Relayed-NOE experiments in the rotating frame for sequence analysis of peptides

Abstract

Proton spectra of peptides consist of a set of spin systems for the individual amino acid residues, which are usually identified via their characteristic J-coupling patterns (I) in 2D correlation spectra. However, the examination of a peptide with unknown amino acid sequence or the presence of several amino acids with identical spin systems requires an assignment of each spin system to its position in the peptide chain. The sequence analysis of a peptide makes use of correlations across the peptide bonds between a-protons and the amide protons of the following amino acid, either via heteronuclear long-range couplings to carbonyl carbons (favorably by COLOC (2) or its inverse variant (3)) or directly via NOE effects (I) (NOESY (4)). However, especially in linear peptides the H, resonances are often not sufficiently resolved for an unambiguous assignment. This serious problem can be circumvented by the use of the Relayed-NOESY experiment (5), advantageously with double-quantum filter (6), where the magnetization transfer via NOE between NH’+’ and HL is followed by a transfer through J coupling from HL to NH’ within the same amino acid residue. This results in a cross peak between NH’ and NH’+’ which appears only on one side of the diagonal due to the asymmetry of the mixing sequence (7) (cf. Fig. 1). Unfortunately, many oligopeptides with molecular weights of about 1000 show only very weak NOE effects in high-field spectrometers due to their unfavorable correlation times. Instead of conventional NOE spectra, ROESY (8, 9) experiments can be used to measure NOE effects in the rotating frame (ROE effects), which are almost independent of the correlation time. We describe here the recently proposed (IO) possibility of using spin-lock sequences in Relayed-NOESY experiments for both the NOE and the Jtransfer part of the mixing time. Figure 2 shows the pulse sequences which are discussed in this paper. The DQF-Relayed-NOESY experiment exists in two versions, which differ in the order of NOE transfer and J trandkr. Since the desired cross peaks always show antiphaae s&uctaue in one dime i$ seems to be favoraMe to have antiphase in w2 because of’tlte bd%cs resduhta in that dimension. This can be achieved by placing ~JVrans~&z&eNOEtrans&ass&owninF~ Z?a.Tbecorrespondingsequence wi~theJ~r~~eNOE~(givingrise~bo~inwl)is~f~ omitted in Fig. 2.