Extended Data Fig. 5: Direct torque measurements during DNA twisting using the nanorotor.
From: The energy landscape for R-loop formation by the CRISPR–Cas Cascade complex
a, Experimental setup for torque measurements during DNA twisting. Torque measurements require the nanorotor to be torsionally constrained with respect to both attachments, that is at the surface and at the bead. Twisting the DNA by magnet rotations changes the torque in the molecule, which in turn displaces the nanorotor. Calibrating the torsional stiffness ktot of the nanorotor system (see Extended Data Fig. 4), which is dominated by the linker on the bottom, allows to obtain the torque from the angular displacement of the nanorotor (see Supplementary Discussion 4). b, DNA length and torque during DNA twisting at different forces. Around zero turns, the DNA length remains approximately constant, while the torque is increasing proportionally with the applied turns. From the slope of this linear part, a torsional persistence length of dsDNA of ptor =102 nm was obtained, which is in good agreement with previous reports44. Once a critical positive torque was reached, the DNA buckles, which is seen as a sudden length jump. This is followed by a linear DNA length decrease with applied turns in which the plectonemic writhe structure (see sketch in torque plot) is further expanding. In the plectonemic regime, the torque remains approximately constant. Both the buckling point as well as the torque plateau are force dependent. For the measurement at 5.7 pN, no length decrease is observed, as the DNA undergoes a transition to p-DNA21. When the negative torque falls below −9 pN nm, another torque plateau is reached, where additionally applied negative turns are absorbed by force-independent local melting or z-DNA formation45. These local structural transitions appear to be dynamically changing throughout the molecule. They are typically found in the long DNA spacer above the nanorotor but can transiently appear as well in the dsDNA part below the nanorotor (see transient spikes towards negative nanorotor angles). Measurements are performed at a resolution of 1 pN nm within 38 ms which allowed to measure the torque continuously while twisting, instead of averaging measurements at set twist values46.
