PCOMPBIOL-D-25-01216

Learning, sleep replay and consolidation of contextual fear memories: A neural network model

PLOS Computational Biology

Dear Dr. Seriès,

Thank you for submitting your manuscript to PLOS Computational Biology. After careful consideration, we feel that it has merit but does not fully meet PLOS Computational Biology's publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript within 60 days Oct 19 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at ploscompbiol@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pcompbiol/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

* A rebuttal letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. This file does not need to include responses to formatting updates and technical items listed in the 'Journal Requirements' section below.

* A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

* An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, competing interests statement, or data availability statement, please make these updates within the submission form at the time of resubmission. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter

We look forward to receiving your revised manuscript.

Kind regards,

Daniel Bush

Academic Editor

PLOS Computational Biology

Marieke van Vugt

Section Editor

PLOS Computational Biology

Additional Editor Comments:

The authors should clarify how robust their results are to different parameter settings, and the relationship between their model and empirical data, as described in more detail below.

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

1) We ask that a manuscript source file is provided at Revision. Please upload your manuscript file as a .doc, .docx, .rtf or .tex. If you are providing a .tex file, please upload it under the item type u2018LaTeX Source Fileu2019 and leave your .pdf version as the item type u2018Manuscriptu2019.

2) Please upload all main figures as separate Figure files in .tif or .eps format. For more information about how to convert and format your figure files please see our guidelines:

https://journals.plos.org/ploscompbiol/s/figures

3) Please upload a copy of Figures ??a, ??d. Fig ??b, and Fig ??c which you refer to in your text on pages 46, and 47. Or, if the figure is no longer to be included as part of the submission please remove all reference to it within the text.

4) We notice that your supplementary Figures, and Tables are included in the manuscript file. Please remove them and upload them with the file type 'Supporting Information'. Please ensure that each Supporting Information file has a legend listed in the manuscript after the references list.

5) Some material included in your submission may be copyrighted. According to PLOSu2019s copyright policy, authors who use figures or other material (e.g., graphics, clipart, maps) from another author or copyright holder must demonstrate or obtain permission to publish this material under the Creative Commons Attribution 4.0 International (CC BY 4.0) License used by PLOS journals. Please closely review the details of PLOSu2019s copyright requirements here: PLOS Licenses and Copyright. If you need to request permissions from a copyright holder, you may use PLOS's Copyright Content Permission form.

Please respond directly to this email and provide any known details concerning your material's license terms and permissions required for reuse, even if you have not yet obtained copyright permissions or are unsure of your material's copyright compatibility. Once you have responded and addressed all other outstanding technical requirements, you may resubmit your manuscript within Editorial Manager.

Potential Copyright Issues:

i) Figures 1, 3, and 5. Please confirm whether you drew the images / clip-art within the figure panels by hand. If you did not draw the images, please provide (a) a link to the source of the images or icons and their license / terms of use; or (b) written permission from the copyright holder to publish the images or icons under our CC BY 4.0 license. Alternatively, you may replace the images with open source alternatives. See these open source resources you may use to replace images / clip-art:

- https://commons.wikimedia.org

- https://openclipart.org/.

6) In the online submission form, you indicated that your data will be submitted to a repository upon acceptance. We strongly recommend all authors deposit their data before acceptance, as the process can be lengthy and hold up publication timelines. Please note that, though access restrictions are acceptable now, your entire minimal dataset will need to be made freely accessible if your manuscript is accepted for publication. This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If you are unable to adhere to our open data policy, please kindly revise your statement to explain your reasoning and we will seek the editor's input on an exemption.

7) Please amend your detailed Financial Disclosure statement. This is published with the article. It must therefore be completed in full sentences and contain the exact wording you wish to be published.

1) State what role the funders took in the study. If the funders had no role in your study, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

2) If any authors received a salary from any of your funders, please state which authors and which funders..

If you did not receive any funding for this study, please simply state: u201cThe authors received no specific funding for this work.u201d

8) Please send a completed 'Competing Interests' statement, including any COIs declared by your co-authors. If you have no competing interests to declare, please state "The authors have declared that no competing interests exist". Otherwise please declare all competing interests beginning with the statement "I have read the journal's policy and the authors of this manuscript have the following competing interests:"

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: The paper presents a biologically grounded neural network model of contextual fear memory formation and consolidation, organized into modules that mirror key brain regions. A sensory cortical module encodes environmental contexts and sends this information in parallel to both a hippocampal module, which encodes contexts rapidly via entorhinal inputs, and a neocortical module that gradually builds long term, generalized context representations through consolidation. A basal amygdala module encodes associations reflecting each context’s valence. Offline replay allows these representations to be strengthened in the neocortex. Replay supports both renewal and time dependent increases in generalization, while synaptic homeostasis trims weak associations to prevent overgeneralization. These dynamics also enable the model to account for stress enhanced fear learning and the maladaptive effects of sleep deprivation.

The manuscript is clear and well-organized, outlining each model component and its empirical rationale in a way that allows readers to understand how replay and synaptic downscaling contribute to the observed effects. The model offers a thorough, operationalized implementation of the hypothesis that fear generalization reflects hippocampal-to-neocortical transfer, providing a strong foundation for future experimental work. However, the conceptual novelty of the framework may be somewhat limited, as systems consolidation has long been proposed as a mechanism underlying generalization. The following comments raise several issues regarding the rationale behind specific architectural choices, the consistency of mechanistic assumptions across simulations, and the model’s alignment with related empirical findings on contextual fear memory consolidation.

Major comments:

1. There appear to be discrepancies between the model’s architecture and established anatomical pathways in the brain. In the model, the sensory cortex feeds separately into the hippocampus and neocortex. However, in the brain, the sensory cortex is part of the neocortex, and hippocampus inputs are processed through neocortical hierarchies that interact with the hippocampus via a bidirectional loop. The inclusion of ECout as an isolated output in the model also departs from the brain, as the entorhinal cortex receives hippocampal outputs and projects them back to the neocortex. The authors should clarify or revise these architectural choices to better reflect the relevant circuitry.

2. The benefits of consolidation for memory retention appear limited in the current model. For example, Figure 4c shows that consolidated cortical memories decay linearly over time and would eventually be lost, suggesting that consolidation provides only a short-term advantage. This seems like an important limitation. The authors should consider discussing or implementing mechanisms that help stabilize cortical representations over longer timescales.

3. Greater consistency in mechanistic assumptions across simulations would strengthen the model. Currently, some implementation details appear only in specific modules or simulations, making it unclear which assumptions apply universally. For instance, it is not clear whether the suppression of fear-inhibiting BA_I cells is learned or hard-coded, whether the higher activation threshold for BA_I versus BA_P cells holds across simulations, or whether the boost in cell recruitability applies only to the amygdala or more broadly. Clarifying which mechanisms are implemented universally and which are specific to particular simulations or model components, and specifying which results rely on these assumptions being simulation- or component-specific, would help readers better interpret the results.

4. The model’s ability to align with related empirical findings on fear memory consolidation remains unclear. For instance, studies have shown that hippocampal replay selectively prioritizes fear-related experiences (Wu et al., 2017), and that strong aversive events can retroactively strengthen neutral memories formed days earlier (Zaki et al., 2025). In the current model, replay and hippocampal-to-amygdala connections do not vary with valence, and valence does not modulate replay. As a result, it is not obvious whether the model could account for such findings. It would be valuable to discuss or explore potential mechanisms that might allow the model to capture these effects.

Minor comments:

1. The claim “Ours is the first to incorporate a neural mechanism enabling” fear memory consolidation could be softened, given that models such as Mattar & Daw (2018) also simulated consolidation of fear memory. A more accurate phrasing might be that this is the first model to implement systems consolidation of fear memory.

2. In Figure 1 (page 5), the legend refers to “Nodes labelled with an S-shape”. I am not sure which nodes are marked that way.

3. On page 9, “In the following, we …” seems incomplete.

4. Also on page 9, please add a space in “inRecall mode” so that it reads “in Recall mode.”

5. In Figure 5c (page 13), what does the y-axis label “CeM” refer to?

6. When discussing insomnia and sleep disruptions (page 14), the motivation would be clearer if these pathologies were tied more directly to fear acquisition and expression.

7. The Kitamura et al. study is described in detail in the Discussion, where readers might expect a higher-level summary. The authors might consider introducing it earlier to help motivate the model’s architecture, or alternatively, making the description in the Discussion more concise.

Reviewer #2: The authors extended a previous computational model proposed by Fiebig and Lansner to address the underlying neural mechanisms of learning and consolidation of contextual fear memories. This model consists of three networks: the hippocampus, the neocortex, and the amygdala. Their simulations provide a mechanistic account of the observed transfer of context–fear memories from hippocampal–amygdalar to amygdalo–cortical circuits. The model also reproduces empirically observed increases in fear generalisation following contextual fear conditioning. In addition, they found that (in their model) impaired synaptic weakening in amygdalar fear circuits—as might plausibly occur due to sleep disruptions or psychological stress—could mechanistically contribute to fear sensitisation. Overall, this is a valuable model, and experimentally testing some of its hypotheses could yield clinically relevant insights.

Major comments:

1, This work focuses mainly on how fear memory acquisition can be enhanced and generalised. This is valuable, as it helps to elucidate potential neural mechanisms underlying these phenomena through computational modelling. However, what is perhaps more interesting—and more relevant for translational neuroscience—is how established fear memories can be weakened, particularly traumatic memories. I would like to see some simulations targeting this point included in the paper. There is a short section on fear extinction, but it is too brief and lacks sufficient detail on how sleep, de-stressing, or other potential factors may lead to fear memory weakening/extinction.

For example:

• Line 334: “…enhancing synaptic sleep homeostasis in the amygdala could be a key therapeutic strategy in PTSD and related disorders.” I would expect the authors to demonstrate this directly in their model.

• Figure 7: the authors show that chronic sleep deprivation leads to accelerated fear acquisition. I would also expect to see conditions that decelerate fear acquisition.

2, I have concerns about the modelling of hippocampal replay. From Ji & Wilson, we know that replay coordinates sequential activity between the hippocampus and visual cortex, but this occurs during specific windows of sleep. In the current context, “replay” is used to establish communication between hippocampus, cortex, and amygdala. This does not align precisely with the timescale of replay, which is usually tens to hundreds of milliseconds. For example, it is unclear what the unit of the time step in Fig. 4b is. In the text you mention “over the course of a sleep phase”, but it is not clear what duration you mean by a sleep phase. Is it an entire NREM episode? At other times the timescale is given as “days” (e.g., Fig. 5d and Fig. 8d). The authors need to make the timescale consistent, or, if that is not possible, provide clearer descriptions.

3, I ran the authors’ provided code and was able to reproduce the figures in the main text. This is a relatively complex model with many parameters, and I give credit to the authors for making this model work as intended. However, I am somewhat concerned about the robustness of the reported results. For example, if a parameter setting were altered, would the results remain qualitatively similar? Or is the current configuration “cherry-picked”? A stability analysis of the model would strengthen confidence in the results.

4, Although the model explains previous empirical findings, makes testable predictions, and links these results to potential neural mechanisms, there remains a significant gap between the computational assumptions/mechanisms and the actual neural mechanisms underlying these phenomena. For example:

• Line 314: “Our model can explain these findings due to our assumption that sleep after fear learning prunes weak synapses onto P-cells.”

• Line 331: “Targeting sleep disruptions … may directly influence the synaptic mechanisms underlying chronic anxiety.”

• Line 383: “…implies that therapeutic interventions targeting either sleep quality or resilience to stress-induced homeostatic dysregulation could reduce vulnerability to pathological fear and anxiety disorders.”

These claims are somewhat too strong, given that they derive from a computational model rather than direct empirical evidence. The language could be moderated to avoid potentially misleading over-interpretation. I understand that some of these issues are difficult to address fully in a computational modelling study. Nonetheless, I find the model valuable.

Minor comments

1. Line 22: “…also exhibits activity coordinated with hippocampal replay during sleep…” — needs citations.

2. Is there empirical evidence supporting the choice to model CTX as a BPCNN and HPC as a WTA network? How do these differ from a Hopfield network, and why was a Hopfield network not used instead?

3. In the author summary, the authors ask: “Why do some fear memories fade while others persist or even grow stronger over time?” After reading the paper, I did not find a clear answer. Can the model explain this? (This relates to my major comment 1.)

4. Line 76: the authors state that representations in the hippocampus are short-lived and overwritten by new learning. In Hopfield-like models, this would be termed catastrophic forgetting. Why is this not the case for the neocortex in the proposed model?

5. Line 79–80: replay of sequential patterns is implemented via inhibitory synapses operating at a fast timescale. Previous works (e.g., Azizi et al., 2013; Ji et al., 2024) used adaptation to generate replay-like dynamics. What is the difference between these approaches, given that the present model does include adaptation (Eq. 7–10)?

6. Line 239: “…fear extinction depends primarily on silencing, rather than erasure…” The model shows that silencing can work, but why not erasure? This was not tested.

7. Fig. 4a: why does the hippocampus rapidly form context engrams? Is this due to fast synapses?

8. Fig. 4d: why is the x-axis labelled “trials” rather than “steps”?

9. Line 300: what does “unchecked” mean here?

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No: The authors include the following note: "All code written in support of this publication will be publicly available on Zenodo."

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .

Reviewer #1: No

Reviewer #2: Yes: Zilong Ji

Figure resubmission:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step. If there are other versions of figure files still present in your submission file inventory at resubmission, please replace them with the PACE-processed versions.

Reproducibility:

To enhance the reproducibility of your results, we recommend that authors of applicable studies deposit laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

PCOMPBIOL-D-25-01216R1

Learning, sleep replay and consolidation of contextual fear memories: A neural network model

PLOS Computational Biology

Dear Dr. Seriès,

Thank you for submitting your manuscript to PLOS Computational Biology. After careful consideration, we feel that it has merit but does not fully meet PLOS Computational Biology's publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Mar 09 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at ploscompbiol@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pcompbiol/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

* A letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'. This file does not need to include responses to formatting updates and technical items listed in the 'Journal Requirements' section below.

* A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

* An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, competing interests statement, or data availability statement, please make these updates within the submission form at the time of resubmission. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Daniel Bush

Section Editor

PLOS Computational Biology

Marieke van Vugt

Section Editor

PLOS Computational Biology

Additional Editor Comments:

Both reviewers are happy to accept the manuscript for publication, but Reviewer 2 has requested a few very minor final edits. I think it is important to make these changes, but I will check those myself - the manuscript will not go back out for review.

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: Overall, the authors’ responses are clear and thorough, and the revisions address most of the concerns. The architectural abstractions are now better motivated, the limitations of cortical memory decay are acknowledged, and parameter consistency across simulations is clarified. The discussion of valence-biased replay and memory linking is appropriate, even if these remain future extensions. All minor comments were handled cleanly. Together with the revisions made in response to Reviewer #2, the manuscript is substantially improved, and the authors’ responses are generally satisfactory.

Reviewer #2: Thank you for the detailed responses to my review. I am satisfied with most of the authors’ replies, particularly the inclusion of parameter sweeps, which strengthen the complex network model. Overall, I consider this work to be valuable and have recommended it for publication.

However, some of the network manipulations remain relatively coarse and lack detailed neurobiological grounding. As a result, there is still a gap between the modelled phenomena and their biological interpretation, and alternative computational mechanisms might plausibly lead to similar results.

I have a few additional suggestions:

1. For figures reporting results averaged over multiple simulation runs, measures of variability (e.g. s.e.m. or s.d.) should be included in the plots, for example in Fig. 4.

2. Line 31: “in PTSD — often emerge gradually, rather than immediately, after intense fearful or traumatic experiences.” Appropriate citations are missing here.

3. Line 50: citations 22–24 do not seem necessary in this context.

4. Line 90: please consider adding a citation to the Hopfield (2010, PNAS) paper on mental exploration.

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: None

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .

Reviewer #1: No

Reviewer #2: No

Figure resubmission:

While revising your submission, we strongly recommend that you use PLOS’s NAAS tool (https://ngplosjournals.pagemajik.ai/artanalysis) to test your figure files. NAAS can convert your figure files to the TIFF file type and meet basic requirements (such as print size, resolution), or provide you with a report on issues that do not meet our requirements and that NAAS cannot fix.

After uploading your figures to PLOS’s NAAS tool - https://ngplosjournals.pagemajik.ai/artanalysis, NAAS will process the files provided and display the results in the "Uploaded Files" section of the page as the processing is complete. If the uploaded figures meet our requirements (or NAAS is able to fix the files to meet our requirements), the figure will be marked as "fixed" above. If NAAS is unable to fix the files, a red "failed" label will appear above. When NAAS has confirmed that the figure files meet our requirements, please download the file via the download option, and include these NAAS processed figure files when submitting your revised manuscript.

Reproducibility:

To enhance the reproducibility of your results, we recommend that authors of applicable studies deposit laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Dear Prof Seriès,

We are pleased to inform you that your manuscript 'Learning, sleep replay and consolidation of contextual fear memories: A neural network model' has been provisionally accepted for publication in PLOS Computational Biology.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Computational Biology.

Best regards,

Daniel Bush

Section Editor

PLOS Computational Biology

Marieke van Vugt

Section Editor

PLOS Computational Biology

***********************************************************