Mutations in the CRE pocket of bacterial RNA polymerase affect multiple steps of transcription

Abstract

During transcription, the catalytic core of RNA polymerase (RNAP) must interact with the DNA template with low-sequence specificity to ensure efficient enzyme translocation and RNA extension. Unexpectedly, recent structural studies of bacterial promoter complexes revealed specific interactions between the nontemplate DNA strand at the downstream edge of the transcription bubble (CRE, core recognition element) and a protein pocket formed by core RNAP (CRE pocket). We investigated the roles of these interactions in transcription by analyzing point amino acid substitutions and deletions in Escherichia coli RNAP. The mutations affected multiple steps of transcription, including promoter recognition, RNA elongation and termination. In particular, we showed that interactions of the CRE pocket with a nontemplate guanine immediately downstream of the active center stimulate RNA-hairpin-dependent transcription pausing but not other types of pausing. Thus, conformational changes of the elongation complex induced by nascent RNA can modulate CRE effects on transcription. The results highlight the roles of specific core RNAP-DNA interactions at different steps of RNA synthesis and suggest their importance for transcription regulation in various organisms.

Figure 1.

RNAP–CRE interactions in the open promoter complex. (A) Structure of T. thermophilus RNAP with a promoter DNA fragment (4Q4Z, (13)). The β flap and β′ clamp domains are indicated. The σ subunit is not shown. (B) RNAP–DNA interactions at the downstream edge of the transcription bubble. The active site Mg²⁺ ions are shown as purple spheres; the +1 and +2 initiating nucleotides are shown in orange and yellow, respectively. The template and nontemplate DNA strands are black and blue, respectively; DNA nucleotides at promoter position +2 are shown in red, nontemplate thymine at position +1 is dark blue. β subunit segments forming the CRE pocket are shown in green (168–176; 443–451; 533–546, E. coli numbering); residues W183, D446 and E546 are shown as stick models.

Figure 2.

Effects of mutations in the CRE pocket on promoter complex stability. (A) Outline of experiment. Heparin was added to preformed promoter complexes, followed by nucleotide addition after increasing time intervals. The reactions were performed with CpA+(UTP/GTP) in the case of T7A1 and CpA+(CTP/GTP)+UTP in the case of rrnB P1* promoter variants (with addition of α-[³²P]-UTP), resulting in the synthesis of 3 and 4 nt RNA products, respectively. (B) Kinetics of promoter complex dissociation for the wild-type and +2G T7A1 promoter variants, measured for wild-type and D446A RNAPs. Averages and standard deviations from two to three independent experiments are shown; the lines correspond to single-exponential fits of the data.

Figure 3.

Comparison of elongation rates of mutant RNAPs. (A) Outline of experiment. (B) Kinetics of accumulation of full-length RNA transcript (in percent of the activity measured at the 10 min time point) for wild-type and mutant RNAP variants (averages and standard deviations from two to three independent experiments).

Figure 4.

Analysis of hairpin-dependent pausing by wild-type and mutant RNAPs. (A) Structure of the hisP transcription template. The major pausing position is indicated below the RNA transcript. The consensus pause-inducing sequence (CP) identified in genome-wide experiments is shown in blue in the same register relative to the pause position. +1G is shown in yellow; the +1C template contained a C:G base pair instead of G:C at this position. (B) Outline of experiment. (C) Kinetics of the pause decay measured on the wild-type and +1C templates for wild-type and D446A RNAPs. Averages and standard deviations from three to four independent experiments are shown; the lines are single-exponential fits of the data.

Figure 5.

Role of RNA duplex in RNAP pausing at the hisP site. (A) Structure of the scaffold template used in the experiments. RNA is red, with the four 3′-terminal nucleotides added during transcription shown in lowercase letters. Positions of the 3′-end in the starting, paused and read-through (R) transcripts are indicated. Antisense RNA (asRNA) oligonucleotide is orange. The +1 G:C base pair is shown in yellow. (B) Outline of experiment. (C) Pausing kinetics measured in the absence or presence of asRNA for wild-type and D446A RNAPs. (D) Pause half-life times and pause efficiencies (P_max, predicted pausing at zero time point) determined from the pause decay curves.

Figure 6.

In vivo analysis of the effects of CRE pocket mutations on cell viability and Rif resistance. (A) E. coli temperature-sensitive strain RL585 was transformed with plasmids bearing inducible copies of wild-type and mutant rpoB genes or a control plasmid (first row) and grown at either restrictive (left) or permissive (right) temperature. (B) The same plasmids were transformed into E. coli strain DH5α and the cells were grown at 37°C in the absence or presence of Rif.

Figure 7.

CRE–RNAP interactions at different steps of transcription. RNAP is schematically shown as a yellow oval; the CRE pocket is green; the β flap is shown as a blue semi-oval; the downstream DNA binding channel is gray; the nontemplate guanine downstream of the active site is red. The i, i+1 and i+2 positions are indicated. (A) +2G bound in the CRE pocket stabilizes the open promoter complex. (B) Consensus pause complexes are stabilized in the pre-translocated conformation defined by several conserved nucleotides, including a G:C pair at the upstream boundary of the RNA:DNA hybrid, pyrimidine:purine pair (Y:R) in the i+1 site and a G:C pair downstream of the active site. (C) Interactions of +1G with the CRE pocket promote forward EC translocation and suppress consensus pausing. (D) RNA hairpins or duplexes formed in the RNA exit channel modulate the effects of +1G/CRE-pocket interactions, likely through changes in the clamp conformation and positions of the downstream duplex and the nontemplate DNA strand. This results in stabilization of the paused state, either by preventing EC translocation (upper) or by inhibiting nucleotide addition in the post-translocated state (lower).