Skip to main content
. 2022 Dec 3;13:7460. doi: 10.1038/s41467-022-35116-5

Fig. 1. Random-walk model for target recognition by Cascade.

Fig. 1

a Target recognition starts with PAM binding (i) followed by R-loop priming adjacent to the PAM. The R-loop can expand in a reversible random-walk like fashion towards the PAM distal end of the target (ii). After reaching the PAM distal end (iii), a conformational change is induced that locks the R-loop in a stable conformation and enables Cas3 recruitment and cleavage of the DNA (iv). b Simplified one-dimensional free energy landscapes for R-loop formation. Local energy states are the unbound state (U), the R-loop states of different length and the locked full R-loop (F). R-loop expansion and shrinkage occurs in successive forward or backward steps to either adjacent state. Shown free energy landscapes correspond to R-loop formation without an energy bias (top), with a negative bias as expected for negative supercoiling or due to favorable molecular interactions (middle) and with a mismatch at position 14, which introduces a local energy barrier (bottom). In the latter case an additional dynamic intermediate R-loop state (I) is introduced. c Real-time detection of the R-loop dynamics on negatively supercoiled DNA using magnetic tweezers. R-loop expansion progressively absorbs part of the introduced supercoiling, which results in a DNA length increase. d R-loop formation on a fully matching target is seen as a single abrupt DNA length increase corresponding to the transition from the unbound to the full-R-loop state. e R-loop formation on a target with a single internal C:C mismatch at position 17 occurs via additional transitions to and from the unstable intermediate state until the full locked R-loop is formed.