Different evolutionary constraints on chemotaxis proteins CheW and CheY revealed by heterologous expression studies and protein sequence analysis

Abstract

CheW and CheY are single-domain proteins from a signal transduction pathway that transmits information from transmembrane receptors to flagellar motors in bacterial chemotaxis. In various bacterial and archaeal species, the cheW and cheY genes are usually encoded within homologous chemotaxis operons. We examined evolutionary changes in these two proteins from distantly related proteobacterial species, Escherichia coli and Azospirillum brasilense. We analyzed the functions of divergent CheW and CheY proteins from A. brasilense by heterologous expression in E. coli wild-type and mutant strains. Both proteins were able to specifically inhibit chemotaxis of a wild-type E. coli strain; however, only CheW from A. brasilense was able to restore signal transduction in a corresponding mutant of E. coli. Detailed protein sequence analysis of CheW and CheY homologs from the two species revealed substantial differences in the types of amino acid substitutions in the two proteins. Multiple, but conservative, substitutions were found in CheW homologs. No severe mismatches were found between the CheW homologs in positions that are known to be structurally or functionally important. Substitutions in CheY homologs were found to be less conservative and occurred in positions that are critical for interactions with other components of the signal transduction pathway. Our findings suggest that proteins from the same cellular pathway encoded by genes from the same operon have different evolutionary constraints on their structures that reflect differences in their functions.

FIG. 1.

Heterologous expression of the A. brasilense CheW protein in E. coli. (A) Swarm plate assay. Conditions were LB medium, 100 μM IPTG, 22°C, and 12 h of incubation. Strains, clockwise from top left, are RP437/pTrc99a, RP437/pIZ101, RP4606/pIZ101, and RP4606/pTrc99a. (B) Inhibition of chemotaxis of wild-type cells by A. brasilense CheW. Relative sizes of swarms are shown. Conditions were LB medium, 500 μM IPTG, 30°C, and 12 h of incubation. (C) Temporal gradient assay. Cells were induced with 100 μM IPTG. Times for adaptation to 10 mM leucine are shown.

FIG. 2.

Heterologous expression of the A. brasilense CheY protein in E. coli. (A) Swarm plate assay. Conditions were LB medium, 500 μM IPTG, 30°C, and 12 h of incubation. Strains, clockwise from top left, are RP437/pProEx, RP437/pIZ103, RP5232/pIZ103, and RP5232/pProEx. (B) Expression of A. brasilense CheY in E. coli. Relative sizes of swarms are shown. (C) Western blot analysis of A. brasilense CheY expression in E. coli RP5232. IPTG concentrations: 0 μM (lane 1), 250 μM (lane 2), and 1 mM (lane 3). (D) Temporal gradient assay. Cells were induced with 500 μM IPTG. Times for adaptation to 10 mM leucine are shown.

FIG. 3.

Plots of sequence similarity between E. coli and A. brasilense proteins. Positions in the sequence alignment are shown along the x axis; residue numbers given are for E. coli sequences. Scores based on the PAM150 amino acid substitution matrix (10) are shown along the y axis; positive scores are given for matches and negative scores are given for mismatches according to chemical groups of amino acids. Dashed line marks the −2 cutoff. (A) CheW proteins. Positions of functionally important residues identified in mutational and biochemical studies are shown in color, as follows: red, null mutations (5); green, mutations that affect CheA and chemoreceptor binding in vitro (6); blue, chemoreceptor-suppressible mutations (21). The secondary structure of CheW shown above the plot is based on the reported structure of the Thermotoga maritima protein (15) and predicted two-dimensional structures for E. coli and A. brasilense proteins. (B) CheY proteins. Positions of functionally important residues identified in mutational, biochemical, and structural studies are shown in color, as follows: red, active-site residues; green, residues proposed to interact with FliM (19, 23), CheZ (23, 45), or CheA (24, 42). The secondary structure of CheY shown above the plot is based on the reported structure of the E. coli protein (41). The A. brasilense protein contains a 4-amino-acid deletion corresponding to residues 120 to 123 of the E. coli homolog. α-Helices and β-strands are indicated.