Fig. 3: Occlusion of 5′ end suppresses folding of threaded conformer.

a Unfolding FECs measured with 1-nt spacer between duplex handle and pseudoknot show same length changes as in Fig. 2. Dashed lines show WLC fits. Inset: schematic showing location of handle duplex. b Unfolding force distribution with 1-nt spacer between handle and pseudoknot (N = 342 FECs from 8 molecules) shows more prominent low-force peak and less prominent high-force peak. c Re-measuring unfolding with handle extending 2 nt into stem 1 of pseudoknot (inset: schematic showing location of handle duplex) yields unfolding force distribution (N = 220 FECs from 3 molecules) with even more prominent low-force peak and even less prominent high-force peak. d Fraction of FECs showing higher-force pseudoknot unfolding (red) decreases systematically as duplex handle is extended towards stem 1 of pseudoknot (or invades it), with compensating increase in fraction showing lower-force pseudoknot unfolding (orange), whereas fraction showing alternative stem-loop unfolding (brown) remains constant. Error bars on occupancies of pseudoknotted (sum of N and N′) and alternate conformers (Alt) represent standard error of proportion from numbers of FECs showing each conformation type; error bars on fraction of pseudoknotted conformers occupying N/N′ represent fitting errors from analysis of unfolding force distributions with two-population landscape model (errors correlated for N and N′, as shown). N = 1825 FECs from 10 molecules for 6-nt spacer, 491 FECs from 8 molecules for 1-nt spacer, and 270 FECs from 3 molecules for −2-nt spacer.