Oversized galactosides as a probe for conformational dynamics in LacY

Abstract

Binding kinetics of α-galactopyranoside homologs with fluorescent aglycones of different sizes and shapes were determined with the lactose permease (LacY) of Escherichia coli by FRET from Trp151 in the binding site of LacY to the fluorophores. Fast binding was observed with LacY stabilized in an outward-open conformation (k_on = 4-20 μM^-1·s^-1), indicating unobstructed access to the binding site even for ligands that are much larger than lactose. Dissociation rate constants (k_off) increase with the size of the aglycone so that K_d values also increase but remain in the micromolar range for each homolog. Phe27 (helix I) forms an apparent constriction in the pathway for sugar by protruding into the periplasmic cavity. However, replacement of Phe27 with a bulkier Trp does not create an obstacle in the pathway even for large ligands, since binding kinetics remain unchanged. High accessibility of the binding site is also observed in a LacY/nanobody complex with partially blocked periplasmic opening. Remarkably, E. coli expressing WT LacY catalyzes transport of α- or β-galactopyranosides with oversized aglycones such as bodipy or Aldol518, which may require an extra space within the occluded intermediate. The results confirm that LacY specificity is strictly directed toward the galactopyranoside ring and also clearly indicate that the opening on the periplasmic side is sufficiently wide to accommodate the large galactoside derivatives tested here. We conclude that the actual pathway for the substrate entering from the periplasmic side is wider than the pore diameter calculated in the periplasmic-open X-ray structures.

Keywords: fluorescence; lactose permease; membrane transport proteins; nanobodies; stopped-flow.

Fig. 1.

Periplasmic-open conformer LacY_WW. The structural model viewed from the side (PDB ID code 5GXB) is shown with N- and C-terminal six-helical bundles colored gray and cyan, respectively, and Nb9039 attached to the periplasmic side shown in red. Residues participating in sugar binding are presented as yellow sticks. Residue Phe27 located in the middle of the periplasmic cavity was replaced with Trp and is shown as pink spheres. Interacting residues in the area of contact between the CDR3 loop of the Nb (Tyr104) and the periplasmic end of helix I of LacY (Lys42) are shown as red and gray spheres, respectively. Structural models of the α-galactosides with different aglycones are also shown.

Fig. 2.

Accessibility of the sugar-binding site in LacY with an outward-open cavity and the effect of the Phe27→Trp mutation. Binding of five α-galactosides was measured directly as FRET from Trp151 of LacY to bound ligand as described in Methods. Two outward-facing constructs were used: the WT LacY/Nb9048 complex (A) and mutant LacY_WW (B). Binding rates (k_obs, filled symbols) were estimated from single-exponential fits of stopped-flow traces at each galactoside concentration (Fig. S2), and k_offs (open symbols) were measured as the rate of displacement of bound ligand with an excess of TDG (Fig. S3). The slopes of the linear fits of the concentration dependencies of binding rates gave estimates of k_on for each galactoside derivative. Black and red symbols represent data obtained before and after replacement of Phe27 with Trp, respectively, for each galactoside (circles: NPG; squares: BNG; triangles: MUG; stars: bodipy-Gal; diamonds: Aldol-Gal) (Table S1). NPG binding rates measured with WT LacY without Nb9048 are shown for comparison (green circles).

Fig. 3.

Accessibility of the sugar-binding site in the LacY_WW/Nb9039 complex. Binding rates of different α-galactosides were measured as described in Fig. 2 using the outward-open LacY conformer with or without Nb9039 (red and black symbols, respectively). The values of k_obs (filled symbols) and k_off (open symbols) were estimated from single-exponential fits of stopped-flow traces (Figs. S4 and S5). The slopes of the linear fits of the concentration dependencies of binding rates gave estimates of the k_on for each galactoside derivative (circles: NPG; squares: BNG; triangles: MUG; stars: bodipy-Gal) (Table S1).

Fig. 4.

Transport of β-MUG. The accumulation of MUG was measured using E. coli T184 cells (ΔlacZY) expressing WT LacY (red), the E325A (green) or LacY_WW (blue) mutants, or no LacY (cyan) as described in the text and in Methods (see also Fig. S6). Cells incubated with β-MUG for a given time were washed with acidic Na-acetate (pH 4.8), disrupted by sonication, and treated with β-galactosidase. (A) Fluorescence change in cells containing WT LacY during a 30-min incubation with 50-μM β-MUG. (B) Fluorescence change in cells containing LacY mutants or no LacY after a 30-min incubation under the same conditions. Dotted lines show fluorescence at time 0. (C) Time course of β-MUG uptake. Transport activity was calculated from fluorescence intensities at 445 nm in A and B using the calibration curve (Fig. S6D).

Fig. 5.

Transport of α-MUG. The accumulation of MUG was recorded as the fluorescence increase with time in intact E. coli T184 cells (ΔlacZY) expressing WT LacY (red), the E325A (green) or LacY_WW (blue) mutants, or no LacY (cyan) as described in the text and in Methods. The arrow indicates the addition of 10-μM α-MUG to cells at 0.6 mg/mL total protein. Uptake of α-MUG was detected as an increase in fluorescence due to enzymatic cleavage of accumulated α-galactoside by intracellular α-galactosidase. As controls, cells containing WT LacY were either pretreated with NEM, an inactivator of lactose transport (10 mM for 20 min) (black dots) or were mixed with 10-μM β-MUG (red dots).

Fig. 6.

Transport of galactopyranosides with different-sized aglycones. (A) Time courses of the accumulation of [¹⁴C] lactose (0.4 mM) by E. coli T184 cells (ΔlacZY) expressing WT LacY (filled circles), mutant LacY F27W (stars), or no LacY (open circles) were measured by scintillation spectrometry as described in Methods. The steady-state level of 100% corresponds to 210 nmol/mg of total protein. (B) Time courses of accumulation of different galactoside derivatives (50 μM) by cells expressing WT LacY (filled symbols) or no LacY (open symbols). Transport was measured with β-MUG (green circles), β-bodipy-Lact (purple squares), β-bodipy-Gal (blue triangles), α-bodipy-Gal (cyan triangles), and β-Aldol-Gal (red diamonds) as described in Fig. 4, but samples containing bodipy-galactosides were not treated with β-galactosidase (Fig. S6). The fluorescence changes in cells during incubation with β-bodipy-Lact, β-bodipy-Gal, α-bodipy-Gal, and β-Aldol-Gal are shown in Fig. S7. Transport activity was calculated from fluorescence intensities using the calibration curves (Fig. S6 E and F for Aldol and bodipy, respectively). The steady-state level of 100% corresponds to 2.7, 0.72, 1.0, 0.86, and 6.2 nmol/mg of total protein for β-MUG, β-bodipy-Lact, β-bodipy-Gal, α-bodipy-Gal, and β-Aldol-Gal, respectively.