Diverse Bacterial Properties Influence Dispersal Along Fungal Networks

Round 1

Reviewer 1 Report

Please see comments in attached file.

Please see comments in attached file.

Comments for author File: Comments.pdf

Author Response

Reviewer comments are in blue, our response is in black
Response to reviewer 1.

Introduction

The introduction starts by mentioning symbiosis of fungal-bacterial interactions (FBIs); however, these are not the only types of interactions that fungi and bacteria participate in. It then moves onto biofilms and then fungal highways. The focus of the introduction, and of the paper in general should be more clearly articulated. What were the main research questions and hypotheses to be addressed? This is unclear at present.

The term “fungal superhighway” is not consistent with literature, and is not used in the seminal article by Kohlmeier et al (Reference 8, Line 43). I would avoid using this terminology; only 1-2 articles have used this term (that I am aware of), more specifically in the context of mycorrhizal fungi (e.g., common mycorrhizal networks; see, https://doi.org/10.1016/j.tplants.2012.06.007). Landmark fungal highways papers are missing, and there is no mention of microfluidics (or the need to use a microfluidic system) in the introduction. These experiments could also be performed in an agar-based format where hyphae traverse an air gap; therefore, what is the advantage of using a microfluidic device?

Line 69-70: “The movement of non-motile bacteria through fungal networks is largely unknown.” There is at least one article on this topic and should be referenced, e.g., https://doi.org/10.1186/s12915-022-01406-z

We thank the reviewer for their insightful comments on the Introduction. We have substantially revised the section to address all points and to more clearly articulate the focus and rationale of our study. We agree that the main research question and hypotheses were not sufficiently clear, and we have revised the Introduction to explicitly state our primary research question: How do different bacterial species utilize a fungal hyphal network for dispersal, and what are the specific molecular mechanisms governing these interactions? We then present two specific hypotheses: first, that fungal hyphal networks facilitate the dispersal of both motile and non-motile bacteria, but the efficiency and mechanism of dispersal are species-dependent; and second, that in some cases, dispersal is dependent on bacterial motility structures, such as flagella, and can be modulated by inter-species quorum sensing [Lines 78 -99]. We also accept the reviewer's critique that the term "fungal superhighway" is not standard literature terminology and have removed it from the manuscript, replacing it with the more conventional "fungal highways" or "fungal networks"[Lines 18, 31, 48, 54, 55, 76, 79, 83, 89, 97, 427, 436, 474, 520, 528].

We disagree with the reviewer on the criticism of using a microfluidic device; it was exactly what we needed for our experiments. Our microfluidic device provided precise environmental control, allowing for the maintenance of a stable, humid microenvironment conducive to hyphal growth; direct, real-time visualization at a high-resolution, single-cell level; and enhanced reproducibility and quantification due to its fixed geometry and physically separated ports.

Finally, we thank the reviewer for bringing a relevant article to our attention and have now cited the paper (https://doi.org/10.1186/s12915-022-01406-z) [Line 77] to support our statement regarding the movement of non-motile bacteria, ensuring our introduction accurately reflects the current state of knowledge. Please note that we said the understanding of movement of non-motile bacteria is largely unknown. That remains true even with this additional citation.

---

Methods

Details surrounding the microfluidic device preparation are not complete; it would not be

possible to make this device from the information provided. There needs to be a thorough

explanation to ensure reproducibility to meet research ethical standards (full transparency).

Lines 119-121: I do not understand the following description:

“Following the protocol established by Guckenberger et al. (2012) [20], a piece of transparency

paper was placed on a hotplate beneath a silicon wafer containing the microfluidic device molds.” The cited reference concerns oxygen plasma treatment only, which is one very specific part of the fabrication process (normally for bonding PDMS to glass/PDMS)

 

We appreciate this comment and agree with the necessity of full transparency of research methods. We would like to apologize for citing an incorrect reference here (Guckenberger et al. (2012)). The design and fabrication methods of the microfluidic device came from a previous collaboration with Dr. David Beebe, a world renown microfluidic researcher, and the citation for their methods was misplaced in our records. We apologize for the mistake and have since consulted with Dr. Beebe’s group and have added his student Jamie Liang and Dr. Beebe to the authorship as they have rewritten this section to accurately reflect the methods. They have confirmed the exact fabrication methods we used and helped revise the method section and citation accordingly. The revised section is as follows:

Line143-168:

2.4* (previously 2.3) Preparation of Microfluidic Device

Microfluidic devices were fabricated using a SYLGARD 184 Silicone Elastomer Base and Curing Agent Kit (Dow, Midland, MI, USA), following the protocols established by Duffy et al. [21]. First, a total of 11 g SYLGARD mixture, in a 10:1 ratio of base to curing agent, was thoroughly mixed and degassed under a vacuum-sealed container for 40 minutes or until all air bubbles were removed. The SYLGARD mixture was then poured onto the silicon wafer master. Two pieces of transparency paper were placed on top of and beneath the master to prevent the SYLGARD mixture from contacting hotplates or other surfaces. A tongue depressor was used to spread the SYLGARD mixture on the master and remove any air bubbles generated during the previous steps. To cure the SYLGARD mixture and ensure uniform thickness, a 13,608 g (30 lb) weight was placed on the upper transparency paper and the entire assembly was heated on a hotplate at 80 °C for 4 hours.

Once cured, the now solid SYLGARD mixture was removed from the wafer and cut into strips of four microfluidic channels using a razor blade. Each individual microfluidic channel comprises an input port, where microbes are introduced, and an output port, connected by a microfluidic channel through which bacteria traverse [Fig 1].  To bond the microfluidic devices in preparation for inoculation, 35 mm glass-bottom petri dishes (Cellvis, D35-20-1.5-N) were used. The microfluidic devices were cleaned of any debris by applying scotch tape to the bottom side of the device. This was the side that made contact with the glass surface of the petri dish. Once the device was completely cleaned, they were placed along with their corresponding petri dish bottom side up in a plasma etcher machine (PlasmaEtch PE-50) for 2 minutes. Upon completion, the devices were carefully removed from the Plasma etcher and placed bottom side down on the glass portion of the petri dish. The devices were gently pressed onto the glass further using a toothpick and allowed to continue bonding to the glass surface for an hour. To ensure sterile conditions, the plates were then UV treated in a biosafety cabinet for 30 minutes

What is “transparency paper” and its purpose?

We apologize for the confusion and have updated the methods to more clearly describe the purpose of using transparency paper. We used the term “transparency paper” to refer to the clear plastic films more commonly called “transparency” in microfluidic fields. The change of terminology was intended as a clearer description of the plastic films for audiences that might be unfamiliar with microfluidics. In our methods, transparency papers were used to prevent the contact of PDMS mixtures with hotplates and the weight that was placed on top of the assembly during PDMS curing. The revised section is as follows:

 

Lines 148-149: …Two pieces of transparency paper were placed on top of and beneath the master to prevent the SYLGARD mixture from contacting hotplates or other surfaces….

Line 121: How are the “microfluidic device moulds” made?

Line 121: Is this an SU 8-coated silicon wafer? Specific details need to be provided, and

appropriate references.

Response: We have added a section describing the fabrication of the moulds as above, which are SU-8 coated silicon wafers used for casting PDMS devices. The added section is as follows:

 

Line 128-140:

 

2.3* Fabrication of Master for Casting Microfluidic Device

            Masters used to repeatedly cast microfluidic devices were fabricated following standard photolithography techniques established by Duffy et al. [21]. Briefly, microfluidic channel designs were made using Computer-Aided Design (CA.D) software and printed onto a piece of transparency paper. The transparency paper containing the designs acts as a photomask for creating positive reliefs of SU-8 (Kayaku, Westborough, MA, USA) on silicon wafers. In this study, two layers of positive reliefs were created sequentially on the same silicon wafer using the photomasks marked as “first layer” and “second layer” (Fig. S1), respectively. The thickness of the two layers was determined by the speed of spin-coating SU-8 on silicon wafer and validated with a digital dial gauge. The thicknesses of the first and second layers were 50 µm and 250 µm, corresponding to the height of the microfluidic channel and the height of the input and output ports, respectively.

Line 121: A CAD figure of the design should be provided.
We thank the reviewer for pointing this out and have included a figure (Fig. S1) for the CAD design.

Line 123-4: “A tongue depressor was used to spread the mixture thinly to ensure proper mold

penetration after curing.”

Please explain the purpose of this. The elastomeric polymer, PDMS, should naturally penetrate

a mold as it is a liquid.

We apologize for the mistake and thank the reviewers for the opportunity to clarify our methods. We agree that PDMS would naturally penetrate the silicon wafer mold and have consulted with Dr. Beebe’s group about the purpose of this step. They have clarified that the tongue presser was used to help spread the PDMS mixture more quickly and remove air bubbles by pushing them to the edges. During the steps of pouring PDMS onto the silicon wafer master and putting a piece of transparency paper on top, air bubbles can be generated and trapped in the devices. By sliding a tongue presser over the top transparency paper and pushing down slightly, air bubbles can be efficiently pushed to the edges of the wafer, where they won’t obscure/interfere with the microfluidic channels. The revised method is as follows:

Lines 134-138: …The SYLGARD mixture was then poured onto the silicon wafer master. Two pieces of transparency paper were placed on top of and beneath the master to prevent the SYLGARD mixture from contacting hotplates or other surfaces. A tongue depressor was used to spread the SYLGARD mixture on the master and remove any air bubbles generated during the previous steps…

Line124: weight should be expressed in g.

We thank the reviewers for pointing this out and have revised the unit:

Line 139: …ensure uniform thickness, a 13,608 g (30 lb) weight was placed on the upper transparency paper…

Line 129-130: Refers to figure 1. However, there are no details about the device height. How

deep are the microchannels, and how has this been characterised?

Please also include a photograph of the setup.

We thank the reviewer for pointing out the omission. The channel height was 50 µm, which was determined by the thickness for the first positive relief layer. We also validated the height by measuring the master thickness using a digital dial gauge after the SU-8 has been developed.

We have revised figure 1 to include the channel height.

Lines 166-168: Hydrophobic recovery of PDMS prevents the filling of microfluidic channels

with aqueous liquid media. Are the authors relying on capillary action to fill the channels? If

yes, how long after plasma bonding is the molten agar medium applied?

We again apologize for citing the wrong reference. Hydrophobic recovery is not performed in this study, and scotch tape was only used before plasma bonding to remove debris on PDMS. The channels were filled by the pressure from pipette dispensing molten agar as well as capillary action. The inner surfaces of the microfluidic channel (including the PDMS surfaces and glass surfaces) were all treated with plasma during the device bonding step, and would remain in a hydrophilic state, facilitating capillary action.

Lines 177-178: The authors use FC-40 oil to prevent dessication; how do the authors ensure

that sterility is maintained (i.e. contamination from the oil)?

We thank the reviewer for this important question regarding sterility. To prevent contamination, the FC-40 oil (Fluorinert™ FC-40, 3M) was sterilized prior to use by being passed through a 0.22 µm syringe filter directly into a sterile centrifuge tube. All subsequent steps involving the oil, including its application to the microfluidic devices, were performed inside a biosafety cabinet to maintain an aseptic environment. This two-step process ensures both the sterility of the oil and the integrity of the experiment.

Results and conclusions

• Result sections 3.1 and 3.2 should be condensed into a single section as 3.1 alone does not present any results.

We agree with the reviewer's assessment. We have condensed Result sections 3.1 and 3.2 into a single, cohesive section as suggested.

Figure 8 and Paragraph 3.8: Do you still retrieve L. monocytogenes Δagr in the outlet with S. aureus? The authors use S. aureus mutants to show the role of quorum sensing. They then use WT S. aureus and L. monocytogenes Δagr to show that L. monocytogenes interferes with S. aureus quorum sensing, preventing it from reaching the outlet. However, they do not show whether the Δagr mutation interferes with L. monocytogenes' ability to cross the channel. This is important to gain a full picture of the interference mechanism between the two species.

We thank the reviewer for this question, as it allows us to clarify the complex, multi-species interactions we observed. The key findings from our experiments are as follows:

First, we established that the fungus A. flavus is essential for dispersal in our system. When either S. aureus or L. monocytogenes (WT or Δagr) are co-cultured individually with the fungus, they are both able to traverse the channel and reach the output port.  The L. monocytogenes ∆agr results were added to fig 8.

The most significant finding emerges from the tri-culture experiments. When all three species are present, wild-type L. monocytogenes actively suppresses the dispersal of S. aureus, preventing it from reaching the output port. This interference is dependent on the agr system of L. monocytogenes, as this suppression is not observed when the Δagr mutant is used instead of the WT. In the tri-culture with the Δagr mutant, both bacterial species are able to co-disperse along the fungal highway.

While we are not making a concrete claim that the native L. monocytogenes AIP is directly suppressing the S. aureus system, our findings are highly suggestive that a similar mechanism may be occurring. This interpretation is supported by literature demonstrating cross-talk between agr systems. As West et al. (2023) showed, synthetic peptides related to the Listeria AIP can antagonize the Staphylococcus agr system. Our in-situ data provides a biologically relevant demonstration of this principle, suggesting that a functional Listeria QS system can directly impede the dispersal of Staphylococcus in a shared microfluidic environment.

Therefore, our data demonstrate a clear hierarchy: the fungus enables dispersal for both species, but a functional Listeria quorum-sensing system can then be used to competitively exclude S. aureus from that same network. This highlights a sophisticated, three-way interaction where the outcome of dispersal is determined by both the presence of the fungal highway and the specific quorum-sensing capabilities of the competing bacteria. We have revised the manuscript to emphasize this key point and to frame our conclusions as a strong indication of active, agr-dependent interference, which warrants further investigation.

Abstract and Line 391-391: S dependent wave-like dispersal – no quantification of this interaction, only a qualitative observation. Should not be included as a major claim in the abstract.

We thank the reviewer for their thoughtful critique regarding our claim of S-dependent, wave-like dispersal. We agree that our initial presentation in the abstract was too strong and have revised it to be more descriptive of the observation. However, we would like to respectfully provide additional context that strengthens the relevance of this finding.

The reviewer is correct that S. aureus exhibits complex motility patterns. Our observation of a wave-like dispersal pattern was not made in a vacuum; it is consistent with a well-documented, quorum-sensing-controlled phenomenon in S. aureus biofilms, driven by phenol-soluble modulin  surfactants (Periasamy et al., 2011). In that study, PSMs were shown to generate "waves of biofilm detachment and regrowth" and facilitate the dissemination of cells.

Our study does not claim to have discovered this mechanism. Rather, we propose that our microfluidic-fungal highway model provides a unique platform to visualize this known PSM-driven dispersal in a completely new context: along a living hyphal network. The wave-like pattern we observed is likely a manifestation of this same quorum-sensing and surfactant-mediated process, now enabled to occur along the hyphal thread.

To address the reviewer's valid concern about the necessity of the fungus, we reiterate our key finding: this wave-like dispersal and successful translocation to the output port only occurs in the presence of the fungal hyphae. In control experiments without the fungus, S. aureus does not traverse the channel.

Therefore, we have revised our manuscript to frame this finding more accurately. We now present it as evidence that the fungal network not only acts as a physical bridge but may also provide a suitable surface or microenvironment that facilitates known, species-specific bacterial dispersal mechanisms, such as PSM-driven wave-like spreading. We believe this strengthens the paper's central argument about the specific nature of fungal-bacterial interactions. We have softened the language in the abstract and results to reflect this more nuanced interpretation.

S. aureus can disperse through saturated environments via gliding motility & comet formation, meaning that one cannot attribute its dispersal to the fungi.

We respectfully disagree with the assertion that the dispersal of S. aureus cannot be attributed to the fungus based on its known motility mechanisms. While we acknowledge that S. aureus can exhibit motility in saturated environments, our experimental results demonstrate that these mechanisms are insufficient for translocation across the length of our microfluidic channel in the absence of a fungal network.

Crucially, our findings are based on multiple, consistently replicated experiments. In all control trials where the fungus was absent, S. aureus consistently failed to disperse from the input port to the output port. The presence of the fungal hyphae was not merely correlated with dispersal; it was a prerequisite for it. This clear, repeatable experimental evidence demonstrates that the fungus provides the necessary structure or microenvironment that enables S. aureus to bridge the gap and reach the output port, a feat it cannot accomplish on its own in our system. Therefore, we maintain that our data directly supports the conclusion that the fungal network is essential for the observed dispersal.

Claiming it is S dependent (density dependent) wave-like dispersal would require a lot of quantification; qualitative observation cannot support this claim. Density-dependent wave-like dispersal can be tested by examining whether colony or population expansion on a surface occurs as a coherent advancing front whose speed depends on local cell density. Evidence could include a clearly defined expansion front that moves linearly through time, faster spread from higher initial inoculum densities, or the presence of a minimum density required for sustained expansion. These patterns can be quantified by tracking the position of a fixed density or fluorescence threshold across space and time and comparing the observed spread to predictions from reaction–diffusion models with density-dependent motility or growth (e.g., chemotaxis- or quorum-regulated movement) versus density-independent diffusion.

• Additionally, there is only one device, therefore one biological replicate, and therefore not enough evidence to draw such conclusions.

We respectfully but firmly disagree with the assertion that our conclusion is unsupported due to a lack of quantification and a single biological replicate. This critique misinterprets the scope and nature of our study, which is a proof-of-concept demonstration of a biological phenomenon, not a quantitative biophysical analysis of dispersal kinetics.

The reviewer's suggestions for quantification—tracking expansion fronts, modeling reaction-diffusion dynamics, and testing varying inoculum densities—are indeed the appropriate methods for a dedicated study on the physics of S. aureus motility. However, demanding this level of analysis for what is one observation among several in our paper is an unreasonable standard that falls outside the scope of our research question. Our goal was to establish if and how different species use a fungal network for dispersal, not to precisely model the physical parameters of that dispersal.

Furthermore, the characterization of our work as having "only one biological replicate" is a misrepresentation. While the experiments shown may use a single device, the observation of wave-like dispersal is one part of a larger, coherent dataset. The core findings of our paper—that different species exhibit distinct dispersal phenotypes (e.g., L. monocytogenes requiring flagella, S. aureus exhibiting a unique pattern, and the critical role of the fungus itself) are based on multiple, repeated experiments that validate the central thesis. The reviewer is isolating one qualitative observation and demanding a level of statistical rigor that is not consistently applied to the other findings in the manuscript.

In our revised manuscript, we have already addressed the valid concern about the strength of the claim by modifying the abstract and text to present the wave-like pattern as a notable observation that warrants further investigation, rather than a definitive, quantified mechanism [Lines 21-23, 26-29, 91-96, 403-404, 409-410, 438-440]. We believe this is a reasonable and proportional response that accurately reflects the data and the scope of our study. To demand more would be to require a different paper entirely.

 

Necessity of flagella for dispersal, Text in abstract: “Mechanistically, L. monocytogenes dispersal required flagella, with dispersal impaired in flagellar mutants but enhanced in ∆mogR 21 strains that upregulate flagellar expression.”

    o This experiment is not informative in terms of fungal highway dynamics. It is a bacterial trait, but this shows only increased or impaired movement due to bacterial swimming, not increased or decreased dispersal using the liquid films surrounding fungal hyphae.

We thank the reviewer for this comment and appreciate the opportunity to clarify our interpretation. We agree that our experiment does not distinguish between the specific physical mechanism of transport along the hyphae (e.g., swimming in liquid films versus navigating the physical cavity). The reviewer is correct that we did not make a definitive claim about the use of liquid films, and we acknowledge that this is a reasonable assumption for them to have made.

Our intent with the flagella experiment was not to define the precise micro-environment of transport, but rather to establish a more fundamental principle: that active bacterial motility is a prerequisite for successful dispersal along the fungal highway. The key finding is that in the context of our system, non-motile mutants fail to traverse the network, while mutants that transcribe flagella at higher rates have an enhanced ability to do so. This demonstrates that the dispersal is an active, energy-dependent process requiring the flagellar machinery, rather than a passive one.

To avoid further confusion and to align our language with the actual scope of the experiment, we have revised the abstract and the relevant text. We now state that flagella are required for "active dispersal within the microfluidic device via the fungal network," which removes any implication about the specific physical medium and focuses on the established biological requirement for motility.

 

Reviewer 2 Report

The manuscript presents a thorough and carefully executed  study on bacterial dispersal along fungal hyphae. The the research question is clear and relevant, the methods are well explained and reproducible, and the results directly address the objectives. The discussion integrates the findings well with existing literature and avoids overinterpretation. Overall, this is a strong contribution to the field

Please refer to the attachment. In general, the manuscript is well written and structured

Comments for author File: Comments.pdf

Author Response

Reviewer comments are in blue, our response is in black
Response to reviewer 2.

The manuscript focuses on two bacterial motility mechanisms related to dispersal along

the fungal superhighway when interacting with Aspergillus flavus. the research question

is well defined, the experimental strategy was designed to answer it and the results

consistently support the conclusions. The main strengths I see, are:

• the microfluidic system used allows quantitative assessment of bacterial movement

in mono-, co-, and triculture conditions and the analysis of motile and non-motile

bacteria strengthens the generality of the observations while including flagellar and

quorum sensing mutants provides molecular and mechanistic depth, all of which is

supported by appropiate controls and statistical analyses.

• motility comparison between Listeria monocytogenes, Pseudomonas aeruginosa,

and Ralstonia solanacearum, suggests that flagellar structure is an important

dispersal determinant, with different degrees of contribution.

• time-lapse imaging and PSM production suggests Staphylococcus aureus densitydependent and QS-regulated surfactant activity, rather than a motile related dispersal mechanism, supported by L. monocytogenes outcompetition.

While the manuscript appropriately focuses on bacterial traits and their regulatory

mechanisms, acknowledging the fungal cell wall influence in the observed dispersal

dynamics could help contextualize the authors findings. In filamentous fungi, such as

Aspergillus species, temperature and growth conditions affect hyphal architecture,

surface properties, and therefore, cell wall organization, which might change bacterial

interaction and movement. However, I do not consider this to be necessary for

publication, since is beyond the scope of the paper.

In summary, this is a solid and well-executed work that provides mechanistic details into

how bacterial species disperse along fungal networks and how interspecies interactions

shape the process. The findings are of interest to microbial ecology, host-pathogen

interactions, and polymicrobial communities research. I endorse for publication.

Response: We sincerely thank the reviewer for their thoughtful evaluation of our manuscript and for highlighting its strengths. We also appreciate the insightful remark about acknowledging the potential influence of fungal cell wall properties on bacterial dispersal. We agree that it could further contextualize our findings. However, as noted by the reviewer, it is beyond the scope of the paper. We thank the reviewer again for the endorsement.

 

Reviewer 3 Report

Major issues:

1. The pattern of result presentation should, in my opinion, be improved. For example, several videos were put in the supplementary files, which are difficult to see, due to the downloaded problem. I suggest that the authors might present the time-lapse images based on the videos they took during the experiments. In such result presentation way, we, as well as the readers might view and get the points of dynamic bacterial dispersal in a more accessible way than that in the submitted version.
2. The section of Materials and Methods should be condensed. For instance, a paragraph of 3-4 sentences is sufficient for 10 Statistical Analysis.  

Minor issues:

1. The scientific name of microbial species should be in italics. P2 Line 45-46-47-48, Line 50, Line 51…
2. P3 Line 109, at the beginning of a sentence, 100µl -> One hundred microliter
3. P5 Line 233, 10 or 20 X magnification ->10 or 20 × magnification
4. P11 Figure 5, “A. flavus” should be in italics. Same thing for Figure 7.
5. P12 Line 378, the subtiltle of the section 3.6 should not contain a dot at the end. Same thing for section 3.8.
6. P13 line 411, without exception -> as expected
7. P15 Line 452-455, please rewrite this sentence.
8. P15 line 470, Attribute->trait
9. P15 line 471-472, the sentence “…, we found lower temperatures promoted increased movement in both monoculture and coculture.” might be modified to “…, we found that dropped temperatures might promote the movement in both monoculture and coculture.”
10. P15 Line 479, “In flavus/L. monocytogenes interaction….”-> “In A. flavus- L. monocytogenes interaction….”
11. As for the References section, the title of the paper should be in lower-case except the first word of this title. The formatting of Refs. 8, 18, 19, 22, 28, 31, 32, 35, 36 needs modifying.

Comments for author File: Comments.pdf

Author Response

Reviewer comments are in blue, our response is in black
Response to reviewer 3.

Regalado et al. investigated how diverse bacterial species disperse in monoculture versus travel in coculture with Aspergillus flavus using microfluid device and molecular biological technology. They found that bacterial motility on the fungal highway stemmed from several species-specific mechanisms, including bacterial flagella, transcriptional regulation, and quorum sensing. It is interesting and attractive. In order to improve the quality of this paper, there are a few points that need to be addressed prior to the acceptance by JoF.

Response: We thank you for your encouraging comments. We address all your points below:

Major issues:

1. The pattern of result presentation should, in my opinion, be improved. For example, several videos were put in the supplementary files, which are difficult to see, due to the downloaded problem. I suggest that the authors might present the time-lapse images based on the videos they took during the experiments. In such result presentation way, we, as well as the readers might view and get the points of dynamic bacterial dispersal in a more accessible way than that in the submitted version.

Response: We thank the reviewer for this thoughtful suggestion. We agree that a clear visualization of dynamic bacterial dispersal is important. Regarding the accessibility of the supplementary videos, JoF informed us that there was a temporary problem with the supplemental files. However, this issue has been solved, and the readers can fully access the files. While we appreciate the suggestion to include time-lapse images, we respectfully prefer to keep the videos as the primary pattern of result presentation, as they better reflect the temporal changes and spatial progression of the bacteria.

2. The section of Materials and Methods should be condensed. For instance, a paragraph of 3-4 sentences is sufficient for 2.10 Statistical Analysis.

Response: We appreciate this comment and agree that this section can be condensed. The Materials and Methods section has been revised and condensed accordingly.

Minor issues:

1. The scientific name of microbial species should be in italics. P2 Line 45-46-47-48, Line 50, Line 51…

Response: We thank the reviewer for pointing this out and have italicized all microbial species throughout the paper.

2. P3 Line 109, at the beginning of a sentence, 100μl -> One hundred microliter.

Response: This sentence has been corrected to “One hundred microliter” at the beginning of the sentence. P4 Lines 121-122.

3. P5 Line 233, 10 or 20 X magnification ->10 or 20 × magnification.
   
   Response: The text has been corrected as suggested.
   

4. P11 Figure 5, “A. flavus” should be in italics. Same thing for Figure 7.

Response: All figures have been adjusted accordingly.

5. P12 Line 378, the subtitle of the section 3.6 should not contain a dot at the end. Same thing for section 3.8.

Response: The subtitles have been revised, and the periods have been removed.

6. P13 line 411, without exception -> as expected.
   

Response: The text has been corrected as suggested.

7. P15 Line 452-455, please rewrite this sentence.

Response: The sentence has been revised and divided into two to improve clarity:

P16 Line 463-466. “We focused our studies on two Gram-positive pathogens that have been associated with fungi. Listeria is found in similar dairy environments as A. flavus [32] and S. aureus in medical settings with A. fumigatus and Candida spp. [33]”

8. P15 line 470, Attribute->trait.

Response: The text has been corrected as suggested.

9. P15 line 471-472, the sentence “…, we found lower temperatures promoted increased movement in both monoculture and coculture.” might be modified to “…, we found that dropped temperatures might promote the movement in both monoculture and coculture.”

Response: We rewrote the sentence as suggested.

P16 Line 481-482. “Following this, we found that dropped temperatures might promote the movement in both monoculture and coculture.”

10. P15 Line 479, “In A. flavus/L. monocytogenes interaction….”-> “In A. flavus- L. monocytogenes interaction….”

Response: The text has been corrected as suggested.

11. As for the References section, the title of the paper should be in lower-case except the first word of this title. The formatting of Refs. 8, 18, 19, 22, 28, 31, 32, 35, 36 needs modifying.

 

Response: The formatting for all references was checked and modified when needed.

Round 2

Reviewer 1 Report

The manuscript has been improved and I commend the authors on this. However, several comments in the original review were not addressed (especially regarding figures) and there still remains much doubt regarding the nature of the experimental replicates used to generate the data presented.

Additionally, the authors did not address concerns of potential pressure gradients being generated across the device that could drive bacteria through the agarose matrix; as the channel is filled with agar, fungal hyphae break down the agarobiose matrix when they grow through it, creating a saturated area around the hyphae that bacteria can swim through or be suspended in. Hence, a pressure gradient could be generated across the channel, driving flow of liquid containing bacteria from the inlet to the outlet (or vice versa). Controls experiments should be included to ensure that this is not the case.

 

Please see comments above in "Recommendations for Authors" and "Major Comments" sections. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors have addressed all the comments I previously made. In my opinion, it would be acceptable for publishing after modifying the minor points related to Fig. 4B and Fig. 6.

In my opinion, it would be acceptable for publishing after modifying the minor points related to Fig. 4B and Fig. 6.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf