Using structural information to change the phosphotransfer specificity of a two-component chemotaxis signalling complex

Abstract

Two-component signal transduction pathways comprising histidine protein kinases (HPKs) and their response regulators (RRs) are widely used to control bacterial responses to environmental challenges. Some bacteria have over 150 different two-component pathways, and the specificity of the phosphotransfer reactions within these systems is tightly controlled to prevent unwanted crosstalk. One of the best understood two-component signalling pathways is the chemotaxis pathway. Here, we present the 1.40 A crystal structure of the histidine-containing phosphotransfer domain of the chemotaxis HPK, CheA(3), in complex with its cognate RR, CheY(6). A methionine finger on CheY(6) that nestles in a hydrophobic pocket in CheA(3) was shown to be important for the interaction and was found to only occur in the cognate RRs of CheA(3), CheY(6), and CheB(2). Site-directed mutagenesis of this methionine in combination with two adjacent residues abolished binding, as shown by surface plasmon resonance studies, and phosphotransfer from CheA(3)-P to CheY(6). Introduction of this methionine and an adjacent alanine residue into a range of noncognate CheYs, dramatically changed their specificity, allowing protein interaction and rapid phosphotransfer from CheA(3)-P. The structure presented here has allowed us to identify specificity determinants for the CheA-CheY interaction and subsequently to successfully reengineer phosphotransfer signalling. In summary, our results provide valuable insight into how cells mediate specificity in one of the most abundant signalling pathways in biology, two-component signal transduction.

Figure 1. Structure of the CheA₃P1.CheY₆ complex.

CheA₃P1 is shown in light blue whereas CheY₆ is in pale green. The phosphorylatable residue of CheA₃P1, His51, and the phosphoacceptor on CheY₆, Asp56, are shown in stick representation. Secondary structure elements are labelled in black. Residues 113–118 of CheY₆ could not be traced and are depicted as a dotted line. (A) Side view onto the CheA₃P1.CheY₆ complex with the interaction between β5 and α5 in CheY₆ and αB, and the following loop connecting αB and αC on CheA₃P1, marked with an asterisk. (B) Top view showing the major site of interaction between the N-terminal end of α1 on CheY₆ and αA/αB on CheA₃P1 marked by an asterisk.

Figure 2. Superposition of the CheA₃P1.CheY₆ complex in the phosphorylated and unphosphorylated conformations.

Structures were aligned onto the RR. Colour coding for the unphosphorylated conformation as in Figure 1; for the phosphorylated conformation, CheY₆ is shown in yellow and CheA₃ P1 in teal. The active site residues His51 (CheA₃ P1) and Asp56/Asn56 (CheY₆ unphosphorylated/phosphorylated complex conformation) are shown in stick representation. (A) Overview highlighting the 2.1 Å rigid body translation of CheA₃ P1 towards CheY₆. (B) Close-up view onto the active site showing the movement of His51 towards Asp56/Asn56.

Figure 3. Close-up view of the two binding sites between CheA₃P1 and CheY₆.

Colour coding as in Figure 1. Residues that are involved in the interaction are shown in stick representation. (A) Detailed view onto the major binding site between CheY₆ and CheA₃P1. Orientation as in Figure 1B, residues 19–53 of CheA₃ P1 are omitted for clarity. (B) Solvent accessible surface of CheY₆ coloured by electrostatic potential contoured at ±10 kT. Met13 fits snugly into the hydrophobic pocket on CheA₃P1. (C) Close-up view onto the second binding site. Orientation as in Figure 1A. The solvent-accessible surface area is coloured by electrostatic potential contoured at ±10 kT. The hydrogen bond between Gly110 of CheY₆ and Val59 of CheA₃P1 is marked with a dotted line.

Figure 4. Structure-guided sequence alignment of the chemotaxis RRs from E. coli and R. sphaeroides.

Alignment is based on the structures of E. coli CheY (PDB code: 3CHY), E. coli CheB (1A2O), R. sphaeroides CheY₃ (C. H. Bell, unpublished data), and R. sphaeroides CheY₆. Secondary structure is shown for CheY₆. Residues involved in binding of CheY₆ to CheA₃P1 are marked with a star (contributes >30% to total buried surface area), square (>15%), or circle (>5%).

Figure 5. Binding of CheA₃P1 to the response regulators.

(A) Table of binding constants (K_d) measured by SPR between CheA₃P1 and CheY₄, CheY₆ and their mutant versions. Data are expressed as mean ± standard error of the mean (s.e.m.). ND, not determinable. (B–D) Binding of CheA₃P1 to CheY₆ wild type (wt), CheY₄ wt, and CheY₄(P12A,S13M). Left, representative sets of experimental sensorgrams from typical equilibrium-based binding experiments, with reference subtraction. Different concentrations of CheA₃P1 were injected over surfaces coupled with the respective RR. For all injections, the experimental traces reached equilibrium and returned to baseline after the injection. Right, plot of the equilibrium binding response (response units [RU]) against CheA₃P1 concentration ranging from 120 nM to 2 mM. Within one experiment, each concentration was measured twice. All experiments were performed in duplicate. Best-fit binding curves corresponding with a 1∶1 binding model are shown as lines.

Figure 6. Changing the phosphotransfer specificity of CheY₆ and CheY₄.

Phosphorimages of SDS-PAGE gels measuring phosphotransfer from CheA₃P1-P to (A) CheY₆, (B) CheY₆(M13S), (C) CheY₆(M13S,L16S,Y17M), (D) CheY₄, (E) CheY₄(P12A,S13M), and (F) CheY₄(P12A,S13M,C16L,M17Y). CheY (5 µM) was added to 1 µM CheA₃P1-³²P. Ten-microlitre reaction samples were taken at the time points indicated and quenched in 20 µl of 1.5× SDS/EDTA loading dye. The quenched samples were analyzed by SDS-PAGE and detected by phosphorimaging.

Figure 7. Changing the phosphotransfer specificity of other CheYs by introduction of A12 and M13.

Phosphorimages of SDS-PAGE gels measuring phosphotransfer from CheA₃P1-P to (A) E. coli CheY, (B) E. coli CheY(S15A,T16M), (C) CheY₁, (D) CheY₁(R12A,T13M), (E) CheY₃, (F) CheY₃(P12A,S13M), (G) CheY₅, and (H) CheY₅(P12A,S13M). CheY (5 µM) was added to 1 µM CheA₃P1-³²P. Ten-microlitre reaction samples were taken at the time points indicated and quenched in 20 µl of 1.5× SDS/EDTA loading dye. The quenched samples were analyzed by SDS-PAGE and detected by phosphorimaging.