Whole set of constitutive promoters for RpoN sigma factor and the regulatory role of its enhancer protein NtrC in Escherichia coli K-12

Abstract

The promoter selectivity of Escherichia coli RNA polymerase (RNAP) is determined by its promoter-recognition sigma subunit. The model prokaryote E. coli K-12 contains seven species of the sigma subunit, each recognizing a specific set of promoters. Using genomic SELEX (gSELEX) screening in vitro, we identified the whole set of 'constitutive' promoters recognized by the reconstituted RNAP holoenzyme alone, containing RpoD (σ⁷⁰), RpoS (σ³⁸), RpoH (σ³²), RpoF (σ²⁸) or RpoE (σ²⁴), in the absence of other supporting regulatory factors. In contrast, RpoN sigma (σ⁵⁴), involved in expression of nitrogen-related genes and also other cellular functions, requires an enhancer (or activator) protein, such as NtrC, for transcription initiation. In this study, a series of gSELEX screenings were performed to search for promoters recognized by the RpoN RNAP holoenzyme in the presence and absence of the major nitrogen response enhancer NtrC, the best-characterized enhancer. Based on the RpoN holoenzyme-binding sites, a total of 44 to 61 putative promoters were identified, which were recognized by the RpoN holoenzyme alone. In the presence of the enhancer NtrC, the recognition target increased to 61-81 promoters. Consensus sequences of promoters recognized by RpoN holoenzyme in the absence and presence of NtrC were determined. The promoter activity of a set of NtrC-dependent and -independent RpoN promoters was verified in vivo under nitrogen starvation, in the presence and absence of RpoN and/or NtrC. The promoter activity of some RpoN-recognized promoters increased in the absence of RpoN or NtrC, supporting the concept that the promoter-bound NtrC-enhanced RpoN holoenzyme functions as a repressor against RpoD holoenzyme. Based on our findings, we propose a model in which the RpoN holoenzyme fulfils the dual role of repressor and transcriptase for the same set of genes. We also propose that the promoter recognized by RpoN holoenzyme in the absence of enhancers is the 'repressive' promoter. The presence of high-level RpoN sigma in growing E. coli K-12 in rich medium may be related to the repression role of a set of genes needed for the utilization of ammonia as a nitrogen source in poor media. The list of newly identified regulatory targets of RpoN provides insight into E. coli survival under nitrogen-depleted conditions in nature.

Keywords: Escherichia coli; NtrC; RNA polymerase; RpoN sigma factor; gSELEX; nitrogen metabolism.

Fig. 1.

gSELEX-chip search for the binding sequences of the RpoN RNAP holoenzyme on the E. coli K-12 genome. gSELEX was performed to search for the binding sites of the RNAP RpoN holoenzyme. The y-axis represents the ratio against the highest peak at the potF promoter region and shows the level of RpoN holoenzyme-bound DNA fragments, whereas the x-axis represents the position on the E. coli K-12 genome in bp. The adjacent gene on the E. coli K-12 genome of the peak position was indicated for high intensity peaks (>60%). A list of binding sites of the RpoN holoenzyme is provided in Table 1 (detailed in Table S2).

Fig. 2.

gSELEX-chip search for the binding sequences of NtrC on the E. coli K-12 genome. gSELEX was performed to search for the binding sites of NtrC, in the presence of acetylphosphate, with respect to NtrC phosphorylation. The y-axis represents the ratio against the highest peak at the tus ORF and shows the level of NtrC-bound DNA fragments, whereas the x-axis represents the position on the E. coli K-12 genome in bp. The adjacent gene on the E. coli K-12 genome of the peak position was indicated for high intensity peaks (>70%). Peaks located within the spacer regions are shown with green labels, while peaks located within ORFs are shown with orange labels. A list of the binding sites of NtrC is provided in Table 2 (detailed in Table S3).

Fig. 3.

gSELEX-chip search for the binding sequences of the RpoN RNAP holoenzyme in the presence of NtrC on the E. coli K-12 genome. gSELEX was performed to search for binding sites of the RpoN holoenzyme in the presence of NtrC. The y-axis represents the ratio against the highest peak at the hyfA promoter region and shows the level of RpoN holoenzyme-bound DNA fragments in the presence of NtrC, whereas the x-axis represents the position on the E. coli K-12 genome in bp. The adjacent gene on the E. coli K-12 genome of the peak position was indicated for high intensity peaks (>60 %). The peaks located within the spacer regions are shown with green labels, while the peaks located within the ORFs are shown with orange labels. A list of the binding sites of the RpoN holoenzyme in the presence of NtrC is described in Table 3 (detailed in Table S4).

Fig. 4.

Correlation diagrams of the targets between RpoN and NtrC. Venn diagram summarizing the correlation of target sites of RpoN holoenzyme and NtrC. The number of binding sites is shown in (a), while the number of regulatory targets is shown in (b). All the 14 sites detected in the RpoN holoenzyme but not in RpoN holoenzyme+NtrC showed over 24% intensity in RpoN holoenzyme+NtrC (for counting the number of targets, the cut-off level was set as 30%) (Table 5).

Fig. 5.

Consensus sequences of the RpoN holoenzyme promoter and NtrC binding. The promoter motif of RpoN holoenzyme, in the presence or absence of NtrC and binding sequences of NtrC, was analysed using the meme program. The sequences are listed in Tables S2–S4, and were subjected to Logo analysis for the determination of the consensus sequences for the following samples: (a) the whole set of RpoN holoenzyme targets (total 71 sequences in Table S2); (b) the whole set of NtrC targets (total 93 sequences in Table S3); (c) the whole set of RpoN holoenzyme targets in the presence of NtrC, not included in the RpoN holoenzyme targets in the absence of NtrC (total 51 sequences in Table S4).

Fig. 6.

NtrC-dependent RpoN holoenzyme–DNA complex formation. The target promoter fragments were mixed with the RpoN holoenzyme (0.3 µM, lane 2), NtrC protein (15 µM, lane 3), or both in combination with the addition of 25 mM acetylphosphate (lane 4). After incubation at 37 °C for 30 min, the reaction mixture was subjected to 3.5% PAGE. Grey triangles indicate the free probe; grey triangles with white frame indicate the NtrC–probe complex; black triangles with white frame indicate the RpoN holoenzyme–probe complex; white triangles indicate the RpoN holoenzyme–NtrC-probe complex.

Fig. 7.

Influence of rpoN or ntrC on the transcription levels of the newly identified NtrC-dependent RpoN targets. Growth curve of E. coli wild-type BW25113, the rpoN-deleted mutant JW3169 and the ntrC-deleted mutant JW3839 in Gutnick medium supplemented with 0.2% glucose with 3 or 20 mM NH₄Cl as the sole nitrogen source (a). The mRNA levels of the known targets (b) and new targets (c, d) of the wild-type, rpoN mutation and ntrC mutation strains under nitrogen depletion. Total RNA was prepared from the wild-type, rpoN mutation and ntrC mutation strains in the exponential phase (OD₆₀₀ 0.4) and the stationary phase (20 min after growth stopped) in Gutnick medium supplemented with 0.2% glucose and 3 mM NH₄Cl and, subsequently, it was subjected to RT-qPCR analysis. RT-qPCR was repeated at least three times, and the mean values of the experiments are shown. The y-axis represents the relative level of mRNA of each NtrC-dependent RpoN target gene between the nitrogen replete and deplete conditions, with the ratio of 16S rRNA used as an internal control in each strain. nd indicates it was not detected as below the detection limit.