Author response:
The following is the authors’ response to the original reviews.
Public Reviews:
Reviewer #1 (Public Review):
In this manuscript authors examined the effect of rif1 knockout on replication timing and transcription in early embryos of zebrafish. Contrary to the expectation, genome-wide replication timing domains did not significantly change upon Rif1 knockout, although the replication timing became less dynamic in the mutant, meaning the entire genomes are replicated toward the mid S. In contrast, transcriptional profiles change by rif1 mutation throughout the embryo stage. These effects were more predominantly observed after gastrulation at the early stages of zebrafish development.
The results presented in this manuscript provide new information on the effects of rif1 mutation on early zebrafish development, although the underlying mechanism has not been explored. The information is useful for researchers in the field of early development, with specific focus on replication and transcription regulation.
The genome wide analyses of replication timing has been conducted and analyzed properly. The transcriptional analyses are conducted by RNA-seq and SLAM-seq (determining the nascent mRNA), and the results convincingly show the overall transcriptional patterns at different developmental stages.
This work shows that Rif1 regulates replication timing and transcription in zebrafish embryos, while the extents of the effects vary during the developmental process. Although the data convincingly illustrate the whole picture of Rif1 KO on replication and transcription during zebrafish development, the mechanistic insight is missing. Especially, how Rif1 may or may not coordinately regulate replication and transcription during the zebrafish development has not been addressed.
We thank the reviewer for recognizing the value of combining genome-wide replication-timing, RNA-seq, and SLAM-seq analyses across zebrafish development. We agree that the original study did not establish a molecular mechanism linking Rif1-dependent transcriptional and replication-timing effects. To address whether these effects are locally coordinated, we added a gene-centred analysis comparing replication-timing values for genes with increased, decreased, or unchanged transcript abundance at Dome (Figure 5--figure supplement 2). Differentially expressed genes did not show a clear enrichment in early- or late-replicating regions, either at Dome or at pre-MBT. These results argue against replication timing state being the primary determinant of the Dome-stage transcriptional changes. We also expanded the Discussion to explain the limitations of the current study and the need for future measurements of origin use, fork progression, chromatin state, and cell-type-specific effects. The new discussion of Nakatani et al. (2025) further places our findings in the context of evidence that Rif1-dependent replication-timing changes can be uncoupled from transcriptional changes.
Reviewer #2 (Public Review):
This study by Masser et al. analyzes global replication timing and gene expression in rif-1 null zebrafish. This work is an extension of their previous report on the normal replication timing pattern during wild-type zebrafish development. The major valuable finding here is that Rif1 is not essential for viability in zebrafish, and - counter to expectation from studies in cultured cells and other species - late replication does not strongly depend on Rif1. Instead, the data suggest that Rif1 subtly sharpens replication timing pattern during normal development rather than function generally to delay replication timing. In the absence of Rif1, the normal pattern establishment is somewhat delayed. The authors also document some changes in expression during development with more genes being repressed by Rif1 than activated at some early stages.
The study and analysis are generally rigorous, and the conclusions are supported by convincing data. The manuscript is well written, though there are aspects of the presentation that could be improved for a broader scientific audience. Given the strong link between replication timing and cell type/development, studying timing in a whole developing organism is important. The experimental approach is technically challenging, particularly the bioinformatic analysis. The scientific advance here is largely confined to documenting the timing of Rif1-affected transcription, the unanticipated effect of the rif1 deletion on replication timing and on sex determination, though the latter is not explored. The work is descriptive and feels like two relatively unconnected studies, transcription and replication plus a small bit of development, and the difference in timing of the transcription phenotypes and replication phenotypes suggests they may be very distinct Rif1 roles. There isn't a lot of new insight into the mechanism of how Rif1 affects either replication timing or gene expression. As such, the overall study is an useful set of findings and detailed data for future work, but it doesn't make a big step forward in understanding the role of Rif1 or the biological processes it affects.
Weaknesses worth addressing include the following:
(1) Loss of Rif1 did not affect viability, but it did strongly influence sex determination, resulting in a lower population of females. This effect is the strongest organismal phenotype, but the study provides no explanation for the loss of females from the data gathered here.
(2) The approach to distinguish nascent zygotically expressed mRNAs from maternal mRNAs is a strength. Are the differentially expressed genes related at all to regions of the genome whose replication timing is most affected? Are any of them related to the sex determination or developmental phenotypes?
We thank the reviewer for recognizing the rigor of the analyses and the value of studying replication timing in a developing vertebrate. We revised the manuscript extensively to make the experimental logic, zebrafish developmental context, replication-timing analyses, and figure legends more accessible to a broad audience. We also quantified the gastrulation phenotype, showing an approximately one-hour delay in completion of epiboly in maternal-zygotic rif1 mutants rather than a persistent developmental arrest.
We agree that the mechanism underlying the sex-ratio phenotype remains unresolved. The transcriptomic experiments were performed in whole embryos at stages much earlier than zebrafish sex determination and therefore cannot resolve changes in primordial germ cells or supporting gonadal somatic cells. We have avoided making a mechanistic connection between the early embryonic transcriptional changes and the adult sex-ratio phenotype and identify this as an important area for future study. To address the relationship between transcription and replication timing, we added Figure 5--figure supplement 2. Genes with increased or decreased transcript abundance at Dome were not preferentially associated with early- or late-replicating regions. Together with the distinct developmental timing of the transcriptional and replication-timing phenotypes, this supports the interpretation that Rif1 has separable roles in the two processes rather than a single local mechanism that directly couples them.
Reviewer #3 (Public Review):
Using the zebrafish model system, this manuscript assessed the roles of Rif1 protein in replication timing control and transcription during early development, and successfully demonstrated the differential impact of Rif1 protein in replication timing control and transcription. Moreover, the comprehensive assessments of the impacts of mutating Rif1 on animal development (including animal survival and sexual development) were assessed. Although there are works that examined Rif1's implications in replication timing and transcription separately, this work is unique in assessing all these points at once.
The strength of this manuscript is the genomic analyses of replication timing and transcription being combined in a single model system. Consequently, this manuscript clearly demonstrates the differential impact of Rif1 in these processes during zebrafish development.
The weakness of this manuscript is, as the authors comment in the Discussion, analyses of replication timing and transcription were performed using bulk embryos. There is a possibility that tissue-specific changes could have been masked. Tissue-specific or single-cell analysis in the future will fill the gap in the knowledge.
Some of the findings presented in this manuscript are consistent with previous findings using different models such as Drosophila and mice, whereas other findings do not necessarily agree. I hope further studies will reveal more clearly what is common in these systems, and what is different.
Also, the suggestion that the Rif1 protein may be implicated in a function similar to Fanconi-Anemia genes/proteins is very intriguing.
Overall, the data presented in this manuscript sufficiently justify the authors' claims. Moreover, this manuscript provides interesting insights into Rif1's function, as well as how development could be controlled.
We thank the reviewer for highlighting the strength of analyzing replication timing, transcription, and developmental phenotypes in the same vertebrate model. We agree that bulk-embryo measurements may mask tissue- or cell-type-specific effects. We now emphasize this limitation and the need for future tissue-specific or single-cell studies, particularly in the cell populations relevant to sex determination. We also expanded the cross-species context by discussing the recent mouse-embryo study by Nakatani et al. (2025), which supports a conserved role for RIF1 in consolidation of the replication-timing program while also indicating that replication-timing and transcriptional effects can be uncoupled. We agree that defining which Rif1 functions are conserved across zebrafish, mouse, Drosophila, and other systems, including possible relationships to Fanconi-anaemia pathways, will be an important direction for future work.
Reviewing Editor:
While the paper was under revision, a relevant paper from the Torres-Padilla lab was published (Nakatani et al., Developmental Cell, 2025). It complements these studies and cites the previous version of this manuscript. I suggest adding a reference in the Discussion to support the conclusions.
We thank the Reviewing Editor for bringing the recent study by Nakatani et al. to our attention. We have added a standalone paragraph near the end of the Discussion explaining how this work complements our findings, and we have added the complete reference to the bibliography. The new Discussion text reads:
“A recent study in mouse embryos independently identified RIF1 as a regulator of the developmental consolidation of the RT program. RIF1 depletion produced a less-defined, developmentally immature RT program, while RIF1-dependent RT changes were not correlated with transcriptional changes (Nakatani et al., 2025). Together with our findings in zebrafish, these results support a conserved role for RIF1 in sharpening replication timing during vertebrate development and indicate that its effects on replication timing can be uncoupled from changes in gene expression.”
Recommendations for the authors:
Reviewer #1 (Recommendations For The Authors):
The results presented in this manuscript provide new information on the effects of rif1 mutation on replication and transcription during early zebrafish development, although the underlying mechanism has not been explored. I suggest authors consider conducting the following experiments.
(1) Does replication timing domains have any role in Rif1-mediated regulation of transcription? It is not clear from the data presented whether transcriptionally affected genes are in the early replicating domains or late replicating domains (that appear after the shield stage). This should be examined.
We thank the reviewer for this helpful suggestion. To address whether transcriptional effects in rif1 mutants are associated with replication timing, we assigned each gene the nearest smoothed replication timing value and compared replication timing distributions for genes whose transcript levels increased at Dome, decreased at Dome, or were not significantly changed. This analysis is now shown in Figure 5—figure supplement 2. Genes with increased or decreased transcript abundance at Dome did not show a clear enrichment for either early- or late-replicating regions relative to genes with no significant transcript change. This was also true when replication timing was examined at pre-MBT, the stage preceding the major transcriptional changes detected at Dome. These results argue against replication timing state being the primary determinant of the Dome-stage transcriptional changes observed in rif1 mutant embryos. We have revised the Results to describe this analysis and added Figure 5—figure supplement 2.
(2) It is of interest whether the Rif1-mediated regulation of transcription and replication are mediated by a common mechanism, e.g. through alteration of chromatin structures. Close look at the data in Figure 3D indicates that some genome segments convert replication timing or undergo significant changes of replication timing. It would be informative to know whether these segments (Rif1-regulated replication domains) are associated with the genes whose expression change upon rif1 knockout.
We thank the reviewer for this insightful suggestion. We agree that an association between Rif1-dependent replication timing changes and Rif1-dependent transcriptional changes would be informative, and we considered this analysis. We attempted to identify Rif1-regulated replication timing domains using the same approach that we previously used to define developmentally regulated timing domains. However, the effect of Rif1 loss differed qualitatively from the developmental timing switches described in our prior work. Rather than producing a limited set of discrete timing-domain transitions, Rif1 loss caused a broad reduction in the dispersion of replication timing values across the genome, consistent with a general flattening of the timing profile. Under these conditions, an unbiased domain-calling approach preferentially identifies genomic regions with the most extreme early or late timing values in wild-type embryos, because these regions show the largest shift toward the mean in rif1 mutants. Thus, the resulting “Rif1-regulated replication domains” largely reflect the strongest wild-type timing domains rather than a discrete set of Rif1-specific regulatory intervals. For this reason, we do not think that assigning differentially expressed genes to such domains would provide a meaningful test of whether Rif1 regulates transcription and replication timing through a common local mechanism. Instead, we have now added a gene-centred analysis comparing replication timing values for genes with increased, decreased, or unchanged transcript abundance at Dome (Figure 5—figure supplement 2), which directly addresses whether transcriptionally affected genes are associated with early- or late-replicating regions.
(3) Replication is analyzed only by timing analysis. Authors need to analyze frequency of origin firing and replication fork rate by DNA fiber analyses to see whether they are affected by rif1 knockout at various stages of development.
We agree that measuring origin firing frequency and replication fork rate would provide valuable additional information about how Rif1 loss affects the replication program. However, performing DNA fibre analyses across multiple zebrafish developmental stages and genotypes would require substantial optimization and experimental expansion beyond the scope of the current revision. The current study was designed to measure genome-wide replication timing and transcript abundance across developmental stages, rather than single-molecule replication dynamics. We therefore have not added DNA fibre experiments. Instead, we have revised the Discussion to acknowledge this limitation and to clarify that replication timing reflects the combined effects of origin usage, fork progression, fork directionality, and fork stability. We added the following text to the Discussion:
“A further limitation of this study is that we concentrated on replication timing without directly measuring other features of the replication program that contribute to this timing. These features include origin usage, replication fork spacing, fork directionality, fork progression, and fork stability. A more comprehensive understanding of how Rif1 loss affects these parameters will be important for defining the relationship between Rif1-dependent changes in replication timing and transcription.”
Figure 4B, D and F: I did not see the blue lines which represent preMBT in the panels shown.
We thank the reviewer for identifying this error. The pre-MBT data were not intended to be shown in Figures 4B, 4D, and 4F. We have corrected the figure legend by removing the reference to the blue pre-MBT line.
Line 270: Figure 4G should be Figure 6G.
We thank the reviewer for identifying this error. We have corrected the figure reference from Figure 4G to Figure 6G.
No description of Figure 6E and 6F in the main text.
We thank the reviewer for noting this omission. We have added text to the Results describing Figures 6E and 6F. The revised text explains that Dome Up-DEGs are normally upregulated from pre-MBT to Shield stages but show earlier upregulation in rif1 mutant embryos, whereas Dome Down-DEGs normally decrease between Dome and Shield stages but show earlier reduction in mutant embryos.
Reviewer #2 (Recommendations For The Authors):
(1) This study is an extension of the lab’s previous work which established the wild-type genome-wide replication timing pattern during zebrafish development. The experimental details and analysis are described in the methods, but the general strategy is sometimes treated very cursorily. A non-expert can only understand parts of it by going back to the Seifert study.
We thank the reviewer for pointing this out. We agree that the replication-timing strategy should be understandable without requiring readers to consult our previous study. We have revised the manuscript to explain the general logic of the assay more clearly. Specifically, we now state that replication timing was inferred from copy-number differences between S-phase and G1-phase genomic DNA: genomic regions that replicate early in S phase are enriched in S-phase DNA relative to G1 DNA, whereas later-replicating regions are less enriched. We also clarified that pre-MBT, dome, and shield embryos were treated as S-phase samples because most cells are in S phase at these stages, whereas nuclei from bud and 24 hpf embryos were sorted by DNA content to isolate G1 and S-phase fractions. These additions make the experimental design and interpretation of the replication-timing profiles clearer in the main text and Methods.
Figure 2 is meant to document developmental delay in early embryos, but the differences between the single wt and mutant examples in 2D are poorly described and labeled. Most readers will be unfamiliar with the specifics of zebrafish development. There is also no quantification of this developmental phenotype, and that quantification should be included along with better labeling and description of 2D.
We thank the reviewer for pointing this out. We agree that the developmental delay shown in Figure 2D required clearer explanation and quantification for readers who are less familiar with zebrafish gastrulation. We have revised the Results to explain that epiboly is the process by which the blastoderm and yolk syncytial layer move toward the vegetal pole to envelop the yolk cell, and that zebrafish gastrulation stages are commonly described by the percentage of yolk coverage. We also added quantification of this phenotype. At 10 hpf, most wild-type embryos had completed epiboly, whereas most rif1 mutant embryos had not: 18 of 24 wild-type embryos, but only 2 of 24 mutant embryos, had reached 100% yolk coverage. By 11 hpf, all wild-type and mutant embryos had completed epiboly. These revisions clarify that rif1 mutant embryos show an approximately 1-hour delay in epiboly completion rather than a persistent arrest in gastrulation.
(3) The presentation could be greatly improved with additional information about the experimental approach and display. As written, the text and figure legends assume readers are intimately familiar with replication timing experiments, zebrafish development, and differential gene expression analysis. Most of the figure legends are not sufficient to understand the figures themselves, and the necessary information is also not always in the results. An example is Figure 3 which is not well described (other than the PCA plots); the term “lag” which is the x-axis in 3C is not defined.
We thank the reviewer for this helpful comment. We agree that several aspects of the replication-timing analysis required clearer explanation for readers who are less familiar with replication-timing experiments. We have revised the Results to explain the logic of the replication-timing assay more clearly and have added a more detailed description of the autocorrelation analysis in Figure 3C. Specifically, we now explain that autocorrelation measures how similar replication-timing values are across increasing genomic distances along the same chromosome, providing a quantitative readout of the peak-and-valley structure of the timing profile. We also clarified that increasing autocorrelation across hundreds of kilobases reflects the progressive establishment of broader replication-timing domains during development. In addition, we changed the x-axis label in Figure 3C from “lag” to “Genomic distance (Mb).” Together, these changes should make the experimental approach and display easier to understand without requiring readers to consult our previous replication-timing study.
Figure 4 is generally poorly described and labelled (4B, D, and F graph legends indicate preMBT in the data, but there are no blue lines on the graphs), and Figures 6 and 7 are quite busy.
We thank the reviewer for pointing this out. We agree that the Figure 4 legend incorrectly described the data shown in panels B, D, and F. The pre-MBT data were not intended to be plotted in these panels, and we have removed the corresponding reference from the figure legend. We recognize that Figures 6 and 7 contain several analyses, but we have retained the current organization because the panels in each figure address a connected set of questions. Figure 6 summarizes how Rif1 loss affects abundance of developmentally regulated transcripts, whereas Figure 7 extends this analysis by directly measuring nascent transcription using SLAM-seq.
Reviewer #3 (Recommendations For The Authors):
I do not think any additional experiments are required to justify the authors’ claims. Well done! However, for readers’ benefit, I propose the following changes or adding more explanations:
(1) Page 2, line 86: I guess “single copy” means “single copy per haploid”. Better to clarify this point.
We thank the reviewer for this helpful clarification. The reviewer is correct that “single copy” refers to a single copy per haploid genome. We have revised the text to state that the zebrafish genome has a single copy of the rif1 gene per haploid genome.
(2) Related to the data presented in Figure 2C, do you have an explanation for why sex determination is affected in the heterozygotes, despite the change in Rif1 expression being subtle (Figure 1C)?
We thank the reviewer for raising this point. We agree that the reduction in whole-embryo rif1 mRNA levels in heterozygotes appears modest relative to the sex-ratio phenotype. At present, we can only speculate about the basis for this difference. One possibility is that whole-embryo mRNA measurements do not accurately reflect Rif1 abundance in the specific cell populations that influence zebrafish sex determination, such as primordial germ cells or their supporting somatic cells. We have therefore avoided making a strong mechanistic conclusion from the heterozygous phenotype.
(3) Related to the data presented in Figure 2D, did you observe a delay in heterozygotes?
We thank the reviewer for this question. We have not quantitatively analyzed epiboly progression in heterozygous embryos. However, we did not observe an obvious developmental delay in heterozygotes during early development. The delay shown in Figure 2D was observed in maternal-zygotic rif1 homozygous mutants.
(4) Figure 3D: it is not easy to distinguish WT and mutant lines, particularly for the Bud stage. Please consider changing the colour schemes or other aspects. For example, making colour lines thinner may help.
We thank the reviewer for this helpful suggestion. We agree that the wild-type and mutant profiles in Figure 3D, particularly at the bud stage, were difficult to distinguish in the original version. We have revised Figure 3D by reducing the line width of the colored profiles, which improves the contrast between the wild-type and mutant traces.
(5) Figure 3E: Could you avoid overlapping of WT and mutant plots?
We thank the reviewer for this suggestion. We considered separating the wild-type and mutant density plots in Figure 3E, but we have retained the overlaid format because the purpose of this panel is to directly compare the distributions of replication timing values between genotypes at each developmental stage. Overlaying the plots makes the reduced dispersion of timing values in the rif1 mutants easier to visualize relative to the corresponding wild-type distribution.
(6) Figure 4C and 4E: the point legends (WT and mutant) do not match the points used in the graph.
We thank the reviewer for noting this potential source of confusion. In Figures 4C and 4E, point shape indicates genotype, with open squares representing wild-type samples and open circles representing rif1 mutant samples. Point color indicates developmental stage. We used separate visual encodings for genotype and stage to avoid a large legend containing every genotype-stage combination. To make this clearer, we have revised the figure legend to state explicitly that point shape denotes genotype and point color denotes developmental stage.
(7) Figure 4D: Very difficult to recognise 24 hr mutant line. Please improve the way there are shown.
We thank the reviewer for this helpful suggestion. We agree that the 24 hpf mutant profile in Figure 4D was difficult to distinguish in the original version. We have revised the figure by changing the appearance of the mutant lines to make them more visible while preserving the stage color scheme.
(8) Related to data presented in Figure 4B. Is it possible to show a statistical evaluation of all (or a reasonably large number of samples from) DARs?
We thank the reviewer for this suggestion. Figure 4A already provides a genome-wide analysis of the DAR set shown by example in Figure 4B. Specifically, Figure 4A plots the change in replication timing from shield to 24 hpf for all 2,498 putative enhancer-associated DARs in both wild-type and rif1 mutant embryos. The strong correlation between wild-type and mutant values indicates that DAR-associated timing changes are largely preserved in rif1 mutants. Because all DARs used for this analysis are included in the scatterplot, we did not add a separate statistical analysis of selected examples from Figure 4B.
(9) Page 8, line 220: It is unclear what “all” means. Is it all the available replication timing values genome-wide? Please clarify.
We thank the reviewer for noting this ambiguity. In this sentence, “all” refers to all genome-wide replication timing values calculated from the genomic windows used in our replication timing analysis. We have revised the text to make this clearer.
(10) Figures 6C and 6D: Colour labels are too dark and it is almost impossible to read texts inside. Please reconsider the colour scheme.
We thank the reviewer for pointing this out. We agree that the labels in Figures 6C and 6D were difficult to read because of insufficient contrast. We have changed the text colour inside the colored boxes to white to improve legibility.
(11) Related to overall transcription studies: Is there any sign that Rif1 mutation affects the transcription of genes involved in sex determination?
We thank the reviewer for raising this interesting question. We have not specifically analyzed whether genes involved in sex determination are differentially expressed in the early embryonic transcriptome data. Because zebrafish sex determination occurs substantially later than the embryonic stages analyzed here, and likely depends on specific cell populations such as primordial germ cells and supporting gonadal somatic cells, we do not think the current whole-embryo RNA-seq data can directly resolve this question. We therefore avoid drawing a mechanistic connection between the early transcriptional changes and the adult sex-ratio phenotype. Determining whether Rif1 mutation affects transcription in the cell populations that regulate zebrafish sex determination will be an important direction for future work.
