Flow cytometry-based isolation of Salmonella -containing phagosomes combined with ultra-sensitive proteomics reveals novel insights into host-pathogen interactions

Article activity feed

1. 

   Review Commons

   Apr 22, 2026

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   We would like to thank the reviewers for their positive and constructive feedback.

   We apologise for the delay in coming back. The first author has moved to the LMB, and the Trost lab has been relocating to the University of Manchester, which delayed our ability to respond quickly.

   * *

   * *

   Reviewer #1 (Evidence, reproducibility and clarity (Required)):

   Reviewer Comments

   The manuscript by Chatterjee et al. describes a novel ultra-sensitive isolation and deep proteomics workflow to investigate phagosome dynamics of bacterium-containing phagosomes. The method enables dual proteome coverage of both host and pathogen, and the authors report quantitative …

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   We would like to thank the reviewers for their positive and constructive feedback.

   We apologise for the delay in coming back. The first author has moved to the LMB, and the Trost lab has been relocating to the University of Manchester, which delayed our ability to respond quickly.

   * *

   * *

   Reviewer #1 (Evidence, reproducibility and clarity (Required)):

   Reviewer Comments

   The manuscript by Chatterjee et al. describes a novel ultra-sensitive isolation and deep proteomics workflow to investigate phagosome dynamics of bacterium-containing phagosomes. The method enables dual proteome coverage of both host and pathogen, and the authors report quantitative changes in the host and bacterial proteomes using Salmonella isogenic mutants defective in intracellular survival. They further leverage these datasets to assess the relevance of selected Salmonella genes in intracellular fitness.

   Overall, the manuscript presents a powerful and technically impressive approach that will be of significant interest to the infection biology community. The study is well conceived and addresses an important gap in the field. However, several clarifications and additions would strengthen the work and improve interpretability of the results.

   * *

   Specific Comments

   Line 76: The authors should consider including the following relevant citations: PMID: 30079117 and PMID: 31009521.

   We thank the reviewer for pointing this out. We have now included the suggested references

   ---

   Line 104: Please define the abbreviation BFP clearly upon first use.

   We thank the reviewer; we have defined the abbreviation upon first instance.

   * *

   Figure 1A, Step 2: From the schematic, it is unclear whether the pellet or the supernatant is used for the subsequent step in which the CellVue dye is added. Please clarify.

   We thank the reviewer for bringing this to our attention. We have now modified Figure 1A.

   * *

   Figure 1B: It would be informative to report the percentage of S. Typhimurium that are double positive, especially in the BFP + Claret condition. A small bar plot for each condition would help visualize and compare the proportion of Claret-labelled bacteria.

   We have now included a figure for the percentage of BFP + Claret for STM in S1H.

   Figure 1C: The distinction between the upper and lower images is unclear. Do they represent different particles or different fields of view of the same sample? Please clarify.

   They both are from different fields of view.

   * *

   Line 122: The statement is not entirely accurate. Cells that lyse via pyroptosis will leave behind cellular remnants, including nuclei, that may still co-sediment with intact cells in such preparations.

   We have modified the sentence accordingly.

   Line 128: CellVue and Claret appear to be used interchangeably-are they the same reagent? Please clarify and use consistent terminology throughout.

   We have rectified this inconsistency in our revised manuscript.

   *Line 136: Please explain the basis for the stated estimates. If this is common knowledge within the field, additional explanation would still be helpful for non-experts. *

   We have clarified this further in the manuscript. Obviously, these numbers are estimates but give the reader an idea with how little material we are working.

   * *

   * *

   Lines 143 & 145: Please define "protein IDs" and indicate how many correspond to host proteins versus Salmonella proteins.

   We have defined this in our revised manuscript. Also, to avoid any confusion, these proteomics methods were optimised using a commercially available HeLa protein digest, and hence no Salmonella proteins are detected here.

   * *

   *Figure 2D: Please specify the number and type of replicates used. Also indicate the plot type (e.g., violin plot) and the statistical test used to determine significance. *

   We have updated figure legend for 2D and 2E stating the number of biological replicates, i.e. n=4 and n=3.

   * *

   *Line 244: Please consider citing PMID: 32514074 and PMID: 23162002. *

   *We have included these references. *

   * *

   *Line 253: Have the authors considered how their observations regarding MHC relate to prior findings (PMID: 27832589)? *

   *Thank you for suggesting this paper and we enjoyed reading it. However, since the paper suggested by the reviewer focusses on cell surface MHC molecules and we are looking at the phagolysosomal compartment, we feel it may be difficult to make connections. *

   * *

   Line 265: Clarify which "cell" is being referred to-the host cell or the bacterial cell.

   We have modified the sentence to reduce confusion.

   * *

   *Line 278: Have the authors considered how their observations on glycolytic proteins relate to earlier work (PMID: 19380470 and PMID: 37594988)? *

   *Thank you for pointing out these papers. We have cited both of these and added another sentence that intracellular STM utilises host metabolites. *

   * *

   Line 285: The claim that "PhoP-dependent effectors actively remodel..." requires clarification. If the authors are referring to all PhoP-regulated genes as "effectors," this terminology may cause confusion, as "effectors" in the Salmonella field typically denotes T3SS-secreted proteins. While some T3SS effectors are PhoP-regulated, PhoP controls many additional genes, and the observed phenotypes may reflect broader defects in intracellular survival rather than absence of secreted effectors specifically. Rewording is recommended.

   Thank you for your suggestion, we have modified the same in text.

   * *

   Line 313: Have the authors examined later time points (e.g., 8 hpi), when the SCV is more established and SPI-2 effector expression is higher?

   We did not test the 8 hpi timepoint because our primary aim was to identify the induction of SPI-2 effectors at earlier stages. Testing later timepoints would be problematic, as PhoP mutants show poor survival at these times, which would confound comparisons between STM WT and PhoP mutants.

   Line 317: Were secreted SPI-2 effectors detectable using PhagoCyt, and if so, how did they behave?

   We detected some of the secreted effectors as well, and they are in accordance with the literature. As expected, most of them were detected only in WT at 4 hpi.

   For example, PipB2, SseL and SctB1 are significantly decreased in the PhoP mutant compared to the STM WT at 4 hpi.

   * *

   Line 319: Have the candidate Salmonella mutants been evaluated at later time points (6-8 hpi)? Stronger phenotypic differences may emerge when intracellular replication relies more heavily on SPI-2 function.

   We acknowledge that there may be larger differences at later time points; However, we wanted to be comparable with the data within the manuscript, i.e. within the 4 hour time-point that we have kept throughput. Moreover, at later timepoint we see increase macrophage cell death and therefore refrain from doing timepoints much longer after the 4 hour mark.

   Figure 5B: For all mutant strains, please also report in vitro growth to determine whether the phenotypes reflect general growth defects or are specific to the intracellular environment.

   We have performed the growth curve for the PhoP mutant, which is in the supplemental figure 1.

   * *

   Line 336: As above, please reconsider the use of the term "effectors." Unless evidence is provided that these are bona fide secreted SPI-2 effectors, an alternative term would avoid confusion.

   * *

   We have modified the sentence to reduce confusion.

   * *

   Supplementary Figure 5: The volcano plots appear pixelated. Please provide higher-resolution versions.

   Thank you for pointing this out. We have rectified this.

   * *

   Reviewer #1 (Significance (Required)):

   General assessment:

   This study introduces a highly sensitive dual host-pathogen proteomics workflow for profiling bacterium-containing phagosomes. Its key strengths are the technical innovation and the mechanistic insight gained using Salmonella mutants. The main areas needing improvement are clarification of methodological details and tighter interpretation of some biological claims.

   Advance:

   To my knowledge, this is the first study to achieve such deep, simultaneous proteomic coverage of both host and intracellular bacteria within purified phagosomes. This represents a notable technical advance and provides new mechanistic insight into intracellular adaptation and immune regulation.

   Audience:

   The work will interest a specialized audience in infection biology, host-pathogen interactions, and proteomics, with broader relevance for researchers studying organelle isolation or intracellular pathogens. The workflow and datasets will be useful as a resource for future studies.

   Reviewer expertise:

   Expertise in host-pathogen interactions, bacterial intracellular survival, macrophage biology, and functional proteomics. Limited expertise in MS instrumentation.

   * *

   * *

   Reviewer #2 (Evidence, reproducibility and clarity (Required)):

   In this work, Chatterjee, Rubio and colleagues use a novel flow cytometry-based method to isolate phagosomes from Salmonella infected macrophages. This method is applied both to wild-type and to a mutant (deletion of phoP) that does not express virulence genes, prior to the proteome characterization of these phagosomes and the bacteria that they contain. The experiments were done at an early point of infection (30 min) and a later time point (4 h). The authors first identified mitochondrial proteins in their analysis, which had previously been considered contaminants from the preparation of phagosomes. However, some Salmonella effector proteins are known to affect mitochondria, and the authors demonstrate that inhibition of Complex I showed decreased Salmonella intracellular viability. Comparing WT and the phoP mutant also highlighted two Salmonella proteins that enhance intracellular survival. In addition, the authors show that their method recapitulates previously known proteins involved in Salmonella infection. The study is well designed and clearly written.

   I have only some minor comments that I hope will strengthen the work:

   * It would be interesting to compare the results with a whole cell proteome analysis, and to other approaches that involve subcellular fractionation (both in the context of Salmonella infection) to: a) highlight proteins that are specifically changing in abundance in the phagosomes (but not necessarily in the cell), and b) to show that this approach is able to capture previously unknown phenomena. To avoid the performing additional experiments, the authors can compare their dataset to previous proteomic datasets of Salmonella infection.* We have compared this with the ultracentrifugation methods STM WT 4h vs STM WT uptake (Figure 6A).

   * A color scale for the heatmap in Fig 2C is needed. I assume that this heatmap shows intensity and not fold-changes, and thus suggest that the authors use a single-color gradient for easier visualization. *

   *This has now been included. *

   * *

   Best regards,

   André Mateus

   * *

   * *

   Reviewer #2 (Significance (Required)):

   General assessment: This study provides a novel approach to study intracellular pathogenic bacteria. The method is applied to Salmonella, but can potentially be used for any bacteria, including non-genetically tractable organisms. A strength of the approach is that it captures the bacterial proteome, which is mostly undetectable when studying infected cells. Further, by enriching phagosomes, it allows measuring the spatial distribution of proteins to these organelles. The study could be improved by distinguishing proteome changes that are caused by trafficking of proteins to phagosomes vs general changes in protein abundance.

   *Advance: Apart from a new methodology, the authors use the approach to identify novel aspects of Salmonella infection biology, e.g., the importance of mitochondrial proteins in host defense or novel Salmonella proteins that are involved in intracellular survival. Audience: The audience for this study is mostly those in the field of infection biology, particularly Salmonella. The dataset generated can be used to identify novel aspects of Salmonella infection, and the described method could be applied to other pathogens. *

   My field of expertise: Proteomics, microbiology.

   * *

   * *

   * *

   * *

   Reviewer #3 (Evidence, reproducibility and clarity (Required)):

   In the manuscript "Flow cytometry-based isolation of Salmonella-containing phagosomes combined with ultra-sensitive proteomics reveals novel insights into host-pathogen interactions", the authors describe a new method for analysis of composition pathogen-containing phagosomes and the pathogens within. Combination of FACS-based single phagosome analysis and sorting combined with optimised highly sensitive proteomic analysis of sorted vesicles has potential for identification of so far overlooked host-pathogen interactions. Although this is well described in the manuscript, some controls are missing.

   Major comments:

   1) The sorting of labelled bacteria is a crucial bottleneck in the whole procedure. The gating strategy presented in the Fig. 1B suggest that the initial "bacterial phagosome size" is limited from the bottom based on the noise signal but not from top. Therefore any not broken THP-1 cell remaining in the sample would be also included in the analysis. In respect to very high sensitivity of the mass spectrometry procedure and high abundance of housekeeping genes in host cells, this contamination could well explain the appearance of mitochondria, ribosome, and nuclear envelope proteins identified in Fig 2B and undermine the following results. Therefore, the gating strategy should be more stringent and data from this more stringent gating shall be compared with the current data sets. Since the authors use BFP+ Salmonella and do not analyse the claret+BFP- events, a BFP vs FSC gating step could help to distinguish free bacteria, bacteria in vesicles, and not or only partially broken host cells.

   We use a series of centrifugations to ensure that we do not have intact cells in the prepared samples. We have also visualised the final samples under the microscope and did not observe any intact cells. Because of the side/forward scatter gating, intact cells are not within the field of sorting. In Figure 1B we show that free bacteria are not within the gating strategy that we used. Finally, we visually inspected >100 pictures of sorted phagosomes by imaging flow cytometry and did not see any intact cells or free bacteria.

   2) Since the authors present data previously well accepted as contaminations from other fractions, these shall be carefully validated by other methods. For example the contact of mitochondria with SCV could be validated using a FRET- or split FP- based assays. Change of abundance of surface proteins on SCV in individual timepoints shall be validated using antibody-based flow cytometry on isolated SCVs. Most relevant antibodies are already described in the manuscript or available commercially (IL4R, IFNgR, integrins, TLRs). Microscopy-based quantification could help with the soluble proteins present within SCVs.

   We agree with the reviewer that this would be very interesting. However, we feel that this is outside of the scope of this paper and will be very laborious and time consuming, practically a whole project in itself.

   * *

   3) Since the authors describe an alternative method to methods used previously, they shall discuss the differences in results obtained by the formerly used methods.

   We have now provided a dataset that is with SCVs isolated using ultracentrifugation as a comparatively analysis to our method (__Figure S6A and Table S8). __The data show that the ultracentrifugation-isolated phagosomes have many more proteins from any organelle (Figure S6B), suggesting that they are less pure than the phagosomes isolated by the PhagoCyt approach.

   4) Only 15 Salmonella proteins downregulated between 0.5 and 4 h timepoints were identified. However, at least genes from SPI-1 and flagella would be expected to be downregulated at 4 h p.i. How do the authors explain this discrepancy? In contrast, are the SPI-2 genes among those identified as upregulated?

   In our supplementary table 6 (comparison between WT 4h vs WT uptake), we see that there are 458 Salmonella proteins that are only present in uptake samples, these were not included in limma analysis since they are completely absent in the WT 4h. We decided to report these as “unique” proteins rather than perform imputation. In Figure 5B, we specifically highlight STM proteins down-regulated, which include flagellar proteins and SPI-1 proteins.

   To answer your second question, yes, several SPI-2 genes (effectors and other regulatory proteins) are upregulated at 4 hpi. 131 Salmonella proteins are significantly upregulated, and 55 proteins are exclusively present in the WT 4hpi samples. Some selected examples are in Figure 5A.

   * *

   Minor comments:

   1) Fig 1, the figure caption seems to remain parts of an older version, mentioning blue bars not present in the current version?

   The figure caption appears to be correct for us; the “blue” is in the unstained BFP Salmonella, which is hidden behind the purple, which is the BFP Salmonella + CellVue Claret.

   2) Fig 1A point 1, how were the dead cells removed? Normal centrifugation is not able to discriminate dead and living cells well enough as percoll gradient centrifugation for example would be. Such gradient centrifugation is not mentioned in the Methods section though.

   We have not used Percoll-based centrifugation to remove dead cells; instead, we have washed the adherent macrophages in dishes 3-4 times with ice-cold PBS to remove dead, floating cells, and then washed the pellet several times with PBS to ensure we are not taking any dead cells into the sample preparation.

   * *

   3) Fig 1A point 2, did the authors check for the composition of the pellet fraction in each centrifugation step? What are the losses and cross contaminations of the other fraction?

   No, we have not checked the composition of each fraction using mass spec; however, we did run some western blots to correctly identify the major organelle contribution in each fraction.

   4) Suppl. Fig 1, caption for panels F and G are missing. The axis in the panel G is misleading - the bacteria obtained in "output" contain proliferating intracellular bacteria that originate only from a fraction of the "input" bacteria. Since the figure clearly show increase in the number of intracellular bacteria and all the extracellular bacteria should be killed by gentamicin, all bacteria in the "output" probably proliferate intracellularly and, therefore, originate from the same fraction of the "input" throughout the whole assay. Showing these results as CFU per well/plate/surface area or cell count would be more exact, in this case the "input" data shall be shown as a separate data point.

   We thank the reviewer for this observation. We have now modified the figure legends. These are normalised per cell, and we think they provide accurate results.

   * *

   5) Fig 1B, could the authors show the percentages in individual quadrants for the green "Sample with BFP Salmonella + claret"?

   Yes, there is the plot that depicts the percentage in Supplementary Figure 1H, this varies between WT and PhoP mutant, and hence, we decided to not show this in one figure.

   ---

   6) All proteins identified as significantly up or down represented shall be listed in a supplementary file.

   They are listed in the supplemental tables.

   * *

   7) Fig 2C suggests that some mitochondrial proteins are similarly present at the SCV containing WT Salmonella at 4h as ∆phoP mutant at 0.5 h p.i. Could the authors speculate how is that? The scale of blue/orange transition shall be shown in Fig 2C.

   We speculate that Salmonella WT alters the maturation of the SCVs is heavily arrested by the pathogen and hence resemble the early SCV of a mutant that is unable to arrest the SCV degradation stages.

   * *

   8) In the Fig 2D, the authors show decrease of CFU obtained from THP-1 cells treated with Rotenone. However, rotenone is known to induce host cell apoptosis. Were the presented data normalized to amount of living host cells in the sample? For example measurement of protein concentration in the sample lysate after washing away the dying host cells should enable this.

   Yes, we have normalised the data to the account for the percentage of live cells using live dead staining. However, in the timepoints used, we did not observe significant cell death.

   9) Microscopy-based observation of mitochondria relocation to SCVs in time shall strengthen the claim that mitochondria-derived ROS are involved in anti-Salmonella host defense.

   There are multiple literature PMID: 38356294, PMID: 41444067, PMID: 15866946, PMID: 41198672 that support our data in this regard.

   10) The Salmonella proteins identified in the Fig 5 shall be validated using qPCR.

   We think that data from qPCR would not be accurate to validate Salmonella proteins, as it has been shown that Salmonella mRNAs can have sub-minute half-lives (PMID: 38527194). We used rather conservative proteomics analysis settings, that have shown in a recent pre-print of our lab to have 0% false discoveries and 0.4% false quantitative rate ( https://doi.org/10.1101/2025.09.22.677725). We acknowledge that another reviewer did not find this experiment to be essential.

   * *

   Reviewer #3 (Significance (Required)):

   The manuscript was reviewed mainly from the Salmonella and flow cytometry/FACS expertise point of view. The main interest in the study lies within its methodological advances - combination of single vesicle analysis using flow cytometry/FACS with highly sensitive mass spectrometry analysis. In comparison to other similar studies in the field, this combination significantly expands the possibilities of sorting of distinct subpopulations of vesicles from the same cells. This will make the article of interest to scientists in the broad field of host-pathogen interactions and immunology.

   * *

   **Referee cross-commenting**

   Reviewer 3 - @Reviewer #1: I see your point and leave it at the editors to judge how important this comment is. My reasoning was this: Fig 5

   serves as a proof of concept that PhagoCyt has the power to make new discoveries in Salmonella biology. While behavior of some of the proteins

   shown if Fig 5 is well described (e.g. flagella or SPI-1 T3SS components and effectors), some are novel and to prove the functionality of the

   method, these results should be confirmed by some other well accepted mean. Given the great sensitivity of PhagoCyt, other proteomic

   approaches are unlikely to help in this case (e.g. flagella or SPI-1 T3SS components and effectors are not detectable by western blot at 4 h p.i.).

   Therefore, I suggest qPCR (but would accept any other method as well) as a very sensitive and well accepted approach, but leave at the authors

   to chose what proteins they want to use for the validation.

   Reviewer 1- I agree with comments raised by the other two reviewers, except the following point from Reviewer 3 '10) The Salmonella proteins

   identified in the Fig 5 shall be validated using qPCR.' It is not clear which proteins are being referred to and it is unclear to this reviewer how this

   experiment(s) would improve the manuscript in its current form.

   Reviewer 3- I agree with all comments raised.

   * *

   * *

   Read the original source
2. 

   Review Commons

   Apr 22, 2026

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #3

   Evidence, reproducibility and clarity

   In the manuscript "Flow cytometry-based isolation of Salmonella-containing phagosomes combined with ultra-sensitive proteomics reveals novel insights into host-pathogen interactions", the authors describe a new method for analysis of composition pathogen-containing phagosomes and the pathogens within. Combination of FACS-based single phagosome analysis and sorting combined with optimised highly sensitive proteomic analysis of sorted vesicles has potential for identification of so far overlooked host-pathogen interactions. Although this is well described in the manuscript, some controls are missing.

   Major comments:

   1. The sorting of …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #3

   Evidence, reproducibility and clarity

   In the manuscript "Flow cytometry-based isolation of Salmonella-containing phagosomes combined with ultra-sensitive proteomics reveals novel insights into host-pathogen interactions", the authors describe a new method for analysis of composition pathogen-containing phagosomes and the pathogens within. Combination of FACS-based single phagosome analysis and sorting combined with optimised highly sensitive proteomic analysis of sorted vesicles has potential for identification of so far overlooked host-pathogen interactions. Although this is well described in the manuscript, some controls are missing.

   Major comments:

   1. The sorting of labelled bacteria is a crucial bottleneck in the whole procedure. The gating strategy presented in the Fig. 1B suggest that the initial "bacterial phagosome size" is limited from the bottom based on the noise signal but not from top. Therefore any not broken THP-1 cell remaining in the sample would be also included in the analysis. In respect to very high sensitivity of the mass spectrometry procedure and high abundance of housekeeping genes in host cells, this contamination could well explain the appearance of mitochondria, ribosome, and nuclear envelope proteins identified in Fig 2B and undermine the following results. Therefore, the gating strategy should be more stringent and data from this more stringent gating shall be compared with the current data sets. Since the authors use BFP+ Salmonella and do not analyse the claret+BFP- events, a BFP vs FSC gating step could help to distinguish free bacteria, bacteria in vesicles, and not or only partially broken host cells.
   2. Since the authors present data previously well accepted as contaminations from other fractions, these shall be carefully validated by other methods. For example the contact of mitochondria with SCV could be validated using a FRET- or split FP- based assays. Change of abundance of surface proteins on SCV in individual timepoints shall be validated using antibody-based flow cytometry on isolated SCVs. Most relevant antibodies are already described in the manuscript or available commercially (IL4R, IFNgR, integrins, TLRs). Microscopy-based quantification could help with the soluble proteins present within SCVs.
   3. Since the authors describe an alternative method to methods used previously, they shall discuss the differences in results obtained by the formerly used methods.
   4. Only 15 Salmonella proteins downregulated between 0.5 and 4 h timepoints were identified. However, at least genes from SPI-1 and flagella would be expected to be downregulated at 4 h p.i. How do the authors explain this discrepancy? In contrast, are the SPI-2 genes among those identified as upregulated?

   Minor comments:

   1. Fig 1, the figure caption seems to remain parts of an older version, mentioning blue bars not present in the current version?
   2. Fig 1A point 1, how were the dead cells removed? Normal centrifugation is not able to discriminate dead and living cells well enough as percoll gradient centrifugation for example would be. Such gradient centrifugation is not mentioned in the Methods section though.
   3. Fig 1A point 2, did the authors check for the composition of the pellet fraction in each centrifugation step? What are the losses and cross contaminations of the other fraction?
   4. Suppl. Fig 1, caption for panels F and G are missing. The axis in the panel G is misleading - the bacteria obtained in "output" contain proliferating intracellular bacteria that originate only from a fraction of the "input" bacteria. Since the figure clearly show increase in the number of intracellular bacteria and all the extracellular bacteria should be killed by gentamicin, all bacteria in the "output" probably proliferate intracellularly and, therefore, originate from the same fraction of the "input" throughout the whole assay. Showing these results as CFU per well/plate/surface area or cell count would be more exact, in this case the "input" data shall be shown as a separate data point.
   5. Fig 1B, could the authors show the percentages in individual quadrants for the green "Sample with BFP Salmonella + claret"?
   6. All proteins identified as significantly up or down represented shall be listed in a supplementary file.
   7. Fig 2C suggests that some mitochondrial proteins are similarly present at the SCV containing WT Salmonella at 4h as ∆phoP mutant at 0.5 h p.i. Could the authors speculate how is that? The scale of blue/orange transition shall be shown in Fig 2C.
   8. In the Fig 2D, the authors show decrease of CFU obtained from THP-1 cells treated with Rotenone. However, rotenone is known to induce host cell apoptosis. Were the presented data normalized to amount of living host cells in the sample? For example measurement of protein concentration in the sample lysate after washing away the dying host cells should enable this.
   9. Microscopy-based observation of mitochondria relocation to SCVs in time shall strengthen the claim that mitochondria-derived ROS are involved in anti-Salmonella host defense.
   10. The Salmonella proteins identified in the Fig 5 shall be validated using qPCR.

   Referee cross-commenting

   Reviewer 3 - @Reviewer #1: I see your point and leave it at the editors to judge how important this comment is. My reasoning was this: Fig 5 serves as a proof of concept that PhagoCyt has the power to make new discoveries in Salmonella biology. While behavior of some of the proteins shown if Fig 5 is well described (e.g. flagella or SPI-1 T3SS components and effectors), some are novel and to prove the functionality of the method, these results should be confirmed by some other well accepted mean. Given the great sensitivity of PhagoCyt, other proteomic approaches are unlikely to help in this case (e.g. flagella or SPI-1 T3SS components and effectors are not detectable by western blot at 4 h p.i.). Therefore, I suggest qPCR (but would accept any other method as well) as a very sensitive and well accepted approach, but leave at the authors to chose what proteins they want to use for the validation.

   Reviewer 1- I agree with comments raised by the other two reviewers, except the following point from Reviewer 3 '10) The Salmonella proteins identified in the Fig 5 shall be validated using qPCR.' It is not clear which proteins are being referred to and it is unclear to this reviewer how this experiment(s) would improve the manuscript in its current form.

   Reviewer 3- I agree with all comments raised.

   Reviewer 2- I agree with the other reviewer's comments/suggestions.

   Significance

   The manuscript was reviewed mainly from the Salmonella and flow cytometry/FACS expertise point of view. The main interest in the study lies within its methodological advances - combination of single vesicle analysis using flow cytometry/FACS with highly sensitive mass spectrometry analysis. In comparison to other similar studies in the field, this combination significantly expands the possibilities of sorting of distinct subpopulations of vesicles from the same cells. This will make the article of interest to scientists in the broad field of host-pathogen interactions and immunology.

   Read the original source
3. 

   Review Commons

   Apr 22, 2026

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   In this work, Chatterjee, Rubio and colleagues use a novel flow cytometry-based method to isolate phagosomes from Salmonella infected macrophages. This method is applied both to wild-type and to a mutant (deletion of phoP) that does not express virulence genes, prior to the proteome characterization of these phagosomes and the bacteria that they contain. The experiments were done at an early point of infection (30 min) and a later time point (4 h). The authors first identified mitochondrial proteins in their analysis, which had previously been considered contaminants from the preparation of phagosomes. However, some Salmonella …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   In this work, Chatterjee, Rubio and colleagues use a novel flow cytometry-based method to isolate phagosomes from Salmonella infected macrophages. This method is applied both to wild-type and to a mutant (deletion of phoP) that does not express virulence genes, prior to the proteome characterization of these phagosomes and the bacteria that they contain. The experiments were done at an early point of infection (30 min) and a later time point (4 h). The authors first identified mitochondrial proteins in their analysis, which had previously been considered contaminants from the preparation of phagosomes. However, some Salmonella effector proteins are known to affect mitochondria, and the authors demonstrate that inhibition of Complex I showed decreased Salmonella intracellular viability. Comparing WT and the phoP mutant also highlighted two Salmonella proteins that enhance intracellular survival. In addition, the authors show that their method recapitulates previously known proteins involved in Salmonella infection. The study is well designed and clearly written.

   I have only some minor comments that I hope will strengthen the work:

   1. It would be interesting to compare the results with a whole cell proteome analysis, and to other approaches that involve subcellular fractionation (both in the context of Salmonella infection) to: a) highlight proteins that are specifically changing in abundance in the phagosomes (but not necessarily in the cell), and b) to show that this approach is able to capture previously unknown phenomena. To avoid the performing additional experiments, the authors can compare their dataset to previous proteomic datasets of Salmonella infection.
   2. A color scale for the heatmap in Fig 2C is needed. I assume that this heatmap shows intensity and not fold-changes, and thus suggest that the authors use a single-color gradient for easier visualization.

   Best regards, André Mateus

   Referee cross-commenting

   Reviewer 3 - @Reviewer #1: I see your point and leave it at the editors to judge how important this comment is. My reasoning was this: Fig 5 serves as a proof of concept that PhagoCyt has the power to make new discoveries in Salmonella biology. While behavior of some of the proteins shown if Fig 5 is well described (e.g. flagella or SPI-1 T3SS components and effectors), some are novel and to prove the functionality of the method, these results should be confirmed by some other well accepted mean. Given the great sensitivity of PhagoCyt, other proteomic approaches are unlikely to help in this case (e.g. flagella or SPI-1 T3SS components and effectors are not detectable by western blot at 4 h p.i.). Therefore, I suggest qPCR (but would accept any other method as well) as a very sensitive and well accepted approach, but leave at the authors to chose what proteins they want to use for the validation.

   Reviewer 1- I agree with comments raised by the other two reviewers, except the following point from Reviewer 3 '10) The Salmonella proteins identified in the Fig 5 shall be validated using qPCR.' It is not clear which proteins are being referred to and it is unclear to this reviewer how this experiment(s) would improve the manuscript in its current form.

   Reviewer 3- I agree with all comments raised.

   Reviewer 2- I agree with the other reviewer's comments/suggestions.

   Significance

   General assessment: This study provides a novel approach to study intracellular pathogenic bacteria. The method is applied to Salmonella, but can potentially be used for any bacteria, including non-genetically tractable organisms. A strength of the approach is that it captures the bacterial proteome, which is mostly undetectable when studying infected cells. Further, by enriching phagosomes, it allows measuring the spatial distribution of proteins to these organelles. The study could be improved by distinguishing proteome changes that are caused by trafficking of proteins to phagosomes vs general changes in protein abundance.

   Advance: Apart from a new methodology, the authors use the approach to identify novel aspects of Salmonella infection biology, e.g., the importance of mitochondrial proteins in host defense or novel Salmonella proteins that are involved in intracellular survival.

   Audience: The audience for this study is mostly those in the field of infection biology, particularly Salmonella. The dataset generated can be used to identify novel aspects of Salmonella infection, and the described method could be applied to other pathogens.

   My field of expertise: Proteomics, microbiology.

   Read the original source
4. 

   Review Commons

   Apr 22, 2026

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   The manuscript by Chatterjee et al. describes a novel ultra-sensitive isolation and deep proteomics workflow to investigate phagosome dynamics of bacterium-containing phagosomes. The method enables dual proteome coverage of both host and pathogen, and the authors report quantitative changes in the host and bacterial proteomes using Salmonella isogenic mutants defective in intracellular survival. They further leverage these datasets to assess the relevance of selected Salmonella genes in intracellular fitness. Overall, the manuscript presents a powerful and technically impressive approach that will be of significant interest to the …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   The manuscript by Chatterjee et al. describes a novel ultra-sensitive isolation and deep proteomics workflow to investigate phagosome dynamics of bacterium-containing phagosomes. The method enables dual proteome coverage of both host and pathogen, and the authors report quantitative changes in the host and bacterial proteomes using Salmonella isogenic mutants defective in intracellular survival. They further leverage these datasets to assess the relevance of selected Salmonella genes in intracellular fitness. Overall, the manuscript presents a powerful and technically impressive approach that will be of significant interest to the infection biology community. The study is well conceived and addresses an important gap in the field. However, several clarifications and additions would strengthen the work and improve interpretability of the results.

   Specific Comments

   Line 76: The authors should consider including the following relevant citations: PMID: 30079117 and PMID: 31009521. Line 104: Please define the abbreviation BFP clearly upon first use. Figure 1A, Step 2: From the schematic, it is unclear whether the pellet or the supernatant is used for the subsequent step in which the CellVue dye is added. Please clarify. Figure 1B: It would be informative to report the percentage of S. Typhimurium that are double positive, especially in the BFP + Claret condition. A small bar plot for each condition would help visualize and compare the proportion of Claret-labelled bacteria. Figure 1C: The distinction between the upper and lower images is unclear. Do they represent different particles or different fields of view of the same sample? Please clarify. Line 122: The statement is not entirely accurate. Cells that lyse via pyroptosis will leave behind cellular remnants, including nuclei, that may still co-sediment with intact cells in such preparations. Line 128: CellVue and Claret appear to be used interchangeably-are they the same reagent? Please clarify and use consistent terminology throughout. Line 136: Please explain the basis for the stated estimates. If this is common knowledge within the field, additional explanation would still be helpful for non-experts. Lines 143 & 145: Please define "protein IDs" and indicate how many correspond to host proteins versus Salmonella proteins. Figure 2D: Please specify the number and type of replicates used. Also indicate the plot type (e.g., violin plot) and the statistical test used to determine significance. Line 244: Please consider citing PMID: 32514074 and PMID: 23162002. Line 253: Have the authors considered how their observations regarding MHC relate to prior findings (PMID: 27832589)? Line 265: Clarify which "cell" is being referred to-the host cell or the bacterial cell. Line 278: Have the authors considered how their observations on glycolytic proteins relate to earlier work (PMID: 19380470 and PMID: 37594988)? Line 285: The claim that "PhoP-dependent effectors actively remodel..." requires clarification. If the authors are referring to all PhoP-regulated genes as "effectors," this terminology may cause confusion, as "effectors" in the Salmonella field typically denotes T3SS-secreted proteins. While some T3SS effectors are PhoP-regulated, PhoP controls many additional genes, and the observed phenotypes may reflect broader defects in intracellular survival rather than absence of secreted effectors specifically. Rewording is recommended. Line 313: Have the authors examined later time points (e.g., 8 hpi), when the SCV is more established and SPI-2 effector expression is higher? Line 317: Were secreted SPI-2 effectors detectable using PhagoCyt, and if so, how did they behave? Line 319: Have the candidate Salmonella mutants been evaluated at later time points (6-8 hpi)? Stronger phenotypic differences may emerge when intracellular replication relies more heavily on SPI-2 function. Figure 5B: For all mutant strains, please also report in vitro growth to determine whether the phenotypes reflect general growth defects or are specific to the intracellular environment. Line 336: As above, please reconsider the use of the term "effectors." Unless evidence is provided that these are bona fide secreted SPI-2 effectors, an alternative term would avoid confusion. Supplementary Figure 5: The volcano plots appear pixelated. Please provide higher-resolution versions.

   Referee cross-commenting

   Reviewer 3 - @Reviewer #1: I see your point and leave it at the editors to judge how important this comment is. My reasoning was this: Fig 5 serves as a proof of concept that PhagoCyt has the power to make new discoveries in Salmonella biology. While behavior of some of the proteins shown if Fig 5 is well described (e.g. flagella or SPI-1 T3SS components and effectors), some are novel and to prove the functionality of the method, these results should be confirmed by some other well accepted mean. Given the great sensitivity of PhagoCyt, other proteomic approaches are unlikely to help in this case (e.g. flagella or SPI-1 T3SS components and effectors are not detectable by western blot at 4 h p.i.). Therefore, I suggest qPCR (but would accept any other method as well) as a very sensitive and well accepted approach, but leave at the authors to chose what proteins they want to use for the validation.

   Reviewer 1- I agree with comments raised by the other two reviewers, except the following point from Reviewer 3 '10) The Salmonella proteins identified in the Fig 5 shall be validated using qPCR.' It is not clear which proteins are being referred to and it is unclear to this reviewer how this experiment(s) would improve the manuscript in its current form.

   Reviewer 3- I agree with all comments raised.

   Reviewer 2- I agree with the other reviewer's comments/suggestions.

   Significance

   General assessment:

   This study introduces a highly sensitive dual host-pathogen proteomics workflow for profiling bacterium-containing phagosomes. Its key strengths are the technical innovation and the mechanistic insight gained using Salmonella mutants. The main areas needing improvement are clarification of methodological details and tighter interpretation of some biological claims.

   Advance:

   To my knowledge, this is the first study to achieve such deep, simultaneous proteomic coverage of both host and intracellular bacteria within purified phagosomes. This represents a notable technical advance and provides new mechanistic insight into intracellular adaptation and immune regulation.

   Audience:

   The work will interest a specialized audience in infection biology, host-pathogen interactions, and proteomics, with broader relevance for researchers studying organelle isolation or intracellular pathogens. The workflow and datasets will be useful as a resource for future studies.

   Reviewer expertise:

   Expertise in host-pathogen interactions, bacterial intracellular survival, macrophage biology, and functional proteomics. Limited expertise in MS instrumentation.

   Read the original source
5. 

   Version published to 10.1101/2025.04.25.650444 on bioRxiv

   Apr 25, 2025