The Phosphohistidine Phosphatase SixA Targets a Phosphotransferase System

Abstract

SixA, a well-conserved protein found in proteobacteria, actinobacteria, and cyanobacteria, is the only reported example of a bacterial phosphohistidine phosphatase. A single protein target of SixA has been reported to date: the Escherichia coli histidine kinase ArcB. The present work analyzes an ArcB-independent growth defect of a sixA deletion in E. coli A screen for suppressors, analysis of various mutants, and phosphorylation assays indicate that SixA modulates phosphorylation of the nitrogen-related phosphotransferase system (PTS^Ntr). The PTS^Ntr is a widely conserved bacterial pathway that regulates diverse metabolic processes through the phosphorylation states of its protein components, EI^Ntr, NPr, and EIIA^Ntr, which receive phosphoryl groups on histidine residues. However, a mechanism for dephosphorylating this system has not been reported. The results presented here suggest a model in which SixA removes phosphoryl groups from the PTS^Ntr by acting on NPr. This work uncovers a new role for the phosphohistidine phosphatase SixA and, through factors that affect SixA expression or activity, may point to additional inputs that regulate the PTS^NtrIMPORTANCE One common means to regulate protein activity is through phosphorylation. Protein phosphatases exist to reverse this process, returning the protein to the unphosphorylated form. The vast majority of protein phosphatases that have been identified target phosphoserine, phosphotheronine, and phosphotyrosine. A widely conserved phosphohistidine phosphatase was identified in Escherichia coli 20 years ago but remains relatively understudied. The present work shows that this phosphatase modulates the nitrogen-related phosphotransferase system, a pathway that is regulated by nitrogen and carbon metabolism and affects diverse aspects of bacterial physiology. Until now, there was no known mechanism for removing phosphoryl groups from this pathway.

Keywords: CvrA; PtsN; YcgO; histidine phosphatase; histidine phosphorylation.

FIG 1

Cells lacking SixA have a growth defect that is independent of ArcB/ArcA. (A and B) Growth curves were measured for the wild-type (MG1655) and ΔsixA (JES13) strains in (A) glycerol minimal medium and (B) LB Miller medium. Optical density at 600 nM (OD₆₀₀) was monitored for two independent cultures of each strain. Symbols represent the average OD₆₀₀; bars indicate ranges and are not visible where smaller than the symbol. (C) Cultures of the wild-type (MG1655) and ΔsixA (JES13) strains, transformed with empty vector pTrc99a, SixA-expressing plasmid pSixA, or SixA(H8A)-expressing plasmid pSixA(H8A) were grown in glycerol minimal medium with ampicillin for 12 h. Symbols represent the OD₆₀₀ for individual cultures, and the horizontal black lines indicate the average OD₆₀₀ from three biological replicates. (D) Strains were grown in glycerol minimal medium for 12 h. Symbols are as described for panel C. Strains are MG1655, JES13, JES47, JES48, JES49, and JES50. (E) β-Galactosidase activity was measured for strains containing a P_1icdA-lacZ reporter following anaerobic growth in minimal medium with 40 mM KNO₃. Symbols represent the activity (Miller units) measured for individual cultures, error bars report standard deviations, and the blue bars indicate the average levels of activity from the three biological replicates. Strains are JES252, JES253, and JES254. As shown in the illustration to the right of the graph, SixA was previously proposed to dephosphorylate ArcB.

FIG 2

Suppressors of the ΔsixA growth defect. (A) Growth curves were measured for strains in glycerol minimal medium. Optical density at 600 nM (OD₆₀₀) was measured for two independent cultures of each strain. Symbols represent the average OD₆₀₀; bars indicate ranges and are not visible where smaller than the symbol. Samples for this growth curve were collected at the same time as the samples for the growth curve in Fig. 3B, so the data for the wild-type and ΔsixA strains are identical between figures. Strains are MG1655, JES13, JES30, and JES31. (B) Growth curves were measured for strains transformed with empty vector pMG91 or YcgO-expressing plasmid pYcgO in glycerol minimal medium with chloramphenicol. Symbols are as described for panel A. Strains are MG1655, JES13, and JES31. (C) Cultures of the wild-type (MG1655) and ΔsixA (JES13) strains, transformed with empty vector pTrc99a or ptsN-3×FLAG-expressing plasmid pEIIA^Ntr, were grown in glycerol minimal medium with ampicillin for 13 h. Symbols represent the OD₆₀₀ for individual cultures, and the horizontal black lines indicate the average OD₆₀₀ from three biological replicates. (D) EIIA^Ntr is part of the nitrogen-related phosphotransferase system (PTS^Ntr). Phosphoryl groups entering the system originate from phosphoenolpyruvate and are passed by successive phosphotransfers between the three PTS^Ntr proteins. According to a previously proposed model (13), unphosphorylated EIIA^Ntr inhibits YcgO, decreasing potassium efflux. (E) Exponential-phase cultures, grown in glycerol minimal medium, were assayed for potassium by inductively coupled plasma mass spectrometry (ICP-MS). Symbols represent the cellular potassium content for individual cultures, error bars report standard deviations, and the purple bars indicate the average cellular potassium content from all biological replicates. Strains are MG1655, JES208, JES13, and JES31. The average cellular potassium content for the wild-type strain is 243 nmol K⁺/OD₆₀₀.

FIG 3

Modulation of the PTS^Ntr by SixA. (A) Western blot of strains expressing a 3×FLAG-tagged EIIA^Ntr from the native ptsN locus and containing empty vector pTrc99a, SixA-expressing plasmid pSixA, or SixA(H8A)-expressing plasmid pSixA(H8A). Cells were grown in LB Miller medium with ampicillin to stationary phase. Extraneous lanes from the blot are not shown. Strains are JES211 and JES212. The Western blot shown is representative of blots from three independent experiments. (B) Growth curves were measured for strains in glycerol minimal medium. Optical density at 600 nM (OD₆₀₀) was monitored for two independent cultures of each strain. Symbols represent the average OD₆₀₀; bars indicate ranges and are not visible where smaller than the symbol. Samples for this growth curve were collected at the same time as samples for the growth curve in Fig. 2A, so the data for the wild-type and ΔsixA strains are identical between figures. Strains are MG1655, JES13, JES208, and JES264. (C) Strains were grown in glycerol minimal medium for 12 h. Symbols represent the OD₆₀₀ for individual cultures, and the horizontal black lines indicate the average OD₆₀₀ from three biological replicates. Strains are MG1655, JES13, JES189, JES190, JES185, and JES186. (D) Schematic showing the phosphoryl group transfers within the PTS^Ntr, potential cross-phosphorylation from the PTS^sugar, and proposed dephosphorylation by SixA. Phosphoryl groups from phosphoenolpyruvate (PEP) are passed by successive phosphotransfers between proteins in the PTS^sugar and the PTS^Ntr. Dotted arrows show cross talk between the PTS^sugar and the PTS^Ntr that has been reported to occur either in vitro or in vivo in some mutant backgrounds. Whereas the PTS^sugar ultimately uses phosphoryl group transfer to facilitate uptake of carbohydrates such as glucose, there are no known phosphoryl group acceptors for the PTS^Ntr. We propose that SixA dephosphorylates NPr, providing a phosphoryl group sink for the PTS^Ntr. (E) Strains were grown in glucose minimal medium for 8 h. Symbols are as described for panel C. Strains are MG1655, JES13, JES189, JES190, JES185, and JES186. (F) Western blot of strains expressing a 3×FLAG-tagged EIIA^Ntr from the native ptsN locus and containing empty vector pTrc99a or SixA-expressing plasmid pSixA. Cultures were grown in LB Miller medium with ampicillin to stationary phase. Strains are JES211 and JES271. The Western blot shown is representative of blots from three independent experiments.