Distinguishing multiple chemotaxis Y protein conformations with laser-polarized 129Xe NMR

Abstract

The chemical shift of the (129)Xe NMR signal has been shown to be extremely sensitive to the local environment around the atom and has been used to follow processes such as ligand binding by bacterial periplasmic binding proteins. Here we show that the (129)Xe shift can sense more subtle changes: magnesium binding, BeF(3)(-) activation, and peptide binding by the Escherichia coli chemotaxis Y protein. (1)H-(15)N correlation spectroscopy and X-ray crystallography were used to identify two xenon-binding cavities in CheY that are primarily responsible for the shift changes. One site is near the active site, and the other is near the peptide binding site.

Figure 1.

Change in ¹²⁹Xe chemical shift (ΔδXe) with protein concentration (C) for apo (○), magnesium-bound (▪), activated (•), FliM peptide-bound (□), and denatured (⋄) CheY. Δ δXe is the difference between the ¹²⁹Xe chemical shift of each titration point and buffer. ¹²⁹Xe chemical shift values and error bars (± 0.01 ppm) were obtained from peak fits. The concentration-normalized ¹²⁹Xe chemical shifts, or slopes of the lines, are 1.6 ± 0.1 ppm mM⁻¹, 1.3 ± 0.1 ppm mM⁻¹, 0.8 ± 0.1 ppm mM⁻¹, 0.7 ± 0.1 ppm mM⁻¹, and 0.4 ± 0.1 ppm mM⁻¹, for apo, magnesium-bound, activated, FliM-bound, and denatured CheY, respectively. ¹²⁹Xe chemical shift is sensitive to four different CheY functional states.

Figure 2.

Histogram plots of the total xenon-induced shift for each assigned residue. (A) Shifts induced by 40 mM xenon for apo CheY. (B) Shifts induced by 45 mM xenon for activated CheY. Similar segments of the protein undergo xenon-induced shifts for both CheY conformations.

Figure 3.

Xenon-induced shifts from Figure 2 ▶ mapped on the backbone of apo CheY (A) (Simonovic and Volz 2001) and activated CheY (B) (this work). Residues are shaded according to the magnitude of their xenon-induced shifts. Unassigned residues are shaded white. For both conformations of CheY shifting residues cluster around two cavities identified by VOIDOO, the H1-β3 cavity (black mesh) and the β4-H4 cavity (gray mesh). Magnesium-binding and BeF₃⁻ activation occur in the active site, directly above the H1-β3 cavity. The side chains of Tyr¹⁰⁶ and Asp⁵⁷ are shown in both structures. Activation at Asp⁵⁷ results in Tyr¹⁰⁶ changing its position from “out” to “in,” filling the β4-H4 cavity. Cavities were calculated with VOIDOO (Jones et al. 1991) and images generated with PyMol (DeLano Scientific LLC).

Figure 4.

X-ray crystal structure of BeF₃⁻-activated CheY bound to xenon. (A) An F_obs–F_calc difference electron density omit map (mesh) contoured at 10σ with the xenon atom removed from the model. Electron density from the xenon atom is clearly visible just below the active site inside the H1-β3 cavity. (B) A 2F_obs–F_calc electron density map (mesh) of the H1-β3 cavity contoured at 1.5σ showing the xenon atom (sphere) and the surrounding residues (sticks).

Figure 5.

(A) The average of the normalized change in chemical shift (Δ δ) vs. xenon concentration, C, for the amide proton resonances of residues Asp¹², Arg¹⁸, Ile²⁰, and Asp⁵⁷, which line the H1-β3 cavity, for the apo (○, solid line) and the activated (•, dashed line) forms of CheY. (B) The average of the normalized change in chemical shift (Δ δ) vs. xenon concentration, C, for the amide proton resonances of residues Tyr¹⁰⁶ and Val¹⁰⁷, which line the β4-H4 cavity, for the apo (□, solid line) and activated (▪, dashed line) conformations of CheY. Chemical shift changes were normalized according to the limiting shift obtained by an initial fit for each residue. These normalized values were averaged for residues unique to the two cavities to obtain the data and error bars shown in this figure. The respective binding constants obtained from these fits are shown in Table 2. Unlike the H1-β3 cavity, the xenon binding affinity of the β4-H4 cavity changes with activation.