Pre‐initiation complex assembly functions as a molecular switch that splits the Mcm2‐7 double hexamer
2017; Springer Nature; Volume: 18; Issue: 10 Linguagem: Inglês
10.15252/embr.201744206
ISSN1469-3178
AutoresMayumi Miyazawa‐Onami, Hiroyuki Araki, Seiji Tanaka,
Tópico(s)Molecular Biology Techniques and Applications
ResumoScientific Report17 August 2017free access Transparent process Pre-initiation complex assembly functions as a molecular switch that splits the Mcm2-7 double hexamer Mayumi Miyazawa-Onami Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan Search for more papers by this author Hiroyuki Araki Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan Department of Genetics, School of Life Science, The Graduate University for Advanced Studies (SOKENDAI), Mishima, Shizuoka, Japan Search for more papers by this author Seiji Tanaka Corresponding Author [email protected] orcid.org/0000-0003-3882-3494 Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan Department of Genetics, School of Life Science, The Graduate University for Advanced Studies (SOKENDAI), Mishima, Shizuoka, Japan School of Environmental Science and Engineering, Kochi University of Technology, Kami, Kochi, Japan Search for more papers by this author Mayumi Miyazawa-Onami Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan Search for more papers by this author Hiroyuki Araki Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan Department of Genetics, School of Life Science, The Graduate University for Advanced Studies (SOKENDAI), Mishima, Shizuoka, Japan Search for more papers by this author Seiji Tanaka Corresponding Author [email protected] orcid.org/0000-0003-3882-3494 Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan Department of Genetics, School of Life Science, The Graduate University for Advanced Studies (SOKENDAI), Mishima, Shizuoka, Japan School of Environmental Science and Engineering, Kochi University of Technology, Kami, Kochi, Japan Search for more papers by this author Author Information Mayumi Miyazawa-Onami1, Hiroyuki Araki1,2 and Seiji Tanaka *,1,2,3 1Division of Microbial Genetics, National Institute of Genetics, Mishima, Shizuoka, Japan 2Department of Genetics, School of Life Science, The Graduate University for Advanced Studies (SOKENDAI), Mishima, Shizuoka, Japan 3School of Environmental Science and Engineering, Kochi University of Technology, Kami, Kochi, Japan *Corresponding author. Tel: +81 887 57 2501; Fax: +81 887 57 2520; E-mail: [email protected] EMBO Rep (2017)18:1752-1761https://doi.org/10.15252/embr.201744206 PDFDownload PDF of article text and main figures.AM PDF Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract Initiation of chromosomal DNA replication in eukaryotes involves two steps: licensing and firing. In licensing, a core component of the replicative helicase, the Mcm2–7 complex, is loaded onto replication origins as an inactive double hexamer, which is activated in the firing step by firing factors. A reaction intermediate called the pre-initiation complex (pre-IC) has been proposed to assemble transiently during firing, but the existence of the pre-IC has not yet been confirmed. Here, we show, by systematic chromatin immunoprecipitation, that a distinct intermediate that fits the definition of the pre-IC assembles during firing in the budding yeast Saccharomyces cerevisiae. Pre-IC assembly is observed in the absence of Mcm10, one of the firing factors, and is mutually dependent on all the firing factors whose association to replication origins is triggered by cyclin-dependent kinase. In the pre-IC, the Mcm2–7 double hexamer is separated into single hexamers, as in the active helicase. Our data indicate that pre-IC assembly functions as an all-or-nothing molecular switch that splits the Mcm2–7 double hexamer. Synopsis This study reveals how the pre-initiation complex (pre-IC) during DNA replication initiation is assembled, and shows that pre-IC assembly separates the core component of the replicative helicase, the Mcm2–7 double hexamer. All firing factors associate with replication origins to assemble pre-IC before Mcm10 functions. Pre-IC assembly is mutually dependent on all the firing factors whose association with replication origins is triggered by CDK. In the pre-IC, not only the formation of tight Cdc45-Mcm2–7-GINS (CMG) complex, but also the separation of Mcm2–7 double hexamer occurs. Introduction In eukaryotes, chromosomal DNA replication initiates from specific regions of chromosomes known as the origins of replication. Replication origins are activated by a two-step reaction to initiate DNA synthesis. In the first step, called licensing, the Mcm2–7 complex, which is the core component of the replicative helicase, is loaded onto replication origins as a double hexamer to form the pre-replicative complex, pre-RC (reviewed in 1). However, the loaded Mcm2–7 complex is inactive at this point. In the second step, which is called initiation, or firing, the Mcm2–7 complex is activated, and bidirectional replication forks including active replicative helicase are formed (reviewed in 2). The active form of replicative helicase consists of at least Cdc45 and the GINS complex (Sld5, Psf1, Psf2, and Psf3) in addition to the Mcm2–7 complex and is called the CMG complex (Cdc45–Mcm2–7–GINS) 3456789. Therefore, the key to understand the firing reaction is to determine how Cdc45 and GINS are integrated with Mcm2–7. In the budding yeast Saccharomyces cerevisiae, all replication factors required for firing have been defined, and the firing reaction has been analyzed both in vivo and in vitro (reviewed in 10). The firing reaction can be divided into three steps, which are regulated by three factors: two conserved protein kinases (Dbf4-dependent kinase, DDK, also known as Cdc7–Dbf4 kinase, and cyclin-dependent kinase, CDK) and Mcm10. In addition to these regulators, other essential replication proteins, such as Cdc45, DNA polymerase ε (Polε), Dpb11, GINS, Sld2, and Sld3, are also required. In the first step, DDK phosphorylates the Mcm4 and Mcm6 subunits of Mcm2–7 1112, promoting interaction between Mcm2–7 and Sld3 13. Sld3 associates with Cdc45, and so DDK promotes Cdc45 loading onto Mcm2–7 in the pre-RC 141516. Next, S-phase-specific CDK (S-CDK) triggers the second step, which defines the timing of origin firing. S-CDK phosphorylates Sld2 and Sld3 171819. Phosphorylated Sld2 interacts with a BRCT-domain-containing protein, Dpb11, and they further interact with Polε and GINS to form the pre-loading complex (pre-LC) 20. Dpb11 also interacts with phosphorylated Sld3 1819. Because Sld3 is loaded onto the pre-RC, GINS in the pre-LC can also be loaded onto Mcm2–7 in the pre-RC. DDK and S-CDK, therefore, enhance the recruitment of Cdc45 and GINS, respectively, onto Mcm2–7 in the pre-RC, to form the CMG complex. In Mcm10-deficient cells, although a tight Cdc45–Mcm2–7–GINS complex is formed, DNA replication does not occur, and association of the single-stranded DNA-binding protein RPA (a hallmark of the unwinding of double-stranded DNA by replicative helicase) is not observed 2122. These observations demonstrate that Mcm10 has a role in the final activation of CMG or the establishment of the replication fork. Although the outline of the firing reaction is known, the precise roles of many of the individual firing factors are still unknown. The protein complex assembled in Mcm10-deficient cells might be identical to the pre-initiation complex (pre-IC), which is proposed to assemble after S-CDK activation and which contains Cdc45 tightly associated with chromatin 23. However, the pre-IC has not been further characterized in detail at a molecular level, although this term is widely used to describe a distinct complex that is assembled during the firing reaction. In vitro reconstitution of DNA replication with purified proteins in budding yeast 24 revealed that all the firing factors (Cdc45, Sld3, Dpb11, Sld2, GINS, and Polε) associate with the replication origin when Mcm10 is omitted from the reaction. This complex fits the definition of the pre-IC, but its existence in vivo and the mode of its assembly are still unknown. We dissected the firing reaction in detail by performing systematic chromatin immunoprecipitation (ChIP) in S. cerevisiae mutants, in which individual firing factors can selectively be depleted. Our results showed that all of the firing factors accumulate on the replication origin in Mcm10-depleted cells, and this accumulation defines the status of the pre-IC. The pre-IC assembly was mutually dependent on all the CDK-dependent factors. Notably, further analysis revealed that the Mcm2–7 double hexamer separates into single hexamers in the pre-IC. Our results show that assembly of the pre-IC acts as a "molecular switch" for origin firing, to increase biological robustness and ensure genome stability. Results and Discussion Individual firing factors can be immunoprecipitated from formaldehyde-cross-linked cell extracts with specific antibodies To capture the behavior of all of the firing factors, we performed ChIP in S. cerevisiae, using antibodies to recognize the individual firing factors Cdc45, Sld3, Dpb11, Sld2, Dpb2 (the second largest subunit of Polε), GINS, and Mcm10 16. As a first step, we confirmed that all antibodies precipitated their target proteins from whole-cell extracts prepared by following the ChIP protocol (Fig EV1A), although the efficiencies of immunoprecipitation varied between antibodies. Click here to expand this figure. Figure EV1. Immunoprecipitation of firing factors and characterization of degron mutants A. Whole-cell extracts (WCE: W) were prepared from formaldehyde-cross-linked, asynchronous wild-type (YST2199) cells. WCE were immunoprecipitated with the indicated antibodies, and immunoprecipitates (IP) and remaining supernatants (S) were analyzed by Western blotting. Concentration ratios of WCE, IP, and S were 1:10:1. * indicates a nonspecific band. B. Wild-type (Wt; YST1351) and cdc45-AID (YST1400) cells were serially diluted 10 times from ˜107 cells/ml, and spotted and grown on the indicated plates for 2 days at 30°C. NAA, 1-naphthaleneacetic acid; YPAGal, YPA medium containing 2% galactose. C, D. Wild-type (Wt at the top; YST2199), sld3-iAID (C) (YST2204), dpb11-iAID (N) (YST2211), sld2-iAID (C) (YST2359), dpb2-iAID (C) (YST2393), sld5-iAID (C) (YST2423), mcm10-iAID (C) (YST2400), pol1-iAID (C) (YST2395), pol3-iAID (C) (YST2397), wild-type (Wt, second from the bottom; YST2363), and pol2-iAID (C) (YST2418) cells were serially diluted and grown on the indicated plates for 2 days at 30°C, as in (B). Dox, doxycycline. E. cdc45-AID (YST1400) cells were grown in YPA-raffinose and collected at the indicated times (0–3 h) after the addition of 2% galactose and 1 mM NAA. * indicates a nonspecific band. F–N. The above-mentioned mutant cells were grown without doxycycline (panel G, and "From Dox = 0" in other panels) or with low-level doxycycline (the concentration of Dox in each case is indicated as "From Dox=" in μg/ml) and collected at the indicated times (0–3 h) after the addition of 20 μg/ml Dox and 1 mM NAA. The samples were analyzed by Western blotting with the indicated antibodies. * indicates a nonspecific band. O. WCE prepared from wild-type yeast cells and purified GINS complex were analyzed with antibodies against GINS. * indicates a nonspecific band. P, Q. Indicated cells were grown, collected and whole-cell extracts were prepared. Samples were analyzed with antibodies against Mcm4 (P) or Mcm7 (Q). Download figure Download PowerPoint Construction and characterization of firing-factor mutants Tight mutants of individual firing factors were generated by a degron-fusion technique, to eliminate specific gene products through targeted ubiquitin-dependent proteolysis. Detailed construction and characterization are described in Materials and Methods and Fig EV1B–N. All mutants used in this study (cdc45-AID, sld3-iAID (C), dpb11-iAID (N), sld2-iAID (C), dpb2-iAID (C), pol2-iAID (C), sld5-iAID (C), and mcm10-iAID (C)) showed tight cell-cycle arrest and did not enter S phase in the single-cell-cycle analysis (Fig 1). Figure 1. Systematic chromatin immunoprecipitation (ChIP) of the firing factors A–K. Wild-type (Wt, YST2199), sld3-iAID (C) (YST2204), cdc45-AID (YST1400), dpb11-iAID (N) (YST2211), sld2-iAID (C) (YST2359), sld5-iAID (C) (YST2423), dpb2-iAID (C) (YST2393), pol2-iAID (C) (YST2418), mcm10-iAID (C) (YST2400), pol1-iAID (C) (YST2395), and pol3-iAID (C) (YST2397) cells were grown as indicated in Fig EV2, and samples were collected and analyzed by flow cytometry (shown on the left of each panel) and ChIP (shown on the right of each panel, with % recovery of the early firing origin, ARS305), which was performed for the indicated firing factors with specific antibodies, as described in Materials and Methods. The same colors in flow cytometry and ChIP analyses represent samples taken at the same time points. Values shown are the average of two independent immunoprecipitation results, each of which is shown as an open circle and was obtained as the average of duplicated measurements. Download figure Download PowerPoint Systematic ChIP of firing factors in a wild-type strain To determine the behavior of each firing factor, systematic ChIP was performed with antibodies to Cdc45, Sld3, Dpb11, Sld2, Dpb2, GINS, and Mcm10 (Fig 1). In wild-type cells, cell-cycle progression was not affected by the addition of doxycycline and the synthetic auxin, 1-naphthaleneacetic acid (NAA), which induces depletion of mutant protein in AID or iAID mutants (Fig 1A). Cells were collected for ChIP 0, 15, and 30 min after release, corresponding to G1, early S phase, and the end of S phase, respectively, and the association of firing factors with an early replication origin, ARS305, was monitored (Fig 1A). As reported previously 14151625, association of Sld3 and Cdc45 with early replication origins was observed in G1 and in early S phase, after which the association disappeared by the end of S phase (Fig 1A). Other firing factors that can assemble in the pre-LC by S-CDK activation (Dpb11, Sld2, Dpb2, and Psf2) only associated with ARS305 in early S phase, as did Mcm10 (Fig 1A). Addition of NAA and doxycycline did not affect assembly of firing factors in the wild type (Figs 1A and EV2A). These results demonstrated that ChIP can systematically trace the behavior of firing factors in vivo. Click here to expand this figure. Figure EV2. Systematic ChIP of the firing factors A. Systematic ChIP of firing factors in wild-type cells (YST2199) grown in the permissive condition (without doxycycline (Dox) or 1-naphthaleneacetic acid (NAA)), in comparison with growth in restrictive conditions, as shown in Fig 1A. B. The growth condition for YST1400 (cdc45-AID) cells in the ChIP experiment shown in Fig 1C. C. Growth conditions for sld3-iAID (C) (YST2204), dpb11-iAID (N) (YST2211), sld2-iAID (C) (YST2359), sld5-iAID (C) (YST2423), dpb2-iAID (C) (YST2393), pol2-iAID (C) (YST2418), mcm10-iAID (C) (YST2400), pol1-iAID (C) (YST2395), and pol3-iAID (C) (YST2397) cells in the ChIP experiments shown in Fig 1B and D–K. D–O. Measurement of ARS306 recovery in systematic ChIP of firing factors in wild-type cells and degron mutants shown in Fig 1. Data information: Values shown are the average of two independent immunoprecipitation results, each of which is shown as an open circle and was obtained as the average of duplicated measurements. Download figure Download PowerPoint Firing factors do not assemble in sld3 or cdc45 mutants To analyze the first step of the firing reaction, cdc45-AID and sld3-iAID (C) cells grown in restrictive conditions were collected at times that corresponded to G1, early S phase, and late S phase under permissive conditions. ChIP demonstrated that none of the firing factors tested accumulated on ARS305, in either cdc45-AID or sld3-iAID (C). Cdc45 and Sld3 association with ARS305 was mutually dependent (Fig 1B and C), as previously suggested by ChIP with temperature-sensitive cdc45 and sld3 mutants 14. These results partly disagree with previous results obtained from the in vitro reconstitution system, in which Cdc45 recruitment to the pre-RC is dependent on Sld3, but Sld3 associates with the pre-RC in the absence of Cdc45 13. The reason for this discrepancy is not known and should be elucidated in future studies. Switch-like, mutually dependent association of pre-LC factors with replication origins The second step of firing is the recruitment of GINS to replication origins, which is triggered by S-CDK and requires all of the pre-LC factors, such as Dpb11, Sld2, Polε, and GINS. Previous studies have shown by ChIP that association of GINS and Dpb11 with replication origins is mutually dependent 26, and Dpb11 has a GINS-interaction domain, which is required for efficient DNA replication 27. All the factors in the pre-LC can interact with each other, and so they might co-associate at replication origins after S-CDK activation. To dissect this S-CDK-dependent step, association of firing factors with ARS305 was monitored in dpb11-iAID (N), sld2-iAID (C), sld5-iAID (C), dpb2-iAID (C), and pol2-iAID (C) (Figs 1D–K and EV2C). Under restrictive conditions, dpb11-iAID (N) cells did not enter S phase, and only Sld3 and Cdc45 associated with ARS305 (Fig 1D), reflecting their association with early replication origins prior to S-CDK activation. The other pre-LC factors did not associate with ARS305 (Fig 1D), indicating dependence on Dpb11. Notably, similar results were observed with sld2-iAID (C), sld5-iAID (C), and dpb2-iAID (C) (Fig 1E–G). In these mutants, depletion of specific replication factors did not affect the expression levels of all other firing factors (Fig EV3), with the exception that expression levels of GINS subunits Psf1 and Psf2 were lowered in sld5-iAID (C) (Fig EV3, lanes 31–33). Expression levels of firing factors other than GINS were unaffected. Therefore, low ChIP signals observed in the mutants were caused by the nonassociation of firing factors, rather than the absence of the corresponding factors. These results indicated that the association of the pre-LC factors with replication origins is mutually dependent and occurs as an "on–off" switch after CDK activation, rather than in a stepwise manner. Click here to expand this figure. Figure EV3. Depletion of a firing factor does not affect the expression of other firing factorsWild-type (YST2199), sld3-iAID (C) (YST2204), dpb11-iAID (N) (YST2211), sld2-iAID (C) (YST2359), dpb2-iAID (C) (YST2393), pol1-iAID (C) (YST2395), pol3-iAID (C) (YST2397), mcm10-iAID (C) (YST2400), pol2-iAID (C) (YST2418), and sld5-iAID (C) (YST2423) cells were grown and collected as in Fig EV1F–N ("From Dox = 0"). cdc45-AID (YST1400) cells were grown and collected as in Fig EV1E. The samples were analyzed by Western blotting with the indicated antibodies. * indicates a nonspecific band. Download figure Download PowerPoint Two essential subunits of Polε, Pol2 and Dpb2, are required for formation of the CMG complex Notably, different binding patterns of firing factors were observed in mutants of the two essential subunits of Polε: dpb2-iAID (C) and pol2-iAID (C) (Fig 1G and H). Association of pre-LC factors with ARS305 was absent in dpb2-iAID (C), as in the case of other pre-LC-factor mutants (Fig 1G). By contrast, in pol2-iAID (C), the association of firing factors with ARS305 was observed at the later time points (Fig 1H). These results suggested the distinct roles of Pol2 and Dpb2 in the firing reaction. Previously reported results show that Dpb2 depletion results in the absence of replisome assembly 2028, and so this deficiency is very likely to be a direct result of the lack of association of pre-LC factors with the replication origin, as observed in dpb2-iIAD (C). By contrast, the exact role of Pol2 in the firing reaction has not been demonstrated previously. In budding yeast, although the amino-terminal portion of Pol2 (which possesses the catalytic activity of Polε) is dispensable 2930, Pol2 is essential for growth. Therefore, we further analyzed the firing reaction in the pol2-iAID (C) mutant in two ways: by immunoprecipitation, to determine whether the CMG complex forms; and by Rfa1 ChIP, to see whether a replication fork is established and origin DNA is unwound, because Rfa1 is a subunit of RPA. For immunoprecipitation, the Psf2 subunit of GINS was tagged to recover the CMG complex. In the wild type, Mcm7 and Cdc45 proteins were co-purified with Psf2-Flag only when the cell extracts were prepared from cells in S phase (20 min after release: Fig 2A and C, lanes 13–16), indicating that the CMG complex was specifically recovered. By contrast, neither Mcm7 nor Cdc45 was co-purified with Psf2-Flag from any extracts prepared from dpb2-iAID (C) and pol2-iAID (C) cells (Fig 2C, lanes 17–24). Therefore, the CMG complex did not form in pol2-iAID (C) cells or dpb2-iAID (C) cells. In Rfa1-ChIP, substantial association of Rfa1 with ARS305 was not observed in pol2-iAID (C) or dpb2-iAID (C), whereas Rfa1 was specifically associated with ARS305 in the wild-type cells in early S phase (Fig 2D and E). Figure 2. CMG assembly and DNA unwinding are absent in Pol2-depleted cells A, B. Flow cytometry profiles of cells used to prepare whole-cell extracts (WCE) for the experiment shown in panel (C). Wild-type (Wt, YST2591), dpb2-iAID (C) (YST3080), and pol2-iAID (C) (YST3085) cells were grown as in Fig 1A, or arrested in G2/M with nocodazole (B), and samples were collected. Time points in which cells were collected for the preparation of WCE are colored in yellow. C. WCE were prepared from cells collected in (A and B), and 3 × FLAG-1 × HA-tagged Psf2 was pulled down with anti-FLAG beads and eluted with the FLAG peptide (Flag). Each sample was analyzed with antibodies against Flag, Mcm7, and Cdc45. N in time indicates nocodazole arrest. The concentration ratio of WCE to FLAG was 1:4. D. Flow cytometry profiles of cells used to prepare WCE for the experiment shown in panel (E). Wild-type (Wt, YST3072), mcm10-iAID (C) (YST3074), dpb2-iAID (C) (YST3076), and pol2-iAID (C) (YST3078) cells were grown as in Fig 1, and samples were collected and analyzed by flow cytometry. Time points in which cells were collected for the preparation of ChIP samples are colored in yellow. E. WCE were prepared from samples collected in (D), and ChIP of 3 × Flag-1 × HA-tagged Rfa1 was performed with or without anti-Flag antibodies, as described in Materials and Methods. Values shown are the average of two independent immunoprecipitation results, each of which is shown as an open circle and was obtained as the average of duplicated measurements. Download figure Download PowerPoint Taken together, our results indicated that, in Pol2-depleted cells, firing factors, including Cdc45 and GINS, can associate with the replication origin, but as the CMG complex is not formed, origin unwinding does not occur, suggesting that Pol2, and especially its essential carboxyl-terminal portion, may have a role in the formation of the complex including CMG. In the fission yeast Schizosaccharomyces pombe, Pol2 is also essential for cell growth, and in ChIP experiments, the association of Cdc45 and GINS with replication origins occurs in a Pol2 carboxyl-terminal mutant (cdc20-ct1) in which the interaction between Dpb2 and Pol2 is weakened. However, Pol2 depletion blocks origin association of Cdc45 and GINS 31. Therefore, although the association pattern of Cdc45 and GINS in Pol2 depletion in S. pombe seems different from that in S. cerevisiae, testing CMG formation in the S. pombe cdc20-ct1 mutant could provide interesting results. Mcm10 is not required for assembly of the complex Results of previous studies in vivo and in vitro indicate that Mcm10 has a late role in the firing reaction in yeasts, in the step between the recruitment of GINS by CDK activation and establishment of the replication fork 15212232. Accordingly, in mcm10-iAID (C) cells, Cdc45, Sld3, GINS, Dpb11, Sld2, and Dpb2 were all associated with ARS305 (Fig 1I). Notably, ChIP signals for Cdc45, GINS, and Dpb2, which travel with the replication fork, were very high (Fig 1I). The protein-binding pattern in Mcm10-depleted cells was reminiscent of the pre-IC, the complex that is proposed to form just before the initiation of DNA replication, as a prerequisite to activation of the replicative helicase 23. However, details of the assembly and composition of the pre-IC have not previously been determined. Cdc45, Mcm2–7, and GINS have been shown to associate tightly in mcm10 mutants, as in the CMG 2122. This complex formation was also observed in our mcm10-iAID (C) mutant, but not in dpb11-iAID (N) or dpb2-iAID (C) (Fig EV4). Therefore, the protein complex that assembles at replication origins in mcm10-iAID (C) cells seems to correspond to the proposed pre-IC. Click here to expand this figure. Figure EV4. Assembly of the CMG complex in firing-factor mutants Flow cytometry profiles of cells used to prepare whole-cell extracts (WCE) for the experiment shown in panel (B). Wild-type (Wt, YST2668), dpb11-iAID (N) (YST2534), dpb2-iAID (C) (YST2597), pol1-iAID (C) (YST2598), pol3-iAID (C) (YST2847), and mcm10-iAID (C) (YST2848) cells were grown as in Fig 1 and collected. WCE were prepared from cells indicated in panel (A). 3 × FLAG-1 × HA-tagged Psf2 was pulled down with control beads or anti-FLAG beads and eluted with the FLAG peptide (shown as "no Ab" and "α-Flag", respectively). Each of the eluates was analyzed with antibodies against Flag, Mcm4, and Cdc45. Long-exposure and short-exposure images are shown for anti-Mcm4. Concentration ratios of WCE to Beads pull-down is 1:4. Download figure Download PowerPoint After the firing reaction, the replication fork is established, and nascent DNA strands are synthesized by the cooperative action of DNA polymerases, Polα, Polδ, and Polε. Polα and Polδ associate with GINS as components of active replication forks, but not in Mcm10-depleted cells 22, and so this association might occur after activation of the CMG helicase. To analyze the roles of POL1 and POL3, which encode the essential catalytic subunits of Polα and Polδ, respectively, pol1-iAID (C) and pol3-iAID (C) mutants were constructed. Association of firing factors with ARS305 in these mutants was similar to that in wild-type cells (Fig 1A, J, and K). Dissociation of firing factors from ARS305 in pol3-iAID (C) cells was also similar to that in wild-type cells. By contrast, factors involved in active replication forks (Cdc45, Dpb2, and GINS) remained in association with ARS305 even at the later time point in pol1-iAID (C) cells (Fig 1J), which suggested that CMG forms in pol1 mutants, but the replication fork stalls very close to the replication origin because of the lack of DNA synthesis. This idea is supported by immunoprecipitation, in which the formation of the CMG complex was observed in pol1-iAID (C) and pol3-iAID (C) (Fig EV4). The above-mentioned binding patterns of firing factors to ARS305 in mutant cells were mostly recapitulated at another early firing origin, ARS306 (Fig EV2D–O), although the ChIP signals with Dpb11 antibodies were nonspecifically increased in some mutants, for unknown reasons. Therefore, the mode of assembly of the pre-IC at replication origins might be governed by a common mechanism. Although the Mcm10-dependent reaction is the last step for the formation of the active replicative helicase, Mcm10 is constitutively expressed throughout the cell cycle 33. The Mcm10-dependent process is very likely to occur automatically after formation of the pre-IC, suggesting that pre-IC assembly is the last step in the firing reaction that is subject to regulation. The mutual dependency of pre-IC assembly not only functions as a molecular switch, but might also increase the biological robustness of the firing reaction by ensuring the readiness of all pre-IC factors, to maintain genome integrity. Mcm2–7 double hexamer dissociates in Mcm10-depleted cells Mcm2–7 is loaded at the replication origin as a head-to-head double hexamer in the pre-RC 34. This double hexamer separates during helicase activation. In the pre-IC in Mcm10-depleted cells, Cdc45, Mcm2–7, and GINS form a complex like the CMG in an active replication fork. To determine whether Mcm2–7 in the pre-IC is in the double-hexamer form, sequential immunoprecipitation was performed in conditions suitable for recovery of the pre-RC 322. First, we confirmed that we could recover Mcm2–7 double hexamer in the pre-RC. For this experiment, we created cells expressing both Myc-tagged and untagged (normal) Mcm7, as well as a sole copy of Mcm4 tagged with Flag and HA. Cells in G1 phase, S phase, and G2/M phase were collected (Fig EV5A), and whole-cell extracts were prepared. In the first purification with Flag tag, both Myc-tagged and untagged Mcm7 were co-purified with Mcm4-FlagHA (Fig EV5B, lanes 4–6). To confirm that the Mcm2–7 double hexamer was recovered, Myc pull-down was performed. The Myc pull-down fractions recovered most of the Myc-tagged Mcm7, and untagged Mcm7 was co-recovered in fractions derived from G1 phase and S phase, but not from G2/M phase (Fig EV5B, lanes 7–9). Notably, the amount of co-recovered, untagged Mcm7 was lower in the S-phase-derived fraction than in the G1-phase-derived fraction (Fig EV5B, lanes 7 and 8), suggesting conversion of the Mcm2–7 double hexamer to the CMG during S phase. Cdc45 and the GINS subunits, Psf1 and Psf2, also co-pre
Referência(s)