Conformational stability and domain coupling in D-glucose/D-galactose-binding protein from Escherichia coli

Abstract

The monomeric D-glucose/D-galactose-binding protein (GGBP) from Escherichia coli (M(r) 33000) is a periplasmic protein that serves as a high-affinity receptor for the active transport and chemotaxis towards both sugars. The effect of D-glucose binding on the thermal unfolding of the GGBP protein at pH 7.0 has been measured by differential scanning calorimetry (DSC), far-UV CD and intrinsic tryptophanyl residue fluorescence (Trp fluorescence). All three techniques reveal reversible, thermal transitions and a midpoint temperature (T(m)) increase from 50 to 63 degrees C produced by 10 mM D-glucose. Both in the absence and presence of D-glucose a single asymmetric endotherm for GGBP is observed in DSC, although each endotherm consists of two transitions about 4 degrees C apart in T(m) values. In the absence of D-glucose, the protein unfolding is best described by two non-ideal transitions, suggesting the presence of unfolding intermediates. In the presence of D-glucose protein, unfolding is more co-operative than in the absence of the ligand, and the experimental data are best fitted to a model that assumes two ideal (two-state) sequential transitions. Thus D-glucose binding changes the character of the GGBP protein folding/unfolding by linking the two domains such that protein unfolding becomes a cooperative, two two-state process. A K(A)' value of 5.6x10(6) M(-1) at 63 degrees C for D-glucose binding is estimated from DSC results. The domain with the lower stability in DSC measurements has been identified as the C-terminal domain of GGBP from thermally induced Trp fluorescence changes.

Figure 1. Representative DSC scans for GGBP in 10 mM Hepes/NaOH, pH 7.0 (at 25 °C)

The data were collected at a 90 °C/h scan rate, corrected for instrument baseline and normalized for protein concentration. Data obtained in the absence or presence of 10 mM D-glucose are shown with deconvolutions into two independent two-state transitions (broken lines) with overall ΔC_p values of 1.7 and 1.3 kcal·K⁻¹·mol⁻¹ respectively. Dotted lines show endotherms obtained in second and third DSC scans with cooling and equilibration at 15 °C for 60 min between scans (for details, see the Materials and methods section).

Figure 2. Deconvolution analysis of DSC data for the partial unfolding of GGBP in the absence of D-glucose

Lower panel: the sigmoidal transitional protein baseline was subtracted from data collected at a 30 °C/h scan rate, normalized for protein concentration, giving a ΔC_p value of 0 for the baseline shown. Experimental data are shown as open circles, and the continuous line is for a non-two-state unfolding model with two transitions. Upper panel: deviations of experimental data points from the best fits to a two-state model (open symbols) either deconvoluting into two independent (▵) or two sequential (○) transitions. Filled squares (▪) represent deviations from a non-two-state model with two transitions shown as a continuous line in the lower panel.

Figure 3. Deconvolution analysis of DSC data for the partial unfolding of GGBP in the presence of 10 mM D-glucose

Lower panel: sigmoidal transitional protein baseline was subtracted from data collected at 30 °C/h scan rate, normalized for protein concentration, giving a ΔC_p of 0 for the baseline shown. Experimental data are shown as open circles and the continuous line is for a two, two-state unfolding model with two sequential transitions. Upper panel: deviations of experimental data points from the best fits to a two-state model assuming a single transition (▵), two independent transitions (○) or two sequential transitions (▪).

Figure 4. Thermally induced changes in tryptophan residue exposure

Open circles represent normalized Trp fluorescence changes of GGBP upon heating at a scan rate of 30 °C/h in the absence or presence of 10 mM D-glucose. The excitation and observation wavelengths were 295 and 342 nm respectively. The fit of each data set to a two-state model of unfolding is shown as a solid line with T_m values of 48.6 and 61.3 °C in the absence or presence of D-glucose respectively.

Figure 5. CD characterization of GGBP protein

Upper panel: far-UV CD spectra of GGBP at 15 °C in the absence (broken line) or presence (continuous line) of 10 mM D-glucose. Lower panel: progress curves for temperature-induced CD changes at 222 nm. Open circles represent mean residue ellipticity ([Θ]) changes of GGBP in the absence or presence of 10 mM D-glucose. The fit of each data set to a two-state model of unfolding is shown as a continuous line with T_m values of 50.8 and 64.5 °C in the absence or presence of 10 mM D-glucose respectively. Data were collected at a 30 °C/h scan rate using a cell with a 0.05 cm path length.

Figure 6. Thermal unfolding of GGBP in the absence of D-glucose

Upper panel: normalized DSC scan (after pre- and post-transitional baseline subtraction) obtained in the range from 20 to 75 °C. The fit of the experimental data points (open circles) to a non-two-state model assuming two unfolding transitions is shown by the broken line. Individual deconvoluted components of this fit are shown by the continuous lines. Vertical dotted lines are drawn at the T_m value of each component and the continuous line at the midpoint of the DSC endotherm. Middle panel: first derivative of CD data from Figure 5 Bottom panel: first derivative of Trp fluorescence (TrpFl) data from Figure 4 CD and fluorescence data were smoothed by adjacent averaging over the range of 1.5 °C prior to differentiation. All data collected at a 30 °C/h scan rate.