The design and optimization of complex NMR experiments. Application to a triple-resonance pulse scheme correlating Hα, NH, and 15N chemical shifts in 15N 13C-labeled proteins

Abstract

Sequence-specific assignments of the backbone resonances of proteins form the basis for further study of the structural and dynamic properties of the molecule under investigation. Traditional assignment strategies have relied on through-bond and through-space connectivities provided by homonuclear COSY, HOHAHA/TOCSY, and NOESY spectra (I-4). While very fruitful for proteins less than 10 kDA, this approach becomes difficult for larger systems, due to extensive overlap and decreasing sensitivity of experiments relying on through-bond connectivities. Recently, we have proposed a novel approach for the sequential assignment of 'H, 13C, and 15N spectra of larger proteins based on triple-resonance three-dimensional NMR spectroscopy (5, 6). These experiments exploit the relatively large one-bond J couplings between the backbone 13C and “N nuclei and between the backbone protons and the 13C and “N nuclei to which they are directly attached. Because the couplings are often large compared to pertinent linewidths for proteins less than 25 kDa in molecular weight, magnetization can be transferred between spins in an efficient manner, with the resultant spectra having good sensitivity. Moreover, because the experiments are recorded in the “3D mode,” spectral overlap is virtually eliminated. The sequential assignment of 13C-15N-labeled proteins using the triple-resonance approach is aided considerably by the HOHAHA-HMQC experiment (7). For example, the HNCA triple-resonance experiment correlates the NH and r5N chemical shifts with the intraresidue Ccu chemical shift. When combined with the HNCA experiment, the HOHAHA-HMQC experiment (which correlates “N, NH, and Ha chemical shifts) firmly establishes intraresidue correlations between pairs of “N-NH and G-Ha backbone resonances, despite significant overlap in 1 D ‘H, 15N, and 13C spectra. Linking the 15N, NH, CCX, and Ha chemical shifts is the first step in the sequential assignment process using this new approach. Unlike the triple-resonance experiments, the HOHAHA-HMQC experiment is very sensitive to the secondary structure of the protein under study, as the efficiency of magnetization transfer between NH and Ha protons depends strongly on the NH-Ha scalar couplings. These couplings are, in turn, strongly related to the backbone angle I#I and are less than 6 Hz in regions of regular a-helical secondary structure. For larger proteins rich in a-helical content, a substantial number of NH-Ha connectivities are often found to be weak or missing. In this paper a triple-