The ability of Escherichia coli to use both nitrate and nitrite as terminal electron acceptors during anaerobic growth is mediated by the dual-acting two-component regulatory systems NarX-NarL and NarQ-NarP. In contrast, Neisseria gonorrhoeae responds only to nitrite: it expresses only NarQ-NarP. We have shown that although N. gonorrhoeae NarQ can phosphorylate E. coli NarL and NarP, the N. gonorrhoeae NarP is unable to regulate gene expression in E. coli. Mutagenesis experiments have revealed residues in E. coli NarQ that are essential for nitrate and nitrite sensing. Chimaeric proteins revealed domains of NarQ that are important for ligand sensing.
In Escherichia coli, the global regulatory protein FNR (fumarate and nitrate reductase regulator) regulates gene expression in response to oxygen limitation . Gene expression is further regulated by the dual-acting two-component regulatory systems NarX-NarL and NarQ-NarP in response to the presence of nitrate and nitrite [2,3]. Neisseria gonorrhoeae encodes a homologue of FNR, but only NarQ-NarP. The gonococcus is insensitive to nitrate but can regulate the expression of genes in response to nitrite [4,5]. E. coli NarX and NarQ are membrane-spanning histidine kinases. They contain a periplasmic 18 amino acid region called the P-box that has been implicated in the binding of nitrate and nitrite . Once the ligand is bound, the histidine kinase phosphorylates its cognate response regulator, which regulates the expression of many genes. The P-boxes of NarX and NarQ differ by three amino acids. Mutations in the P-box of NarX and NarQ result in sensor proteins that either no longer respond to nitrate or nitrite or have an altered response [7,8]. However, no mutation has been made that has enabled the protein to sense nitrite but not nitrate. The first aim of the present study was to investigate whether the gonococcal NarQP proteins could function in E. coli. The second aim was to change the E. coli NarQ P-box to the sequence of the gonococcal NarQ P-box to see whether the E. coli NarQ would lose its response to nitrate. Lastly, chimaeric proteins were constructed to investigate the importance of the different domains of NarQ in ligand sensing.
Materials and methods
Media and growth of strains
Strains used during the present study were JCB391 (Δlac narX narL narQ pcnB) and JCB12 (Δlac frdA::lacZ narX narQ, where frd is fumarate reductase). Strain JCB391, transformed with plasmid p7150, which encodes the nirB (NADH-dependent nitrite reductase B) promoter fused to lacZ, was grown anaerobically at 37°C in LB (Luria–Bertani) medium+0.4% glucose and either 20 mM nitrate or 2.5 mM nitrite. Strain JCB12 was grown anaerobically at 30°C in a minimal medium supplemented with 40 mM fumarate, 10% LB medium, 0.4% glycerol and either 20 mM nitrate or 2.5 mM nitrite. For β-galactosidase assays, the method described in  was used.
Plasmid construction and mutagenesis
The E. coli and gonococcal narQ genes were subcloned individually into the pBAD/myc-hisA plasmid (Invitrogen) to make plasmids pBADecQ and pBADgcQ. The gonococcal narQ and narP genes were subcloned together into the pBAD/myc-hisA plasmid to make plasmid pBADgcQP. The QuikChange Site-Directed Mutagenesis kit from Stratagene was used for the site-directed mutagenesis of the P-box of the E. coli narQ gene.
Chimaeric protein construction
Genes encoding eight chimaeric proteins with fusions within the P′-box (pRNW200), TMII (second transmembrane region) (pRNW300), HAMP (named after histidine kinases, adenylate cyclases, methyl-accepting proteins and phosphatases) linker (pRNW400) [10,11] and Y-box (pRNW500)  were constructed by exchanging different segments of narQ from E. coli and the gonococcus (Figure 1). For the HAMP-linker chimaera, an NdeI site was introduced at codon 217 of the E. coli narQ (resulting in the substitution E217H) and the corresponding codon 228 in the gonococcal narQ gene (resulting in the substitution E228H). The Y-box chimaeric proteins were constructed by introducing a BstBI site at codon 229 of E. coli narQ (resulting in the substitution L229F) and codon 240 of the gonococcal narQ (resulting in the substitution L240F). The TMII chimaeric proteins were constructed by introducing a BamHI restriction site at codons 167 and 168 of E. coli narQ (resulting in the substitutions F167W and T168I) and codons 179 and 180 of the gonococcal narQ (resulting in the substitutions L179R and M180I). The P′-box chimaeric proteins were constructed by introducing a BamHI restriction site at codons 143 and 144 of E. coli narQ (resulting in substitutions A143G and E144C) and codons 155 and 156 of gonococcal narQ (resulting in substitutions G155E and E156C).
Construction of chimaeric proteins
The effects initiated at the E. coli nirB promoter by the gonococcal NarQ and NarP
To investigate whether the gonococcal NarQ and NarP proteins could activate transcription at the E. coli nirB promoter, strain JCB391 was transformed with the reporter plasmid p7150 and either pBADgcQ or pBADgcQP. The transformants were grown in LB medium anaerobically with or without nitrate or nitrite and the β-galactosidase activities were measured. The untransformed strain gave low levels of activity under all three conditions due to the absence of a histidine kinase. Transformation with the plasmid encoding the wild-type gonococcal NarQ gave high activity under all three conditions, a phenotype known as ‘locked-on’, so the gonococcal NarQ must be able to phosphorylate E. coli NarP. The strain expressing both the gonococcal NarQ and NarP proteins gave similar activities to the untransformed strain, indicating not only that the gonococcal NarP could not activate transcription at the E. coli nirB promoter, but that it also prevents gonococcal NarQ from phosphorylating E. coli NarP.
Effect of mutating the E. coli NarQ P-box
The wild-type E. coli NarQ protein gave a small activation in the presence of nitrite and a 4-fold increase in activation in the presence of nitrate. Mutations to change the E. coli narQ P-box sequence to the gonococcal narQ P-box sequence resulted in a NarQ protein that was insensitive to both nitrate and nitrite. This ‘locked-off’ phenotype indicates that the P-box is essential for nitrite and nitrate sensing. The E. coli NarQ with the entire gonococcal P-box was unable to sense nitrite, indicating that the P-box alone is insufficient for sensing or discriminating between the ligands.
Roles for the HAMP linker and Y-box in signal transduction
Strain JCB12 was transformed with plasmids encoding wild-type E. coli NarQ, wild-type gonococcal NarQ, all eight chimaeric proteins and the wild-type proteins containing the restriction site mutations. The β-galactosidase activity reporting transcription at the chromosomal frdA::lacZ fusion was measured (Figure 2). The wild-type E. coli NarQ responded to the presence of nitrate and phosphorylated NarL, causing repression of the frdA promoter. The gonococcal NarQ phosphorylated NarL even in the absence of ligand (the ‘locked-on’ phenotype). Introduction of the restriction sites had no effect on the gonococcal NarQ protein, but substitutions resulting from the introduction of restriction sites at TMII, the HAMP linker and the Y-box increased the sensitivity of E. coli NarQ to nitrite. The restriction site in the P′-box decreased the sensitivity to nitrate. All of the chimaeras with an E. coli N-terminus and gonococcal C-terminus were ‘locked-on’. The chimaeras with a gonococcal N-terminus and E. coli C-terminus and junctions in the P′-box or TMII were ‘locked-off’ or responded weakly to nitrate. The chimaeras with the junction in the HAMP linker or Y-box were ‘locked-on’.
Effects of chimaeric NarQ proteins at the E. coli frdA promoter
Three conclusions can be drawn from the results: (i) the gonococcal NarQ is a ligand-insensitive kinase; (ii) the sequence of the HAMP linker determines the response of E. coli NarQ to nitrite; and (iii) the Y-box and HAMP linker of E. coli NarQ must be intact for NarL to be dephosphorylated (either by suppressing the kinase activity of NarQ or by dephosphorylation of NarL-phosphate).
The 11th Nitrogen Cycle Meeting 2005: Independent Meeting held at Estación Experimental del Zaidín, Granada, Spain, 15–17 September 2005. Organized and Edited by E.J. Bedmar (Granada, Spain), M.J. Delgado (Granada, Spain) and C. Moreno-Vivián (Córdoba, Spain).