Proton-detected hetero-TOCSY experiments with application to nucleic acids

Abstract

Homonuclear polarization transfer in liquids based on isotropic mixing (I, 2) has become a method of choice for two-dimensional correlation spectroscopy. Sensitivity, the capacity to produce pure in-phase absorption spectra, and, most importantly, the ability to correlate resonances via multiple steps OF net coherence transfer have all contributed to its popularity. Although one of its common names, “HOHAHA” (2) is derived from its analogy to the rotating-frame heteronuclear cross-polarization experiments performed on solids (3-5), heteronuclear polarization transfer in liquids has relied almost exclusively on laboratory-frame methods such as INEPT (6) and DEPT (7). The reasons for this are clear. Scalar coupling constants in liquids are generally much smaller than dipolar coupling constants in solids, which makes the requirements for energy matching stringent. Since RF inhomogeneity severely degrades contact efficiency, double-tuned, single-coil probe assemblies are usually needed for J cross-polarization experiments. Additionally, many low-y nuclei resonate over a considerably wider bandwidth than protons, so that exceedingly high power levels are needed to overcome off-resonance effects in nonselective experiments. Following earlier suggestions (8, 9), several authors recently demonstrated (10-14) that very efficient heteronuclear J cross polarization in liquids can be achieved by using synchronous composite pulse sequences such as WALTZ16 (1.5) or MLEV16 (16) instead of single-pulse spin locking. These sequences were designed for heteronuclear decoupling over wide frequency ranges with moderate power requirements and compensate for the effects of offset and RF inhomogeneity. They generate efficient polarization transfer over bandwidths wider than those of one-pulse methods and bypass the need for single-coil assemblies. Heteronuclear “isotropic mixing” (10, 14) [or, better, hetero-TOCSY (12)] generates polarization transfer as efficient as that of INEPT or DEPT (14), with the additional advantage of enhancing long-range (14) and relayed (vide infra) couplings simply by lengthening the contact time. My interest in heteronuclear J cross polarization was sparked by problems encountered in assigning proton and “P resonances in RNA. In both RNA and DNA, 3’P chemical shifts and 31P-‘Hst coupling constants contain valuable conformational information (17). However, sequence-specific assignment of 3’P resonances is often limited by difficulties in assigning 3’ proton resonances using standard ‘H-‘H NMR