Substrate conformational dynamics facilitate structure-specific recognition of gapped DNA by DNA polymerase

Abstract

DNA-binding proteins utilise different recognition mechanisms to locate their DNA targets; some proteins recognise specific DNA sequences, while others interact with specific DNA structures. While sequence-specific DNA binding has been studied extensively, structure-specific recognition mechanisms remain unclear. Here, we study structure-specific DNA recognition by examining the structure and dynamics of DNA polymerase I Klenow Fragment (Pol) substrates both alone and in DNA-Pol complexes. Using a docking approach based on a network of 73 distances collected using single-molecule FRET, we determined a novel solution structure of the single-nucleotide-gapped DNA-Pol binary complex. The structure resembled existing crystal structures with regards to the downstream primer-template DNA substrate, and revealed a previously unobserved sharp bend (∼120°) in the DNA substrate; this pronounced bend was present in living cells. MD simulations and single-molecule assays also revealed that 4-5 nt of downstream gap-proximal DNA are unwound in the binary complex. Further, experiments and coarse-grained modelling showed the substrate alone frequently adopts bent conformations with 1-2 nt fraying around the gap, suggesting a mechanism wherein Pol recognises a pre-bent, partially-melted conformation of gapped DNA. We propose a general mechanism for substrate recognition by structure-specific enzymes driven by protein sensing of the conformational dynamics of their DNA substrates.

Figure 1.

Measuring distances within single polymerase–DNA complexes in heterogeneous mixtures with dynamic species. (A) Schematic of a 1-nt gapped DNA substrate showing the template (red lettering) and non-template strands (black). Red stars represent acceptor labelled dT bases. Split red/green stars indicate positions labelled with donor for DNA–DNA FRET, or acceptor for DNA–protein FRET. (B) DNA Polymerase I (Klenow Fragment; Pol) structural schematic (grey – pdb 1KLN) and donor labelling positions (green stars). (C) Apparent FRET histograms for the doubly-labelled substrate T(–12) B(+11) at increasing concentrations of Pol. The data (grey bars) were fitted with up to three Gaussians (black, red and blue dashed lines), yielding apparent FRET efficiencies, E* of 0.35, 0.55 and 0.75. (D) Corrected ES histogram for a DNA–DNA FRET measurement (here, for T(–12)B(+11) in the presence of 3 nM Pol). Data (grey bars) were fitted by the sum of three Gaussians (solid black lines) centered on E = 0.41 (black dash), E = 0.63 (red dash) and E = 0.90 (blue dash) respectively. (E) Corrected ES histogram for a protein–DNA FRET measurement (here, for C907-Cy3B B6-Atto647N). Data were fitted with a single Gaussian function, centered on E = 0.48. See also Supplementary Figure S1.

Figure 2.

Pol-DNA binary structure from rigid-body docking. (A) Pol structure showing the fingers (blue), thumb (purple) and palm (wheat) subdomains and the proof-reading exonuclease domain (grey). Example DNA–DNA and protein DNA-distances (black dashed lines) are shown between mean dye positions (green and red spheres). Example accessible volumes of a donor (pale green cloud) and an acceptor (pale red cloud) dye are also shown, along with the full sequences of the docked DNAs; the shaded region indicating the DNA not used for the docking. (B) Results of the rigid-body docking: template DNA (red), non-template DNA (black), subdomains coloured as in (A). (C) Position of Y719 relative to downstream DNA. (D) Clash between full-length downstream DNA and the fingers subdomain (cyan). See also Supplementary Figure S2 and Supplementary Table S1.

Figure 3.

Binary complex structure and dynamics. (A) Representative snapshot of the DNA–Pol binary complex from a 100-ns MD simulation, showing the volume accessed by the DNA over the simulation (pale pink). The plot shows the DNA end-to-end distance fluctuations over the same simulation, with the ends taken as the terminal non-hydrogen atoms of the template strand. The time point corresponding to the snapshot is indicated with an arrowhead. (B) Representative snapshot of the conformation of the 6-nt non-template flap, with its volumetric map during a 100-ns simulation (pale orange). The plot shows the frequency of the number of hydrogen bonds formed between the flap and the template strand of downstream DNA during the entire 1-μs (10 × 100 ns) simulation time. (C) Overview of Pol residues involved in strand separation or interactions with downstream DNA. See also panels (D) to (F). (D) Involvement of Y719 in strand separation of downstream DNA. Top - a representative snapshot of the position of Y719 relative to the template DNA strand. The three DNA residues positioned closest to Y719 during the time course of the simulation are highlighted in CPK colouring. The position of the three-helix bundle is shown for reference; the rest of the protein is omitted for clarity. Lower panels, two different views of the volumetric maps of Y719 (yellow), template (red) and non-template DNA strands (black) during a 100-ns simulation. (E) A representative snapshot of the interactions between R779 (green) and R784 (cyan) with phosphate groups (orange spheres) in the non-template strand of downstream DNA. The plot shows the minimum distance between the side-chain nitrogen atoms of R779 (green) or R784 (cyan) to any phosphorous atom in the non-template strand of downstream DNA during a 100 ns simulation. Arrowhead denotes the time point corresponding to the snapshot, and dashed lines indicate the distance corresponding to an interaction. (F) A representative snapshot of the interaction between residue K549 (purple) with phosphate groups in the template strand of downstream DNA. See also Supplementary Figure S3, Movie M1 and PBD File P1.

Figure 4.

Downstream DNA is melted in the binary complex. Quenchable FRET data are plotted, showing the number of events with mid-stoichiometry (0.4 < S < 0.8) versus the apparent FRET efficiency, for DNA substrate T(+1)B(+4) alone (grey bars) and in the presence of 3 nM Pol (red bars). Inset: Schematics of the labeling positions and DNA structures for the unbound (left; B-DNA) and bound conformations (right; snapshot from MD simulations, atomic coordinates provided as SI), and the related accessible volumes of the donor (green) and acceptor (red) dyes, quoting the percentage overlap between them (see main text). See also Supplementary Figure S4.

Figure 5.

Gapped DNA substrate adopts highly bent conformations due to base unstacking. (A) Schematic of the 1-nt gapped DNA substrate. Green and red stars indicate donor and acceptor labelled dT bases, respectively. (B) Comparison of experimental (red) and simulated (black) corrected FRET values for 34 FRET measurements in the 1-nt gapped substrate (Labelling index listed in Supplementary Table S2). Inset: residual histogram, ΔE = E_model – E_measured. (C) Simulated FRET time trace for substrate T(+8)B(–11), labelling index 1. Stacked (blue circles) and unstacked (green circles) states are identified. The average FRET efficiency across the entire simulation is shown for each state individually (blue and green lines), for the sum of both states (black line) and for the experimentally determined FRET value (red line). (D) Typical snapshots of the stacked and unstacked states indicating the stacking (blue) and cross-stacking (orange) interactions present. (E) Free-energy profiles as a function of the bend angle (θ = 0 corresponds to a straight duplex) for 1nt-gap (black), nicked (blue) and duplex (magenta) DNAs. The total free energy (solid lines) and the contribution due to unstacked states alone (broken lines) are shown. The vertical dotted lines and cartoons correspond to the most stable bend angle (20°) and the bend angle observed in the binary complex structure (120°). See also Supplementary Figure S5, Supplementary Table S2 and Movie M2.

Figure 6.

DNA bending detected in live cells. (A) Schematic showing internalization of doubly-labelled gapped DNA fragments into live E. coli using electroporation and single-molecule imaging (left to right). Example cell (bottom right) is shown in white-light image, donor fluorescence channel, FRET fluorescence channel, and latter both combined in overlay image. The overlay image is color-coded such that intermediate-FRET molecules appear orange and high-FRET molecules appear red; two example molecules are highlighted accordingly. Scale bar: 1 μm. (B) FRET histogram of tracked gapped DNA trajectories in vivo. Two major FRET species were observed for the T(–12)T(+8) substrate, which were attributed to unbent DNA (black dots; E* = 0.40) and bent DNA (red dots; E* = 0.83). The number of trajectories (N) is stated for each experiment. (C) FRET histogram of tracked duplex DNA trajectories in vivo. A single low-FRET species, was observed (black, E* = 0.38). See also Supplementary Figure S6.

Figure 7.

Gapped DNA recognition: conformational capture followed by an ‘on-protein’ rearrangement. Gapped DNA is dynamic adopting bent and frayed states (orange haze). Pol can bind to the upstream DNA when the downstream DNA conformation is not impeding the Pol (conformational capture of slightly bent states). Following binding of the upstream DNA, the downstream DNA now docks and is further melted, beginning the process of strand-displacement.