Solution-state NMR investigations of triosephosphate isomerase active site loop motion: ligand release in relation to active site loop dynamics

Abstract

Product release is partially rate determining in the isomerization reaction catalyzed by Triosephosphate Isomerase, the conversion of dihydroxyacetone phosphate to d-glyceraldehyde 3-phosphate, probably because an active-site loop movement is necessary to free the product from confinement in the active-site. The timescale of the catalytic loop motion and of ligand release were studied using 19F and 31P solution-state NMR. A 5′-fluorotryptophan was incorporated in the loop N-terminal hinge as a reporter of loop motion timescale. Crystallographic studies confirmed that the structure of the fluorinated enzyme is indistinguishable from the wild-type; the fluorine accepts a hydrogen bond from water and not from a protein residue, with minimal perturbation to the flexible loop stability. Two distinct loop conformations were observed by 19F NMR. Both for unligated (empty) and ligated enzyme samples a single species was detected, but the chemical shifts of these two distinct species differed by 1.2 ppm. For samples in the presence of subsaturating amounts of a substrate analogue, glycerol 3-phosphate, both NMR peaks were present, with broadened lineshapes at 0°C. In contrast, a single NMR peak representing a rapid average of the two species was observed at 30°C. We conclude that the rate of loop motion is less than 1400 s−1 at 0°C and more than 1400 s−1 at 30°C. Ligand release was studied under similar sample conditions, using 31P NMR of the phosphate group of the substrate analogue. The rate of ligand release is less than 1000 s−1 at 0°C and more than 1000 s−1 at 30°C. Therefore, loop motion and product release are probably concerted and likely to represent a rate limiting step for chemistry.