In recent years, heteronuclear J cross-polarization experi-pled proton(s) via 1 H-1 H NOE when the source carbon is ments have been revitalized as a preferable alternative to protonated (13). Both selective one-dimensional and nonsetechniques based on pulse-interrupted free precession to lective multidimensional experiments have been employed. transfer coherence between heteronuclei. The rather ineffi-When only a few coupling constants are of interest (such cient coherence-transfer method using a continuous wave as interglycosidic 3 J CH in oligosaccharides), selective onespin lock has been replaced by multiple-pulse sequences (1dimensional experiments using shaped pulses are more effi-4) such as MLEV16, WALTZ16, and DIPSI2 borrowed cient, providing better signal-to-noise ratio and higher digital from heteronuclear decoupling. Essentially, the 15). Hahn match between heteronuclei is sought over a wide For many long-range proton-carbon couplings (for examspectral width to facilitate coherence transfer in a manner ple, interglycosidic 3 J CH in carbohydrates, and the C i O to similar to the homonuclear TOCSY or HOHAHA method. C a H i/1 , C i O to NH i01 couplings in peptides and proteins), The heteronuclear J cross polarization has found numerous either the source or the bridging heteronucleus is unprotonapplications in multidimensional NMR studies of proteins ated, and method (i) frequently becomes the only choice. and nucleic acids . Here, we propose a doubly selective However, for most experiments based on method (i) (7, 9version of the heteronuclear J cross-polarization experiment 11), the active long-range proton-carbon couplings are antiand apply it to the measurement of long-range proton-carphase, while passive proton-proton couplings are in-phase bon coupling constants.
in the resultant spectrum. The possible cancellation of reso-The difficulty in extracting long-range proton-carbon nance lines obviously complicates the interpretation of speccoupling constants using inverse detection from 13 C naturaltra. Besides, the peak-to-peak separation in an antiphase abundance samples is well known. Methods based on carbon splitting is always somewhat greater than the true coupling detection are becoming obsolete due to their poor sensitivity. constant, since the linewidth is an appreciable fraction of All published inverse-detection methods rely on pulse-interthe coupling. Therefore, it is very desirable to measure dirupted free precession to transfer coherence between protons rectly 3 J CH from multiplets with both active and passive couand carbons such as that used in the INEPT experiment. The plings in-phase. This will also make it possible to use deconbasic scheme is to correlate carbon with its long-range scavolution methods currently available when very complex lar-coupled proton partner(s). multiplets are present. Although antiphase long-range pro-There are three general classes of methods to obtain correton-carbon couplings can be refocused (8) to produce all lation between remotely coupled proton-carbon pairs. These in-phase peaks, this is at the expense of further signal loss are based on (i) defocusing the long-range 1 H-13 C couplings due to relaxation as well as the evolution of passive protoninto antiphase spin state (7-12); (ii) propagating coherence proton couplings during the additional refocusing delay. to long-range coupled proton(s) via 1 H-1 H TOCSY when Since in-phase coherence is transferred between carbon both the source and the bridging heteronuclei are protonated and proton in the heteronuclear J cross-polarization experi-(13-15); and (iii) transferring population to long-range coument, it is possible to use it as an alternative method to obtain in-phase long-range proton-carbon coupling con-* Current address: Magnetic Resonance Center, Department of Molecular stants. However, as in the homonuclear TOCSY experi-