Synergism between CMG helicase and leading strand DNA polymerase at replication fork

Abstract

The replisome that replicates the eukaryotic genome consists of at least three engines: the Cdc45-MCM-GINS (CMG) helicase that separates duplex DNA at the replication fork and two DNA polymerases, one on each strand, that replicate the unwound DNA. Here, we determined a series of cryo-electron microscopy structures of a yeast replisome comprising CMG, leading-strand polymerase Polε and three accessory factors on a forked DNA. In these structures, Polε engages or disengages with the motor domains of the CMG by occupying two alternative positions, which closely correlate with the rotational movement of the single-stranded DNA around the MCM pore. During this process, the polymerase remains stably coupled to the helicase using Psf1 as a hinge. This synergism is modulated by a concerted rearrangement of ATPase sites to drive DNA translocation. The Polε-MCM coupling is not only required for CMG formation to initiate DNA replication but also facilitates the leading-strand DNA synthesis mediated by Polε. Our study elucidates a mechanism intrinsic to the replisome that coordinates the activities of CMG and Polε to negotiate any roadblocks, DNA damage, and epigenetic marks encountered during translocation along replication forks.

Fig. 1. Detailed interactions of Polε with CMG helicase.

a–d Side views of the cryo-EM density map of the replisome, showing the docking sites of Polε on CMG (dashed squares). For clarity, Mcm6/4/7 subunits and Tof1-Csm3 are removed in (d), highlighting the interaction of Polε with Mcm5. Relevant subunits are color-coded and labeled as indicated. Measured buried surface areas for each interface in (a–d) are labeled. BSA: buried surface area. e Schematic domain organization of Dpb2. f Interaction between Dpb2 and Psf1 with Dpb2-NTD shown in cartoon presentation and GINS displayed with its cryo-EM density map. The α helices and connecting loops of Dpb2-NTD are labeled as indicated. α: helix, L: loop; NT: N-terminus; CT: C-terminus. g, h Magnified views of the boxed region in (f), showing the detailed interaction between Dpb2-NTD and Psf1 B domain in cartoon presentation. The hydrogen bonds are indicated by dashed lines in blue; Salt bridge between E177 of Psf1 and R27 of Dpb2 is indicated by the dashed line in green. i, j Density map of Dpb2 showing the interaction between the NTD and PDE domain of Dpb2. k Atomic model of Dpb2 with α4 contacting loop and helix highlighted in green and red, respectively. l Schematic domain organization of Pol2. Dashed lines denote highly disordered segments that are not resolved in the replisome structure. m Magnified view of the boxed region in (b) highlighting the interaction details of Pol2 with the AAA+ domain of Mcm2 in cartoon presentation. n Magnified view of the boxed region in (b) illustrating the interaction of Pol2 with Cdc45. o Magnified view of the boxed region in (c) showing the Dpb2 OB domain contacting the CTD interface of Mcm3:5. Hydrogen bond between R277 of Dpb2 and E563 of Mcm3 is indicated by the blue dashed line. p Interaction between Pol2-ZF2 (yellow) and Mcm5-CTD (blue). q Magnified view of the boxed region in (p). The hydrogen bond between R2166 of Pol2 and D383 of Mcm5 is indicated by the red dashed line. Zn²⁺ atom is shown as a gray sphere. r Pol2 (yellow) docking to Mcm5-WHD (blue). Cation-π interaction between R764 of Mcm5 and W1711 of Pol2 is indicated by the dashed line in red. s Same as (r) but shown with the electrostatic surface potential map of Mcm5-WHD, highlighting Mcm5-WHD binding to a region of Pol2 rich in hydrophobic residues. Positively charged residues are in blue; negatively charged residues in red; nonpolar and hydrophobic residues in white. t A replisome conformer exhibiting a highly flexible Mcm5-WHD.

Fig. 2. The physiological function of Polε-MCM coupling in replication initiation and S phase progression.

a In vitro CMG pulldown assays to examine the binding affinity of Polε, PolεΔ2N, and the related mutants (Pol2ΔDS2 and Pol2ΔDS2+4) to CMG in the presence or absence of DNA. Similar results were obtained in two independent experiments. b Schematic illustration of pol2 mutant construction. c Polε-MCM coupling is essential for cell viability. The indicated truncation mutants of Pol2 were tested for complementation of a depletion of Pol2-AID, the expression of which is under the control of Tet promoter. YPD: yeast extract-peptone-dextrose, NAA: 1-naphthalene acetic acid, Dox: doxycycline. d Flow cytometry profiles of cells collected from the α factor block-and-release assays with the indicated pol2-iAID strains under restrictive conditions. e The lysates of the indicated samples were prepared for Flag IP with anti-FLAG beads. Each sample was analyzed with antibodies against Flag, Cdc45, and Mcm2. Similar results were obtained in two independent experiments. Source data are provided as a source data file. f Flow cytometry profiles of cells collected from the hydroxyurea (HU) block-and-release assays. α factor was added after cells were released from HU arrest.

Fig. 3. Polε stability in the replisome.

a–d Four different conformations of Polε in the replisome. e–h Same as (a–d) but with 90° rotation. Replisome components are colored as indicated.

Fig. 4. Polε stability and rotational movement of DNA around MCM pore.

Comparison of the leading-strand DNA conformations around MCM pore from different states of the replisome: State I (a, b), State II (c, d), State III (e, f), State IV (g, h), State V (i, j), and State VI (k, l). Relevant density maps (a, c, e, g, i, and k) shown with MCM subunits free of DNA engagement colored in gray. b, d, f, h, j, and l Bottom CTD views of the leading-strand replisome with DNA and Polε in different conformational states (I-V) shown with the atomic models. State VI with no DNA bound in MCM pore.

Fig. 5. The conformational change in the MCM motor domains regulates Polε docking site formation.

Structural comparison of replisome in State I and State V, showing large displacement in relevant Pol2 (a), Dpb2-OB (b), and Pol2-ZF2 (c) contacting α helices from Mcm2 (a), Mcm3:5 interface (b) and Mcm5-CTD (c).

Fig. 6. The effect of Polε docking to MCM ring on CMG helicase and leading strand DNA synthesis in vitro.

a Schematic illustration of the unwinding of a 20-bp forked DNA by CMG. b, c Native PAGE of CMG helicase assays performed as in (a) in the absence (b) or presence (c) of Polε. d Quantification of the results; assays were performed in triplicate. Data are mean ± SEM (n = 3). Source data are provided as a source data file. e Scheme of reconstituting leading-strand DNA synthesis by Polε (see details in “Methods”). f Purified factors used for DNA replication assays analyzed by SDS-PAGE with Coomassie staining. g Alkaline agarose gel of reaction products from the replication assays performed as in (e). Similar results were obtained in two independent experiments.

Fig. 7. Model illustrating potential functions of Polε cycling on and off MCM ring.

a Polε binding to the MCM ring facilitates its engagement with the PCNA-DNA complex to drive strand extension. b DNA translocating away the Mcm2:5 interface releases Polε from MCM. c Polε flips over CMG helicase to target parental histones via Dpb3-4 for redisposition of epigenetic information from parental DNA strand to newly synthesized leading strand. d Upon DNA damage, a flexible Polε can provide access to other polymerases for damage repair and/or re-priming for fork restart. e Leading-strand DNA relocating to the Mcm2:5 interface enables Polε binding to MCM to re-establish strand extension. f DNA translocation by CMG initiates a new round of Polε cycling on and off the MCM ring.