Fig. 3

Titration of Sen1 HD activity. a Time-trace and accompanying sketch showing cycles of Pol II stalling and subsequent removal by Sen1 HD. The total lifetime of the transcription bubble is noted t_total. b Typical pulses showing transcription initiation by Pol II and removal of resulting stalled Pol II by Sen1-HD. Sen1-HD, in a small majority of cases, removes stalled Pol II without any visible intermediates (e.g. first pulse, required time denoted t_total). However, for a subset of ~40% of events, removal of Pol II from DNA can be seen to initiate after time t₁ via formation of an intermediate in which the transcription bubble is apparently rewound by ~3.6 ± 0.1 bp (see Supplementary Fig. 8) (e.g. second pulse). Complete rewinding of the transcription bubble is observed after an additional time t₃. c Mean lifetime of stalled Pol II on DNA as a function of Sen1-HD concentration under negative supercoiling (−sc, σ = −0.027) is well-fit to a Michaelis–Menten model (blue for t_total which is the total lifetime of all transcription bubbles; red for t₁ which is the lifetime prior to intermediate formation when it can be detected). The orange data point represents the mean lifetime of stalled Pol II terminated by 40 pM Sen1 HD for positive supercoiling (+sc, σ = 0.016, n = 37 events). Error bars represent SEM. d Distribution of lifetimes (t_total) of Pol II on DNA as a function of Sen1-HD concentration is well-fit to a single-molecule recasting of the Michaelis–Menten equation (see Supplementary notes) giving a chi-square of 211.5 and a reduced chi-square of 1.47. e Distribution of duration t₃ is well-fit to single-exponential statistics across the concentration range tested, giving time constants of 16.4 ± 2.4 s (SEM, n = 86), 18.1 ± 1.8 s (SEM, n = 131), 19.0 ± 2.4 s (SEM, n = 105), and 16.3 ± 1.9 s (SEM, n = 107) for 10, 20, 40, and 100 pM, respectively, of Sen1 HD. Source data are provided as a Source Data file