Double-quantum (2Q) experiments represent a powerful refocusing period is hence missing in this version. The last three PFGs are used for selection of the desired coherence and useful tool for the proton resonance assignments of crowded spectra (1-9). However, they have not encoun-pathway, whereas the first PFG is a crusher which destroys all magnetization not aligned along the z axis. tered, in the biomolecular NMR community, the same success as other NMR experiments. This is mainly due to the Two data sets which correspond to the echo and antiecho coherence pathway selections are recorded for these presentation of the 2Q spectra along v 1 which makes the comparison with other two-or three-dimensional spectra experiments by reversing the sign of the second and third PFG located in t 2 . These data are then properly added, more difficult. Recently, we have devised a new class of homonuclear 3D experiments (10) in which the spins are resulting in spectra with pure absorption character for the direct 2Q peaks. Quadrature detection in t 1 is achieved labeled in one dimension with their characteristic 2Q frequency. The other two axes correspond to single-quantum instead with the TPPI method, where the phase of the first 90ะ pulse is incremented in steps of 90ะ in concert with the (1Q) frequencies. Planes extracted from these 3D spectra along the 2Q frequency axis are easy to analyze and allow t 1 evolution period. The H 2 O suppression in this type of experiment is good, allowing the observation of resonances rapid proton resonance assignments of unlabeled biomolecules. These experiments make use of pulsed field gradients close to the water signal. Remarkable H 2 O suppression is achieved by using for G2, G3, and G4 gradients tilted at (PFGs) (11-13) for the selection of 2Q coherence.
In the original version of the 3D PFG 2Q-NOESY and the magic angle ( 14, 15 ) , as has been shown for the 2Q experiment ( ) . In the original version of the experiment 2Q-TOCSY experiments, 2Q coherence is created first. This coherence is subsequently converted to single-quantum anti-( ) , the peaks are in-phase in all dimensions, whereas phase magnetization and then refocused to in-phase singlehere they are in-phase in v 1 and v 2 and antiphase in v 3 . quantum coherence before the NOESY or TOCSY step. At However, a 90ะ shift in the phase of the peaks along v 3 the end of the refocusing period, not all the antiphase magneresults in peaks which are mostly positive with small negatization will be rephased to in-phase magnetization. The tive lobs on both sides ( 17 ). amount of rephased single-quantum magnetization depends
We have tested this new version of the experiment with on the spin-system coupling constants, and, for certain values chicken egg white lysozyme, a protein consisting of 129 of J, it can be very small. Furthermore, there are magnetizaamino acids and for which resonance assignments are tion losses due to the spin-spin relaxation (T 2 ) occurring known (18) . Only four scans were recorded for each of the during the refocusing period. The experiment is hence applit 1 and t 2 increments. Figure shows two typical v 1 v 3 planes cable only to concentrated samples.
taken at two different v 2 (2Q) NHaH frequencies. The planes We report here a modification of the original experiment in ( a, b) and ( c, d) are taken at the (2Q) frequency of 14.2 which results in an improvement in sensitivity. The pulse and 12.9 ppm, respectively. The planes in ( a, c) and ( b, d) sequences of the new 3D experiments are shown in Fig. . have been extracted respectively from the 3D PFG NOESY-The first step is the NOESY (a, c) or TOCSY (b) step. The 2Q and from the 3D PFG TOCSY-2Q spectra. Only the in-phase, single-quantum coherence is then converted, with amide region of the spectra is shown along v 3 . In this specthe excitation multiple-quantum step, to 2Q coherence and tral region, different types of peaks are observed which labeled during t 2 with its characteristic chemical shift. The originate from different magnetization pathways. The peaks 90ะ pulse at the end of the evolution period t 2 converts this centered on the diagonal at v 1 ร
v 3 ร
v NH are the so-called 2Q coherence to antiphase magnetization which is, after the cross-diagonal peaks. The peaks located at v 1 ร
v aH and last refocusing PFG, immediately detected during t 3 . The v 3 ร
v NH lie on a line with slope 01, which is perpendicular 186