arg_1
stringlengths 4
5.08k
| round_1
float64 2
8
⌀ | ann_1
float64 1
2
⌀ | arg_2
stringlengths 8
2.19k
| round_2
float64 1
7
⌀ | ann_2
float64 1
2
⌀ | annotation_name
stringclasses 131
values | is_attacks
int64 0
1
|
|---|---|---|---|---|---|---|---|
Interesting point, which we have added to the revised Conclusion and Implications section.
| 2 | 1 |
The authors consider a number of factors associated with long COVID and psychological symptoms. I would suggest another factor - frustration with not getting well. Normally, you get sick, you get better (if you don't die). You get injured moderately and you get better with time. Long COVID breaks that pattern - you get sick and stay sick. Like being sick and tired of being sick and tired. Some uncommon illnesses may act that way, but COVID has become endemic and common. What is your common cold lasted for ten months? People would get really irritated and discouraged by such a situation.
| 1 | 2 |
life12060901_perova
| 1 |
Thank you for the kind response.
| 2 | 1 |
I accept the manuscript for its publication in this journal.
| 1 | 2 |
life12060901_perova
| 1 |
The first mentioned statement has been removed altogether in the revised manuscript. The second statement has been modified in accordance with the reviewer’s view (see revised section 4.3).
| 2 | 1 |
The authors said in the statistic section that “statistical significance was set at p<0.05.” (page 5, line 212). However, later they wrote “The descriptively higher proportion of females with long COVID bordered statistical significance (p=0.05)” page 5, line 230) and “However, the different proportions of men and women with long COVID (22% and 33%, respectively) bordered towards statistical significance […]” (page 10, line 318). As a statistician, I felt very mad about this, because the results speak a different language, this is non-significant result.
| 1 | 2 |
life12060901_perova
| 1 |
Thank you for pointing this out. We have removed the relevant section in accordance with this guidance.
| 2 | 1 |
The authors wrote later in the conclusion section “[…] but a non-significant trend was found for the association between female gender and long COVID.” (page 11, line 389). This is a “no-go” in scientific research, the interpretation of “trends” resulted from non-significant results is strongly misleading. Honestly, there is no empirical evidence that there is an association between female gender and long COVID, there was no significant difference observed between men and women (p=0.05). That is the true story. Hence, I urge the authors not to mislead readers and to re-write these sections.
| 1 | 2 |
life12060901_perova
| 1 |
Thank you for pointing this out, we have revised the sentence (see section 4.1).
| 2 | 1 |
The next major issue targets the interpretation of interaction effects. The authors did three models with three different outcomes, namely psychological distress, fatigue, and perceived stress. Authors introduced an interaction effect long COVID x gender into the model in order to test their third hypothesis (“(iii) whether gender moderated the associations between long COVID status and the health outcomes”, page 3, line 119). In the results section, only two (and not three!) interaction effects became significant (psychological distress: p<0.001, fatigue: p<0.05, and perceived stress: p=0.36, see Table 4, page 7, line 269). However, the authors wrote “While participants with long COVID generally perceived more psychological distress, fatigue, and stress than those without long COVID, differences were larger for men than for women” page 9, line 303), Authors’ response:
| 1 | 2 |
life12060901_perova
| 1 |
We have revised the sentence; see section 4.4.
| 2 | 1 |
“[…] men’s perception of poorer health when having long COVID appears to include higher levels of psychological distress, fatigue and perceived stress” (page 11, line 363),
| 1 | 2 |
life12060901_perova
| 1 |
We have revised the sentence; see section 5.
| 2 | 1 |
and finally “Third, long COVID appears to have a stronger effect on men than on women” (page 11, line 401). Again, the authors are overselling their results. All these statements are not true, only for two out of three outcomes, namely psychological distress and fatigue, but not for perceived stress (see also Figure 3). Hence, I strongly urge the authors to re-write these sentences so that readers are not misled.
| 1 | 2 |
life12060901_perova
| 1 |
Alignment has been fixed.
| 2 | 1 |
Chapter 2 Materials and Methods: Something strange happened to the alignment of heading “2.2 Sample” (page 3, line 128).
| 1 | 2 |
life12060901_perova
| 1 |
Table 1 has been revised accordingly.
| 2 | 1 |
Please fix this. Table 1 (page 3, line 137): I detected an error in the third column (“COVID-19 infection”). The total sample is reported here as 303. However, the single numbers do not sum up to 303 but rather to 310 (13+74+220+3=310). I know, there were 7 missing values within the 310. However, this table is still wrong, it must be 310 and not 303. The percentage of 87 long COVID (28.7%) then refers to 303 (87/303=28.7%). Hence, I urge the authors to re-calculate percentages in Table 1. Otherwise, this is misleading the readership.
| 1 | 2 |
life12060901_perova
| 1 |
Italics have been removed, see revised section 2.3.
| 2 | 1 |
Chapter 2.3 Measures: Sometimes the verbal scales are set in italic (e.g., “0=better than ususal”, page 4, line 173) and sometimes not (e.g., “4=very often”, page 5, line 187).
| 1 | 2 |
life12060901_perova
| 1 |
We are unsure how this problem occurred. We have fixed the problems in the revised tables.
| 2 | 1 |
Table 3, Table 4, and Table 5: The authors use a very inconsistent style of bold. For example, see Table 4 (page 7, line 269): Column Fatgigue, why is ES not printed in bold? Please use a consistent writing style.
| 1 | 2 |
life12060901_perova
| 1 |
Thank you for pointing out, we have addressed this issue throughout the revised manuscript.
| 2 | 1 |
Please use a consistent writing style throughout the whole manuscript. Results (page 7, line 274): Sometimes fatigue (page 7, line 274) and sometimes with a capital F (Fatigue, page 7, line 283).
| 1 | 2 |
life12060901_perova
| 1 |
We have corrected the two tables according to this guidance.
| 4 | 1 |
Before a final publication of this paper, I have two minor points left: Wrong column headings in tables: I detected that the authors replaced “GHQ” with “psychological distress” in the text (also “PSS” replaced with “perceived stress”). This is correct. Psychological distress is the latent construct which is measured with the instrument GHQ. Unfortunately, the authors missed to replace GHQ by psychological distress in Table 4 (page 7, line 280) and Table 5 (page 7, line 297; see column headings: not “GHQ”, “Fatigue”, and “Perceived stress” but rather “Psychological distress”, “Fatigue”, and “Perceived stress”). Please correct the two tables.
| 3 | 2 |
life12060901_perova
| 1 |
Thank you for noticing. We have corrected the manuscript according to this guidance.
| 4 | 1 |
Wrong use of abbreviations: The authors introduced the abbreviation “GHQ-12” for the 12-item General Health Questionnaire (see page 4, line 161), but wrote in the text only "GHQ" (e.g., page 5, line 207; page 6, line 256; page 7, line 274). Please replace every single “GHQ” in the text with the correct abbreviation “GHQ-12”. Please repeat this procedure for “PSS-10” (see page 4, line 185), i.e. replace “PSS” with “PSS-10” in the whole text.
| 3 | 2 |
life12060901_perova
| 1 |
In the paper we justify (at some length) the bending modulus we used, which we extrapolated from various sources, and explain why we believe the value in Boal and Ng’s study is likely too low. The reviewer did not question the logic or the sources used to arrive at this value, but nevertheless seemed surprised at the result. The reviewer claims that the parameter falls outside of the physical regime, but doesn’t offer a justification other than that the implied persistence length “seems too large”. However, the large persistence length simply implies that the cyanobacteria are stiff enough that they are largely unaffected by thermal fluctuations. Despite this, the trichomes in our simulations are still quite flexible in practice, as can be seen in Figure 4. Even if the virtual trichomes are somewhat too stiff, we do not feel that this would have a major impact on the results.
| 2 | 1 |
The literature seems to indicate a low value of the bending modulus (alpha) that corresponds to a persistence length of 500 microns. The authors instead use alpha = 2 × 10−21 Nm2, which is 1000× larger. With this value, the persistence length = alpha/(kbT) = 0.5 metres. This seems orders of magnitude too large, especially in the face of the results of Boal and Ng [44]. The authors need to be in a physical regime of parameter space in order to confront their results with observed patterns. Otherwise, they are just fitting a non-linear phenomenon with a (potentially inappropriate) non-linear model— in this case agreement does not provide firm insight into the physical system.
| 1 | 2 |
life4030433_makarova
| 1 |
We added the following statement to the end of Section 3.3: “The simulation results of the previous section for β > 0.5 are unlikely to be affected by the domain size, however, since the features in those simulations are on a much smaller scale than the domain size and are also more chaotic, as seen in the correlation length and the global alignment (Figure 5a,d).”
| 2 | 1 |
In Section 3.3 the authors describe how an increase in system size indicated that one of the results presented earlier in the paper (at beta = 0.5) was an artifact of a smaller system size. This brings into question all results at the smaller system size, potentially including all the quantitative results of the paper. The authors should be able to argue that the other results will not depend on system size. The correlation length in Figure 3b is by eye more than the system size, indicating a potential qualitative change with increasing system size (which the authors find). The correlation length in other figures is less than the system size, indicating that they may be fine. The authors need to add some of this discussion to reassure the reader.
| 1 | 2 |
life4030433_makarova
| 1 |
We added the following clarification to the end of the non-dimensionalization section: “In the subsequent sections, simulations for β = 0 use a modified interaction force such that there is no cohesion, but the hard core repulsion is maintained, i.e., 0 0 c F = for 0 h ≥ , otherwise 0 c c F F = with β = 0.125.” (4)
| 2 | 1 |
The authors find that results at beta = 0 are most similar to the experimental reticulate patterns. In the force model, there are three forces: elastic, gliding, and contact/cohesion. The contact/cohesion force is presented in equation 12, and is proportional to epsilon. After Equation 23, the authors define beta = epsilon/(zeta⋅v0), implying that there will be no contact/cohesion forces when beta = 0. However, in the results of Figures 3, 4, and 5 the results for beta = 0 clearly show interactions between the filaments. The fourth paragraph of section 2.9 describes a hard-core interaction, which appears to be the source of the patterns when beta = 0. This hard-core interaction should be mentioned after Equation 12, since it is not simply an implementation issue but the dominant interaction with beta = 0.
| 1 | 2 |
life4030433_makarova
| 1 |
We added two references to justify using the Lennard–Jones potential.
| 2 | 1 |
All the results depend on the cohesion forces, which are described by a Lennard–Jones function in Equation 12, in Section 2.5. However, this way of approximating cohesion is not well justified, and in Section 2.9 (several sections later) the authors mention that in reality the attractive force “does not really exist” and that the trichomes cohere after contact. The authors are using an LJ interaction to approximate contact adhesion. They should say why, and address how good of an approximation they expect this to be.
| 1 | 2 |
life4030433_makarova
| 1 |
We added the following clarification: “This is simulated by generating a pseudo-random number x following a uniform distribution X ~ U(0,1), and reversing the gliding direction if x < ω⋅∆t.
| 2 | 1 |
In Section 2.4, it is stated that a stochastic process determines P, but no details are given. There should be some description, if brief, for this process. What distribution is used?
| 1 | 2 |
life4030433_makarova
| 1 |
We amended the following sentences of the discussion: “These differences may be due to the fact that we use a shallow domain (7.5 microns) to reduce the simulation run time, which may cause the pattern to be “squashed” as the domain ceiling prevents ridges from growing vertically” “Restricting the virtual trichomes such that they may only glide when in contact with the substratum or another trichome may promote further aggregation of the trichomes from streams into ridges, as a trichome would be less likely to successfully break away from a stream because it would lose much of its propulsive force as it lost contact with neighboring trichomes and/or the substratum.”
| 2 | 1 |
The authors mention in the discussion that they may get stronger results for a deeper system, or for trichomes that may only glide freely near a surface or other trichome. In terms of the understanding gained from the modelling/simulations, how would these lead to qualitative changes in the observed patterns?
| 1 | 2 |
life4030433_makarova
| 1 |
We added the following sentence to Section 2.4: “Gliding requires contact with some (semi-)solid substrate in order to provide a reaction force to the gliding mechanism. We assume that the trichomes are immersed in highly viscous medium that allows them to glide freely in all directions. This medium could consist of the EPS the trichomes produce copiusly when gliding [Hoiczyk2000].” (8)
| 2 | 1 |
The viscosity mu = 1 Pa s is used. This value is approximately 1000 times higher than the viscosity of water, which is the environment of filamentous cyanobacteria. A reference is given, but there should also be a brief justification in the text.
| 1 | 2 |
life4030433_makarova
| 1 |
The exact force is now given.
| 2 | 1 |
The authors do not include a description of the reaction force for the top and bottom planes. A brief explanation would be appropriate.
| 1 | 2 |
life4030433_makarova
| 1 |
The bibliography now has volume and page numbers.
| 2 | 1 |
None of the references at the end of the paper contain journal volume or page numbers. They should.
| 1 | 2 |
life4030433_makarova
| 1 |
The reviewer did not provide any examples where he/she felt a reference was warranted. We cited 55 sources, which we feel offers reasonable background and justification for our assumptions.
| 2 | 1 |
In general, and especially in the introduction, many statements are made without accompanying references.
| 1 | 2 |
life4030433_makarova
| 1 |
References to the movies were added to the results section.
| 2 | 1 |
There are supplemental movies attached. These are not referred to in the text.
| 1 | 2 |
life4030433_makarova
| 1 |
We added the following sentence to the end of Section 2.8: “This value corresponds to a persistence length of 0.49 m, which implies that the trichomes would be practically unaffected by thermal fluctuations.”
| 4 | 1 |
At the end of the discussion of the bending modulus (in Section 2.8, on page 11), the implied persistence length and the irrelevance of thermal fluctuations should be explicitly mentioned.
| 3 | 2 |
life4030433_makarova
| 1 |
We added the following paragraph to the discussion: “Many of the model parameters, such as trichome length, diameter, gliding speed, reversal frequency, etc. are easily measured and are well known. However, the bending modulus is more difficult to measure directly. We attempted to use the bending modulus implied from the relation α= kBT ε and Boal and Ng’s (2010, p. 4625) measurements of the trichome persistence length, but the resulting value seemed too low compared to measurements of other bacteria and in practice the virtual trichomes appeared flaccid during simulations. It is possible that the flexure seen in Boal and Ng’s trichomes was due more to the motility of the trichomes than random thermal fluctuations, in which case the above relation would no longer be valid, and a more complex model would be required to associate the observed geometry of the trichomes to their bending modulus. For example, Wolgemuth (2005) used an elastic model to estimate the bending modulus of M. xanthus by fitting the model to the flailing motions of a Myxobacterium stuck at one end.”
| 4 | 1 |
In Section 4, the discussion section of the paper, the topic of the bending modulus is returned to for a paragraph. This paragraph needs to explicitly address why the measurements of Boal and Ng were not used, i.e., why Boal and Ng misinterpreted their data. In particular the authors should suggest how the measurements are significant measurements to not use. In particular this paragraph needs to R8 suggest how the measurement should be done to properly test their suggested values for the bending modulus, and so resolve the disagreement with Boal and Ng.
| 3 | 2 |
life4030433_makarova
| 1 |
The symbol used to denote trichome length was changed to an uppercase lambda. The , ⋅⋅ is often used to denote a tuple in computer science, in this case a pair. The superscripts should have been subscripts. The notation used in Equations (5)–(8) was adopted from Bergou et al’s paper. h-hat is in fact a unit vector parallel to h, and we’ve clarified the text to this point.
| 2 | 1 |
• ˆ h looks like a unit vector parallel to h. Is there a reason that h only has a ˆ ⋅ in Equation (12)?
| 1 | 2 |
life4030433_makarova
| 1 |
We added a few sentences to the introduction that expand a bit on similar structures in the Petroff and Walter references.
| 2 | 1 |
Walter (ref 7) describes the formation of “clumps” which are very similar to the structures describes by Shepard. Shepard and Sumner provide better images (Walter only shows sketches), but it might be worth mentioning that the phenomenon is more general. As I recall, Petroff et al. (ref 13) show similar patterns.
| 1 | 2 |
life4030433_makarova
| 1 |
Actually, as Reviewer 1 noted, we use a viscosity that is 1000 times greater than that of water (following Wolgemuth et al. 2005), for the very reasons the reviewer mentioned.
| 2 | 1 |
In the force balance, the drag coeffcient is assumed to be the viscosity of water. Since the cells glide over a surface and next to other cells, it would seem that cell–cell friction and cell–surface friction might be at least as important as hydrodynamic drag. What about the viscosity of the material used in the “slime jet”?
| 1 | 2 |
life4030433_makarova
| 1 |
The statement has been clarified and is now: “The sum of the interaction’s opposing forces and torques is zero, thereby ensuring Newton's third law is respected.”
| 2 | 1 |
Following Equation (13), it is stated that “The interpolation ensures the net force and net torque of the interaction forces are null”. This confuses me so I suspect that I have misunderstood something. In an overdamped system, such as this one, all forces and torques are balanced by drag on the cell. Is this all that is happening in this derivation? The text reads like certain forces and torques are assumed to vanish. Are these forces and torques introduced by the interpolation and the parameterization is chosen to cancel artifacts? Or, are there internal forces and and torques acting on the vertices that cancel as a result of Newton’s second law? Please clarify this point.
| 1 | 2 |
life4030433_makarova
| 1 |
Our estimate is simplistic, however we were unable to find a reliable measurement of this parameter in the literature, and so we resorted to some simple extrapolations.
| 2 | 1 |
In the estimation of the Young’s modulus it is assumed that the bending resistance is due to the cell wall. I suspect that this is a very poor approximation, accounting for the two order of magnitude difference between the estimated and benchmark values of Y. It would seem that the resistance is controlled or at least influenced by additional factors such as protein expression (notably MreB) and osmotic pressure. If my intuition is wrong (I am not an expert on cell morphology), I would appreciate knowing why.
| 1 | 2 |
life4030433_makarova
| 1 |
There are indeed additional parameters. We have added a complete list to Section 2.8. Many are explored to some extent in the results (beta, reversal frequency, domain dimensions, trichome R7 density), whereas most of the physical characteristics of the trichomes are known from published sources (diameter, length, gliding speed, bending modulus), and finally N was simply chosen so that trichomes had a “smooth” appearance.
| 2 | 1 |
If I understand correctly, there are six important length scales in the problem, the size of the domain DH , the spacing between trichomes 13 ρ − , 0 , , L l αν ζ = Θ, and the trichome length L = Nl (also denoted L in the manuscript), giving five dimensionless numbers. However, in the non-dimensionalization, the authors state that there are only two parameters: ρ and Nl γ θ = . Unless I am mistaken, there are additional parameters like l/L, Nl/D, and N that the authors hold fixed but may influence the details of the numerics.
| 1 | 2 |
life4030433_makarova
| 1 |
To avoid confusion on this point we have replaced this statement with the following: “We find that these errors are infrequent and acceptable as a trade-off for increased computational performance.” We have also added a comment to Section 2.5 regarding the Lennard–Jones interaction: “The attractive force has a short range, and is practically zero at h = 2θ.”
| 2 | 1 |
It seems a little odd to posit an attractive force in the derivation (Equation (12)) and then ignore/dismiss it in the numerics (“… in reality the attractive force does not really exist, trichomes only cohere after they have come into contact” page 12). Clearly, this is done to (quite reasonably) avoid modeling steric interactions with hard bodies. It was a little distracting to think that sticky adhesion was decaying like 6 ( ) h Θ . Perhaps note after Equation (12) that the attractive force decreases to zero at a finite distance in the numerics.
| 1 | 2 |
life4030433_makarova
| 1 |
It would be interesting to explore the approach the reviewer suggested; we have not yet explored more coarse grained models.
| 2 | 1 |
Can you coarse grain this model to make it analytically tractable? The basic phenomena suggest something like an active shear-thinning uid? This would make it easier to understand the bifurcations in the dynamics and may require fewer parameters. Just something to think about.
| 1 | 2 |
life4030433_makarova
| 1 |
Shepard et al. claim that the formation of the reticulate pattern occurs on a faster scale than cell growth and division, and so we did not consider this in our model. Including this might make the results more robust, however, and we added a comment to the discussion.
| 2 | 1 |
Notably, the introduction of reproduction does not introduce another parameter. As I recall, these cyanobacteria reproduce (with doubling time τ ) by adding cells to the filament and that filaments periodically break (rate k+). Thus, the length of a filament is not a model parameter. Rather, length is determined dynamically as ~ N k τ + (10)
| 1 | 2 |
life4030433_makarova
| 1 |
We had plans to include a quantitative comparison, however we were not able obtain a good dataset for doing so. It would be interesting to replicate Shepard’s experiments taking care to collect enough good quality images/movies for a quantitative analysis on the patterns.
| 2 | 1 |
This project would benefit immeasurably from a quantitative comparison to experiments. I understand that this is beyond the scope of this paper, but I strongly encourage future work in this direction.
| 1 | 2 |
life4030433_makarova
| 1 |
Replicating Shepard and Sumner’s results was our primary objective for this paper, as we state in the discussion. In future work we would like to increase the scale of the simulations by adding more trichomes and a deeper domain to (hopefully) get more robust reticulate formation. We would also like to do a quantitative analysis on experimental data to in order to fit the model to the data. Finally, we would like to see whether this same minimal system plus photomovement is sufficient to produce the cone-shaped structures documented by Walter, Petroff and others. If successful, we could then link macroscopic features of similar stromatolites back to the parameters of the trichomes, as well as understand the contribution of each factor in the model to building these structures. We have added this explanation to the conclusions section.
| 2 | 1 |
In the introduction, this project is put into the context of stromatolites. In the discussion/conclusion, I would appreciate to hear the authors thoughts on what this model teaches us about stromatolites. In the current form I am left to conclude only that communities of filamentous bacteria can develop reticulated patterns similar to forms observed in the fossil record. This conclusion does not offer much more than Shepard’s experimental observation of this result. Can we use the details of this model to, for example, estimate the cohesion β of ancient cells? Do these results provide any insight into the identity of stromatolite building cells. As far as I can tell, nothing in this model suggests that photosynthesis (much less oxygenic photosynthesis) is required to form these patterns. It seems that this model shows that polygonal patters are a generic feature of elongated gliding cells and is independent of metabolism. A quick google search produced this video of mat of the sulfur-oxidizing bacteria Beggiatoa spp. http://vimeo.com/57205513 (note: I have no idea of the details of these observations so any similarity may be spurious).
| 1 | 2 |
life4030433_makarova
| 1 |
We initially attempted to use the bending modulus implied from Boal and Ng’s study, however we found that during simulations this value seemed to be too low. The trichomes were completely flaccid and would collapse into a heap as soon as one collided with another trichome, unlike the smooth sinuous shape filamentous cyanobacteria often display. In the paper we justify (at some length) the bending modulus we used, which we extrapolated from various sources, and explain why we believe the value in Boal and Ng’s study is likely too low. The reviewer did not question the logic or the sources used to arrive at this value, but nevertheless seemed surprised at the result. The reviewer claims that the parameter falls outside of the physical regime, but doesn’t offer a justification other than that the implied persistence length “seems too large”. However, the large persistence length simply implies that the cyanobacteria are stiff enough that they are largely unaffected by thermal fluctuations. Despite this, the trichomes in our simulations are still quite flexible in practice, as can be seen in Figure 4. Even if the virtual trichomes are somewhat too stiff, we do not feel that this would have a major impact on the results.
| 2 | 1 |
The literature seems to indicate a low value of the bending modulus (alpha) that corresponds to a persistence length of 500 microns. The authors instead use alpha = 2 × 10−21 Nm2, which is 1000× larger. With this value, the persistence length = alpha/(kbT) = 0.5 metres. This seems orders of magnitude too large, especially in the face of the results of Boal and Ng [44]. The authors need to be in a physical regime of parameter space in order to confront their results with observed patterns. Otherwise, they are just fitting a non-linear phenomenon with a (potentially inappropriate) non-linear model— in this case agreement does not provide firm insight into the physical system.
| 1 | 2 |
life4030433_perova
| 1 |
We added the following statement to the end of Section 3.3: “The simulation results of the previous section for β > 0.5 are unlikely to be affected by the domain size, however, since the features in those simulations are on a much smaller scale than the domain size and are also more chaotic, as seen in the correlation length and the global alignment (Figure 5a,d).”
| 2 | 1 |
In Section 3.3 the authors describe how an increase in system size indicated that one of the results presented earlier in the paper (at beta = 0.5) was an artifact of a smaller system size. This brings into question all results at the smaller system size, potentially including all the quantitative results of the paper. The authors should be able to argue that the other results will not depend on system size. The correlation length in Figure 3b is by eye more than the system size, indicating a potential qualitative change with increasing system size (which the authors find). The correlation length in other figures is less than the system size, indicating that they may be fine. The authors need to add some of this discussion to reassure the reader.
| 1 | 2 |
life4030433_perova
| 1 |
We added the following clarification to the end of the non-dimensionalization section: “In the subsequent sections, simulations for β = 0 use a modified interaction force such that there is no cohesion, but the hard core repulsion is maintained, i.e., 0 0 c F = for 0 h ≥ , otherwise 0 c c F F = with β = 0.125.”
| 2 | 1 |
The authors find that results at beta = 0 are most similar to the experimental reticulate patterns. In the force model, there are three forces: elastic, gliding, and contact/cohesion. The contact/cohesion force is presented in equation 12, and is proportional to epsilon. After Equation 23, the authors define beta = epsilon/(zeta⋅v0), implying that there will be no contact/cohesion forces when beta = 0. However, in the results of Figures 3, 4, and 5 the results for beta = 0 clearly show interactions between the filaments. The fourth paragraph of section 2.9 describes a hard-core interaction, which appears to be the source of the patterns when beta = 0. This hard-core interaction should be mentioned after Equation 12, since it is not simply an implementation issue but the dominant interaction with beta = 0.
| 1 | 2 |
life4030433_perova
| 1 |
We added two references to justify using the Lennard–Jones potential.
| 2 | 1 |
All the results depend on the cohesion forces, which are described by a Lennard–Jones function in Equation 12, in Section 2.5. However, this way of approximating cohesion is not well justified, and in Section 2.9 (several sections later) the authors mention that in reality the attractive force “does not really exist” and that the trichomes cohere after contact. The authors are using an LJ interaction to approximate contact adhesion. They should say why, and address how good of an approximation they expect this to be.
| 1 | 2 |
life4030433_perova
| 1 |
We added the following clarification: “This is simulated by generating a pseudo-random number x following a uniform distribution X ~ U(0,1), and reversing the gliding direction if x < ω⋅∆t.
| 2 | 1 |
In Section 2.4, it is stated that a stochastic process determines P, but no details are given. There should be some description, if brief, for this process. What distribution is used?
| 1 | 2 |
life4030433_perova
| 1 |
We amended the following sentences of the discussion: “These differences may be due to the fact that we use a shallow domain (7.5 microns) to reduce the simulation run time, which may cause the pattern to be “squashed” as the domain ceiling prevents ridges from growing vertically” “Restricting the virtual trichomes such that they may only glide when in contact with the substratum or another trichome may promote further aggregation of the trichomes from streams into ridges, as a trichome would be less likely to successfully break away from a stream because it would lose much of its propulsive force as it lost contact with neighboring trichomes and/or the substratum.”
| 2 | 1 |
The authors mention in the discussion that they may get stronger results for a deeper system, or for trichomes that may only glide freely near a surface or other trichome. In terms of the understanding gained from the modelling/simulations, how would these lead to qualitative changes in the observed patterns?
| 1 | 2 |
life4030433_perova
| 1 |
We added the following sentence to Section 2.4: “Gliding requires contact with some (semi-)solid substrate in order to provide a reaction force to the gliding mechanism. We assume that the trichomes are immersed in highly viscous medium that allows them to glide freely in all directions. This medium could consist of the EPS the trichomes produce copiusly when gliding [Hoiczyk2000].” (8)
| 2 | 1 |
The viscosity mu = 1 Pa s is used. This value is approximately 1000 times higher than the viscosity of water, which is the environment of filamentous cyanobacteria. A reference is given, but there should also be a brief justification in the text.
| 1 | 2 |
life4030433_perova
| 1 |
The exact force is now given.
| 2 | 1 |
The authors do not include a description of the reaction force for the top and bottom planes. A brief explanation would be appropriate.
| 1 | 2 |
life4030433_perova
| 1 |
The bibliography now has volume and page numbers.
| 2 | 1 |
None of the references at the end of the paper contain journal volume or page numbers. They should.
| 1 | 2 |
life4030433_perova
| 1 |
The reviewer did not provide any examples where he/she felt a reference was warranted. We cited 55 sources, which we feel offers reasonable background and justification for our assumptions.
| 2 | 1 |
In general, and especially in the introduction, many statements are made without accompanying references.
| 1 | 2 |
life4030433_perova
| 1 |
References to the movies were added to the results section.
| 2 | 1 |
There are supplemental movies attached. These are not referred to in the text.
| 1 | 2 |
life4030433_perova
| 1 |
We added the following sentence to the end of Section 2.8: “This value corresponds to a persistence length of 0.49 m, which implies that the trichomes would be practically unaffected by thermal fluctuations.”
| 4 | 1 |
At the end of the discussion of the bending modulus (in Section 2.8, on page 11), the implied persistence length and the irrelevance of thermal fluctuations should be explicitly mentioned.
| 3 | 2 |
life4030433_perova
| 1 |
We added the following paragraph to the discussion: “Many of the model parameters, such as trichome length, diameter, gliding speed, reversal frequency, etc. are easily measured and are well known. However, the bending modulus is more difficult to measure directly. We attempted to use the bending modulus implied from the relation α= kBT ε and Boal and Ng’s (2010, p. 4625) measurements of the trichome persistence length, but the resulting value seemed too low compared to measurements of other bacteria and in practice the virtual trichomes appeared flaccid during simulations. It is possible that the flexure seen in Boal and Ng’s trichomes was due more to the motility of the trichomes than random thermal fluctuations, in which case the above relation would no longer be valid, and a more complex model would be required to associate the observed geometry of the trichomes to their bending modulus. For example, Wolgemuth (2005) used an elastic model to estimate the bending modulus of M. xanthus by fitting the model to the flailing motions of a Myxobacterium stuck at one end.”
| 4 | 1 |
In Section 4, the discussion section of the paper, the topic of the bending modulus is returned to for a paragraph. This paragraph needs to explicitly address why the measurements of Boal and Ng were not used, i.e., why Boal and Ng misinterpreted their data. In particular the authors should suggest how the measurements are significant measurements to not use. In particular this paragraph needs to R8 suggest how the measurement should be done to properly test their suggested values for the bending modulus, and so resolve the disagreement with Boal and Ng.
| 3 | 2 |
life4030433_perova
| 1 |
The symbol used to denote trichome length was changed to an uppercase lambda. The , ⋅⋅ is often used to denote a tuple in computer science, in this case a pair. The superscripts should have been subscripts. The notation used in Equations (5)–(8) was adopted from Bergou et al’s paper. h-hat is in fact a unit vector parallel to h, and we’ve clarified the text to this point.
| 2 | 1 |
Notation: • L is used both for the length of the trichome Nl and in the non-dimensionalization as 0 αν ζ . • I have never seen , ⋅⋅used to define an edge. It looks like an inner product. Is this standard? • In Equations (5)–(8), is there any difference between subscripts and superscripts? • It is confusing that b κ is a vector that is not simply related to b. I initially thought that κ was a new parameter. Similarly, it is odd that e is a vector and [e] is a matrix. Why not use the Levi–Civita tensor? • ˆ h looks like a unit vector parallel to h. Is there a reason that h only has a ˆ ⋅ in Equation (12)?
| 1 | 2 |
life4030433_perova
| 1 |
We added a few sentences to the introduction that expand a bit on similar structures in the Petroff and Walter references.
| 2 | 1 |
Walter (ref 7) describes the formation of “clumps” which are very similar to the structures describes by Shepard. Shepard and Sumner provide better images (Walter only shows sketches), but it might be worth mentioning that the phenomenon is more general. As I recall, Petroff et al. (ref 13) show similar patterns.
| 1 | 2 |
life4030433_perova
| 1 |
Actually, as Reviewer 1 noted, we use a viscosity that is 1000 times greater than that of water (following Wolgemuth et al. 2005), for the very reasons the reviewer mentioned.
| 2 | 1 |
In the force balance, the drag coeffcient is assumed to be the viscosity of water. Since the cells glide over a surface and next to other cells, it would seem that cell–cell friction and cell–surface friction might be at least as important as hydrodynamic drag. What about the viscosity of the material used in the “slime jet”?
| 1 | 2 |
life4030433_perova
| 1 |
The statement has been clarified and is now: “The sum of the interaction’s opposing forces and torques is zero, thereby ensuring Newton's third law is respected.”
| 2 | 1 |
Following Equation (13), it is stated that “The interpolation ensures the net force and net torque of the interaction forces are null”. This confuses me so I suspect that I have misunderstood something. In an overdamped system, such as this one, all forces and torques are balanced by drag on the cell. Is this all that is happening in this derivation? The text reads like certain forces and torques are assumed to vanish. Are these forces and torques introduced by the interpolation and the parameterization is chosen to cancel artifacts? Or, are there internal forces and and torques acting on the vertices that cancel as a result of Newton’s second law? Please clarify this point.
| 1 | 2 |
life4030433_perova
| 1 |
Our estimate is simplistic, however we were unable to find a reliable measurement of this parameter in the literature, and so we resorted to some simple extrapolations.
| 2 | 1 |
In the estimation of the Young’s modulus it is assumed that the bending resistance is due to the cell wall. I suspect that this is a very poor approximation, accounting for the two order of magnitude difference between the estimated and benchmark values of Y. It would seem that the resistance is controlled or at least influenced by additional factors such as protein expression (notably MreB) and osmotic pressure. If my intuition is wrong (I am not an expert on cell morphology), I would appreciate knowing why.
| 1 | 2 |
life4030433_perova
| 1 |
There are indeed additional parameters. We have added a complete list to Section 2.8. Many are explored to some extent in the results (beta, reversal frequency, domain dimensions, trichome R7 density), whereas most of the physical characteristics of the trichomes are known from published sources (diameter, length, gliding speed, bending modulus), and finally N was simply chosen so that trichomes had a “smooth” appearance.
| 2 | 1 |
If I understand correctly, there are six important length scales in the problem, the size of the domain DH , the spacing between trichomes 13 ρ − , 0 , , L l αν ζ = Θ, and the trichome length L = Nl (also denoted L in the manuscript), giving five dimensionless numbers. However, in the non-dimensionalization, the authors state that there are only two parameters: ρ and Nl γ θ = . Unless I am mistaken, there are additional parameters like l/L, Nl/D, and N that the authors hold fixed but may influence the details of the numerics.
| 1 | 2 |
life4030433_perova
| 1 |
To avoid confusion on this point we have replaced this statement with the following: “We find that these errors are infrequent and acceptable as a trade-off for increased computational performance.” We have also added a comment to Section 2.5 regarding the Lennard–Jones interaction: “The attractive force has a short range, and is practically zero at h = 2θ.”
| 2 | 1 |
It seems a little odd to posit an attractive force in the derivation (Equation (12)) and then ignore/dismiss it in the numerics (“… in reality the attractive force does not really exist, trichomes only cohere after they have come into contact” page 12). Clearly, this is done to (quite reasonably) avoid modeling steric interactions with hard bodies. It was a little distracting to think that sticky adhesion was decaying like 6 ( ) h Θ . Perhaps note after Equation (12) that the attractive force decreases to zero at a finite distance in the numerics.
| 1 | 2 |
life4030433_perova
| 1 |
It would be interesting to explore the approach the reviewer suggested; we have not yet explored more coarse grained models.
| 2 | 1 |
The parameter space is rather large, and only a small subspace is explored in the numerics. This is fine, but it leaves a lot of interesting questions un-addressed. For example, is the transition from broad streams (Figure 4a) to loops (Figure 4d) continuous or are there bifurcations/phase transitions? Can you coarse grain this model to make it analytically tractable? The basic phenomena suggest something like an active shear-thinning uid? This would make it easier to understand the bifurcations in the dynamics and may require fewer parameters. Just something to think about.
| 1 | 2 |
life4030433_perova
| 1 |
Shepard et al. claim that the formation of the reticulate pattern occurs on a faster scale than cell growth and division, and so we did not consider this in our model. Including this might make the results more robust, however, and we added a comment to the discussion.
| 2 | 1 |
In this model, the density of cells is assumed to be constant. However, the timescale of the simulations is comparable to the doubling time of many bacteria. Because cell density is found to R6 be important in determining the morphology of communities, I suspect that reproduction is important in understanding the patterns. This point might be worth addressing in the text.
| 1 | 2 |
life4030433_perova
| 1 |
We had plans to include a quantitative comparison, however we were not able obtain a good dataset for doing so. It would be interesting to replicate Shepard’s experiments taking care to collect enough good quality images/movies for a quantitative analysis on the patterns.
| 2 | 1 |
This project would benefit immeasurably from a quantitative comparison to experiments. I understand that this is beyond the scope of this paper, but I strongly encourage future work in this direction.
| 1 | 2 |
life4030433_perova
| 1 |
Replicating Shepard and Sumner’s results was our primary objective for this paper, as we state in the discussion. In future work we would like to increase the scale of the simulations by adding more trichomes and a deeper domain to (hopefully) get more robust reticulate formation. We would also like to do a quantitative analysis on experimental data to in order to fit the model to the data. Finally, we would like to see whether this same minimal system plus photomovement is sufficient to produce the cone-shaped structures documented by Walter, Petroff and others. If successful, we could then link macroscopic features of similar stromatolites back to the parameters of the trichomes, as well as understand the contribution of each factor in the model to building these structures. We have added this explanation to the conclusions section.
| 2 | 1 |
In the introduction, this project is put into the context of stromatolites. In the discussion/conclusion, I would appreciate to hear the authors thoughts on what this model teaches us about stromatolites. In the current form I am left to conclude only that communities of filamentous bacteria can develop reticulated patterns similar to forms observed in the fossil record. This conclusion does not offer much more than Shepard’s experimental observation of this result. Can we use the details of this model to, for example, estimate the cohesion β of ancient cells? Do these results provide any insight into the identity of stromatolite building cells. As far as I can tell, nothing in this model suggests that photosynthesis (much less oxygenic photosynthesis) is required to form these patterns. It seems that this model shows that polygonal patters are a generic feature of elongated gliding cells and is independent of metabolism. A quick google search produced this video of mat of the sulfur-oxidizing bacteria Beggiatoa spp. http://vimeo.com/57205513 (note: I have no idea of the details of these observations so any similarity may be spurious).
| 1 | 2 |
life4030433_perova
| 1 |
We are not sure if type IV pili are visible in our EM figures. They are very thin (less than 1 nm) and normally visualized by negative staining. In the published EM figures of Phormidium (conventional and cryo EM), type IV pili have not been noted. The filaments that we see in our EM figures could be components of sheath materials.
| 2 | 1 |
As depicted in some of the electron micrographs in Figure 3, the authors describe the presence of thin filaments within the sheaths surrounding the motile filaments. Is it possible that these are in fact type IV pili? How does the structure compare to other published reports of type IV pili?
| 1 | 2 |
life4040819_makarova
| 1 |
This was corrected.
| 2 | 1 |
In Section 3.5, Analysis of Movement of Filaments, the descriptions were confusing at times. For instance Line 253 states “only in some exceptional cases (the straight line in Figure 6D), net displacement was achieved”. Net displacement presumably refers to any final position after a cycle that differs from the starting position. In other words, net displacement would refer to any value other than zero in Figure 6D. All of the points on this graph appear to indicate net displacement, albeit some less than others. Do the authors mean to state that only in a few instances was a large net displacement achieved? Otherwise the authors’ statement does not appear correct. Perhaps using numerical values instead of statements such as “the average velocity was quite limited” would help to clarify.
| 1 | 2 |
life4040819_makarova
| 1 |
We did not find such an indication in the literature. This point was added in the text.
| 2 | 1 |
The authors observe that the velocity of movement was maximal immediately after a reversal and then declined exponentially until the next reversal. The authors then suggest that this observation is theoretically inconsistent with type IV pili or focal adhesion complexes as the driving force for movement. Are there any available published reports quantifying the motility of organisms using these systems that would corroborate this empirically? In short, has it been demonstrated that bacteria moving by twitching motility or focal adhesion do not show a similar pattern for velocity of movement?
| 1 | 2 |
life4040819_makarova
| 1 |
The expression was modified.
| 2 | 1 |
The authors state that supracellular structures emerge from unordered motion and are therefore not encoded for genetically. While their model for spiral formation does not require a genetic program for controlling motility, it may not always be the case for other supercellular structures. In other words, it may not always be true that supercellular structures are not genetically determined.
| 1 | 2 |
life4040819_makarova
| 1 |
Line 45—This was corrected.
| 2 | 1 |
Line 45—suggest adding “to” before “back-and-forth motion...”
| 1 | 2 |
life4040819_makarova
| 1 |
Lines 268 and 269—These were corrected. We found many similar cases, too.
| 2 | 1 |
Lines 268 and 269—“(D)” and “(E)” should be bolded to be consistent
| 1 | 2 |
life4040819_makarova
| 1 |
Line 287—Two values were added.
| 2 | 1 |
Line 287—appears to be missing e-values. The beginning of this sentence indicates 10 genes and then only gives 8 respective e-values for the 10 genes.
| 1 | 2 |
life4040819_makarova
| 1 |
Line 290—Some description was added.
| 2 | 1 |
Line 290—do orf303 and orf297 have homology to any characterized proteins? Are they conserved hypotheticals? Is there any indication of the function for these?
| 1 | 2 |
life4040819_makarova
| 1 |
Line 415—This was corrected.
| 2 | 1 |
Line 415—“single” appears to be formatted differently than the rest of the text
| 1 | 2 |
life4040819_makarova
| 1 |
The choice of EM technique might have been wrong in visualizing the so-called oscillin fibrils, but was effective in visualizing membrane structures. The absence of “oscillin” was based on the genomic analysis but not on EM images. In fact, “oscillin” is a glycine-rich large protein which is poorly conserved in bacteria. In the reported case of P. uncinatum, oscillin might be important in forming fibrils, but in other organisms, other proteins could function as surface fibrils aligning the flow of slime. We agree that the method of fixation was not good for preserving surface structures such as oscillin fibrils, if present.
| 2 | 1 |
About 20 years ago it was shown that in order to preserve crucial ultrastructural features of cyanobacterial cell walls, the cells have to be processed using cryo-procedures (Envelope structure of four gliding filamentous cyanobacteria. J. Bacteriol. 1995, 177, 2387–2395). The authors use an outdated and very artifact-laden method to process their samples. This choice is responsible for a number of shortcomings that need to be addressed: The lack of certain features such as the S-layer and the oscillin fibrils in the reported pictures are attributable to the choice of preparation, as they are clearly present in two species of the same genus described in the above cited paper. Moreover, these structures would explain the rotation of the cells that now is somewhat “mysteriously” explained by a helical slime ejection that is not supported by the observed arrangement of the nozzles. Therefore the authors should do the following: Response:
| 1 | 2 |
life4040819_makarova
| 1 |
This point was added in Section 3.7, in which oscillin is discussed. In general, pore-inclination and oscillin fibrils are not exclusive.
| 2 | 1 |
(a) Clearly spell out in the text why there are no oscillin fibrils and surface structures visible and compare their results with the published results of the cell wall structure of the species of the same genus prepared using cryo-procedures.
| 1 | 2 |
life4040819_makarova
| 1 |
The quality of Figure 1A was rather bad. The contrast was corrected and a new figure was inserted. In the new Figure 1A, the sheath is quite visible. The quality of EM figure is very low as embedded in an MS word file. We provide a better quality figure as supplemental Figure S1. The use of “sheath” is therefore correct in the paragraph describing the figure. In the model, the word “sliding” was not understood in the sense in which we wanted it to be. This is lateral sliding that provokes curvature of the filament. This point was clarified in the text.
| 2 | 1 |
(b) There is no clear distinction in the text between the “sheath” and the secreted “slime” of the cells, which are two completely different structures both physically and chemically (Structural and biochemical analysis of the sheath of Phormidium uncinatum. J. Bacteriol. 1998, 180, 3923– 3932). Figure 1A shows a filament that lacks the “sheath”, while in other pictures the structure is visible. The reason for this is the following: cyanobacteria of the genus Phormidium build over time a carbohydrate layer on their surface that is physically attached to their cell surface, usually called the sheath. Filaments that are ensheathed are non-motile! In contrast, the slime that is secreted by gliding filaments of the same species is not visible in TEM preparations, even when cryo-preservation methods are used (see Figure 1A of the ms and Envelope structure of four gliding filamentous cyanobacteria. J. Bacteriol. 1995, 177, 2387–2395 and Structural and biochemical analysis of the sheath of Phormidium uncinatum. J. Bacteriol. 1998, 180, 3923– 3932). Therefore, the authors should go through their ms and make a careful distinction between these two structures. For example, in the discussion it sounds as if the “sheath” is preventing the R5 cells from sliding, however this is the slime tube that is secreted that is not physically attached to the cell surface at all.
| 1 | 2 |
life4040819_makarova
| 1 |
What is “the basal body part”? Basal body is an organelle of eukaryotic cell. If this means the connection of junctional pore tube and inner membrane, it is clearly seen in the figure, as well as in old paper by Halfen and Castenholz. A high quality figure is provided as Supplementary Figure S1. The “bulges” in the inner membrane are also visible. In the Line 183, the description was revised.
| 2 | 1 |
(c) Please omit the description of the basal body part of the junctional pores or provide substantially better pictures. Even after careful inspection of your images, I cannot convince myself of seeing these parts. Moreover, in cryo-preserved and freeze-fractured cells there are no such structures visible. That does not mean that they don’t exist, it may only mean that none of the so far used preservation methods reveals these structures.
| 1 | 2 |
life4040819_makarova
| 1 |
In Figure 3D, the junctional pores are seen as an array of tilted tubes. The mechanism of rotation is still a mystery because oscillin itself is not present in the strain KS. This is discussed in Section 3.7 with reservations on our hypothesis of inclined junctional pores. On the other hand, the involvement of fibrils in the rotation and locomotion was proposed a long time ago, whether the fibrils are made of oscillin or other proteins. In this respect, we still have to work a lot to identify the real mechanism of rotation.
| 2 | 1 |
(d) Please omit the description about helical tilted junctional pores. These structures are clearly not arranged in this way and the reasons for rotation are the unfortunately not-preserved surface structures mentioned above.
| 1 | 2 |
life4040819_makarova
| 1 |
This is not a good way of naming headings. A heading should not be a phrase or conclusion, but should be descriptive words.
| 2 | 1 |
Results section headings and figure captions should be phrased as conclusions, supported by the data presented, when appropriate (can be a minor change, for example 3.2 could be titled, “Phormidium colonies form divergent structures,” or larger, for example 3.5 could be titled, “Individual filaments undergo directional reversals, with maximum velocity immediately following a reversal.”
| 1 | 2 |
life4040819_makarova
| 1 |
This was described in Section 3.7.
| 2 | 1 |
In the discussion there are further points that need to be addressed: (a) Please clarify the source of the helical flow of the slime, which is most likely due to the presence of the not-preserved helically arranged cell surface proteins not the arrangement or tilt of the junctional pores.
| 1 | 2 |
life4040819_makarova
| 1 |
This point was corrected.
| 2 | 1 |
(b) In the model that describes the potential activity of the junctional pores (How myxobacteria glide. Curr. Biol. 2002, 12, 369–377) it is not suggested that the nozzle actually “contracts” as mentioned in the ms. The swelling of the slime material in the nozzle fills the nozzle and eventually will generate a counter-force of the nozzle walls of the nozzle that results in the ejection of the slime. Please rephrase the text accordingly.
| 1 | 2 |
life4040819_makarova
| 1 |
There might be a misunderstanding in the first part. The rotation of filament does not drive macroscopic rotation or spiral formation. What we describe in the text was the switch to turn to the left was governed by the filament rotation, but the real formation of a spiral is driven by the locomotion of the filament. The cited paper described the clumping of Anabaena cylindria in a dense culture. The clumping or aggregation in Arthrospira (Ohmori group) is mediated by cAMP, but A. cylindrica might be different. I have been using Anabaena for about 40 years, but I have never seen Anabaena filaments form a spiral on agar plates. Spiral formation and clumping are different phenomena. It is difficult to discuss the relationship (if any) between clumping and rotation.
| 2 | 1 |
(c) When the cells form spirals, there are two components to this process: the formation of spirals and their macroscopic sense of rotation: clock- or counterclockwise. In the discussion it is assumed that the rotation of the filaments is crucial for both of these components. If this would be true, then no spirals should be observed in non-rotating filamentous gliding cyanobacteria. However, this is not the case. Anabaena spec. a non-rotating species can generate spirals (see Figure X; Mucilage secretion and the movements of blue-green-algae. Protoplasm 1968, 65, 223–238). This means that the sense of rotation may only be responsible for the direction of the spiral but not for its emergence in the first place. It would be good if the authors would distinguish between these two phenomena and discuss them accordingly.
| 1 | 2 |
life4040819_makarova
| 1 |
The current paper is focused on Phormidium. We never used Myxobacteria, and we have no idea about the motility in Myxobacteria. To clarify the situation, the mention to Myxobacteria was added in the text.
| 2 | 1 |
(d) In Lines 309–315: References 17 and 22 are accurately described, however these models relate to gliding in Myxobacteria, a distinction not noted in the text. Thus, comparisons between these models (including the focal adhesion model) and the current model are only valid if the mechanism of motility in these microorganisms is the same. The authors should (briefly) explain their reasoning on the validity of comparing these two systems, and the implications of their findings to motility in Myxobacteria.
| 1 | 2 |
life4040819_makarova
| 1 |
Lines 94–95: We have determined genomic sequences of many organisms, but it is not easy to publish the genomes as genome paper. The sequences should be connected by PCR, and annotated. In the current study, we are interested in the genes involved in motility. We annotated the related genes but we would not connect the contigs and annotate all the genes. Nowadays, every researcher can sequence his/her own materials quite easily. It is not necessary to deposit all the raw sequence data.
| 2 | 1 |
Lines 94–95: Is it standard to sequence an entire genome, but only deposit several clusters, or should the entire genome be accessible to allow other researchers access to this source?
| 1 | 2 |
life4040819_makarova
| 1 |
Lines 227–230: This was corrected.
| 2 | 1 |
Lines 227–230: The interpretation of the counterclockwise spirals here is somewhat distracting. It can be left as an observation here, and the model explained later.
| 1 | 2 |
life4040819_makarova
| 1 |
Lines 316–317: We did not notice this point. Polysaccharide chain may not be very long. Many of the products of the hps gene cluster encode glycosyltransferases and pseudopilins, which are, respectively, involved in the synthesis and secretion of the slime. The secretion of slime is likely mediated by some molecular machinery that is otherwise involved in type II secretion/motility machinery. Reference 24 was added.
| 2 | 1 |
Lines 316–317: This model for slime secretion suggests that rather than slime being comprised of a single, long, polysaccharide, it is made of smaller subunits that are secreted in a step-wise fashion. The authors should connect this idea to the finding that pseudopilins appear to be part of the molecular machinery, as the model and the data are coherent on this point.
| 1 | 2 |
life4040819_makarova
| 1 |
We thank the reviewers for giving us helpful comments. This study is our new exploration into a new field, and the comments from different fields help us to revise our manuscript. The text was revised according to the comments. In addition, Figure 3A was replaced by a figure of better contrast. Because the EM figures are difficult to view within the Word document, we provide a high-quality figure as a supplement. Within the text, corrected words are highlighted.
| 2 | 1 |
A few major comments can be found below concerning R2 the possible presence of type IV pili in their EM images, some confusion regarding descriptions of motility analysis and others.
| 1 | 2 |
life4040819_perova
| 1 |
We are not sure if type IV pili are visible in our EM figures. They are very thin (less than 1 nm) and normally visualized by negative staining. In the published EM figures of Phormidium (conventional and cryo EM), type IV pili have not been noted. The filaments that we see in our EM figures could be components of sheath materials.
| 2 | 1 |
As depicted in some of the electron micrographs in Figure 3, the authors describe the presence of thin filaments within the sheaths surrounding the motile filaments. Is it possible that these are in fact type IV pili? How does the structure compare to other published reports of type IV pili?
| 1 | 2 |
life4040819_perova
| 1 |
This was corrected.
| 2 | 1 |
In Section 3.5, Analysis of Movement of Filaments, the descriptions were confusing at times. For instance Line 253 states “only in some exceptional cases (the straight line in Figure 6D), net displacement was achieved”. Net displacement presumably refers to any final position after a cycle that differs from the starting position. In other words, net displacement would refer to any value other than zero in Figure 6D. All of the points on this graph appear to indicate net displacement, albeit some less than others. Do the authors mean to state that only in a few instances was a large net displacement achieved? Otherwise the authors’ statement does not appear correct. Perhaps using numerical values instead of statements such as “the average velocity was quite limited” would help to clarify.
| 1 | 2 |
life4040819_perova
| 1 |
We did not find such an indication in the literature. This point was added in the text.
| 2 | 1 |
The authors observe that the velocity of movement was maximal immediately after a reversal and then declined exponentially until the next reversal. The authors then suggest that this observation is theoretically inconsistent with type IV pili or focal adhesion complexes as the driving force for movement. Are there any available published reports quantifying the motility of organisms using these systems that would corroborate this empirically? In short, has it been demonstrated that bacteria moving by twitching motility or focal adhesion do not show a similar pattern for velocity of movement?
| 1 | 2 |
life4040819_perova
| 1 |
The expression was modified.
| 2 | 1 |
The authors state that supracellular structures emerge from unordered motion and are therefore not encoded for genetically. While their model for spiral formation does not require a genetic program for controlling motility, it may not always be the case for other supercellular structures. In other words, it may not always be true that supercellular structures are not genetically determined.
| 1 | 2 |
life4040819_perova
| 1 |
Line 45—This was corrected.
| 2 | 1 |
Line 45—suggest adding “to” before “back-and-forth motion...”
| 1 | 2 |
life4040819_perova
| 1 |
Lines 268 and 269—These were corrected.
| 2 | 1 |
Lines 268 and 269—“(D)” and “(E)” should be bolded to be consistent
| 1 | 2 |
life4040819_perova
| 1 |
Line 287—Two values were added.
| 2 | 1 |
Line 287—appears to be missing e-values. The beginning of this sentence indicates 10 genes and then only gives 8 respective e-values for the 10 genes.
| 1 | 2 |
life4040819_perova
| 1 |
Line 290—Some description was added.
| 2 | 1 |
Line 290—do orf303 and orf297 have homology to any characterized proteins? Are they conserved hypotheticals? Is there any indication of the function for these?
| 1 | 2 |
life4040819_perova
| 1 |
Line 415—This was corrected.
| 2 | 1 |
Line 415—“single” appears to be formatted differently than the rest of the text
| 1 | 2 |
life4040819_perova
| 1 |
The choice of EM technique might have been wrong in visualizing the so-called oscillin fibrils, but was effective in visualizing membrane structures. The absence of “oscillin” was based on the genomic analysis but not on EM images. In fact, “oscillin” is a glycine-rich large protein which is poorly conserved in bacteria. In the reported case of P. uncinatum, oscillin might be important in forming fibrils, but in other organisms, other proteins could function as surface fibrils aligning the flow of slime. We agree that the method of fixation was not good for preserving surface structures such as oscillin fibrils, if present.
| 2 | 1 |
About 20 years ago it was shown that in order to preserve crucial ultrastructural features of cyanobacterial cell walls, the cells have to be processed using cryo-procedures (Envelope structure of four gliding filamentous cyanobacteria. J. Bacteriol. 1995, 177, 2387–2395). The authors use an outdated and very artifact-laden method to process their samples. This choice is responsible for a number of shortcomings that need to be addressed: The lack of certain features such as the S-layer and the oscillin fibrils in the reported pictures are attributable to the choice of preparation, as they are clearly present in two species of the same genus described in the above cited paper. Moreover, these structures would explain the rotation of the cells that now is somewhat “mysteriously” explained by a helical slime ejection that is not supported by the observed arrangement of the nozzles. Therefore the authors should do the following: Response:
| 1 | 2 |
life4040819_perova
| 1 |
This point was added in Section 3.7, in which oscillin is discussed. In general, pore-inclination and oscillin fibrils are not exclusive.
| 2 | 1 |
Clearly spell out in the text why there are no oscillin fibrils and surface structures visible and compare their results with the published results of the cell wall structure of the species of the same genus prepared using cryo-procedures.
| 1 | 2 |
life4040819_perova
| 1 |
There is no clear distinction in the text between the “sheath” and the secreted “slime” of the cells, which are two completely different structures both physically and chemically (Structural and biochemical analysis of the sheath of Phormidium uncinatum. J. Bacteriol. 1998, 180, 3923– 3932). Figure 1A shows a filament that lacks the “sheath”, while in other pictures the structure is visible. The reason for this is the following: cyanobacteria of the genus Phormidium build over time a carbohydrate layer on their surface that is physically attached to their cell surface, usually called the sheath. Filaments that are ensheathed are non-motile! In contrast, the slime that is secreted by gliding filaments of the same species is not visible in TEM preparations, even when cryo-preservation methods are used (see Figure 1A of the ms and Envelope structure of four gliding filamentous cyanobacteria. J. Bacteriol. 1995, 177, 2387–2395 and Structural and biochemical analysis of the sheath of Phormidium uncinatum. J. Bacteriol. 1998, 180, 3923– 3932). Therefore, the authors should go through their ms and make a careful distinction between these two structures. For example, in the discussion it sounds as if the “sheath” is preventing the R5 cells from sliding, however this is the slime tube that is secreted that is not physically attached to the cell surface at all.
| 2 | 1 |
The quality of Figure 1A was rather bad. The contrast was corrected and a new figure was inserted. In the new Figure 1A, the sheath is quite visible. The quality of EM figure is very low as embedded in an MS word file. We provide a better quality figure as supplemental Figure S1. The use of “sheath” is therefore correct in the paragraph describing the figure. In the model, the word “sliding” was not understood in the sense in which we wanted it to be. This is lateral sliding that provokes curvature of the filament. This point was clarified in the text.
| 1 | 2 |
life4040819_perova
| 1 |
What is “the basal body part”? Basal body is an organelle of eukaryotic cell. If this means the connection of junctional pore tube and inner membrane, it is clearly seen in the figure, as well as in old paper by Halfen and Castenholz. A high quality figure is provided as Supplementary Figure S1. The “bulges” in the inner membrane are also visible. In the Line 183, the description was revised.
| 2 | 1 |
Please omit the description of the basal body part of the junctional pores or provide substantially better pictures. Even after careful inspection of your images, I cannot convince myself of seeing these parts. Moreover, in cryo-preserved and freeze-fractured cells there are no such structures visible. That does not mean that they don’t exist, it may only mean that none of the so far used preservation methods reveals these structures.
| 1 | 2 |
life4040819_perova
| 1 |
In Figure 3D, the junctional pores are seen as an array of tilted tubes. The mechanism of rotation is still a mystery because oscillin itself is not present in the strain KS. This is discussed in Section 3.7 with reservations on our hypothesis of inclined junctional pores. On the other hand, the involvement of fibrils in the rotation and locomotion was proposed a long time ago, whether the fibrils are made of oscillin or other proteins. In this respect, we still have to work a lot to identify the real mechanism of rotation.
| 2 | 1 |
Please omit the description about helical tilted junctional pores. These structures are clearly not arranged in this way and the reasons for rotation are the unfortunately not-preserved surface structures mentioned above.
| 1 | 2 |
life4040819_perova
| 1 |
This is not a good way of naming headings. A heading should not be a phrase or conclusion, but should be descriptive words.
| 2 | 1 |
Results section headings and figure captions should be phrased as conclusions, supported by the data presented, when appropriate (can be a minor change, for example 3.2 could be titled, “Phormidium colonies form divergent structures,” or larger, for example 3.5 could be titled, “Individual filaments undergo directional reversals, with maximum velocity immediately following a reversal.”
| 1 | 2 |
life4040819_perova
| 1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.