Review Reports - Deletion of Clock Gene <i>Period1</i> (<i>Per1</i>) in Neurons but Not in Astrocytes Shortens Clock Period and Diminishes Light-Mediated Rapid Phase Advances in Mice

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In the mammalian circadian system, Period (Per) 1 and 2 genes play an important role in the core transcription-translation feedback loops that maintain circadian rhythms. Previous studies showed that mutation of the Per2 gene in all cells of mice shortens the circadian period, diminishes phase delays, and enhances phase advances. In contrast, deletion of Per1 shortens the period and induces only phase advances. Moreover, cell-type-specific deletion of Per2 in astrocytes or neurons also shortens the clock period, but only neuronal deletion affects phase shifts. However, the cell-type-specific contribution of Per1 is still unclear. In this study, the authors aimed to address this topic and further investigate the role of Per1 in neurons and astrocytes in modulating circadian period and phase shifts. The results show that neuron‐specific deletion of Per1 leads to a shorter circadian period, slightly reduced phase advances, and reduced circadian amplitude following a light pulse given late in subjective night (CT22) for mice maintained under constant darkness. In contrast, astrocyte-specific deletion of Per1 does not produce these effects. The study has the potential to contribute to current understanding of how Per1 deficiency in specific cell types affects circadian clock regulation and behavioral rhythms. Overall, the study conveys interesting findings. However, some points need to be clarified throughout, and methodological details to improve the manuscript further, as detailed below, came into my mind.

Comment #1: It would be great if the authors could rephrase the end of the introduction (lines 48-55) to present scientific hypotheses beyond the objectives and what remains unclear. Also, briefly explain how the experiments were conducted and highlight the key findings of this study.

Comment #2: The description of the IHC experiments (lines 316-317, and Figure 1) implies the impression that separate cohorts of mice were used for tissue collection at Zeitgeber 12. But at (lines 344-345) it stated that the tissue collection was performed after a recovery period of 2 weeks in LD conditions after the end of all treatments. Did you confirm the same findings with behaviorally tested animals under constant darkness conditions at the end of each treatment and at different circadian times? How many SCN were analyzed from each condition at the end of the trial? Figure 1 shows Per1 absence qualitatively; however, I am wondering whether fluorescence intensity was quantified. Did you confirm the successful deletion of Per1 using other quantitative methods such as western blot or in situ hybridization in the behaviorally tested mice?

Comment #3: The description of the light-pulse experiments (lines 331-343) suggests that the same mice were repeatedly tested at different circadian times (CT10, CT14, CT22), with ≥10 days of constant darkness between pulses. Please clarify whether the same cohort of mice was tested across all circadian times or whether a side-by-side design was also used. I am asking because I see a discrepancy in the sample sizes across conditions. If the same set of animals were used repeatedly, then I would expect the number of animals to be consistent across the circadian times, but this is not the case. If I am not mistaken, from the scatter plot points and the raw data, for example, it started with 12 measurements at NCO and per1NKO at CT10, but later at CT22, the number became 20 and 16, respectively. In comparison, the other groups showed similar counts (n =10) across the circadian times. Please verify and explain why the sample sizes within the same group differ. If so, adjust the figure legends to specify whether repeated or independent measures were used, and provide the exact number of animals contributing to each CT condition and each genotype.

Comment #4: For Figure 3A, why are the actograms for the wild type at different circadian times not shown? Particularly, five groups were compared in Figure 3B. Also, I see that multiple unpaired tests were used, which could be a valid method for handling two groups at a time from multiple comparisons, but correction adjustments for multiple comparisons when performing several pairwise tests, such as the Bonferroni correction, should be considered unless comparisons are independent.

Comment #5: Related to statistical analysis: Although the authors state that statistical significance was determined by ANOVA analysis or t-tests (lines 389-397), there are more details missing. Were the datasets normally distributed? Which test has been used to assess the data distribution? In Figure 4 legend, for example, nonparametric tests were used, the Wilcoxon matched-pairs test between two groups. Also, I recommend adding to the results section or as a supplementary table about the statistical significance and test values of the differences among the different conditions of the data you are presenting (e.g., t, F, df, etc.) which could help to understand the effect sizes and the significance of the findings (not just stating P < 0.05 or symbols). Additionally, please provide all information on the statistical tests used for two-independent or multiple-comparison analyses for each figure legend. Specify whether the same animals were used across circadian times, explain discrepancies in sample size, and adjust statistical analyses accordingly.

Comment #6: The way of the presentation of the within-group or between-genotype comparison is not consistent. In Figure 4, for example, all genotypes were shown together in a single plot, whereas in Figures 5-7, genotypes were split into subplots. Please unify and explain how the statistical tests were performed, either independently or across treatments. In Figure 4, for example, nonparametric tests were used, the Wilcoxon matched-pairs test to compare two groups. In Figure 5-7, ANOVA tests were performed for multiple comparisons between groups or before and after light pulses. Then I am wondering how you test the difference in the left plots (wt), where only two comparisons were shown. Please clarify and run the appropriate test.

Comment #7: I am missing information about the sex of the tested mice. (In line 301): It stated both male and female, without specifying which gender was used for the experiments. Please include the sex of the tested mice and analyze for sex-dependent effects if you used both. Additionally, I could not find a statement of the background of littermates and strains. Are all parental strains maintained in the same C57Bl/6 background or mixed backgrounds, and for how many generations have they been backcrossed? Even with mixed backgrounds, it is valid to compare effects within each cross. However, it should be clarified in the manuscript or in the limitations of the study, especially when comparing different lines, that some of that difference could reflect the strain background rather than solely the cell-type specificity of Per1.

Comment #8: Just a minor typo issue in figures 5-7, lower left bottom. I think it should be “LP CT22” instead of “LP CR22”.

Comment #9: The discussion section primarily elaborates on repeating the results and figures without offering novel perspectives. I am missing sufficient mechanistic clarification of the potential reasons for the results, which suggest that astrocytic Per1 has limited importance for circadian regulation, or of other factors that may have influenced the tested behavioral paradigms and parameters through neuronal mechanisms rather than astrocytic ones. Also, there are no mentions of potential limitations. What are the implications and future research that might benefit from understaanding the cell‐type specificity of molecular clock genes in the brain?

Comment #10: It is a bit confusing in the appendix when writing, “with and without LP.” Do you mean "before and after the light pulses were applied"? Or do you mean side-by-side treatments?

Author Response

We thank the reviewer for their time and for the constructive feedback given to improve this manuscript.

We have added three sentences at the end of the introduction to accommodate the reviewers suggestion.

The tissue collection was performed after the 2 week recovery period, only after the behaviorally tested mice have completed all the light pulses. The text present in line 316-317 was removed, to avoid the confusion created.

Only one time point was analysed, at ZT12 after the mice were transferred back into LD. This was to account for the highest Per1 presence in the SCN, as shown by Yan and Silver, 2002. In this case only one SCN was analysed per genotype of the mouse. We can observe Per1 absence qualitatively in the SCN of the Per1 Nko mouse, however it is not so easy to observe the diffences between the Per1Gko mouse and its control. Hence, this is why we used a higher magnification and decided to do a 3D reconstruction of an astrocyte. The fluorescence was not quantified in this experiment. Other methods such as western blot or in situ hybridisation will not allow us to visualise the deletion of Per1 in the exact specific cell type. FACS analysis is not sensitive enough as well. We highlight this wekaness now in the discussion.

Thank you very much for this observation. A side-by-side design was used to reach our desired number of mice in this study. Furthermore, in certain cases, not all mice performed well under the wheel-running experiments. The wheel-running activity of some mice was becoming too weak to be able to quantify the results. For some mice this was happening right from the start or after the administration of one or two light pulses. This could be due to age or potential sickness of those mice.

The reason more mice have been used for the Nco and Per1Nko mice at CT22 is because we have initially observed a difference between the male and female mice. So, to investigate that, we needed a higher number across both sexes, which resulted in no difference between the sexes. This is further explained in comment 7.

The figures have now been adjusted to include this information and the correct number of mice across all conditions.

The actograms for the WT mice are not present as they show similar shifting responses to the other control lines (Ncre and Gcre). Adding the wild type actogramms would make the figure less readable with no additional relevant information. That is why we decided to not show them. However, we are attaching actograms of WT mice below.

see attached files

CT10:

CT14:

CT22:

The comparisons were intented to be independent. Our goal was to compare the knock-out lines against their controls.

However, we have now performed a normality tests to compare the WT, Nco and Gco lines across all the light pulses, and parameters investigated (amplitudes, relative power of phase (FFT) etc.). If the results of the Shapiro-Wilk test showed that all groups have passed the normality test, we have procedeed to perform a one-way anova analysis, followed by Bonferonni’s multiple comparisons between these groups. If one of the groups did not pass the Shapiro-Wilk test, then we procedeed to perform a Kruskal-Wallis test to compare the groups. The tables with the results of the normality tests, one-way Anova and Kruskal-Wallis tests are shown in the document provided.

Thank you very much for this observation and the suggestion.

As per the answer to the previous comment we have now performed normality tests to assess the distribution between the Nco and Per1Nko and between Gco and Per1GKO across all the parameters investigated. If the results of the Shapiro-Wilk test were ok, then we performed the One-Way Anova followed by Bonferroni’s multiple comparisons, to asssess the control lines against the knock-out lines (with or withour light pulses). On the contrary a Kruskal-Wallis test was performed to check for the same differences.

The comparison of each mouse line, before and after the light pulse, across all parameters investigated, was performed via paired t-tests, whenever the datasets were normally distributed or via Wilcoxon matched paired tests when the data sets were not normally distributed.

The tables with the results of all normality, anova, and paired t-tests that showed statistically significant results have been attached to the document provided. The figure legends have also been updated to include the missing information.

The Figure 4 has now been changed to match figures 5-7.

As explained in comments 4 and 5 we have now performed normality of the WT, Nco and Gco followed by the correct analysis across all parameters investigated.

The same approach was applied to investigate the control lines against the Per knock-out lines. The differences between each mouse line before and after the LP (across all parameters investigated) have been analysed using paired tests. Hence, the WT mice on the left side of each figure were investigated in a similar way. As mentioned the tables showing all the statistically significant changes have now been provided in the document attached.

Both male and female mice were used across all groups. This has now been modified in the manuscript to be more clear. Differences beween male and female mice were investigated. Initially differences between male and female were observed in the Per1Nko mice after the LP at CT22. We increased the number of animals to deper investigate our initial observation. We did not observe statistically significant differences between male and female mice.

Regarding the background of the mice, they have all been maintained in the C57Bl/6 background, and crossed for at least 6 generations in our facility. This has now been clarified in the manuscript in the methods section.

Comment #8: Just a minor typo issue in figures 5-7, lower left bottom. I think it should be “LP CT22” instead of “LP CR22”.

Thank you very for this observation. This has now been corrected in the manuscript.

We have followed the reviewers recommendations.

Comment #10: It is a bit confusing in the appendix when writing, “with and without LP.” Do you mean "before and after the light pulses were applied"? Or do you mean side-by-side treatments?

We apologise for this confusion. Yes, we mean before and after the light pulses were applied, this has now been adjusted in the manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This work highlights the role of Per1 in modulating the neuronal and astrocytic contribution to the regulation of circadian rhythms. I think these results are very interesting and it will be a nice contribution to the field, however I have one main concern along with some comments/suggestions.

The wording of the results section should be improved as often the message communicated does not fully reflect the results and/or describes the changes, the authors often begin with the sentence “Quantitative analysis revealed no significant effect of the LP at CT XX across all strains” but then this is not what is actually show and it can be misleading and confusing.

Since GFAP is not expressed in all glia cell types, the authors should be more specific and refer to the KO mice as astrocyte-specific Per1 knockout (KO) mice; the expressions “glial-specific; glial Knock-out; glial control” should be changed to astrocytic - ……

The sentence “glial fibrillary acidic protein (GFAP) specific to glial cells including astrocytes” is not a completely correct statement. In the CNS, GFAP is a marker for astrocytes and some other cell types for instance radial glia, however many glia cell lineages, as an example oligodendrocytes, do not express GFAP, so the authors may want to be more specific

In the KO mice, Was Per1 absent in neurons and astrocytes in all brain regions or only in the SCN? Please clarify.

Fig 1: The red-green combination should be avoided to be mindful of readers who might be color vision deficient. Maybe the crossing strategy should be depicted in this figure as well.

Fig 2 legend, please add the light intensity.

How did the authors decide the times of the LP? Two seems to be based on the peak and trough of Per1 protein but how about CT 14? It might be good to add a sentence about this.

Line 75: GFAP and Per1 do not really co-localise as their cellular distribution is different, perhaps co-express or co-label are a better word choice.

My main concern is about the statistical tests chosen by the authors and the lack of comparison of the effects of the LP in the two Per1 controls with the WT.

The experimental design appears to have three variables: noLP/LP; genotype; CT or at least two variables noLP/LP and genotype, which should be considered when chosing the statistical test and data presentation.

Figure 3:

The statistical tests chosen do not seem appropriate, aren’t the authors comparing all the groups? They do repeatedly mention determining the effects on circadian parameters across all strains; hence, they should use at least One-Way ANOVA followed by a multiple comparisons test, or even better two-way ANOVA followed by a multiple comparisons test if both CT and exp groups are considered as varaibles (comparing changes across different CTs), these could be reported in a table.

Figure 4:

Same for the results in figure 4, if the authors want to claim that there was a significant effect of the LP at different times on all the groups, the test should be a two-way ANOVA followed by a multiple comparisons test

It should also be shown that the LP effects on the two controls NCO and GCO (again this needs to be changed to clarify that is targeting astrocytes) are not significantly different from the WT for all the measurements of the circadian parameters.

The authors wrote: “An LP at CT10 had no significant impact on period across all genotypes (Fig. 4A)” however in panel A: there is a significant effect on period at CT 10 in the Per1NKo mice vs their control.

“Panel C: Interestingly, an LP at CT22 resulted in a significant shortening of the circadian period in all genotypes examined (Fig. 4C, * symbols).“ were the periods of the NCo and GCo significantly different from WT with and without LP?

The title of some of the panels in the figure legends are not accurate and do not describe the results, for instance one example, Fig 5 panel A: “ Amplitude before and after an LP at CT10 is similar in all genotypes investigated;……..panel C: Amplitude before and after an LP at CT22 is similar in all genotypes investigated. “ These do not well reflect the results, the descriptions should be a bit more specific, are the authors just referring to the controls? There is no report of a statistical test for this claim, and it is misleading.

Figure 5 and 6, the F values should be reported for the one-way ANOVA, also are the p values ((*p < 0.05, **p < 0.01, n = 12, one-way ANOVA) reported from the Tukey’s post test?

Discussion: “Our experiments revealed that deletion of Per1 in astrocytes had no measurable impact on these parameters. “ This does not really seems to fully reflect the results obtained as Per1 absence in astrocytes had some mild but measurable effects.

“Our immunohistochemistry data support this interpretation (Fig. 1): we observed high levels of PER1 protein in SCN neurons (Fig. 1A), while PER1 signal in astrocytes was markedly lower (Fig. 1B). “ how many mice/genotype were used to support this claim? Why did the authors choose to perfuse the mice at ZT 12?

“Glial and neuronal Per1 deletion was verified by genotyping, “ please show these results and describe how it was performed, primers? A schematic of the breeding strategy and genotyping results should be added and shown in Figure 1.

Author Response

We thank the reviewer for their time and for the constructive feedback given to improve this manuscript.

We have addressed this justified concern and have adapted the results section accordingly.

We agree that GFAP is not expressed in all glial cells, but it is not exclusively expressed in astrocytes either. Hence, the suggestion to use astrocytes instead of glia is as incorrect as only to say glia. Since we state that the cre-driver is specific for GFAP positive cells, the correct formulation would be GFAP-positive cells. The reader can find this information in the methods section.

We state now that it is specific to some subtypes of glial cells including astrocytes …

In the KO mice, Was Per1 absent in neurons and astrocytes in all brain regions or only in the SCN? Please clarify.

Yes, Per1 was absent in all brain regions and not just in the SCN. This is described in the methods section where we use cre-driver lines that delete Per1 in Per1floxed animals from nestin-positive or GFAP-positive cells.

Fig 1: The red-green combination should be avoided to be mindful of readers who might be color vision deficient. Maybe the crossing strategy should be depicted in this figure as well.

We can understand this comment, however, Green/Red combinations are standard practice. The crossing strategy for Per1 floxed animals has already been published see reference 15.

Fig 2 legend, please add the light intensity.

The light intensity has now been included in the figure legend, and also in the methodology with the following text being mentioned in brakets “using 2 visible light spectrum neon light sources: Lumilux cool daylight, 1000 Lux, 18 W, OSRAM, Munich, Germany”

How did the authors decide the times of the LP? Two seems to be based on the peak and trough of Per1 protein but how about CT 14? It might be good to add a sentence about this.

The LP at CT14 was selected, as it is the time point in the subjective night, where mice exhibit a strong phase delay responses. Hence, we considered that if any changes in the phase delaying response were observed, then it should be reflected best at this time point. The manuscript has been adapted to include this information, inlcuding reference 4 for the PRC that is underlying our decision.

Line 75: GFAP and Per1 do not really co-localise as their cellular distribution is different, perhaps co-express or co-label are a better word choice.

We thank the reviewer for the suggestion. The wording has now been adapted in the manuscript.

My main concern is about the statistical tests chosen by the authors and the lack of comparison of the effects of the LP in the two Per1 controls with the WT.

Figure 3:

The aim of this study was to investigate the effect of Per1 deletion (in neurons or astrocytes) on the phase shifting response and other circadian parameters following different light pulses in the subjective night or day. We consider than in order adress that, the comparison of their responses should be made against their control mouse lines (Ncre and Gcre).

However, we have now performed a normality tests to compare the WT, Nco and Gco lines across all the light pulses, and parameters investigated (amplitudes, relative power of phase (FFT) etc.). If the results of the Shapiro-Wilk test showed that all groups have passed the normality test, we have procedeed to perform a one-way anova analysis between these groups. If one of the groups did not pass the Shapiro-Wilk test, then we procedeed to perform a Kruskal-Wallis test to compare the groups. The tables with the results of the normality tests, one-way Anova and Kruskal-Wallis tests are shown in the document provided.

Normality tests were also used to assess the distribution between the Nco and Per1Nko and between Gco and Per1GKO across all the parameters investigated. If the results of the Shapiro-Wilk test were positive, then we performed the One-Way Anova followed by with Bonferroni’s multiple comparisons, to assess the control lines against the knock-out lines (with or withour light pulses). On the contrary a Kruskal-Wallis test was performed to check for the same differences.

The comparison of each mouse line before and after the light pulse across all parameters investigated, was performed via paired t-tests, whenever the datasets were normally distributed or via Wilcoxon matched paired tests when the data sets were not normally distributed.

Figure 4:

In this figure we wanted to show how the LP at CT22 is able to reduce the period legth in each genotypes used in this study. We chose to perform a paired t-test, in order to compare the period length of each mouse before and after the LP in each genotype. The aim is not to compare all genotypes and their responses to light at the same time. Of interest in only either light or genotypes and their corresponding controls. So for light, paired t-test, for genotype and the corresponding control, one-was ANOVA.

Thank you very much for this comment. We have performed one-way anova analysis between the WT, Nco and Gco, across all light pulses to see if these mice show similar results before proceeding with our independent comparisons.

We have observed statistically significant differences in the period lengths only between the WT and Ncre mouse lines at the LPs CT14 and CT22. These results are now shown in the tables which are present in the document provided.

The significant effect observed in this comparison is between the Per1Nko and their control group which have not received a light pulse. Therefore the effect on the period observed in this case is due to the Per1 deletion in the neurons and not due to the LP administered at CT10.

We have performed one-way anova analysis between the WT, Nco and Gco, across all light pulses (with or without).

As mentioned earlier, we have observed statistically significant differences in the period lengths only between the WT and Ncre mouse lines at the LPs CT14 and CT22. These results are now shown in the tables which are present in the document provided.

Thank you very much for this observation. The titles of the figure legends have been changed to reflect the significant changes (if any) observed.

Figure 5 and 6, the F values should be reported for the one-way ANOVA, also are the p values ((*p < 0.05, **p < 0.01, n = 12, one-way ANOVA) reported from the Tukey’s post test?

The figure legends have now been updated to include the exact statistically significant values. The number of mice was reported for each condition as well as the F, df etc. values.

This is correct and we adapted the text accordingly.

Only one time point was analysed, at ZT12 after the mice were transferred back into the LD. This was to account for the highest Per1 presence in the SCN, as shown by Yan and Silver, 2002. In this case only one SCN was analysed per genotype of the mouse. We can observe Per1 absence qualitatively in the SCN of the Per1 Nko mouse, however is not so easy to observe the diffences between the Per1Gko mouse and its control. Hence, why we used a higher magnification and decided to do a 3D reconstruction of an astrocyte. This limitation in our expereimental set up is now mentioned in the discussion.

Genotyping was performed to assure successful cre-recombination as described in reference 15. The primers and the breeding strategy are described in that reference. We have clarified this also in the text of the methods section.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript addresses an interesting topic in the field of neuroscience: the role of the Per1 clock gene in modulating the circadian period and light-induced phase changes in both neurons and glial cells. While the study of and interest in this topic are not recent, the authors' proposal provides relevant information that makes the work a valuable contribution to the field. Overall, the manuscript is well-structured and written, allowing for a guided reading supported by robust evidence. The conclusions are appropriate to the study's objective and the scientific arguments presented. Regarding the cited literature, fundamental studies and recognized reviews in the field are mentioned; approximately one-fifth of the references are up-to-date. Minor revisions are suggested:

The diagram in Figure 2 is missing the information that, at the end of the last DD entrainment, the mice were returned to a 12:12 light:dark cycle.

The methodology fails to indicate the intensity of the light pulse used for light stimulation.

Words originating from Latin should be written in italics. For example, ad libitum.

Correct the wording of the following statement: The experimental designed (Fig. 2) was designed and performed according to the Aschoff Type I protocol.

Author Response

We thank this reviewer for taking the time to read our manuscript and for the constructive feedback to improve the manuscript.

The diagram in Figure 2 is missing the information that, at the end of the last DD entrainment, the mice were returned to a 12:12 light:dark cycle.

Thank you for the observation. We changed not the diagram in Figure 2 to show this missing part.

The methodology fails to indicate the intensity of the light pulse used for light stimulation.

The methodology has been adapted to include the missing information regarding the intensity of the light pulse. The following information has been included: 'using 2 visible light spectrum neon light sources: Lumilux cool daylight, 1000 Lux, 18 W, OSRAM, Munich, Germany'

Words originating from Latin should be written in italics. For example, ad libitum.

This has been changed in the text now.

Correct the wording of the following statement: The experimental designed (Fig. 2) was designed and performed according to the Aschoff Type I protocol.

Our wording sounds awkward indeed. The wording has been corrected.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have significantly improved the manuscript, and all my comments have been addressed and reflected in the updated version.

Author Response

The authors have significantly improved the manuscript, and all my comments have been addressed and reflected in the updated version.

We thank the reviewer for his time and constructive comments.

Reviewer 2 Report

Comments and Suggestions for Authors

It would have been easier to assess the manuscript if the authors had provided a copy with the changes made highlighted....

The authors have decided not to address several of my comments, which can be understandable, thus there are still several unclear sentences or some of the additional explanation requested have been added in a minimal form.

The authors need to clearly state their goals at the end of the introduction as well as the groups of comparison in the beginning, (the methods are at the end of the manuscript, the reader should not have to look for this information).

Previous Question: "How did the authors decide the times of the LP? Two seems to be based on the peak and trough of Per1 protein but how about CT 14? It might be good to add a sentence about this." The authors replied that they have modified the manuscript to explain this point but all I found is"The LPs were selected according to the phase response curve described in [4]." this sentence does not describe how they have chosen the time points just refers the readers to another paper, not helping to understand the experimental design and rationale.

Statistical test and data analyses: I suggest that the authors consider consulting an experienced statistician as some of the tests still do not seem appropriate:

AUTHORS response: (The comparison of each mouse line before and after the light pulse across all parameters investigated, was performed via paired t-tests, whenever the datasets were normally distributed or via Wilcoxon matched paired tests when the data sets were not normally distributed.)

The new graphs are much better at showing the different groups of comparison, however:

Fig 4 through 7 middle and right panels present with a 2X2 design (genotype and before/after LP), a two way anova followed by a multiple comparison test should be used. AS for the before and after LP presentation the authors could use the before and after paired t-test they should use a better presentation and add those graphs as an additional figure, in these, each before value is connected to the after by a line and gives a better idea of the variability in the animals' response (see attached image from graphpad/prism and this link: https://www.graphpad.com/support/faq/creating-a-color-coded-before-after-graph/)

There are too many symbols in the graphs and the legends are confusing and with missing information:

Fig 3 legend: "(*p = 0.0286, t = 2.551, DF = 44, n = 12, One-Way ANOVA, followed by Bonferroni’s multiple comparisons test)."

Fig 7 legend: "(*p = 0.0386, t = 2.397, DF= 68, n =16-20, for before the LP CT22 and, *p = 0.0363, t = 2.421, DF= 68, n = 16-20, for after the LP CT22, One-Way ANOVA, followed by Bonferroni’s multiple comparisons test) My question is Are the 't' values from the one-way ANOVA as that is what it seems from the wording.

Where are the F and DF/df values from the one-way ANOVA reported?

Thanks for adding the tables to the manuscript, however, these need better captions/legends and information on the numbers reported, what are these?

"We have performed one-way anova analysis between the WT, Nco and Gco, across all lightpulses (with or without). As mentioned earlier, we have observed statistically significant differences in the period lengths only between the WT and Ncre mouse lines at the LPs CT14 and CT22. These results are now shown in the tables which are present in the document provided." Where are these results reported in the manuscript?

The tables and values provided in the response to the reviewer comments should be also added to the manuscript, unless I missed them.

How many male and female mice were in each group? was there an equal number of male and females? Males and females may respond differently , the lack of sex differences analyses should be acknowledged and considered a limitation.

Results:

As mentioned in the previous comments the authors did not measure the levels of Per1 in neurons or astrocytes, and they only have a qualitative observation from one mouse, hence in the following sentence the word "expression" is more appropriate :
"Our immunohistochemistry data support this interpretation (Fig. 1): we observed high levels of PER1 protein in SCN neurons (Fig. 1A), while PER1 signal in astrocytes was markedly lower (Fig. 1B)."

Suggested "we observed higher expression/signal of PER1 protein in SCN neurons (Fig. 1A), which was markedly lower in astrocytes (Fig. 1B)."

The manuscript still contains a series of uncorrect statements:

"We generated glia-specific (Per1G; Per1/Gfap-Cre)....." the authors did not generate a "glia-specific" but an astroglia KO. In the CNS, GFAP is expressed early in development in radial glia (please note that Nestin is also a marker for radial glia), there is evidence that these cells then become precursors for different cell lineages in the CNS :

https://www.sciencedirect.com/science/article/pii/S0959438825000777#fig1 https://pmc.ncbi.nlm.nih.gov/articles/PMC4559283/

https://journals.biologists.com/dev/article/140/3/483/46074/Radial-glia-from-boring-cables-to-stem-cell-stars

https://www.sciencedirect.com/science/article/abs/pii/S0955067415000137?via%3Dihub

https://www.sciencedirect.com/science/article/pii/S0165027023002443

In the adult brain, GFAP is a marker for astrocytes.

Throughout the manuscript and in the title, the authors use the word "astrocytes". Their goal was "To assess the impact of light on mice lacking Per1 in neurons or astrocytes,........", so why keep calling the mouse line "glia-specific"?

Sentences like the two below are misleading :

"Lack of Per1 in neuronal or glial knock-out mice".

"Consistent with these findings, most behavioral circadian parameters assessed in mice lacking Per1 in glial cells (Per1GKo) were similar from controls."

Likely lack of Per1 in Oligodendrocytes, microglia, Ependymal cells, NG2 glia/polydendrocytes will have very different effects. Since the other glia lineages are still expressing Per1, these can't be assessed.

If they argue that: "Hence, the suggestion to use astrocytes instead of glia is as incorrect as only to say glia. Since we state that the cre-driver is specific for GFAP positive cells, the correct formulation would be GFAP-positive cells. " than they are not assessing the effects of lack of Per1 in astrocytes but in all the GFAP-expressing cells...... i.e.: radial glia in the developing central nervous system and Schwann cells in the peripheral nervous system.

Perhaps they should use another marker (Aldh1) for astrocytes and see how much the two populations overlap, so than they can be less general.
They should to change their objective as their argumentation contradicts it and stop using glia and astrocytes as synonyms as they are not: "Consistent with these findings, most behavioral circadian parameters assessed in mice lacking Per1 in glial cells (Per1GKo) were similar from controls."

"(GFAP) specific to certain subtypes of glial cells including astrocytes." the subtypes or lineages should be listed, this statement also need references.

This might seem silly but in regard to the use of red and green in the images, this USED to be standard practice, many journals now ask not to use such combination anymore.

The limitation paragraph should be revised to include more.

In my opinion the manuscript still needs thorough revision before being accepted.