Competition for the nascent leading strand shapes the requirements for PCNA loading in the replisome

Abstract

During DNA replication, the DNA polymerases Pol δ and Pol ε utilise the ring-shaped sliding clamp PCNA to enhance their processivity. PCNA loading onto DNA is accomplished by the clamp loaders RFC and Ctf18-RFC, which function primarily on the lagging and the leading strand, respectively. RFC activity is essential for lagging-strand replication by Pol δ, but it is unclear why Ctf18-RFC is required for leading-strand PCNA loading and why RFC cannot fulfil this function. Here, we show that RFC cannot load PCNA once Pol ε has been incorporated into the budding yeast replisome and commenced leading-strand synthesis, and this state is maintained during replisome progression. By contrast, we find that Ctf18-RFC is uniquely equipped to load PCNA onto the leading strand and show that this activity requires a direct interaction between Ctf18 and the CMG (Cdc45-MCM-GINS) helicase. Our work uncovers a mechanistic basis for why replisomes require a dedicated leading-strand clamp loader.

Keywords: CMG Helicase; Clamp Loader; Ctf18-RFC; PCNA Loading; Replisome.

Figure 1. Ctf18-RFC accelerates replication in the regulated system.

(A) Schematic of the regulated in vitro DNA replication system. (B) Denaturing agarose gel analysis of replication reactions using the regulated system as in (A) in the absence or presence of Ctf18-RFC, with RFC present throughout. (C) Lane profiles of the 4 min timepoints from reactions in the absence and presence of Ctf18-RFC as in (B). (D) Quantification of maximal leading-strand synthesis rates from replication reactions in the regulated system as in (B). Linear regression is fit to the mean of three experiments. The error bars represent the standard error of the mean (s.e.m.) and the mean is indicated by filled circles. (E) Denaturing agarose gel analysis of replication reactions using the regulated system with wild-type (WT) or mutant Ctf18-RFC complexes present where indicated. Reactions were analysed after 4 min. In this and subsequent figures, CMG-independent template labelling is indicated. (F) Denaturing agarose gel analysis of replication reactions using the regulated system in the absence or presence of RFC and Ctf18-RFC. Source data are available online for this figure.

Figure 2. Ctf18-RFC, but not RFC, supports PCNA loading at preassembled CMGE.

(A) Schematic of the CMG-based in vitro DNA replication system. (B) Denaturing agarose gel analysis of replication reactions using the CMG-based system as in (A), in the absence or presence of PCNA, RFC, and Ctf18-RFC. Reactions were analysed after 3 min. (C) Denaturing agarose gel analysis of Pol ε primer extension reactions performed as in Fig. EV1B, in the presence of RFC or Ctf18-RFC at the concentrations indicated. Reactions were performed in the presence of 250 mM potassium glutamate and analysed after 6 min. Source data are available online for this figure.

Figure 3. Transient uncoupling promotes PCNA loading by RFC.

(A) Denaturing agarose gel analysis (right) of replication reactions using the CMG-based system under pulse-chase conditions as outlined (left). RFC and Ctf18-RFC were included during the pulse or after the chase where indicated. (B) Schematic of the staged CMG-based DNA replication system. Pol ε is added 1 min after initiation of template unwinding. (C) Denaturing agarose gel analysis of replication reactions using the standard CMG-based system as in Fig. 2A or the staged CMG-based system as in (B). RFC was included from the outset at the concentrations indicated. In this and subsequent staged CMG-based system reactions, timepoints are from initiation of template unwinding. Source data are available online for this figure.

Figure 4. Pol ε functions with RFC-loaded PCNA after switch from Pol δ.

(A) Schematic representing polymerase exchange during replication initiation. Upon collision-release of extending Pol δ with CMGE, PCNA could either be transferred along with the 3’ DNA end (i) or loaded by RFC (ii). (B) Denaturing agarose gel analysis of replication reactions using the staged CMG-based system as in Fig. 3B, but with Pol ε, Pol δ, and Pol α-primase added 1 min after initiation of template unwinding. Ctf18-RFC was present where indicated, with RFC present throughout. (C) Quantification of maximal leading-strand synthesis rates from replication reactions in the staged CMG-based system as in (B). Linear regression is fit to the mean of three experiments. The error bars represent the s.e.m. and the mean is indicated by filled circles. (D) Denaturing agarose gel analysis of replication reactions using the staged CMG-based system with Pol ε, Pol δ and Pol α-primase added 1 min after initiation of template unwinding, in the absence or presence of RFC or Ctf18-RFC. Source data are available online for this figure.

Figure 5. Structural analysis of the budding yeast replisome with and without Ctf18-RFC.

(A) Cryo-EM map of the replisome lacking density corresponding to Ctf18-RFC (EMD-52459). In this reconstruction, the Pol ε catalytic domain occupies a stable position beneath the Mcm2-7 C-tier. The map was generated via homogenous refinement and filtered according to local resolution. The map is coloured according to protein occupancy. (B) Cryo-EM maps of the replisome containing density corresponding to Ctf18-RFC. The individual maps displayed were obtained using local refinement following signal subtraction and filtered according to local resolution. Both maps are displayed on the same origin as the parental consensus refinement and describe the following replisome components respectively: CMG, Ctf4, DNA and Tof1-Csm3 (EMD-52107), and Pol ε and Ctf18-RFC (EMD-52116). The approximate boundary between these local refinements is indicated by a dotted line. Maps are coloured according to protein occupancy. (C) Cryo-EM map of the Pol ε catalytic domain bound by Ctf18, Dcc1, and Ctf8 (EMD-52120). This map was obtained via focused refinement following signal subtraction and sharpened using a B-factor of -300. PDB:6S2E (Stokes et al, 2020) was rigid body docked into the volume and the map coloured according to protein occupancy. (D) Cryo-EM map of Pol ε engaged with the Mcm5 WH domain and Ctf18-RFC (EMD-52116). This map was obtained via focused refinement following signal subtraction and filtered according to local resolution. The map was coloured according to protein occupancy. An AlphaFold 3 prediction for Ctf18-Rfc2-5 was manually docked into the cryo-EM density which remained unmodelled after the docking of Pol ε, the Mcm5 WH, and the Ctf18-1-8 module.

Figure 6. Interaction of Ctf18 with Mcm7 is involved in PCNA loading.

(A) AlphaFold 3 predicted aligned error plot of S. cerevisiae Ctf18, Rfc2-5 and Mcm7 730–845. (B) AlphaFold 3 prediction of S. cerevisiae Ctf18, Rfc2-5 and Mcm7 730–845 corresponding to (A), coloured by subunit. (C) Denaturing agarose gel analysis of replication reactions using the CMG-based system, with wild-type (WT) or mutant CMG and Ctf18-RFC complexes present where indicated. Reactions were analysed after 3 min. Source data are available online for this figure.

Figure EV1. Ctf18-RFC mutants and analysis of Ctf18-RFC in primer extension assays.

(A) Coomassie-stained 4–12% SDS-PAGE gel of purified S. cerevisiae wildtype (WT) or mutant Ctf18-RFC complexes. Subunits are labelled. (B) Schematic of a DNA polymerase primer extension reaction on circular ssDNA. (C) Denaturing agarose gel analysis of primer extension reactions as in (B) using Pol ε in the absence or presence of Ctf18-RFC. Reactions were performed at 100 mM potassium glutamate. (D) Denaturing agarose gel analysis of primer extension reactions as in (B) using Pol δ in the absence or presence of RFC or Ctf18-RFC, analysed after 1.5 min. The ssDNA template was coated with either S. cerevisiae RPA or E. coli SSB. Reactions were performed at 100 mM potassium glutamate. Source data are available online for this figure.

Figure EV2. (Related to Figs. 1 and 2). Analysis of leading-strand synthesis acceleration by Ctf18-RFC in the regulated and CMG-based systems.

(A) Lane profiles of 4 min timepoints from replication reactions using the regulated system, with wildtype (WT) or mutant Ctf18-RFC complexes present where indicated as in Fig. 1E. (B) Lane profiles of 5 min timepoints from replication reactions using the regulated system, in the absence and presence of RFC and Ctf18-RFC as in Fig. 1F. Arrows indicate ‘left’ leading-strand populations. (C) Denaturing agarose gel analysis of replication reactions using the CMG-based system, with wildtype (WT) or mutant Ctf18-RFC complexes present where indicated. Reactions were analysed after 3 min. (D) Denaturing agarose gel analysis of replication reactions using the CMG-based system with Pol ε or Pol ε^PIP, in the absence or presence of Ctf18-RFC. Reactions were analysed after 3 min. Source data are available online for this figure.

Figure EV3. (Related to Figs. 3 and 4). Analysis of reduced dNTP and staged CMG-based system assays.

(A) Denaturing agarose gel analysis of replication reactions using the CMG-based system in the absence or presence of RFC or Ctf18-RFC, with reduced dNTP concentrations as indicated. Reactions were analysed after 2.5 min. (B) Denaturing agarose gel analysis of replication reactions using the staged CMG-based system with Pol ε addition 1 min after initiation of template unwinding. PCNA and RFC were included where indicated. Reactions were analysed 4 min after initiation of template unwinding. (C) Denaturing agarose gel analysis of replication reactions using the staged CMG-based system but with Pol ε and Pol δ/ Pol δ^CAT-DEAD (Pol δ^CD) added 1 min after initiation of template unwinding where indicated. Pol α-primase was included with polymerase addition throughout. Reactions were analysed 5 min after initiation of template unwinding. Source data are available online for this figure.

Figure EV4. (Related to Fig. 5). Cryo-EM analysis of a budding yeast replisome prepared with Ctf18-RFC.

(A) Silver-stained SDS-PAGE gels analysing 100 μl fractions taken across 10-30% glycerol gradients, either in the absence (top) or presence (bottom) of crosslinking agents. Fractions 11-12 used for cryo-EM sample preparation are indicated with a red bar labelled “pooled”. Protein annotations are based on the position of bands in lane 12. (B) Representative cryo-EM micrograph obtained using a K3 direct electron detector (Gatan) at a nominal pixel size of 0.93 Å/pixel. Scale bar: 30 nm (inset). (C) Representative 2D class averages with corresponding particle numbers. Derived from 47,823 particle subset used to obtain EMD maps 52107, 52116, 52120. Mask diameter 500 Å. Obtained using cryoSPARC-3 2D classification. (D–I) Cryo-EM reconstructions obtained using homogeneous or local refinement, coloured according to local resolution according to the key in (D). Local resolution estimation and filtering was performed in cryoSPARC-3. (D) Replisome with Ctf18-RFC bound to the Pol ε catalytic domain. The 47,823 particle subset used for this homogenous refinement was also used to derive maps EMD-52107, EMD-52116 and EMD-52120. (E) Local refinement following particle subtraction showing CMG, Tof1-Csm3, dsDNA, Ctf4 and Dpb2_N-term. EMD-52107, derived from particle subset in (D). (F) Local refinement following particle subtraction, showing Pol ε, Ctf18-RFC and Mcm5 WH. EMD-52116, derived from particle subset in (D). (Top) map at high threshold, (bottom) same map at low threshold highlighting presence of unmodelled density. (G) Local refinement following particle subtraction, showing the Pol ε catalytic domain and Ctf18-RFC. EMD-52120, derived from particle subset in (D). (H) Replisome lacking Ctf18-RFC bound to the Pol ε catalytic domain. EMD-52459, obtained from a homogenous refinement of a 20,462 particle subset. (I) Local refinement following particle subtraction, showing Pol ε and Mcm5 WH. EMD-52505, derived from particle subset in (H). (J) Fourier shell correlation (FSC) graph describing the maps in (D–I). Resolution was calculated using the FSC = 0.143 cut-off with values reported in Appendix Figs. S1–S3. (K, L) Viewing direction plots. 2D-histograms that show the number of particles with a viewing direction at a particular elevation/azimuth bin. (K) Consensus refinement of the replisome with Ctf18-RFC bound, as in (D). (L) Consensus refinement of the replisome without Ctf18-RFC bound, as in (H). Source data are available online for this figure.

Figure EV5. (Related to Fig. 5). Structural comparisons of the replisome bound to Ctf18-RFC (this study) with relevant previously published structures.

(A) Model for Pol ε catalytic domain–DNA–Ctf18-RFC–PCNA complex (PDB:8TWA (Yuan et al, 2024)) docked into the cryo-EM density for Pol ε–Ctf18-RFC obtained in this study (EMD-52116). The Pol ε catalytic domain from 8TWA was rigid body docked into EMD-52116 using the “fit-in-map” command in ChimeraX. Docking highlights how the Ctf18-RFC ATPase module in 8TWA adopts an alternative position relative to Pol ε compared to that observed in this study. (B) Model for Pol ε catalytic domain–DNA–Ctf18-RFC–PCNA complex (PDB:8TWA) aligned to a model for the budding yeast replisome, where the Pol ε catalytic domain is positioned below the Mcm2-7 C-tier. To generate this replisome model, the previously published structure of the budding yeast replisome (PDB:6SKL (Baretic et al, 2020)) was rigid body docked into EMD-52107, and the Pol ε catalytic domain bound to Ctf18-RFC (PDB:6S2E (Stokes et al, 2020)) was docked into EMD-52116. The Pol ε catalytic domain of 8TWA was then aligned using Matchmaker in ChimeraX to the Pol ε catalytic domain of 6S2E. This alignment reveals how the conformation of the Ctf18-RFC ATPase module in 8TWA clashes with the C-tier of the Mcm2-7 helicase when the replisome adopts this configuration.

Figure EV6. (Related to Fig. 6). Analysis of mutants designed to disrupt the interface between Ctf18 and the Mcm7 WH.

(A) Detailed views of the AlphaFold predicted interaction between S. cerevisiae Ctf18 (red) and Mcm7 730–845 (blue). In the third panel, the Mcm7 WH molecular surface is rendered by hydrophobicity according to the Kyte-Doolittle scale, with hydrophilic regions in cyan and hydrophobic regions in grey. (B) Coomassie-stained 4–12% SDS-PAGE gel of purified S. cerevisiae wildtype (WT) or mutant Ctf18-RFC and CMG complexes. Subunits are labelled. (C) Denaturing agarose gel analysis of a Pol ε primer extension reaction with wildtype (WT) or mutant Ctf18-RFC complexes present where indicated. Reactions were performed at 100 mM potassium glutamate. (D) Denaturing agarose gel analysis of replication reactions using the staged CMG-based system with Pol ε addition 1 min after initiation of template unwinding. Wildtype (WT) or mutant CMG complexes and RFC were present where indicated. Source data are available online for this figure.