The role of polyproline motifs in the histidine kinase EnvZ

Abstract

Although distinct amino acid motifs containing consecutive prolines (polyP) cause ribosome stalling, which necessitates recruitment of the translation elongation factor P (EF-P), they occur strikingly often in bacterial proteomes. For example, polyP motifs are found in more than half of all histidine kinases in Escherichia coli K-12, which raises the question of their role(s) in receptor function. Here we have investigated the roles of two polyP motifs in the osmosensor and histidine kinase EnvZ. We show that the IPPPL motif in the HAMP domain is required for dimerization of EnvZ. Moreover, replacement of the prolines in this motif by alanines disables the receptor's sensor function. The second motif, VVPPA, which is located in the periplasmic domain, was found to be required for interaction with the modulator protein MzrA. Our study also reveals that polyP-dependent stalling has little effect on EnvZ levels. Hence, both polyP motifs in EnvZ are primarily involved in protein-protein interaction. Furthermore, while the first motif occurs in almost all EnvZ homologues, the second motif is only found in species that have MzrA, indicating co-evolution of the two proteins.

Fig 1. Schematic overview of the EnvZ/OmpR signaling cascade.

EnvZ is located in the cytoplasmic membrane and senses alterations in osmolarity. It transduces the signal via phosphorylation to its cognate response regulator OmpR, which in turn reciprocally adjusts the expression of the target genes ompC and ompF, which code for outer membrane porins with different pore diameters. MzrA also resides in the cytoplasmic membrane and modulates the activity of EnvZ. PolyP motifs and their localization are shown in red. CM–cytoplasmic membrane; PP–periplasm; OM–outer membrane.

Fig 2. The role of polyP_c in EnvZ dimerization.

(A) The 3D structure of the HAMP domain of EnvZ in E. coli K-12, modelled with Phyre. Proline residues of the polyP_c motif are marked in red. (B) Sequence conservation of the EnvZ HAMP domain based on the alignment of 63 EnvZ homologues (exhibiting >44% sequence identity to E. coli K-12 EnvZ) from a phylogenetic tree of representative Gammaproteobacteria [52]. (C) Two-hybrid analysis (BACTH assay) of the significance of EnvZ polyP_c for EnvZ dimerization, based on the complementation of T25 and T18 adenylate cyclase fragments fused N-terminally to EnvZ. The histograms depict β-galactosidase activities after transformation of the reporter strain BTH101 with plasmids encoding the indicated hybrids. Cells were grown in LB medium and harvested 12 h after induction with IPTG. The data is based on biological triplicates, and error bars indicate standard deviations of the mean. (D) Determination of β-galactosidase activity of BTH101 cells, transformed with plasmid encoded T18/T25-EnvZ variants, as revealed by blue staining of colonies grown on X-Gal/IPTG agar plates for 24 h. The experiment was repeated three times and a representative plate for each interaction condition is shown.

Fig 3. The role of polyP_c in EnvZ function.

An E. coli wild-type control (EnvZ_WT), the envZ deletion strain (ΔenvZ) and the mutant in which the IPPPL motif was replaced by IAAAL (EnvZ_P/A(c)) were characterized. (A) Analysis of EnvZ/OmpR target gene expression in response to osmotic stress caused by the addition of 0.2 or 0.4 M NaCl to the growth medium (M9 medium). Cells were grown to the mid-exponential growth phase. Outer membrane proteins were isolated from E. coli K-12 EnvZ_WT, and the EnvZ_P/A(c) and ΔenvZ mutants, fractionated on an SDS-urea gel and stained with Coomassie Blue. The experiment was repeated three times and a representative gel is shown. (B) Quantification of OmpF and OmpC band intensities of the gel is shown in (A). (C) Levels of OmpC-CFP fluorescence were measured in E. coli EPB273a [27] reporter strains deleted for envZ (ΔenvZ) or expressing wild-type EnvZ or the EnvZ_P/A(c) variant, following exposure to osmotic stress imposed by added NaCl or sucrose (Suc) for approximately 3 hours (OD₆₀₀ = 0.2). The addition of 0.2 M NaCl and 0.4 M sucrose, respectively, corresponded to an increase in the medium osmolality from 0.2 to 0.460 Osmol/kg. In the presence of 0.8 M sucrose the medium osmolarity was determined with 1.080 Osmol/kg. The results are based on the analysis of biological triplicates and values were normalized to the fluorescence level of wild-type EnvZ grown in M9 medium (value 1.0). The standard deviations are indicated. (D) Western blot analysis using anti-EnvZ antibodies. Aliquots (200 μg) of cytoplasmic membrane proteins obtained from wild-type E. coli K-12 (EnvZ_WT), EnvZ polyP_P/A(C) or E. coli K-12 Δefp were separated on a SDS polyacrylamide gel. The values for relative band intensities are derived from biological triplicates. The E. coli K-12 ΔenvZ strain served as the negative control and was complemented with plasmid-encoded envZ (pEnvZ_WT [26]) for use as the positive control). (E) Western blot analysis of membrane proteins prepared from wild-type E. coli K-12 or a Δefp mutant harboring plasmid-encoded envZ (pBAD24_EnvZ-FLAG). EnvZ-FLAG was detected with anti-FLAG antibodies. The values for relative band intensities represent biological duplicates.

Fig 4. The role of polyP_p in the EnvZ-MzrA interaction.

(A) The 3D structure of the periplasmic domain of (E. coli K-12) EnvZ [63]. Proline residues in the PolyP_p motif are marked in red. (B) Sequence conservation of the EnvZ periplasmic domain based on multiple sequence alignment of EnvZ homologues (for details see Fig 2). (C) BACTH analysis of the significance of the polyP_p and polyP_c motifs of EnvZ for its interaction with MzrA, based on the complementation of T25 and T18 adenylate cyclase fragments fused N-terminally to EnvZ variants. β-Galactosidase activities of the reporter strain BTH101 after transformation with plasmids encoding the indicated hybrids and growth in LB-medium. (D) BACTH assay (constructs and cultivation conditions as in (C)) to analyze the effect of polyP_p on EnvZ dimerization, quantified by measuring β-galactosidase activities. (E) Determination of β-galactosidase induced blue staining of colonies of BTH101, transformed with the described constructs, and grown on X-Gal/IPTG agar plates for 24 hours. All data are based on biological triplicates. Error bars indicate standard deviations of the mean, and representative plates are shown.

Fig 5. Distribution and evolution of EnvZ and MzrA.

(A) Gene tree based on multiple sequence alignment of 793 EnvZ sequences (exhibiting >44% identity to E. coli K-12 EnvZ) identified by UniProt BLAST search. Strains harboring homologues of E. coli K-12 MzrA (> 31% sequence identity) appear on a red background or otherwise on a blue background. The presence of polyP_c and polyP_p is indicated by the branch and letter color. (B) Species subtree of Gammaproteobacteria from a “tree of life” phylogeny [52]. EnvZ homologues of the bacterial species are indicated according to their degree of sequence identity to E. coli K-12 EnvZ (letter color). The presence of MzrA homologues in some species is marked with a blue background. PolyP motifs are labelled with colored background according to their predicted stalling strength.