Transcription through the roadblocks: the role of RNA polymerase cooperation

Abstract

During transcription, cellular RNA polymerases (RNAP) have to deal with numerous potential roadblocks imposed by various DNA binding proteins. Many such proteins partially or completely interrupt a single round of RNA chain elongation in vitro. Here we demonstrate that Escherichia coli RNAP can effectively read through the site-specific DNA-binding proteins in vitro and in vivo if more than one RNAP molecule is allowed to initiate from the same promoter. The anti-roadblock activity of the trailing RNAP does not require transcript cleavage activity but relies on forward translocation of roadblocked complexes. These results support a cooperation model of transcription whereby RNAP molecules behave as 'partners' helping one another to traverse intrinsic and extrinsic obstacles.

Fig. 1. Cooperation between RNAP molecules in overcoming EcoRQ111 roadblock. (A) Stimulating effect of trailing RNAP molecules on transcription through DNA-bound EcoRQ111. An EcoRI restriction site is located at position +98 from the +1 start of transcription of template 1. A step-by-step diagram of the experiment is shown at the top (see details in the text and Materials and methods). The autoradiogram displays RNA products synthesized during a single round of transcription by RNAP1. The washed roadblocked complexes were chased in the presence of RNAP2 with or without rifampicin (Rif) for the indicated time intervals (lanes 2–7). (B) Dissociation of EcoRQ111 from DNA during transcription. The auto radiogram shows end-labeled DNA template fragments (template 1) recovered from transcription reactions with RNAP1 alone (lanes 1 and 2) or together with RNAP2 (lanes 3–5). The experiment was performed as shown in (A) except that native EcoRI was also added prior to the chase reaction (lanes 2–5). The increase in EcoRI-mediated DNA cleavage (% cut) during multi-round transcription (lane 4) reflects the increased accessibility of the restriction site, i.e. dissociation of EcoRQ111. (C) Comparison of the anti-roadblock effect of RNAP2 with that of GreB. The experiment is performed as in (A) except that equal molar amounts of RNAP2 and GreB (3 pmol) were added prior to the chase reaction. The amount of the roadblocked complexes (% block) was calculated by dividing the amount of radioactivity in the ‘EcoRQ111’ band by the total radioactivity present in all readthrough bands.

Fig. 2. Cooperation between RNAP molecules in overcoming the lac repressor roadblock. A step-by-step diagram of the experiment is shown at the top. The autoradiogram displays RNA products synthesized during a single round of transcription by RNAP1. The washed roadblocked complexes were chased in the presence of RNAP2 holoenzyme [with or without rifampicin (Rif)], core RNAP2 (core) or IPTG (200 µM) for the indicated time intervals. The amount of the roadblocked complexes (% block) was calculated by dividing the amount of radioactivity in the ‘lac repressor’ band by the total radioactivity present in all readthrough bands. Note that not only the roadblock band, but also specific arrest and pausing bands disappear much faster if transcription performed in the presence of RNAP2 (lanes 4 and 5).

Fig. 3. The anti-roadblock mechanism. (A) Analysis of the conformation of the roadblocked EC. A step-by-step diagram of the experiment is shown at the top. The autoradiogram displays RNA products synthesized during a single round of transcription by RNAP1 with (lanes 2–5) or without EcoRQ111 (lane 1). All conditions are as in Figure 1. The washed roadblocked complex was treated with GreB before (lanes 2 and 3) or after washing with 1 mM KCl followed by equilibration with standard TB (lanes 4 and 5). Numbers on the left indicate the size of RNA transcripts. (B) The model for the anti-roadblock mechanism. The scheme summarizes results of Figure 1 and this figure. Trailing EC first reactivates the roadblocked EC by translocating it forward. Next, the active blocked EC has an opportunity to move through the block as soon as the latter dissociates. In addition, the local distortion of DNA and/or steric clash between EC and the block may facilitate dissociation of the block. R, EcoRQ111 roadblock. The model can be generalized for other types of blocking molecules.

Fig. 4. Physical interaction between RNAP molecules in vivo. (A) Schematic representation of RNAP molecules roadblocked by the lac repressor within the two plasmids, pATC10 and its terminator-containing variant pATC10Ter. (B) Primer extension analyses of the in situ CAA modifications on the non-template strand in pATC10 and pATC10Ter. The positions of the operator (lac) and the terminator (Ter) are depicted, as well as the locations of EC1 and EC2 within the two plasmids. The arrows mark the positions of the transcription start sites in each construct. The + and – IPTG indicate the presence or absence of the inducer in the growth media during the footprinting experiment. Densitometer scans of lanes 1 and 3 are shown on the right. (C) S1 mapping of the 3′ ends of the RNA transcripts produced in vivo by the roadblocked ECs from the two plasmid derivatives. Bands corresponding to positions –6 and –9 are indicated by arrows. Densitometer scans of lanes 3 and 4, obtained as in (B), are shown on the right.

Fig. 5. Anti-roadblock activity by the trailing EC in vivo. (A) Comparison of the transcriptional readthrough within plasmids pATC10 and pATC10Ter. Autoradiogram of the extension products obtained after in vitro reverse transcription of the RNAs with ³²P primers hybridizing to either cat or bla transcripts. Note, that the reverse transcriptase does not read efficiently across the terminator hairpin. As a result, the extension reaction with the terminator-containing RNAs generates two products (indicated by arrowheads), the runoff at +1 of the transcript, and additional products resulting from the extension up to the base of the terminator hairpin stem. (B) Quantification of the primer extension products from (A) using STORM PhosphorImager. The value plotted in each lane indicates the level of cat mRNA relative to the internal control bla transcript. (C) Determination of the readthrough efficiency within the two plasmids. The readthrough was calculated using the following formula: [(cat/bla transcripts)_–IPTG/(cat/bla transcripts)_+IPTG] × 100. The error bars reflect the standard deviation from the mean of three independent experiments.

Fig. 6. Transcriptional readthrough in vivo as a function of a distance between the promoter and roadblock. (A) Schematic comparison of the pATC10 and pATC21 constructs and their capacity for ECs. (B) The amount of cat mRNA produced by the plasmids was determined by the reverse transcriptase assay, as in Figure 5B. The efficiencies of transcriptional readthrough were calculated and compared as in Figure 5C. The error bars reflect the standard deviation from the mean of three independent experiments.