Characterization of the internal motions of a chimeric protein by C-13 NMRhighlights the important dynamic consequences of the engineering on a millisecond time scale

By transferring the central curaremimetic beta hairpin of the snake toxin a lpha into the scaffold of the scorpion charybdotoxin, a chimeric protein wa s constructed that reproduced the three-dimensional structure and partially reproduced the function of the parent beta hairpin, without perturbing the three-dimensional structure of the scaffold [1]. Picosecond to hour time s cale motions of charybdotoxin and the engineered protein were observed, in order to evaluate the dynamic consequences of the six deletions and eight m utations differentiating the two molecules. The chimeric protein dynamics w ere also compared to that of toxin alpha, in order to examine the beta hair pin motions in both structural contexts. Thus, C-13 R-1, R-1 rho and H-1 -- > C-13 nOe were measured for all the CalphaHalpha and threonine CbetaHbeta vectors. As the proteins were not labeled, accordion techniques combined to coherence selection by pulsed field gradients and preservation of magnetiz ation following equivalent pathways were used to considerably reduce the sp ectrometer time needed. On one hand, we observed that the chimeric protein and charybdotoxin are subjected to similar picosecond to nanosecond time sc ale motions except around the modified beta sheet region. The chimeric prot ein also exhibits an additional millisecond time scale motion on its whole sequence, and its beta structure is less stable on a minute to hour time sc ale. On the other hand, when the beta hairpin dynamics is compared in two d ifferent structural contexts, i.e. in the chimeric protein and the curaremi metic toxin alpha, the picosecond to nanosecond time scale motions are fair ly conserved. However, the microsecond to millisecond time scale motions ar e different on most of the beta hairpin sequence, and the beta sheet seems more stable in toxin alpha than in the chimera. The slower microsecond to h our time scale motions seem to be extremely sensitive to the structural con text, and thus poorly transferred from one protein to another.