Salmonella Promotes Its Own Survival in B Cells by Inhibiting Autophagy

Round 1

Reviewer 1 Report

The authors suggest that Salmonella promote their survival also in B cells due to the inhibition of autophagy, a mechanism that theseintracellular bacteria also utilize in other cell populations. While the experiments appear properly organized, the relevance of this effect in vivo remains unclear.

Specific points:

How many cells are indeed B cells following purification? 

How many B cells are indeed infected? The authors should indicate the percentages of infected B cells and provide respective microscopic pictures. How viable are the infected B cells. 

Can the authors  separate the signaling in infected and uninfected B cells.

Are Salmonella-specific IgG and IgM responses similar in WT and CD19 Cre x raptor mice? Why did the authors not use littermate raptor mice?   

Author Response

The authors suggest that Salmonella promote their survival also in B cells due to the inhibition of autophagy, a mechanism that these intracellular bacteria also utilize in other cell populations. While the experiments appear properly organized, the relevance of this effect in vivo remains unclear.

• A) We agree with the reviewer that Figure 4 shows results supporting the in vitro data, but they do not address to which extent the phenomenon contributes to infection. Further experiments will need to be conducted to evaluate this vital part.

 

How many cells are indeed B cells following purification?

• A) A purity of 98% was obtained according to flow cytometry analysis using CD19 molecule as a B cell marker. Supplementary figure S1A shows a representative dot plot and data from 5 purifications).
•  

How many B cells are indeed infected? The authors should indicate the percentages of infected B cells and provide respective microscopic pictures. How viable are the infected B cells?

• A) The percentage of B cell infection is around 15% (Supplementary Figure S1B). Thanks to the reviewer for requesting a microscopic picture, but we published it in a previous paper [27] (Rosales-Reyes, R et al. Microbial Pathogenesis 52 (2012) 367e374). Because Salmonella does not induce pyroptosis in B cells, viability of 98% is obtained after purification and remains viable during the experimental procedure [14].

 

Can the authors separate the signaling in infected and uninfected B cells?

• A) Thanks to the reviewer for this question. In the current manuscript, we did not separate infected and non-infected cells; we sorted infected and non-infected B cells in our previous published paper. We found a downregulation of NLRC4 gene in both infected and non-infected cells [13] (Perez-Lopez, A. et al. J Immunol 190: 9058-9064, 2013), and the fact that Salmonella effector SopB is responsible for the reduction of the transcription of this gene. Further, we demonstrated that SopB drives the activation of the PI3K-AKT pathway to downregulate NLRC4 and IL-1β Remarkably, we observed an accumulation of PIP3 in the cell membrane in infected and non-infected B cells [14] (Garcia-Gil, A. et al. Virulence 9: 1390-1402, 2018). Thus, we also expected an activation of AKT-mTORC1 in infected as in non-infected B cells in the present results. For this reason, we consider that it is possible to detect the phosphorylation of S6 and ULK1 in the lysate of bulk B cells.

 

Are Salmonella-specific IgG and IgM responses similar in WT and CD19 Cre x raptor mice?

• A) Thanks to the reviewer for this question. We have not evaluated the antibody production in infected mice to date. But we agree that it would be a reasonable control for the function of B cells in conditional knockout mice.

 

Why did the authors not use littermate raptor mice?  

• A) We agree with the reviewer that it would be good to also include the littermate raptor mice as a control. But it was less complex to breed the CD19 heterozygous-raptor knockout. Therefore, we did not generate a foxp-raptor-foxp-CD19 wild type.

 

Reviewer 2 Report

Salmonella is a gram negative bacterium causes gastrointestinal and systemic infections in several hosts. Salmonella has a wide variety of tropism, can infect epithelial cells, macrophages, dendritic cells, B cells and several other cell types. Unlike in other cell types salmonella infection in B cells do not induce pyroptosis, so that B cells survive and salmonella takes advantage of this phenomenon. The mechanism by which Salmonella evades B cell immune responses is not well understood. Luis et al have identified salmonella effector molecule SopB activates mTORC1 and it phosphorylates ULK1, which inhibits autophagy and resulting in successful infection in B cells. But when the mTORC1 is inhibited through pharmacological inhibitor or mutated version of SopB leads to the control of salmonella infection in B cells. This is also followed up with a B cell specific mTORC1 signaling deficient mice.

 

I find this is a well-designed and executed study. Results are very convincing. But there are some concerns that need to be addressed.

Comments:

 

1. Authors mentioned about the Perez-Lopez et al 2013 findings in the abstract but did not check if inhibition or mTORC1 or PI3k signaling had any effect on Il1b secretion. In turn that might have effected Salmonella survival in B cells. I think this needs to be clarified to establish that salmonella survival in B cells is dependent on mTORC1 mediated inhibition of autophagy.  
2. Do authors have assessed the myeloid and lymphoid cellularity in spleen of Raptor^fl/flXCD19Cre negative and positive mice. Comment if that could affect the survival of salmonella in B cells.
3. In figure 4 B authors mentioned about promoting mice survival is that a speculation or authors have evaluated the survival status of mice during salmonella infection in Raptor^fl/flXCD19Cre negative and positive mice. If so you might want to include the survival curve in figure 4.
4. Typos need to be taken care of throughout the manuscript 

 

 

 

                                       

Author Response

Authors mentioned about the Perez-Lopez et al 2013 findings in the abstract but did not check if inhibition or mTORC1 or PI3k signaling had any effect on Il1b secretion. In turn that might have affected Salmonella survival in B cells. I think this needs to be clarified to establish that salmonella survival in B cells is dependent on mTORC1 mediated inhibition of autophagy. 

• A) Thanks to the reviewer for this recommendation. Thus, we revised the abstract (lines 30-31), introduction( Lines: 58-61), and discussion ((lines 371-372) as the reviewer suggested.

 

Do authors have assessed the myeloid and lymphoid cellularity in spleen of Raptorfl/flXCD19Cre negative and positive mice. Comment if that could affect the survival of Salmonella in B cells.

• A) We used the expression of CD11b to evaluate the frequency of myeloid cells and CD19 for B cells. For the former, the spleen of cd19^+/creraptor^fl/fl mice have a similar percentage as wild-type mice. Because knockout mice are heterozygous for the CD19 gene, the percentage of B cells is around half of the wild type (Supplementary Figure S3B and S3C). Taken into consideration, the frequency of CD11b marker is similar in both mice strains; we expected that these cells have a similar role during Salmonella Also, the ratio of the bacteria-B cells might be higher in knockout mice. Although B cells of these mice control Salmonella infection (figure 4), supporting the role of autophagy in eliminating the bacteria.

 

In figure 4 B authors mentioned about promoting mice survival is that a speculation or authors have evaluated the survival status of mice during salmonella infection in Raptorfl/flXCD19Cre negative and positive mice. If so you might want to include the survival curve in figure 4.

• A) We apologize to the reviewer for our error; we did remove the mice survival in the writing.

Typos need to be taken care of throughout

• A) Thanks to the reviewer for pointing out flaws.

 

Reviewer 3 Report

This manuscript tried to investigate the effect of Salmonella SopB on survival in B-cell. This effect is mediated by inhibitory action of SopB on mTORC1. Authors investigated effect of autophagy on this mechanism.  However, I have some concern as follows in this manuscript. This manuscript has not included critical information in some methods (statistical analysis) and results.

 

1. Methods.

Data and statistical analysis is missing.

 

2. Page 4 line 150

“gen” is “gene”?

 

3. Figure 1

Authors should use other than PI3K inhibitor to show role of PI3K. Because inhibitors have side effects. Changes in activity of PI3K is also supports these results.

Why authors calculated salmonella survival % by dividing CFU’s recovery at 24 hours by CFU’s recovery at 1 hour? CFU at 1hour may involve in intervention. Some intervention affects within 1 hour but another did not affect within 1 hour. In this state, this calculate did not suitable to compare the results. Authors should use CFU before infection to B cells.

In addition, direct effect of inhibitor on bacteria should show.

Add sopB complement strain in Fig 1C.

 

4. Fig 2

“beta actina” is “beta actin”?

Expression levels of total pS6 should show.

Direct effect of inhibitor on bacteria should show.

Expression of SopB in bacteria should show.

Salmonella deltha sopB plus Rapamycin should be included in Fig 2B.

 

5. Fig 3

LC3 expression levels could not be explained by ULK1 phosphorylation especially in sopB complemented strain. pULK1 expression was similar to that of WT mice but LC3 expression of sopB complemented strain was higher than that of WT mice. How authors discuss this change?

 

6. Fig 4

Generally, n = 2 data can not be compared in biology. In other word, data for n=2 is inadequate

In legend, 1) UFC is CFU?

2) No (C ) figure.

The percentage (number) of CD19+GFP+ cell is too low at 4 days post-infection. Too few cells measured by flow cytometer.

 

7. Page 6 line 224

Authors did not show “inhibits autophagy in B lymphocytes to promote its survival and dissemination in the organism.

Author Response

Methods.

Data and statistical analysis is missing.

• A) We apologize for this omission. We included the analysis in this revised version,

 

Page 4 line 150

1. R) "gen" is "gene"?

• A) Thanks to the reviewer for pointing out this error; we corrected it.

 

Figure 1

1. R) Authors should use other than PI3K inhibitor to show the role of PI3K. Because inhibitors have side effects. Changes in activity of PI3K is also supports these results.

• A) We agree with the reviewer about the side effects of inhibitors; in this context, we tested cell viability, and we did not find a significant impact (Supplementary Figure S2). Instead of focusing on PI3K, we look downstream in the signaling pathway. Therefore, to avoid using pharmacological inhibitors to support the role of mTORC1 in blocking autophagy in Salmonella-infected B cells, we generate a raptor conditional knockout mice.

 

Why authors calculated salmonella survival % by dividing CFU's recovery at 24 hours by CFU's recovery at 1 hour? CFU at 1hour may involve in intervention. Some intervention effects within 1 hour but another did not affect within 1 hour. In this state, this calculation did not suitable to compare the results. Authors should use CFU before infection to B cells.

• A) We agree with the reviewer that after 1 hour, there is some intervention. There is a rapid control of bacteria by infected B cells. Due the fact that we want to know to which extent the bacteria survive after the initial control we calculated the % of bacterial survival by dividing the CFUs recovery at the final time-point between CFU after 1 hour.

 

In addition, the direct effect of inhibitor on bacteria should show.

• A) As the reviewer requested. We include the direct effect of inhibitors on bacterial growth and CFUs recovery (Supplementary Figure S2A).

 

Add sopB complement strain in Fig 1C.

• A) As the reviewer requested. In the revised version, SopB complement strain Figure 1 C includes the SopB complement strain.

 

Fig 2

"beta actina" is "beta actin"?

• A) Thanks to the reviewer for pointing out this mistake. We corrected it.

 

Expression levels of total pS6 should show.

• A) As the reviewer requested. In the revised version, Figure 2A shows total S6.

 

Direct effect of inhibitor on bacteria should show.

• A) Supplementary Figure S2A shows the direct effect of Rapamycin on Salmonella.

 

Expression of SopB in bacteria should show.

• A) We agree with the reviewer that it would be ideal to have the expression levels of SopB protein. Regrettably, we have not yet generated an antiserum against the protein, and there is not any commercial antibody available.

 

Salmonella deltha sopB plus Rapamycin should be included in Fig 2B.

• A) As the reviewer requested. Figure 2B of the revised version includes delta sopB added with Rapamycin.

 

Fig 3

1. R) LC3 expression levels could not be explained by ULK1 phosphorylation especially in sopB complemented strain. pULK1 expression was similar to that of WT mice but LC3 expression of sopB complemented strain was higher than that of WT mice. How authors discuss this change?

• A) As reviewer requested we discussed this point (lines 316-321)

 

1. R) Fig 4

Generally, n = 2 data cannot be compared in biology. In other word, data for n=2 is inadequate.

• A) We agree with the reviewer, unfortunately, mice production was not sufficient to increase the number of mice necessary to perform statistical analysis. However, we consider worthying to show the current result.

 

In legend, 1) UFC is CFU?

• A) Thanks to the reviewer for pointing out this mistake. We corrected it.

 

2) No (C) figure.

• A) Thanks to the reviewer for pointing out this mistake. We corrected it

 

The percentage (number) of CD19+GFP+ cell is too low at 4 days post-infection. Too few cells measured by flow cytometer.

• A) The reviewer is correct; the frequency of Salmonella-infected B cells is overly low. Thus, this also impacts its measure on flow cytometry. We believe that the main reason for the low frequency of infected B cells is that we infected mice with 2000 CFU. We could not accomplish a dose-response curve because we did not have enough mice.

 

1. R) Page 6 line 224

Authors did not show "inhibits autophagy in B lymphocytes to promote its survival and dissemination in the organism.

• A) We agree with the reviewer, and we deleted " and dissemination in the organism."

Round 2

Reviewer 1 Report

The authors should also address the concerns raised by the reviewers by adding data for the suggested experiments.

Author Response

We would like to address all the suggested experiments. But, in the first revised version, we include three supplementary figures showing the percentage of B cell purity and Salmonella-infected B cells (S1). The effect of inhibitors on Salmonella growth (S2). Characterization of Raptor KO mice and the frequency of B cells and macrophages of these mice (S3). Unfortunately, the production of conditional knockout mice was insufficient to evaluate the antibody production in infected mice, as suggested by the reviewer. In this revised version, we eliminated Figure 4 because it was impossible to increase the number of infected mice. An experiment with two infected mice is insufficient for statistical analysis. Therefore, in this second revised version,  a graphic abstract of the in vitro data replaced our previous Figure 4.

Additionally, MDPI's Author Services edited the manuscript in the English language.

Reviewer 2 Report

Authors addressed the questions raised during revision. Good luck with publishing the article. 

Author Response

Thanks for your good wishes and help. 

Reviewer 3 Report

My recommendation of this manuscript is “Minor revision”.

Sample size is too small to show in valuable paper.

I could not accept n =2 data in Fig4.

Author Response

The reviewer is correct. We should increase the number of mice for Figure 4 data, but as we mentioned in the rebutted of the first round of comments; Unfortunately, mice production was not sufficient to increase the number of mice. The results of Figure 4 are described as follows:

To reinforce the results obtained in vitro as described above, the cd19+/cre raptorfl/fl mice were infected via orogastric with Salmonella WT-GFP. Preliminary results show that cd19+/cre raptorfl/fl mice have a lower bacterial load in the spleen and liver than WT mice. Furthermore, when measuring the percentage of infection in mouse spleen B lymphocytes, a decrease in the percentage of CD19+ GFP+ cells was observed in cells of cd19+/cre raptorfl/fl mice compared those in WT mice (data not shown). Therefore, in this second revised version, a graphic abstract of the in vitro data replaced our previous Figure 4.

Additionally, MDPI's Author Services edited the manuscript in the English language.