Figure 1

Top: Schematics of the experimental setup; the bead is tethered to the bottom of the flow cell by the probe DNA; its position is monitored using evanescent wave scattering. Bottom: The 76 bases long probe DNA containing the IHF binding site is grafted to the $1\mu \mathrm{m}$ diameter bead and slid through adapter oligomers. When IHF binds, the bead is pulled towards the slide.Reuse & Permissions

Figure 2

Time series of the bead’s position (in nm) with respect to the surface, showing that single IHF molecules bind to the DNA and fall off. In the bound state, the DNA is bent and the bead is pulled closer to the surface. For each binding event, we can measure the duration of the bound state, $\Delta$.Reuse & Permissions

Figure 3

(a) Cumulative distribution of the duration of the bound state $\Delta$ obtained from a time trace such as shown in Fig. 2. With a total of $N(0)$ observed binding events, $N(0)-N(\mathrm{t})$ is the number of time intervals with $\Delta <t$. The slope in a semilog plot gives the off rate. We observe two different off rates [$(5.3\pm 0.7)\times {10}^{-2}{\mathrm{s}}^{-1}$ and $0.54\pm 0.10{\mathrm{s}}^{-1}$ in this case], corresponding to two distinct modes of binding of the IHF. The straight lines are linear fits; the whole data set is shown in the inset. (b) The bead-slide interaction potential measured from the thermal motion of the bead in the “off” state. The continuous line is a fit used in the analysis. When the IHF binds, shortening of the molecular tether pulls the bead up the repulsive part of the potential. The derivative of the potential at the “bound” position (see Fig. 2) gives the tension on the tether. (c) Different tensions in the DNA give rise to different off rates. The plot shows the statistics of the “fast” time scale for two different experiments corresponding to measured tensions of 3 and 8.3 pN. The slopes give off rates of $4.9\times {10}^{-2}$ and $0.24{\mathrm{s}}^{-1}$, respectively.Reuse & Permissions

Figure 4

The lifetime of the IHF bound state vs tension in the DNA for the two different modes of binding. The data are consistent with an exponential dependence on tension. Vertical error bars represent the statistical error originating from the finite sample size and were determined from a simulation (see [19]); the horizontal error bars originate from the spread in the bead’s position in the bound state and the uncertainty in the derivative of the potential. The slope of the line is the same for the two modes of binding and defines a characteristic length scale (the position of the barrier in the energy landscape confining the bound state) $s\approx 1.4\pm 0.2\mathrm{nm}$. A more detailed analysis, taking into account that the external force varies in space, yields the more accurate values $s\approx 1.7\pm 0.2\mathrm{nm}$ (see Fig. S3 in the supplementary material [19]). Extrapolating the data to zero tension, we recover the life times measured in ensemble experiments.Reuse & Permissions