The preprint of Bouquet et al. aims to establish a FACS approach for detection, sorting, and quantification of polyphosphate-accumulating bacteria in environmental samples. PolyP is an interesting energy-rich substance accumulated intracellularly by many organisms. However, we still lack a clear understanding of which taxa accumulate PolyP and under what environmental circumstances. Consequently, I agree with the authors that such an approach, and especially a relatively high-throughput approach via FACS, would be very helpful for future research.

General comments:

The authors grew a known PolyP-accumulating strain (T. elongata) and used a second strain as a control (their own isolate), that still showed “low” levels of PolyP accumulation. This is one of the first flaws of the experimental approach, as the control strain did in fact also accumulate polyphosphates. The authors tested a multitude of conditions for staining and storage of cells, although it does not become clear why all these conditions were tested in the first place. More severely, most of the resulting data is either not shown or not conclusive (see comments Figure 2 and 3). Following their tests, the authors applied their FACS approach to environmental samples and find that the FACS detection highly overestimates the cell numbers/proportions as compared to microscopy (which is the benchmarking tool). Surprisingly, the authors conclude that their approach is robust. I strongly disagree with this conclusion. Also, I fail to understand how the authors classified JC-D7 as a useful PolyP-stain when it failed to show similar numbers to DAPI (which is also has a questionable PolyP-specificity). The methods section is not detailed enough to allow for reproduction of the presented approach and tests. Many data are not shown despite the possibility to add data to a supplementary file. I ask the authors to resubmit a carefully revised version where all necessary data is shown, where the specificity of DAPI as a PolyP stain is critically discussed which will then also allow to critically interpret the data produced in this study. Currently, I cannot support the notion that the presented FACS-approach is a robust method to quantify and sort PolyP-accumulating cells from environmental samples. Please also include more literature including (https://enviromicro-journals.onlinelibrary.wiley.com/doi/10.1046/j.1462-2920.2001.00164.x, https://journals.asm.org/doi/10.1128/aem.02592-12, https://doi.org/10.1038/s41396-019-0399-7)

Specific comments:

Please do not use abbreviations in the abstract

It seems that JC-D7 and DAPI could be applied together as their spectrum is different from each other. This would also allow to see how specific both stains are.

L65f: add some results please

L77: I guess it should be MgSO4 x 7 H2O?

L901. 1500 x g does not sound much? How were the authors sure this sample did not contain particulate organic matter or non-microbial particles that could cause autofluorescence in the FACS?

L94: remove “The” at the beginning of the sentence please

L97f: please remove the ‘.’ Between µg or mg and ml-1. It should be µm ml-1. This appears throughout the manuscript

L100: what was the solvent used for DAPI stocks and why were they stored at -20°C?

L103: please choose to use either concentration or molarity throughout the methods section

113: what does qsp stand for?

L120: µm not µM

L125: Darmstadt

L126: how were the cells stored at these different temperatures? This is critical information missing

L131: again, please remove the dot ‘.’ between cell and s-1

L129 and L140: is it correct that two different machines were used for the FACS analysis (counting and sorting)?

L150f: it sounds as if the cells were filtered onto a black membrane, DAPI-stained, washed and DAPI-stained again. Is this correct? If so, why were the cells stained twice?

L153: Did the authors not use a mounting medium between filters and coverslips (e.g. Citifluor or Vectashield)?

L165 and also before: please indicate what solution was used to prepare the formaldehyde fixative. Water, 1 x PBS, or something else? Same applies for all stains (e.g. JC-D7) and solutions used. Please be more diligent in adding information

L181: it does not become clear what independent staining refers to and how that helped to establish thresholds. Please elaborate.

L185f: what software was used to perform statistical analyses?

Figure 1: Please try to use another micrograph for Figure 1B where the cells are magnified in a similar way as in Figure 1C.

L246: It does not become clear what the authors mean with “population structure”. Why would the staining buffer interfere with population structure in the first place?

L254: I am convinced that in the era of almost unlimited online space there is no need for “data not shown” anymore. I would like to ask the authors to show these data as well.

L288: why is this reference showing up in the brackets again? It is clear from the methods how the labelling was performed.

L290: Again, it is pretty unacceptable to not show these results. Please add them to the note or to a supplementary file so the readers can judge your conclusions.

L290: How did the authors decide if there was a strong or a weak fluorescent signal. This needs to be clarified in order to judge the results, e.g. the data shown in Figure 2C (comparison of HEPES and PBS).

L294f and Figure 1C: The interpretation of these results is pretty unclear to me. The FCM detection of DAPI-stained PolyP signal in the strain that supposedly did accumulate low values of PolyP in HEPES buffer showed a PolyP signal for 99% of the cells. This is surprising. The authors then state that in PBS, the number of cells with a PolyP signal was around 1%. Given the size of the bar in Figure 1C, this value should be around 5-10%. Please clarify. The authors then state that under PBS and microscopy a “correct” ratio was observed which is why they excluded HEPES buffer from further experimentation. First of all, how do the authors know what the correct value is if their “control” strain accumulated PolyP as well? Also, HEPES buffer without any P should in principle be more suited for PolyP detection than PBS, which introduces P to the cells. In addition, why does the microscopy data and the FCM data not agree more with each other? Microscopy should be the control here so this means that FCM highly overestimates the PolyP-detection?

Also I would like to ask the authors to also show the data of Figure 2C with TE cells. What are the proportions here? How did HEPES buffer fare in comparison to PBS?

L304: The data referred to here is not shown

Figure 3B shows the proportion of PolyP detected in cells RX (low-accumulators) for different fixation treatments (8-14%). These proportions are similar to the values shown in Figure 2C where they were referred to as 1%. Again, the data does not really add up and disagrees with the microscopy data shown in Figure 2C. In addition, I want to see the proportion of TE cells according to these tests as they were also counted (see Figure 3A). Why were they not shown in the first place?

L332: I am highly surprised to see that the storage at -80°C did not affect cell morphology, and thus detectability of PolyP. But again the authors chose to not show these data.

L334: How did the authors prepare the 50/50 mixture relative abundance for these tests?

L383f: The overestimation was already visible in Figures 2 and 3 which indicates that FCM as established in this study is not very useful for the quantitative detection of PolyP accumulating strains

L424: I strongly disagree with the conclusion that the “present study established the basis of a robust protocol for the detection and enrichment of PAB by flow cytometry”. The current study, in my opinion, fails to establish the FACS analysis in a comparable manner to fluorescence microscopy which serves as the benchmark.

The paper submitted by Bouquet et al. tries to establish a protocol to efficiently stain, detect and and sort polyphosphate accumulating bacteria using flow cytometry. Considering the importance of poly-P in wastewater treatment systems and the limited knowledge on poly-P cycling in natural environments such a protocol would be very useful.

General comments:

I have three major issues with this submission.

1. The experimental set-up; using T. elongata as a positive control is excellent and the use of environmental samples is good and necessary since a high-throughput method that is able to detect poly-P in natural samples could potentially give much more insights into the spatial and temporal dynamics of polyphosphate cycling. However, no real negative control has been used. A strain named “RX” was used as a negative control, even though it was shown to accumulate poly-P. Why not use an actual negative control? The data that has been collected for the “RX” strain could be an excellent addition to show how the protocol functions for species with minor amounts of poly-P, which might be relevant in natural settings.

2. A lot of the data is not shown. Add this data, either to the main results or the supplement.

3. The authors claim that this is a robust protocol whereas the data tells a different story. The observed differences between the proportion of poly-P cells detected by either FCM or Epifluorescence microscopy in sediment samples (Fig 2C, Fig 4BEF, Fig 5C and Fig 6C) shows that this is not a robust protocol and that FCM grossly overestimates the amount of cell accumulating poly-P.

Overall, I would conclude that the combination of DAPI staining and FCM would not work and that this method should not be used to detect and quantify polyP. However, since DAPI staining combined with fluorescence microscopy is often used to detect polyP, it is still important to get the knowledge out there to show that it does not work when combining with FCM. The data shown here where the dye FC-D7 was used seems to be more promising but requires more data (i.e., fluorescence microscopy to show that it works in prokaryotes) and a similar validation protocol as presented here for DAPI.

Specific comments:

Introduction

I think the introduction could elaborate a bit more or the functions and ubiquitousness of polyphosphate and put it in an environmental context. Almost all bacteria have the enzymatic potential for poly-P built-up and breakdown and besides being used as a storage compounds/energy reserve, many functions are known and none of them are mentioned in the introduction here. The following papers are excellent references:

1. Brown, M. R. W., and Kornberg, A. (2004). Inorganic polyphosphate in the origin and survival of species. Proc. Natl. Acad. Sci. 101, 16085–16087. doi: 10.1073/ pnas.0406909101

2. Kornberg, A. (1995). Inorganic polyphosphate - toward making a forgotten polymer unforgettable. J. Bacteriol. 177, 491–496. doi: 10.1128/jb.177.3.491496.1995

3. Kornberg, A., Rao, N. N., and Ault-riché, D. (1999). Inorganic polyphosphate: a molecule of many functions. Annu. Rev. Biochem. 68, 89–125. doi: 10.1146/ annurev.biochem.68.1.89

4. Rao, N. N., Gómez-García, M. R., and Kornberg, A. (2009). Inorganic polyphosphate: essential for growth and survival. Annu. Rev. Biochem. 78, 605–647. doi: 10.1146/annurev.biochem.77.083007.093039

Material and methods

L75: What are “large amounts” of polyP? Is it a large fraction of the cell volume? And if so, how much? Do most cells accumulate poly-P? And if so, how much?

L78: What is “a very small amount of intracellular polyP? Only few cells or does polyP only take up a small amount of the cell volume?

Results

Figure 1A: Beautiful TEM image, the A is on a weird spot in the figure. I would like to see false quantitative color scales for the EDX analyses if that is possible.

Figure 1BC: Make the figures bigger.

L247: What is meant by “population structure”?

L254 & 255: Show the data somewhere.

L290: Show the data.

L292 and Figure 2C: There are significant differences between the proportion of polyP+ cells observed by labeling with DAPI in PBS using the FCM or epifluorescence microscopy. Why is this? Isn’t this already a first insight the DAPI in combination with FCM is not an appropriate method to detect polyP accumulating bacteria because there are too many false positives?

Also, do you have any idea why the use of HEPES would give you so many polyP+ cells? Please elaborate.

Figure 3: Both staining periods, DAPI concentrations, and fixation with formaldehyde seem to significantly affect the detection of polyP+ cells with FCM. How was this with epifluorescence microscopy? Is the detection of polyP+ cells using fluorescence microscopy also affected by the staining period and DAPI concentration? This seems especially relevant since the differences in detection by epifluorescence microscopy when compared to FCM all seem to be significant according to the data presented in figure 1C, and later in Figure 4B, E, F, Figure 5B and Figure 6C. Also, even though there are no significant differences observed between fixed and unfixed polyP+ cells for T. elongata (L307), still show the data.

L332: Show the data. Especially because storage is highly relevant for natural samples.

Figure 4B, 4E, 4F: Where are the statistics between FCM and Epifluorescence? By eyeballing it, the differences seem significant and FCM appears to overestimate the amount of polyP+ cells before sorting and underestimates the amount op polyP+ cells after sorting.

L337: What is significant here? Which values are compared? FCM before sorting to FCM after sorting?

L339: PAB represented less than 10% of the polyP- fraction according to Epifluorescence but the proportion found by FCM is much lower. Why is this?

L378: I do not think that the method has been validated. There are significant differences observed between the FCM and Epifluorescence microscopy in almost every figure shown. The only thing that is validated is that this protocol can be used to enrich polyP+ cells (Figure 4 and Figure 5A).

Figure 5B: This really shows that this is not a robust method to be used for environmental samples.

L399-400: Show the data. Was this fluorescence intensity measured with FCM or epifluorescence microscopy.

Figure 6: Why was there no (epi)fluorescence microsocopy performed with JC-D7 labeling? And what is the significance between JC-D7 and DAPI-epifluo in the sediment? Again, the enormous difference between DAPI stained poly+ cells observed via FCM and Epifluorescence show that this is not a robust method.

The focus of the paper lies on the use of DAPI but looking at Figure 6, the dye JC-D7 looks much more promising for the use in natural samples (or at least for sediment samples). I think validating JC-D7 as a polyP specific dye in combination with epifluorescence microscopy that can also be used in prokaryotes would be a big step forward.

Discussion

L423: This line does not seem to be relevant here.

L458-462: I think this is an overestimation of what this protocol can do. In almost all the data shown, the observations with FCM and epifluorescence are significant, and especially in the natural samples, the differences are striking (Figure 5B and Figure 6C). The only thing that is shown is that polyP+ cells can be enriched. I would not recommend DAPI staining in combination with FCM to quantify the amount of polyP accumulating bacteria in a natural sample based on this data.

L469-472: This conclusion cannot be made without comparing the different staining periods and DAPI concentrations with epifluorescence microscopy.

L481: Show the results. The publication of negative results is important so other researchers do not have to try it for themselves.

L484-485: The data obtained shows promising results but it does not show the specific nature of PAB labeling. This would require fluorescence microscopy to visualize the JC-D7 dye in combination with for instance SEM-EDS so co-localization analysis can be performed