A salt bridge-mediated resistance mechanism to FtsZ inhibitor PC190723 revealed by a cell-based screen

  1. Ajay Kumar Sharma
  2. Sakshi Mahesh Poddar
  3. Joyeeta Chakraborty
  4. Bhagyashri Soumya Nayak
  5. Srilakshmi Kalathil
  6. Nivedita Mitra
  7. Pananghat Gayathri
  8. Ramanujam Srinivasan

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Bacterial cell division proteins, especially the tubulin homolog FtsZ, have emerged as strong targets for developing new antibiotics. Here, we have utilized the fission yeast heterologous expression system to develop a cell-based assay to screen for small molecules that directly and specifically target the bacterial cell division protein FtsZ. The strategy also allows for simultaneous assessment of the toxicity of the drugs to eukaryotic yeast cells. As a proof-of-concept of the utility of this assay, we demonstrate the effect of the inhibitors sanguinarine, berberine and PC190723 on FtsZ. Though sanguinarine and berberine affect FtsZ polymerization, they exert a toxic effect on the cells. Further, using this assay system, we show that PC190723 affects Helicobacter pylori FtsZ function and gain new insights into the molecular determinants of resistance to PC190723. Based on sequence and structural analysis and site-specific mutations, we demonstrate that the presence of salt-bridge interactions between the central H7 helix and beta-strands S9 and S10 mediate resistance to PC190723 in FtsZ. The single-step in vivo cell-based assay using fission yeast enabled us to dissect the contribution of sequence-specific features of FtsZ and cell permeability effects associated with bacterial cell envelopes. Thus, our assay serves as a potent tool to rapidly identify novel compounds targeting polymeric bacterial cytoskeletal proteins like FtsZ to understand how they alter polymerization dynamics and address resistance determinants in targets.

Article activity feed

  1. Version published to 10.1101/2022.04.06.487355v2 on bioRxiv
  2. Review Commons

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

    Learn more at Review Commons

    Reply to the reviewers

    Response to Reviewers:

    1. General Statements

    We thank the reviewers for the comments and the suggestions. We hope that we have addressed all the queries raised by the reviewers in the revised manuscript. We provide a point-by-point response below. Please note that the line numbers indicated in parentheses correspond to the pdf file without the track changes display.

    2. Point-by-point description of the revisions

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

    Summary: Srinivasan and co-workers developed an alternative screening method for defining the ability of FtsZ inhibitor to affect FtsZ polymerization. This alternative assay was defined considering the expertise of the authors on the topic, and they use Schizosaccharomyces pombe as a model for studying the effect of PC190723, sanguinarine and berberine on FtsZ assembly. The use of a heterologous expression system is useful for the evaluation of FtsZ coming from different strains, both Gram - and Gram +. The same model could gain insights also on the capability of FtsZ inhibitors to affect eukaryotic cell physiology. Finally, authors resulted also in suggesting a possible cause to suspected resistance to PC190723 from Gram - strains as E. coli.

    Major comments: • The conclusions are included in the discussion section and are quite convincing, for a general audience.

    We thank the reviewer for the positive comments.

    In my opinion, the authors should define which could be the limits of their method, since no data on the possible weaknesses are reported.

    RESPONSE: We have discussed the limitations of the methods as well. The discussion has been modified and the following sentences have been now included in the revised manuscript.

    “However, one of the major disadvantages of using fission yeast could be the need to use much higher concentrations of drugs than normally used for mammalian cell cultures to achieve an inhibitory effect. This could probably be due to the poor permeability of certain drugs in fission yeast because of its thick cell wall (Benko et al. 2017; Pérez and Ribas 2004). A similar effect of toxicity might arise at much lower concentrations in other eukaryotic cells, such as human cells. Consistently, while sanguinarine and berberine are known to affect the eukaryotic microtubules at 10 μΜ – 20 μM concentrations (Lopus and Panda 2006; Wang et al. 2016; Raghav et al. 2017), morphological effects on yeast cells were observed only at concentrations > 100 μM. However, yeast microtubules were not affected by berberine and sanguinarine. Differences in membrane lipid profiles and MDR efflux pumps between yeasts and mammalian cells might also contribute to differential resistance to the drugs being tested (Balzi and Goffeau 1991). Conversely, an inhibitory effect in yeast cells may not necessarily translate into toxicity in a human cell. These and the permeability of drugs in yeast cells represent an important caveat in using such heterologous expression systems for the screening of compounds against target molecules.”

    [Lines 498-513]

    As suggested in the later sections, we have also elaborated on the pros and cons of various methods including the yeast-based screening methods. [Lines 462-523]

    No additional experiments are required to support the claims.

    The suggested experiments could be quite easy to be realized for authors working in the microbiological field, and familiar with protein expression and purification, as well as bacteria and yeast growth.

    From my side, even if I am not so expert in microbiology and plasmid/protein purification, the methods presented could be reproduced with no significant doubt.

    Statistical analysis was done and seems to be adequate.

    RESPONSE: We thank the reviewer for these encouraging comments.

    Minor comments: • Prior studies should be deepened, especially for the state of art authors referred to. Additional paper, both reviews on the possible methods for evaluating FtsZ inhibition, as well as research papers on FtsZ inhibitors targeting E. coli and other Gram negative strains should be mentioned, since, in my opinion, these could move authors in changing a little bit the overall text of the manuscript.

    RESPONSE: We have now elaborated the state-of-art methods used for evaluation of FtsZ inhibition and cited the relevant papers and reviews. We have also included papers on development of FtsZ inhibitors, especially the ones similar to PC190723, targeting Gram-negative bacteria. The following sentences have been included in the revised manuscript.

    “Several approaches have been used to screen small molecules targeting bacterial cell division and FtsZ. While in vitro methods such as NMR (Domadia et al. 2007; Sun et al. 2014; Araújo‑Bazán et al. 2019) and crystallography (Läppchen et al. 2008; Fujita et al. 2017) are valuable and offer information on distinct binding sites, these are not efficient for screening. Electron microscopic examination can distinguish the effects of the compounds being tested on the FtsZ protofilament assembly and lateral associations (Nova et al. 2007; Kaul et al. 2012; Anderson et al. 2012; Sun et al. 2014; Huecas et al. 2017; Kumar et al. 2011; Park et al. 2014). Other techniques that are routinely used include fluorescence anisotropy (Ruiz‑Avila et al. 2013; Park et al. 2014), 90º light-scattering assay (Mukherjee and Lutkenhaus 1999) and dynamic light scattering (Hou et al. 2012; Di Somma et al. 2020) for assessing inhibition of FtsZ assembly (Kaul et al. 2012; Nova et al. 2007; Lui et al. 2019; Anderson et al. 2012, (Irwin et al. 2015). Other easily scalable high-throughput assays include FCS/FCCS and FRET-based methods (Hernández‑Rocamora et al. 2015; Mikuni et al. 2015; Reija et al. 2011).

    In vivo assays relying on cell filamentation phenotype coupled with the localization of Z-ring might be a good indicator of FtsZ being the direct target. However, since bacteria can undergo cell filamentation and not assemble FtsZ rings in response to a variety of conditions, including DNA damage (Mukherjee et al. 1998) and disruption of membrane potential (Strahl and Hamoen 2010), the in vivo assay is not so useful unless combined with the in vitro assays mentioned above. Finally, the isolation of resistance mutants in FtsZ to the drug can provide strong evidence of FtsZ being the direct target.

    Reconstitution systems are powerful and provide excellent control over the system, but they are emerging technologies and are technically challenging. Reconstitution systems include a variety of methods, such as the use of membrane nanodiscs, microbeads of different materials, supported bi-layer membranes (SLBs) and biomimetic systems that provide cell-like environments (Monterroso et al. 2013; Rivas et al. 2014).”

    [Lines 462-487]

    “Several compounds have been evaluated for their activity against FtsZ from both Gram-positive bacteria and Gram-negative bacteria. Although many exhibited only weak activity in vivo against Gram-negative bacteria, derivatives could be promising. These include benzamides (Haydon et al. 2008; Adams et al. 2011; Straniero et al. 2017, 2020a), trisubstituted benzimidazoles (Kumar et al. 2011), 4-bromo-1H-indazole derivatives (Wang et al. 2015), cinnamaldehyde and its derivatives (Domadia et al. 2007; Li et al. 2015), curcumin (Rai et al. 2008), heterocyclic molecules like guanidinomethyl biaryl compounds (Kaul et al. 2012), pyrimidine-quinuclidine scaffolds (Chan et al. 2013), 3-phenyl substituted 6,7-dimethoxyisoquinoline (Kelley et al. 2012), thiazole orange derivatives (Sun et al. 2017), viriditoxin (Wang et al. 2003), N-heterocycles such as zantrins and derivatives (Margalit et al. 2004; Nepomuceno et al. 2015).”

    [Lines 69-80]

    “Several efforts have been made to target Gram-negative bacteria with derivatives of benzamide. Examples include difluorobenzamides, substituted benzodioxanes, heterocyclic and non-heterocyclic derivatives (Straniero et al. 2017; Chai et al. 2020; Straniero et al. 2020a, 2020b). Although many exhibited promising activity in vitro, most were substrates for the AcrAB class of efflux pumps (Chai et al. 2020; Kaul et al. 2014; Straniero et al. 2020a, 2020b; Casiraghi et al. 2020). Thus, the poor membrane permeability, signature outer membrane, particularly lipopolysaccharide (LPS) structure (Wang et al. 2021), the presence of multiple efflux pumps in species such as E. coli, Klebsiella pneumonia and Pseudomonas aeruginosa (Piddock 2006), and differences in FtsZ sequences in the binding-site (Kaul et al. 2013b; Miguel et al. 2015) have been cited as reasons for lack of susceptibility of Gram-negative bacteria to benzamide derivatives (Casiraghi et al. 2020). More recently, two molecules, TXA6101 and TXY6129, with substituted 2,6-difluorobenzamide scaffold, have been shown to inhibit the polymerization of both E. coli and Klebsiella pneumoniae FtsZ. Moreover, despite being substrates for efflux pumps, TXA6101 induced morphological changes in K. pneumoniae (Rosado‑Lugo et al. 2022). Studies in the past on the effects of PC190723 on E. coli have been confusing, with a few reports suggesting an effect on FtsZ polymerization resulting in cell filamentation (Kaul et al. 2014), while others did not find any effect on EcFtsZ (Andreu et al. 2010; Anderson et al. 2012; Khare et al. 2019)⁠. The outer membrane has been shown to be a permeability barrier for PC190723 in E. coli (Khare et al. 2019; Chai et al. 2020). In addition, the Resistance-Nodulation-Division (RND) family of efflux pumps has been attributed to resistance against 2,6-difluorobenzamide derivatives, including TX436 (a prodrug of PC190723) in Gram-negative bacteria (Kaul et al. 2014).”

    [Lines 527-550]

    The whole text requires a deep check for grammar and word choice. Some sentences should be re-written since it is not so easy to understand their meaning. Figures are clear, even if I am not so convinced on the need of including Figure 1.

    RESPONSE: We have now deleted Figure 1 and 2 (as also suggested by Reviewer #2), revised the manuscript and have re-written certain long sentences. We have used Grammarly to check for grammatical errors. We hope the manuscript is easier to follow with these changes.

    Reviewer #1 (Significance (Required)):

    • In my opinion, the outcome coming from this work could move researchers in evaluating an alternative method for assessing FtsZ inhibition. Nevertheless, the actual state of art, a few reviews of the last years confirm this, already underlined a huge number of possible assays, both microbiological, biochemical, biophysical, physiological, or other. As a result, the authors did not result in convincing me about the importance of their methods, when compared to others. They may include some other possible assays and comment of the differences, pros and cons.

    RESPONSE: Several alternative methods have been evaluated and several excellent reviews published in the recent past have underlined the importance of these multiple methods to screen and validate small molecules targeting FtsZ. As suggested by the reviewer here and above, we have now discussed these methods including the yeast-based assay we describe, their advantages and limitations in the revised manuscript.

    The following lines have now been included in Introduction.

    “Several methods have been used to ascertain FtsZ as the target of the drug, and the various approaches have been reviewed in detail by many (Kusuma et al. 2019; Silber et al. 2020; Zorrilla et al. 2021; Andreu et al. 2022). Andreu et al. (2022) have recently proposed a streamlined experimental protocol for the screening and characterization of FtsZ inhibitors.”

    Introduction – [Lines 113-117]

    The following paragraphs, including ones as mentioned above have included in the discussion sections of the revised manuscript.

    “Several approaches have been used to screen small molecules targeting bacterial cell division and FtsZ. While in vitro methods such as NMR (Domadia et al. 2007; Sun et al. 2014; Araújo‑Bazán et al. 2019) and crystallography (Läppchen et al. 2008; Fujita et al. 2017) are valuable and offer information on distinct binding sites, these are not efficient for screening. Electron microscopic examination can distinguish the effects of the compounds being tested on the FtsZ protofilament assembly and lateral associations (Nova et al. 2007; Kaul et al. 2012; Anderson et al. 2012; Sun et al. 2014; Huecas et al. 2017; Kumar et al. 2011; Park et al. 2014). Other techniques that are routinely used include fluorescence anisotropy (Ruiz‑Avila et al. 2013; Park et al. 2014), 90º light-scattering assay (Mukherjee and Lutkenhaus 1999) and dynamic light scattering (Hou et al. 2012; Di Somma et al. 2020) for assessing inhibition of FtsZ assembly (Kaul et al. 2012; Nova et al. 2007; Lui et al. 2019; Anderson et al. 2012, (Irwin et al. 2015). Other easily scalable high-throughput assays include FCS/FCCS and FRET-based methods (Hernández‑Rocamora et al. 2015; Mikuni et al. 2015; Reija et al. 2011).

    In vivo assays relying on cell filamentation phenotype coupled with the localization of Z-ring might be a good indicator of FtsZ being the direct target. However, since bacteria can undergo cell filamentation and not assemble FtsZ rings in response to a variety of conditions, including DNA damage (Mukherjee et al. 1998) and disruption of membrane potential (Strahl and Hamoen 2010), the in vivo assay is not so useful unless combined with the in vitro assays mentioned above. Finally, the isolation of resistance mutants in FtsZ to the drug can provide strong evidence of FtsZ being the direct target.

    Reconstitution systems are powerful and provide excellent control over the system, but they are emerging technologies and are technically challenging. Reconstitution systems include a variety of methods, such as the use of membrane nanodiscs, microbeads of different materials, supported bi-layer membranes (SLBs) and biomimetic systems that provide cell-like environments (Monterroso et al. 2013; Rivas et al. 2014). While in vitro biochemical assays and reconstitution systems are useful to find molecules that directly target FtsZ, they are cumbersome and need to be performed at optimal physiological pH and ionic conditions, which can be considerably variable among FtsZ from different species.

    Our results on the ability of sanguinarine and berberine to specifically affect the assembly of FtsZ and not MreB in fission yeast highlight the utility of the heterologous expression system as a platform to identify molecules that specifically affect FtsZ polymerization. The yeast platform offers a cellular context mimicking the cytoplasm for cytoskeletal assembly. The system is simple to replicate in any laboratory, including those focused on chemical synthesis with minimum microbiological expertise and can be easily reproduced and scaled up as well. However, one of the major disadvantages of using fission yeast could be the need to use much higher concentrations of drugs than normally used for mammalian cell cultures to achieve an inhibitory effect. This could probably be due to the poor permeability of certain drugs in fission yeast because of its thick cell wall (Benko et al. 2017; Pérez and Ribas 2004). A similar effect of toxicity might arise at much lower concentrations in other eukaryotic cells, such as human cells. Consistently, while sanguinarine and berberine are known to affect the eukaryotic microtubules at 10 μΜ – 20 μM concentrations (Lopus and Panda 2006; Wang et al. 2016; Raghav et al. 2017), morphological effects on yeast cells were observed only at concentrations > 100 μM. However, yeast microtubules were not affected by berberine and sanguinarine. Differences in membrane lipid profiles and MDR efflux pumps between yeasts and mammalian cells might also contribute to differential resistance to the drugs being tested (Balzi and Goffeau 1991). Conversely, an inhibitory effect in yeast cells may not necessarily translate into toxicity in a human cell. These and the permeability of drugs in yeast cells represent an important caveat in using such heterologous expression systems for the screening of compounds against target molecules. However, notwithstanding this caveat, the heterologous system provides significant advantages in assessing the direct effects of the drug on FtsZ assembly. Moreover, fission yeast-based high-throughput platform screening methods using imaging have been successfully adapted to the screening of drugs against HIV-1 proteases by large-scale screening facilities such as the NIH Molecular Libraries Probe Production Centers Network in the Molecular Libraries Program, leading to several candidate drugs (Benko et al. 2017, 2019).”

    Discussion - [Lines 462-519]

    “A powerful emerging technique based on cytological profiling has been successfully used to identify the cellular pathways targeted by the inhibitors (Nonejuie et al. 2013; Martin et al. 2020), including cell division inhibition by FtsZ (Araújo‑Bazán et al. 2016). The recent advances in computational image analysis and deep learning approaches (von Chamier et al. 2021; Spahn et al. 2022) could further advance image-based screening for FtsZ inhibitors (Andreu et al. 2022).”

    Discussion – [Lines 581-586]

    As I mentioned before, there are a lot of reviews including the possible tests to perform for assessing FtsZ inhibition. A recent one was not cited, but, from my side, it should be mentioned (10.3390/antibiotics10030254).

    The suggested article is an excellent review that in addition to providing an overview of the state-of-art methods currently in practice for screening drugs targeting FtsZ, also suggests other emerging technologies suitable for assay development. We had cited this article (Zorrilla et al., 2021; doi: 10.3390/antibiotics10030254) in other contexts in our original manuscript but inadvertently missed in the text while mentioning the methods for screening.

    We have now cited Zorrilla et al., 2021 at all appropriate places in the revised manuscript. In addition, we have also cited (Monterroso 2013; https://doi.org/10.1016/j.ymeth.2012.12.014); (Rivas 2014; https://doi.org/10.1016/j.cbpa.2014.07.018); Kusuma 2019 (doi: 10.1021/acsinfecdis.9b00055); Schaffner-Barbero 2012 (doi: 10.1021/cb2003626); Silber et al 2020 (doi: 10.2217/fmb-2019-0348); Li et al., 2015 (doi: 10.1016/j.ejmech.2015.03.026); Casiraghi et al 2020 (doi: 10.3390/antibiotics9020069); Andreu et al., 2022 (10.3390/biomedicines10081825)

    Moreover, I think authors should reconsidered novel research papers, in which researchers evaluated the reason behind the apparent inactivity of benzamide derivatives, similar to PC190723, towards Gram negative strains.

    RESPONSE: Several novel papers that have reported reason for the inactivity of benzamide derivatives towards Gram-negative bacteria, including PC190723 have now been cited. The following sentences have been now included in the revised manuscript.

    “Several efforts have been made to target Gram-negative bacteria with derivatives of benzamide. Examples include difluorobenzamides, substituted benzodioxanes, heterocyclic and non-heterocyclic derivatives (Straniero et al. 2017; Chai et al. 2020; Straniero et al. 2020a, 2020b). Although many exhibited promising activity in vitro, most were substrates for the AcrAB class of efflux pumps (Chai et al. 2020; Kaul et al. 2014; Straniero et al. 2020a, 2020b; Casiraghi et al. 2020). Thus, the poor membrane permeability, signature outer membrane, particularly lipopolysaccharide (LPS) structure (Wang et al. 2021), the presence of multiple efflux pumps in species such as E. coli, Klebsiella pneumonia and Pseudomonas aeruginosa (Piddock 2006), and differences in FtsZ sequences in the binding-site (Kaul et al. 2013b; Miguel et al. 2015) have been cited as reasons for lack of susceptibility of Gram-negative bacteria to benzamide derivatives (Casiraghi et al. 2020). More recently, two molecules, TXA6101 and TXY6129, with substituted 2,6-difluorobenzamide scaffold, have been shown to inhibit the polymerization of both E. coli and Klebsiella pneumoniae FtsZ. Moreover, despite being substrates for efflux pumps, TXA6101 induced morphological changes in K. pneumoniae (Rosado‑Lugo et al. 2022). Studies in the past on the effects of PC190723 on E. coli have been confusing, with a few reports suggesting an effect on FtsZ polymerization resulting in cell filamentation (Kaul et al. 2014), while others did not find any effect on EcFtsZ (Andreu et al. 2010; Anderson et al. 2012; Khare et al. 2019)⁠. The outer membrane has been shown to be a permeability barrier for PC190723 in E. coli (Khare et al. 2019; Chai et al. 2020). In addition, the Resistance-Nodulation-Division (RND) family of efflux pumps has been attributed to resistance against 2,6-difluorobenzamide derivatives, including TX436 (a prodrug of PC190723) in Gram-negative bacteria (Kaul et al. 2014).”

    [Lines 527-550]

    Researchers working on FtsZ inhibitors could be interested in this paper, especially microbiologists.

    I specifically work on the design, synthesis and evaluation of the microbiological assays performed by others on my compounds.

    ========================================================================

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

    Dr. Srinivasin and colleagues previously developed a system where they expressed bacterial FtsZ in yeast and showed that it could assemble into polymers related to the Z rings. Here they develop this system further as a way to assay for drugs that may poison FtsZ, which would be candidates for new antibiotics. They test three drugs against three species of FtsZ. The results suggest that this system should be useful in screening new drugs that may target FtsZ. I would recommend publication after addressing a number of concerns and apparent contradictions.

    Fig. 1 showing chemical formulas of the drugs, and Fig. 2 showing a schematic of the yeast expression system, are probably not needed.

    RESPONSE: Reviewer #1 had also made a similar suggestion and we have now deleted these two figures (Fig. 1 and Fig. 2 in the older version).

    The authors make a point that sanguinarine and berberine inhibit eukaryote cell morphology. In fact, what they show is that they affect yeast cell morphology. This may or may not extend to other eukaryotes. Also, other eukaryotic cells may be more sensitive to drugs than yeast. They should me more conservative in this claim that the system also screens for drugs effects on eukaryotes.

    RESPONSE: We agree with the reviewer’s suggestions here that other eukaryotic cells may be more sensitive to drugs than yeast. We have modified the statements pertaining to these claims in the revised manuscript.

    We have made the following changes in the revised version.

    The title of the manuscript has been now modified as “A salt bridge-mediated resistance mechanism to FtsZ inhibitor PC190723 revealed by a cell-based screen”.

    Lines 23-24 in the abstract has been modified to read as “The strategy also allows for simultaneous assessment of the toxicity of the drugs to eukaryotic yeast cells.”

    Other sentences modified in the revised version are:

    “We find that although sanguinarine and berberine affected FtsZ polymerization, they also affected yeast cell physiology”. [Lines 146-147]

    “In this study, we have attempted to develop a cell-based assay using fission yeast (S. pombe) as a heterologous expression host, which would enable the screening of compounds that could directly affect FtsZ polymerization as well as identify potential toxicity to yeast (or eukaryotic) cells simultaneously”. [Lines 444-447]

    “However, one of the major disadvantages of using fission yeast could be the need to use much higher concentrations of drugs than normally used for mammalian cell cultures to achieve an inhibitory effect. This could probably be due to the poor permeability of certain drugs in fission yeast because of its thick cell wall (Benko et al. 2017; Pérez and Ribas 2004). A similar effect of toxicity might arise at much lower concentrations in other eukaryotic cells, such as human cells”. [Lines 498-503]

    “Conversely, an inhibitory effect in yeast cells may not necessarily translate into toxicity in a human cell. These and the permeability of drugs in yeast cells represent an important caveat in using such heterologous expression systems for the screening of compounds against target molecules”. [Lines 510-513]

    Fig. 3 has some new structural data that should be explored more quantitatively. My quick measurement gave 0.5 and 0.8 µm for the outside diameters of Ec and Sa rings. The spirals of Hp seem to be 0.8 µm outside diameter, similar to SA rings. These spirals may be related to those reported by Popp and by Andreu under certain buffer conditions. This should be explored and referenced.

    RESPONSE: We have now quantitatively measured the diameters of the rings formed by EcFtsZ and SaFtsZ and the diameter and pitch of the spiral polymers of HpFtsZ. These have been now included in the results section and presented as a graph in a new figure (Supplementary Fig. S2). Please also note that the scale bar in Figure 1 (previously Figure 3) was erroneously marked as 5 µm. This has been corrected in the revised version to 2.5 µm.

    Also, the possibility that these spiral polymers may be related to those described by Popp and Andreu have been discussed. We included the following sentences in the discussion.

    “Previous studies have shown that various factors such as molecular crowding, variable C-terminal regions and bound nucleotide state lead to the formation of supramolecular structures like twisted helical structures, toroids and rings similar to those that have been observed in vivo (Popp et al. 2009; Huecas et al. 2017). Thus, the molecular crowding due to the dense cytoplasm of the yeast cells could have possibly induced the spiral and ring-like assembly of FtsZ polymers (Erickson et al. 2010).”

    [Lines 456-461]

    But Fig. 4 presents a contradiction. Here the Hp control cells show long smooth polymers, not helical. This seems an important difference and needs to be addressed. Are the polymers sometimes straight and sometimes helical? After finishing the paper I see that in some experiments the HP is helical, and in others the polymers are straight and smooth. I think it would be important to determine what favors the two forms. If this remains a mystery, at least address it openly.

    RESPONSE: This was definitely an oversight from the authors. We should have clearly mentioned this in the manuscript but completely missed the description of different polymers assembled by HpFtsZ.

    We have now described this clearly in the results and added a new Figure (Supplementary Fig. S1) showing a time course for the appearance of spiral and linear polymers. We have also replaced the images in Figure 5E.

    We have modified the results to read as:

    “Interestingly, HpFtsZ assembled into linear cable-like structures as well as twisted polymers that were curled and spiral in appearance (Fig. 1D). The spiral filaments were more clearly visualized by deconvolution of the images (Fig. 1D iii and 1E). Further, super-resolution imaging using 3D-SIM clearly revealed that HpFtsZ assembles into spiral filaments in fission yeast (Fig. 1F).”

    [Lines 171-175]

    We have also added the following lines in the results section:

    “Spiral polymers appeared early, at 16 – 18 hours after induction of expression (absence of thiamine), and linear cables appeared later at 20 – 22 hours (Fig. S1). The smooth linear polymers possibly arise from lateral association and bundling of FtsZ filaments (Monahan et al. 2009), but the factors determining the two forms in yeast cells remain unclear.”

    [Lines 175-179]

    I am concerned that the quantitation of drug inhibition in Fig 4, 5 is flawed. Visually from 4A it looks like ~90-100% of control cells have polymers, and sang reduces polymers by 70% for Sa and Ec and 100% for Hp: this is based on the number of spots and filaments I see in Fig. 4 Aii. But the quantitation in D shows only 17-23% reduction for all three. These numbers were based on determining the fraction of cells that showed polymer (spots or lines) vs diffuse. It seems that cells are counted as containing polymer even if they had a great reduction in spots or lines, but still had a few. E.g., 4Aii Sa has 4 cells, two of them with no spots, one with only 2, and one with ~7, which totals ~1/3 the spots in control cells. Categorizing cells with only a couple of spots as polymerized, seems to be a poor way to quantitate. Would it not be better to count all spots in all cells, or measure the total length of line polymers, as a measure of inhibition.

    RESPONSE: We agree with the reviewer here that number of spots or the length of the polymers would be a better quantitative measure of the effect of the drugs than the percentage of cells presented. In the revised manuscript, we now present quantified data as suggested.

    We have quantitated the number of spots per cell for SaFtsZ and total polymer length per cell for HpFtsZ to elucidate the effect of drugs on FtsZ polymers. The number of spots per cell were counted using built-in ImageJ macro OPS threshold IJ1 script which combines the otsu thresholding method and analyse particles plugin. The total polymer length per cell in the case HpFtsZ, was measured using used the lpx-plugins as described by Higaki (Higaki et al., 2017).

    In addition, using the lpx-plugins, we also quantify density, a measure of the amount cytoskeleton per unit area in a given cell (Henty-Ridilla et al., 2014; Higaki et al., 2017). We had previously used this measure successfully to quantify assembly of Spiroplasma citri MreB in fission yeast (Pande et al., 2022).

    The methodology has been described in detail in the Materials and Methods section under the heading – “Quantitation of the number of spots, polymer length and density”

    Lines [665-689]

    The new data has been included in the results (lines 207-231 and 275-284) and new Figures (Fig. 2 E, G and Fig. 3 G, H) have been added.

    Fig. 5 makes a convincing case that PC19 accelerates or enhances the polymerization of Sa and Hp. Fig. S2 shows that the structures of polymers are not changed when PC19 is added at 20 hrs, after polymers have already formed. It would have been nice to see for both 5A and S2A that the round spots had holes in the center, when imaged by SIM. Again the quantitation of cells as polymer vs diffuse seems ill suited, because it misses cells with a reduced number of spots.

    RESPONSE: We have imaged the FtsZ polymers of Sa and Hp in the presence of PC190723 using SIM and included these images as new panels in the figures. Figure 3C, 3F and Figure S4 in the revised manuscript.

    Again, for Figure 5 (Fig. 3 in the revised version), we have provided the quantitation as number of spots per cell, polymer length per cell and density (amount of cytoskeleton per unit area) as described above (new Figures - Fig. 3 G, H) in the revised manuscript.

    [Lines 275-284]

    Fig. 6 uses FRAP to show that PC reduces the dynamic exchange of Sa polymers by a factor of 3. It is remarkable to me that rapid exchange is not completely eliminated by PC. Regardless, it would be very important to reference the previous study of Adams..Errington 2011, where they showed the same thing for Foci in Bacillus. PC19 reduced the exchange from 3 to 10 s, but the foci were still very dynamic.

    RESPONSE: We had referenced this work in the original submission in the discussion section – “These results are also consistent with the earlier findings that PC190723 acts to induce FtsZ polymerization and stabilize FtsZ filaments (Andreu et al. 2010; Elsen et al. 2012; Miguel et al. 2015; Fujita et al. 2017) and its derivative compound, 8j acting to slow down FtsZ-ring turnover by 3-fold in B. subtilis (Adams et al. 2011).”

    [Lines 563-567] in revised manuscript

    We have now added the following statement and referenced Adams et al., 2011 in the results section as well.

    “Interestingly, compound 8j, a related benzamide derivative, has been shown to slow down FtsZ-ring turnover by 3-fold in B. subtilis (Adams et al. 2011).”

    [Lines 324-326]

    The analysis of the salt bridge as opposed to a single Arg or His being the cause of resistance to PC19 is an interesting addition to the study. In Fig. 8D some numbers do not agree between the caption and figure (R309/7; S226/7). The whole figure should be carefully checked.

    RESPONSE: We thank the reviewer for pointing to these. We have corrected these errors now in the revised version (Fig. 6).

    I am not familiar with the Gram -ve and Gram +ve nomenclature. Why not simply gram- and gram+?

    RESPONSE: We agree that Gram -ve / +ve are not standard notations and inappropriate.

    We have now written them as Gram-negative and Gram-positive throughout the text.

    The Discussion is quite long largely because it repeats items from Results and Introduction. It is also redundant to hype the value of this system in both Introduction and Discussion; The Introduction should be sufficient. The Discussion should be pared down by eliminating repetition and focusing on relating results to previous literature, in particular items that have not been referenced previously in the paper. Also, I think we don't need the final "In summary" paragraph. That is already nicely presented in the Abstract.

    RESPONSE: We have omitted the repetitive statements from the discussion. We have also deleted the final summary paragraph. We had added new paragraphs [lines 462-519] pertaining to previous literature (also suggested by Reviewer #1) to the discussion section in the revised manuscript.

    The authors should probably provide references to other studies that have used yeast expression to study assembly of FtsZ. I am thinking in particular of papers from the Osteryoung lab looking at chloroplast FtsZ.

    RESPONSE: We have now referenced other papers that have used yeast expression to study assembly of FtsZ.

    The following statement has been added to the introduction:

    “Moreover, the dynamics of chloroplast FtsZs have also been successfully studied using the heterologous fission yeast expression system (TerBush and Osteryoung 2012; Yoshida et al. 2016; TerBush et al. 2018).”

    Lines [132-134]

    NO PAGE NUMBERS. Authors should be penalized a week delay for submitting a mss without page numbers.

    RESPONSE: We sincerely apologise for this gross error and oversight and thank the reviewer for patiently reading through and reviewing a manuscript with no page numbers and line numbers. We are truly sorry for having submitted a manuscript as such and have now included page numbers and line numbers in the manuscript.

    Reviewer #2 (Significance (Required)):

    This work should be of interest to the broad field of research on FtsZ. The authors present it as a new platform for assaying drugs targeting FtsZ, and researchers in this area will certainly be interested. It will also be of broader interest for the novel assay of assembly and exchange dynamics and how they may be modulated by small molecules.

    ========================================================================

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

    Summary: The authors established a proof-of-concept assay to investigate the bacterial cytoskeletal protein FtsZ in fission yeast, and this heterologous yeast system is useful for compounds identification targeting FtsZ. The authors used this system to understand the mechanism of FtsZ's resistance to drug PC190723. Major comments:

    1. From the study, the pombe seems to be a good system for investigating the bacterial cytoskeleton proteins and testing the drugs for them. However, to my knowledge it is not convincing that this is the proper system can be used to assessing the eukaryotic toxicity, since no toxicity to pombe does not mean no toxicity to human cells and vice versa.

    RESPONSE: We agree with the reviewer that toxicity to S. pombe cannot be directly extended to assessing toxicity to other eukaryotic cells such as human cells. As suggested by Reviewer#2 as well, we have modified these claims in the revised manuscript, discussed the possibilities and limited the scope of this work to assessing toxicity in yeast cells.

    We have made the following changes in the revised version.

    The title of the manuscript has been now modified as “A salt bridge-mediated resistance mechanism to FtsZ inhibitor PC190723 revealed by a cell-based screen”.

    Lines 23-24 in the abstract has been modified to read as “The strategy also allows for simultaneous assessment of the toxicity of the drugs to eukaryotic yeast cells.”

    Other sentences modified in the revised version are:

    “We find that although sanguinarine and berberine affected FtsZ polymerization, they also affected yeast cell physiology”. [Lines 146-147]

    “In this study, we have attempted to develop a cell-based assay using fission yeast (S. pombe) as a heterologous expression host, which would enable the screening of compounds that could directly affect FtsZ polymerization as well as identify potential toxicity to yeast (or eukaryotic) cells simultaneously”. [Lines 444-447]

    “However, one of the major disadvantages of using fission yeast could be the need to use much higher concentrations of drugs than normally used for mammalian cell cultures to achieve an inhibitory effect. This could probably be due to the poor permeability of certain drugs in fission yeast because of its thick cell wall (Benko et al. 2017; Pérez and Ribas 2004). A similar effect of toxicity might arise at much lower concentrations in other eukaryotic cells, such as human cells”. [Lines 498-503]

    “Conversely, an inhibitory effect in yeast cells may not necessarily translate into toxicity in a human cell. These and the permeability of drugs in yeast cells represent an important caveat in using such heterologous expression systems for the screening of compounds against target molecules”. [Lines 510-513]

    From figure 4A to 4C, there seems no big difference of cell morphology between control and drug treatment, except for Berberine treatment of SaFtsZ-GFP. Under the low concentration of Sanguinarine (20 µM) and Berberine (53.791 µm), the FtsZ polymerization was disrupted and seems no effect on cell morphology. Why would the authors use much higher Sanguinarine (135.95 µM) and Berberine (134.45 µM) to prove there two drugs are toxic to pombe cells?

    RESPONSE: Earlier reports had shown that sanguinarine and berberine affect mammalian microtubules (Lopus and Panda 2006 - DOI: 10.1111/j.1742-4658.2006.05227.x; Raghav et al., 2017 - DOI: 10.1021/acs.biochem.7b00101). While, we did not observe any growth defect in yeast cells, earlier studies have suggested that yeasts possibly require higher concentrations of certain drugs than used for mammalian cells due to the presence of the cell wall, particularly S. pombe (Perez and Ribas 2004 - https://doi.org/10.1016/j.ymeth.2003.11.020; Benko et al., 2017 - DOI: 10.1186/s13578-016-0131-5). We had thus explored the possibility of cell toxicity to yeast cells at higher concentrations of the drugs.

    The following lines have thus been added to the results section in the revised manuscript.

    “Although we did not observe any growth defect in yeast cells at lower concentrations of the drugs, earlier studies have suggested that yeast cells possibly require higher concentrations of drugs than used for mammalian cells due to the presence of the cell wall, which is particularly thick in S. pombe (Benko et al. 2017; Pérez and Ribas 2004). We thus explored the possibility of cell toxicity to yeast cells at higher concentrations of the drugs.”

    Lines [234-239]

    Sanguinarine and Berberine are FtsZ disruption drugs, do these drugs have effect on microtubule?

    RESPONSE: We have now examined the effect of Sanguinarine and Berberine on yeast microtubules as well and did not find any visible differences between the control and inhibitor (either low or high concentrations) treated cells. This data has been added as a new figure (Supplementary Fig. S3 A and B) in the revised manuscript and the following line added to the results.

    “However, even at higher concentrations, neither of the drugs showed any visible effect on yeast microtubules (Fig. S3 A and B).”

    [Lines 241-242]

    The discussion has been modified as follows:

    “However, one of the major disadvantages of using fission yeast could be the need to use much higher concentrations of drugs than normally used for mammalian cell cultures to achieve an inhibitory effect. This could probably be due to the poor permeability of certain drugs in fission yeast because of its thick cell wall (Benko et al. 2017; Pérez and Ribas 2004). A similar effect of toxicity might arise at much lower concentrations in other eukaryotic cells, such as human cells. Consistently, while sanguinarine and berberine are known to affect the eukaryotic microtubules at 10 μΜ – 20 μM concentrations (Lopus and Panda 2006; Wang et al. 2016; Raghav et al. 2017), morphological effects on yeast cells were observed only at concentrations > 100 μM. However, yeast microtubules were not affected by berberine and sanguinarine. Differences in membrane lipid profiles and MDR efflux pumps between yeasts and mammalian cells might also contribute to differential resistance to the drugs being tested (Balzi and Goffeau 1991). Conversely, an inhibitory effect in yeast cells may not necessarily translate into toxicity in a human cell. These and the permeability of drugs in yeast cells represent an important caveat in using such heterologous expression systems for the screening of compounds against target molecules.”

    [Lines 498-513]

    There are very few SaFtsZ-GFP dot structure in fig 5B, and this is inconsistent with the SaFtsZ-GFP dot structure in fig 4A. Fig 5D has the same issue compare to Fig 4Ci

    RESPONSE: We had probably not made it very clear the experimental differences between Figure 4 and 5 (Figure 2 and 3 in the revised manuscript), which has led to this apparent inconsistency.

    The strong nmt1 promoter (thiamine repressible) takes about 18 hours for full-induction in the absence of thiamine (Forsburg 1993 - https://doi.org/10.1093/nar/21.12.2955). We have utilised the medium strength nmt41 promoter in our studies and hence, in Figure 2, expression of FtsZ-GFP fusions were allowed for longer periods of time (22 – 24 hours) in the experiments concerning sanguinarine and berberine treatments.

    This has been now clearly mentioned in the revised version of the manuscript in the results section (lines 196-199) as well as in figure legends.

    In contrast the very few dot structures or polymers in Figure 3 (revised manuscript) is because of a shorter period of expression of FtsZ-GFP (12 – 14 hours in the absence of thiamine). The shorter period of expression time in these experiments allowed us to test if PC190723 indeed induced the polymerisation of FtsZ, at a stage when the control cells still exhibited diffuse fluorescence and had minimal FtsZ assembly. Thus, the cultures were allowed to express FtsZ for a shorter period of time and imaged in the case of experiments presented in Figure 3.

    This has been now clearly mentioned in the results (lines 259-263) as well as in figure legends in the revised manuscript.

    We hope that we have now made these experimental differences clear and provide more clarity. We have also included this information (hours of induction) in the figure panel.

    The concentration of PC190723 the author used is 20 µg/ml, which is enough for disrupting FtsZ function, however according to the Sanguinarine and Berberine experiments, the author may use higher concentration of PC190723 to assess its toxicity to pombe cells. Same as Sanguinarine and Berberine, does PC190723 has effect on microtubule?

    RESPONSE: As suggested by the reviewer, we have tested the effect of PC190723 at a higher concentration (140.6 µM) similar to that of Sanguinarine and Berberine. We did not find any morphological changes in yeast upon treatment with higher concentrations of PC190723. Also, the drug did not seem to affect the yeast microtubules. These have been now included in the results section and new images have been added in the figure (Supplementary Fig. S3).

    The following lines have been added in the revised manuscript to the results section:

    “Earlier studies had reported that PC190723 was non-toxic to eukaryotic cells, including budding yeast (Haydon et al. 2008). We further tested if PC190723 resulted in morphological defects in S. pombe, like sanguinarine and berberine, at higher concentrations. However, consistent with the earlier reports, PC190723 was inactive against S. pombe at both 56.2 μM and 140.6 μM and did not cause any morphological changes (Fig. 2H iv). Further, PC190723 did not disrupt the yeast microtubules at either of the concentrations (Fig. S3 A iv and B iv).”

    [Lines 294-300]

    The authors mentioned much higher concentrations of drugs than normally used for mammalian cell cultures have to be used for fission yeast. Is there any criterion for this?

    RESPONSE: In the discussion section, we had mentioned that “Much higher concentrations of drugs than normally used for mammalian cell cultures have to be used for fission yeast probably due to permeability issues because of the presence of a thick cell wall (Benko 2017 - DOI: 10.1186/s13578-016-0131-5).

    This has now been mentioned in the results as well in the revised manuscript.

    “Although we did not observe any growth defect in yeast cells at lower concentrations of the drugs, earlier studies have suggested that yeast cells possibly require higher concentrations of drugs than used for mammalian cells due to the presence of the cell wall, which is particularly thick in S. pombe (Benko et al. 2017; Pérez and Ribas 2004). We thus explored the possibility of cell toxicity to yeast cells at higher concentrations of the drugs.”

    [Lines 234-239]

    The following lines in the discussion have been modified in the revised manuscript to read as – “However, one of the major disadvantages of using fission yeast could be the need to use much higher concentrations of drugs than normally used for mammalian cell cultures to achieve an inhibitory effect. This could probably be due to the poor permeability of certain drugs in fission yeast because of its thick cell wall (Benko et al. 2017; Pérez and Ribas 2004). A similar effect of toxicity might arise at much lower concentrations in other eukaryotic cells, such as human cells.”

    [Lines 498-503]

    Minor comments:

    1. There are two units used for drug concentration µM for Sanguinarine and Berberine and µg/ml for PC190723, I think they should be consistent.

    We have now used µM for all drugs.

    Check the units (µM and µg/ml) italic in text and figure legend.

    We have now used µM for all drugs and corrected the italics. We apologise for the erroneous usage of italics in the text for µM.

    Reviewer #3 (Significance (Required)):

    The authors provided a proof-of-concept assay for studying bacterial cytoskeleton proteins in yeast cells. This idea will facilitate people to investigate the bacterial cytoskeleton proteins and also find compounds targeting them without affecting the yeast cells. This study will provide different perspectives to people who study cell biology and secondary metabolites discovery.

    We hope that we have satisfactorily addressed all the concerns raised by the reviewers in the revised manuscript.

    Thanking you,

    With Regards

    Dr. Ramanujam Srinivasan

    Dr. Pananghat Gayathri

  3. Review Commons

    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

    Summary:

    The authors established a proof-of-concept assay to investigate the bacterial cytoskeletal protein FtsZ in fission yeast, and this heterologous yeast system is useful for compounds identification targeting FtsZ. The authors used this system to understand the mechanism of FtsZ's resistance to drug PC190723.

    Major comments:

    1. From the study, the pombe seems to be a good system for investigating the bacterial cytoskeleton proteins and testing the drugs for them. However, to my knowledge it is not convincing that this is the proper system can be used to assessing the eukaryotic toxicity, since no toxicity to pombe does not mean no toxicity to human cells and vice versa.
    2. From figure 4A to 4C, there seems no big difference of cell morphology between control and drug treatment, except for Berberine treatment of SaFtsZ-GFP. Under the low concentration of Sanguinarine (20 µM) and Berberine (53.791 µm), the FtsZ polymerization was disrupted and seems no effect on cell morphology. Why would the authors use much higher Sanguinarine (135.95 µM) and Berberine (134.45 µM) to prove there two drugs are toxic to pombe cells? Sanguinarine and Berberine are FtsZ disruption drugs, do these drugs have effect on microtubule?
    3. There are very few SaFtsZ-GFP dot structure in fig 5B, and this is inconsistent with the SaFtsZ-GFP dot structure in fig 4A. Fig 5D has the same issue compare to Fig 4Ci
    4. The concentration of PC190723 the author used is 20 µg/ml, which is enough for disrupting FtsZ function, however according to the Sanguinarine and Berberine experiments, the author may use higher concentration of PC190723 to assess its toxicity to pombe cells. Same as Sanguinarine and Berberine, does PC190723 has effect on microtubule?
    5. The authors mentioned much higher concentrations of drugs than normally used for mammalian cell cultures have to be used for fission yeast. Is there any criterion for this?

    Minor comments:

    1. There are two units used for drug concentration µM for Sanguinarine and Berberine and µg/ml for PC190723, I think they should be consistent.
    2. Check the units (µM and µg/ml) italic in text and figure legend.

    Significance

    The authors provided a proof-of-concept assay for studying bacterial cytoskeleton proteins in yeast cells. This idea will facilitate people to investigate the bacterial cytoskeleton proteins and also find compounds targeting them without affecting the yeast cells.
    This study will provide different perspectives to people who study cell biology and secondary metabolites discovery.

  4. Review Commons

    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

    Dr. Srinivasin and colleagues previously developed a system where they expressed bacterial FtsZ in yeast and showed that it could assemble into polymers related to the Z rings. Here they develop this system further as a way to assay for drugs that may poison FtsZ, which would be candidates for new antibiotics. They test three drugs against three species of FtsZ. The results suggest that this system should be useful in screening new drugs that may target FtsZ. I would recommend publication after addressing a number of concerns and apparent contradictions.

    Fig. 1 showing chemical formulas of the drugs, and Fig. 2 showing a schematic of the yeast expression system, are probably not needed.

    The authors make a point that sanguinarine and berberine inhibit eukaryote cell morphology. In fact, what they show is that they affect yeast cell morphology. This may or may not extend to other eukaryotes. Also, other eukaryotic cells may be more sensitive to drugs than yeast. They should me more conservative in this claim that the system also screens for drugs effects on eukaryotes.

    Fig. 3 has some new structural data that should be explored more quantitatively. My quick measurement gave 0.5 and 0.8 µm for the outside diameters of Ec and Sa rings. The spirals of Hp seem to be 0.8 µm outside diameter, similar to SA rings. These spirals may be related to those reported by Popp and by Andreu under certain buffer conditions. This should be explored and referenced.

    But Fig. 4 presents a contradiction. Here the Hp control cells show long smooth polymers, not helical. This seems an important difference and needs to be addressed. Are the polymers sometimes straight and sometimes helical? After finishing the paper I see that in some experiments the HP is helical, and in others the polymers are straight and smooth. I think it would be important to determine what favors the two forms. If this remains a mystery, at least address it openly.

    I am concerned that the quantitation of drug inhibition in Fig 4, 5 is flawed. Visually from 4A it looks like ~90-100% of control cells have polymers, and sang reduces polymers by 70% for Sa and Ec and 100% for Hp: this is based on the number of spots and filaments I see in Fig. 4 Aii. But the quantitation in D shows only 17-23% reduction for all three. These numbers were based on determining the fraction of cells that showed polymer (spots or lines) vs diffuse. It seems that cells are counted as containing polymer even if they had a great reduction in spots or lines, but still had a few. E.g., 4Aii Sa has 4 cells, two of them with no spots, one with only 2, and one with ~7, which totals ~1/3 the spots in control cells. Categorizing cells with only a couple of spots as polymerized, seems to be a poor way to quantitate. Would it not be better to count all spots in all cells, or measure the total length of line polymers, as a measure of inhibition.

    Fig. 5 makes a convincing case that PC19 accelerates or enhances the polymerization of Sa and Hp. Fig. S2 shows that the structures of polymers are not changed when PC19 is added at 20 hrs, after polymers have already formed. It would have been nice to see for both 5A and S2A that the round spots had holes in the center, when imaged by SIM. Again the quantitation of cells as polymer vs diffuse seems ill suited, because it misses cells with a reduced number of spots.

    Fig. 6 uses FRAP to show that PC reduces the dynamic exchange of Sa polymers by a factor of 3. It is remarkable to me that rapid exchange is not completely eliminated by PC. Regardless, it would be very important to reference the previous study of Adams..Errington 2011, where they showed the same thing for Foci in Bacillus. PC19 reduced the exchange from 3 to 10 s, but the foci were still very dynamic.

    The analysis of the salt bridge as opposed to a single Arg or His being the cause of resistance to PC19 is an interesting addition to the study. In Fig. 8D some numbers do not agree between the caption and figure (R309/7; S226/7). The whole figure should be carefully checked.

    I am not familiar with the Gram -ve and Gram +ve nomenclature. Why not simply gram- and gram+?

    The Discussion is quite long largely because it repeats items from Results and Introduction. It is also redundant to hype the value of this system in both Introduction and Discussion; The Introduction should be sufficient. The Discussion should be pared down by eliminating repetition and focusing on relating results to previous literature, in particular items that have not been referenced previously in the paper. Also, I think we don't need the final "In summary" paragraph. That is already nicely presented in the Abstract.

    The authors should probably provide references to other studies that have used yeast expression to study assembly of FtsZ. I am thinking in particular of papers from the Osteryoung lab looking at chloroplast FtsZ.

    NO PAGE NUMBERS. Authors should be penalized a week delay for submitting a mss without page numbers.

    Significance

    This work should be of interest to the broad field of research on FtsZ. The authors present it as a new platform for assaying drugs targeting FtsZ, and researchers in this area will certainly be interested. It will also be of broader interest for the novel assay of assembly and exchange dynamics and how they may be modulated by small molecules.

  5. Review Commons

    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

    Summary:

    Srinivasan and co-workers developed an alternative screening method for defining the ability of FtsZ inhibitor to affect FtsZ polymerization. This alternative assay was defined considering the expertise of the authors on the topic, and they use Schizosaccharomyces pombe as a model for studying the effect of PC190723, sanguinarine and berberine on FtsZ assembly. The use of a heterologous expression system is useful for the evaluation of FtsZ coming from different strains, both Gram - and Gram +. The same model could gain insights also on the capability of FtsZ inhibitors to affect eukaryotic cell physiology. Finally, authors resulted also in suggesting a possible cause to suspected resistance to PC190723 from Gram - strains as E. coli.

    Major comments:

    • The conclusions are included in the discussion section and are quite convincing, for a general audience.
    • In my opinion, the authors should define which could be the limits of their method, since no data on the possible weaknesses are reported.
    • No additional experiments are required to support the claims.
    • The suggested experiments could be quite easy to be realized for authors working in the microbiological field, and familiar with protein expression and purification, as well as bacteria and yeast growth.
    • From my side, even if I am not so expert in microbiology and plasmid/protein purification, the methods presented could be reproduced with no significant doubt.
    • Statistical analysis was done and seems to be adequate.

    Minor comments:

    • Prior studies should be deepened, especially for the state of art authors referred to. Additional paper, both reviews on the possible methods for evaluating FtsZ inhibition, as well as research papers on FtsZ inhibitors targeting E. coli and other Gram negative strains should be mentioned, since, in my opinion, these could move authors in changing a little bit the overall text of the manuscript.
    • The whole text requires a deep check for grammar and word choice. Some sentences should be re-written since it is not so easy to understand their meaning. Figures are clear, even if I am not so convinced on the need of including Figure 1.

    Significance

    • In my opinion, the outcome coming from this work could move researchers in evaluating an alternative method for assessing FtsZ inhibition. Nevertheless, the actual state of art, a few reviews of the last years confirm this, already underlined a huge number of possible assays, both microbiological, biochemical, biophysical, physiological, or other. As a result, the authors did not result in convincing me about the importance of their methods, when compared to others. They may include some other possible assays and comment of the differences, pros and cons.
    • As I mentioned before, there are a lot of reviews including the possible tests to perform for assessing FtsZ inhibition. A recent one was not cited, but, from my side, it should be mentioned (10.3390/antibiotics10030254). Moreover, I think authors should reconsidered novel research papers, in which researchers evaluated the reason behind the apparent inactivity of benzamide derivatives, similar to PC190723, towards Gram negative strains.
    • Researchers working on FtsZ inhibitors could be interested in this paper, especially microbiologists.
    • I specifically work on the design, synthesis and evaluation of the microbiological assays performed by others on my compounds.
  6. Version published to 10.1101/2022.04.06.487355v1 on bioRxiv