Crl facilitates RNA polymerase holoenzyme formation

Abstract

The Escherichia coli Crl protein has been described as a transcriptional coactivator for the stationary-phase sigma factor sigma(S). In a transcription system with highly purified components, we demonstrate that Crl affects transcription not only by the Esigma(S) RNA polymerase holoenzyme but also by Esigma(70) and Esigma(32). Crl increased transcription dramatically but only when the sigma concentration was low and when Crl was added to sigma prior to assembly with the core enzyme. Our results suggest that Crl facilitates holoenzyme formation, the first positive regulator identified with this mechanism of action.

FIG. 1.

Crl activates transcription when preincubated with σ^S or σ⁷⁰ before the addition of core RNAP. (A) Schematic representation of the promoters and transcription terminators on plasmid pBP5. The cc-35con transcript is terminated by the T7 Te terminator, the rrnB P1 and rrnB P2 transcripts are terminated by the rrnB T₁T₂ terminators, and the RNA-I transcript is terminated by the RNA-I terminator. (B) Transcription in the presence or absence of Crl. Lanes 1, 2, 5, and 6, 16 μM Crl, or buffer as a control, was incubated with 32 nM σ^S or σ⁷⁰ in transcription buffer (40 mM Tris-Cl, pH 7.9, 100 mM KCl, 10 mM MgCl₂, 1 mM dithiothreitol, 100 μg/ml BSA) for 10 min. This mixture was added to an equal volume of 333 nM core RNAP for 10 min, and then 2 μl of this reaction mixture was added to 8 μl transcription buffer containing 200 μM ATP, GTP, CTP, and 10 μM UTP with 1 μCi [α-³²P]CTP and the plasmid template. All steps were performed at 22°C. The final concentrations of σ, core RNAP, and Crl in the transcription reaction mixture were therefore 3.2 nM, 33.4 nM, and 1.6 μM, respectively. The reactions were terminated after 15 min and examined on 5.5% polyacrylamide denaturing gels (10). Lanes 3, 4, 7, and 8, holoenzyme was formed by mixing core RNAP with σ⁷⁰ or σ^S for 20 min, and then the reaction mixture was diluted into transcription buffer with or without 8 μM Crl for 10 min. Two microliters of this reaction mixture was then used for each 10-μl transcription reaction mixture (as indicated above). The final concentrations of σ, core RNAP, and Crl per reaction were therefore the same as in the above-described reactions. Care was always taken to keep reaction conditions and times of incubation exactly the same when samples were compared in the order-of-addition experiments. High σ and/or Crl concentrations resulted in radioactivity that remained in the sample wells. This may derive from protein aggregation, since it was partially eliminated by heat treatment (5 to 10 min at 95°C). In addition, the responses of additional RNA species migrating more slowly than the test transcripts also suggest that there are effects of Crl on the plasmid-derived RNA-II promoter, on the bla promoter, and/or on termination readthrough products (Fig. 2E and 3A; data not shown). (C) Transcripts were quantified by phosphorimaging and are presented as ratios with or without Crl. Data from two experiments are averaged, and the ranges are shown.

FIG. 2.

Crl activates transcription only at low σ concentrations. Representative gels are shown for illustrative purposes, but each experiment was performed at least three times with very similar results. (A) σ^S-dependent transcription. Transcription was performed by incubation of Crl (8 μM) with the indicated concentrations of σ^S before addition to core RNAP (333 nM) for 10 min. Two microliters of this mixture was used in each 10-μl transcription reaction mixture, as for Fig. 1. (B) Quantitation of the data from panel A. (C) σ⁷⁰-dependent transcription. Transcription was performed by incubation of Crl (8 μM) with the indicated concentrations of σ⁷⁰ before addition to core RNAP (88 nM) for 10 min. Two microliters of this mixture was used in each 10-μl transcription reaction mixture, as for Fig. 1. (D) Quantitation of the data from panel C. (E) σ³²-dependent transcription. Transcription was performed by incubation of Crl (8 μM) with the indicated concentrations of σ³² before addition to core RNAP (333 nM) for 10 min. Two microliters of this mixture was used in each 10-μl transcription reaction mixture. The template was plasmid pJET40, a pUC19 derivative encoding the dnaK P1 promoter as well as the RNA-I promoter (22). A control reaction was performed with Eσ⁷⁰ (left lane). This holoenzyme produced little or no transcript from the dnaK P1 promoter, and the Eσ³² preparation did not make a transcript corresponding to RNA-I, confirming that Crl affected Eσ³²-dependent transcription. (F) Quantitation of the data from panel E.

FIG. 3.

Concentration dependence of activation of transcription by Crl. σ (8 nM) and Crl (0 to 16 μM) were incubated for 10 min and added to an equal volume of core RNAP (167 nM). The plasmid template and other components of the transcription reaction mixture were added, and transcripts were measured as described above. A representative gel is shown for each form of the holoenzyme, and the ratio of transcription from the RNA-I promoter in these reactions with or without Crl at different Crl concentrations is quantified below for each gel. (A) σ^S. (B) σ⁷⁰. At very high concentrations of Crl, we sometimes observed inhibition of the rrnB P1 promoter. However, we did not observe effects of inactivation of the crl gene on rRNA promoter activity in vivo (data not shown), suggesting that this phenomenon is unlikely to be physiologically relevant.