Virtual reality reveals fine-scale alterations in behaviour following loss of the ADHD-linked gene adgrl3.1 in zebrafish

  1. Phoebe Reynolds
  2. Emily Read
  3. Caitlin Daly-East
  4. Matthew O. Parker
  5. Robert Hindges

  • Curated by eLife

    eLife logo

    eLife Assessment

    The study presents valuable findings on early behavioral phenotypes that arise in an ADHD-associated adgrl3.1 mutant zebrafish, using a new behavioral approach with a closed-loop OMR assay. The evidence supporting the claims of the authors is solid; however, the validation of the method and analysis is incomplete and would benefit from more rigorous approaches and validation practices. This work will be of broad interest to developmental biologists and behavioral neuroscientists.

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Zebrafish have been used a prominent model for high-throughput phenotypic screens of candidate risk gene mutations for several disorders. This also includes models for attention deficit/hyperactivity disorder (ADHD). Traditional behavioural tests, such as the forced light/dark assay, concentrate on basic locomotion measures. However, recently developed visually-driven locomotion assays, for example closed-loop systems using virtual reality, have allowed extraction of richer data on animal locomotion and decision-making under different sensory stimuli. Here, we have used such a system to assess the behaviour in adgrl3.1 mutant fish, an established model for ADHD. Our results show that mutants exhibit a higher baseline excitability and a lower threshold for initiating motor events, demonstrating that collecting behavioural responses in an interactive environment enables a more precise characterisation of ADHD-relevant phenotypes associated with adgrl3.1 disruption. More generally, we establish a scalable translational platform to screen gene-function relationships and possible therapeutic interventions, not only for ADHD but multiple neurodevelopmental disorders.

Article activity feed

  1. eLife

    eLife Assessment

    The study presents valuable findings on early behavioral phenotypes that arise in an ADHD-associated adgrl3.1 mutant zebrafish, using a new behavioral approach with a closed-loop OMR assay. The evidence supporting the claims of the authors is solid; however, the validation of the method and analysis is incomplete and would benefit from more rigorous approaches and validation practices. This work will be of broad interest to developmental biologists and behavioral neuroscientists.

  2. eLife

    Reviewer #1 (Public review):

    Summary:

    Reynolds and colleagues provide a deep phenotypic analysis of behavior in adgrl3.1 mutant zebrafish at larval stages using a closed-loop optomotor response (OMR) assay. The analyses conducted are interesting and extract new locomotor phenotypes with possible relevance to the role of adgrl3.1 in ADHD. Reduced interbout interval (both in the OMR assay and in dark rest periods) and increased distance moved provide greater resolution on hyperactivity phenotypes already described in these mutants. Reduced variation in interbout interval and reduced variation in swim speeds throughout the assay provide new insights into how behavior is altered; the authors suggest that these findings reflect more stereotyped, less flexible behavior in adgrl3.1 mutant animals. Analyses of task performance are interesting and could help understand how / whether animals maintain vigilance over time in the OMR assay and reveal trends in adgrl3.1 mutants relative to siblings, but ultimately do not identify significant phenotypes for adgrl3.1 mutants. While methods are extremely clear and analyses and phenotypic insights are solid, the authors do not provide sufficient support for assertions that their paradigm separates anxiety from locomotor activity or extracts phenotypes central to ADHD (impulsiveness, attention, etc). In some instances, interpretation of behavioral phenotypes in the context of disease is difficult to follow or not well supported with citations, etc.

    Strengths:

    (1) Deeper phenotypic analysis of adgrl3.1 locomotor phenotypes reveals changes to bout timing/initiation of locomotion as potentially causative for broader hyperactivity phenotypes previously reported.

    (2) Interesting dissection of OMR performance over time and variability in locomotor parameters, assessment of OMR performance in high- and low-contrast.

    (3) Methods are clearly described and considered to be rigorous.

    Weaknesses:

    (1) The introduction does not clearly spell out why the closed-loop OMR assay is expected to capture phenotypes central to ADHD (impulsiveness, attention, etc). Similarly, it's stated in the discussion that hyperactivity is driven by shorter inter-bout intervals and longer bout lengths...reflecting a reorganization of locomotor timing," and that "such fine-scale insights are not possible in standard light/dark paradigms." But in fact, each of these parameters was examined in the dark periods and could be assessed in a standard light/dark assay. As explained at the end of the discussion, this work provides a detailed analysis of locomotion and extracts new and interesting phenotypes, but the assertion that this is a function of the assay / that the assay is uniquely relevant to ADHD is not well-supported. The final statement of the introduction more accurately captures the advantages of the assay used: "this allowed us to assess whether loss of adgrl3.1 alters not only overall locomotor drive...but also specific visuomotor behavioral responses under different stimulus demands."

    (2) The statement early in the results that "this approach extends beyond classical locomotor assays conducted in static light / dark environments, where locomotor activity may conflate with anxiety-related responses" and later that the closed-loop OMR assay "disentangles hyperactivity from anxiety-related responses" are not well-supported. Anxiety states could influence performance on OMR (Braun et al., 2024, Molec Psychiatry).

    (3) Some interpretations of the phenotypes are not well supported by citations and may be overstated. For example, "adgrl3.1 elevates baseline arousal...producing a phenotype of heightened but less exploratory visuomotor activation." Since bout duration is increased alongside reduced interbout interval and increased total distance traveled, reduced exploration is not well-supported by the data. Later in the results, it's suggested that the increase in distance traveled reflects "over compensatory hyperactivity under ambiguous sensory conditions, consistent with attentional deficits." It's not clear what this means - references would be helpful to create links between hyperactivity and detection of ambiguous sensory conditions, and also between hyperactivity under these conditions and attentional deficits.

  3. eLife

    Reviewer #2 (Public review):

    In the study by Reynolds et al., the authors propose a new behavioral approach for ADHD evaluation using a genetically modified zebrafish model. The study is interesting and has potentially important implications for the field. However, there are several methodological, analytical, and validation-related issues that should be addressed before the study can be considered scientifically rigorous.

    Comments:

    (1) Introduction section

    What is the epidemiological evidence supporting the prevalence of ADGRL3 dysfunction in the human population? The authors should consider adding this information, as well as clarifying where ADGRL3 mutations rank among other genetic variants associated with ADHD.

    I suggest reconsidering the sentence "quantifiable behavioural repertoires that complement rodent approaches." Zebrafish studies do not simply complement rodent studies; they can serve as independent pharmacological and toxicological tools that may be used in parallel with rodent models.

    The statement "forced light/dark (FLD) locomotion test" is too broad. Are the authors referring to the Visual Motor Response Test? If so, this is a robust assay that can evaluate not only anxiety-like responses but also locomotor state, arousal, decision-making, and potential cognitive impairment in larvae. A clearer description of the assay is necessary, especially to justify the statement that "while useful for detecting overall activity differences, it cannot determine whether increased movement reflects hyperactivity, altered arousal, disrupted behavioural control, or anxiety-like responses." In contrast, subtle behavioral changes across light and dark phases can be highly informative when velocity, time moving, and anxiety-like responses are analyzed together.

    (2) Methods section

    The zebrafish husbandry section lacks several essential details. The authors should include fundamental information, such as the embryo medium used, how embryos were obtained, the age of the breeding adults, and how larval age was determined in hours post-fertilization. These details are necessary for proper interpretation and reproducibility of the data.

    Why was the mutant DNA not sequenced? Although agarose gel electrophoresis can provide useful preliminary evidence of mutation, it cannot precisely determine the number or nature of base-pair changes. This information is essential because different mutations can have distinct impacts on gene function.

    The sentence "All statistical analyses and tests were completed on Prism10 (GraphPad)" is insufficient. It should be specified which statistical tests were used, including assumptions tested, post hoc comparisons, correction methods, and how experimental replicates or batch effects were handled.

    Overall, the methods section lacks sufficient information to support the scientific rigor of the study. It is unclear whether every individual evaluated behaviorally was injected and then only a subset was genetically confirmed, or whether stable breeding matrices were generated and all experimental individuals were derived from these parents. The manuscript mentions "2-6 parent batches per experiment, with batches collected and run on separate days," but the genetic origin and validation of these batches remain unclear.

    If embryos were injected for each batch, how did the authors ensure that the mutation was homogeneous enough across individuals to consider them equivalent? How did the authors confirm that the mutation was homozygous or present across all relevant cells? Zebrafish embryos remain at the single-cell stage for only a short period before mitosis begins. Without detailed information regarding breeding timing, embryo collection, injection timing, and sequencing validation, it is difficult to determine whether the injected embryos developed homogeneous mutations or mosaic patterns.

    Additionally, to claim a knockout model, protein-level validation, such as Western blotting or another protein expression assay, should be provided. At present, there appears to be some confusion between a knockout and a knockdown model.

    To validate a new behavioral protocol, the authors should compare their assay with an established gold-standard behavioral paradigm using the same experimental batch. They should clarify why this comparison was not performed.

  4. eLife

    Reviewer #3 (Public review):

    Summary:

    The study provides an in-depth phenotyping of a novel zebrafish larval model of ADHD. This topic is interesting, and the model and the approach are relevant and well-justified. While the paper has a massive amount of high-quality data, the general structure and presentation of this material lack focus and a clearly articulated rationale.

    Strengths:

    The paper is methodologically sound, well-presented, and well- illustrated. It has a clear logical rationale and reasonable experimental design.

    Weaknesses:

    The amount of high-quality data is impressive, yet the general structure and presentation of this material lack focus and a clearly articulated rationale.

    (1) First, it is unclear why VR is necessary here. It needs a better explanation in both the abstract and the intro section of the manuscript.

    (2) Second, data need to be better presented (most important things first, least important - shorter or move to the Supplementary materials). Currently, it is too much to be clear and easy to follow.

    (3) Discussion needs to better state the novelty and the significance of these findings. What does the study offer that is new? Why was it important to perform? What big questions does it address?

    (4) The authors should better discuss the study limitations and future directions of research.

    (5) There should be a stronger conclusion with a take-home message to emphasize what new information the study brings and why it is important.

    (6) The overall style of the paper should be improved. Currently, it reads like a dry bulleted CRO report, not a usual scholarly paper.

    (7) Optimize the text flow. Currently, the overall flow of the discussion needs to be smoother - it now reads as a selection of bulleted paragraphs, with few connections between them.

  5. eLife

    Author response:

    We would like to thank the editorial team and the reviewers for their thoughtful assessment of our manuscript. We are highly encouraged that the reviewers found our closed-loop OMR virtual reality assay to be a valuable new behavioural paradigm, and that our findings regarding the early behavioural phenotypes in adgrl3.1 mutant zebrafish provide solid, high-quality data of broad interest to the community. We also appreciate your constructive feedback regarding the structural flow of our manuscript, the need for tighter conceptual framing, and the request for more rigorous methodological explanation and improved presentation. In our upcoming revision, we plan to directly address the suggestions raised to ensure clarity of our work.

    For our revised manuscript, we will ensure to focus on fully contextualising our conceptual rationale and unique utility of our behavioural assay, ensuring a clear link to clinical relevance. This means we will provide further clarity on the rationale and relevance of the paradigm and interpretations of behaviours, while ensuring we are not overstating the absolute separation of anxiety-like states from hyperactivity and contextualising our findings within the broader literature. In addition, as suggested by the reviewers, we will improve the Abstract to explicitly articulate the unique advantages of the closed-loop OMR virtual reality paradigm over standard static assays and provide definitions. In addition, we will also revise our interpretations in the Discussion, such as around swim speed, exploration, and visual sensitivity, to ensure they are fully grounded and supported with clear flow.

    Equally important to our revision is to clarify genetic validation, methods and statistical rigour as recommended by the reviewers. Therefore, we will update the Methods to explicitly detail further husbandry details, such as the embryo medium used, breeding protocols, and the exact timings in hours post-fertilisation. To resolve any ambiguity surrounding our genetic model, we will provide further explanations and include details on genotype and sequencing protocols. Where needed, we will also update the statistical analysis and expand the Methods accordingly to detail all statistical reporting for clarity.

    Finally, as suggested, we will also improve the overall structure, flow, and accessibility. In the Results, we will ensure the core behavioural phenotypes take primary focus throughout the writing. Furthermore, we will also improve the Discussion to ensure it is a unified, cohesive narrative that smoothly integrates our main behavioural findings with genetic model validation, limitations, and future directions. Additionally, we will update all visual presentations, figure formatting, and labelling to ensure accessibility and improved readability.

    We are incredibly grateful to the reviewers for their insightful recommendations. We are confident that by integrating their revisions, we will substantially strengthen the clarity of the work and better demonstrate its impact.

  6. Version published to 10.64898/2026.04.20.719162 on bioRxiv