Quality control mechanisms exclude incorrect polymerases from the eukaryotic replication fork

Abstract

The eukaryotic genome is primarily replicated by two DNA polymerases, Pol ε and Pol δ, that function on the leading and lagging strands, respectively. Previous studies have established recruitment mechanisms whereby Cdc45-Mcm2-7-GINS (CMG) helicase binds Pol ε and tethers it to the leading strand, and PCNA (proliferating cell nuclear antigen) binds tightly to Pol δ and recruits it to the lagging strand. The current report identifies quality control mechanisms that exclude the improper polymerase from a particular strand. We find that the replication factor C (RFC) clamp loader specifically inhibits Pol ε on the lagging strand, and CMG protects Pol ε against RFC inhibition on the leading strand. Previous studies show that Pol δ is slow and distributive with CMG on the leading strand. However, Saccharomyces cerevisiae Pol δ-PCNA is a rapid and processive enzyme, suggesting that CMG may bind and alter Pol δ activity or position it on the lagging strand. Measurements of polymerase binding to CMG demonstrate Pol ε binds CMG with a K_d value of 12 nM, but Pol δ binding CMG is undetectable. Pol δ, like bacterial replicases, undergoes collision release upon completing replication, and we propose Pol δ-PCNA collides with the slower CMG, and in the absence of a stabilizing Pol δ-CMG interaction, the collision release process is triggered, ejecting Pol δ on the leading strand. Hence, by eviction of incorrect polymerases at the fork, the clamp machinery directs quality control on the lagging strand and CMG enforces quality control on the leading strand.

Keywords: DNA polymerase; PCNA; clamp loader; replication; replisome.

Fig. 1.

RFC inhibits Pol ε, but not Pol δ, on a model lagging strand. RFC is titrated into either (A) Pol ε or (B) Pol δ reactions on primed, RPA-coated ssDNA, and products were analyzed in an alkaline agarose gel. “FL” indicates the full length (5.4 kb) RFII product.

Fig. S1.

RFC inhibition still occurs even when RFC is added after Pol ε-PCNA has been initiated. (A) Schematic of reaction corresponding to the experiment in group 4 (explained below). Pol ε reactions on primed, RPA-coated ssDNA are allowed to progress for 4 min, and 60-nM RFC is subsequently added to the reaction. (B) Alkaline agarose gel of extension products of Pol ε in the presence of the indicated reaction components, then stopped at the indicated times and visualized by phosphorimaging. As indicated, the four groups in this gel are: (1) Control Pol ε reaction with 5 nM (stimulatory) RFC. (2) 60 nM RFC is added to the mixture before Pol ε, before initiation. (3) 60 nM RFC is added after Pol ε, before initiation. (4) A reaction using a low (5 nM) RFC concentration is used to start the reaction as in group 1; one fraction is removed and stopped at 4′, and 60 nM RFC is subsequently added to the reaction.

Fig. S2.

RFC displays a detectable inhibition of the DNA polymerase activity of Pol α-primase on DNA primed model lagging stand substrates. (A) Schematic of reaction. RFC is titrated into Pol α-primase reactions on DNA primed and RPA-coated ssDNA. (B) Alkaline agarose gel of extension products of Pol α-primase in the presence of the indicated RFC concentrations, visualized by phosphorimaging. Reactions were stopped at the indicated times.

Fig. 2.

Pol ε and RFC bind a primer terminus with similar affinity and preferentially bind primed DNA. (A) Fluorescence anisotropy binding curves are compared for Pol ε (black) and RFC (red) binding to the FITC-labeled template/primer (T/P) depicted above. The curves are the respective Kd_app value fits to the data. (B) RFC and Pol ε bind primed DNA more tightly than ssDNA or dsDNA. Kd_app values for the indicated binding partners are presented as mean ±SD.

Fig. S3.

Anisotropy measurements of RFC and Pol ε with fluorescently tagged dsDNA or ssDNA. Fluorescence anisotropy binding curves corresponding to the dsDNA and ssDNA Kd values shown in Fig. 2B are shown for Pol ε (solid black circles) and RFC (empty circles) binding to the FITC-labeled dsDNA (A) or ssDNA (B) substrates (substrates are indicated above the figures).

Fig. S4.

Competition predictions for RFC and Pol ε at estimates of their physiological concentrations. Using the calculated Kd values for RFC and Pol ε and an estimate of the concentration of origins in the S. cerevisiae nucleus (Competitive Binding Model), an equation for the occupancy of Pol ε in the presence of RFC competitor was solved for the indicated Pol ε and RFC concentrations. The binding occupancy of Pol ε is expressed as a color gradient (scale bar shown on Right). Circled are nuclear concentrations derived from protein counts reported in the indicated reference, with the circle size scaled to reflect reported error in both Pol ε and RFC estimates, i.e., Pol ε and RFC concentrations were respectively calculated as 2.0 ± 0.7 μ and 1.1 ± 0.5 μM from Ghaemmaghami et al. (21), or 2.1 ± 0.8 and 1.1 ± 0.5 mM from Kulak et al. (44). The full details of these calculations and the model upon which this figure was generated are given in Competitive Binding Model.

Fig. S5.

Pol δ does not bind primed template DNA as detected using fluorescence anisotropy. Fluorescence anisotropy binding curves corresponding to Pol δ (filled circles) binding to the T/P substrate used in Fig. 2A. RFC binding to primed template (squares) is included for comparison. Data are presented as mean ±SD.

Fig. S6.

Pol ε has comparable activity on primed ssDNA using two different primer lengths and either RPA or SSB. (Top) Reaction scheme. Alkaline agarose gels of ³²P-dC incorporation by Pol ε on primed ϕX174 ssDNA, stopped at the indicated time points, are visualized by phosphorimaging. (A) Pol ε is active with RPA. RPA or E. coli SSB are preincubated with ssDNA before replication initiation at a ratio of 1.5:1 of [protein]:[footprint], where [footprint] is the expected concentration of the ssDNA footprint available for either RPA (∼20 nt; ref. 38) or SSB (∼35 nt; ref. 39) binding. (B) Pol ε is functional on short primed sites. Extension by Pol ε of the indicated ϕX174 ssDNA primer lengths is shown.

Fig. S7.

Pol ε is inactive on naked DNA and RFC inhibits Pol ε on SSB-coated ssDNA. (Top) Schematic of reaction. RFC is titrated into Pol ε reactions on primed, E. coli SSB-coated ssDNA. Alkaline agarose gel of extension products of Pol ε in the presence of the indicated RFC concentrations, visualized by phosphorimaging. All reactions were stopped at 10 min.

Fig. S8.

CMG stimulates Pol ε activity on a model lagging strand. (A) Reaction scheme. CMG is titrated into Pol ε reactions on primed, RPA-coated ssDNA. (B) Alkaline agarose gel of extension products in the presence of the indicated CMG concentrations, visualized by phosphorimaging. Reactions were stopped at the indicated times. “FL” indicates the full length (5.4 kb) product.

Fig. 3.

RFC does not inhibit Pol ε during function with CMG in leading strand synthesis. (A) Schematic of the reaction. RFC is titrated into CMG-directed Pol ε (leading strand) reactions on a 3-kb linear substrate ligated to a forked junction primed with a 5′-³²P labeled primer. (B) Alkaline agarose gel of extension products in the presence of the indicated RFC concentrations. Reactions were stopped at the indicated times. “FL” indicates the full length (3.2 kb) product.

Fig. 4.

CMG tightly binds Pol ε but does not bind Pol δ. (A) Schematic of MST assay. Pol ε or Pol δ is mixed with fluorescently labeled CMG. A localized IR beam creates a temperature gradient, and thermophoresis causes depletion of molecules from the observation volume. During equilibrium, some molecules return toward the heat gradient. (B) MST traces for CMG binding to Pol ε. Arrow indicates increasing Pol ε concentration. F_cold and F_hot are depicted as blue and red regions. (C) Binding isotherm for CMG and Pol ε, obtained from plotting ΔF_norm (the change in the F_cold/F_hot ratio) vs. Pol ε concentration. The fitted Kd_app value is shown. (D) MST traces and (E) binding curve for CMG binding to Pol δ. Data are presented as mean ±SD.

Fig. S9.

Representative capillary tube scans from an MST experiment. Overlayed are 16 normalized cross-sectional fluorescence intensity scans of capillary tubes containing CMG-Cy5 and various concentrations of Pol ε. The Gaussian shape of the scan indicates there were no insoluble aggregates adsorbed to the walls of the tubes.

Fig. S10.

Pol ε and Pol δ are PCNA dependent. (Top) Reaction scheme. PCNA is titrated into Pol ε or Pol δ primer extension reactions at the indicated concentrations. (Bottom) Alkaline agarose gel of extension products in the presence of the indicated PCNA concentrations, visualized by phosphorimaging. Reactions were stopped at the indicated times. “FL” indicates the full length (5.4 kb) product.

Fig. 5.

Pol δ binds PCNA–DNA with over 20-fold higher affinity than Pol ε binds PCNA–DNA. MST binding curves of primer-loaded PCNA–Cy5 binding to (A) Pol ε and (B) Pol δ. Reaction schemes are shown above each plot: Fluorescently tagged PCNA is loaded by RFC on to a primed template substrate and reacted with the indicated polymerase. Isotherms are obtained by plotting ΔF_norm vs. Pol ε or Pol δ concentration. The fitted Kd_app value is shown.

Fig. 6.

RFC does not inhibit Pol δ during lagging strand synthesis in reconstituted replication forks. (A) Reaction scheme. RFC is titrated into replication reactions on a nucleotide-biased 3-kb linear DNA substrate ligated to an unprimed fork. Okazaki fragments are visualized by strand-selective ³²P-dG incorporation (red arrows). (B) Alkaline agarose gel of lagging strand synthesis products in the presence of the indicated RFC concentrations. Reactions contained CMG, PCNA, RPA, the indicated RFC, and either 20-nM Pol α and/or 20-nM Pol ε where indicated, and the indicated concentrations of Pol δ. All reactions were stopped at 20 min.

Fig. 7.

Model of quality control reactions at the fork. (A) Pol ε and Pol δ are shown on the incorrect strands. (B) Quality control reactions described in the text eject incorrectly placed Pol ε and Pol δ. (C) Pol ε and Pol δ are shown on the correct strands, in which Pol ε is stabilized by CMG on the leading strand and Pol δ targets PCNA on the lagging strand. For clarity, Ctf4, Pol α, and other replisome proteins are not shown.