Activation by NarL at the Escherichia coli ogt promoter

Abstract

The Escherichia coli NarX/NarL two-component response-regulator system regulates gene expression in response to nitrate ions and the NarL protein is a global transcription factor, which activates transcript initiation at many target promoters. One such target, the E. coli ogt promoter, which controls the expression of an O6-alkylguanine-DNA-alkyltransferase, is dependent on NarL binding to two DNA targets centred at positions -44.5 and -77.5 upstream from the transcript start. Here, we describe ogt promoter derivatives that can be activated solely by NarL binding either at position -44.5 or position -77.5. We show that NarL can also activate the ogt promoter when located at position -67.5. We present data to argue that NarL-dependent activation of transcript initiation at the ogt promoter results from a direct interaction between NarL and a determinant in the C-terminal domain of the RNA polymerase α subunit. Footprinting experiments show that, at the -44.5 promoter, NarL and the C-terminal domain of the RNA polymerase α subunit bind to opposite faces of promoter DNA, suggesting an unusual mechanism of transcription activation. Our work suggests new organisations for activator-dependent transcription at promoters and future applications for biotechnology.

Keywords: NarL; molecular interactions; promoters; transcription activation.

Figure 1.. Activities of the ogt promoter derivatives used in this study.

(A) The panel shows the base sequence of the E. coli K-12 ogt promoter fragment from position −105 to +1 relative to the transcript start site (+1), which is indicated by a bent arrow. Inverted arrows show the two DNA sites for NarL and the centre of each site is indicated. The −10, extended −10 and −35 elements are underlined and in bold. (B) A schematic representation of the ogt100, ogt102, ogt1052, ogt104 and ogt1041 promoter fragments used in this study. The NarL binding sites are shown as inverted arrows, −35 and −10 promoter elements are shown as boxes and the transcript start site (+1), is indicated by a bent arrow. Improved NarL binding sites are shown as filled black arrows. (C) The panel shows base sequences of the DNA sites for NarL in the ogt100, ogt102, ogt1052, ogt104 and ogt1041 promoter fragments. The NarL 7-2-7 consensus binding sequence is shown in boldface above (Y = C/T, M = A/C, K = G/T, R = A/G) [23]. The underlined sequences indicate the bases that have been changed in each promoter fragment. (D) The panel shows β-galactosidase activities measured in wild type E. coli K-12 JCB387 cells, carrying the ogt100, ogt102, ogt1052, ogt104 and ogt1041 promoters as lacZ fusions, cloned into expression vector pRW50 [17]. Cells were grown in minimal salts media supplemented with 20 mM sodium nitrate, where indicated. β-galactosidase activities are expressed as nmol ONPG hydrolysed min⁻¹ mg⁻¹ dry cell mass and represent the average of three independent experiments.

Figure 2.. NarL- and NarP-dependent activation of ogt promoter fragments.

The figure shows β-galactosidase activities measured in wild type JCB387, JCB3875 (ΔnarP), JCB3883 (ΔnarL) and JCB3884 (ΔnarL ΔnarP) cells, carrying the (A) ogt100, (B) ogt1052 and (C) ogt1041 promoters, cloned into the lacZ expression vector pRW50. (D) The panel details β-galactosidase activities measured in JCB3884 (ΔnarL ΔnarP) cells carrying the ogt1052 promoter, cloned into pRW50, with either pLG339 (empty vector control) or pDLC5 (narXL). For all panels, cells were grown in minimal salts media supplemented with 20 mM nitrate, where indicated (black-filled bars). β-galactosidase activities are expressed as nmol ONPG hydrolysed min⁻¹ mg⁻¹ dry cell mass and represent the average of three independent experiments.

Figure 3.. Position-dependent effects of altering the location of NarL I at the ogt1041 promoter.

(A) The panel shows the partial base sequences of promoters, based on the ogt1041 promoter, in which the improved NarL I binding site was moved to different positions. Inverted arrows show the location of NarL I in each fragment and the relevant sequence is bold and underlined. The sequence is numbered with respect to the transcript start site (+1), which is indicated by a bent arrow. The extended −10 and −35 elements are also indicated and in bold. (B) The panel shows β-galactosidase activities measured in wild type E. coli K-12 JCB387 cells, carrying the promoters detailed in (A) as lacZ fusions, cloned into expression vector pRW50 [17]. Cells were grown in minimal salts media supplemented with 20 mM sodium nitrate, where indicated. β-galactosidase activities are expressed as nmol ONPG hydrolysed min⁻¹ mg⁻¹ dry cell mass and represent the average of three independent experiments.

Figure 4.. Residues of αCTD required for transcription activation at ogt promoter derivatives.

The figure shows β-galactosidase activities measured in JCB3875 (ΔnarP) cells that carried either (A) ogt100, (B) ogt1052 or (C) ogt1041 cloned into pRW50, together with pHTf1 or pREII plasmids, encoding derivatives of α with a single alanine substitution (residues 255−329). Cells were grown in minimal salts media supplemented with 20 mM sodium nitrate and promoter activities are presented as percentages of the activity measured in cells that carry plasmids encoding the wild type α subunit.

Figure 5.. Epistasis studies of the interaction between NarL and the αCTD of RNAP at the ogt1052 promoter.

Panels (A) and (B) show β-galactosidase activities measured in JCB3884 (ΔnarL ΔnarP) cells, which contain pRW50/ogt1052 and pDLC5, carrying various NarL derivatives. In (B) cells also contain either pREIIα (encoding the wild type α subunit) or pREIIα-273 (encoding an α derivative with an alanine substitution at residue 273). In both cases, cells were grown in minimal salts media supplemented with 20 mM nitrate. β-galactosidase activities are expressed as nmol ONPG hydrolysed min⁻¹ mg⁻¹ dry cell mass and represent the average of three independent experiments. In panel (B), * indicates P < 0.05 and ** P = 0.05 (Student's t-test).

Figure 6.. In vitro analysis of NarL and RNAP binding to the ogt1052 p35T promoter.

(A) The binding of NarL and RNAP to the ogt1052 p35T promoter was investigated using DNAse I footprinting. Different concentrations of phospho-NarL and RNAP were incubated with P³² end-labelled ogt1052 p35T promoter fragment and treated with DNAse I. The concentrations of NarL was as follows: lanes 1, 6 and 7, no protein; lane 2, 0.4 µM; lane 3, 0.8 µM; lanes 4 and 8 to 10, 1.6 µM; lane 5, 3.2 µM. The concentrations of RNAP holoenzyme was: lanes 1 to 6, no protein; lane 7 and 10, 160 nM; lane 8, 40 nM; lane 9, 100 nM. The extent of NarL and RNAP protections is indicated by grey boxes and hypersensitivity sites are starred. (B) The panel shows DNA cleavage patterns resulting from an FeBABE experiment with the P³² end-labelled ogt1052 p35T promoter fragment and RNAP carrying an FeBABE attached to residue 302 of the α subunits of RNA polymerase. The reactions contained the following proteins: Lane 1, RNAP FeBABE and 3.2 µM phosho-NarL; Lane 2, RNAP FeBABE; Lane 3, no protein. The DNA sequence of the NarL site in the ogt1052 p35T fragment is shown and the bases modified by FeBABE footprinting are shaded grey. In both panels, gels were calibrated using Maxam–Gilbert G + A sequence reactions (GA) and relevant positions are indicated.

Figure 7.. Structure-based model of the transcript initiation complex at the ogt promoter.

(A) The panel shows two views of a structure-based model detailing the binding of a NarL dimer, an αCTD of RNAP and Domain 4 of the σ subunit of RNAP (σD4) bound to NarL II at the ogt1052 promoter. The αCTD is located immediately upstream of the boundary of the −35 hexamer, thus, it orients residue E273 to interact with residue R178 of NarL, which is located at position −45.5 on the opposite face of DNA helix from αCTD. This action causes residue 261 of αCTD to point downstream toward the σ subunit and interact with R603 in Domain 4 of the σ subunit [41,42]. The model was created using WebLab Viewer (Accelrys). (B) The panel shows the architectures of (i) simple and (ii) unusual Class I activator-dependent promoters. At ‘classic’ Class I promoters, a single transcription factor, bound to an upstream site, interacts with the αCTD of RNAP on the same face of the DNA helix to recruit RNAP to the promoter. At unusual Class I promoters (e.g. the ogt1052 promoter), transcription factors, such as NarL, bind to the opposite face of the DNA helix to contact αCTD and recruit polymerase.