DNA Replication in Time and Space: The Archaeal Dimension
Andrei Kuzminov
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript written by Serdyuk and Allers describes the history of research　　　　　　　into the elucidation of the mechanisms of DNA replication from the dawn of molecular biology to the present day. It is well written and is a good reference for general readers of the journal ”DNA” to study DNA replication from a broad perspective. I think there are no major revisions to be made, but would like to point out some minor corrections to improve the manuscript.
1. Oric is first appeared in line 295. Then, "OriC" (Origin of Chromosomal Replication) is appeared in line 313 after "OriC" was used several times before that. In addition, "OriC" is used many times in this manuscript. However, "OriC" is the gene, but not Protein, and therefore, "oriC" (italic) is appropriate rather than "OriC".
2. Line 500. "In fact, the first archaeal gene encoding an initiator protein – a distinct sequence located downstream of the replication origin (OriC) [59,60]–was first identified in the hyperthermophile Pyrococcus genome. Due to it sequence homology to regions of the eukaryotic Orc1 and Cdc6 – it was subsequently named Cdc/Orc1" This description is not precisely correct. The archaeal gene encoding an eukaryotic initiator(Orc1/Cdc6/Cdc18)-like sequence was reported in 1987 (Uemori T, Sato Y, Kato I, Doi H, and Ishino Y. (1997) A novel DNA polymerase in the hyperthermophilic archaeon, Pyrococcus furiosus. Gene cloning, expression, and characterization. Genes Cells 2, 499-512.) This paper clearly predicted that "The presence of cdc18þ/CDC6 family homologues upstream of the pol genes and their co-transcription with the Pol II genes, indicates that the products of these genes may all work together in the initiation of DNA replication. For instance, it may be possible that the ORF1 protein binds to an origin of replication and—through a molecular interaction—attracts DNA polymerase II to the origin for the initiation of DNA synthesis. Please revise this part appropriately.
3. Line 593. (DnaG in bacteria, PriSL in eukaryotes, and its homologue PriXLS in archaea). Primase in the most of archaea is PriSL, and PriXLS is present in Sulfolobus, for example, but is exceptional.
4. Line 594. "Ligase" should be "DNA ligase"
5. Line 614. "through rolling circle replication (i.e, mode of RDR), where a simple nick generated by RCR (Rolling Circle Replication)" should be revised.
6. Line 642. "R-loops possess an advantage over D-loops due to their ability to serve as direct primers." Please add more explanation for the general readers to understand this statement.
7. Line 646. "DBSs" should be "DSBs". Also, this is the first time to use this term here. "DBS (Double Strand Break)" is appeared later in Line 704, but DBS should be DSB.
8. Line 659. "recA" should be "RecA".
9. Line 663. What is "RecA119"?
10. Line 698. Why pol-δ has hyphen?
11. Line 760. radA recombinase should be RadA recombinase.
12. Line 710. MCM (Minichromosome Maintenance Complex) is appeared after appearance of MCM many times.
13. Please check the used terms and their abbreviations in the whole manuscript again.
14. Fig. 1. The text in each Figure is too small to read. For example, I can't read the text next to the photo at the top in Fig. 1. Fig. 2 is also difficult to read the texts. a, b, and c of the nulceotide should be α,β, and γ.
15. It is not easy for the readers to understand what is different between the left and right panels in Figs. 7 and 8. Please revise to more reader-friendly figures. For example, how to process from the second to the third step in the right panels. In addition, "rad52" should be "Rad52". Fig. 8 is also important, but it does not show the essence of the difference between iSDR and cSDR. Please put more effort into your drawing.
16. I think the recent progress of our understandings in the structure and functions of replisome complex from the archaeal studies are missing in this review. For example, functional interactions of PolD-Primase, Primase-PolD-CMG helicase are very interesting with comparison with bacterial and eukaryotic replisomes. However, it takes time and space to describes this issue precisely, and I understand that this subject will be written somewhere in the other review papers.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors
This is an unusual review — both in its breadth and in its arguments. I definitely had fun checking the references. Although I found myself frequently disagreeing with what the authors say, I think it is a valuable addition to the overall thinking about replication initiation. Its archaeal perspective makes it unique.
Specific suggestions/comments
Page 2, lines 70-72 — It may sound surprising these days that the Hershey-Chase experiment was so readily accepted, despite so much of the phage protein (probably tail fibers) staying with the infected cells. The easy acceptance was likely due to so much other evidence already pointing out in this direction. For example, one could mention the cell biology research of the same Mirsky (your ref. #5) — PMID: 1812635; PMID: 14808154; — or the works on DNA stability in vivo (please find references yourself).
Page 2, lines 73-77 — You could leave this as it is, since this is common knowledge anyway, but FYI consider the following:
— Since the transformation results of Hotchkiss and the Hershey-Chase experiment, “everybody” knew that the secret of genetic heredity is in the DNA structure (BTW, “tertiary DNA structure” sounds weird).
— “The available crystallographic data” of Franklin [11] indicated a double-helix — so do not give credit to W+C for it. They acted like petty thieves there, and everybody knows it now, since pompous ass Watson boasted about it himself in his book “The Double Helix”.
— “The biochemical data [13]” is the infamous triple helix model of Pauling. There is nothing biochemical about it — just a pure fantasy (which makes no sense, BTW). If you really need to cite pre-1953 biochemistry that provided insights into the DNA structure, give the 1947 pH-titration data of Gulland et al. (https://scholar.google.com/scholar_lookup?title=Deoxypentose%20nucleic%20acids.%20Part%20II.%20Electrometric%20titration%20of%20the%20acidic%20and%20the%20basic%20groups%20of%20the%20deoxypentose%20nucleic%20acid%20of%20calf%20thymus&publication_year=1947&author=J.M.%20Gulland) or DNA solution viscosity measurements of the same lab, same year (https://pubs.rsc.org/en/content/articlelanding/1947/jr/jr9470001141). (Sorry for the long URLs — these gems are not available on PubMed.)
Page 2, lines 82-87 — Bloch seems to be the first to propose how the conservative outcome of DNA replication could be achieved, but the test-stimulating comparison of the three replication models was actually formulated by Cyrus Levinthal in 1956 (PMID: 16589875) (and his pictures are available on the Internet) exactly in response to Max Delbrück concern about DNA unwinding. In fact, your Figure 1 is a version of Levinthal’s figure. And you do not have to add Gunther (Stent) to the dispersive model — it was Max’s own invention (PMID: 16589559) — pure genius (as always) and wrong by definition (Max failed to propose a single correct idea).
Page 2, lines 87-88 — The Meselson+Stahl (1958) experiment was a breakthrough, but please also mention the 1957 experiment of Taylor et al. (PMID: 16589984) that reached the same conclusion for higher eukaryotes a year earlier.
Page 4, the top half of the page — several problems should be fixed.
— First, define enzymatic DNA synthesis reaction as always performed at a particular partially-duplex DNA structure made of the template strand and a complementary to it, but shorter PRIMER strand, which can be extended along the single-strand region of the template strand. The statements like “if the template substrate served as a simple primer” will only confuse the unprepared reader.
— Second, inform the reader that in vitro nucleic acid synthesis started with a blunder by Severo Ochoa (got the Nobel for it anyway, together with Kornberg) — exactly because he did not insist on templated synthesis!
— Third, do not make it sounds as if accuracy of templated NA synthesis is mostly due to proofreading (lines 144-146). RNA polymerases are pretty accurate copiers, and none of them proof-reads. You correctly recognize “polymerase selectivity” as the main factor, but this happens later (page 6, line 225).
— Forth: but, importantly, RNA polymerases initiate synthesis without a primer — because they have a site for internal one-nucleotide primer — which in DNA polymerases is occupied by the proofreading site. This is why DNA polymerases require a primer to initiate, while RNA polymerases cannot proofread.
Page 4, lines 154-157 — several problems with this statement.
— First, the title of the corresponding citation (#22) is incomplete. Please add “XXXVI. A PROOFREADING FUNCTION FOR THE 3’ ïƒ 5’ EXONUCLEASE ACTIVITY IN DEOXYRIBO-
NUCLEIC ACID POLYMERASES*”
— Second, as the number shows, this is paper #36 from Kornberg’s lab about DNA pol I — not “their second manuscript” (LOL).
— Third, the problem was not as much the proofreading exonuclease (3’ïƒ 5’), as the more powerful 5’ïƒ 3’ (forward) exonuclease of this repair DNA polymerase. Separation by proteolysis really helped (PMID: 4982877).
Page 10, lines 374-382 — a single archaeal example of robust chromosome replication without origins should not be taken as a general rule that can question the role of origins in initiation of DNA replication. Everybody knows from daily experience with most organisms that a DNA molecule without a single replication origin fails to replicate. A balanced discussion (for example, PMID: 24644021) is required.
Page 12, the central paragraph (lines 446-466) — a couple of things here.
— First, some bacterial genomes are up to 15 Mbp, and still replicate from a single origin, so the size difference cannot be the only explanation of multiple origins in the eukaryotic genomes. A typical explanation is the speed of replication forks: in the naked bacterial DNA replication forks are ~10 times faster than in the chromatinized DNA of eukaryotes.
— Second, the argument in the middle (about instability at the origins) countered by relaxation of initiation specificity is not a logical one: it should be the other way around. Indeed, the paragraph ends with statements about “tight control mechanisms” that seem to contradict what was discussed in the middle.
Page 17, lines 602-604 — “template independent RNA” polymerization is incompatible with RNA being “genetic material”. The essence of any genetic material is template-dependent replication/transcription!
Page 17, lines 612-614 — T4 is not a “simple replicator”; T4 does not practice rolling-circle replication (RCR), and RCR is not a mode of RDR!
Pages 19-20, the BIR section — it is important to mention that BIR is the replication happening in G2, outside of S-phase.
Minor points
Page 2, line 56 — reference #3 is irrelevant to the transforming principle.
Page 2, line 70 — Levene’s tetranucleotide idea and its subsequent versions, leading to suggestion of the structural role of DNA in the nucleus, needs to be introduced early and spelled out.
Page 2, line 72, “The same year” — actually, the next year.
Page 3, lines 97-98 — no, in the conservative mechanism, the original duplex is preserved. ‘Both parental strands are conserved in the semi-conservative replication as well — they just paired differently.
Page 3, lines 101-102 — the reference for the Phage group results is needed (for example, PMID: 16589732).
Page 3, lines 104-105 — not Hershey+Chase labeling, which was radioactive. In contrast, M+S used density label and CsCl gradient density centrifugation to separate “heavy” DNA from “light” DNA. No radioactivity was involved.
Page 4, line 147 — DNA pol I (Kornberg enzyme) is a repair polymerase, not a replicative one.
Page 5, lines 191-193 — non-sequitur about transition to DNA
Page 5, lines 194-195 — how could the 5’-triphosphate of NTPs impose any polarity on NA synthesis? Check Kornberg’s book on directionality of polymer biosynthesis — the precursors can be activated on either side, it does not matter! And why this would be an artefact of the RNA world?
The conventional answer to how the 5’-triphosphates may favor the 5’ïƒ 3’ synthesis is given later (page 6, lines 219-222), but it is still cryptic. Just say that if with 5’-triphosphates, the NA synthesis would be going 3’ïƒ 5’, then in case of proofreading, an additional pyrophosphate-recharging step would be required, to reactivate the 5’-end for synthesis.
Page 5, lines 198-200 — this sentence only applies to self-replicating ribozymes. No such (RNA-polymerizing) enzymes are known.
Page 5, line 226 — I liked this 102-103-fold. To avoid mis-formatting, try 10e2-10e3.
Page 7, lines 269-270 — why the plasmids would be in contrast to the replicon idea, outlines in the preceding sentence?
Page 7, lines 270-273 — the replicon concept does not directly explain plasmid incompatibility — otherwise multiple origins of replication would not be able to fire at the same time in the same cell. The incompatibility arises because plasmid DNA molecules carrying the same replicon segregate from each other.
Page 8, line 312-314 — The stressful growth conditions are not due to the block to protein synthesis — which would simply block growth, making it non-growth conditions — but due to DNA damage (for example, via thymine-starvation in this case). Also, “SDR” (stable DNA replication) needs to be explained at first encounter. Also, citation is needed, for example, Kogoma’s last review PMID: 9184011.
Page 8, lines 330-333 — over-replication in eukaryotes is prevented by a single origin licensing even in G1, rather than by differential timing of origin firing in S-phase. The authors themselves describe this on page 12, lines 471-475.
Page 9, line 336 — First, Table 1 is not a table, but what is usually described as a “box” — which contains a specialized text, like useful terms or concepts, that are not integral part of the main text.
— Second, it is unclear how these terms could be misinterpreted (remove the explanation in the footnote).
— Third, mini-chromosome is not a synonym of mega-plasmid (it can be quite small) — rather, it is simply a plasmid driven by oirC (the chromosomal replication origin).
Page 9, line 344 — “10% of bacterial genome differ from E. coli” sounds funny. Just say, 10% of bacterial genomes reside on more than one chromosome. There are very few split genomes that have “numerous” replicons.
Page 10, line 358 — the MFA abbreviation should be defined here, rather than on page 22.
Page 10, line 395 — LUCA abbreviation should be defined here, rather than on page 17.
Page 11, lines 417-420 — this statement needs citations.
Page 11, lines 427-429 — before mentioning in passing DnaA-ATP and -ADP complexes, these need to be introduced. The same goes for “domain III” — without the short description of DnaA structure, this mentioning makes no sense.
Page 12, lines 435-436 — It is not the DnaA polymerization on the origin that opens DUE through increased supercoiling, but the combination of sharp bending by IHF and the secondary ssDNA-binding site on DnaA, that opens the ssDNA bubble and stabilizes it (PMID: 29312202).
Page 13, line 498 — the “multiple replication origins on most chromosomes” statement needs references. As far as I know, some archaeal lineages prefer to have a single origin, like in bacteria, but some others, like haloarchaea, always have several. We need examples here. You can start with PMID: 24808892.
Page 17, lines 583-587 — in fact, Okazaki consistently showed, throughout his (unfortunately) brief career, that in vivo both strands are synthesized in pieces. The semi-discontinuous replication was observed only in vitro, with purified enzymes (and precursors). The conundrum was resolved only recently, when the apparently discontinuous synthesis on the leading strand in vivo was shown to be due to ribonucleotide incorporation and excision (PMID: 30617079).
Page 17, lines 592-594 — RNase H is not required for DNA replication in either prokaryotes or eukaryotes — although the enzyme may participate in removing RNA primers (not primases!). Typically, the primers are removed by DNA polymerase I (in bacteria) or by flap-endonucleases in eukaryotes.
Page 17, line 597 — what is this: “non-living viruses”? Vegetative viruses within cells are a form of life.
Page 17, line 622 — a reference is needed there (for example, PMID: 6305581)
Page 18, line 623 — “the cell cycle of T4 bacteriophage” sounds weird. Try ‘the life cycle”?
Page 18, line 629 — say “the very 3’-end of the lagging strand template cannot be replicated”
Page 18, line 635 — which break? (to repair) There was no break introduced above.
Page 18, line 653 — RDR does not restrain — in fact, it relieves any restrains. Better say: “RDR maximizes phage replication”
Page 18, line 663 — what is RecA119?
Page 18, lines 674-675 (also, page 23, lines 828-830) — what are you talking about here? This was all figured out in bacteria long time ago (as your references 124 and 125 would suggest). Please read PMID: 29898897. And eukaryotes do not restart replication forks after DSB repair, because they do not need to, saved by their multiple replication bubbles.
Page 20, lines 725-726 — the oriC/DnaA-initiation is indeed efficient, but it is preferred for a different reason — it can be controlled!
Page 20, lines 728-729 — SOS and iSDR have not been introduced. And low-nutrient environments do not induce iSDR! Inducible stable DNA replication is strictly induced by DNA damage and is your RDR! Again, read and CITE Kogoma here (PMID: 9184011).
Page 21, lines 747-748 — First, “DnaA-dependent initiation”. Second, what “energy-saving advantage of SDR”? Initiation does not take much energy compared with the replication of the whole chromosome. Stable DNA replication tends to overreplicate.
Page 24, references #18, #19 and #22 — these are incomplete. Please complete the titles:
I. PREPARATION OF SUBSTRATES AND PARTIAL PURIFICATION OF AN ENZYME FROM ESCHERICHIA COLI
II. GENERAL PROPERTIES OF THE REACTION
36. A proofreading function for the 3' leads to 5' exonuclease activity in deoxyribonucleic acid polymerases
Page 24, reference #44 has four more authors (not listed)
Miscellaneous
Page 1, line 8 — insert “DNA” before “molecule”. Not all molecules can self-replicate.
Page 1, line 17 — replace “every domain” with “all cellular kingdoms”
Page 1, line 21 — add “an” before “alternative”
Page 1, line 36 — replace “to be distributed equally, and remaining identical to the original copy” with “for its two identical copies to be segregated”
Page 1, line 40 — replace “in the form of” with “resulting in”
Page 3, line 94 — add “biology” after “molecular”
Page 3, line 97 — delete “conserved”, since confusing
Page 5, line 176 — replace “replication” with “DNA synthesis”
Page 5, line 229 — insert “replication” after “at the same time”
Page 7, lines 300 and 301 — “9 bp” DnaA box and “250 bp” oriC.
Page 10, line 384 — insert “in prokaryotes” after “multireplicon genomes”, because eukaryotes all have multiple chromosomes.
Page 10, line 398 — insert “the reader” after “familiarize”
Page 13, line 487 — “checkpoint control point” sounds redundant
Page 14, line 554 — insert “Methanothermobacter” after “from”
Page 17, line 580 — replace “photographic micrograph” with “autoradiograph”
Page 17, lines 591 and 592 — move “RNA” from before “primase” on 591 to before “primers” on 592
Page 18, line 647 — “UvsX has”
Page 18, line 652 — replace “Syeda” with “McGlynn and colleagues”
Page 18, line 659 — say “bacterial RecA, which belongs to”
Page 23, line 828 — secondary initiation mechanisms?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsI appreciate your detailed responses to my previous comments/suggestions.
Please do not spend your time on another detailed response to the new comments: just think about them and act on them, — and I will be happy to know that you did — OK?
And please do not add new material to your review — it is already bursting at seams!
Specific suggestions/comments
Page 2, lines 70-73, “As a response, Chargaff proposed <…> overthrowing Levene’s hypothesis.” — First, the Chargaff-1950 paper could not be “in response” to the 1915 paper by Levene. Second, Chargaff “experimentally established” the ratios, rather than “proposed” them. And third, Levene’s idea was not “overthrown” — instead, its simple version was incompatible with the new data. But one could easily “complicate it a bit” to make it compatible with Chargaff’s findings. Levene was simply not around any more…
Page 2, line 77, “[10]. This assertion was further corroborated…” — Actually, ref #11 was published in 1950, while #10 — in 1951. Thus, #11 could not corroborate #10.
Page 2, lines 95-97, the last sentence spilling on page 3 — it looks like the reference there should be Levinthal-1956 (your ref #21), rather than Bloch-1955 (your ref #17)?
Page 4, line 127 — Kornberg polymerase started being called DNA pol I only much later, when DNA pol II and III were reported in 1971 (PMID: 4927672; PMID: 4943556).
Page 4, line 137 — replace “faithful” with “complementary”. This is important, as “faithful” usually means “identical”. The two daughter DNA duplexes are indeed faithful copies of the mother duplex, but Kornberg used ssDNA in the early experiments, — and the synthesized “primer” strand is not identical to the template strand, but is complementary to it (as in your red text on lines 141-143).
This confusion continues (for example, lines 155) — so you may want to clarify the distinction between the duplex replication (exact, faithful) and primer strand synthesis over the template strand (complementary). “Exact” has a double meaning here, unfortunately…
Page 4, lines 166-170 — the insert in red should be simplified to avoid confusion. First, the 3’ïƒ 5’ exonuclease is the proofreading activity that acts EXACTLY by excising the mispaired nucleotide at the 3’-end of the primer; moreover, in the presence of all four dNTPs, this “backward” nuclease is suppressed on completely paired template-primer pair. That is, this DNA degradation activity was not observed in DNA synthesis conditions and could not be their “problem”. Their problem was the 5’ïƒ 3’ “forward” nuclease, which is active no matter what, including during DNA synthesis — and you eventually also say it “was indeed the culprit”.
I suggest you mention the forward nuclease first (as the culprit for the problem), and then, after saying it was eventually removed by proteolysis (as introduced in your ref #27), explain that this trick allowed to reveal the second, proofreading nuclease.
Page 4, lines 170-173 “When the second replicative enzyme – called DNA polymerase III (Pol III)…” — in fact, DNA pol I is a repair polymerase, not a replicative one — as became clear when De Lucia and Carins isolated the mutant in it that was growing like WT (PMID: 4902142). Also, “second enzyme called pol III” sounds weird. Better say, “When the true replicative enzyme, DNA pol III, …” — and then restructure the remaining sentence accordingly.
“DNA (i.e., replicative) polymerases” — First, not all DNA polymerases are replicative enzymes (as you yourself explain on page 5); DNA pol I and II are good examples of repair polymerases which can still proof-read. There are also TLS polymerases that do not even proofread. Second, there are replicative RNA polymerases (usually found in RNA viruses) — and some of them likely proofread, because some RNA viruses are known for their very low mutation rates (like coronaviruses, for example). Perhaps you should just say “proofreading-capable polymerases”?
Page 5, lines 188-189, “Ochoa demonstrated in vitro RNA synthesis by RNA polymerase to be template independent.” — LOL, please do not say that! Ochoa reported that the known RNase, called “polynucleotide phosphorylase” (PNPase) — one of the strongest in bacteria, by the way, — could, under certain non-physiological conditions, “polymerize” NDPs into random polymers. Not exactly the case to award the Nobel Prize, — but whatever…
Since it was both his blunder (not to insist on templated synthesis), as well as the blunder of the Nobel Prize committee, — it is decided, to save all the important faces, not to advertise this as a blunder…
Page 6, Fig. 2, the kinetics scheme — Pol-DNA complex should be without dNTP in the middle of the top row.
Page 8, the bottom paragraph, lines 347 and 351 — what “clusters of origins” do you refer to? “Cluster” means a “group”, and eukaryotic origins are not grouped along the chromosome — they are actually spaced quite evenly. I suggest to drop “cluster” and use “multiple origins” instead.
Page 9, Figure 3 legend — in describing the eukaryotic initiation model, it is important to mention that the initiation signal itself is generated by the cell cycle machinery, by cyclin-dependent kinases (CDKs) — as you yourself explain later on.
Page 10, the glossary.
— Check the mini-chromosome definition. I suggest a simple “a plasmid driven by the oriC/DnaA replicon of the chromosome.”
— “Extrachromosomal element” in bacteria IS a synonym of plasmid (by the original plasmid definition — PMID: 13003535). By the way, extrachromosomal elements cannot replicate OUTSIDE the host cell, — in fact, nothing replicates outside cells (in nature).
— Replicon is normally defined as “a nucleic acid molecule, or part of one, which replicates as a unit, beginning at a specific site within it.” (here I agree with Google). Therefore, I suggest instead of discouraging its use in Archaea, point out that archaeal chromosomes frequently have several replicons. Especially since this is how you use it yourself later (for example, page 11, line 402 or page 14, line 541).
— Replication fork — I suggest deleting “for replication to proceed bidirectionally from the replication bubble” as confusing.
Page 10, lines 377-379 — please check it, but, if I remember it right, 10% of bacteria have more than one chromosome (that is the DNA molecule carrying essential genes = cannot be cured of). The total number does not have to be two; for example, Deinococcus radiodurans has four DNA molecules in its genome equivalent: two chromosomes (DR_Main [2.65 Mbp] and DR412 [412 kbp]), one megaplasmid (DR177 [177 kbp]), and one plasmid (46 kbp).
By the way, “chromids” are different from true chromosomes in that they do not carry essential genes (=can be cured, although the host may be unhappy about it). At the same time, chromids are different from plasmids in that they do carry house-keeping genes or niche-specific genes.
Page 10, lines 380-381 “Advances in genome sequencing during the 1970s have led to the identification of <…> Archaea” — I had to LOL at this one! Actually, the two competing DNA sequencing methods became available in the 1977-1980 time frame, and I remember what a bombshell it was when pBR322 plasmid sequence was reported in 1979! The first bacterial genome sequence would be reported in some 15 more years (1995).
Woese made his discovery back in 1977, not by sequencing, but by 16S RNA “fingerprinting” (PMID: 24296570) — also a heroic effort in its way. By the way, this erroneous claim is repeated, in different words, on page 25 (see below).
Page 10, line 385 — ref #64 cannot be the right one. Try PMID: 8688087.
Page 10, line 387, “codon (GGTC) skew analysis” — first, codons have three nucleotides, not four; second, it was “cumulative oligomer skew” analysis.
Page 11, line 418 — RDR pops here without a definition — which comes much later, on page 19. Please define it at the first appearance.
Page 11, line 436 — “regular GC skew profile”
Page 11, lines 436-438 — citations are needed (PMID: 9733700; PMID: 32411117 - ?). Also, what is this: “GC profiles have shown asynchronous replication initiation from multiple chromosomal sites”??? Did you mean that most Synechocystis have irregular GC skew profiles in their chromosomes?
Page 12, lines 452-457 — the argument in this paragraph is valid, but appears incomplete. Origin acquisition onto chromosomes by HGT implies that origins were initially developed on DNA molecules other than chromosomes. This needs to be spelled out, and also clarified (or speculated) why such origins were not initially needed on the chromosomes, but were critical for the extrachromosomal elements.
Page 12, lines 470-471 — LUCA is suddenly introduced here, but without the abbreviation (???). And the abbreviation itself is used only once in the whole text, on page 19 — (???).
Page 13, lines 480-483 — so, four stages or five (V)?
Page 13, line 501 — replace “Per single” with “In E. coli”. What follows is the description of E. coli oriC. Origins in other bacteria will have varying numbers of DnaA boxes, and various arrangements of these, too.
Page 17, Table 1 — at the DNA duplex melting step in eukaryotes and archaea the MCM helicase is listed, instead of ORC/Cdc6. I understand it is still a matter of controversy, and MCM does have a potential for origin melting. However, theoretically, it should be able to do it only as a dimer of two back-to-back hexamers (like in your Fig. 5A). So, consider modifying your table entries accordingly.
Page 19, lines 713-717 — I am stressing (again!) that Okazaki only observed fully discontinuous synthesis, even in his WT cells (when he used the shortest pulse times, like 2 or 7 seconds (and, by the way, he only worked in vivo). After a few more seconds he would observe the small fragment “maturation” into longer species and, within 60 seconds into the full-length DNA strands. In the ligase-deficient conditions, this maturation was blocked, and replication intermediates stayed low-molecular weight.
The semi-discontinuous replication was only observed in vitro, with reconstituted replication forks (PMID: 2838481; PMID: 2545703; PMID: 1740451). This happened much later, after Okazaki was long gone.
Page 19, lines 725-728 — you forgot to mention a DNA repair polymerase or flap-endonuclease, to remove the RNA primers from the 5’-ends of Okazaki fragments. DNA ligases are sterically prohibited to link 3’-DNA with 5’-RNA ends.
Page 20, lines 777-781 — NO, there is no advantage of R-loop over D-loop for priming: DNA polymerases extend RNA primers and DNA primers equally well. Please remove these two misleading sentences.
Page 21, Fig. 6 — in “B” you cleave the DNA junction, but “C” still shows no cleavage. If you do not know how to resolve these intermediates, I suggest you do not show the cleavage (I am sure it is not critical for RDR).
Page 22, Fig. 7 (as well as its legend on page 23) — The “A” part is correct, while the “B” part is not. The “B” part will be correct by itself, if it is disconnected from the one-ended DSB at the top, because, while the left half of the diagram shows a replication fork reassembly (one-ended break repair), the right part explains the RNA-assisted repair of a frank (two-ended) double-strand break. I see that your confusion is coming from ref #163, which does not have a scheme like this, but has a couple of schemes that you tried to put together under “one-ended DSB repair”. I suggest you remove the text on lines 853-858, as well as Fig. 7B as having no relation to DNA replication.
Page 24, Fig. 8 (the two SDR pathways) — at least three problems here (not your fault, though, because the SDR ideas are confusing — courtesy of the SDR researchers).
— First, the basic replication (re) initiation mechanisms should be similar for both iSDR and cSDR: strand invasion (D-loop, R-loop), followed by PriA-loading of the replisome. Therefore, the two schemes should look more similar (and simpler, too) — while yours look quite different. The main difference is that D-loop is catalyzed by the RecABCD-promoted strand invasion, while the R-loop forms spontaneously, during transcription.
— Second, Kogoma’s way of thinking about iSDR did not clearly explain where the additional initiation potential would come from. Everybody understood that, after DNA damage, replication forks disintegrate and need to be restored — but, in itself, this does not explain additional replication potential. It was pointed out that, when existing forks are inactivated by DNA damage, there is overinitaition from the origin, which, in principle, could explain why new initiations could be blocked a couple of hours after DNA damage, yet replication still continues (PMID: 7661854). However, Kogoma instead speculated about “double-strand breaks at oriC”, without explaining their possible nature or providing any evidence that they actually happen…
— Third, lines 918-921 claim that R-loops are catalyzed by RecA in vivo, which cannot be true. If they were, they would be resistant to RNase HI, and they are not. In fact, R-loops are more stable than the corresponding DNA duplexes and form spontaneously in the negatively-supercoiled DNA. Sure, one can find in vitro conditions for RecA to promote their formation even further — and Kowalczykowski lab and Kogoma lab did — but this does not mean that this happens in vivo. Not everything demonstrated in vitro transpires into bona fide in vivo mechanisms — remember Ochoa’s blunder with the major bacterial RNase — which he thought was the RNA polymerase!
Page 25, lines 973-977 — again, multiple problems here.
— First, the reference for Sanger sequencing [173] is incorrect — it is supposed to be 1977, not 1965! The correct one is PMID: 271968.
— Second, Woese could not use Sanger’s DNA sequencing method to “sequence” 16S rRNA! Especially if the method was just published (in 1977), while Woese was already reporting on archaebacteria in the same year (1977).
— Third, instead, Woese used the rRNA “fingerprinting” method of Sanger, for which ref #173 is correct.
— And fourth — the reference for Woese and Fox paper (#174) is incorrect. The correct reference is your #63.
Miscellaneous
Page 2, line 51 — insert “Biology” between “Molecular” and “Revolution”
Page 2, lines 55-56 — say “made of DNA, rather than proteins”
Page 2, line 89 — insert “that” after “Assuming”
Page 4, line 155 — insert “synthesis” after “DNA”
Page 5, line 200 — replace “an” with “the”
Page 7, line 294 — replace “element” with “enzyme”?
Page 11, line 415 — replace “rendered strains that are capable” with “resulted in strains that are still capable”
Page 11, line 436 — insert “skew” after “GC”
Page 13, line 495 — insert “typically” before “encoded”. For example, in E. coli, it is not.
Page 15, line 611 — delete the first “to”
Page 17, line 680 — “nonessential”
Page 17, lines 689-690 — replace “proceeds in a 3’ to 5’ direction to translocate the leading strand through” with “translocates in the 3’ïƒ 5’ direction along the leading strand template through”
Page 19, line 710 — insert ”chromosome in the act of” after “E. coli”
Page 20, line 776 — replace “due to aborted” with “during”
Page 20, line 799 — replace “also” with “that”
Page 23, line 903 — insert “genome distribution of” after “characterize”. DRIP results do not tell anything about the nature of the DNA:RNA hybrids — only their distribution along the chromosome.
Page 26, line 1005 — delete “with”
Page 27, ref #10 needs year (195
Author Response
Thank you again for your helpful suggestions. Please see the Word file for our responses.
Author Response File: 
Author Response.docx
