19F NMR relaxation studies on 5-fluorotryptophan- and tetradeutero-5-fluorotryptophan-labeled E. coli glucose/galactose receptor

Abstract

19F NMR relaxation studies have been carried out on a fluorotryptophan-labeled E. coli periplasmic glucose/galactose receptor (GGR). The protein was derived from E. coli grown on a medium containing a 50:50 mixture of 5-fluorotryptophan and [2,4,6,7-2H4]-5-fluorotryptophan. As a result of the large gamma-isotope shift, the two labels give rise to separate resonances, allowing relaxation contributions of the substituted indole protons to be selectively monitored. Spin-lattice relaxation rates were determined at field strengths of 11.75 T and 8.5 T, and the results were analyzed using a model-free formalism. In order to evaluate the contributions of chemical shift anisotropy to the observed relaxation parameters, solid-state NMR studies were performed on [2,4,6,7-2H4]-5-fluorotryptophan. Analysis of the observed 19F powder pattern lineshape resulted in anisotropy and asymmetry parameters of delta sigma = -93.5 ppm and eta = 0.24. Theoretical analyses of the relaxation parameters are consistent with internal motion of the fluorotryptophan residues characterized by order parameters S2 of approximately 1, and by correlation times for internal motion approximately 10(-11)s. Simultaneous least squares fitting of the spin-lattice relaxation and line-width data with tau i set at 10 ps yielded a molecular correlation time of 20 ns for the glucose-complexed GGR, and a mean order parameter S2 = 0.89 for fluorotryptophan residues 183, 127, 133, and 195. By contrast, the calculated order parameter for FTrp284, located on the surface of the protein, was 0.77. Significant differences among the spin-lattice relaxation rates of the five fluorotryptophan residues of glucose-complexed GGR were also observed, with the order of relaxation rates given by: R1F183 > R1F127 approximately R1F133 approximately R1F195 > R1F284. Although such differences may reflect motional variations among these residues, the effects are largely predicted by differences in the distribution of nearby hydrogen nuclei, derived from crystal structure data. In the absence of glucose, spin-lattice relaxation rates for fluorotryptophan residues 183, 127, 133, and 195 were found to decrease by a mean of 13%, while the value for residue 284 exhibits an increase of similar magnitude relative to the liganded molecule. These changes are interpreted in terms of a slower overall correlation time for molecular motion, as well as a change in the internal mobility of FTrp284, located in the hinge region of the receptor.

In-cell thermodynamics and a new role for protein surfaces

“…The internal correlation time (τ e ) and the order parameter (S 2 ) were set to 20 ps and 0.82 for the folded state and 1,200 ps and 0.34 for the unfolded state, respectively (24). Chemical shift anisotropy and asymmetry terms were set to −93.5 ppm and 0.24, respectively (30). No μs-ms motion was observed in buffer, and a 950-Hz effective field was used for the R 2 measurements.…”

Dynamic Frequency Shifts of Complexed Ligands: An NMR Study ofd-[1-13C,1-2H]Glucose Complexed to theEscherichia coliPeriplasmic Glucose/Galactose Receptor

“…both the crystallographic B factors (19) and in the 19 F relaxation times of fluorotryptophan-labeled GGR complexed with D-glucose (12); both approaches indicate that the complex is most highly ordered in the region surrounding the glucose molecule.…”

“…The overall isotropic stant, which results in sufficiently low deuterium spin-latcorrelation time at 25ЊC was set at 20 ns based on our previtice relaxation rates such that the perturbed multiplet is not ous studies of fluorotryptophan-labeled GGR performed uncollapsed for typical rotational correlation times encounder similar conditions (12), and was assumed to vary in tered. Additionally, the alignment of the principal axes of proportion to the viscosity of water as a function of temperthe deuterium quadrupolar and carbon-deuterium dipolar ature.…”

Cross-Correlated 19F Relaxation Measurements for the Study of Fluorinated Ligand–Receptor Interactions

“…1. Following the conventions of Luck et al (9) we assign the z, y, and x principal axes of the CSA tensor with corresponding principal values of σ zz = −75 ppm, σ yy = 0 ppm, and σ xx = 50 ppm. Figure 1 depicts the orientations of the CSA principal axes relative to the DD interaction vector r FH .…”