Who Is Leading the Replication Fork, Pol ε or Pol δ?
2016; Elsevier BV; Volume: 61; Issue: 4 Linguagem: Inglês
10.1016/j.molcel.2016.01.017
ISSN1097-4164
AutoresPeter Burgers, Dmitry A. Gordenin, Thomas A. Kunkel,
Tópico(s)Genomics and Chromatin Dynamics
ResumoSeveral studies in the past decade support a model wherein DNA polymerase ε (Pol ε) carries out the majority of leading-strand DNA replication of the undamaged eukaryotic nuclear genome. Now a recent paper in Molecular Cell from the Prakash laboratory challenges this model, claiming instead that Pol δ is the major replicase for both strands and that Pol ε's primary role is only to proofread errors made by Pol δ during leading-strand replication (Johnson et al., 2015Johnson R.E. Klassen R. Prakash L. Prakash S. Mol. Cell. 2015; 59: 163-175Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar). While we fully subscribe to the idea that the replication fork is plastic and that its composition can adapt to various challenges, we believe the foundation for an unchallenged replication fork remains as established before the Prakash paper, for the following reasons. Our studies with mutants of the POL3 and POL2 genes encoding the catalytic subunits of Pol δ and Pol ε, respectively, revealed that they have two unique properties that lend themselves to studying strand-specific incorporation during replication. The first is that they are asymmetric mutators. As but one example, for the two mispairs that lead to AT→GC transitions, budding yeast Pol δ containing a L612M mutation in the polymerase active site misincorporates dGMP opposite template T much more frequently than dCMP opposite template A. When we eliminated error correction by deleting the mismatch repair gene MSH2 (Lujan et al., 2014Lujan S.A. Clausen A.R. Clark A.B. MacAlpine H.K. MacAlpine D.M. Malc E.P. Mieczkowski P.A. Burkholder A.B. Fargo D.C. Gordenin D.A. Kunkel T.A. Genome Res. 2014; 24: 1751-1764Crossref PubMed Scopus (104) Google Scholar, Nick McElhinny et al., 2008Nick McElhinny S.A. Gordenin D.A. Stith C.M. Burgers P.M. Kunkel T.A. Mol. Cell. 2008; 30: 137-144Abstract Full Text Full Text PDF PubMed Scopus (363) Google Scholar), the msh2Δ pol3-L612M strain exhibited a large synergistic increase in AT→GC transitions compared to the wild-type and msh2Δ strains, and relative to known replication origins, these mutations occurred in a pattern consistent with a model wherein Pol δ's primary role is in lagging-strand replication. Importantly, the synergistic increase was very strong, providing confidence that the large majority of the mutations in the msh2Δ pol3-L612M strain were indeed due to the Pol δ-L612M change. However, strong mutators rapidly accumulate suppressor mutations that result in wide variations in mutation rates in the mitotic progeny of double mutants. This was obvious in our study of the msh2Δ pol3-L612M mutant (see Figure S1 in Nick McElhinny et al., 2008Nick McElhinny S.A. Gordenin D.A. Stith C.M. Burgers P.M. Kunkel T.A. Mol. Cell. 2008; 30: 137-144Abstract Full Text Full Text PDF PubMed Scopus (363) Google Scholar), leading us to quantify mutation rates in cultures obtained by limited outgrowth of haploid spores germinated from meiotic progeny of the heterozygous diploids pol3-L612M/pol3-L612M MSH2/msh2Δ. This was not done in the Prakash study. Instead, they obtained double-mutant strains by two subsequent transformations involving several rounds of outgrowth from single cells, and obtained mutation rates that were only a few percent of the high rates we determined. Thus, whether from accumulating suppressor mutations during extensive propagation or for other reasons, their strains lack the high mutation rates we used to assign mutations specifically to Pol δ-L612M errors. In addition, their mutational spectrum for one orientation of URA3 in the S288c background shows strong GC→TA hotspots at base pairs 679 and 706 in URA3 that together constitute a third of the observed mutations. The authors suggest that those substitutions preferentially originated from template G-dAMP mismatches. However, those two hotspots are missing in a strain having URA3 in the opposite orientation. This would lead to the paradoxical suggestion that Pol δ-L612M does not replicate the lagging strain, yet neither they nor we imply that this is the case. Instead, our data suggest that the majority of GC→TA mutations in undamaged cells actually result from the complementary C-dTMP mismatch (Lujan et al., 2014Lujan S.A. Clausen A.R. Clark A.B. MacAlpine H.K. MacAlpine D.M. Malc E.P. Mieczkowski P.A. Burkholder A.B. Fargo D.C. Gordenin D.A. Kunkel T.A. Genome Res. 2014; 24: 1751-1764Crossref PubMed Scopus (104) Google Scholar, Nick McElhinny et al., 2008Nick McElhinny S.A. Gordenin D.A. Stith C.M. Burgers P.M. Kunkel T.A. Mol. Cell. 2008; 30: 137-144Abstract Full Text Full Text PDF PubMed Scopus (363) Google Scholar). Therefore, we conclude that those two mutation hotspots, and by extension other hotspots, support, rather than disprove, our favored model. A second useful feature of Pol δ and Pol ε variants is that they increase the incorporation of ribonucleotides into DNA. This occurs in vitro and in yeast in which newly incorporated ribonucleotides remain in the genomes of RNase H2-deficient (rnh201Δ) strains that are defective for ribonucleotide excision repair. Mapping data using next-generation sequencing shows increased ribonucleotide incorporation into the nascent lagging strand for pol3-L612M/G budding yeast variants and for an equivalent fission yeast variant, and increased ribonucleotide incorporation into the nascent leading strand of the pol2-M644G variant and its fission yeast equivalent (Clausen et al., 2015Clausen A.R. Lujan S.A. Burkholder A.B. Orebaugh C.D. Williams J.S. Clausen M.F. Malc E.P. Mieczkowski P.A. Fargo D.C. Smith D.J. Kunkel T.A. Nat. Struct. Mol. Biol. 2015; 22: 185-191Crossref PubMed Scopus (132) Google Scholar, Daigaku et al., 2015Daigaku Y. Keszthelyi A. Müller C.A. Miyabe I. Brooks T. Retkute R. Hubank M. Nieduszynski C.A. Carr A.M. Nat. Struct. Mol. Biol. 2015; 22: 192-198Crossref PubMed Scopus (106) Google Scholar, Koh et al., 2015Koh K.D. Balachander S. Hesselberth J.R. Storici F. Nat. Methods. 2015; 12 (253 p following 257): 251-257Crossref PubMed Scopus (86) Google Scholar, Reijns et al., 2015Reijns M.A. Kemp H. Ding J. de Procé S.M. Jackson A.P. Taylor M.S. Nature. 2015; 518: 502-506Crossref PubMed Scopus (157) Google Scholar). The straightforward explanation of these results is that Pol ε primarily replicates the leading strand and Pol δ primarily replicates the lagging strand. The Prakash study offers the alternative explanation that Pol ε does not incorporate ribonucleotides during replication, but rather only proofreads ribonucleotides incorporated by pol3-L612M, and does so only during leading-strand replication. However, Pol δ itself has little or no ability to proofread DNA termini containing ribonucleotides, but it can extend them, so there is no reason for Pol δ to dissociate to allow Pol ε to gain access to these termini. Moreover, while Pol ε can intrinsically proofread its own mistakes, current evidence suggests that it has little or no ability to extrinsically proofread mistakes made by Pol δ (Flood et al., 2015Flood C.L. Rodriguez G.P. Bao G. Shockley A.H. Kow Y.W. Crouse G.F. PLoS Genet. 2015; 11: e1005049Crossref PubMed Scopus (41) Google Scholar). Finally, if Pol ε was important only for proofreading ribonucleotides incorporated by Pol δ into the nascent leading strand, then in an rnh201Δ strain with wild-type polymerases, the ribonucleotide density in the nascent leading strand should be lower than in the nascent lagging strand, whereas it is actually higher (Clausen et al., 2015Clausen A.R. Lujan S.A. Burkholder A.B. Orebaugh C.D. Williams J.S. Clausen M.F. Malc E.P. Mieczkowski P.A. Fargo D.C. Smith D.J. Kunkel T.A. Nat. Struct. Mol. Biol. 2015; 22: 185-191Crossref PubMed Scopus (132) Google Scholar). These facts do not fit a model in which Pol δ is normally the primary leading-strand replicase. However, they do fit a model wherein Pol ε is the major leading-strand replicase, and this role is supported by elegant DNA replication studies in vitro (e.g., see Georgescu et al., 2014Georgescu R.E. Langston L. Yao N.Y. Yurieva O. Zhang D. Finkelstein J. Agarwal T. O'Donnell M.E. Nat. Struct. Mol. Biol. 2014; 21: 664-670Crossref PubMed Scopus (141) Google Scholar). As discussed (Kunkel and Burgers, 2008Kunkel T.A. Burgers P.M. Trends Cell Biol. 2008; 18: 521-527Abstract Full Text Full Text PDF PubMed Scopus (209) Google Scholar), our favored model wherein Pol ε is the primary leading-strand replicase does not exclude an important role for Pol δ in leading-strand replication in certain regions of the undamaged genome and/or when the genome is under stress. A Major Role of DNA Polymerase δ in Replication of Both the Leading and Lagging DNA StrandsJohnson et al.Molecular CellJuly 2, 2015In BriefThe current model of eukaryotic DNA replication suggests that DNA polymerase (Pol)ε primarily replicates the leading strand while Polδ replicates the lagging strand. Johnson et al. provide genetic evidence that Polδ replicates both strands, while Polε's proofreading activity is important for removing Polδ-generated errors from the leading strand. Full-Text PDF Open ArchiveResponse to Burgers et al.Johnson et al.Molecular CellFebruary 18, 2016In BriefIn our study (Johnson et al., 2015), we concluded that DNA polymerase (Pol) δ replicates both the leading and lagging DNA strands and that Polε plays no significant role in leading-strand replication. In their Letter in this issue of Molecular Cell, Burgers et al. (2016) contend that their model wherein Polε primarily replicates the leading strand still remains valid and suggest that (1) our strains contain suppressors, (2) our observed G→T mutations originate in the lagging strand, and (3) ribonucleotide incorporation data support their model. Full-Text PDF Open Archive
Referência(s)