The FlgN chaperone activates the Na+-driven engine of the Salmonella flagellar protein export apparatus

The bacterial flagellar protein export machinery consists of a transmembrane export gate complex and a cytoplasmic ATPase complex. The gate complex has two intrinsic and distinct H+-driven and Na+-driven engines to drive the export of flagellar structural proteins. Salmonella wild-type cells preferentially use the H+-driven engine under a variety of environmental conditions. To address how the Na+-driven engine is activated, we analyzed the fliJ(Δ13–24) fliH(Δ96–97) mutant and found that the interaction of the FlgN chaperone with FlhA activates the Na+-driven engine when the ATPase complex becomes non-functional. A similar activation can be observed with either of two single-residue substitutions in FlhA. Thus, it is likely that the FlgN-FlhA interaction generates a conformational change in FlhA that allows it to function as a Na+ channel. We propose that this type of activation would be useful for flagellar construction under conditions in which the proton motive force is severely restricted.

Page 5, first paragraph: Figure 1b shows that fliC is not expressed in HI0117 and HIJ0117 indicating that FlgM secretion is impaired. The authors should mention this. Line 109: same thing. The authors state that transcription is not increased, but clearly fliC transcription must be increased with 100 mM NaCl. How else does one explain the difference in cellular FliC levels?
Guo at el. Showed that flagellar secretion increased with either NaCl or KCl: Guo, S., I. Alshamy, K.T. Hughes, and F.F.V. Chevance. Analysis of factors that affect FlgMdependent type III secretion for protein purification with Salmonella Typhimurium. J. Bacteriol. Sentence lines 227-229. There is no evidence that FlgN turns on sodium-coupled export.
Reviewer #2 (Remarks to the Author): This work by Minamino et al. examines how the flagellar export apparatus of Salmonella enterica can be driven by the sodium motive force SMF when the coupling to the proton motive force is somehow disrupted. The message is not nearly as easily extracted from the manuscript as my simple description might suggest.
First, I think the authors should explain in the Introduction, and mention in the Abstract, conditions under which having flagellar protein export driven by the SMF might be relevant. They leave any mention of this until the last paragraph of the discussion, in which they propose that within biofilms cells may have a lower proton motive force and profit from having another way to construct flagella that does not depend on the PMF. One then wonders why they did not look at flagellar synthesis under conditions in which the PMF is depleted and see whether the SMF becomes important under those circumstances.
Instead, they utilize a fliJD13-34 fliHD96-97 mutant (MM104-3) to present evidence that the SMF is responsible for energizing the export of flagellar proteins in this strain. That is all right, as the mutant is no doubt easier to manipulate than the PMF. They also show that deletion of the flgN gene compromises the ability of the mutant to use the SMF and that FlgN physically interacts with FliJ and FlhA.
Here is what I understand of the central conclusion. In the wild type, FliJ interacts with FlhA to activate it as a proton channel. In the double mutant, the altered FliJ protein can still interact with FlhA, but primarily to activate it as a sodium ion channel. This activation relies on the FlgN protein in a second function in addition to its role as a chaperone for the export of FlgK and FlgL. FlgN is shown to interact with both FliJ and FlhA, data that support its role in mediating the activation of FlhA as a sodium channel. (The information may be in the text somewhere, but I was not clear whether FlgN is important for the activation of of the proton channel activity of FlhA in the wildtype strain.) It seems that this is the important information. However, the Results section goes on with an additional mutant strain (MMHI0117) and its derivative MMHIJ0117 and pseudorevertants that arise through mutations in flhA or by elimination if FlgM, the negative transcriptional regulator of FliA, the flagella-specific sigma factor. All of this became totally confusing to me so that I lost sight of what the real purpose of the study was. Fortunately, things calmed down a bit and became more clear in the Discussion.
I strongly suggest that the authors take stock of what it is that they really want the reader to understand. The Results section is far too much of a data dump, and I am not convinced that much of that data helps make their main point. Rather, I think it obscures it. The Abstract seems straightforward, and the data results presented in the paper should all be relevant to providing evidence for the conclusion that "the interaction of FlgN with FlhA opens the Na+ channel in the export engine, thereby maintaining the protein export activity in the absence of the active ATPase complex." A sentence might be added to the Abstract to consider when this sort of activation would be useful, as per the last paragraph in the Discussion.
Reviewer #3 (Remarks to the Author): Minamino et al. demonstrate extensive evidence for the mechanism by which sodium-powered flagellar export is activated. This is of great interest in the flagellar community, but also more widely, as an example of built-in redundancy in biology as well as dual-powered protein machinery.
The methods are a nicely blended mix of microbiology, biochemistry and microscopy, and the work is highly suitable for publication in Communications Biology. I enjoyed learning about aspects of the export apparatus and the level of detail that had gone into the experimental design to establish interactions between the components of the system. I have the following minor comments: P5L23: Why, in general, was 100 mM NaCl chosen as the sodium concentration of choice? It is OK, but I was just wondering why; much previous SMF work in the BFM has been done at 85 mM NaCl. In particular for Fig. 1 I was wondering if an extra plate measurement at, eg, 20 mM or 5mM NaCl could be done to show some trend in swim plate size vs sodium. This wouldn't be such a hard experiment to do and it might give us some idea if it was an 'all or nothing' approach, or if there was some tunability in the system with SMF. P6L5: I was wondering in general in Fig. 2 also what the timing of the experiment was, in that surely filament length depends somewhat on cell cycle (time after division). We should be sampling a distribution of this, but could some comment be made here regarding this? You have measured the WT distribution of filament length and number, and perhaps this is somewhat correlated with cell shape and size, and life cycle? For example, I am unsure if published yet but there is some info of relevance in this preprint (and could be discussed from this manuscript's authors' previous work also) (https://doi.org/10.1101/767319). P6L21: I am unfamiliar with the use of the phrase 'bait' here, perhaps that is my ignorance but not sure if it is totally clear, can we clarify? P7L5 typo: mediated P14L10-17: The flow here seems a little surprising, and it is the last paragraph? Why are we suddenly talking about biofilms and subpopulations? I felt that I needed a couple more sentences to transition from the last paragraph here, it just caught me a bit by surprise. Perhaps add some more info here to contextualise? P15L10: for motility plates 7 independent measurements were done. Is this data included somewhere? Might be nice to plot mean/SD or scatter plot of all measurements to see variation/reproducibility of these motility plates, as an SI Figure. P23Fig1, as above, can you comment on dependence on Na, or include a midpoint Na measurement for the plates. Also, I appreciate the strain names are what they are because you made them but it is a little hard to follow at times, because we have 104-3, HI0117 and HIJ0117, but I found myself having to constantly refer to caption or text to remember what was what. Is it perhaps possible to include phenotype on the figure somehow? I realise this is subtle but it does slow the reading down so if there is a way to define a phenotype descriptor throughout would help flow. P24Fig.2: For panel C it would be better I think if this data was presented as a scatter plot with overlay Mean + SD, because we aren't able to tell easily from this what the distribution looks like, most relevantly if there are a small subpopulation of cells with very small filaments, or any difference between WT and 104-3 in terms of distribution. The number of measurements is high (Supp . Table 1/2), so it would be good to explore how this was distributed. Fig3: This is a very nice figure, well put together, that combines EM with blot data. Well done. SI Fig 2: Alexa594 is used both here and in Fig2, but in SI Fig 2 it is coloured red vs magenta in Fig. 2. Perhaps make consistent, I personally prefer red but it doesn't matter just a little odd to have two colours for same fluorophore.
Our responses are listed below.

To Reviewer #1:
This builds on previous work from the same group, which made the surprising discovery that protein secretion by the flagellar type III secretion could use a sodium gradient in addition it a proton gradient to energize protein secretion when the FliHIJ ATPase complex associated with secretion is not functioning properly. The authors previously described a fliJ mutant deleted for residues 13-24 that was defective in interaction with FlhA, which is the ion conduit for coupling ion import to protein export. A two-residue deletion in  suppressed this defect and restored some motility to the fliJ mutant. The authors speculated that the fliJ mutant effect on FlhA might be on its sodium channel function.
Re: Thank you so much for your comments.
In the first chapter, the authors show that addition of 100 mM NaCl improved motility for three mutant strains: MM104-3 (fliJ(D13-24) fliH(D96-97)), MMHI0117 (DfliHI flhB(P28T) and MMHIJ0117 (DfliHIJ flhB(P28T). The main difference appears to be the MM104-3 is already motile and the presence of FliJ in MMHI0117 enhances motility compared to the FliHIJ deletion in MMHI0117. Not surprising enhance motility in 100 mM NaCl correlates with enhances secretion of substrates. Figure 1b shows that fliC is not expressed in HI0117 and HIJ0117 indicating that FlgM secretion is impaired.
Re: Both MMHI0117 and MMHIJ0117 strains did not secret FlgM into the culture media in the absence of 100 mM NaCl. Addition of 100 mM KCl did not facilitate the secretion of FlgM by these two strains, either. Thus, only Na + has a positive impact on FlgM secretion by these two strains (Please see Supplementary Fig. 4

of the revised manuscript).
The effect of pH 7.5 is said to diminish the proton gradient. However, in the referenced article the effect of external pH on internal pH was with addition of sodium benzoate and at pH 7.5 the internal pH was the same with or without added benzoate (Figure 1 of referenced paper). I fail to find data to support the assertion that pH 7.5 results in significant reduction in the proton gradient.
Re: Please see Supplementary Table S1 of the Supplemental Information of the referenced paper, which we extracted and attached below. The intracellular pH was measured to be about 7.5 when external pH was 7.5. Thus, a chemical gradient of protons (DpH) is gone at an external pH value of 7.5. To confirm this, we measured intracellular pH of Salmonella cells using a pH sensor probe, pHluorin(M153R), and found that the intercellular pH was 7.41 ± 0.05 when the cells were exponentially grown in T-broth (pH 7.5). We now mentioned this point in text.

The authors next show that cells have more and longer flagella with 100 mM NaCl. It is known that addition of NaCl turns on Spi1 including HilD which increases flhDC transcription and could easily account for more and longer flagella. What is the effect of 100 mM NaCl on flhDC transcription? The authors assume an SMF effect without considering other possibilities based on published work.
Re: Thank you very much for your comments. Because it has been reported that increased ionic strength facilitates the export of a flagellum-specific anti-sigma factor, FlgM, by wild-type cells, enhancing motility in soft agar In wild-type cells, neither Na + , Li + , K + nor Mg 2+ affected the secretion level of FlgD. In the DfliH-fliI flhB(P28T) mutant, Na + dramatically enhanced FlgD secretion whereas either Li + , K + or Mg 2+ did not. Because Na + did not increased the cellular level of FlgD at all, this suggests that Na + is specific for this positive impact on flagellar protein export by the DfliH-fliI flhB(P28T) mutant. We also showed that the secretion level of FlgD by the DfliH-fliI flhB(P28T) mutant increased with an increase in external NaCl concentration and reached a maximum at 100 mM NaCl although their cellular levels were almost constant over a wide range of external NaCl concentration (Please see Fig. 1B  Because we used SJW1103 as a wild-type strain for motility and chemotaxis studies, we assumed that Na + does not activate FlhDC transcription in the SJW1103 background. The next chapter predicts that a flagellar protein is required for the export engine to prefer the SMF to the PMF. A GST-FliJ fusion inhibits FlhA-dependent export in the MMHI0117 strain. A GST-FliJ was able to co-purify with the FlgN protein and known secretion chaperone for FlgK and FlgL by preventing degradation of FlgK and FlgL in the cytoplasm. However, it was shown by a plasmid clone of the flgKL genes would alleviate the requirement for FlgN (see : Aldridge, et al. Mol. Microbiol. (2003) 49: 1333

previously that FlgN interacts with FlhA. The authors want to test if FlgN is somehow involved in the SMF coupled to export. Deletion of flgN in the 104-3 strain resulted in no motility. Loss of FlgN was specific to
FlgKL export as the authors observe no effect on early substrate export FlgD and FlgE. The authors need to look at FlgM export since a defect in FlgM export would inhibit fliC expression and would explain no filaments in Sup Fig 2A and  Re: Thank you so much for your comments. FlgN is known to act as a cytoplasmic chaperone specific for two hook-filament junction proteins, FlgK and FlgL, not only to protect them from degradation by proteases but also to support their efficient docking to FlhA. We cited the Aldridge Mol. Microbiol. 2003 paper and described this important information in the revised manuscript. In our present study, we found that removal of FlgN from the MM104-3 strain inhibited the secretion of FlgD and FlgE into the culture media (Please see Fig. 3b of the revised manuscript). It has been reported that FlgM is secreted into the culture media upon completion of the hook structure, allowing FliA to induce the transcription of class III genes such as FlgK, FlgL and FliC (Hughes et al. Science 1993; Kutsukake, MGG 1994). We analyzed the cellular levels of FlgK, FlgL and FliC by immunoblotting and found that deletion of FlgN results in a considerable decrease in the cellular levels of FlgK, FlgL and FliC, indicating that FlgN deletion inhibits the secretion of FlgM as well.

Consistently
The next chapter describes the isolation of pseudorevertants form the 104-3 flgN deletion strain. Two previously described alleles of flhA that suppressed flgN motility defect were tested on the 104-3 flgN deletion and as expected also suppressed in this background. The authors suspect the FlgN-FlhA interaction is needed for FlgD and FlgE secretion. That stream of logic is not apparent to this reviewer. Somehow, they conclude FlgN activates sodium driven export. I have no idea how that conclusion is come to -absolutely no idea. Not surprising another suppressor was a flgM null allele. Re: Thank you so much for your understanding.
The next chapter reports to look at the effect of flhA alleles, which suppress the secretion defect in the absence of FlgN, on sodium-coupled export. There is nothing in this chapter that looks at sodium driven export. They show that FliJ has a positive effect of flagellar synthesis. This has been reported by others. Again, they suggest the sodium effect is export when they haven't ruled out other factors. I bet 200 mM KCl would give a similar result. That this is a salt concentration effect.
Re: As mentioned above, we previously showed that K + did not promote the secretion of FlgD by the DfliH-fliI flhB(P28T) strain (Please see Fig. 1A of Minamino et al. PLOS Pathog. 12, e1005495, 2016, which is attached above). We reexamined whether KCl facilitates motility of and flagellar protein export by the DfliH-fliI flhB(P28T) and DfliH-fliI-fliJ flhB(P28T) strains in a way similar to NaCl and again found that K + did not enhance motility of and flagellar protein export by these two strains (Please see Supplementary Fig. 3 of the revised manuscript).
Overall this manuscripts reads as though the authors are trying to subjectively prove a sodium-export hypothesis, based on suppressor analysis of export-defective mutants and a positive effect of 100 mM NaCl on secretion rather than objectively test this hypothesis and try to invalidate the sodium-dependent export hypothesis.
Re: Because we confirmed that KCl did not affect motility of and flagellar protein export by the DfliH-fliI flhB(P28T) and DfliH-fliI-fliJ flhB(P28T) mutants (Please see Supplementary Fig. 3 of the revised manuscript), we believe that the transmembrane export gate complex prefers to use Na + as the coupling ion to drive flagellar protein export and assembly in the absence of the cytoplasmic ATPase complex and that the FlgN chaperone contributes to the efficient Na + -coupled protein export by the export gate complex. Re: Cited this reference.
Page 5, first paragraph: Figure 1b shows that fliC is not expressed in HI0117 and HIJ0117 indicating that FlgM secretion is impaired. The authors should mention this.

Re: Agreed and mentioned in text.
Line 109: same thing. The authors state that transcription is not increased, but clearly fliC transcription must be increased with 100 mM NaCl. How else does one explain the difference in cellular FliC levels?
Re: We used PfliC::luxCDABE to measure the fliC promoter activity in the presence and absence of 100 mM NaCl or KCl in the SJW1103 background. The fliC promotor activities were measured to be 1.36 ± 0.33 x 10 7 RLU and 1.57 ± 0.33 x 10 7 RLU in the presence of 100 mM KCl and 100 mM NaCl, respectively, whereas the promotor activity was 2.45 ± 0.44 x 10 7 RLU in the absence of either KCl or NaCl. This is in agreement with our previous report showing that the transcription levels of flagellar genes were not increased by adding 100 mM NaCl (Minamino et al. PLOS Pathog.  12, e1005495, 2016). Consistently, we found that neither Na + nor K + affected the steady cellular levels of flagellar proteins. However, we do not know why neither Na + nor K + exert the positive impact on flagellar gene transcription in the SJW1103 background. To avoid confusion, we would not like to show our transcription assay data until we will have a clear reason. Re: We carried out motility analysis and flagellar protein export assays as suggested by this reviewer and found that KCl did not facilitate motility of and flagellar protein export by the wild-type, MM104-3, DfliH-fliI flhB(P28T) and DfliH-fliI-fliJ flhB(P28T) strains (Please see Supplementary Fig. 3 of the revised manuscript). The motility ring size of wild-type cells was slightly larger in the presence of 100 mM NaCl and 100 mM KCl than in their absence presumably due to the slow growth rate in their absence because the motility ring size depends on the bacterial growth rate (Please see Supplementary Fig. 2

of the revised manuscript).
Line 143: change "caused" to "resulted in"

Line 159-160: No, that is not what the results indicate. The results most likely indicate that FlgM secretion is defective.
Re: Removal of FlgN from the MM104-3 strain inhibited the secretion of FlgD and FlgE into the culture media (Please see Fig. 3b of the revised manuscript).
Consistently, the MM104-3 DflgN strain produced only the MS-C ring structures (Please see Fig. 3c of the revised manuscript). In contrast, the removal of FlgN from the SJW1103 strain considerably reduced the secretion levels of FlgK and FlgL but did not affect the secretion levels of FlgD and FlgE at all (Please see Fig. 3b of the revised manuscript). Consistently, the DflgN strain produced only hook-basal bodies in a way similar to the wild-type strain. These results indicate that FlgN becomes essential for the export of all flagellar structural subunits in the MM104-3 strain background rather than for FlgM secretion.

Sentence lines 227-229. There is no evidence that FlgN turns on sodium-coupled export.
Re: Pull-down assays by GST affinity chromatography revealed that FlgN did not improve the interaction between FliJ(D13-24) and FlhAC (Please see Fig. 4 of the revised manuscript). Unlike the FliS and FliT chaperones, which require their cognate substrates, FliC and FliD, respectively to bind to FlhAC, FlgN alone can bind to FlhAC (Please see Fig. 4

of the revised manuscript). Because FlgN is required
for the activation of the Na + -driven export engine of the transmembrane export gate complex in the MM104-3 strain, we suppose that the FlgN-FlhAC interaction turns on Na + -coupled protein export by the MM104-3 strain.

This work by Minamino et al. examines how the flagellar export apparatus of Salmonella enterica can be driven by the sodium motive force SMF when the coupling to the proton motive force is somehow disrupted. The message is not nearly as easily extracted from the manuscript as my simple description might suggest.
Re: Thank you so much for your useful comments and suggestions. We tried to make such message get clearer and easily extracted as much as possible.

First, I think the authors should explain in the Introduction, and mention in the Abstract, conditions under which having flagellar protein export driven by the SMF might be relevant. They leave any mention of this until the last paragraph of the discussion, in which they propose that within biofilms cells may have a lower proton motive force and profit from having another way to construct flagella that does not depend on the PMF. One then wonders why they did not look at flagellar synthesis under conditions in which the PMF is depleted and see whether the SMF becomes important under those circumstances.
Re: Thank you so much for your useful comments and suggestions. We described this point as suggested by this reviewer.
Instead, they utilize a fliJD13-34 fliHD96-97 mutant (MM104-3) to present evidence that the SMF is responsible for energizing the export of flagellar proteins in this strain. That is all right, as the mutant is no doubt easier to manipulate than the PMF.

Re: Thank you so much for your understanding.
Here is what I understand of the central conclusion. In the wild type, FliJ interacts with FlhA to activate it as a proton channel. In the double mutant, the altered FliJ protein can still interact with FlhA, but primarily to activate it as a sodium ion channel. This activation relies on the FlgN protein in a second function in addition to its role as a chaperone for the export of FlgK and FlgL. FlgN is shown to interact with both FliJ and FlhA, data that support its role in mediating the activation of FlhA as a sodium channel.

(The information may be in the text somewhere, but I was not clear whether FlgN is important for the activation of the proton channel activity of FlhA in the wildtype strain.)
Re: Thank you so much for your understanding. Re: Thank you so much for your comments. We needed to provide these mutant strains' data because these data fully support that the export gate complex becomes a Na + -coupled export engine when the cytoplasmic ATPase complex becomes nonfunctional and that the FlgN chaperone is required for the activation of the Na + -driven export engine. We agree, however, that the background information for the experimental design may have been rather limited for readers to fully understand the purpose of this study. We added necessary background information in the Abstract, Introduction and Results.

It seems that this is the important information. However, the Results section goes on with an additional mutant strain (MMHI0117) and its derivative MMHIJ0117 and pseudorevertants that arise through mutations in flhA or by elimination if
I strongly suggest that the authors take stock of what it is that they really want the reader to understand. The Results section is far too much of a data dump, and I am not convinced that much of that data helps make their main point. Rather, I think it obscures it. The Abstract seems straightforward, and the data results presented in the paper should all be relevant to providing evidence for the conclusion that "the interaction of FlgN with FlhA opens the Na+ channel in the export engine, thereby maintaining the protein export activity in the absence of the active ATPase complex." A sentence might be added to the Abstract to consider when this sort of activation would be useful, as per the last paragraph in the Discussion.
Re: Thank you so much for your comments. We rewrote the Results section to make our main point clearer as much as possible. We also rewrote the Abstract as follows: "The bacterial flagellar protein export machinery is composed of a transmembrane export gate complex and a cytoplasmic ATPase complex. The export gate complex intrinsically has two distinct, H + -and Na + -driven export engines to facilitate the export of flagellar structural proteins. Planktonic Salmonella wild-type cells prefer to use the H +driven export engine to construct flagella on the cell surface under a variety of environmental conditions. However, it remains unknown when and how the Na + -driven export engine is activated. To clarify this question, we analyzed the fliJ(D13-24) fliH(D96-97) mutant and provided evidence that the interaction of the FlgN chaperone with FlhA activates the Na + -driven export engine, thereby maintaining the protein export activity when the ATPase complex becomes non-functional. We propose that this type of activation would be useful for flagellar construction under conditions in which proton motive force across the cytoplasmic membrane is severely restricted."

To Reviewer #3:
Minamino et al. demonstrate extensive evidence for the mechanism by which sodium-powered flagellar export is activated. This is of great interest in the flagellar community, but also more widely, as an example of built-in redundancy in biology as well as dual-powered protein machinery.
The methods are a nicely blended mix of microbiology, biochemistry and microscopy, and the work is highly suitable for publication in Communications Biology. I enjoyed learning about aspects of the export apparatus and the level of detail that had gone into the experimental design to establish interactions between the components of the system.
Re: Thank you very much for your supportive comments.
I have the following minor comments: P5L23: Why, in general, was 100 mM NaCl chosen as the sodium concentration of choice? It is OK, but I was just wondering why; much previous SMF work in the BFM has been done at 85 mM NaCl. In particular for Fig. 1 I was wondering if an extra plate measurement at, eg, 20 mM or 5mM NaCl could be done to show some trend in swim plate size vs sodium. This wouldn't be such a hard experiment to do and it might give us some idea if it was an 'all or nothing' approach, or if there was some tunability in the system with SMF.
Re: We previously measured the levels of FlgD secreted by the ΔfliH-fliI flhB(P28T) mutant cells grown in T-broth (pH 7.5) containing 10, 25, 50 and 100 mM NaCl and showed that FlgD secretion levels increased with increasing external Na + concentration and reached a maximum at a Na + concentration of 100 mM (Please see Fig. 1B of Minamino et al. PLOS Pathog. 12, e1005495, 2016, attached above). Therefore, we analyzed the motility in the presence and absence of 100 mM NaCl. Fig. 2 also what the timing of the experiment was, in that surely filament length depends somewhat on cell cycle (time after division). We should be sampling a distribution of this, but could some comment be made here regarding this? You have measured the WT distribution of filament length and number, and perhaps this is somewhat correlated with cell shape and size, and life cycle? For example, I am unsure if published yet but there is some info of relevance in this preprint (and could be discussed from this manuscript's authors' previous work also) (https://doi.org/10.1101/767319).

P6L5: I was wondering in general in
Re: Thank you so much for your comments. The number of flagellar filaments and filament length were measured when the cells reached the stationary growth phase. We described this point in the figure legend.

P6L21: I am unfamiliar with the use of the phrase 'bait' here, perhaps that is my ignorance but not sure if it is totally clear, can we clarify?
Re: When searching for the target protein from a huge number of cellular proteins, the protein with a specific tag become a "bait" to catch its binding partner(s), just like fishing. Therefore, "bait" is often used in Biochemistry.

P7L5 typo: mediated
Re: Corrected P14L10-17: The flow here seems a little surprising, and it is the last paragraph? Why are we suddenly talking about biofilms and subpopulations? I felt that I needed a couple more sentences to transition from the last paragraph here, it just caught me a bit by surprise. Perhaps add some more info here to contextualise?
Re: Thank you so much for your comments. We added the following sentences before start talking about biofilms.
"In Salmonella planktonic wild-type cells, the transmembrane export gate complex prefers to use PMF to transport flagellar structural subunits to the cell exterior 16-18,21 . However, when the cytoplasmic ATPase ring complex becomes non-functional 24 , which occurs in the biofilm formation, the export gate complex prefers to use SMF over a wide range of external pH 10 ." P15L10: for motility plates 7 independent measurements were done. Is this data included somewhere? Might be nice to plot mean/SD or scatter plot of all measurements to see variation/reproducibility of these motility plates, as an SI Figure. Re: We measured a diameter of the motility ring of 10 colonies of each strain in the presence and absence of 100 mM NaCl and provided relative diameter of each motility ring in Fig. 1a. The average diameter of the motility ring of each strain grown in the presence of 100 mM NaCl was set to 1.0. P23Fig1, as above, can you comment on dependence on Na, or include a midpoint Na measurement for the plates.
Re: As mentioned above, we already showed that the secretion levels of flagellar building blocks increased with an increase in the external NaCl concentration and reached to a maximum at 100 mM NaCl (Please see Fig. 1B  Re: To make it easier to read the text, we changed "104-3", "HI0117" and "HIJ0117" to "fliJ(D13-24) fliH*", "DfliHI flhB*" and "DfliHIJ flhB*", respectively. Re: Thank you so much for this comment. Fig. 2. Perhaps make consistent, I personally prefer red but it doesn't matter just a little odd to have two colours for same fluorophore.

SI Fig 2: Alexa594 is used both here and in Fig2, but in SI Fig 2 it is coloured red vs magenta in
Re: Thank you for the point. The color of the flagellar filaments was unified to "magenta" for the readers with color blindness.
My concerns have all been adequately addressed, nice work! Reviewer #2 (Remarks to the Author): I raised several points in my original review. One of them, leaving an explanation of the context in which a second, sodium-coupled export channel might be important for Salmonella, is now addressed in the Introduction. This research addresses an important problem and provides a convincing answer. The difficulty lies in the inaccessibility of the information to any but the most tenacious, diligent and experienced reader.
One point I raised in the original review was the confusion caused by all of the different strain names, which were repeated again and again as genotypes, which became very cumbersome and confusing. This point has not been dealt with adequately. I suggest that the authors include a table in which they assign each genotype they include in the results a simple name that they then use after the first description in the text. The reader will be able to consult the table to refresh memory. The important properties of each strain with respect to motility, flagellation pattern, and protein export capacity could also be summarized in this table or in an additional table or tables.
In the current manuscript, it is sometimes not clear which results are new and which are repeats from previous publications. It is annoying to read, "in agreement with previous results" followed by a literature citation, as it raises the question of what is really new. I understand that these are control experiments, but in the Results section they are often treated as though they are novel until that qualifying phrase appears.
The biggest problem is that the Results section is still almost unreadable. I will outline below what I think are the important findings, but I had to read and reread the manuscript and then synthesize for myself the major conclusions. Most readers will not be willing to do that. The authors should get the help of an native English speaking scientist who is familiar with bacterial flagellar assembly to assist with the production of a readable document. I am convinced that the data can be presented in a coherent fashion.
Here is my understanding of the results.
Motility of the fliJ(D13-24) fliH(D96-97) strain is poor because of a compromised interaction of FliJ of the FliHIJ ATPase with FlhAC. This interaction is important for activating proton conduction through FlhA to drive flagellar protein export. Motility is enhanced by 100 mM NaCl, leading to the conclusion that a sodium-coupled channel is important for export of flagellar proteins when the function of the ATPase complex is compromised. FlgN, the known chaperone for the capping proteins FlgK and FlgL, appears to be important for activation of this "bypass" sodium-coupled export. Unlike FliJ, which interacts with the linker region, FlhAL, that joins the transmembrane and cytoplasmic (FlhAC) domains of FlhA, FlgN interacts directly with FlhAC. Two residue replacements, D456V and T490M, in the region of FlhAC that interacts with FlgN allow some sodium-coupled export of flagellar proteins in the absence of FliHIJ and FlgN. This result suggests that interaction of FlgN with FlhAC causes a conformational change in FlhA that converts it from a protonconducting into a sodium-conducting channel. The D456V and T490M substitutions in FlhAC partially mimic the conformational change induced by FlgN binding to FlhAC. FlgN-mediated activation of sodium-coupled export of flagellar proteins may be important to allow for the presence of some motile cells under conditions of low proton motive force and high cyclic-di-GMP, as within biofilms. (There is some more involving the P28T substitution in FlhB and the L244R substitution in FliI that I found hard to understand and integrate into the central story.) If my understanding is correct, then it should be possible to present the data in a logical and digestible fashion. That is, in my opinion, what the authors need to do.

1
Our responses are listed below. We highlighted all changes in red in the revised manuscript (Marked Up version).

To Reviewer #1
My concerns have all been adequately addressed, nice work! Re: Thank you so much for your support for publication of our manuscript in Communications Biology.
To Reviewer #2 I raised several points in my original review. One of them, leaving an explanation of the context in which a second, sodium-coupled export channel might be important for Salmonella, is now addressed in the Introduction. This research addresses an important problem and provides a convincing answer. The difficulty lies in the inaccessibility of the information to any but the most tenacious, diligent and experienced reader.
Re: Thank you so much for your useful comments.
One point I raised in the original review was the confusion caused by all of the different strain names, which were repeated again and again as genotypes, which became very cumbersome and confusing. This point has not been dealt with adequately. I suggest that the authors include a table in which they assign each genotype they include in the results a simple name that they then use after the first description in the text. The reader will be able to consult the table to refresh memory. The important properties of each strain with respect to motility, flagellation pattern, and protein export capacity could also be summarized in this table or in an additional  table or tables. Re: Thank you so much for your comments. We agreed with this reviewer and provided abbreviations for all strains which we used in this study and a table summarizing their important properties regarding motility, flagellar protein export and flagellar assembly.
In the current manuscript, it is sometimes not clear which results are new and which are repeats from previous publications. It is annoying to read, "in agreement with previous results" followed by a literature citation, as it raises the question of what is really new. I understand that these are control experiments, but in the Results section they are often treated as though they are novel until that qualifying phrase appears.
Re: Agreed and so we rewrote the Results section to make our new discoveries as clear as possible.
The biggest problem is that the Results section is still almost unreadable. I will outline below what I think are the important findings, but I had to read and reread the manuscript and then synthesize for myself the major conclusions. Most readers will not be willing to do that. The authors should get the help of an native English speaking scientist who is familiar with bacterial flagellar assembly to assist with the production