DNA binding sites for the Mlc and NagC proteins: regulation of nagE, encoding the N-acetylglucosamine-specific transporter in Escherichia coli

Abstract

The NagC and Mlc proteins are homologous transcriptional regulators that control the expression of several phosphotransferase system (PTS) genes in Escherichia coli. NagC represses nagE, encoding the N:-acetylglucosamine-specific transporter, while Mlc represses three PTS operons, ptsG, manXYZ and ptsHIcrr, involved in the uptake of glucose. NagC and Mlc can bind to each others operator, at least in vitro. A binding site selection procedure was used to try to distinguish NagC and Mlc sites. The major difference was that all selected NagC binding sites had a G or a C at positions +11/-11 from the centre of symmetry. This is also the case for most native NagC sites, but not the nagE operator, which thus looks like a potential Mlc target. The nagE operator does exhibit a higher affinity for Mlc than NagC, but no regulation of nagE by physiological concentrations of Mlc was detected in vivo. Regulation of wild-type nagE by NagC is achieved because of the chelation effect due to a second high affinity NagC operator covering the nagB promoter. Replacing the A/T at +11/-11 with C/G allows repression by NagC in the absence of the nagB operator.

Figure 1

Comparison of NagC and Mlc DNA binding motifs and operators. (A) Alignment of the HTH motifs in the N-terminal regions of NagC and Mlc from E.coli and the XylR repressor from Bacillus subtilis. The locations of the motifs relative to the N-terminal of the proteins are given. Amino acids in the recognition helix (helix 2 shown in yellow) are numbered 1–11. Identical amino acids are indicated by shading and conservative replacements by hatching. (B) Alignment of native NagC and Mlc operator sites. The conserved TT/AA bases at positions –5,6/+5,6 are shown in yellow. AT bases in positions 7–10 (NagC) or 7–11 (Mlc) are shown in green, while C or G at positions +11/–11 of NagC are in blue. C or G residues in pink are in positions –4 to +4, which conforms to the CGCGNCGCG pattern, while those in mauve are in the GCGCNGCGC order.

Figure 2

Alignment of NagC (A) and Mlc (B) DNA binding sites selected in this study after nine cycles of selection and amplification. The Mlc binding sequences are derived from two independent selections, new sequences found in the second experiment are indicated by a number greater than 100. Some sequences were selected many times; the total number of times each sequence was found is shown in brackets after the sequence. Bases derived from the N25 randomised starting sequence are shown in upper-case letters while those derived from the flanking oligonucleotide sequences are shown in lower-case. The colour coding of the homologous sequences is as described in Figure 1B. Sequences tested for NagC or Mlc binding in vivo by operator titration (Table 3) are indicated by a cross.

Figure 3

Footprint of NagC and Mlc binding to the Nag14E–Nag15B fragment labelled at Nag15B (A) or Nag14E (B). The DNA was incubated with extracts containing overproduced NagC (lanes A3 and B4), Mlc (lanes A2 and B2) or an empty plasmid vector control (lane A4) (~100 µg/ml) in binding buffer, 50 mM HEPES, 100 mM Na glutamate, pH 8.0, 0.5 mg/ml BSA. Lanes labelled with an F show DNA in the absence of protein extracts. After DNaseI digestion, the reactions were extracted with phenol and analysed on an 8% denaturing polyacylamide gel as described previously (22). The nagE and nagB operators (BoxE and BoxB) are indicated as are the seven hypersensitive cleavages produced by NagC binding. (C) The organisation of the nagE-B promoter region with the location of the NagC operators (BoxE and BoxB) and the CAP site. The sequence is numbered from the nagE transcription site (+1) so that the nagB transcription start is at –133. The 5′ ends of the oligonucleotides are indicated by an asterisk and are at +46 for Nag14E, –118 for Nag42B and –187 for Nag15B. HphI cuts at –56/57 and the XbaI site was created by insertion of 6 bp at –91.

Figure 4

Footprint of NagC and Mlc binding to the single nagB and nagE operators. (A) Labelled DNA fragments (~6 nM) with BoxB, Nag15B*–HphI (lanes 1–5), BoxE, Nag14E*–XbaI (lanes 6–10) and BoxE s.o. Nag14E*–Nag42B with the –11C,+11G mutation (lanes 11–15) were mixed with different amounts of NagC or Mlc extracts. The locations of the protected nagE and nagB operators (BoxE and BoxB) are indicated. The relative concentrations of Mlc and NagC are shown at the top of each lane. ‘1’ corresponds to ~33 µg/ml total extract protein of which Mlc or NagC represents a few percent. (B) The relative protection of each operator by Mlc or NagC was calculated as described in Materials and Methods and is reported as percentage occupancy data for each site by the different amounts of the proteins tested. A dash indicates that that concentration was not tested on the operator indicated. (Note, since the absolute amounts of NagC or Mlc in the extracts are unknown it is only possible to compare the relative binding of the same extract with the different operators and not the binding of NagC and Mlc with the same operator.)