To pass on genetic information from one generation to the next, the DNA in a cell must be precisely copied. DNA is made of two strands and genetic information is encoded by sequences of molecules called bases in the strands. The bases from one strand form pairs with complementary bases on the other strand. However, errors in the copying process result in unmatched pairs of bases. Such errors are corrected by a repair system called mismatch repair.

When DNA is copied, the two strands are separated and used as templates to make new complementary strands. This means that errors only arise on the new strands. Mismatch repair must therefore target the new strands to maintain the original information encoded by the template DNA. The repair needs to happen before the copying process is complete because the template strands and the new strands become indistinguishable afterwards. However, it is not clear how the two processes communicate with each other.

Previous studies have identified a ring-shaped molecule called the replication clamp – which is essential for the copying process – as a prime candidate for the molecule responsible for this communication. This molecule binds to the DNA to promote the copying process, and afterwards it is removed from the DNA by other molecules. Furthermore, a group of proteins called the MutSα complex, which recognizes unmatched bases in DNA molecules, physically interacts with the replication clamp.

Kawasoe et al. used eggs from African clawed frogs to study how the replication clamp connects the copying process and mismatch repair in more detail. The experiments show that when the replication clamp is bound to the DNA, it is able to direct mismatch repair to a specific DNA strand. When MutSα recognizes unmatched bases, it prevents the replication clamp from being removed from the DNA. By doing so, MutSα prevents the information about the new DNA strand from being lost until mismatch repair has taken place.

These findings reveal new interactions between DNA copying and the correction of errors by mismatch repair. The next steps will be to understand how MutSα is able to keep the replication clamp on the DNA and to clarify its role in protecting DNA from gaining mutations.

The evolutionarily conserved mismatch repair (MMR) system corrects replication errors post-replicatively to prevent their being fixed as mutations in the next round of complementary DNA synthesis (Iyer et al., 2006; Jiricny, 2013; Kunkel and Erie, 2015). Since erroneously inserted nucleotides are present on newly synthesized DNA strands, identification of the newly synthesized strand is a critical step in MMR. By rectifying replication errors, MMR increases the replication fidelity by ~2 orders of magnitude in yeast (Lujan et al., 2014). In humans, genetic or epigenetic inactivation of MMR genes elevates the risk of tumorigenesis in both hereditary and sporadic manners (Jiricny, 2013).

In Escherichia coli, MMR distinguishes the parental and daughter strands mainly by using adenine methylation on GATC sites (Lahue et al., 1989; Iyer et al., 2006). Mismatches are recognized by the mismatch sensor MutS homodimer. MutS and the MutL homodimer then activate the latent nicking-endonuclease MutH, which cleaves the unmethylated strand at the hemi-methylated GATC sequences. MMR is possible from the time of hemi-methylated GATC generation by DNA synthesis until full methylation of the GATC sites. Maintaining this temporal window is critical for efficient MMR, because over-expression of the Dam methylase, by which full-methylation of the GATC sites is accelerated, significantly elevates the mutation frequency (Herman and Modrich, 1981; Marinus et al., 1984). The E. coli MMR system can also correct replication errors through a methylation-independent mechanism, where strand discontinuities can substitute for GATC methylation both in vivo and in vitro (Laengle-Rouault et al., 1986; Lahue et al., 1987; 1989).

Eukaryotic MMR is directed by strand discontinuities such as nicks or gaps in vitro (Holmes et al., 1990; Thomas et al., 1991). Two MutS heterodimers, MutSα (Msh2-Msh6) and MutSβ (Msh2-Msh3) recognize replication errors; MutSα has a biased preference for base-base mismatches and small insertion/deletion loops (IDLs), while MutSβ preferentially recognizes large IDLs (Iyer et al., 2006; Jiricny, 2013; Kunkel and Erie, 2015). After mismatch binding, MutSα/β are converted into closed ‘clamp-like’ forms, by which they can translocate along DNA. They then recruit the latent nicking-endonuclease MutLα (Mlh1-Pms2 in vertebrates and Mlh1-Pms1 in yeast) to initiate the removal of the error-carrying DNA strand. Two other eukaryotic MutL homologs, MutLβ (Mlh1-Pms1 in vertebrates and Mlh1-Mlh2 in yeast) and MutLγ (Mlh1-Mlh3) are suggested to play minor roles in somatic MMR (Jiricny, 2013; Campbell et al., 2014). Successful reconstitutions of eukaryotic MMR have shown that MutSα, or MutSβ, and MutLα, the Proliferating Cell Nuclear Antigen (PCNA) sliding clamp, the clamp loader Replication Factor C (RFC), the Exo1 exonuclease, and the DNA synthesis components promote MMR when a strand discontinuity is present (Genschel and Modrich, 2003; Dzantiev et al., 2004; Constantin et al., 2005; Zhang et al., 2005). A strand discontinuity, which can occur on either 5’- or 3’-side of the mismatch, activates MutLα through a MutSα- and PCNA-dependent mechanism to induce successive rounds of nicking on the error-carrying strand (Kadyrov et al., 2006; 2007; Pluciennik et al., 2010). Single-strand DNA termini such as 5’-ends of the Okazaki fragments serve as entry points for Exo1 and strand discrimination signals in vivo as well (Pavlov et al., 2003; Nick McElhinny et al., 2010; Liberti et al., 2013; Lujan et al., 2014; Liu et al., 2015). Recent studies have revealed that ribonucleotides embedded by replicative DNA polymerases serve as strand signals for MMR in vitro and contribute to a sub-fraction of leading strand MMR in vivo, after converted into single-strand gaps through RNaseH2-dependent ribonucleotide excision repair (Ghodgaonkar et al., 2013; Lujan et al., 2013).

PCNA has also been presumed to be the strand discrimination marker in eukaryotes (Umar et al., 1996; Chen et al., 1999; Pavlov et al., 2003; Dzantiev et al., 2004; Kadyrov et al., 2006; Pluciennik et al., 2010; Hombauer et al., 2011b; Peña-Diaz and Jiricny, 2012; Georgescu et al., 2015; Kunkel and Erie, 2015). PCNA is a ring-shaped homo-trimer that supports various DNA transactions including DNA replication and repair (Georgescu et al., 2015). PCNA is loaded onto DNA from the template-primer junction by RFC, and likely unloaded by an RFC-like complex containing Elg1 after the completion of DNA synthesis (Kubota et al., 2013; 2015). Since its DNA binding is asymmetric with respect to polarities of the parental and daughter strands, DNA-bound PCNA can hold information for the newly synthesized strand (Bowman et al., 2004; Georgescu et al., 2015). PCNA plays an essential role in an early MMR step that precedes degradation of the error-carrying strand (Umar et al., 1996). PCNA loaded from a strand discontinuity induces strand-specific, mismatch- and MutSα-dependent activation of the MutLα endonuclease (Kadyrov et al., 2006; 2007; Pluciennik et al., 2010). When PCNA is loaded onto closed circular DNA without defined orientation, it induces unbiased nicking on both DNA strands (Pluciennik et al., 2010; 2013). These findings have led to a proposal that orientation of DNA-bound PCNA is a critical determinant for the nicking specificity of MutLα.

In addition to its proposed role in strand discrimination, PCNA is also involved in multiple steps of MMR. Numbers of PCNA mutants in yeast exhibit mutator phenotypes that are epistatic to MMR mutations (Johnson et al., 1996; Umar et al., 1996; Chen et al., 1999; Amin et al., 2001; Lau et al., 2002; Goellner et al., 2014). It interacts with MutSα/β through their PCNA-interacting peptide (PIP) motifs, which reside at the N-termini of Msh6 and Msh3 (Clark et al., 2000; Flores-Rozas et al., 2000; Kleczkowska et al., 2001). In both cases, the PIP motifs and the mispair binding domains are connected through long linkers, which are predicted to be disordered in yeast (Shell et al., 2007). Mutants of the PIP motifs in Msh6 and Msh3 show substantial but not complete reduction of the MMR activity, indicating that the PIP motif plays an important, yet assistive, non-essential role in MMR (Clark et al., 2000; Flores-Rozas et al., 2000). The PIP motif has been proposed to tether MutSα to the replication fork to assist its mismatch recognition (Kleczkowska et al., 2001; Hombauer et al., 2011a; Haye and Gammie, 2015). A recent study has shown that this motif in MutSα is important for a late MMR step(s) involving degradation of the error-carrying strand, especially by the Exo1-independent mechanism (Goellner et al., 2014). These findings suggest that PCNA coordinates multiple reactions in MMR, yet the exact role of PCNA in MMR is still not clearly defined.

An important remaining question in eukaryotic MMR is how (and how long) the strand discrimination signals are retained after DNA synthesis. It has been shown in yeast that MutSα must be present during S-phase to suppress mutations (Hombauer et al., 2011b). Therefore, the generation and retention of the strand signals for MMR should be intimately coupled with DNA replication. Since PCNA transiently remains on DNA after the completion of DNA synthesis, DNA-bound PCNA could mediate the coupling of MMR with DNA replication. However, because experimental evidence that DNA-bound PCNA induces strand-specific MMR in the absence of strand discontinuities is currently lacking, it is still uncertain whether PCNA by itself can mediate the coupling of MMR with DNA replication. In addition, how functional interaction between the MMR system and DNA-bound PCNA, which should be unloaded from DNA shortly after the completion of DNA synthesis, is ensured remains highly unclear.

Here, we demonstrate that MutSα and PCNA play a critical role in maintenance of the MMR capability after DNA synthesis. We show that nucleoplasmic extract (NPE) of Xenopus eggs efficiently induces gap-directed, strand-specific MMR, whose capability remains even after sealing of the gap. We further show that DNA-bound PCNA induces strand-specific MMR in the absence of strand discontinuities. Strikingly, we found that MutSα attenuates unloading of PCNA and retains the MMR capability largely through its PIP motif. Our data thus identify PCNA as the eukaryotic strand discrimination molecule that retains the MMR capability after DNA synthesis, and delineate a critical role of MutSα in maintenance of the temporal window for eukaryotic MMR.

Identification of the newly synthesized DNA strand is vital to post-replicative mismatch repair. Since DNA replication generates identical copies of parental DNA, the strand discrimination reaction in MMR necessarily requires direct or indirect interplay with the replication complex (Wagner and Meselson, 1976). Because of the asymmetric nature of its binding to DNA, its critical roles in an early step of MMR and in the activation of MutLα, PCNA is presumed to be the mediator of the interplay between the replication complex and MMR (Umar et al., 1996; Chen et al., 1999; Pavlov et al., 2003; Dzantiev et al., 2004; Kadyrov et al., 2006; Pluciennik et al., 2010; Hombauer et al., 2011b; Peña-Diaz and Jiricny, 2012; Georgescu et al., 2015; Kunkel and Erie, 2015). In this study, we provided experimental evidence that DNA-loaded PCNA indeed directs strand-specific MMR in the absence of strand discontinuities. We further discovered that MutSα inhibits PCNA unloading using its PIP motif on the N-terminus of Msh6, and that this reaction significantly extends the temporal window during which MMR is possible. Our results thus uncovered a novel mechanism in which interplay between PCNA and MutSα maintains the post-replicative temporal window for MMR.

NPE has been extensively used as a physiological model system that recapitulates various DNA repair reactions coupled with DNA synthesis (Räschle et al., 2008; Olivera Harris et al., 2015). As seen in other systems, we observed efficient bidirectional MMR in NPE (frequently >80% efficient). Curiously, we noticed some difference between NPE and other systems. Both in reconstitution systems and in human cell extracts, 5’-nick directed MMR can occur in the absence of MutLα, whereas 3’-nick directed MMR depends on MutLα (Drummond et al., 1996; Genschel and Modrich, 2003; Dzantiev et al., 2004; Constantin et al., 2005; Bowen et al., 2013). In contrast, not only 3’-gap directed but also 5’-gap directed MMR was dependent on MutL complexes in NPE (Figure 1B–E). Although this could possibly reflect difference in the 5’ to 3’ strand excision mechanism between species, we assume that the difference largely come from highly efficient DNA synthesis and ligation in NPE; quick sealing of a gap would necessitate MutLα-dependent incision for MMR even when a gap was placed on the 5’-side of the mismatch. Interestingly, DNA synthesis at the 3’-gap clearly precedes MMR (Figure 1F), indicating that strand degradation does not progressively occur from the 3’-terminus of the gap. PCNA molecules loaded at the 3’ terminus, perhaps more than one trimer, could be used by both MMR and DNA polymerases without significantly interfering with each other (see below). Consistently, a nick is suggested to function as a strand signal rather than an entry point for exonucleases in crude extracts of Xenopus eggs (Varlet et al., 1996).

Our stepwise incubation experiments demonstrated that MMR is possible even after the complete sealing of the gap (Figure 2). Three lines of evidence indicate that the molecule responsible for memorizing strand information is PCNA. First, directional loading of PCNA efficiently bypassed the need for strand discontinuities in strand-specific MMR (Figure 3 and 4). Second, the amount of DNA-bound PCNA correlated with the MMR capability (Figure 6). Third, MutSα maintained the MMR capability derived from a gap largely through the PIP motif (Figure 9). The fact that PCNA was essential for the gap-filling reaction in NPE also supports this conclusion, because at least one PCNA trimer must be involved in DNA synthesis at the gap. It is unlikely that RFC determines the strand specificity of MMR in our experiments, because amounts of RFC remained on DNA were insufficient to explain the efficiency of MMR. Based on the above evidence, we suggest that PCNA would function as the strand discrimination signal for MMR after the disappearance of local strand discontinuities during DNA replication. A series of seminal studies from the Modrich lab has led to the model in which DNA-bound PCNA activates the MutLα nicking endonuclease to initiate strand specific MMR (Kadyrov et al., 2006; 2007; Pluciennik et al., 2010; 2013). Our data are consistent with this prevailing model, since 1) DNA-bound PCNA induced strand-specific MMR in NPE, and 2) PCNA-directed MMR in NPE was completely dependent on the MutL complexes. As gap-directed MMR in NPE was bidirectional and preferentially degrades the DNA strand through the shortest path from the gap to the mismatch, we assume that the PCNA-directed MMR would function in the same manner. If so, the strand excision complex presumably built upon a specific face of PCNA would catalyze strand degradation in either direction depending on the relative orientation toward a mismatch. How this can be achieved is an important next question, and the system established here will be a useful tool to investigate this mechanism. However, since our plasmid substrates carry multiple PCNA molecules with free sliding ability, we currently do not exclude a possibility that PCNA-directed MMR catalyzes only a specific direction of strand excision, and PCNA molecules localized on a specific side toward a mismatch are preferentially used for MMR.

Unexpectedly, the 'strand memory' experiments revealed that MutSα/β significantly extend the duration of strand memory. Our data clearly indicate that this retention involves inhibition of PCNA unloading. A recent study showed that DNA polymerase δ captures DNA-bound PCNA which is otherwise quickly unloaded by the clamp-unloading activity of RFC (Hedglin et al., 2013). It would be possible that MutSα inhibits PCNA unloading in a similar manner; binding of the PIP motif onto PCNA may physically prevent the engagement of PCNA with RFC. Since in vivo PCNA unloading during S-phase largely depends on the Elg1-containing RFC-like complex (Elg1-RLC) both in humans and in yeast (Kubota et al., 2013; Lee et al., 2013; Shiomi and Nishitani, 2013; Kubota et al., 2015), the PIP-motif in MutSα may also limit the access of Elg1-RLC to DNA-bound PCNA. Detailed understanding of the molecular mechanism of this inhibition of PCNA unloading must await reconstitution of this reaction in vitro.

Interestingly, the PIP motif and the mispair-binding domain are connected by a long (~300 a.a.) linker that is predicted to be disordered in yeast (Shell et al., 2007). Although a significant portion of this linker in human MutSα seems to adopt a globular conformation (Iyer et al., 2008), MutSα might be able to retain PCNA that is located far from MutSα using this long 'PIP-arm', possibly even over nucleosomes. In this scenario, strand removal would initiate when MutLα reaches to PCNA that is retained by the PIP-arm of MutSα. This could be accomplished either by sliding of MutSα-MutLα complexes along the DNA contour, or through loading of multiple MutLα molecules on DNA (Hombauer et al., 2011a; Qiu et al., 2015). Interaction between PCNA that is retained by MutSα and MutLα may invoke a specific strand degradation pathway, because a recent study in yeast has suggested that recruitment or retention of PCNA by MutSα around mismatches activates a specific strand removal pathway that is independent of Exo1 (Goellner et al., 2014). As seen for MutSα, we predict that MutSβ may inhibit PCNA unloading as well, since the NTR of yeast Msh6 and Msh3 are inter-exchangeable without significant deleterious effects (Shell et al., 2007). However, because NTRs of MutSα and MutSβ likely possess different characteristics such as the PWWP domain found only in vertebrate Msh6 (Clark et al., 2007; Laguri et al., 2008; Iyer et al., 2010; Li et al., 2013), this point will need rigorous investigation.

To our surprise, xMutSα^PIP and xMutSα^ΔN retained partial strand memory. A possibility would be that, although our PCNA unloading experiment failed to detect attenuation of PCNA unloading with these mutants (Figure 8D), they might have the ability to maintain DNA-bound PCNA that is below our detection limit. Considering the fact that we quantified PCNA after several washing steps, our quantification almost certainly underestimated the number of PCNA molecules on DNA. Alternatively, PCNA and/or a gap could induce structural alteration of MutSα (and possibly other chromatin binding factors), to maintain the strand information. How MutSα retains strand memory independently of the NTR will be an interesting question for future studies.

From the data presented in this work, we propose a model wherein two different modes of strand memory contribute to MMR (Figure 10). After the completion of local DNA synthesis, e.g. after ligation of Okazaki fragments, PCNA remains on DNA for a certain period until unloaded by an Rfc3-containing complex(es) (t_1/2 = ~2 min in NPE; Figure 10, [2]). The MMR system utilizes such PCNA as strand memory. However, if PCNA is not available closely adjacent to mismatches, MutSα (and possibly MutSβ also) retains available PCNA with its long 'PIP-arm', to maintain strand memory for subsequent MMR (Figure 10, [3]). This mode of memory survives for >30 min in NPE, and perhaps even longer in cells. In addition to the PIP-dependent tethering of MutSα/β to the replication fork (Kleczkowska et al., 2001; Hombauer et al., 2011a; Haye and Gammie, 2015), we suggest that the PIP motif would contribute to the replication fidelity through this mode of strand memory. Yeast studies have shown that the PIP motif has an assistive but not essential role in MMR (Clark et al., 2000; Flores-Rozas et al., 2000). The PIP-deficient MutSα complexes were proficient in repairing a mismatch in our in vitro MMR system (Figure 8—figure supplement 1), and similar observations were already reported both in human cell extracts and in a yeast in vitro reconstitution system (Kleczkowska et al., 2001; Iyer et al., 2008; Bowen et al., 2013). Therefore, a simple MMR reaction involving a mismatch and a closely-placed strand-discrimination signal, either PCNA or a strand discontinuity, would not require retention of PCNA by MutSα. However, when a stepwise incubation, which mimics transient arrest of the MMR reaction, was involved, the PIP motif became important for MMR (Figure 9). We speculate that inhibition of PCNA unloading is needed for MMR only under specific situations, e.g. when a strand-discrimination signal is buried within chromatin. In such difficult-to-repair situations, successful MMR may rely on the maintenance of the strand signals by MutSα. Since the leading strand seems to have less PCNA than the lagging strand (Yu et al., 2014), such situations could happen more frequently in the leading strand, in which MMR is less efficient than in the lagging strand (Pavlov et al., 2003; Lujan et al., 2014). An interesting possibility worth testing in the future would be that MutSα may take over PCNA from other factors such as DNA polymerases. Hijacking PCNA from polymerases may not be overly deleterious to DNA synthesis, because new PCNA loading at the 3’-terminus would quickly restore the PCNA-polymerase complex. Similar hijacking might occur on other PCNA interactors such as CAF-1 and FEN-1, both of which are reported to show complex interplay with MutSα (Kadyrova et al., 2011; Schöpf et al., 2012; Kadyrova et al., 2015; Liu et al., 2015). On the leading strand that has less PCNA, such competition could be important for effective MMR.

Figure 10

Download asset Open asset

Much of our understanding of MMR has come from bacterial studies including E. coli, whose MMR utilizes adenine methylation as the strand discrimination signal. Although E. coli MMR seems to function with strand discontinuities occurring at the replication fork, GATC methylation plays a critical role in maintaining the temporal window permissive to MMR. Strand discontinuities are clearly strand discrimination signals in eukaryotic MMR in vitro (Holmes et al., 1990; Thomas et al., 1991). There is also concrete evidence that single-strand DNA termini such as 5’-ends of the Okazaki fragments or strand breaks generated by ribonucleotide excision repair contribute to MMR in vivo(Pavlov et al., 2003; Nick McElhinny et al., 2010; Ghodgaonkar et al., 2013; Liberti et al., 2013; Lujan et al., 2013; 2014; Liu et al., 2015). We suggest that, in addition to such DNA termini, two modes of strand memory, somewhat resembling short-term and long-term memories in neuroscience, would operate as functional parallels of GATC methylation in E. coli. Analogous reactions could operate in bacteria and archaea lacking the methyl-dependent MMR (Pillon et al., 2015), and it would be interesting to study dynamics of the replication clamp in such organisms. Further investigation of the regulation of PCNA dynamics by MMR both in NPE and in vivo will lead to a more detailed understanding of complex interplay between the MMR system and the replication complex.

1. 1. Amin NS
   2. Nguyen MN
   3. Oh S
   4. Kolodner RD

   (2001) exo1-Dependent mutator mutations: model system for studying functional interactions in mismatch repair

   Molecular and Cellular Biology 21:5142–5155.

   https://doi.org/10.1128/MCB.21.15.5142-5155.2001

   • Google Scholar
2. 1. Bowen N
   2. Smith CE
   3. Srivatsan A
   4. Willcox S
   5. Griffith JD
   6. Kolodner RD

   (2013) Reconstitution of long and short patch mismatch repair reactions using Saccharomyces cerevisiae proteins

   Proceedings of the National Academy of Sciences of the United States of America 110:18472–18477.

   https://doi.org/10.1073/pnas.1318971110

   • Google Scholar
3. 1. Bowers J
   2. Tran PT
   3. Joshi A
   4. Liskay RM
   5. Alani E

   (2001) MSH-MLH complexes formed at a DNA mismatch are disrupted by the PCNA sliding clamp

   Journal of Molecular Biology 306:957–968.

   https://doi.org/10.1006/jmbi.2001.4467

   • Google Scholar
4. 1. Bowman GD
   2. O'Donnell M
   3. Kuriyan J

   (2004) Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex

   Nature 429:724–730.

   https://doi.org/10.1038/nature02585

   • Google Scholar
5. 1. Campbell CS
   2. Hombauer H
   3. Srivatsan A
   4. Bowen N
   5. Gries K
   6. Desai A
   7. Putnam CD
   8. Kolodner RD

   (2014) Mlh2 is an accessory factor for DNA mismatch repair in Saccharomyces cerevisiae

   PLoS Genetics 10:e1004327.

   https://doi.org/10.1371/journal.pgen.1004327

   • Google Scholar
6. 1. Chen C
   2. Merrill BJ
   3. Lau PJ
   4. Holm C
   5. Kolodner RD

   (1999) Saccharomyces cerevisiae pol30 (proliferating cell nuclear antigen) mutations impair replication fidelity and mismatch repair

   Molecular and Cellular Biology 19:7801–7815.

   https://doi.org/10.1128/MCB.19.11.7801

   • Google Scholar
7. 1. Clark AB
   2. Deterding L
   3. Tomer KB
   4. Kunkel TA

   (2007) Multiple functions for the N-terminal region of Msh6

   Nucleic Acids Research 35:4114–4123.

   https://doi.org/10.1093/nar/gkm409

   • Google Scholar
8. 1. Clark AB
   2. Valle F
   3. Drotschmann K
   4. Gary RK
   5. Kunkel TA

   (2000) Functional interaction of proliferating cell nuclear antigen with MSH2-MSH6 and MSH2-MSH3 complexes

   The Journal of Biological Chemistry 275:36498–36501.

   https://doi.org/10.1074/jbc.C000513200

   • Google Scholar
9. 1. Constantin N
   2. Dzantiev L
   3. Kadyrov FA
   4. Modrich P

   (2005) Human mismatch repair: reconstitution of a nick-directed bidirectional reaction

   The Journal of Biological Chemistry 280:39752–39761.

   https://doi.org/10.1074/jbc.M509701200

   • Google Scholar
10. 1. Drummond JT
    2. Anthoney A
    3. Brown R
    4. Modrich P

    (1996) Cisplatin and adriamycin resistance are associated with MutLalpha and mismatch repair deficiency in an ovarian tumor cell line

    The Journal of Biological Chemistry 271:19645–19648.

    https://doi.org/10.1074/jbc.271.33.19645

    • Google Scholar
11. 1. Dzantiev L
    2. Constantin N
    3. Genschel J
    4. Iyer RR
    5. Burgers PM
    6. Modrich P

    (2004) A defined human system that supports bidirectional mismatch-provoked excision

    Molecular Cell 15:31–41.

    https://doi.org/10.1016/j.molcel.2004.06.016

    • Google Scholar
12. 1. Flores-Rozas H
    2. Clark D
    3. Kolodner RD

    (2000) Proliferating cell nuclear antigen and Msh2p-Msh6p interact to form an active mispair recognition complex

    Nature Genetics 26:375–378.

    https://doi.org/10.1038/81708

    • Google Scholar
13. 1. Fukuda K
    2. Morioka H
    3. Imajou S
    4. Ikeda S
    5. Ohtsuka E
    6. Tsurimoto T

    (1995) Structure-function relationship of the eukaryotic DNA replication factor, proliferating cell nuclear antigen

    The Journal of Biological Chemistry 270:22527–22534.

    https://doi.org/10.1074/jbc.270.38.22527

    • Google Scholar
14. 1. Genschel J
    2. Modrich P

    (2003) Mechanism of 5'-directed excision in human mismatch repair

    Molecular Cell 12:1077–1086.

    https://doi.org/10.1016/S1097-2765(03)00428-3

    • Google Scholar
15. 1. Georgescu R
    2. Langston L
    3. O'Donnell M

    (2015) A proposal: Evolution of PCNA's role as a marker of newly replicated DNA

    DNA Repair 29:4–15.

    https://doi.org/10.1016/j.dnarep.2015.01.015

    • Google Scholar
16. 1. Ghodgaonkar MM
    2. Lazzaro F
    3. Olivera-Pimentel M
    4. Artola-Borán M
    5. Cejka P
    6. Reijns MA
    7. Jackson AP
    8. Plevani P
    9. Muzi-Falconi M
    10. Jiricny J

    (2013) Ribonucleotides misincorporated into DNA act as strand-discrimination signals in eukaryotic mismatch repair

    Molecular Cell 50:323–332.

    https://doi.org/10.1016/j.molcel.2013.03.019

    • Google Scholar
17. 1. Goellner EM
    2. Smith CE
    3. Campbell CS
    4. Hombauer H
    5. Desai A
    6. Putnam CD
    7. Kolodner RD

    (2014) PCNA and Msh2-Msh6 activate an Mlh1-Pms1 endonuclease pathway required for Exo1-independent mismatch repair

    Molecular Cell 55:291–304.

    https://doi.org/10.1016/j.molcel.2014.04.034

    • Google Scholar
18. 1. Haye JE
    2. Gammie AE

    (2015) The eukaryotic mismatch recognition complexes track with the replisome during DNA synthesis

    PLoS Genetics 11:e1005719.

    https://doi.org/10.1371/journal.pgen.1005719

    • Google Scholar
19. 1. Hedglin M
    2. Perumal SK
    3. Hu Z
    4. Benkovic S

    (2013) Stepwise assembly of the human replicative polymerase holoenzyme

    eLife 2:e00278.

    https://doi.org/10.7554/eLife.00278

    • Google Scholar
20. 1. Herman GE
    2. Modrich P

    (1981)

    Escherichia coli K-12 clones that overproduce dam methylase are hypermutable

    Journal of Bacteriology 145:644–646.

    • Google Scholar
21. 1. Higashi TL
    2. Ikeda M
    3. Tanaka H
    4. Nakagawa T
    5. Bando M
    6. Shirahige K
    7. Kubota Y
    8. Takisawa H
    9. Masukata H
    10. Takahashi TS

    (2012) The prereplication complex recruits XEco2 to chromatin to promote cohesin acetylation in Xenopus egg extracts

    Current Biology  22:977–988.

    https://doi.org/10.1016/j.cub.2012.04.013

    • Google Scholar
22. 1. Holmes J
    2. Clark S
    3. Modrich P

    (1990) Strand-specific mismatch correction in nuclear extracts of human and Drosophila melanogaster cell lines

    Proceedings of the National Academy of Sciences of the United States of America 87:5837–5841.

    https://doi.org/10.1073/pnas.87.15.5837

    • Google Scholar
23. 1. Hombauer H
    2. Campbell CS
    3. Smith CE
    4. Desai A
    5. Kolodner RD

    (2011a) Visualization of eukaryotic DNA mismatch repair reveals distinct recognition and repair intermediates

    Cell 147:1040–1053.

    https://doi.org/10.1016/j.cell.2011.10.025

    • Google Scholar
24. 1. Hombauer H
    2. Srivatsan A
    3. Putnam CD
    4. Kolodner RD

    (2011b) Mismatch repair, but not heteroduplex rejection, is temporally coupled to DNA replication

    Science 334:1713–1716.

    https://doi.org/10.1126/science.1210770

    • Google Scholar
25. 1. Iyer RR
    2. Pluciennik A
    3. Burdett V
    4. Modrich PL

    (2006) DNA mismatch repair: functions and mechanisms

    Chemical Reviews 106:302–323.

    https://doi.org/10.1021/cr0404794

    • Google Scholar
26. 1. Iyer RR
    2. Pluciennik A
    3. Genschel J
    4. Tsai MS
    5. Beese LS
    6. Modrich P

    (2010) MutLalpha and proliferating cell nuclear antigen share binding sites on MutSbeta

    The Journal of Biological Chemistry 285:11730–11739.

    https://doi.org/10.1074/jbc.M110.104125

    • Google Scholar
27. 1. Iyer RR
    2. Pohlhaus TJ
    3. Chen S
    4. Hura GL
    5. Dzantiev L
    6. Beese LS
    7. Modrich P

    (2008) The MutSalpha-proliferating cell nuclear antigen interaction in human DNA mismatch repair

    The Journal of Biological Chemistry 283:13310–13319.

    https://doi.org/10.1074/jbc.M800606200

    • Google Scholar
28. 1. Jiricny J

    (2013) Postreplicative mismatch repair

    Cold Spring Harbor Perspectives in Biology 5:a012633.

    https://doi.org/10.1101/cshperspect.a012633

    • Google Scholar
29. 1. Johnson RE
    2. Kovvali GK
    3. Guzder SN
    4. Amin NS
    5. Holm C
    6. Habraken Y
    7. Sung P
    8. Prakash L
    9. Prakash S

    (1996) Evidence for involvement of yeast proliferating cell nuclear antigen in DNA mismatch repair

    The Journal of Biological Chemistry 271:27987–27990.

    https://doi.org/10.1074/jbc.271.45.27987

    • Google Scholar
30. 1. Kadyrov FA
    2. Dzantiev L
    3. Constantin N
    4. Modrich P

    (2006) Endonucleolytic function of MutLalpha in human mismatch repair

    Cell 126:297–308.

    https://doi.org/10.1016/j.cell.2006.05.039

    • Google Scholar
31. 1. Kadyrov FA
    2. Holmes SF
    3. Arana ME
    4. Lukianova OA
    5. O'Donnell M
    6. Kunkel TA
    7. Modrich P

    (2007) Saccharomyces cerevisiae MutLalpha is a mismatch repair endonuclease

    The Journal of Biological Chemistry 282:37181–37190.

    https://doi.org/10.1074/jbc.M707617200

    • Google Scholar
32. 1. Kadyrova LY
    2. Blanko ER
    3. Kadyrov FA

    (2011) CAF-I-dependent control of degradation of the discontinuous strands during mismatch repair

    Proceedings of the National Academy of Sciences of the United States of America 108:2753–2758.

    https://doi.org/10.1073/pnas.1015914108

    • Google Scholar
33. 1. Kadyrova LY
    2. Dahal BK
    3. Kadyrov FA

    (2015) Evidence that the DNA mismatch repair system removes 1-nucleotide Okazaki fragment flaps

    The Journal of Biological Chemistry 290:24051–24065.

    https://doi.org/10.1074/jbc.M115.660357

    • Google Scholar
34. 1. Kleczkowska HE
    2. Marra G
    3. Lettieri T
    4. Jiricny J

    (2001) hMSH3 and hMSH6 interact with PCNA and colocalize with it to replication foci

    Genes & Development 15:724–736.

    https://doi.org/10.1101/gad.191201

    • Google Scholar
35. 1. Kubota T
    2. Katou Y
    3. Nakato R
    4. Shirahige K
    5. Donaldson AD

    (2015) Replication-Coupled PCNA Unloading by the Elg1 Complex Occurs Genome-wide and Requires Okazaki Fragment Ligation

    Cell Reports 12:774–787.

    https://doi.org/10.1016/j.celrep.2015.06.066

    • Google Scholar
36. 1. Kubota T
    2. Nishimura K
    3. Kanemaki MT
    4. Donaldson AD

    (2013) The Elg1 replication factor C-like complex functions in PCNA unloading during DNA replication

    Molecular Cell 50:273–280.

    https://doi.org/10.1016/j.molcel.2013.02.012

    • Google Scholar
37. 1. Kunkel TA
    2. Erie DA

    (2015) Eukaryotic mismatch repair in relation to DNA replication

    Annual Review of Genetics 49:291–313.

    https://doi.org/10.1146/annurev-genet-112414-054722

    • Google Scholar
38. 1. Laengle-Rouault F
    2. Maenhaut-Michel G
    3. Radman M

    (1986)

    GATC sequence and mismatch repair in Escherichia coli

    The EMBO Journal 5:2009–2013.

    • Google Scholar
39. 1. Laguri C
    2. Duband-Goulet I
    3. Friedrich N
    4. Axt M
    5. Belin P
    6. Callebaut I
    7. Gilquin B
    8. Zinn-Justin S
    9. Couprie J

    (2008) Human mismatch repair protein MSH6 contains a PWWP domain that targets double stranded DNA

    Biochemistry 47:6199–6207.

    https://doi.org/10.1021/bi7024639

    • Google Scholar
40. 1. Lahue RS
    2. Au KG
    3. Modrich P

    (1989) DNA mismatch correction in a defined system

    Science 245:160–164.

    https://doi.org/10.1126/science.2665076

    • Google Scholar
41. 1. Lahue RS
    2. Su SS
    3. Modrich P

    (1987) Requirement for d(GATC) sequences in Escherichia coli mutHLS mismatch correction

    Proceedings of the National Academy of Sciences of the United States of America 84:1482–1486.

    https://doi.org/10.1073/pnas.84.6.1482

    • Google Scholar
42. 1. Lau PJ
    2. Flores-Rozas H
    3. Kolodner RD

    (2002) Isolation and characterization of new proliferating cell nuclear antigen (POL30) mutator mutants that are defective in DNA mismatch repair

    Molecular and Cellular Biology 22:6669–6680.

    https://doi.org/10.1128/MCB.22.19.6669-6680.2002

    • Google Scholar
43. 1. Lau PJ
    2. Kolodner RD

    (2003) Transfer of the MSH2.MSH6 complex from proliferating cell nuclear antigen to mispaired bases in DNA

    The Journal of Biological Chemistry 278:14–17.

    https://doi.org/10.1074/jbc.C200627200

    • Google Scholar
44. 1. Lebofsky R
    2. Takahashi T
    3. Walter JC

    (2009) DNA replication in nucleus-free Xenopus egg extracts

    Methods in Molecular Biology 521:229–252.

    https://doi.org/10.1007/978-1-60327-815-7_13

    • Google Scholar
45. 1. Lee KY
    2. Fu H
    3. Aladjem MI
    4. Myung K

    (2013) ATAD5 regulates the lifespan of DNA replication factories by modulating PCNA level on the chromatin

    The Journal of Cell Biology 200:31–44.

    https://doi.org/10.1083/jcb.201206084

    • Google Scholar
46. 1. Lee SD
    2. Alani E

    (2006) Analysis of interactions between mismatch repair initiation factors and the replication processivity factor PCNA

    Journal of Molecular Biology 355:175–184.

    https://doi.org/10.1016/j.jmb.2005.10.059

    • Google Scholar
47. 1. Li F
    2. Mao G
    3. Tong D
    4. Huang J
    5. Gu L
    6. Yang W
    7. Li GM

    (2013) The histone mark H3K36me3 regulates human DNA mismatch repair through its interaction with MutSα

    Cell 153:590–600.

    https://doi.org/10.1016/j.cell.2013.03.025

    • Google Scholar
48. 1. Liberti SE
    2. Larrea AA
    3. Kunkel TA

    (2013) Exonuclease 1 preferentially repairs mismatches generated by DNA polymerase α

    DNA Repair 12:92–96.

    https://doi.org/10.1016/j.dnarep.2012.11.001

    • Google Scholar
49. 1. Liu S
    2. Lu G
    3. Ali S
    4. Liu W
    5. Zheng L
    6. Dai H
    7. Li H
    8. Xu H
    9. Hua Y
    10. Zhou Y
    11. Ortega J
    12. Li GM
    13. Kunkel TA
    14. Shen B

    (2015) Okazaki fragment maturation involves α-segment error editing by the mammalian FEN1/MutSα functional complex

    The EMBO Journal 34:1829–1843.

    https://doi.org/10.15252/embj.201489865

    • Google Scholar
50. 1. Lujan SA
    2. Clausen AR
    3. Clark AB
    4. MacAlpine HK
    5. MacAlpine DM
    6. Malc EP
    7. Mieczkowski PA
    8. Burkholder AB
    9. Fargo DC
    10. Gordenin DA
    11. Kunkel TA

    (2014) Heterogeneous polymerase fidelity and mismatch repair bias genome variation and composition

    Genome Research 24:1751–1764.

    https://doi.org/10.1101/gr.178335.114

    • Google Scholar
51. 1. Lujan SA
    2. Williams JS
    3. Clausen AR
    4. Clark AB
    5. Kunkel TA

    (2013) Ribonucleotides are signals for mismatch repair of leading-strand replication errors

    Molecular Cell 50:437–443.

    https://doi.org/10.1016/j.molcel.2013.03.017

    • Google Scholar
52. 1. Marinus MG
    2. Poteete A
    3. Arraj JA

    (1984) Correlation of DNA adenine methylase activity with spontaneous mutability in Escherichia coli K-12

    Gene 28:123–125.

    https://doi.org/10.1016/0378-1119(84)90095-7

    • Google Scholar
53. 1. Mattock H
    2. Jares P
    3. Zheleva DI
    4. Lane DP
    5. Warbrick E
    6. Blow JJ

    (2001) Use of peptides from p21 (Waf1/Cip1) to investigate PCNA function in Xenopus egg extracts

    Experimental Cell Research 265:242–251.

    https://doi.org/10.1006/excr.2001.5181

    • Google Scholar
54. 1. Nick McElhinny SA
    2. Kissling GE
    3. Kunkel TA

    (2010) Differential correction of lagging-strand replication errors made by DNA polymerases {alpha} and {delta}

    Proceedings of the National Academy of Sciences of the United States of America 107:21070–21075.

    https://doi.org/10.1073/pnas.1013048107

    • Google Scholar
55. 1. Ohta S
    2. Shiomi Y
    3. Sugimoto K
    4. Obuse C
    5. Tsurimoto T

    (2002) A proteomics approach to identify proliferating cell nuclear antigen (PCNA)-binding proteins in human cell lysates. Identification of the human CHL12/RFCs2-5 complex as a novel PCNA-binding protein

    The Journal of Biological Chemistry 277:40362–40367.

    https://doi.org/10.1074/jbc.M206194200

    • Google Scholar
56. 1. Olivera Harris M
    2. Kallenberger L
    3. Artola Borán M
    4. Enoiu M
    5. Costanzo V
    6. Jiricny J

    (2015) Mismatch repair-dependent metabolism of O6-methylguanine-containing DNA in Xenopus laevis egg extracts

    DNA Repair 28:1–7.

    https://doi.org/10.1016/j.dnarep.2015.01.014

    • Google Scholar
57. 1. Pavlov YI
    2. Mian IM
    3. Kunkel TA

    (2003) Evidence for preferential mismatch repair of lagging strand DNA replication errors in yeast

    Current Biology 13:744–748.

    https://doi.org/10.1016/S0960-9822(03)00284-7

    • Google Scholar
58. 1. Peña-Diaz J
    2. Jiricny J

    (2012) Mammalian mismatch repair: error-free or error-prone?

    Trends in Biochemical Sciences 37:206–214.

    https://doi.org/10.1016/j.tibs.2012.03.001

    • Google Scholar
59. 1. Pillon MC
    2. Babu VM
    3. Randall JR
    4. Cai J
    5. Simmons LA
    6. Sutton MD
    7. Guarné A

    (2015) The sliding clamp tethers the endonuclease domain of MutL to DNA

    Nucleic Acids Research 43:10746–10759.

    https://doi.org/10.1093/nar/gkv918

    • Google Scholar
60. 1. Pluciennik A
    2. Burdett V
    3. Baitinger C
    4. Iyer RR
    5. Shi K
    6. Modrich P

    (2013) Extrahelical (CAG)/(CTG) triplet repeat elements support proliferating cell nuclear antigen loading and MutLα endonuclease activation

    Proceedings of the National Academy of Sciences of the United States of America 110:12277–12282.

    https://doi.org/10.1073/pnas.1311325110

    • Google Scholar
61. 1. Pluciennik A
    2. Dzantiev L
    3. Iyer RR
    4. Constantin N
    5. Kadyrov FA
    6. Modrich P

    (2010) PCNA function in the activation and strand direction of MutLα endonuclease in mismatch repair

    Proceedings of the National Academy of Sciences of the United States of America 107:16066–16071.

    https://doi.org/10.1073/pnas.1010662107

    • Google Scholar
62. 1. Qiu R
    2. Sakato M
    3. Sacho EJ
    4. Wilkins H
    5. Zhang X
    6. Modrich P
    7. Hingorani MM
    8. Erie DA
    9. Weninger KR

    (2015) MutL traps MutS at a DNA mismatch

    Proceedings of the National Academy of Sciences of the United States of America 112:10914–10919.

    https://doi.org/10.1073/pnas.1505655112

    • Google Scholar
63. 1. Räschle M
    2. Knipscheer P
    3. Knipsheer P
    4. Enoiu M
    5. Angelov T
    6. Sun J
    7. Griffith JD
    8. Ellenberger TE
    9. Schärer OD
    10. Walter JC

    (2008) Mechanism of replication-coupled DNA interstrand crosslink repair

    Cell 134:969–980.

    https://doi.org/10.1016/j.cell.2008.08.030

    • Google Scholar
64. 1. Repmann S
    2. Olivera-Harris M
    3. Jiricny J

    (2015) Influence of oxidized purine processing on strand directionality of mismatch repair

    The Journal of Biological Chemistry 290:9986–9999.

    https://doi.org/10.1074/jbc.M114.629907

    • Google Scholar
65. 1. Schöpf B
    2. Bregenhorn S
    3. Quivy JP
    4. Kadyrov FA
    5. Almouzni G
    6. Jiricny J

    (2012) Interplay between mismatch repair and chromatin assembly

    Proceedings of the National Academy of Sciences of the United States of America 109:1895–1900.

    https://doi.org/10.1073/pnas.1106696109

    • Google Scholar
66. 1. Shell SS
    2. Putnam CD
    3. Kolodner RD

    (2007) The N terminus of Saccharomyces cerevisiae Msh6 is an unstructured tether to PCNA

    Molecular Cell 26:565–578.

    https://doi.org/10.1016/j.molcel.2007.04.024

    • Google Scholar
67. 1. Shiomi Y
    2. Nishitani H

    (2013) Alternative replication factor C protein, Elg1, maintains chromosome stability by regulating PCNA levels on chromatin

    Genes to Cells : Devoted to Molecular & Cellular Mechanisms 18:946–959.

    https://doi.org/10.1111/gtc.12087

    • Google Scholar
68. 1. Shiomi Y
    2. Shinozaki A
    3. Sugimoto K
    4. Usukura J
    5. Obuse C
    6. Tsurimoto T

    (2004) The reconstituted human Chl12-RFC complex functions as a second PCNA loader

    Genes to Cells : Devoted to Molecular & Cellular Mechanisms 9:279–290.

    https://doi.org/10.1111/j.1356-9597.2004.00724.x

    • Google Scholar
69. 1. Thomas DC
    2. Roberts JD
    3. Kunkel TA

    (1991)

    Heteroduplex repair in extracts of human HeLa cells

    The Journal of Biological Chemistry 266:3744–3751.

    • Google Scholar
70. 1. Umar A
    2. Buermeyer AB
    3. Simon JA
    4. Thomas DC
    5. Clark AB
    6. Liskay RM
    7. Kunkel TA

    (1996) Requirement for PCNA in DNA mismatch repair at a step preceding DNA resynthesis

    Cell 87:65–73.

    https://doi.org/10.1016/S0092-8674(00)81323-9

    • Google Scholar
71. 1. Varlet I
    2. Radman M
    3. Brooks P

    (1990) DNA mismatch repair in Xenopus egg extracts: repair efficiency and DNA repair synthesis for all single base-pair mismatches

    Proceedings of the National Academy of Sciences of the United States of America 87:7883–7887.

    https://doi.org/10.1073/pnas.87.20.7883

    • Google Scholar
72. 1. Varlet I
    2. Canard B
    3. Brooks P
    4. Cerovic G
    5. Radman M

    (1996) Mismatch repair in Xenopus egg extracts: DNA strand breaks act as signals rather than excision points

    Proceedings of the National Academy of Sciences of the United States of America 93:10156–10161.

    https://doi.org/10.1073/pnas.93.19.10156

    • Google Scholar
73. 1. Vashee S
    2. Cvetic C
    3. Lu W
    4. Simancek P
    5. Kelly TJ
    6. Walter JC

    (2003) Sequence-independent DNA binding and replication initiation by the human origin recognition complex

    Genes & Development 17:1894–1908.

    https://doi.org/10.1101/gad.1084203

    • Google Scholar
74. 1. Wagner R
    2. Meselson M

    (1976) Repair tracts in mismatched DNA heteroduplexes

    Proceedings of the National Academy of Sciences of the United States of America 73:4135–4139.

    https://doi.org/10.1073/pnas.73.11.4135

    • Google Scholar
75. 1. Walter J
    2. Sun L
    3. Newport J

    (1998) Regulated chromosomal DNA replication in the absence of a nucleus

    Molecular Cell 1:519–529.

    https://doi.org/10.1016/S1097-2765(00)80052-0

    • Google Scholar
76. 1. Yu C
    2. Gan H
    3. Han J
    4. Zhou ZX
    5. Jia S
    6. Chabes A
    7. Farrugia G
    8. Ordog T
    9. Zhang Z

    (2014) Strand-specific analysis shows protein binding at replication forks and PCNA unloading from lagging strands when forks stall

    Molecular Cell 56:551–563.

    https://doi.org/10.1016/j.molcel.2014.09.017

    • Google Scholar
77. 1. Zhang Y
    2. Yuan F
    3. Presnell SR
    4. Tian K
    5. Gao Y
    6. Tomkinson AE
    7. Gu L
    8. Li GM

    (2005) Reconstitution of 5'-directed human mismatch repair in a purified system

    Cell 122:693–705.

    https://doi.org/10.1016/j.cell.2005.06.027

    • Google Scholar

Author details

1. Yoshitaka Kawasoe

   Graduate School of Science, Osaka University, Toyonaka, Japan

   Contribution

   YK, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0003-0925-2004

2. Toshiki Tsurimoto

   Department of Biology, Faculty of Sciences, Kyushu University, Fukuoka, Japan

   Contribution

   TT, Drafting or revising the article, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0001-7597-2216

3. Takuro Nakagawa

   Graduate School of Science, Osaka University, Toyonaka, Japan

   Contribution

   TN, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0003-3455-8224

4. Hisao Masukata

   Graduate School of Science, Osaka University, Toyonaka, Japan

   Contribution

   HM, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

5. Tatsuro S Takahashi

   Graduate School of Science, Osaka University, Toyonaka, Japan

   Contribution

   TST, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   tatsuro_takahashi@bio.sci.osaka-u.ac.jp

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-1947-7680

• 3,489

  views

• 540

  downloads

• 38

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.