Structural characterization of Escherichia coli sensor histidine kinase EnvZ: the periplasmic C-terminal core domain is critical for homodimerization

Abstract

Escherichia coli EnvZ is a membrane sensor histidine kinase that plays a pivotal role in cell adaptation to changes in extracellular osmolarity. Although the cytoplasmic histidine kinase domain of EnvZ has been extensively studied, both biochemically and structurally, little is known about the structure of its periplasmic domain, which has been implicated in the mechanism underlying its osmosensing function. In the present study, we report the biochemical and biophysical characterization of the periplasmic region of EnvZ (Ala38-Arg162). This region was found to form a dimer in solution, and to consist of two well-defined domains: an N-terminal a-helical domain and a C-terminal core domain (Glu83-Arg162) containing both a-helical and b-sheet secondary structures. Our pull-down assays and analytical ultracentrifugation analysis revealed that dimerization of the periplasmic region is highly sensitive to the presence of CHAPS, but relatively insensitive to salt concentration, thus suggesting the significance of hydrophobic interactions between the homodimeric subunits. Periplasmic homodimerization is mediated predominantly by the C-terminal core domain, while a regulatory function may be attributed mainly to the N-terminal a-helical domain, whose mutations have been shown previously to produce a high-osmolarity phenotype.

Figure 1. Structural characteristics of the EnvZ osmosensor

(A) Schematic representation of subdomain composition of E. coli histidine kinase EnvZ. The conserved active site His²⁴³ and G-boxes in the cytoplasmic domain and the function of each domain are highlighted. (B) Sequence alignment of EnvZ_pers from different γ-proteobacterial species (Shigella flexneri, Salmonella enterica, Yersinia pestis, Pseudomonas luminescens, Proteus vulgaris, Vibrio cholerae and Vibrio vulnificus). Residues that are conserved among the different species are highlighted, and the predicted secondary structure of the periplasmic domain is shown above the sequence (h indicates α-helical structure, and e indicates β-sheets). The predicted N-terminal α-helix (Ser⁴²–Leu⁵⁷) and the trypsin-resistant CT domain (Glu⁸³–Arg¹⁶²) are also indicated.

Figure 2. Biophysical characterization of EnvZ_per

(A) Far-UV CD analysis of 100 μM refolded EnvZ_per in 5 mM sodium acetate, pH 5, and 50 mM NaCl; 5 mM potassium phosphate, pH 6, and 50 mM NaCl; 5 mM potassium phosphate, pH 7, and 50 mM NaCl; and 5 mM potassium phosphate, pH 8, and 50 mM NaCl. (B) Near-UV CD analysis of 300 μM native (5 mM sodium acetate, pH 5, and 50 mM NaCl) and denatured (3 M guanidinium chloride and 50 mM NaCl) EnvZ_per. (C) Intrinsic tryptophan fluorescence analysis of 50 μM EnvZ_per at pH 5–8. (D) Thermal stability of EnvZ_per as measured by a change in molar ellipticity at 222 nm. T_m, melting point.

Figure 3. Structural analysis of EnvZ_per in the presence and absence of CHAPS

(A) Far UV CD analysis of 100 μM EnvZ_per in the presence (◆) and absence (□) of 1 mM CHAPS in 5 mM sodium acetate, pH 5, and 50 mM NaCl buffer. (B) ¹⁵N-edited HSQC spectra of 0.3 mM EnvZ_per in 20 mM sodium acetate, pH 5, and 50 mM NaCl, and (C) 0.9 mM EnvZ_per in 20 mM sodium acetate, pH 5, 50 mM NaCl, and 10 mM CHAPS.

Figure 4. Domain mapping of EnvZ_per

(A) SDS/PAGE analysis of limited tryptic proteolysis of EnvZ_per. EnvZ_per was incubated with trypsin at a 20:1 molar ratio at 4 °C in 20 mM potassium phosphate, pH 8, and 50 mM NaCl, and aliquots were collected at 0, 2, 5,10, 15, 30, 60, 120, 240 and 480 min. (B) Far-UV CD analysis of 100 μM CT domain (■) and 100 μM EnvZ_per (◆) in 5 mM potassium Tris, pH 8, and 50 mM KCl. (C) Thermal stability of EnvZ_per CT domain as measured by a change in molar ellipticity at 215 nm. The denaturation curve is denoted by (■), while the renaturation curve by (▼). The insert is a CD spectrum of the protein sample after thermal denaturation and renaturation. T_m, melting temperature.

Figure 5. Dimerization of EnvZ_per in solution

(A and B) Chromatogram of size-exclusion analysis of 100 μM full-length EnvZ_per and CT domain on a Superdex 75 column respectively. The elution volume of protein standards is indicated on top of the chromatogram. (C and D) SDS/12% PAGE analysis of pull-down assays of His₆-tagged EnvZ_per and UT-EnvZ_per at pH 5. The assays in (C) and (D) were performed at pH 5 at various NaCl concentrations. The assays were performed on the following samples: lane 2, His₆–EnvZ_per; lane 3, UT-EnvZ_per; lanes 4–9, His₆–EnvZ_per and UT-EnvZ_per at the indicated NaCl concentrations. (E) SDS/PAGE analysis of pull-down assay of EnvZ_per in the presence and absence of CHAPS at various pHs. (F) Representative sedimentation equilibrium data obtained for 50 μM EnvZ_per at pH 8 in the presence and absence of 10 mM CHAPS. Only the data at 40000 rev./min is represented.

Figure 6. Role of conserved leucine residues in N-alpha in EnvZ_per oligomerization

(A) Helical wheel representation of the predicted N-terminal α-helix of EnvZ_per. Hydrophobic residues are shaded. Arrows indicate the position of alanine substitution mutations. (B) SDS/PAGE analysis of pull-down assays of mutant EnvZ_per. Assays were performed in 20 mM potassium phosphate, pH 8, and 50 mM NaCl. Quantification of the ratio of UT/His₆-tagged protein by densitometry is represented by the histogram. Results are means±S.D.

Figure 7. Model of osmosensing mediated by N-alpha in EnvZ_per

The CT domain maintains the dimeric structure of EnvZ_per, while the N-alpha dimer interface formed by the putative hydrophobic interaction acts as the signal transducer. In this model, high-osmolarity signals bring the two N-alpha domains close together, and low-osmolarity signals increase the distance between them, resulting in a fine-tuning of the overall structure of the periplasmic domain, which in turn leads to signal propagation across the membrane via sliding, rotating or tilting of the transmembrane helices. As a result, the topological relationship between domains A and B is changed, thus adjusting the kinase/phosphatase activity ratio of EnvZ.