Extended Data Fig. 10: R-loop free energy landscapes before and after locking for the T6, T2 and T1 targets.
From: The energy landscape for R-loop formation by the CRISPR–Cas Cascade complex
a-d, (left) R-loop length trajectories including a locking event measured on targets with (a) six, (b) two, (c) one and (d) zero terminal mismatches (T6, T2, T1 and WT targets). The targets sequence is shown below (Colours as in Extended Data Fig. 6). Data was taken at 3947 fps (light colours) and smoothed with a 100-point sliding average (dark colours). The extensive R-loop length fluctuations before locking became suddenly quenched upon irreversible locking indicating a tightly constrained R-loop length (grey shaded area). (Right) Histograms of the measured R-loop lengths as well as apparent (solid lines) and deconvolved (lines and circles) energy landscapes before (black, red, blue and green colours) and after locking (purple colour). For the T2, T1 and WT targets, a pronounced shift of the predominant R-loop length from 27 to 30 bp (T2, T1 target) or 32 bp (WT target) is observed upon locking. For the T6 target, the predominant R-loop length of ~26 bp remains unchanged. e, Deconvolved free energy landscapes of the locked state for the T6, T2 and T1 targets (solid lines). We added a mismatch penalty of 2.26 kBT to the T2 and T1 landscapes for each additional mismatch that the T6 target comprises (dashed lines) to reconstruct the T6 energy landscape. The landscapes are shifted along the energy axis for better comparison. f, g, Scheme of the observed free energy landscapes of R-loop formation including the unbound and the locked state (blue and grey shaded areas, respectively) without (f) and with (g) an internal mismatch (before position 27). For a fully matching target the terminal barrier is slightly lower than the initial PAM binding barrier such that full R-loop formation and locking occur at a higher probability than R-loop collapse. The local penalty induced by an internal mismatch also lifts the terminal barrier such that R-loop collapse becomes favoured over full R-loop formation and locking, typically resulting in rejection of the target. Such energetics of R-loop formation is indicative of a kinetic target discrimination mechanism25. Thus, similar heights of the initial and terminal barriers on a fully matching target ensure highly efficient R-loop formation and high specificity at the same time.