Talk:PLOS/Origins of DNA Replication

Reviewer 1: Alan Leonard

This review by Alan C. Leonard is based on the version of Origins of Replication topic-page by Babatunde Ekundayo and Franziska Bleichert as of 03:26, 27 January 2019

This is a nice overview on the topic of replication origins. It provides broad coverage of origins in bacteria, archaea, and eukaryotes. The writing is clear and the content is well-organized. The figures are appropriate for the topic and the authors have done a particularly good job building the bibliography. To improve the article, I suggest a few changes below:

1. Fig. 2A is out of date and should be updated to reflect current knowledge. It appears to be based on a figure in ref. 15 which also did not reflect the state of the field at the time of its publication. Only 3 high affinity DnaA boxes with sequence TTATCCACA (show this in the figure key) exist in E. coli. They are designated R1, R2, and R4 and should be noted as the only “red” boxes. R5(M) is a low affinity site. R3 is only identified by its sequence similarity to a perfect box, but it is a low affinity site and DnaA interactions with it have become less clear based on high resolution footprints. For a more accurate and recent view of the architecture of E. coli oriC, please look over the figure in reference 46 as well as the figures in two references that are not cited in the topic page: 1) Grimwade et al., 2018, Origin recognition is the predominant role for DnaA-ATP in initiation of chromosome replication, Nucleic Acids Research, 6140-6151 and 2) Sakiyama et al. 2017, Regulatory dynamics in the ternary DnaA complex for initiation of chromosomal replication in Escherichia coli, Nucleic Acids Research, 12354-12373. These references should be included in the bibliography. The authors might also want to briefly address the finding in the first reference that the predominant role of DnaA-ATP in E. coli is in origin recognition (DnaA-ADP can replace the ATP form if allowed access to low affinity sites).

It is not clear what the two different colors represent in the key for low affinity sites in Fig. 2A. tau and I sites appear to be distinctive, but it is not clear why unless the authors are just distinguishing tau from I. The naming of these sites was never very meaningful and it would be appropriate to treat them all as simply low affinity DnaA binding sites, most of which have a preference for DnaA-ATP.

The sites for the interaction of DNA bending proteins are also not noted in Fig. 2A although IHF is shown in 2C. Since this is an important feature of the loop back model, these binding sites might be shown in 2A and an appropriate reference to the dynamic interaction of IHF with E. coli oriC might be added (perhaps Ryan et al. 2004, Molecular Microbiology, 1347-1359)

2. The authors provide no other examples of bacterial replication origins other than E. coli. Since they mention bipartite versions and the fact that other arrangements of bacterial origins exist, it would help to show a few examples. These could be limited to a bipartite version (H. pylori) and perhaps one other that looks different than E. coli. These could be added to Fig. 2 as part D.

3. Figure 4 is done in an interesting way to designate the interaction of ORC protein domains with features within eukaryotic origins. However, the proteins represented by the purple blob on the left half of the metazoan origin (HMGA1a, LRWD1 etc) are described in the text as partner proteins that interact with ORC. I assume there is no designation of interaction (a colored line) because the exact subunit involved Is not known, but perhaps some way to show ORC interaction could be devised so that it was obvious from the figure that these are ORC interactive proteins as opposed to features where direct ORC interactions do not exist or are yet to be determined.

4. I wonder if it is appropriate to include mention of the recent paper by Sima et al. (Cell, 2019, 176, 1-15, https://doi.org/10.1016/j.cell.2018.11.036) where cis-elements are identified on mammalian chromosomes that regulate the timing of replication origin activity during the cell cycle? I leave this up to the discretion of the authors, but it appears to be another feature that could be added to Figure 4.

5. A minor issue; the title of some references are presented in all caps. Is this of special significance?

Reviewer 2: Ralf Wellinger

The authors provide a comprehensive Topic Page that summarizes most of what we know about the control of DNA replication onset. I only have a few comments that might aid to relate this interesting topic to other research topics such as transcription, genome instability, epigenetics or biological engeneering :

1. The authors mention that replisome has to cope with other DNA-dependent processes such as transcription, and that the collision between the replication and transcription machinery could lead to genome instability. In addition to reviews, it might be worth to include some original research papers on this topic including early work in E. coli or S. cerevisiae (doi.org/10.1016/0092-8674(88)90086-4; DOI: 10.1126/science.272.5264.1030).

2. I am missing a comment on the identification and use of replication origins for shuttle vectors that allow the propagation of the same plasmids in different organisms. One example would be DOI: 10.1128/MCB.2.3.221.

3. The authors show epigenetic features involved in ORC recruitment and origin function. I would include original references on the observation that CpG islands in mammalian chromosomes function as replication origins (see work from Adrian Bird’s lab).

4. The authors could include a comment on the possibility of non-canonical DNA replication events that bypass the need for replication initiation at DNA replication origins. These events have been described in E. coli, S. cerevisiae or H. volcanii (see the following review: https://doi.org/10.1016/j.tim.2017.12.001).

Reviewer 3: Michael R. Botchan

The topics page essay written by Ekundayo and Bleichert is scholarly, timely and will be of interest to readers at all levels. Textbooks are still trying to catch up with the ever deeper and evolving fields that contribute to the focus of this piece that summarizes advances in the latest research on the structures of the protein machines and the genome information that duplicate DNA in the three kingdoms of life. I particularly enjoyed the historical perspective and the pun received from the title. The piece is more than a review of the cis acting elements that allow for DNA replication. I had but 3 points to be considered in minor revisions.

1. Along the lines of history, I do think it would be worth a few more sentences as to how and why Jacob, Brenner and Cuzin purposed the "replicon model" for DNA replication. It would serve more than to give context as I believe it feeds into present situations about complexity with regard to eukaryote DNA replication and the dangers of oversimplification, without taking away the enormous credit that our pioneers deserve. Back in that day some plasmids where incompatible with each other- hence without selection one might overtake another and said to be incompatible, others not. Thus a positive replicator with a cis acting element shared by incompatible plasmids were postulated and compatibles had different "ori's" and a different replicator. Genetics provided a mechanism. Jacob and Monod thought that every control for transcription would be negative ( before evidence for positive pushed forward) and DNA replication might be positive. Work eventually uncovered for the E. coli chromosome (and plasmids) showed that both negative and positive regulatory pathways exist. I do appreciate the concise focussed discussion in the present piece.

2. Figure 4 is great, especially the expanded view illustrating the various ways that different elements of ORC might recognize DNA. As complex and current this picture is there are still other layers of control suggested by the data in the literature that mediate DNA replication initiation. For example DNA replication mediated amplification of the Chorion protein coding genes are mediated by cis acting elements that do not bind ORC specifically but ORC itself is spread throughout a zone. Negative regulation provided by the Myb-MuvB complex keeps licensed Mcm's and cis acting ORC's from executing downstream replication and loss of Myb leads to no replication. The earliest work was published in Beall et al 2002, Beall et al 2004 and Lewis et al 2004. Myb is needed to activate and counter the repressive effects of E2F2. Further work form others shows that E2F1 is also required for positive replication perhaps interacting directly with Orc in very early but repressed stages of oocyte development! The point really is that in metazoans- especially when one is dealing with real systems not in a petri dish most DNA replication controls at the cis level are multilayered and staged developmentally. I point out that there is no genetic evidence to date that really resolves what the highly conserved BAH domain of ORC1 is doing although it does touch nucleosome in a clear way as indicated. Deletions of the domain in both budding and fission yeasts are viable as they may be so in metazoans as well. To digress to make a subsequent point: The ABF1 binding site in ARS 1 is not found at other ARS elements yet the data is pretty good that it plays a role in efficiency for that ARS. Similarly ORC interactions with nucleosomes through the BAH domain probably will be substantiated with better genetics. So- I think some some indication as to the need for case by case examination of "Ori's" might be something to anticipate in the future. This long paragraph links to the second paragraph above.

3. The references are current and dealt with very clearly. I would suggest that this paper be referenced: Dynamics of DNA replication in a eukaryotic cell. Kelly T, Callegari AJ. Proc Natl Acad Sci U S A. 2019 Mar 12;116(11):4973-4982. doi: 10.1073/pnas.1818680116. Epub 2019 Feb 4. This recent work provides a compelling meta-analysis and then a computational model for replication origin site use in S.pombe. The requirements for A/T rich sites mediated by the ORC4 extension unique to this pombe ORC's is indicated obliquely in the review now submitted to PLOS and might be highlighted a bit more. One conclusion from this new paper from Tom Kelly is that I find important is that "there are many more potential initiation sites in the S. pombe genome than previously identified and that the distribution of these sites is primarily determined by two factors: the sequence preferences of the origin recognition complex (ORC), and the interference of transcription with the assembly or stability of prereplication complexes (pre-RCs)."

Response to Reviewers

We would like thank Drs. Leonard, Wellinger and Botchan for their careful review of our Topic Page – Origins of DNA Replication. We very much appreciate their positive and helpful feedback, and we have addressed their suggestions in our revised version. A point-by-point response to the reviewers’ comments is included below.

Reviewer 1: Alan Leonard

1a) Fig. 2A is out of date and should be updated to reflect current knowledge. It appears to be based on a figure in ref. 15 which also did not reflect the state of the field at the time of its publication. Only 3 high affinity DnaA boxes with sequence TTATCCACA (show this in the figure key) exist in E. coli. They are designated R1, R2, and R4 and should be noted as the only “red” boxes. R5(M) is a low affinity site. R3 is only identified by its sequence similarity to a perfect box, but it is a low affinity site and DnaA interactions with it have become less clear based on high resolution footprints. For a more accurate and recent view of the architecture of E. coli oriC, please look over the figure in reference 46 as well as the figures in two references that are not cited in the topic page: 1) Grimwade et al., 2018, Origin recognition is the predominant role for DnaA-ATP in initiation of chromosome replication, Nucleic Acids Research, 6140-6151 and 2) Sakiyama et al. 2017, Regulatory dynamics in the ternary DnaA complex for initiation of chromosomal replication in Escherichia coli, Nucleic Acids Research, 12354-12373. These references should be included in the bibliography.

We thank Dr. Leonard for this comment and have modified Figure 2A accordingly. We have also included the two references as suggested by the reviewer.

1b) The authors might also want to briefly address the finding in the first reference that the predominant role of DnaA-ATP in E. coli is in origin recognition (DnaA-ADP can replace the ATP form if allowed access to low affinity sites).

We have modified the main manuscript text and now include a statement that under certain conditions, ADP-DnaA can substitute for ATP-DnaA during initiation.

1c) It is not clear what the two different colors represent in the key for low affinity sites in Fig. 2A. tau and I sites appear to be distinctive, but it is not clear why unless the authors are just distinguishing tau from I. The naming of these sites was never very meaningful and it would be appropriate to treat them all as simply low affinity DnaA binding sites, most of which have a preference for DnaA-ATP.

The different colors for low-affinity sites were used to differentiate between tau and I-sites. We have now changed the color of the I-sites to that of the tau sites in Figs. 2A and 2C.

1d) The sites for the interaction of DNA bending proteins are also not noted in Fig. 2A although IHF is shown in 2C. Since this is an important feature of the loop back model, these binding sites might be shown in 2A and an appropriate reference to the dynamic interaction of IHF with E. coli oriC might be added (perhaps Ryan et al. 2004, Molecular Microbiology, 1347-1359).

We have added the IHF binding site to Figure 2A and also the reference to the figure legend.

2) The authors provide no other examples of bacterial replication origins other than E. coli. Since they mention bipartite versions and the fact that other arrangements of bacterial origins exist, it would help to show a few examples.

We have expanded Figure 2A to illustrate origin architectures in different bacterial organisms. Specifically, we included a schematic of Thermotoga maritima oriC and of the bipartite origin of Helicobacter pylori.

3) Figure 4 is done in an interesting way to designate the interaction of ORC protein domains with features within eukaryotic origins. However, the proteins represented by the purple blob on the left half of the metazoan origin (HMGA1a, LRWD1 etc) are described in the text as partner proteins that interact with ORC. I assume there is no designation of interaction (a colored line) because the exact subunit involved Is not known, but perhaps some way to show ORC interaction could be devised so that it was obvious from the figure that these are ORC interactive proteins as opposed to features where direct ORC interactions do not exist or are yet to be determined.

While some partner proteins clearly directly associate with ORC subunits, this behavior has not been firmly established for all proteins listed and interactions could be mediated by bridging proteins in some cases. We clarify this issue now in the main text and the legend of Figure 4.

4) I wonder if it is appropriate to include mention of the recent paper by Sima et al. (Cell, 2019, 176, 1-15, https://doi.org/10.1016/j.cell.2018.11.036) where cis-elements are identified on mammalian chromosomes that regulate the timing of replication origin activity during the cell cycle.

We thank the reviewer for this suggestions. The paper was not published at the time of our original submission but we have now included a brief discussion of Sima et al.’s findings in the fifth paragraph of the section describing eukaryotic origins of replication.

5) A minor issue; the title of some references are presented in all caps. Is this of special significance?

The references on Wikipedia are directly imported from Pubmed. Since some titles are capitalized within Pubmed, they are consequently formatted as all-caps in Wikepedia as well.

Reviewer 2: Ralf Wellinger

1) The authors mention that replisome has to cope with other DNA-dependent processes such as transcription, and that the collision between the replication and transcription machinery could lead to genome instability. In addition to reviews, it might be worth to include some original research papers on this topic including early work in E. coli or S. cerevisiae (doi.org/10.1016/0092-8674(88)90086-4; DOI: 10.1126/science.272.5264.1030).

We thank the reviewer for this suggestion and have included several references in the introduction and the last paragraph of eukaryotic replication origin section.

2) I am missing a comment on the identification and use of replication origins for shuttle vectors that allow the propagation of the same plasmids in different organisms. One example would be DOI: 10.1128/MCB.2.3.221.

We appreciate this request and have included a comment on the use of origins for shuttle vectors in the "Replicon Model" section.

3) The authors show epigenetic features involved in ORC recruitment and origin function. I would include original references on the observation that CpG islands in mammalian chromosomes function as replication origins (see work from Adrian Bird’s lab).

We apologize for this oversight and have added several references that reported correlations of CpG islands with replication origins.

4) The authors could include a comment on the possibility of non-canonical DNA replication events that bypass the need for replication initiation at DNA replication origins. These events have been described in E. coli, S. cerevisiae or H. volcanii (see the following review: https://doi.org/10.1016/j.tim.2017.12.001).

We agree with the referee that DNA replication in the absence of defined origins is an interesting topic and we had mentioned such events in H. volcanii, viruses and bacteriophages in our original submission in the “Concluding Remarks” section. We have now expanded on this topic to also include examples in bacteria and yeast.

Reviewer 3: Michael Botchan

1) Along the lines of history, I do think it would be worth a few more sentences as to how and why Jacob, Brenner and Cuzin purposed the "replicon model" for DNA replication. It would serve more than to give context as I believe it feeds into present situations about complexity with regard to eukaryote DNA replication and the dangers of oversimplification, without taking away the enormous credit that our pioneers deserve. Back in that day some plasmids where incompatible with each other- hence without selection one might overtake another and said to be incompatible, others not. Thus a positive replicator with a cis acting element shared by incompatible plasmids were postulated and compatibles had different "ori's" and a different replicator. Genetics provided a mechanism. Jacob and Monod thought that every control for transcription would be negative ( before evidence for positive pushed forward) and DNA replication might be positive. Work eventually uncovered for the E. coli chromosome (and plasmids) showed that both negative and positive regulatory pathways exist. I do appreciate the concise focussed discussion in the present piece.

We appreciate the referee’s suggestion and have elaborated more on the replicon hypothesis in an additional paragraph in the respective section.

2a) Figure 4 is great, especially the expanded view illustrating the various ways that different elements of ORC might recognize DNA. As complex and current this picture is there are still other layers of control suggested by the data in the literature that mediate DNA replication initiation. For example DNA replication mediated amplification of the Chorion protein coding genes are mediated by cis acting elements that do not bind ORC specifically but ORC itself is spread throughout a zone. Negative regulation provided by the Myb-MuvB complex keeps licensed Mcm's and cis acting ORC's from executing downstream replication and loss of Myb leads to no replication. The earliest work was published in Beall et al 2002, Beall et al 2004 and Lewis et al 2004. Myb is needed to activate and counter the repressive effects of E2F2. Further work form others shows that E2F1 is also required for positive replication perhaps interacting directly with Orc in very early but repressed stages of oocyte development! The point really is that in metazoans- especially when one is dealing with real systems not in a petri dish most DNA replication controls at the cis level are multilayered and staged developmentally.

We thank the reviewer for this detailed comment. We also think that Figure 4 is already quite complex and have therefore opted to include information about the chorion locus and its regulation in the main text. In addition, we have expanded on the discussion that origin features are variable and origin activity developmentally regulated.

2b) I point out that there is no genetic evidence to date that really resolves what the highly conserved BAH domain of ORC1 is doing although it does touch nucleosome in a clear way as indicated. Deletions of the domain in both budding and fission yeasts are viable as they may be so in metazoans as well.

We agree with the referee that the role of the Orc1-BAH domain in replication initiation in metazoans needs to be clarified further in future studies. We have modified the manuscript to reflect this view.

2c) The ABF1 binding site in ARS 1 is not found at other ARS elements yet the data is pretty good that it plays a role in efficiency for that ARS. Similarly ORC interactions with nucleosomes through the BAH domain probably will be substantiated with better genetics. So- I think some some indication as to the need for case by case examination of "Ori's" might be something to anticipate in the future. This long paragraph links to the second paragraph above.

We mention in the main text that ABF1 binding site is not found at all yeast origins and have now also expanded on our discussion on the variability of origin features. Please see also our response to point 2a.

3) The references are current and dealt with very clearly. I would suggest that this paper be referenced: Dynamics of DNA replication in a eukaryotic cell. Kelly T, Callegari AJ. Proc Natl Acad Sci U S A. 2019 Mar 12;116(11):4973-4982. doi: 10.1073/pnas.1818680116. Epub 2019 Feb 4. This recent work provides a compelling meta-analysis and then a computational model for replication origin site use in S.pombe. The requirements for A/T rich sites mediated by the ORC4 extension unique to this pombe ORC's is indicated obliquely in the review now submitted to PLOS and might be highlighted a bit more. One conclusion from this new paper from Tom Kelly is that I find important is that "there are many more potential initiation sites in the S. pombe genome than previously identified and that the distribution of these sites is primarily determined by two factors: the sequence preferences of the origin recognition complex (ORC), and the interference of transcription with the assembly or stability of prereplication complexes (pre-RCs)."

We appreciate this comments. The paper by Kelly and Callegari was not yet published at the time of our original submission. We have now updated the main text and incorporate these new published findings, as well as the reference.