Extended Data Fig. 7: Procedure for DNA molecule calibration and dgRNA-SpCas9 injection using single-molecule magnetic trapping.

(a) Typical extension vs. supercoiling curve showing a DNA molecule rotated from 20 turns of positive supercoiling to 20 turns of negative supercoiling under a constant ~0.2 pN extending force, allowing us to identify un-nicked dsDNA molecules and also to calibrate the change in DNA extension for a unit change in DNA writhe (Strick et al. 1996), here typically ~55 nm/turn. (b) Extension time-trace for the molecule calibrated in (a), providing real-time references at a constant 0.2 pN extending force (blue line) for its extension at (i) 8 turns of negative supercoiling, (ii) zero supercoiling (that is the torsionally-relaxed state), and (iii) 5 turns of positive supercoiling before (iv) increasing the applied force to 1 pN and injecting S1_T0 SpCas9 RNPs (seen as a downward spike in extension as flow pushes the bead towards the surface). Once flow ends and the bead returns to its extension prior to injection (v) the applied force is returned to 0.2 pN and the extension is verified to return to that see in (iii). After a few minutes (vi) DNA molecules are returned to 8 turns of negative supercoiling. (vii) R-loop formation and (viii) supercoil loss are followed by (ix) terminal loss of the magnetic bead. (c, d) as in (a,b) but with injection of only the S1_T0 guide (that is without SpCas9) shows that no R-loops are formed within 2000 seconds of detection window. This experiment was repeated on a total of 72 molecules.