PREreview of Developmental analysis of the cone photoreceptor-less little skate retina

Summary:

This manuscript investigates retinal development in the cone-less little skate, focusing on how Onecut1 (OC1) and cone phototransduction genes might underlie the absence of cones. The central question, as I interpret it, is: Does the loss of cones in little skate arise from changes in OC1 function/regulation and/or absence of cone-specific genes, and what does this reveal about photoreceptor evolution? 

The authors combine embryonic histology, electroporation, bulk RNA-seq, and molecular assays to explore the regulation of OC1 - a transcription factor central to cone specification in other vertebrates. They identify a novel OC1 splice isoform (LSOC1X2) containing a 48-amino-acid “spacer” sequence between its CUT and homeodomain regions, which is developmentally regulated and capable of activating a ThrbCRM1 reporter in a mouse retina assay. The study suggests that alternative OC1 splicing may underlie the skate’s cone-loss phenotype while preserving partial cone-like functional traits in rods.

Overall, the paper presents a creative molecular characterization of an evolutionarily informative model. The combination of comparative genomics, functional assays, and careful developmental staging adds novelty to our understanding of retinal evolution. However, some of the core conclusions, especially those connecting the OC1 isoform and the developmental data to cone loss - remain indirect. In several cases, the methods generate rich descriptive data but do not yet fully answer the central mechanistic question. With additional functional controls in skate, cell-type–resolved expression data, and clarified narrative structure, the manuscript could make a much stronger and more coherent contribution to developmental and evolutionary biology.

Major Comments

1. Functional significance of LSOC1X2 would benefit from a deeper interpretation - Both Onecut1 isoforms activate ThrbCRM1 in the mouse retina (Fig. 4), suggesting that the spacer insertion does not overtly disrupt canonical Onecut1 function. However, to improve reader understanding, the authors could consider clarifying what mechanistic or evolutionary advantage the spacer might provide, and how its developmental regulation could influence OC1 activity - for example, by altering cofactor-binding preferences, protein stability, or DNA-binding kinetics. A concise discussion outlining these plausible mechanisms would significantly strengthen the interpretation, even if the authors are not able to perform additional experiments. That said, functional tests (such as assessing whether the spacer modulates cofactor interactions or affects OC1 kinetics) would further support the proposed roles, should the authors choose to pursue them. Without such contextualization, it remains difficult to link this isoform to potential developmental or evolutionary roles.

2. It may help readers if the interpretation of the ThrbCRM1 reporter results were supported by some validation within the skate system. Showing that the reporter can be activated in skate by introducing mouse OC1 and OTX2, followed by skate OC1 with or without the spacer together with skate OTX2, would confirm that the assay context is permissive. Including a second OC1/OTX2-responsive enhancer as an orthogonal control would further ensure that any observed differences are biological rather than reporter-specific.

3. The RNA-seq data are bulk, so cell-specific differences in Onecut1 expression are inferred but not demonstrated. The Discussion notes this limitation, but it would strengthen the paper to add either in situ hybridization, single-cell validation, or at least computational deconvolution to identify likely expressing cell types for OC1, OTX2, Thrb/TRβ2, RXRγ, CRX, and NRL.

4. The alphafold models are visually appealing but lack quantitative metrics (e.g., confidence scores). Including these and clarifying whether differences between isoforms are within confidence thresholds would make this analysis more convincing.

Minor Comments

1. Figure 1 labeling: Fig. 1A–E currently lacks clear labels; adding channel labels, and staging info on the figure would help readers interpret the developmental characterization.

2. Introduction context: A brief paragraph summarizing the established roles of OC1 and OTX2 in cone and horizontal cell specification in chick/mouse would orient readers who are less familiar with this literature and clarify why these factors were chosen as candidates in skate.

3. The manuscript would benefit from clearer signposting of the question (“We ask whether…”), a short explanation of the OC1–OTX2–ThrbCRM1 framework for readers outside this niche, and a more linear connection between “question → experiment → result → interpretation.” Tightening paragraphs, reducing method details in the Results (with cross-references to Methods), and explicitly indicating why each experiment was done would significantly improve readability without changing the underlying science.

4. Nomenclature and formatting: Consistent formatting for genes vs. proteins, italicization of species names, and uniform use of “ThrbCRM1” will give the manuscript a more polished presentation.

5. Supplementary figures: Short, explicit captions describing how each supplemental figure supports specific main-text claims (e.g., “supports Fig. 2 by…”) would guide readers through the additional analyses.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.