Osmolytes modulate conformational exchange in solvent-exposed regions of membrane proteins

Abstract

Site-directed spin labeling (SDSL) was used to investigate local structure and conformational exchange in two bacterial outer-membrane TonB-dependent transporters, BtuB and FecA. Protecting osmolytes, such as polyethylene glycols (PEGs) are known to modulate a substrate-dependent conformational equilibrium in the energy coupling motif (Ton box) of BtuB. Here, we demonstrate that a segment that is N-terminal to the Ton box in BtuB, is in conformational exchange between ordered and disordered states with or without substrate. Protecting osmolytes shift this equilibrium to favor the more ordered, folded state. However, a segment of BtuB that is C-terminal to the Ton box that is not solvent exposed is insensitive to PEGs. Protecting osmolytes also modulate a conformational equilibrium in the Ton box of FecA, with larger molecular weight PEGs producing the largest shifts in the conformational free energy. These data indicate that solvent-exposed regions of these transporters undergo conformational exchange and that regions of these transporters that are involved in protein-protein interactions sample multiple conformational substates. The sensitivity to solute provides an explanation for differences seen between two high-resolution structures of BtuB, which each likely represent one conformation from a subset of states that are normally sampled by the protein. This work also illustrates how SDSL and osmolytes may be used to characterize and quantitate conformational equilibria in membrane proteins.

Figure 1

Three models for BtuB. (a) Apo-structure of BtuB obtained from crystals in a micelle phase (PDB ID: 1NQE), (b) apo-structure of BtuB obtained from crystals in mesophase (PDB ID: 2GUF), and (c) model for the Ton box of BtuB in the presence of substrate obtained by EPR spectroscopy.

Figure 2

Structure of the nitroxide side chain R1. The label is linked to the protein backbone through five rotatable bonds; however, interactions between the distal sulfur and the Cα proton restrict rotameric conversions to X4 and X5.

Figure 3

(a) Structure of the N-terminal segment encompassing residues 1–5 in the model for BtuB obtained from mesophase (PDB ID: 2GUF). (b) EPR spectra from spin labels on residues 1–5 in the absence of substrate. In most of these spectra, both a mobile (m) and immobile (i) component are revealed. The immobile component in Q1R1 is near the rigid limit of nitroxide motion at X-band.

Figure 4

EPR spectra obtained from spin labels placed at positions 1–5 in the N-terminus of BtuB in the absence (black traces) and presence (red traces) of 30% w/v PEG 3350. (a) EPR spectra of sites 1–5 in BtuB reconstituted into POPC in the absence of substrate. (b) EPR spectra of sites 1–5 in the presence of substrate. The amplitudes of these spectra are normalized relative to the same spin number, except where the scale is adjusted as indicated.

Figure 5

(a) EPR spectra from T3R1 in BtuB as a function of the concentration (% w/v) of PEG3350. Increasing PEG concentration alters the EPR spectrum and increases the fraction of immobilized label. (b) Plot of −lnK versus solution osmolality for T3R1, where K is the apparent equilibrium constant for the N-terminal conformational equilibrium [see Eq. (1)].

Figure 6

(a) EPR spectra obtained from positions 18–24 in BtuB in the absence (black trace) and presence (red trace) of vitamin B₁₂. (b) Location of the spin label P18R1 and (c) location of the spin label L23R1 within the crystal structure of BtuB (PDB ID: 1NQE). The EPR spectra indicate the existence of a substrate-driven conformational change; however, unlike the BtuB Ton box, these spectra are insensitive to the addition of osmolytes.

Figure 7

(a) Spectra from sites 82 and 85 within the Ton box of FecA. (b) Spectra in the presence of substrate (ferric citrate or FC) indicate an increase in backbone dynamics and a disordering of the Ton box. (c) Spectra in the presence of substrate with the addition of PEG 1000 (30% w/v) indicate a decrease in nitroxide motion reflecting an ordering in the Ton box. (d) EPR spectra in the presence of substrate but with the addition of Ficoll 400 at a solution viscosity (∼17% w/v) that matches the viscosity of the PEG solution.

Figure 8

(a) Position of the Ton box (highlighted in a stick configuration) in the apo-crystal structure of FecA (PDB ID: 1KMO). (b) EPR spectra from V85R1 at increasing concentrations of PEG 1000. Both a slow or immobile (i) and fast or mobile (m) motional components are revealed in the EPR spectra, and PEG 1000 addition shifts the conformational equilibrium and increases the fraction of Ton box in the more ordered (immobile) form. (c) Plots of the apparent equilibrium constant measured from V85R1 as a function of solution osmolality for PEG 1000.