Three-dimensional structure of the histidine-containing phosphocarrier protein (HPr) from Enterococcus faecalis in solution

The histidine-containing phosphocarrier protein (HPr) transfers a phosphate group between components of the prokaryotic phosphoenolpyruvate-dependent phosphotransferase system (PTS), which is finally used to phosphorylate the carbohydrate transported by the PTS through the cell membrane. Recently it has also been found to act as an intermediate in the signaling cascade tha t regulates transcription of genes related to the carbohydrate-response sys tem. Both functions involve phosphorylation/dephosphorylation reactions, bu t at different sites. Using multidimensional H-1-NMR spectroscopy and angul ar space simulated annealing calculations, we determined the structure of H Pr from Enterococcus faecalis in aqueous solution using 1469 distance and 4 4 angle constraints derived from homonuclear NMR data. It has a similar ove rall fold to that found in HPrs from other organisms. Four beta strands, A, B, C, D, encompassing residues 2-7, 32-37, 40-42 and 60-66, form an antipa rallel beta sheet lying opposite the two antiparallel alpha helices, a and c (residues 16-26 acid 70-83). A short alpha helix, b, from residues 47-53 is also observed. The pairwise root mean square displacement for the backbo ne heavy atoms of the mean of the 16 NMR structures to the crystal structur e is 0.164 nm. In contrast with the crystalline state, in which a torsion a ngle strain in the active-center loop has been described [Jia, Z., Vandonse laar, M., Quail, J.W. & Delbaere, L.T.J. (1993) Nature (London) 361, 94-97] , in the solution structure, the active-site His15 rests on top of helix a, and the phosphorylation site N-delta1 of the histidine ring is oriented to wards the surface, making it easily accessible to the solvent. Back calcula tion of the 2D NOESY NMR spectra from both the NMR and X-ray structures sho ws that the active-center structure derived by X-ray crystallography is not compatible with experimental data recorded in solution. The observed torsi onal strain must either be a crystallization artefact or represents a confo rmational state that exists only to a small extent in solution.