Figure 3.
Transient binding of protein S4 to pre-16S rRNA during transcription. A) In co-transcriptional experiments, stalled TECs (left) are restarted in the presence of Cy5-labeled S4 and co-localization of Cy5-S4 with Cy3-TEC is monitored during (middle) and after (right) transcription. B) For post-transcriptional binding experiments, unlabeled stalled TECs were immobilized on the surface (left), and transcription was restarted by addition of NTPs for 2 min. Free NTPs were washed away and a Cy3-SA5_aadU3 oligomer complementary to the 3’ end of the pre-16S RNA was added (middle). RNAs were allowed to fold for 30 min before Cy5-S4 was added to the slide chamber during imaging (right). C) Rastergram of S4 binding (black bars) during and just after transcription; each horizontal line represents a single TEC (N = 54). The lengths of horizontal bars indicate the lifetimes of individual complexes. On the x-axis, t = 0 represents the time of transcription restart marked by the flow of Cy5-S4 and NTPs into the slide chamber. Green circles represent the center of the PIFE peak marking the end of RNA transcription. D) Rastergram of S4 binding (black bars) 30 min post-transcription (N = 54). On the x-axis, t = 0 represents the time when Cy5-S4 is added to the slide chamber. E) Probability density histogram of S4 binding lifetimes 30 min after transcription (post-txn; Nobs = 497 on 104 molecules) and during and shortly following transcription (co-txn; Nobs = 1879 on 103 molecules). Fits represent maximum likelihood fitting to a function with three exponential terms. Error bars represent the variance in a binomial distribution. F) S4 binding probabilities as a function of time after transcription restart (x = 0), as in C,D. Binding events were clustered according to the lifetime τ of the complex: Transient binding (left axis), τ < 1 s (black line and grey circles); productive binding (right axis), τ > 10 s (red line and squares) and τ > 50 s (red dotted line and diamonds). Linear and exponential fits are shown to visualize the trends and do not represent physical models of the data. See also Table S2 for fit parameters.