctO mutants is most likely due to the consumption of these compounds by Spn.Response: Thank you for this comment, we have performed this suggested experiment.This data set overall (Fig. 6F) is referring to the capacity Hb-Fe 3+ oxidation by Spn-H 2 O 2 has on DNA degradation.If there is capacity to degrade bacterial DNA, there is also capacity to degrade alveolar DNA, which is seen in Fig. 6E.Also, adding in a control of Spn+DNA would remove any questions of if DNA damage is caused by the bacteria itself.In the updated figure 6F and within the results section, there is no DNA loss when R6 Spn is inoculated with TIGR4 DNA.As there is no significant difference seen in loss of DNA when R6 Spn is introduced, we would also expect there to be no difference as well with a spxBlctO strain, so it was not included in the experiment.During the time of this experiment, H 2 O 2 was not measured but R6 is known to produce similar H 2 O 2 levels to that of TIGR4.

Line 593: Again, what is this concentration based off?
Response: This concentration of 56M is based off of quantifiable Soret and alpha and beta chain peaks pre and post oxidation by Spn-H 2 O 2 when measured on the spectrophotometer.Lower concentrations of Cytochrome c do not exhibit quantifiable/measurable peaks where we are able to see the oxidation phenomenon.
23. Figure 1: Showing the results of statistical comparison between two conditions in graph seems to be chosen somewhat randomly, as in 1I for examples.
Response: We agree, the statistical comparisons have been reformatted so that the comparisons being made are clear to the reader and aren't chosen randomly, thank you.

24
. Figure 1: 1C: Were CFU counts/ODs measured to determine if the bacterial load was similar?Otherwise the effect could be due to the bacterial growth being affected by catalase.If it is cytotoxic for A549 cells in those concentrations, perhaps it is the same for pneumococci.
Response: CFU counts/OD's were not measured at the time of this experiment, but we agree with your statement here.If catalase was cytotoxic to A549 cells, there is a possibility that it is also cytotoxic to Spn.To test this and to supplement the data in 1C, TIGR4 was cultured with 400U and 1000U catalase.Notably, there is significant increase in the growth found when TIGR4 is grown in the presence of either 400U or 1000U compared to control.This signifies that catalase is not cytotoxic to Spn and does not affect its growth or ability to cause cytotoxicity and the phenomenon observed here is due to Spn-H 2 O 2 's inability to cause cytotoxicity as it is being scavenged by catalase.A new figure incorporating this data is reflected in the revised manuscript.
25. Figure 1: 1D/1H: Why were two different timepoints chosen for the different strains?Additionally, there seems to be no bacteria in the TIGR4 DAPI strain?
Response: For the differing in timepoints between the strains, please see response under comment 10.We also apologize for the oversight in the addition of the wrong confocal image, we appreciate you catching this.This image has been replaced with the correct version with a representative image showing bacterial presence in the DAPI stain.Response: We agree with your statement here, this concern has been addressed, and the response can be found in both comment #8 and #14.Response: Thank you for noting this, the Figure 2 legend has been fixed and the edited version of the is reflected in the revised manuscript.
• 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

1 :
1G: Would an effect of H 2 O 2 on tubulin not be expected if that is what is the main difference between WT and the double mutant?The expected concentration is different though than the ~700M that was mentioned previously.The authors do not make note of this observation or try to explain what is seen.
Line 191: Why was this concentration of Hb used?Does this have any physiological basis?It is used throughout the paper.Although it is possibly difficult to determine the level of Hb the pneumococcus is exposed to, it is important to understand the relevance of this amount of Hb in the in vivo situation.Line 224-226: where the nuclei counted manually or using software algorithms?Line 239: Why was the 100 µM H 2 O 2 concentration used?Especially since you mention in line 379-380 that the TIGR4 strain used produces 700µM within 6h of incubation.Line 383-384: What was the conclusion?Line 418: 8h for TIGR4 but 6h for EF3030 Line 434-435: H 2 O 2 itself is already a ROS.It increases other intracellular ROS? Line 439-440: Does it not just visualize an increase of intracellular H 2 O 2 ?That the mutants do not produce?Have you tested the same experiment with just extracellular H2O2 in similar concentration to see whether the observed ROS is not just the increased H2O2?H2O2 alone did not affect tubulin, suggesting another actor, but a control for H2O2 would have been more correct.Line 442-444: Have you quantified this observation?Line 446-447: If H 2 O 2 is the cause of cytotoxicity, why did the 100 µM H 2 O 2 not affect cytoskeleton, why were no other concentrations tested to see if this is the result of just H 2 O 2 or if other factors are involved?Should the oxidation of cell molecules also not be induced by the extracellularly added 100µM H 2 O 2 .Line 459-460: This is worth showing Line 481: Or that the amount of H 2 O 2 produced (+/-200µM, fig.

434-435: H 2 O 2 itself is already a ROS. It increases other intracellular ROS?
Response: The increase in ROS that H 2 O 2 produced by Spn would cause is our hypothesis.This is a proof of concept that Spn is able to produce H 2 O 2 intracellularly after invasion of cells and/or that the H 2 O 2 produced by Spn extracellularly are able to diffuse into cells.

12. Line 439-440: Does it not just visualize an increase in intracellular ROS? That the mutants do not produce? Have you tested the same experiment with just extracellular H 2 O 2 in similar concentration to see whether the observed ROS is not just the increased H 2 O 2 ? H 2 O 2 alone did not affect tubulin, suggesting another actor, but a control for H 2 O 2 would have been more correct.
Response: This kit (CellRox Green Reagent, for oxidative stress detection, Invitrogen) detects all ROS present within the cell, including H 2 O 2 .We agree that this experiment could include a H 2 O 2 control and within the revised manuscript, there now includes a H 2 O 2 control with similar concentrations that Spn would produce.With a similar level of H 2 O 2 (750 µM), there is a small increase in ROS presence but no significant difference in measurable ROS.Reasoning for this mimic what most likely occurs in the Western blot experiments and explanation can be found in response to comment #8.Response: Thank you for this suggestion, we have not quantified this observation.Though it would provide supplemental data to what is currently in Fig 2C,we believe that visual representation of increase in ROS represented by an increase in green fluorescence is agreeable to its conclusion.

2 is the cause of cytotoxicity, why did the 100 µM H 2 O 2 not affect cytoskeleton, why were no other concentrations tested to see if this is the result of just H 2 O 2 or if other factors are involved? Should the oxidation of cell molecules also not be induced by the extracellularly added 100 µM H 2 O 2 .
Response: We agree with your statement here that H 2 O 2 would show an effect.We believe you are referring to the Western blot figure in Figure1.At the 100M concentration, there is no loss of tubulin cytoskeleton, but this does not reflect the amount of H 2 O 2 Spn is capable of producing.Experiments have since been performed with other H 2 O 2 concentrations and have reported the 750M concentration as this is most similar to the concentration that Spn would produce.A new figure1Hreflects new data with this concentration.Explanation for what is seen in this new figure can be seen in the manuscript as well as in response to comment 8. Due to the response, we have seen at 100M, A549 cells seem to be hardy against lower levels of H 2 O 2 added once at the beginning of experiments.As well, Spn is a continuous producer of H 2 O 2 capable of causing continuous repetitive damage to cells.

15. Line 459-460: This is worth showing Response
: Thank you for the suggestion here.The oxidation that is seen during the initial steps of the experiments exhibits similar oxidation curves as what is seen in Fig.3D, 4A, and 4C.We believe the manuscript provides ample data to support our hypothesis and therefore, we have decided to not include this data within this section of results.

Line 485: It is also possible the cells reduced the produced H 2 O 2 , although increase in catalase would have been expected in the TIGR4 16h condition in fig 1G.
Fe 3+ , as there is a decrease in measurable H 2 O 2 .This would indicate that there is reduction of H 2 O 2 by cells or targets within cells to create radicals leading to cytotoxicity.
Response: Yes, this experiment here is addressing that the Spn-H 2 O 2 is reacting with targets within the cells and/or the cells themselves in a similar reaction with what is seen with Spn-H 2 O 2 and Hb-