Specific nucleotide binding and rebinding to individual DNA polymerase complexes captured on a nanopore

Abstract

Nanoscale pores are a tool for single molecule analysis of DNA or RNA processing enzymes. Monitoring catalytic activity in real time using this technique requires that these enzymes retain function while held atop a nanopore in an applied electric field. Using an alpha-hemolysin nanopore, we measured the dwell time for complexes of DNA with the Klenow fragment of Escherichia coli DNA polymerase I (KF) as a function of the concentration of deoxynucleoside triphosphate (dNTP) substrate. We analyzed these dwell time measurements in the framework of a two-state model for captured complexes (DNA-KF binary and DNA-KF-dNTP ternary states). Average nanopore dwell time increased without saturating as a function of correct dNTP concentration across 4 orders of magnitude. This arises from two factors that are proportional to dNTP concentration: (1) The fraction of complexes that are in the ternary state when initially captured predominantly affects dwell time at low dNTP concentrations. (2) The rate of binding and rebinding of dNTP to captured complexes affects dwell time at higher dNTP concentrations. Thus there are two regimes that display a linear relationship between average dwell time and dNTP concentration. The transition from one linear regime to the other occurs near the equilibrium dissociation constant (K(d)) for dNTP binding to KF-DNA complexes in solution. We conclude from the combination of titration experiments and modeling that DNA-KF complexes captured atop the nanopore retain iterative, sequence-specific dNTP binding, as required for catalysis and fidelity in DNA synthesis.

Figure 1. Detection of capture events with the nanopore device

(a) Schematic of the nanopore device. A patch-clamp amplifier supplies voltage and measures ionic current through a single α-HL channel inserted in a ∼25-μm diameter lipid bilayer (trans-side positive). Current through the nanopore is carried by K⁺ and Cl^- ions. (b) Representative current trace at 180 mV for nanopore capture of a ternary complex of KF, dGTP, and the 5ab(12,16) 14 bphp DNA substrate. Cartoons in b illustrate the molecular events that correspond to each current level. Red circles in the DNA strand represent the segment of five abasic residues in the 5ab(12,16) 14 bphp. A dNTP molecule is illustrated with its nucleobase base-paired to the n=0 template position in the KF active site, and its triphosphate moiety indicated as -PPP. b(i), The initial long blockade at 31 pA is the enzyme bound state (EBS). It arises from capture of binary or ternary complex, with the duplex DNA held atop the pore vestibule due to association with KF. b(ii), The shorter ∼ 21 pA terminal step occurs when KF dissociates and the duplex DNA drops into the nanopore vestibule. b(iii), The DNA hairpin unfolds and translocates through the nanopore, resulting in a return to the open channel current. (c) Representative current trace for capture of a ternary complex of KF, dGTP, and the standard DNA 14 bphp substrate. The amplitude of the initial blockade corresponding to the EBS with the standard DNA substrate is 24 pA. (d) Dwell time vs. amplitude plot showing the EBS current segments (black points) and their corresponding the terminal current step segments (blue points) for a representative experiment in which 1 μM DNA (5ab(12,16) 14 bphp), 2 μM KF, and 4 μM dGTP were present in the nanopore chamber. The red points indicate events that lack terminal current steps.

Figure 2. Effect of dGTP concentration on nanopore dwell time of captured DNA-KF complexes

Histograms of log dwell time (in ms) are shown for the EBS of nanopore capture events measured in experiments with the 5ab(12,16) 14 bphp DNA substrate (1 μM), KF (2 μM), and dGTP at the concentrations indicated in the figure panels. The histograms for dGTP concentrations from 0 μM to 1000 μM represent no less than 105, and up to 653, EBS events (Table S1). Data for dwell times longer than 4700 ms is not displayed in the histograms for 3000 μM dGTP (89 total EBS events) and 10,000 μM dGTP (69 total EBS events) due to space limitations. See the Supporting Information for the calculation of the 95% confidence interval for the observed number of samples in each individual bin of the histograms, and plots of histograms with error bars (representing 95% confidence intervals) for sixteen dGTP concentrations (Figs. S5, S6, S7 and S8).

Figure 3. Two-state model for dNTP binding to DNA-KF complexes captured in the nanopore

Illustration of a model for two states of nanopore-captured DNA-KF complexes (binary and ternary) and the probabilities and rates that govern transitions to and from these two states. Probabilities of initial capture of either binary or ternary complex are indicated as p₁(0) and p₂(0), respectively, and are dependent upon the bulk phase equilibrium of binary and ternary complexes. Captured complexes can exchange between the binary and ternary states while they reside atop the pore due to dNTP binding (k_on[dNTP]) and unbinding (k_off). KF dissociates (k₁) only from binary complexes. KF dissociation is irreversible because the double strand-single strand DNA junction drops into the nanopore vestibule where it is inaccessible to KF.

Figure 4. Comparison of dGTP concentration-dependent EBS mean dwell times predicted by the model and the experimentally observed mean dwell times

(a) EBS mean dwell times measured for the four lowest dGTP concentrations ([dGTP]≪ K_d^(B)) were fit to a linear function using the least square formulation. (b) EBS mean dwell times measured for the six highest dGTP concentrations ([dGTP]≫ K_d^(B)) were fit to another linear function using the least square formulation on a logarithmic scale. Slopes and y-axis intercept values from (a) and (b) were used to determine kinetic parameters k₁, k_off, K_d, and K_d^(B). (c) Experimentally observed mean dwell times (squares) and those predicted from a model generated using the kinetic parameters (solid line). The dGTP-concentration dependence of the measured and predicted mean dwell times across the entire range of dGTP concentrations tested is shown. The dashed lines show the two linear functions from (a) and (b) (which are not straight lines when plotted in log-log scale). The model curve shows the transition from one linear function to the other, which occurs in a region around [dGTP]= K_d^(B). Error bars indicate the 95% confidence interval for each sample mean. The dotted line indicates the experimental mean dwell time for the binary complex of KF and DNA (0 μM dGTP).