A low-level Cdkn1c/p57 ^kip2 expression in spinal progenitors drives the transition from proliferative to neurogenic modes of division

Article activity feed

1. 

   Review Commons

   Apr 14, 2025

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   Reviewer #1 (Evidence, reproducibility and clarity (Required)):

   __* SUMMARY

   This study utilizes the developing chicken neural tube to assess the regulation of the balance between proliferative and neurogenic divisions in the vertebrate CNS. Using single-cell RNAseq and endogenous protein tagging, the authors identify Cdkn1c as a potential regulator of the transition towards neurogenic divisions. Cdkn1c knockdown and overexpression experiments suggest that low Cdkn1c expression enhances neurogenic divisions. Using a combination of clonal analysis and sequential knockdown, the authors find that Cdkn1c lengthens the G1 phase of the cell cycle via …

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   Reviewer #1 (Evidence, reproducibility and clarity (Required)):

   __* SUMMARY

   This study utilizes the developing chicken neural tube to assess the regulation of the balance between proliferative and neurogenic divisions in the vertebrate CNS. Using single-cell RNAseq and endogenous protein tagging, the authors identify Cdkn1c as a potential regulator of the transition towards neurogenic divisions. Cdkn1c knockdown and overexpression experiments suggest that low Cdkn1c expression enhances neurogenic divisions. Using a combination of clonal analysis and sequential knockdown, the authors find that Cdkn1c lengthens the G1 phase of the cell cycle via inhibition of cyclinD1. This study represents a significant advance in understanding how cells can transition between proliferative and asymmetric modes of division, the complex and varying roles of cycle regulators, and provides technical advance through innovative combination of existing tools.

   MAJOR AND MINOR COMMENTS *__

   * Overall * *Sample numbers are missing or unclear throughout for all imaging experiments. The authors should add numbers of cells analysed and/or numbers of embryos for their results to be appropriately convincing. *

   This information is now provided in the figure legends (numbers of cells analyzed and/or numbers of embryos) except for data in Figure 5, which are presented in a new Supplementary Table

   *Values and error bars on graphs must be defined throughout. Are the values means and error bars SD or SEM? *

   We have used SD throughout the study. This information has now been added in figure legends.

   Results 2

   * ____A reference should be provided for cell type distribution in spinal neural tube, where the authors state that cell bodies of progenitors reside within the ventricular zone. *

   We now cite a recent review on spinal cord development (Saade and E. Marti, Nature Reviews Neuroscience, 2025) to illustrate this point

   *The authors state that Cdkn1c "was expressed at low levels in a salt and pepper fashion in the ventricular zone, where the cell bodies of neural progenitors reside, and markedly increased in a domain immediately adjacent to this zone which is enriched in nascent neurons on their way to the mantle zone. In contrast, the transcript was completely excluded from the mantle zone, where HuC/D positive mature neurons accumulate." It is not clear if this is referring only to E4 or also to E3 embryos. Indeed, Cdkn1c expression appears to be much more salt and pepper at E3 and only resolves into a clear domain of high expression adjacent to the mantle zone at E4. It may be helpful if this expression pattern could be described in a bit more detail highlighting the changes that occur between E3 and E4. *

   We have now reformulated this paragraph as follows: "At E3, the transcript was expressed at low levels in a salt and pepper fashion in the ventricular zone, where the cell bodies of neural progenitors reside (Saade and Marti, 2025)). One day later, at E4, this salt and pepper expression was still detected in the ventricular zone, while it markedly increased in the region of the mantle zone that is immediately adjacent to the ventricular zone. This region is enriched in nascent neurons on their way to differentiation that are still HuC/D negative. In contrast, the transcript was completely excluded from the more basal region of the mantle zone, where mature HuC/D positive neurons accumulate.

   *It would be useful to annotate the ISH images in Fig 2A to show the ventricular and mantle zones as defined by immunofluorescence. *

   Thank you for the suggestion. We have now added a dotted line that separates the ventricular zone from the mantle zone at E3 and E4 in Figure 2A

   *Reference should be included for pRb expression dynamics. *

   This section has been rewritten in response to comments from Reviewer #3, and now contains several references regarding pRb expression dynamics. See detailed response to Reviewer #3 for the new version

   *Could the Myc tag insertion approach disrupt protein function or turnover? ____Why was the insertion target site at the C terminus chosen? *

   The first reason was practical: at the time when we decided to generate a KI in Cdkn1c, we had already generated several successful KIs at C-termini of other genes, in particular using the P2A-Gal4 approach (see Petit-Vargas et al, 2024), and had not yet experimented with N-terminal Gal4-P2A. We therefore decided to use the same approach for Cdkn1c.

   We also chose to target the C-terminus to avoid affecting the active CKI domain which is located at the N-terminus.

   Nevertheless, the C-terminal targeting may have an impact on the turnover: it has been described that CDK2 phosphorylation of a Threonin close to the C-terminus of Cdkn1c leads to its targeting for degradation by the proteasome from late G1 (Kamura et al, PNAS, 2003; doi: 10.1073/pnas.1831009100). We can therefore not rule out that the addition of the Myc tags close to this phosphorylation site modulates the dynamics of Cdkn1c degradation. We note, however, that we observed little overlap between the Cdkn1c-Myc and pRb signals in cycling progenitors, suggesting that Cdkn1c is effectively degraded from late G1.

   *OPTIONAL Could a similar approach be used to tag Cdkn1c with a fluorescent protein to enable live imaging of dynamics? *

   Although it could be done, we have not attempted to do this for CDKN1c because our current experience of endogenous tagging of several genes with a similar expression level (based on our scRNAseq data) and nuclear localization (Hes5, Pax7) with a fluorescent reporter shows that the fluorescent signal is extremely low or undetectable in live conditions; Therefore we favored the multi-Myc tagging approach, and indeed we find that the Myc signal in progenitors is also very low even though it is amplified by the immunohistology method; this suggests that most likely, the only signal that would be detected -if any- with a fluorescent approach would be the peak of expression in newborn neurons.

   *In suppl Fig 1C nlsGFP-positive cells are shown in the control shRNA condition. How can this be explained and does it impact the interpretation of the findings? *

   The reviewer refers to the control gRNA condition in panel C, that shows that two small patches of GFP-positive cells are visible in the whole spinal cord of this particular embryo.

   Technically, the origin of these "background" cells could be multiple. A spontaneous legitimate insertion at the CDKN1c locus by homologous recombination is possible, although we tend to think it is unlikely, given the extremely short length of the arms of homology; illegitimate insertions of the Myc-P2A-Gal4 cassette at off-target sites of the control gRNA is a possibility. Alternatively, a low-level leakage of Gal4 expression from the donor vector could lead to a detectable nls-GFP expression in a few cells via Gal4-UAS amplification.

   In any case, these cells are observed at a very low frequency (1 or 2 patches of cells/embryo) relative to the signal obtained in presence of the CDKN1c gRNA#1 (probably several thousand positive cells per embryo). This suggests that if similar "background" cells are also present in presence of the CDKN1c gRNA, they would not significantly contribute to the signal, and would not impact the interpretation.

   *In Fig 2B, there are a number of Myc labelled cells in the mantle zone, whereas the in situ images show no appreciable transcript expression. Is this because the protein but not the transcript is present in these cells? Could the authors comment on this? *

   It is indeed possible that the CDKN1c protein is more stable than the transcript in newborn neurons and remains detectable in the mantle zone after the mRNA disappears. In Gui et al, 2006, where they use an anti-CDKN1c antibody to label the protein in mouse spinal cord transverse sections at E11.5 (Figure 1B), a few positive cells are also visible basally. They could correspond to neurons that have not yet degraded CDKN1c, although it is unclear in the picture whether these cells are really in the mantle zone or in the adjacent dorsal root ganglion; we note that a similar differential expression dynamics between mRNA and protein has been described for Tis21/Btg2 in the developing mouse cortex, where the protein, but not the mRNA, is detected in some differentiated bIII-tubulin-positive neurons (Iacopetti et al, 1999).

   However, related to our response above to a previous comment from the same reviewer, we cannot rule out the possibility that the Myc tags modulate the turnover of CDKN1c protein and slow down the dynamics of its degradation in differentiating neurons.

   We have added a sentence to indicate the presence of these cells: "In addition, a few Myc-positive cells were located deeper in the mantle zone, where the transcript is no more present, suggesting that the protein is more stable than the transcript."

   *Results *

   *It should be mentioned how mRNA expression levels were quantified in the shRNA validation experiment (supp Fig 2A). *

   We did not quantify the level of mRNA reduction, it was just evaluated by eye. The reason for choosing shRNA1 for the whole study was dictated by 1) the fact that we more consistently saw (by eye) a reduction in the signal on the electroporated side with this construct than with the other shRNAs, and 2) that the effect on neurogenesis was also more consistent.

   We will perform additional experiments to provide some quantitation of the shRNA effect, as this is also requested by Reviewer #3.

   As our Cdkn1c KI approach offers a direct read-out of the protein levels in the ventricular and mantle zones, and since our shRNA strategy of "partial knock-down" is based on the idea that the shRNA effect should be more complete in progenitors expressing Cdkn1c at low levels than in newborn progenitors that express the protein at a higher level, we propose to validate the shRNA in the Cdkn1c-Myc knock-in background, by comparing the Myc signal intensity between control and Cdkn1c shRNA conditions

   *Figure panels are not currently cited in order. Citation or figure order could be changed. *

   We have now added a common citation of the panels referring to analyses at 24 and 48 hours after electroporation (now Figure 3A-F), allowing us to display the experimental data on the figure according to the timing post electroporation, while the text details the phenotype at the later time point first.

   *The authors should provide representative images for the graphs shown in Fig 3A and 3B. These could go into supplementary if the authors prefer. *

   We have added images in a revised version of the Figure 3, as requested

   *A supplementary figure showing the Caspase3 experiment should be added. *

   We have added data showing Caspase3 experiments in Supplementary Figure 3D

   *OPTIONAL. Identification of sister cells in the clonal analysis experiments is based on static images and cannot be guaranteed. Could live imaging be used to watch divisions followed by fixation and immunostaining to confirm identity? *

   We agree with the reviewer that direct tracking is the most direct method for the identification of pairs of sister cells. However, it remains technically challenging, and the added value compared to the retrospective identification would be limited, while requiring a great workload, especially considering the many different experimental conditions that we have explored in this study.

   Results 4

   *How did the authors quantify the intensity of endogenous Myc-tagged Cdkn1c to confirm the validity of the Pax7 locus knock in? Can they show that the expression level was consistently lower than the endogenous expression in neurons? Quantification and sample numbers should be shown. *

   We have not done these quantifications in the original version of the study. We will add a quantification of the signal intensity in the ventricular and mantle zones for the revised version of the manuscript, as also requested by reviewer #3.

   *In Fig 4B, the brightness of row 2 column 1 is lower than the same image in row 2 column 2, which is slightly misleading, since it makes the misexpressed expression level look lower than it is compared with endogenous in column 3. Is this because only a single z-section is being displayed in the zoomed in image? If so, this should be stated in the figure legend. *

   All images in the figure are single Z confocal images. Images in Column 2 (showing both electroporated sides of the same tube) were acquired with a 20x objective, whereas the insets shown in Columns 1 and 3 are 100x confocal images. 100x images on both sides were acquired with the same acquisition parameters, and the display parameters are the same for both images in the figure. The signal intensity can therefore be compared directly between columns 1 and 3.

   We have modified the legend of the Figure to indicate these points: "The insets shown in Columns 1 and 3 are 100x confocal images acquired in the same section and are presented with the same display parameters".

   *In Fig 4D, the increase in neurogenic divisions is mainly because of the rise in terminal NN divisions according to the graph, but no clear increase in PN divisions. Could the authors comment on the significance of this? *

   Our interpretation is that Pax7-CDKN1c misexpression experiments cause both PP to PN and PN to NN conversions. This is coherent with the classical idea of a progressive transition between these three modes of division in the spinal cord. Coincidentally, in our experimental conditions (timing of analysis and level of overexpression), the increase in PN resulting from PP to PN conversions is perfectly balanced by a decrease resulting from PN to NN conversions, giving the artificial impression that the PN compartment is unaffected. A less likely hypothesis would be that misexpression directly transforms symmetric PP into symmetric NN divisions, and that asymmetric PN divisions are insensitive to CDKN1c levels. We do not favor this hypothesis, because one would expect, in that case, that the shRNA approach would also not affect the PN compartment, and it is not what we have observed (see Figure 3H - previously 3F).

   We have modified the manuscript to elaborate on our interpretation of this result: "We observed an increase in the proportion of terminal neurogenic (NN) divisions and a decrease in proliferative (PP) divisions (Figure 4D). This suggests that CDKN1c premature expression in PP progenitors converts them to the PN mode of division, while the combined endogenous and Pax7-driven expression of CDKN1c converts PN progenitors to the NN mode of division. Coincidentally, at the stage analyzed, PP to PN conversions are balanced by PN to NN conversions, leaving the PN proportion artificially unchanged. The alternative interpretation of a direct conversion of symmetric PP into symmetric NN divisions is less likely, because the PN compartment was affected in the reciprocal CDKN1c shRNA approach (see Figure 3H)."

   *Results 5 ____The proportion of pRb-positive progenitors having entered S phase was stated to be higher at all time points; however, it is not significantly higher until 6h30 and is actually trending lower at 2h30. *

   Thank you for pointing this out. We have modified the sentence in the main text.

   "We found that the proportion of pRb positive progenitors having entered S phase (EdU positive cells) was significantly higher at all time points examined more than 4h30 after FT injection in the Cdkn1c knock-down condition compared to the control population (Figure 5D)"

   *OPTIONAL Could CyclinD1 activity be directly assessed? *

   This is an interesting suggestion. For example, using the fluorescent CDK4/6 sensor developed by Yang et al (eLife, 2020; https://doi.org/10.7554/eLife.44571) in a CDKN1c shRNA condition would represent an elegant experimental alternative to complement our rescue experiments with the double CDKN1c/CyclinD1 shRNA. However, we fear that setting up and calibrating such a tool for in vivo usage in the chick embryo represents too much of a challenge for incorporation in this study.

   *General ____Scale bars missing fig s1c s4d. *

   Thanks for pointing this out. Scale bars have been added in the figures and corresponding legends

   OPTIONAL Some of the main findings be replicated in another species, for example, mouse or human to examine whether the mechanism is conserved.

   *OPTIONAL Could use approaches other than image analysis be used to reinforce findings, for example biochemical methods, RNAseq or FACS? *

   We agree that it will be interesting and important that our findings are replicated in other species, experimental systems, and even tissues, or by alternative experimental approaches. Nevertheless, it is probably beyond the scope of this study.

   *A model cartoon to summarise outcomes would be useful. *

   We thank the reviewer for the suggestion. We will propose a summary cartoon for the revised version of the manuscript.

   *Unclear how cells were determined to be positive or negative for a label. Was this decided by eye? If so, how did the authors ensure that this was unbiased? *

   Positivity or negativity was decided by eye. However, for each experiment, we ensured that all images of perturbed conditions and the relevant controls were analyzed with the same display parameters and by the same experimenter to guarantee that the criteria to determine positivity or negativity were constant.

   *Reviewer #1 (Significance (Required)): *

   *SIGNIFICANCE *

   Strengths: This manuscript investigates the mechanisms regulating the switch from symmetric proliferative divisions to neurogenic division during vertebrate neuronal differentiation. This is a question of fundamental importance, the answer to which has eluded us so far. As such, the findings presented here are of significant value to the neurogenesis community and will be of broad interest to those interested in cell divisions and asymmetric cell fate acquisition. Specific strengths include:

   • * Variety of approaches used to manipulate and observe individual cell behaviour within a physiological context.*
   • * A limitation of using the chicken embryo is the lack of available antibodies for immunostaining. The authors take advantage of recent advances in chicken embryo CRISPR strategy to endogenously tag the target protein with Myc, to facilitate immunostaining. *
   • * Innovative combination of genetic and labelling tools to target cells, for example, use of FlashTag and EdU in combination to more accurately assess G1 length than the more commonly used method. *
   • Premature misexpression demonstrates that the previously observed dynamics indeed regulate cell fate.
   • * Mechanistic insight by examining downstream target CyclinD1. *
   • * Clearly presented with useful illustrations throughout. *
   • * Logic is clear and examination thorough. *
   • * Conclusions are warranted on the basis of their findings. * * ____Limitations ____T____his study primarily used visual analysis of fixed tissue images to assess the main outcomes. To reinforce the conclusions, these could be supplemented with live imaging to appreciate dynamics, or biochemical techniques to look at protein expression levels.*

   Some aspects of quantification require explanation in order for the experiments to be replicated.

   *It is imperative that precise sample sizes are included for all experiments presented. *

   Advance: ____First functional demonstration role for Cdkn1c in regulating neurogenic transition in progenitors.

   Conceptual advance suggesting Cdkn1c has dual roles in driving neurogenesis: promoting neurogenic divisions of progenitors and the established role of mediating cell cycle exit previously reported.

   *Technical advances in the form of G1 signposting and endogenous Myc tagging using CRISPR in chicken embryonic tissue. *

   Audience:

   *Of broad interest to developmental biologists. Could be relevant to cancer, since Cdkn1c is implicated. *

   *Please define your field of expertise with a few keywords to help the authors contextualize your point *

   Developmental biology, vertebrate embryonic development, neuronal differentiation, imaging. Please note that we have not commented on RNAseq experiments as these are outside of our area of expertise.

   *Reviewer #2 (Evidence, reproducibility and clarity (Required)): *

   *The work by Mida and colleagues addresses important questions about neurogenesis in the embryo, using the chicken neural tube as their model system. The authors investigate the mechanisms involved in the transition from stem cell self-renewal to neurogenic progenitor divisions, using a combination of single cell, gene functional and tracing studies. *

   The authors generated a new single cell data set from the embryonic chicken spinal cord and identify a transitory cell population undergoing neuronal differentiation, which expresses Tis21, Neurog2 and Cdkn1c amongst other genes. They then study the role of Cdkn1c and investigate the hypothesis that it plays a dual role in spinal cord neurogenesis: low levels favour transition from proliferative to neurogenic divisions and high levels drive cell cycle exit and neuronal differentiation.

   Major comments

   I have only a general comment related to the main point of the paper. The authors claim that Cdkn1c onset in cycling progenitor drives transition towards neurogenic modes of division, which is different from its role in cell cycle exit and differentiation. Figures 3F and 4D are key figures where the authors analysed PP, PN and NN mode of divisions via flash tag followed by analysis of sister cell fate. If their assumption is correct, shouldn't they also see, for example in Fig. 4D, an increase in PN or is this too transient to be observed or is it bypassed?

   As already stated in our response to a similar question from reviewer #1, our interpretation is that Pax7-CDKN1c misexpression experiments cause both PP to PN and PN to NN conversions. This is coherent with the classical idea of a progressive transition between these three modes of division in the spinal cord. Coincidentally, in our experimental conditions (timing of analysis and level of overexpression), the increase in PN resulting from PP to PN conversions is perfectly balanced by a decrease resulting from PN to NN conversions, giving the artificial impression that the PN compartment is unaffected. A less likely hypothesis would be that misexpression directly transforms symmetric PP into symmetric NN divisions, and that asymmetric PN divisions are insensitive to CDKN1c levels. We do not favor this hypothesis, because one would expect, in that case, that the shRNA approach would also not affect the PN compartment, and it is not what we have observed (see Figure 3H - previously 3F).

   *At the moment, the calculations of PN and NN frequencies are merged in the text, so perhaps describing PN and NN numbers separately will help better understand the dynamics of this gradual process (especially since there is little to no difference in PN). *

   Regarding the results of Pax7 overexpression presented in figure 4D (now Figure 4E in the revised version), we had made the choice to merge PN and NN values in the main text to focus on the neurogenic transition from PP to PN/NN collectively. We agree with this reviewer, as well as with reviewer #1, that it should be more detailed and better discussed. We therefore propose to modify the paragraph as follows (and as already indicated above in the response to reviewer #1):

   "We observed an increase in the proportion of terminal neurogenic (NN) divisions and a decrease in proliferative (PP) divisions (Figure 4D). This suggests that Cdkn1c premature expression in PP progenitors converts them to the PN mode of division, while the combined endogenous and Pax7-driven expression of Cdkn1c converts PN progenitors to the NN mode of division. Coincidentally, at the stage analyzed, PP to PN conversions are balanced by PN to NN conversions, leaving the PN proportion artificially unchanged. The alternative interpretation of a direct conversion of symmetric PP into symmetric NN divisions is less likely, because the PN compartment was affected in the reciprocal Cdkn1c shRNA approach (see Figure 3F, now 3H)."

   *Could the increase in NN be compatible also with a role in cell cycle exit and differentiation, for example from cells that have been targeted and are still undergoing the last division (hence marked by flash tag) or there won't be any GFP cells marked by flash tag a day after expression of high levels of Cdkn1c? *

   It is likely that a proportion of cells that would normally have done a NN division are pushed to a direct differentiation that bypasses their last division in the Pax7-CDKN1c condition, and that they contribute to the general increase in neuron production observed in our quantification 48hae (Figure 3F -previously 3C). However, these cases would not contribute to the increase in the NN quantification in pairs of sister cells 6 hours after division at 24hae (Figure 4E - previously 4D), because by design they would not incorporate FlashTag. The rise in NN is therefore the result of a PN to NN conversion.

   *Basically, what would the effect of expressing higher levels of Cdkn1c be? I guess this will really help them distinguish between transition to neurogenic division rather than neuronal differentiation. If not experimentally, any further comments on this would be appreciated. *

   These experiments have been performed and presented in the study by Gui et al., 2007, which we cite in the paper. Using a strong overexpression of CDKN1c from the CAGGS promoter, they showed a massive decrease in proliferation, assessed by BrdU incorporation, 24hours after electroporation. We will cite this result more explicitly in the main text, and better explain the difference of our approach. We propose the following modification

   « We next explored whether low Cdkn1c activity is sufficient to induce the transition to neurogenic modes of division. A previous study has shown that overexpression of Cdkn1c driven by the strong CAGGS promoter triggers cell cycle exit of chick spinal cord progenitors, revealed by a drastic loss of BrdU incorporation 1 day after electroporation (Gui et al., 2007). As this precludes the exploration of our hypothesis, we developed an alternative approach designed to prematurely induce a pulse of Cdkn1c in progenitors, with the aim to emulate in proliferative progenitors the modest level of expression observed in neurogenic progenitors.* We took advantage of the Pax7 locus, which is expressed in progenitors in the dorsal domain at a level similar to that observed for Cdkn1c in neurogenic precursors (Supplementary Figure 6A)."*

   Minor comments

   *Fig 3C my understanding is that HuC/D should be nuclear, but in fig 3C it seems more cytoplasmic (any comment?) *

   Some studies suggest that HuC/D can, under certain conditions, be observed in the nucleus of neurons. However, HuC/D is a RNA binding protein whose localization is mainly expected to be cytoplasmic. In our experience (Tozer et al, 2017), and in other publications using the antibody in the chick spinal cord (see, for example, le Dreau et al, 2014), it is observed in the cell body of differentiated neurons, as in the current manuscript.

   *Fig Suppl 3E (and related 4B), immuno for Cdkn1c-Myc: to help the reader understand the difference between the immuno signals when looking at the figure, I would suggest writing on the panel i) Pax7-Cdkn1c-Myc and ii) endogenous Cdkn1c-Myc, rather than 'misexpressed' and 'endogenous', which is slightly confusing (especially because what it is called endogenous expression is higher). *

   This has now been modified in the figures.

   *Literature citing: Introduction and discussion are very nicely written, although they could benefit from some more recent literature on the topic. For example, Cdkn1c role as a gatekeeper of stem cell reserve in the stomach, gut, (Lee et al, CellStemCell 2022 PMID: 35523142) or some other work on symmetric/asymmetric divisions and clonal analysis in zebrafish (Hevia et al, CellRep 2022 PMID: 35675784, Alexandre et al, NatNeur PMID: 20453852), mammals (Royal et al, Elife 2023 37882444, Appiah et al, EMBO rep 2023 PMID: 37382163). Also, similar work has been performed in the developing pancreatic epithelium, where mild expression of Cdkn1a under Sox9rtTa control was used to lengthen G1 without overt cell cycle exit and this resulted in Neurog3 stabilization and priming for endocrine differentiation (Krentz et al, DevCell 2017 PMID: 28441528), so similar mechanisms might be in in place to gradually shift progenitor towards stable decision to differentiate. Moreover, in the discussion, alongside Neurog2 control of Cdkn1c, it could be mentioned that the feedback loop between Cdk inhibitors and neurogenic factor is usually established via Cdk inhibitor-mediated inhibition of proneural bHLHs phosphorylation by CDKs (Krentz et al, DevCell 2017 PMID: 28441528, Ali et al, 24821983, Azzarelli et al 2017 - PMID: 28457793; 2024 - PMID:39575884). Further, in the discussion, could they mention anything about the following open questions: is there evidence for Cdkn1c low/high expression in mammalian spinal cord? Or maybe of other Cdk inhibitors? Is Cdkn1c also involved in cell cycle exit during gliogenesis? Or is there another Cdk inhibitor expressed at later developmental stages, hence linking this with specific cell fate decisions? *

   We will modify the introduction and discussion in several instances, in order to address the above suggestions and we will:

   • add references to its role in other contexts and/or species.

   • expand the discussion on the cross talk between neurogenic factors and CDK inhibitors in other cellular contexts.

   • add a dedicated paragraph in the discussion to answer reviewer#2's questions: is there evidence for Cdkn1c low/high expression in mammalian spinal cord? Or maybe of other Cdk inhibitors? Is Cdkn1c also involved in cell cycle exit during gliogenesis or is there another Cdk inhibitor expressed at later developmental stages?

   Reviewer #2 (Significance (Required)):

   The work here presented has important implications on neural development and its disorders. The authors used the most advanced technologies to perform gene functional studies, such as CRISPR-HDR insertion of Myc-tag to follow endogenous expression, or expression under endogenous Pax7 promoter, often followed by flash tag experiments to trace sister cell fate, and all of this in an in vivo system. They then tested cell cycle parameters, clonal behaviour and modes of cell division in a very accurate way. Overall data are convincing and beautifully presented. The limitation is potentially in the resolution between the events of switching to neurogenic division versus neuronal differentiation, which might just warrant further discussion. This work advances our knowledge on vertebrate neurogenesis, by investigating a key player in proliferation and differentiation.

   * ____I believe this work will be of general interest to developmental and cellular biologists in different fields. Because it addresses fundamental questions about the coordination between cell cycle and differentiation and fate decision making, some basic concepts can be translated to other tissues and other species, thus increasing the potential interested audience.*

   *My work focuses on stem cell fate decisions in mammalian systems, and I am familiar with the molecular underpinnings of the work here presented. However, I am not an expert in the chicken spinal cord as a model and yet the manuscript was interesting. I am also not sufficiently expert in the bioinformatic analysis, so cannot comment on the technical aspects of Figure 1 and the way they decided to annotate their data. *

   __*

   Reviewer #3 (Evidence, reproducibility and clarity (Required)): *__

   *Summary: In this study, Mida et al. analyze large-scale single-cell RNA-seq data from the chick embryonic neural tube and identify Cdkn1c as a key molecular regulator of the transition from proliferative to neurogenic cell divisions, marking the onset of neurogenesis in the developing CNS. To confirm this hypothesis, they employed classical techniques, including the quantification of neural cell-specific markers combined with the flashTAG label, to track and isolate isochronic cohorts of newborn cells in different division modes. Their findings reveal that Cdkn1c expression begins at low levels in neurogenic progenitors and becomes highly expressed in nascent neurons. Using a classical knockdown strategy based on short hairpin RNA (shRNA) interference, they demonstrate that Cdkn1c suppression promotes proliferative divisions, reducing neuron formation. Conversely, novel genetic manipulation techniques inducing low-level CDKN1c misexpression drive progenitors into neurogenic divisions prematurely. *

   *By employing cumulative EdU incorporation assays and shRNA-based loss-of-function approaches, Mida et al. further show that Cdkn1c extends the G1 phase by inhibiting cyclin D, ultimately concluding that Cdkn1c plays a dual role: first facilitating the transition of progenitors into neurogenic divisions at low expression levels, and later promoting cell cycle exit to ensure proper neural development. *

   *This study presents several ambiguities and lacks precision in its analytical methodologies and quantification approaches, which contribute to confusion and potential bias. To enhance the reliability of the conclusions, a more rigorous validation of the methods employed is essential. *

   *This study introduces a novel approach to tracking the fate of sister cells from neural progenitor divisions to infer the division modes. While previous methods for analyzing the division mode of neural progenitor cells have been implemented, rigorous validation of the approach introduced by Mida et al. is necessary. Furthermore, the concept of cell cycle regulators interacting to control the duration of specific cell cycle stages and influencing progenitor cell division modes has been explored before, potentially limiting the novelty of these findings. *

   Major comments:

   *1.-The study presents ambiguity and lacks precision in quantifying neural precursor division modes. The authors use phosphorylated retinoblastoma protein (pRb) as a marker for neurogenic progenitors, claiming its reliability in identifying neurogenic divisions. *

   *However, they do not provide a thorough characterization of pRb expression in the developing chick neural tube, leaving its suitability as a neurogenic division marker unverified. *

   Throughout their comments on the manuscript, this reviewer raises several points regarding the characterization of pRb expression in our model and of our use of this marker in our study. We take these comments into account and propose to expand on pRb characteristics in the first occurrence of pRb as a marker of cycling cells in the manuscript. The modifications rely on:

   • the quotation of several studies showing that phosphorylation of Rb is regulated during the cell cycle, and that "it is not detectable during a period of variable length in early G1 in several cell types (Moser et al, 2018;Spencer et al, 2013; Gookin et al, 2017), including neural progenitors in the developing chick spinal cord (Molina et al, 2022). Apart from this absence in early G1, pRb is detected throughout the rest of the cell cycle until mitosis".

   • a more detailed description of our own characterization of pRb dynamics in a synchronous cohort of cycling cells, which reveals a similar heterogeneity in the timing of the onset of Rb phosphorylation after mitosis. This description was initially shown in supplementary figure 3 and will be transferred to a new supplementary figure 2 to account for the fact that it will now be cited earlier in the manuscript.

   Regarding the specific question the "suitability (of pRb) as a neurogenic division marker": we do not directly "use phosphorylated retinoblastoma protein (pRb) as a marker for neurogenic progenitors", but we use Rb phosphorylation to discriminate between progenitors (pRb+) and neurons (pRb-) identity in pairs of sister cells to retrospectively identify the mode of division of their mother.

   Given that Rb is unphosphorylated during a period of variable length after mitosis (see references above), pRb is not a reliable marker of ALL cycling progenitors. We developed an assay to identify the timepoint (the maximal length of this "pRb-negative" phase) after which Rb is phosphorylated in all cycling progenitors (new Supplementary Figure 2). This assay relies on a time course of pRb detection in cohorts of FlashTag-positive pairs of sister cells born at E3. This time course experiment allowed us to identify a plateau after which the proportion of pRb-positive cells in the cohort remains constant. From this timepoint, this proportion corresponds to the proportion of cycling cells in the cohort. Rb phosphorylation therefore becomes a discriminating factor between cycling progenitors (pRb+) and non-cycling neurons (pRb-).

   We are confident that this provides a solid foundation for the determination of the identity of pairs of sister cells in all our Flash-Tag based assays, which retrospectively identify the mode of division of a progenitor on the basis of the phosphorylation status of its daughter cells 6 hours after division.

   We propose to modify the main text to describe the strategy and protocol more explicitly, by introducing the sentence highlighted in yellow in the following paragraph where the paired-cell analysis is first introduced (in the section on CDKN1c knock-down):

   "This approach allows to retrospectively deduce the mode of division used by the mother progenitor cell. We injected the cell permeant dye "FlashTag" (FT) at E3 to specifically label a cohort of progenitors that undergoes mitosis synchronously (Baek et al., 2018; Telley et al., 2016 and see Methods), and let them develop for 6 hours before analyzing the fate of their progeny using pRb immunoreactivity (Figure 3D). Our characterization of pRb immunoreactivity in the tissue had established beforehand that 6 hours after mitosis, all progenitors can reliably be detected with this marker (Supplementary Figure 2, Methods). *Therefore, at this timepoint after FT injection, *two-cell clones selected on the basis of FT incorporation can be categorized as PP, PN, or NN based on pRb positivity (P) or not (N) (see Methods, new Figure 3G and new Supplementary Figures 2 and 4)."

   We also modified accordingly the legend to Supplementary Figure 2 (previously Supplementary Figure 3, which describes the identification of the plateau of pRb.

   *Furthermore, retinoblastoma protein (Rb) and cyclin D interact crucially to regulate the G1/S phase transition of the cell cycle, with cyclin D/CDK complexes phosphorylating Rb. Since the authors conclude that CDKN1c primarily acts by inhibiting the cyclin D/CDK6 complex, it is likely that CDKN1c influences pRb expression or phosphorylation state. This raises the possibility that pRb could be a direct target of CDKN1c, whose expression and phosphorylation would be altered in gain-of-function (GOF) and loss-of-function (LOF) analyses of CDKN1c. *

   *In light of this, it would be more appropriate to consider pRb as a CDKN1c target and discuss the molecular mechanisms regulating cell cycle components. *

   We agree with the reviewer that Rb phosphorylation may be a direct or indirect target of Cdkn1c activity, and exploring the molecular aspects of the cellular and developmental phenomena that we describe in our manuscript would represent an interesting follow up study.

   * ____A more precise approach would involve using other markers or targets to quantify neural precursor division modes at earlier stages of neurogenesis. *

   To complement our analyses of the modes of division, we propose to use a positive marker to assess neural identity in parallel to the absence of pRb within pairs of cells. This approach may be the most meaningful in the gain of function context (Pax7 driven expression of Cdkn1c) because in this context, the time-point to reach the plateau of Rb phosphorylation used in our FT-based assay may indeed be delayed. On the opposite, in the context of loss of functions, the plateau may be reached earlier, which would have no effect on this assay.

   *2.-Furthermore, the study employs FlashTag labeling to track daughter cells post-division, but the 16-hour post-injection window may result in misidentification of sister cells due to the potential presence of FlashTagged cells that did not originate from the same division. *

   *This introduces a risk of bias in quantification, data misinterpretation, and potential errors in defining division modes. A more rigorous validation of the FlashTag strategy and its specificity in tracking division pairs is necessary to ensure the reliability of their conclusions. *

   The reviewer probably mistyped and meant 6-hour post injection, which is the duration that we use for paired cell tracking. We would like to emphasize that in addition to the FlashTag label, we benefit from the electroporation reporter to assess clonality. Altogether, we combine 5 criteria to define a clonal relationship :

   • 2 cells are positive for Flash Tag
   • The Flash Tag intensity is similar between the 2 cells
   • The 2 cells are positive for the electroporation reporter
   • The electroporation reporter intensity is similar between the two cells
   • the position of the two cells is consistent with the radial organization of clones in this tissue (Leber and Sanes, 1995;__; __Loulier et al, 2014): they are found on a shared line along the apico-basal axis, and share the same Dorso-Ventral and Antero-Posterior position . This combination is already described in the Methods section. We propose to modify the paragraph to include the sentence highlighted in yellow in the text below;

   "Cell identity of transfected GFP positive cells was determined as follows: cells positive for pRb and FT were classified as progenitors and cells positive for FT and negative for pRb as neurons. In addition, a similar intensity of both the GFP and FT signals within pairs of cells, and* a relative position of the two cells consistent with the radial organization of clones in this tissue (Leber and Sanes, 1995; Loulier et al, 2014)* were used as criteria to further ascertain sisterhood. This combination restricts the density of events fulfilling all these independent criteria, and can confidently be used to ensure a robust identification of pairs of sister cells."

   *3.- The knock-in strategy used to tag the endogenous CDKN1c protein in Figure 2 is an elegant tool to infer protein dynamics in vivo. However, since strong conclusions regarding CDKN1c dynamics during the cell cycle are drawn from this section, it would be advisable to strengthen the results by including quantification with adequate replication and proper statistical analysis, as the current findings are preliminary and somewhat speculative. *

   *- "Although pRb is specific for cycling cells, it is only detected once cells have passed the point of restriction during the G1 phase." Please provide literary reference confirming this observation. *

   We have entirely remodeled this section, which describes the expression of Myc-tagged Cdkn1c relative to pRb and now provide several references that describe the generally accepted view that pRb is specific of cycling cells, regulated during the cell cycle, and in particular absent in early G1. We also remove the mention of the "Restriction point" in the main text to avoid any confusion on the timing of phosphorylation, as the notion of restriction point is not useful in our study. The section now reads as follows:

   *"To ascertain that **Cdkn1c **is translated in neural progenitors, we used an anti-pRb antibody, recognizing a phosphorylated form of the Retinoblastoma (Rb) protein that is specifically detected in cycling cells (Gookin et al., 2017; Moser et al., 2018; Spencer et al., 2013) , including neural progenitors of the developing chick spinal cord (Molina et al., 2022). In the ventricular zone of transverse sections at E4 (48hae), we detected triple Cdkn1c-Myc/GFP/pRb positive cells (arrowheads in Figure 2B), providing direct evidence for the Cdkn1c protein in cycling progenitors. We also observed many double GFP/pRb positive cells that were Myc negative (arrowheads in Figure 2B). The observation of UAS-driven GFP in these pRb-positive cells is evidence for the translation of Gal4 and therefore provides a complementary demonstration that the **Cdkn1c **transcript is translated in progenitors. The absence of Myc detection in these double GFP/pRb positive cells also suggests that Cdkn1c/Cdkn1c-Myc stability is regulated during the cell cycle. *

   *Finally, we observed double Myc/GFP-positive cells that were pRb-negative (Figure 2B; asterisks). One characteristic of Rb phosphorylation as a marker of cycling cells is a period in early G1 during which it is not detectable, as described in several cell types (Gookin et al., 2017; Moser et al., 2018; Spencer et al., 2013) including chick spinal cord neural progenitors (Molina et al., 2022). Using a method that specifically labels a synchronous cohort of dividing cells in the neural tube, we similarly observed a period in early G1 during which pRb is not detectable in some progenitors at E3 (See Supplementary Figure 2 and Methods). *Hence, the double Myc/GFP positive and pRb negative cells may correspond to progenitors in early G1. Alternatively, they may be nascent neurons whose cell body has not yet translocated basally (see Figure 2C). Finally, we observed a pool of GFP positive/pRb negative nuclei with a strong Myc signal in the region of the mantle zone that is in direct contact with the ventricular zone (VZ), corresponding to the region where the transcript is most strongly detected (see Figure 2A). This pool of cells with a high Cdkn1c expression likely corresponds to immature neurons exiting the cell cycle and on their way to differentiation (Figure 2B; double asterisks). In addition, a few Myc positive cells were located deeper in the mantle zone, where the transcript is no more present, suggesting that the protein is more stable than the transcript.

   In summary, our dual Myc and Gal4 knock-in strategy which reveals the history of Cdkn1c transcription and translation confirms that Cdkn1c is expressed at low level in a subset of progenitors in the chick spinal neural tube, as previously suggested (Gui et al., 2007; Mairet-Coello et al., 2012). In addition, the restricted overlap of Cdkn1c-Myc detection with Rb phosphorylation suggests that in progenitors, Cdkn1c is degraded during or after G1 completion. "

   This section will again be remodeled in a future revised version of the manuscript, in which we will add quantifications of Myc levels, as requested by Reviewer 1 above, and also by Reviewer #3 below.

   *Given that pRb immunoreactivity is used as a marker for cycling progenitors to base many of the results of this study, it would be very valuable to characterize the dynamics of pRb in cycling cells in the studied tissue, for instance combined with the cell cycle reporter used by Molina et al. (Development 2022). *

   In the original version of the manuscript, the section describing the dynamics of CDKN1c-Myc in the KI experiments presented in Figure 2 relied on the idea that the dynamics of pRb in chick spinal progenitors is similar to what I described in other tissues and cell types, without providing any references to substantiate this fact. Actually, Molina et al provide a characterization of pRb in combination with their cell cycle reporter and conclude that pRb negative progenitors are in G1 ("We also verified that phospho-Rb- and HuC/D-negative cells were in G1 by using our FUCCI G1 and PCNA reporters"). We will now cite this reference to support our claim. In addition, our characterization of Rb progressive phosphorylation in the synchronic Flash-Tag cohort of newborn sister cells provides a complementary demonstration that a fraction of the progenitors are pRb-negative when they exit mitosis (i.e. in early G1). This analysis was initially only introduced in the supplementary Figure 3, as support for the section that presents the Paired-cell assay used in Figure 3. We propose to introduce the data from Supplementary Figure 3 earlier in the manuscript (now Supplementary Figure 2), in order to better introduce the reader with the dynamics of pRb in cycling cells in our model. This will better support our description of the Cdkn1c-Myc dynamics in relation with pRb. We therefore propose to reformulate this whole section as follows.

   - It would be valuable to analyse the dynamics of Myc immunoreactivity in combination of pRb in all three gRNAs (highlighted in Supplementary Figure 1), as it would be a strong point in favour that the dynamics reflect the endogenous CDKN1c dynamics.

   *- It would be very valuable to provide a quantification of said dynamics (e.g. plotting myc intensity / pRb immunoreactivity along the apicobasal axis of the tissue). *

   These are two interesting suggestions. To complement our data with guide #1, we have performed Myc-immunostaining experiments on transverse sections in the context of guide #3, showing exactly the same pattern of Myc signal, with low expression in the VZ, and a peak of signal in the part of the mantle zone that is immediately touching the VZ. This confirms the specificity of the spatial distribution of the Cdkn1c-Myc signal. These data have been added in a revised version of Supplementary Figure 1.

   We will perform the suggested quantifications using guides #1 and #3, which both show a good KI efficiency. We do not think it is useful to do these experiments with guide #2, whose efficiency is much lower, and which would lead to a very sparse signal.

   *- The characterization of dynamics is performed only with one of the gRNAs (#1) on the basis that it produces the strongest NLS-GFP signal, as a proxy for guide efficiency. It would be nice if the authors could validate guide cutting efficiency via sequencing (e.g. using a Cas9-T2A-GFP plasmid and sorting for positive cells). *

   We will perform these experiments to validate guide cutting efficiency using the Tide method (Brinkman et al, 2014)

   *- In order to make sure that the dynamics inferred from Myc-tag immunoreactivity do reflect the cell cycle dynamics of CDKN1c-myc, it would be advisable to confirm in-frame insertion of the myc-tag sequence. *

   We will perform genomic PCR experiments to confirm in-frame insertion of the Myc tags at the Cdkn1c locus

   4.- In Figure 3, the authors use a short-hairpin-mediated knock-down strategy to decrease the levels of Cdkn1c, and show that this manipulation leads to an increase percentage of cycling progenitors and a decrease in the number of neurons in electroporated cells.

   *The authors claim that their shRNA-based knockdown strategy aims to reduce low-level Cdkn1c expression in neurogenic progenitors while minimally affecting the higher expression in newborn neurons required for cell cycle exit. However, several factors need consideration. Electroporation introduces variability in shRNA delivery, making it difficult to achieve consistent gene inhibition across all cells, especially for dose-dependent genes like Cdkn1c. *

   *Additionally, Cdkn1c generates multiple isoforms, which may not be fully annotated in the chick genome, raising the possibility that the shRNA targets specific isoforms, potentially explaining the observed low expression. *

   All the predicted isoforms in the chick genome contain the sequence targeted by shRNA1, which is located in the CKI domain, the region of the protein that is most conserved between species. Besides, all the isoforms annotated in the mouse and human genomes also contain the region targeted by shRNA1. We are therefore confident that shRNA1 should target all chick isoforms.

   *A more rigorous approach, such as qPCR analysis of sorted electroporated cells, would better validate the expression levels, rather than relying on in situ hybridization, presenting electroporated and non-electroporated cells in the same section (Supp. Figure 2). *

   This approach (qRT-PCR on sorted cells) would enable us to focus solely on electroporated cells, but it would result in an averaged quantification of Cdkn1c depletion. In order to obtain additional information on the shRNA-dependent decrease in Cdkn1C in the different neural cell populations (progenitor versus differentiating neuron), we propose an alternative approach consisting in monitoring the level of Cdkn1c protein, assessed through Cdkn1c-Myc signal in knock-in cells, in the presence versus absence of Cdkn1c shRNA.

   *- As the authors note, "Unambiguous identification of cycling progenitors and postmitotic neurons is notoriously difficult in the chick spinal cord". "markers of progenitors usually either do not label all the phases of the cell cycle (eg. Phospho-Rb, thereafter pRb), or persist transiently in newborn neurons (eg. Sox2)." Given that pRb immunoreactivity is used as the basis for a lot of the conclusions in this study, it would be valuable to add a characterization of its dynamics as mentioned in Figure 2, as well as provide literary references/proof that Sox2 expression persists in newborn neurons. *

   We have addressed the case of pRb dynamics in progenitors above and added a reference documented pRb expression during the cell cycle of chick neural progenitors (Molina et al, 2022).

   Regarding Sox2 persistence: we consistently detect a small fraction of double positive Sox2+/HuC/D+ cells in chick spinal cord transverse sections. We have shown that this marker of differentiating neurons (HuC/D) only becomes detectable more than 8 hours after mitosis in newborn neurons at E3 (Baek et al, 2018), indicating that Sox2 protein can persist for up to at least 8 hours in newborn neurons.

   We now cite a paper showing that a similar persistence of Sox2 protein is reported in differentiating neurons of the human neocortex, where double Sox2/NeuN positive cells are frequently observed in cerebral organoids (Coquand et al, Nature Cell Biology 2024__)__

   *- The undefined population (pRb-/HuCD-) introduces an unknown that assumes that the percentage of progenitors in G1 phase before the restriction point and the number of newborn neurons are equal for both conditions in an experiment. Can the authors provide explanation for this assumption? *

   We do not think that these numbers are equal for both conditions, and we did not formulate this assumption. We only indicate (in the methods section) that this undefined/undetermined population (based on negativity for both markers) is a mix of two possible cell types. However, we do not offer any interpretation of the CDKN1c phenotypes based on the changes in this population. Indeed, our interpretation of the knock-down phenotype is solely based on the increase in pRb-positive and decrease in HuC/D-positive cells, which both suggest a delay in neurogenesis. We understand from the reviewer's comment that depicting an "undefined" population on the graph may cause some confusion. We therefore propose to present the data on pRb and HuC/D in different graphs, rather than on a combined plot, and to remove the reference to undefined cells in Figure 3, as well as in Figures 4 and 5 depicting the gain of function and double knock-down experiments. We have implemented these changes in updated versions of the figures.

   *- In Gui et al. (Dev Biol 2006), authors showed that a knockdown of Cdkn1c leads to a failure of nascent neurons to exit the cell cycle and causes them to re-entry the cell cycle, shown by ectopic mitoses. In that study, cells born from those ectopic mitoses eventually leave the cell cycle leading to an increase in the number of neurons. Can the authors check for ectopic mitoses at 24hpe and 48hpe? *

   We have now performed experiments with an anti phospho Histone 3 antibody, which labels mitotic cells, at 24 and 48 hours post electroporation. We do not see any ectopic mitoses upon Cdkn1c knock-down with this marker, and we have produced a Supplementary Figure with these data. This is consistent with the fact that we also do not see ectopic pRb or Sox2 positive cells in the mantle zone in the knock-down experiments. These data (pH3 and Sox2) have been added in the new Supplementary Figure 3E and F.

   We have now modified the main text to include these data:

   "In the context of a full knock-out of Cdkn1c in the mouse spinal cord, a reduction in neurogenesis was also observed, which was attributed to a failure of prospective neurons to exit the cell cycle, resulting in the observation of ectopic mitoses in the mantle zone (Gui et al, 2007). In contrast with this phenotype, using an anti phospho-Histone3 antibody, we did not observe any ectopic mitoses 24 or 48 hours after electroporation in our knock-down condition (Supplementary Figure 3E-F). This is consistent with the fact that we also do not observe ectopic cycling cells with pRb (Figure 3A and D) and Sox2 (Supplementary Figure 3E-F) antibodies. We therefore postulated that the reduced neurogenesis that we observe upon a partial Cdkn1c knock-down may result from a delayed transition of progenitors from the proliferative to neurogenic modes of division."

   *- The authors then address the question of whether the decrease in neuron number is due to the failure of newborn neurons to exit the cell cycle or to a delay in the transition from proliferative to neurogenic divisions. For that, they implement a strategy to label a synchronized cohort of progenitors based of incorporation of a FlashTag dye. *

   *- Given that this strategy is the basis of many of the experiments in this article, it would be very valuable to expand on the validation of this technique as cited in major comment #2. In figure 3E, the close proximity of cell pairs in PP and PN clones shown in the pictures makes their sibling status apparent. However, this is not the case for the NN clone. Can the authors further explain with what criteria they determined the clonal status of two FlashTag labelled cells? *

   The key criterion for cells that are not directly touching each other is that their relative position corresponds to the classical "radial" organization of clones in this tissue (Leber and Sanes, 1995__; __Loulier et al, Neuron, 2014). In other words, we make sure that they are located on a same apico-basal axis, as is the case for the NN clone presented on the figure. As stated above in our response to major comment #2, we have modified the Methods section accordingly.

   *Can they provide further image examples of different types of clones? *

   We now provide additional examples in a new Supplementary Figure 4

   *- Can the authors show that the plateau reached in Sup Figure 3 for pRb immunoreactivity corresponds to a similar dynamic for HuC/D immunoreactivity? *

   The plateau for Rb phosphorylation in progenitors is reached before 6 hours post mitosis at E3. At the same age, we have previously shown (Baek et al, PLoS Biology 2018) in a similar time course experiment in pairs of FT+ cells that the HuC/D signal is not detected in newborn neurons 8 hours after mitosis. HuC/D only starts to appear between 8 and 12 hours, and still increases between 8 and 16 hours. The plateau would therefore be very delayed for HuC/D compared to pRb. This long delay in the appearance of this « positive » marker of neural differentiation is the main reason why we chose to use Rb phosphorylation status for the analysis of synchronous cohorts of pairs of sister cells, because pRb becomes a discriminating factor much earlier than HuC/D after mitosis.

   *- In order to further validate the strategy, could the authors use it at different stages to validate if they can replicate the different percentages of PP/PN/NN reported in the literature (e.g. Saade Cell Rep 2013)? *

   We have carried out similar experiments at E2, showing a plateau of 95% of pRb-positive cells in the FT-positive population (see graph on the right). This provides a retrospective estimate of the mode of division of the mother cells at this stage (roughly 90% of PP and 10% of PN) which is consistent with the vast majority of PP divisions described by Saade et al (2013, see Figure S1) at this stage.

   *5.- In Figure 4, the strategy used to induce a low-dose overexpression of CDKN1c is an elegant method to introduce CDKN1c-Myc expression under the control of the endogenous Pax7 promoter, active in proliferative progenitors. The main point to address is: *

   *- Please provide proof that Pax7 expression is not altered in guides with a successful knock-in event (e.g. sorting and WB against the Pax7 protein) or the immunohistochemistry as performed in the Pax7-P2A-Gal4 tagging in Petit-Vargas et al., 2024. *

   We have now performed Pax7 immunostainings on transverse sections at 24 and 48 hours post electroporation, both with the Pax7-CDKN1c-Gal4 and with the Pax7-Gal4 control constructs. We present these data in the new supplementary figure 7. In both conditions, we find that the Pax7 protein is still present in KI-positive cells. We observe a modest increase in Pax7 signal intensity in these cells, suggesting either that the insertion of exogenous sequences stabilizes the Pax7 transcript, or that the C-terminal modification of Pax7 protein with the P2A tag increases its stability. This does not affect the interpretation of the CDKN1c overexpression phenotype, because we used the Pax7-Gal4 construct that shows the same modification of Pax7 stability as a control for this experiment. We have introduced this comment in the legend of Supplementary Figure 7.

   *- Given the cell cycle regulated expression and activity of CDKN1c, can the authors elaborate on whether this is regulated at the promoter level? *

   Cdkn1c transcription is regulated by multiple transcription factors and non-coding RNAs (see for example Creff and Besson, 2020, or Rossi et al, 2018 for a review). To our knowledge, these studies focus more on the regulation of Cdkn1c global expression than on the regulation of its levels during cell cycle progression. Although it is very likely that transcriptional regulation contributes, post-translational regulation, and in particular degradation by the proteasome, is also a key factor in the cell cycle regulation of Cdkn1c activity

   *If so, how does this differ from the promoter activity of Pax7? *

   The transcriptional regulation of Pax7 and Cdkn1c is probably controlled by different regulators, since their expression profiles are very different. Regardless of the mechanisms that control their expression, the rationale for choosing Pax7 as a driver for Cdkn1c expression was that Pax7 expression precedes that of Cdkn1c in the progenitor population, and that it disappears in newborn neurons, when that of Cdkn1c peaks. This provided us with a way to advance the timing of Cdkn1c expression onset in proliferative progenitors.

   *- It would be advisable to characterize the dynamics along the cell cycle for the overexpressed form of CDKN1c-Myc relative to pRb, similarly to what was done in Figure 2B. *

   We will carry out experiments similar to those shown in Figure 2B in order to characterise the dynamics of Cdkn1c in a context of overexpression, in relation to pRb.

   In addition, we will include a more precise quantification of the "misexpressed" compared to "endogenous" Cdkn1c -Myc levels, as already mentioned in the answer to a request by reviewer1.

   6.-In figure 5, the authors use a double knock-down strategy to test the hypothesis that the effect of Cdkn1c in G1 length is partially at least through its inhibition of CyclinD1. Results show that double shRNA-mediated knock-down of CyclinD1 and Cdkn1c counteracts the effects of Cdkn1c-sh alone on EdU incorporation, PP/PN/NN cell divisions and overall rations of progenitors and neurons.

   *- In the measurement of progenitor cell cycle length in Figure 5A, it would be more appropriate to present the nonlinear regression method described by Nowakowski et al. (1989), as has been commonly used in the field (Saade et al., 2013, PMID: 23891002, Le Dreau et al., 2014, PMID: 24515346, Arai et al., 2011, PMID: 21224845). *

   The Nowakowski non linear regression method has been used often in the literature in the same tissue, and is generally used to calculate fixed values for Tc, Ts, etc... This method is based on several selective criteria, and in particular the assumption that "all of the cells have the same cycle times". Yet, many studies have documented that cell cycle parameters change during the transition from proliferative to neurogenic modes of division during which our analysis is performed; live imaging data in the chick spinal cord have illustrated very different cell cycle durations at a given time point (see Molina et al). We therefore think that the proposed formulas do not reflect the heterogenous reality of neural progenitors of the embryonic spinal cord. However, the cumulative approach described by Nowakowski is useful to show qualitative differences between populations (e.g. a global decrease of the cycle length, like in our comparison between control and shRNA conditions). For these reasons, we prefer to display only the raw measurements rather than the regression curves.

   *- Cumulative EdU incorporation in spinal progenitors (pRb-positive) at E3 (24 hours after injection) showed that the proportion of EdU-positive progenitors reached a plateau at 14 hours in control conditions, which is later than what has been reported in Le Dreau et al., 2014 (PMID: 24515346). Can you explain why? *

   Le Dreau et al count the EdU+ proportion of cells in the total population of electroporated cells located in the VZ (which includes progenitors, but also future neurons that have been labelled during the previous cycles -at least for the time points after 2hours- and have not yet translocated to the mantle zone), whereas we only consider pRb+ progenitors in the analysis. In addition, the experiments are not performed at the same developmental stage. Altogether, this may account for the different curves obtained in our study.

   *- It would be interesting to measure G1 length as in Figure 5D for the double cdkn1c-sh - ccnd1-sh knock down condition, to see if it rescues G1 length. As well as in the Ccnd1 knock down condition alone to see if it increases G1 length in this context as well. *

   We will perform cumulative EDU incorporation experiments similar to that shown in Figure 5D to measure G1 length for the cdkn1c-sh - ccnd1-sh knock down double conditions, as well as in the Ccnd1 knock down condition alone.

   *Minor comments *

   __*Introduction:

   • The introduction should include references of studies of the role of Cdkn1c in cortical development (Imaizumi et al. Sci Rep 2020, Colasante et al. Cereb Cortex 2015, Laukoter et al. ____Nature Communications 2020).*__

   We will modify the introduction in several instances, in order to address suggestions by Reviewers #2 (see above) and #3, in particular to expand the description of the role of Cdkn1c during cortical development

   1) Transcriptional signature of the neurogenic transition (Figure 1).

   *- In the result section, it would be informative to include the genes used to determine the progenitor and neuron score (instead of in Methods). *

   We have now listed the genes used to determine the progenitor and neuron score in the main text of the result section

   *- Figure 1A. It would be informative to add in the diagram what "filtering" means (eg. Neural crest cells). *

   We have now added the detail of what 'filtering' means in the diagram

   *- In the result section, "However, while Tis21 expression is switched off in neurons, Cdkn1c transiently peaks at high levels in nascent neurons before fading off in more mature cells." Missing literary reference or data to clearly demonstrate this point. *

   We have reworded this sentence, adding a reference to the expression profile of Tis 21. The paragraph now reads as follows:

   « However, Cdkn1c expression is maintained longer and transiently peaks at high levels after Tis21 expression is switched off. Given that Tis21 is no more expressed in neurons (Iacopetti et al, 1999), this suggests that Cdkn1c expression is transiently upregulated in nascent neurons before fading off in more mature cells. »

   *- "Interestingly, the gene cluster that contained Tis21 also contained genes encoding proteins with known expression and/or functions at the transition from proliferation to differentiation, such as the Notch ligand Dll1, the bHLH transcription factors Hes6, NeuroG1 and NeuroG2, and the coactivator Gadd45g." Missing references. *

   We have now added references linking the function and/or expression profile of these genes to the neurogenic transition: Dll1 (Henrique et al., 1995), the bHLH transcription factors Hes6 (Fior and Henrique, 2005), NeuroG1 and NeuroG2 (Lacomme et al., 2012; Sommer et al., 1996) and the coactivator Gadd45g (Kawaue et al., 2014).

   *- There is an error in the color code in Cell Clusters in Figure 1C (cluster 4 yellow in the legend but ocre in the figure) *

   - Figure Sup3B colour code is switched (green for PP and red for NN) compared to the rest of the paper.

   We have corrected the colour code errors in Figure 1c and Supp Figure 3B (now changed to Supplementary Figure 5 in the modified revision)

   * ____It would be valuable to assign cell cycle stage to neural progenitor cells (based on cell cycle score) and determine whether cdkn1c at the transcript level also shows enrichment in G1 cells considered to be progenitors. *

   We have so far refrained from performing the suggested combined analysis based on cell cycle and cell type scores, as the "neurogenic progenitor population" (based on neurogenic progenitor score values) in which Cdkn1c expression is initiated represents a small number of cells in our scRNAseq, and felt that the significance of such an analysis is uncertain. We will perform this analysis in the revised version

   *2) Progressive increase in Cdkn1c/p57kip2 expression underlie different cellular states in the embryonic spinal neural tube (Figure 2). *

   *- Figure 2A. Scale bar is missing in E3 and E4. It is important to consider the growth of the developing spinal cord and present it accordingly (E3 transverse section, Figure 2). *

   The scale bar is actually valid for the whole panel A. The E2 section in the original figure appeared as "large" as the E3 section along the DV axis probably because the cutting angle was not perfectly transverse at E2, artificially lengthening the section. In a new version of the figure, we have replaced the E2 images with another section from the same experiment. The scale bar remains valid for the whole panel.

   *- Figure 2 could use a diagram of the knock-in strategy used, similar as the one in Figure 4A. *

   We have now added a diagram for the knock-in strategy in Figure 2B, and modified the legend of the figure accordingly.

   *- Indicate hours post-electroporation. Indicate which guide is used in the main text. *

   We have now added the post-electroporation timing and guide used in the main text.

   *3) Downregulation of Cdkn1c in neural progenitors delays the transition from proliferative to neurogenic modes of division (Figure 3). *

   *- In methods: "Thus, to reason on a more homogeneous progenitor population, we restricted all our analysis to the dorsal one half or two thirds of the neural tube." Indicate when and depending on what one half or two thirds of the neural tube were analysed. *

   *- Are the clonal analysis experiments (Fig 3D, E and F) also restricted to the dorsal region? *

   * We have modified this sentence as follows: "*Thus, to reason on a more homogeneous progenitor population, we restricted all our analysis to the dorsal two thirds of the neural tube, except for the Pax7-Cdkn1c misexpression analysis, which was performed in the more dorsal Pax7 domain."

   This is valid both for the whole population and clonal analyses

   *- Figure 3. Would have a better flow if 3C preceded 3A and 3B. *

   We have modified the Figure accordingly.

   *- Figure 3C. it would be informative to show pictures of the electroporated NT at both 24hpe and 48hpe, as well as highlighting the dorsal part of the neural tube that was used for quantification. *

   We have modified the Figure accordingly

   *- In methods "At each measured timepoint (1h, 4h, 7h, 10h, 12h, 14 and 17h after the first EdU injection), we quantified the number of EdU positive electroporated progenitors (triple positive for EdU, pRb and GFP) over the total population of electroporated progenitor cells (pRb and GFP positive) (Figure 3B)." Explanation does not correspond to Figure 3B. *

   This explanation corresponds indeed to Figure 5A. We have corrected this mistake in the new version of the manuscript.

   4) Inducing a premature expression of Cdkn1c in progenitors triggers the transition to neurogenic modes of division (Figure 4.).

   *- "We took advantage of the Pax7 locus, which is expressed in progenitors in the dorsal domain at a level similar to that observed for Cdkn1c in neurogenic precursors (Supplementary Figure 4A)". Missing reference or data showing that Pax7 is restricted to the dorsal domain. *

   We have added references to the expression profile of Pax7 in the dorsal neural tube (Jostes et al, 1990). In addition, the new Supplementary Figure 7 shows anti-Pax7 staining that confirm this expression pattern at E3 and E4

   *- "its intensity was similar to the one observed for endogenous Myc-tagged Cdkn1c in progenitors (Figure 4B and Supplementary Figure 4E), and remained below the endogenous level of Myc-tagged Cdkn1c observed in nascent neurons, confirming the validity of our strategy". It would be valuable to add a quantification to demonstrate this point, either by fluorescence levels or WB of nls-GFP cells. *

   As stated in the response to Major Point 5 above, we will perform a quantification based on Myc immunofluorescence to compare endogenous Cdkn1c expression versus Cdkn1c expression upon overexpression.

   *- "At the population level, at E4, Cdkn1c expression from the Pax7 locus resulted in a strong reduction in the number of progenitors (pRb positive cells)". Indicate in the main text that this is 48hpe. *

   We have added in the main text that the quantification was performed 48hae.

   *- Legend of figure 4D should indicate that the quantification has been done 24hpe. *

   We have added the timing of quantification in the legend of Figure 4D.

   *- "To circumvent the cell cycle arrest that is triggered in progenitors by strong overexpression of Cdkn1c (Gui et al., 2007)". It would be advisable to expand on this reference on the text, or ideally to include a simple Cdkn1c overexpression experiment. *

   These experiments have been performed and presented in the study by Gui et al., 2007, which we cite in the paper. Using a strong overexpression of CDKN1c from the CAGGS promoter, they showed a massive decrease in proliferation, assessed by BrdU incorporation, 24hours after electroporation. We will cite this result more explicitly in the main text, and better explain the difference of our approach. We propose the following modification:

   « We next explored whether low Cdkn1c activity is sufficient to induce the transition to neurogenic modes of division. A previous study has shown that overexpression of Cdkn1c driven by the strong CAGGS promoter triggers cell cycle exit of chick spinal cord progenitors, revealed by a drastic loss of BrdU incorporation 1 day after electroporation (Gui et al., 2007). As this precludes the exploration of our hypothesis, we developed an alternative approach designed to prematurely induce a pulse of Cdkn1c in progenitors, with the aim to emulate in proliferative progenitors* the modest level of expression observed in neurogenic progenitors. We took advantage of the Pax7 locus, which is expressed in progenitors in the dorsal domain at a level similar to that observed for Cdkn1c in neurogenic precursors (Supplementary Figure 4A)."*

   *- "We observed a massive increase in the proportion of neurogenic (PN and NN) divisions rising from 57% to 84% at the expense of proliferative pairs (43% PP pairs in controls versus 16% in misexpressing cells, Figure 4D)." adding the percentages in the main text is a bit inconsistent with how the rest of the data is presented in the rest of the sections. *

   This whole section has been modified in response to a question from reviewer 1. The new version does not contain percentages in the main text, and reads as follows:

   « Using the FlashTag cohort labeling approach described above, we traced the fate of daughter cells born 24 hae. We observed an increase in the proportion of terminal neurogenic (NN) divisions and a decrease in proliferative (PP) divisions (Figure 4D). This suggests that CDKN1c premature expression in PP progenitors converts them to the PN mode of division, while the combined endogenous and Pax7-driven expression of CDKN1c converts PN progenitors to the NN mode of division. Coincidentally, at the stage analyzed, PP to PN conversions are balanced by PN to NN conversions, leaving the PN proportion artificially unchanged. The alternative interpretation of a direct conversion of symmetric PP into symmetric NN divisions is less likely, because the PN compartment was affected in the reciprocal CDKN1c shRNA approach (see Figure 3F). Overall, these data show that inducing a premature low-level expression of Cdkn1c in cycling progenitors is sufficient to accelerate the transition towards neurogenic modes of division. »

   *- Figure sup 4C includes references to 3 gRNAs even when only one is used in the study. *

   The three guides listed in the original Supplementary Figure 4C correspond to the guides that we tested in Petit-Vargas et al. 2024. In this study, we only used the most efficient of these three guides. We have modified Figure 4C by quoting only this guide.

   *5) The proneurogenic activity of Cdkn1c in progenitors is mediated by modulation of cell cycle dynamics (Figure 5) *

   *- "we targeted the CyclinD1/CDK4-6 complex, which promotes cell cycle progression and proliferation, and is inhibited by Cdkn1c." reference missing *

   We have included references related to the activity of the CyclinD1/CDK4-6 complex in the developing CNS, and the antagonistic activities of CyclinD1 and Cdkn1c in this model

   - "we targeted the CyclinD1/CDK4-6 complex, which promotes cell cycle progression and proliferation in the developing CNS (Lobjois et al, 2004, 2008, Lange 2009, Gui et al 2007), and is inhibited by Cdkn1c (Gui et al, 2007)."

   *- It would be informative to include experimental set-up information (e.g. hae) in Figures 5A, 5B, 5F and 5G. *

   We have added the experimental set-up information in Figure 5.

   *- Clarify if analysis is restricted to the dorsal progenitors or the whole dorsoventral length of the tube. *

   The analyses were carried out on two thirds of the neural tube (dorsal 2/3), excluding the ventral zone, as specified above (and in the Methods section)

   *- It would be valuable to add an image to illustrate what is quantified in Figure 5D, Figure F and Figure G. *

   *- For Figure 4C and D, it would be valuable to add images to illustrate the quantification. *

   We have added images:

   • in Supplementary Figure 7C to illustrate what is quantified in Figures 4C (now 4C and 4D);
   • In Figure 5E to illustrate what is quantified in Figure 5D
   • In Supplementary Figure 8B to illustrate what is quantified in Figure 5G (now Figure 5H and 5I) Regarding the requested images for Figures 4D and 5F, they correspond to the same types of images already shown in Figure 3E. Since we have now added several additional examples of representative pairs of each type of mode of division in the new Supplementary Figure 4, we do not think that adding more of these images in figures 4 and 5 would strengthen the result of the quantifications.

   Discussion:

   *- "Nonetheless, studies in a wide range of species have demonstrated that beyond this binary choice, cell cycle regulators also influence the neurogenic potential of progenitors, i.e the commitment of their progeny to differentiate or not (Calegari and Huttner, 2003; FUJITA, 1962; Kicheva et al., 2014; Lange et al., 2009; Lukaszewicz and Anderson, 2011a; Pilaz et al., 2009; Smith and Schoenwolf, 1987; Takahashi et al., 1995)." Should include maybe references to Peco et al. Development 2012, Roussat et al. J Neurosci. 2023). *

   We have now included the references suggested by the reviewer.

   *- "This occurs through a change in the mode of division of progenitors, acting primarily via the inhibition of the CyclinD1/CDK6 complex." The data shown in the paper does not demonstrate that Cdkn1c is inhibiting CyclinD1, only that knocking down both mRNAs counteracts the effect of knocking down Cdkn1c alone at the general tissue level and in the percentage of PP/PN/NN clones. This statement should be qualified. *

   We propose to reformulate this paragraph in the discussion as follows to take this remark into account

   "This allows us to re-interpret the role of Cdkn1c during spinal neurogenesis: while previously mostly considered as a binary regulator of cell cycle exit in newborn neurons, we demonstrate that Cdkn1c is also an intrinsic regulator of the transition from the proliferative to neurogenic status in cycling progenitors. This occurs through a change in their mode of division, and our double knock-down experiments suggest that the onset of Cdkn1c expression may promote this change by counteracting a CyclinD1/CDK6 complex dependent mechanism."

   *Other comments: *

   - To improve clarity for the reader, it would help if electroporation was shown consistently on the same side of the neural tube. If electroporation has been performed at different sides and this is reflected in the figures, it would be advisable to explain on the figure legend.

   We have modified the figures to systematically show the electroporated side of the neural tube on the same side of the image for single electroporations.

   * ____- Figure legends should include the number of embryos/tissue sections analysed for each experiment, as well as information on whether the sections were cryostat or vibratome. *

   This information is now provided in the figure legends (numbers of cells analysed and/or numbers of embryos), except for data in Figure 5, which are presented in a new Supplementary Table 1.

   All experiments were performed on vibratome sections, except for in situ hybridization experiments, which were performed on cryostat sections. This last information was already indicated in the relevant figure legends

   *- Overall, there is a lack of consistency in the figures regarding how much information is available to the reader (e.g. Sup Figure 2A, in the panel mRNA in situ hybridisation of Cdkn1c is referred to only as Cdkn1c whereas in Sup figure 5 the in situ reads as CCND1 mRNA). Readability would improve a lot if figures included information on what is an electroporated fluorescent tag or an immunostaining (similar to the label in sup 4D) as well as the exact stage and hours after electroporation where relevant. *

   *- There is a general lack of consistency in indicating the timing of the experiments, both in terms of embryonic stage/day and in terms of hours-post-electroporation. *

   We have now homogenized the nomenclature in the figures.

   *- "Primary antibodies used are: chick anti-GFP (GFP-1020 - 1:2000) from Aves Labs; goat antiSox2 (clone Y-17 - 1:1000) from Santa Cruz". There is no Sox2 immunostaining in the article. *

   In the original version of the manuscript, the anti-Sox2 antibody was not used; we have now added experiments using this antibody in the modified version of the manuscript; this sentence in the Methods thus remains unchanged.

   *Reviewer #3 (Significance (Required)): *

   __*Significance:

   In neural development, there is a progressive switch in competence in neural progenitor cells, that transition from a proliferative (able to expand the neural progenitor pool) to neurogenic (able to produce neurons). Several factors are known to influence the transition of neural progenitor cells from a proliferative to a neurogenic state, including the activity of extracellular signalling pathways (e.g. SHH) (Saade et al. 2013, Tozer et al. 2017). In this study, the authors perform scRNA-seq of the cervical neural tube of chick at a stage of both proliferative and neurogenic progenitors are present, and identify transcriptional differences between the two populations. Among the differently expressed transcripts, they identify Cdkn1c (p57-Kip2) as enriched in neurogenic progenitors. Initially characterized as a driver of cell cycle exit in newborn neurons, the authors investigate the role of Cdkn1c in cycling progenitors. *__

   *The authors find that knock-down of Cdkn1c leads to an increase in proliferative divisions at the expense of neurogenic divisions. Conversely, misexpression of Cdkn1c in proliferative progenitors leads to a switch to neurogenic divisions. Furthermore, they find that knock-down of Cdkn1c shortens G1 phase of the cell cycle, suggesting a link between G1 length and neurogenic competence in neural progenitor cells. Cell cycle length has previously been linked to competence of neural progenitors, and it has been described that longer G1 duration is linked to neurogenic competence (e.g. Calegari F, Huttner WB. 2003). *

   * The strengths of the study include:*

   *The identification of a subset of genes enriched in neurogenic vs. proliferative progenitors. Since the transition from proliferative to neurogenic competence is a gradual process at the tissue level, the classification of proliferative vs. neurogenic progenitors based on a score of transcripts and the identification of a subset of transcripts that are enriched in neurogenic progenitors is a valuable contribution to the neurodevelopmental field. *

   *- The somatic knock-in strategy used to induce low-level overexpression of Cdkn1c in proliferative progenitors is an elegant strategy to induce overexpression in a subset of cells in a controlled manner and is a valuable technical advance. *

   - The characterization of a specific role of Cdkn1c in regulating cell cycle length in cycling progenitors is novel and valuable knowledge contributing to our understanding of how regulation of cell cycle length impacts competence of neural progenitors.

   * The aspects to improve: *

   __*

   • The sc-RNAseq isolated genes enriched in neurogenic versus proliferative progenitors, providing valuable insight into the gradual transition from proliferative to neurogenic competence at the tissue level. However, this gene subset requires clearer representation and detailed characterization. Additionally, the full scRNA-seq dataset should be made publicly available to support further research in neurodevelopment. *__

   The sequencing dataset has been deposited in NCBI's Gene Expression Omnibus database. It is currently under embargo, but will be made available upon acceptance and publication of the peer reviewed manuscript. Access is nonetheless available to the reviewers via a token that can be retrieved from the Review Commons website.

   The following information will be added in the final manuscript.

   Data availability

   Single cell RNA sequencing data have been deposited in NCBI's Gene Expression Omnibus (GEO) repository under the accession number GSE273710, and are available at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE273710."

   - The characterization of Cdkn1c dynamics in cycling progenitors using endogenous tagging of the Cdkn1c transcript with a Myc tag is an elegant way to investigate the dynamics of Cdkn1c-myc along the cell cycle. However, it would be much more powerful if combined with a careful characterization of pRb immunostaining along the cell cycle in this tissue, as well as the quantifications and controls proposed. __*

   • Retinoblastoma protein (Rb) and cyclin D play a key role in regulating the G1/S transition, with cyclin D/CDK complexes phosphorylating Rb. Given that CDKN1c primarily inhibits the cyclin D/CDK6 complex, it likely affects pRb expression or phosphorylation. This suggests pRb may be a direct target of CDKN1c, making it an unreliable marker for tracking and quantifying neurogenic progenitors through CDKN1c modulation. In light of this, it would be more appropriate to consider pRb as a CDKN1c target and discuss the molecular mechanisms regulating cell cycle components. A more precise approach would involve using other markers or targets to quantify neural precursor division modes at earlier stages of neurogenesis. *__

   __*

   • Many of the conclusions of the study are based on experiments performed using the FlashTag dye in order to perform clonal analysis of proliferative vs. neurogenic divisions. It would be very valuable to further characterize the reliability of this tool as well as to provide more information on the criteria used to determine the fate of the pairs of sister cells. *__

   __*

   • The somatic knock-in strategy used to induce low-level overexpression of Cdkn1c in proliferative progenitors is an elegant strategy to induce overexpression in a subset of cells in a controlled manner. It would be valuable to further characterize the dynamics of Cdkn1c expression using this too and to provide proof that Pax7 expression is not altered in guides with the knock-in event. *__

   __*

   • The presentation of the existing literature could be more up to date. *__

   __*

   • The presentation of the data in the figures could be improved for readability. The sc-RNA seq data and the technical advances could be of interest for an audience of researchers using chick as a model organism, and working on neurodevelopment in general. Furthermore, the characterization of Cdkn1c as a regulator of G1 length in cycling progenitors and its implications for neurogenic competence could be of general interest for people working on basic research in the neurodevelopmental field. *__

   * Field of expertise of the reviewer: neural development, cell biology, embryology.*

   Read the original source
2. 

   Review Commons

   Apr 14, 2025

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #3

   Evidence, reproducibility and clarity

   Summary:

   In this study, Mida et al. analyze large-scale single-cell RNA-seq data from the chick embryonic neural tube and identify Cdkn1c as a key molecular regulator of the transition from proliferative to neurogenic cell divisions, marking the onset of neurogenesis in the developing CNS. To confirm this hypothesis, they employed classical techniques, including the quantification of neural cell-specific markers combined with the flashTAG label, to track and isolate isochronic cohorts of newborn cells in different division modes. Their findings reveal that Cdkn1c expression begins at low levels in neurogenic progenitors and becomes …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #3

   Evidence, reproducibility and clarity

   Summary:

   In this study, Mida et al. analyze large-scale single-cell RNA-seq data from the chick embryonic neural tube and identify Cdkn1c as a key molecular regulator of the transition from proliferative to neurogenic cell divisions, marking the onset of neurogenesis in the developing CNS. To confirm this hypothesis, they employed classical techniques, including the quantification of neural cell-specific markers combined with the flashTAG label, to track and isolate isochronic cohorts of newborn cells in different division modes. Their findings reveal that Cdkn1c expression begins at low levels in neurogenic progenitors and becomes highly expressed in nascent neurons. Using a classical knockdown strategy based on short hairpin RNA (shRNA) interference, they demonstrate that Cdkn1c suppression promotes proliferative divisions, reducing neuron formation. Conversely, novel genetic manipulation techniques inducing low-level CDKN1c misexpression drive progenitors into neurogenic divisions prematurely. By employing cumulative EdU incorporation assays and shRNA-based loss-of-function approaches, Mida et al. further show that Cdkn1c extends the G1 phase by inhibiting cyclin D, ultimately concluding that Cdkn1c plays a dual role: first facilitating the transition of progenitors into neurogenic divisions at low expression levels, and later promoting cell cycle exit to ensure proper neural development.

   This study presents several ambiguities and lacks precision in its analytical methodologies and quantification approaches, which contribute to confusion and potential bias. To enhance the reliability of the conclusions, a more rigorous validation of the methods employed is essential.

   This study introduces a novel approach to tracking the fate of sister cells from neural progenitor divisions to infer the division modes. While previous methods for analyzing the division mode of neural progenitor cells have been implemented, rigorous validation of the approach introduced by Mida et al. is necessary. Furthermore, the concept of cell cycle regulators interacting to control the duration of specific cell cycle stages and influencing progenitor cell division modes has been explored before, potentially limiting the novelty of these findings.

   Majors comments:

   1. The study presents ambiguity and lacks precision in quantifying neural precursor division modes. The authors use phosphorylated retinoblastoma protein (pRb) as a marker for neurogenic progenitors, claiming its reliability in identifying neurogenic divisions. However, they do not provide a thorough characterization of pRb expression in the developing chick neural tube, leaving its suitability as a neurogenic division marker unverified. Furthermore, retinoblastoma protein (Rb) and cyclin D interact crucially to regulate the G1/S phase transition of the cell cycle, with cyclin D/CDK complexes phosphorylating Rb. Since the authors conclude that CDKN1c primarily acts by inhibiting the cyclin D/CDK6 complex, it is likely that CDKN1c influences pRb expression or phosphorylation state. This raises the possibility that pRb could be a direct target of CDKN1c, whose expression and phosphorylation would be altered in gain-of-function (GOF) and loss-of-function (LOF) analyses of CDKN1c. In light of this, it would be more appropriate to consider pRb as a CDKN1c target and discuss the molecular mechanisms regulating cell cycle components. A more precise approach would involve using other markers or targets to quantify neural precursor division modes at earlier stages of neurogenesis.
   2. Furthermore, the study employs FlashTag labeling to track daughter cells post-division, but the 16-hour post-injection window may result in misidentification of sister cells due to the potential presence of FlashTagged cells that did not originate from the same division. This introduces a risk of bias in quantification, data misinterpretation, and potential errors in defining division modes. A more rigorous validation of the FlashTag strategy and its specificity in tracking division pairs is necessary to ensure the reliability of their conclusions.
   3. The knock-in strategy used to tag the endogenous CDKN1c protein in Figure 2 is an elegant tool to infer protein dynamics in vivo. However, since strong conclusions regarding CDKN1c dynamics during the cell cycle are drawn from this section, it would be advisable to strengthen the results by including quantification with adequate replication and proper statistical analysis, as the current findings are preliminary and somewhat speculative.
      • "Although pRb is specific for cycling cells, it is only detected once cells have passed the point of restriction during the G1 phase." Please provide literary reference confirming this observation. Given that pRb immunoreactivity is used as a marker for cycling progenitors to base many of the results of this study, it would be very valuable to characterize the dynamics of pRb in cycling cells in the studied tissue, for instance combined with the cell cycle reporter used by Molina et al. (Development 2022).
      • The characterization of dynamics is performed only with one of the gRNAs (#1) on the basis that it produces the strongest NLS-GFP signal, as a proxy for guide efficiency. It would be nice if the authors could validate guide cutting efficiency via sequencing (e.g. using a Cas9-T2A-GFP plasmid and sorting for positive cells).
      • In order to make sure that the dynamics inferred from Myc-tag immunoreactivity do reflect the cell cycle dynamics of CDKN1c-myc, it would be advisable to confirm in-frame insertion of the myc-tag sequence.
      • It would be valuable to analyse the dynamics of Myc immunoreactivity in combination of pRb in all three gRNAs (highlighted in Supplementary Figure 1), as it would be a strong point in favour that the dynamics reflect the endogenous CDKN1c dynamics.

   • It would be very valuable to provide a quantification of said dynamics (e.g. plotting myc intensity / pRb immunoreactivity along the apicobasal axis of the tissue).

   1. In Figure 3, the authors use a short-hairpin-mediated knock-down strategy to decrease the levels of Cdkn1c, and show that this manipulation leads to an increase percentage of cycling progenitors and a decrease in the number of neurons in electroporated cells.

   The authors claim that their shRNA-based knockdown strategy aims to reduce low-level Cdkn1c expression in neurogenic progenitors while minimally affecting the higher expression in newborn neurons required for cell cycle exit. However, several factors need consideration. Electroporation introduces variability in shRNA delivery, making it difficult to achieve consistent gene inhibition across all cells, especially for dose-dependent genes like Cdkn1c. Additionally, Cdkn1c generates multiple isoforms, which may not be fully annotated in the chick genome, raising the possibility that the shRNA targets specific isoforms, potentially explaining the observed low expression. A more rigorous approach, such as qPCR analysis of sorted electroporated cells, would better validate the expression levels, rather than relying on in situ hybridization, presenting electroporated and non-electroporated cells in the same section (Supp. Figure 2). - As the authors note, "Unambiguous identification of cycling progenitors and postmitotic neurons is notoriously difficult in the chick spinal cord". "markers of progenitors usually either do not label all the phases of the cell cycle (eg. Phospho-Rb, thereafter pRb), or persist transiently in newborn neurons (eg. Sox2)." Given that pRb immunoreactivity is used as the basis for a lot of the conclusions in this study, it would be valuable to add a characterization of its dynamics as mentioned in Figure 2, as well as provide literary references/proof that Sox2 expression persists in newborn neurons. - The undefined population (pRb-/HuCD-) introduces an unknown that assumes that the percentage of progenitors in G1 phase before the restriction point and the number of newborn neurons are equal for both conditions in an experiment. Can the authors provide explanation for this assumption? - In Gui et al. (Dev Biol 2006), authors showed that a knockdown of Cdkn1c leads to a failure of nascent neurons to exit the cell cycle and causes them to re-entry the cell cycle, shown by ectopic mitoses. In that study, cells born from those ectopic mitoses eventually leave the cell cycle leading to an increase in the number of neurons. Can the authors check for ectopic mitoses at 24hpe and 48hpe? - The authors then address the question of whether the decrease in neuron number is due to the failure of newborn neurons to exit the cell cycle or to a delay in the transition from proliferative to neurogenic divisions. For that, they implement a strategy to label a synchronized cohort of progenitors based of incorporation of a FlashTag dye. - Given that this strategy is the basis of many of the experiments in this article, it would be very valuable to expand on the validation of this technique as cited in major comment #2. In figure 3E, the close proximity of cell pairs in PP and PN clones shown in the pictures makes their sibling status apparent. However, this is not the case for the NN clone. Can the authors further explain with what criteria they determined the clonal status of two FlashTag labelled cells? Can they provide further image examples of different types of clones? - Can the authors show that the plateau reached in Sup Figure 3 for pRb immunoreactivity corresponds to a similar dynamic for HuC/D immunoreactivity? - In order to further validate the strategy, could the authors use it at different stages to validate if they can replicate the different percentages of PP/PN/NN reported in the literature (e.g. Saade Cell Rep 2013)?.

   1. In Figure 4, the strategy used to induce a low-dose overexpression of CDKN1c is an elegant method to introduce CDKN1c-Myc expression under the control of the endogenous Pax7 promoter, active in proliferative progenitors. The main point to address is:
      • Please provide proof that Pax7 expression is not altered in guides with a successful knock-in event (e.g. sorting and WB against the Pax7 protein) or the immunohistochemistry as performed in the Pax7-P2A-Gal4 tagging in Petit-Vargas et al., 2024.
      • Given the cell cycle regulated expression and activity of CDKN1c, can the authors elaborate on whether this is regulated at the promoter level? If so, how does this differ from the promoter activity of Pax7?
      • It would be advisable to characterize the dynamics along the cell cycle for the overexpressed form of CDKN1c-Myc relative to pRb, similarly to what was done in Figure 2B.
   2. In figure 5, the authors use a double knock-down strategy to test the hypothesis that the effect of Cdkn1c in G1 length is partially at least through its inhibition of CyclinD1. Results show that double shRNA-mediated knock-down of CyclinD1 and Cdkn1c counteracts the effects of Cdkn1c-sh alone on EdU incorporation, PP/PN/NN cell divisions and overall rations of progenitors and neurons.
      • In the measurement of progenitor cell cycle length in Figure 5A, it would be more appropriate to present the nonlinear regression method described by Nowakowski et al. (1989), as has been commonly used in the field (Saade et al., 2013, PMID: 23891002, Le Dreau et al., 2014, PMID: 24515346, Arai et al., 2011, PMID: 21224845).
      • Cumulative EdU incorporation in spinal progenitors (pRb-positive) at E3 (24 hours after injection) showed that the proportion of EdU-positive progenitors reached a plateau at 14 hours in control conditions, which is later than what has been reported in Le Dreau et al., 2014 (PMID: 24515346). Can you explain why?
      • It would be interesting to measure G1 length as in Figure 5D for the double cdkn1c-sh - ccnd1-sh knock down condition, to see if it rescues G1 length. As well as in the Ccnd1 knock down condition alone to see if it increases G1 length in this context as well.

   Minor comments

   Introduction:

   • The introduction should include references of studies of the role of Cdkn1c in cortical development (Imaizumi et al. Sci Rep 2020, Colasante et al. Cereb Cortex 2015, Laukoter et al. Nature Communications 2020).

   1. Transcriptional signature of the neurogenic transition (Figure 1).
      • In the result section, it would be informative to include the genes used to determine the progenitor and neuron score (instead of in Methods).
      • Figure 1A. It would be informative to add in the diagram what "filtering" means (eg. Neural crest cells).
      • In the result section, "However, while Tis21 expression is switched off in neurons, Cdkn1c transiently peaks at high levels in nascent neurons before fading off in more mature cells." Missing literary reference or data to clearly demonstrate this point.
      • "Interestingly, the gene cluster that contained Tis21 also contained genes encoding proteins with known expression and/or functions at the transition from proliferation to differentiation, such as the Notch ligand Dll1, the bHLH transcription factors Hes6, NeuroG1 and NeuroG2, and the coactivator Gadd45g." Missing references.
      • There is an error in the color code in Cell Clusters in Figure 1C (cluster 4 yellow in the legend but ocre in the figure)

   It would be valuable to assign cell cycle stage to neural progenitor cells (based on cell cycle score) and determine whether cdkn1c at the transcript level also shows enrichment in G1 cells considered to be progenitors.

   1. Progressive increase in Cdkn1c/p57kip2 expression underlie different cellular states in the embryonic spinal neural tube (Figure 2).
      • Figure 2A. Scale bar is missing in E3 and E4. It is important to consider the growth of the developing spinal cord and present it accordingly (E3 transverse section, Figure 2).
      • Figure 2 could use a diagram of the knock-in strategy used, similar as the one in Figure 4A.
      • Indicate hours post-electroporation. Indicate which guide is used in the main text.
   2. Downregulation of Cdkn1c in neural progenitors delays the transition from proliferative to neurogenic modes of division (Figure 3).
      • In methods: "Thus, to reason on a more homogeneous progenitor population, we restricted all our analysis to the dorsal one half or two thirds of the neural tube." Indicate when and depending on what one half or two thirds of the neural tube were analysed.
      • Figure 3. Would have a better flow if 3C preceded 3A and 3B.
      • Figure 3C. it would be informative to show pictures of the electroporated NT at both 24hpe and 48hpe, as well as highlighting the dorsal part of the neural tube that was used for quantification.
      • Are the clonal analysis experiments (Fig 3D, E and F) also restricted to the dorsal region?
      • Figure Sup3B colour code is switched (green for PP and red for NN) compared to the rest of the paper.
      • In methods "At each measured timepoint (1h, 4h, 7h, 10h, 12h, 14 and 17h after the first EdU injection), we quantified the number of EdU positive electroporated progenitors (triple positive for EdU, pRb and GFP) over the total population of electroporated progenitor cells (pRb and GFP positive) (Figure 3B)." Explanation does not correspond to Figure 3B.
   3. Inducing a premature expression of Cdkn1c in progenitors triggers the transition to neurogenic modes of division (Figure 4.).
      • "We took advantage of the Pax7 locus, which is expressed in progenitors in the dorsal domain at a level similar to that observed for Cdkn1c in neurogenic precursors (Supplementary Figure 4A)". Missing reference or data showing that Pax7 is restricted to the dorsal domain.
      • "its intensity was similar to the one observed for endogenous Myc-tagged Cdkn1c in progenitors (Figure 4B and Supplementary Figure 4E), and remained below the endogenous level of Myc-tagged Cdkn1c observed in nascent neurons, confirming the validity of our strategy". It would be valuable to add a quantification to demonstrate this point, either by fluorescence levels or WB of nls-GFP cells.
      • For Figure 4C and D, it would be valuable to add images to illustrate the quantification.
      • "At the population level, at E4, Cdkn1c expression from the Pax7 locus resulted in a strong reduction in the number of progenitors (pRb positive cells)". Indicate in the main text that this is 48hpe.
      • Legend of figure 4D should indicate that the quantification has been done 24hpe.
      • "To circumvent the cell cycle arrest that is triggered in progenitors by strong overexpression of Cdkn1c (Gui et al., 2007)". It would be advisable to expand on this reference on the text, or ideally to include a simple Cdkn1c overexpression experiment.
      • "We observed a massive increase in the proportion of neurogenic (PN and NN) divisions rising from 57% to 84% at the expense of proliferative pairs (43% PP pairs in controls versus 16% in misexpressing cells, Figure 4D)." adding the percentages in the main text is a bit inconsistent with how the rest of the data is presented in the rest of the sections.
      • Figure sup 4C includes references to 3 gRNAs even when only one is used in the study.
   4. The proneurogenic activity of Cdkn1c in progenitors is mediated by modulation of cell cycle dynamics (Figure 5)
      • "we targeted the CyclinD1/CDK4-6 complex, which promotes cell cycle progression and proliferation, and is inhibited by Cdkn1c." reference missing
      • It would be valuable to add an image to illustrate what is quantified in Figure 5D, Figure F and Figure G.
      • It would be informative to include experimental set-up information (e.g. hae) in Figures 5A, 5B, 5F and 5G.
      • Clarify if analysis is restricted to the dorsal progenitors or the whole dorsoventral length of the tube.

   Discussion:

   • "Nonetheless, studies in a wide range of species have demonstrated that beyond this binary choice, cell cycle regulators also influence the neurogenic potential of progenitors, i.e the commitment of their progeny to differentiate or not (Calegari and Huttner, 2003; FUJITA, 1962; Kicheva et al., 2014; Lange et al., 2009; Lukaszewicz and Anderson, 2011a; Pilaz et al., 2009; Smith and Schoenwolf, 1987; Takahashi et al., 1995)." Should include maybe references to Peco et al. Development 2012, Roussat et al. J Neurosci. 2023).
   • "This occurs through a change in the mode of division of progenitors, acting primarily via the inhibition of the CyclinD1/CDK6 complex." The data shown in the paper does not demonstrate that Cdkn1c is inhibiting CyclinD1, only that knocking down both mRNAs counteracts the effect of knocking down Cdkn1c alone at the general tissue level and in the percentage of PP/PN/NN clones. This statement should be qualified.

   Other comments:

   • There is a general lack of consistency in indicating the timing of the experiments, both in terms of embryonic stage/day and in terms of hours-post-electroporation.
   • To improve clarity for the reader, it would help if electroporation was shown consistently on the same side of the neural tube. If electroporation has been performed at different sides and this is reflected in the figures, it would be advisable to explain on the figure legend.
   • Figure legends should include the number of embryos/tissue sections analysed for each experiment, as well as information on whether the sections were cryostat or vibratome.
   • Overall, there is a lack of consistency in the figures regarding how much information is available to the reader (e.g. Sup Figure 2A, in the panel mRNA in situ hybridisation of Cdkn1c is referred to only as Cdkn1c whereas in Sup figure 5 the in situ reads as CCND1 mRNA). Readability would improve a lot if figures included information on what is an electroporated fluorescent tag or an immunostaining (similar to the label in sup 4D) as well as the exact stage and hours after electroporation where relevant.
   • "Primary antibodies used are: chick anti-GFP (GFP-1020 - 1:2000) from Aves Labs; goat antiSox2 (clone Y-17 - 1:1000) from Santa Cruz". There is no Sox2 immunostaining in the article.

   Significance

   In neural development, there is a progressive switch in competence in neural progenitor cells, that transition from a proliferative (able to expand the neural progenitor pool) to neurogenic (able to produce neurons). Several factors are known to influence the transition of neural progenitor cells from a proliferative to a neurogenic state, including the activity of extracellular signalling pathways (e.g. SHH) (Saade et al. 2013, Tozer et al. 2017). In this study, the authors perform scRNA-seq of the cervical neural tube of chick at a stage of both proliferative and neurogenic progenitors are present, and identify transcriptional differences between the two populations. Among the differently expressed transcripts, they identify Cdkn1c (p57-Kip2) as enriched in neurogenic progenitors. Initially characterized as a driver of cell cycle exit in newborn neurons, the authors investigate the role of Cdkn1c in cycling progenitors. The authors find that knock-down of Cdkn1c leads to an increase in proliferative divisions at the expense of neurogenic divisions. Conversely, misexpression of Cdkn1c in proliferative progenitors leads to a switch to neurogenic divisions. Furthermore, they find that knock-down of Cdkn1c shortens G1 phase of the cell cycle, suggesting a link between G1 length and neurogenic competence in neural progenitor cells. Cell cycle length has previously been linked to competence of neural progenitors, and it has been described that longer G1 duration is linked to neurogenic competence (e.g. Calegari F, Huttner WB. 2003).

   The strengths of the study include:

   The identification of a subset of genes enriched in neurogenic vs. proliferative progenitors. Since the transition from proliferative to neurogenic competence is a gradual process at the tissue level, the classification of proliferative vs. neurogenic progenitors based on a score of transcripts and the identification of a subset of transcripts that are enriched in neurogenic progenitors is a valuable contribution to the neurodevelopmental field.

   • The somatic knock-in strategy used to induce low-level overexpression of Cdkn1c in proliferative progenitors is an elegant strategy to induce overexpression in a subset of cells in a controlled manner and is a valuable technical advance.
   • The characterization of a specific role of Cdkn1c in regulating cell cycle length in cycling progenitors is novel and valuable knowledge contributing to our understanding of how regulation of cell cycle length impacts competence of neural progenitors.

   The aspects to improve:

   • The sc-RNAseq isolated genes enriched in neurogenic versus proliferative progenitors, providing valuable insight into the gradual transition from proliferative to neurogenic competence at the tissue level. However, this gene subset requires clearer representation and detailed characterization. Additionally, the full scRNA-seq dataset should be made publicly available to support further research in neurodevelopment.
   • The characterization of Cdkn1c dynamics in cycling progenitors using endogenous tagging of the Cdkn1c transcript with a Myc tag is an elegant way to investigate the dynamics of Cdkn1c-myc along the cell cycle. However, it would be much more powerful if combined with a careful characterization of pRb immunostaining along the cell cycle in this tissue, as well as the quantifications and controls proposed.
   • Retinoblastoma protein (Rb) and cyclin D play a key role in regulating the G1/S transition, with cyclin D/CDK complexes phosphorylating Rb. Given that CDKN1c primarily inhibits the cyclin D/CDK6 complex, it likely affects pRb expression or phosphorylation. This suggests pRb may be a direct target of CDKN1c, making it an unreliable marker for tracking and quantifying neurogenic progenitors through CDKN1c modulation. In light of this, it would be more appropriate to consider pRb as a CDKN1c target and discuss the molecular mechanisms regulating cell cycle components. A more precise approach would involve using other markers or targets to quantify neural precursor division modes at earlier stages of neurogenesis.
   • Many of the conclusions of the study are based on experiments performed using the FlashTag dye in order to perform clonal analysis of proliferative vs. neurogenic divisions. It would be very valuable to further characterize the reliability of this tool as well as to provide more information on the criteria used to determine the fate of the pairs of sister cells.
   • The somatic knock-in strategy used to induce low-level overexpression of Cdkn1c in proliferative progenitors is an elegant strategy to induce overexpression in a subset of cells in a controlled manner. It would be valuable to further characterize the dynamics of Cdkn1c expression using this too and to provide proof that Pax7 expression is not altered in guides with the knock-in event.
   • The presentation of the existing literature could be more up to date.
   • The presentation of the data in the figures could be improved for readability. The sc-RNA seq data and the technical advances could be of interest for an audience of researchers using chick as a model organism, and working on neurodevelopment in general. Furthermore, the characterization of Cdkn1c as a regulator of G1 length in cycling progenitors and its implications for neurogenic competence could be of general interest for people working on basic research in the neurodevelopmental field.

   Field of expertise of the reviewer: neural development, cell biology, embryology.

   Read the original source
3. 

   Review Commons

   Apr 14, 2025

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   The work by Mida and colleagues addresses important questions about neurogenesis in the embryo, using the chicken neural tube as their model system. The authors investigate the mechanisms involved in the transition from stem cell self-renewal to neurogenic progenitor divisions, using a combination of single cell, gene functional and tracing studies.

   The authors generated a new single cell data set from the embryonic chicken spinal cord and identify a transitory cell population undergoing neuronal differentiation, which expresses Tis21, Neurog2 and Cdkn1c amongst other genes. They then study the role of Cdkn1c and investigate the …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   The work by Mida and colleagues addresses important questions about neurogenesis in the embryo, using the chicken neural tube as their model system. The authors investigate the mechanisms involved in the transition from stem cell self-renewal to neurogenic progenitor divisions, using a combination of single cell, gene functional and tracing studies.

   The authors generated a new single cell data set from the embryonic chicken spinal cord and identify a transitory cell population undergoing neuronal differentiation, which expresses Tis21, Neurog2 and Cdkn1c amongst other genes. They then study the role of Cdkn1c and investigate the hypothesis that it plays a dual role in spinal cord neurogenesis: low levels favour transition from proliferative to neurogenic divisions and high levels drive cell cycle exit and neuronal differentiation.

   Major comments

   I have only a general comment related to the main point of the paper. The authors claim that Cdkn1c onset in cycling progenitor drives transition towards neurogenic modes of division, which is different from its role in cell cycle exit and differentiation. Figures 3F and 4D are key figures where the authors analysed PP, PN and NN mode of divisions via flash tag followed by analysis of sister cell fate. If their assumption is correct, shouldn't they also see, for example in Fig. 4D, an increase in PN or is this too transient to be observed or is it bypassed? At the moment, the calculations of PN and NN frequencies are merged in the text, so perhaps describing PN and NN numbers separately will help better understand the dynamics of this gradual process (especially since there is little to no difference in PN). Could the increase in NN be compatible also with a role in cell cycle exit and differentiation, for example from cells that have been targeted and are still undergoing the last division (hence marked by flash tag) or there won't be any GFP cells marked by flash tag a day after expression of high levels of Cdkn1c? Basically, what would the effect of expressing higher levels of Cdkn1c be? I guess this will really help them distinguish between transition to neurogenic division rather than neuronal differentiation. If not experimentally, any further comments on this would be appreciated.

   Minor comments

   Fig 3C my understanding is that HuC/D should be nuclear, but in fig 3C it seems more cytoplasmic (any comment?)

   Fig Suppl 3E (and related 4B), immuno for Cdkn1c-Myc: to help the reader understand the difference between the immuno signals when looking at the figure, I would suggest writing on the panel i) Pax7-Cdkn1c-Myc and ii) endogenous Cdkn1c-Myc, rather than 'misexpressed' and 'endogenous', which is slightly confusing (especially because what it is called endogenous expression is higher).

   Literature citing: Introduction and discussion are very nicely written, although they could benefit from some more recent literature on the topic. For example, Cdkn1c role as a gatekeeper of stem cell reserve in the stomach, gut, (Lee et al, CellStemCell 2022 PMID: 35523142) or some other work on symmetric/asymmetric divisions and clonal analysis in zebrafish (Hevia et al, CellRep 2022 PMID: 35675784, Alexandre et al, NatNeur PMID: 20453852), mammals (Royal et al, Elife 2023 37882444, Appiah et al, EMBO rep 2023 PMID: 37382163). Also, similar work has been performed in the developing pancreatic epithelium, where mild expression of Cdkn1a under Sox9rtTa control was used to lengthen G1 without overt cell cycle exit and this resulted in Neurog3 stabilization and priming for endocrine differentiation (Krentz et al, DevCell 2017 PMID: 28441528), so similar mechanisms might be in in place to gradually shift progenitor towards stable decision to differentiate. Moreover, in the discussion, alongside Neurog2 control of Cdkn1c, it could be mentioned that the feedback loop between Cdk inhibitors and neurogenic factor is usually established via Cdk inhibitor-mediated inhibition of proneural bHLHs phosphorylation by CDKs (Krentz et al, DevCell 2017 PMID: 28441528, Ali et al, 24821983, Azzarelli et al 2017 - PMID: 28457793; 2024 - PMID:39575884). Further, in the discussion, could they mention anything about the following open questions: is there evidence for Cdkn1c low/high expression in mammalian spinal cord? Or maybe of other Cdk inhibitors? Is Cdkn1c also involved in cell cycle exit during gliogenesis or is there another Cdk inhibitor expressed at later developmental stages, hence linking this with specific cell fate decisions?

   Significance

   The work here presented has important implications on neural development and its disorders. The authors used the most advanced technologies to perform gene functional studies, such as CRISPR-HDR insertion of Myc-tag to follow endogenous expression, or expression under endogenous Pax7 promoter, often followed by flash tag experiments to trace sister cell fate, and all of this in an in vivo system. They then tested cell cycle parameters, clonal behaviour and modes of cell division in a very accurate way. Overall data are convincing and beautifully presented. The limitation is potentially in the resolution between the events of switching to neurogenic division versus neuronal differentiation, which might just warrant further discussion. This work advances our knowledge on vertebrate neurogenesis, by investigating a key player in proliferation and differentiation.

   I believe this work will be of general interest to developmental and cellular biologists in different fields. Because it addresses fundamental questions about the coordination between cell cycle and differentiation and fate decision making, some basic concepts can be translated to other tissues and other species, thus increasing the potential interested audience.

   My work focuses on stem cell fate decisions in mammalian systems, and I am familiar with the molecular underpinnings of the work here presented. However, I am not an expert in the chicken spinal cord as a model and yet the manuscript was interesting. I am also not sufficiently expert in the bioinformatic analysis, so cannot comment on the technical aspects of Figure 1 and the way they decided to annotate their data.

   Read the original source
4. 

   Review Commons

   Apr 14, 2025

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   Summary

   This study utilizes the developing chicken neural tube to assess the regulation of the balance between proliferative and neurogenic divisions in the vertebrate CNS. Using single-cell RNAseq and endogenous protein tagging, the authors identify Cdkn1c as a potential regulator of the transition towards neurogenic divisions. Cdkn1c knockdown and overexpression experiments suggest that low Cdkn1c expression enhances neurogenic divisions. Using a combination of clonal analysis and sequential knockdown, the authors find that Cdkn1c lengthens the G1 phase of the cell cycle via inhibition of cyclinD1. This study represents a …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   Summary

   This study utilizes the developing chicken neural tube to assess the regulation of the balance between proliferative and neurogenic divisions in the vertebrate CNS. Using single-cell RNAseq and endogenous protein tagging, the authors identify Cdkn1c as a potential regulator of the transition towards neurogenic divisions. Cdkn1c knockdown and overexpression experiments suggest that low Cdkn1c expression enhances neurogenic divisions. Using a combination of clonal analysis and sequential knockdown, the authors find that Cdkn1c lengthens the G1 phase of the cell cycle via inhibition of cyclinD1. This study represents a significant advance in understanding how cells can transition between proliferative and asymmetric modes of division, the complex and varying roles of cycle regulators, and provides technical advance through innovative combination of existing tools.

   Major and Minor Comments:

   Overall

   • Sample numbers are missing or unclear throughout for all imaging experiments. The authors should add numbers of cells analysed and/or numbers of embryos for their results to be appropriately convincing.
   • Values and error bars on graphs must be defined throughout. Are the values means and error bars SD or SEM?

   Results 2

   • A reference should be provided for cell type distribution in spinal neural tube, where the authors state that cell bodies of progenitors reside within the ventricular zone.
   • The authors state that Cdkn1c "was expressed at low levels in a salt and pepper fashion in the ventricular zone, where the cell bodies of neural progenitors reside, and markedly increased in a domain immediately adjacent to this zone which is enriched in nascent neurons on their way to the mantle zone. In contrast, the transcript was completely excluded from the mantle zone, where HuC/D positive mature neurons accumulate." It is not clear if this is referring only to E4 or also to E3 embryos. Indeed, Cdkn1c expression appears to be much more salt and pepper at E3 and only resolves into a clear domain of high expression adjacent to the mantle zone at E4. It may be helpful if this expression pattern could be described in a bit more detail highlighting the changes that occur between E3 and E4.
   • It would be useful to annotate the ISH images in Fig 2A to show the ventricular and mantle zones as defined by immunofluorescence.
   • Reference should be included for pRb expression dynamics.
   • Could the Myc tag insertion approach disrupt protein function or turnover?
   • Why was the insertion target site at the C terminus chosen?
   • OPTIONAL Could a similar approach be used to tag Cdkn1c with a fluorescent protein to enable live imaging of dynamics?
   • In suppl Fig 1C nlsGFP-positive cells are shown in the control shRNA condition. How can this be explained and does it impact the interpretation of the findings?
   • In Fig 2B, there are a number of Myc labelled cells in the mantle zone, whereas the in situ images show no appreciable transcript expression. Is this because the protein but not the transcript is present in these cells? Could the authors comment on this?

   Results 3

   • It should be mentioned how mRNA expression levels were quantified in the shRNA validation experiment (supp Fig 2A).
   • Figure panels are not currently cited in order. Citation or figure order could be changed.
   • The authors should provide representative images for the graphs shown in Fig 3A and 3B. These could go into supplementary if the authors prefer.
   • A supplementary figure showing the Caspase3 experiment should be added.
   • OPTIONAL. Identification of sister cells in the clonal analysis experiments is based on static images and cannot be guaranteed. Could live imaging be used to watch divisions followed by fixation and immunostaining to confirm identity?

   Results 4

   • How did the authors quantify the intensity of endogenous Myc-tagged Cdkn1c to confirm the validity of the Pax7 locus knock in? Can they show that the expression level was consistently lower than the endogenous expression in neurons? Quantification and sample numbers should be shown.
   • In Fig 4B, the brightness of row 2 column 1 is lower than the same image in row 2 column 2, which is slightly misleading, since it makes the misexpressed expression level look lower than it is compared with endogenous in column 3. Is this because only a single z-section is being displayed in the zoomed in image? If so, this should be stated in the figure legend.
   • In Fig 4D, the increase in neurogenic divisions is mainly because of the rise in terminal NN divisions according to the graph, but no clear increase in PN divisions. Could the authors comment on the significance of this?

   Results 5

   • The proportion of pRb-positive progenitors having entered S phase was stated to be higher at all time points; however, it is not significantly higher until 6h30 and is actually trending lower at 2h30.
   • OPTIONAL Could CyclinD1 activity be directly assessed?

   General

   • Scale bars missing fig s1c s4d.
   • OPTIONAL Some of the main findings be replicated in another species, for example, mouse or human to examine whether the mechanism is conserved.
   • OPTIONAL Could use approaches other than image analysis be used to reinforce findings, for example biochemical methods, RNAseq or FACS?
   • A model cartoon to summarise outcomes would be useful.
   • Unclear how cells were determined to be positive or negative for a label. Was this decided by eye? If so, how did the authors ensure that this was unbiased?

   Significance

   Strengths:

   This manuscript investigates the mechanisms regulating the switch from symmetric proliferative divisions to neurogenic division during vertebrate neuronal differentiation. This is a question of fundamental importance, the answer to which has eluded us so far. As such, the findings presented here are of significant value to the neurogenesis community and will be of broad interest to those interested in cell divisions and asymmetric cell fate acquisition. Specific strengths include:

   • Variety of approaches used to manipulate and observe individual cell behaviour within a physiological context.
   • A limitation of using the chicken embryo is the lack of available antibodies for immunostaining. The authors take advantage of recent advances in chicken embryo CRISPR strategy to endogenously tag the target protein with Myc, to facilitate immunostaining.
   • Innovative combination of genetic and labelling tools to target cells, for example, use of FlashTag and EdU in combination to more accurately assess G1 length than the more commonly used method.
   • Premature misexpression demonstrates that the previously observed dynamics indeed regulate cell fate.
   • Mechanistic insight by examining downstream target CyclinD1.
   • Clearly presented with useful illustrations throughout.
   • Logic is clear and examination thorough.
   • Conclusions are warranted on the basis of their findings.

   Limitations

   • This study primarily used visual analysis of fixed tissue images to assess the main outcomes. To reinforce the conclusions, these could be supplemented with live imaging to appreciate dynamics, or biochemical techniques to look at protein expression levels.
   • Some aspects of quantification require explanation in order for the experiments to be replicated.
   • It is imperative that precise sample sizes are included for all experiments presented.

   Advance:

   • First functional demonstration role for Cdkn1c in regulating neurogenic transition in progenitors.
   • Conceptual advance suggesting Cdkn1c has dual roles in driving neurogenesis: promoting neurogenic divisions of progenitors and the established role of mediating cell cycle exit previously reported.
   • Technical advances in the form of G1 signposting and endogenous Myc tagging using CRISPR in chicken embryonic tissue.

   Audience:

   Of broad interest to developmental biologists. Could be relevant to cancer, since Cdkn1c is implicated.

    Please define your field of expertise with a few keywords to help the authors contextualize your point Developmental biology, vertebrate embryonic development, neuronal differentiation, imaging. Please note that we have not commented on RNAseq experiments as these are outside of our area of expertise.

   Read the original source
5. 

   Version published to 10.1101/2024.10.10.617342v2 on bioRxiv

   Jan 15, 2025
6. 

   Version published to 10.1101/2024.10.10.617342v1 on bioRxiv

   Oct 12, 2024