NMR has become a standard tool of the biophysicist and biochemist to study structure and activity of biomolecules. The first step in any structural determination is the sequence-specific assignments of proton resonances. These assignments rely on through-space and through-bond connectivities provided by homonuclear (proton) COSY, TOCSY, and NOESY experiments. Unfortunately, as the molecules become larger, the sensitivity of through- bond experiments decreases and the complexity of the proton spectra makes it impossible to resolve unique resonances. The problem of spectral resolution is not limited to large molecules, for even smaller ones can be NMR """"""""unfriendly"""""""" with many overlapping resonances. For example, in small proteins where there can be significant CalphaH overlap in 13C and 1H dimensions. This overlap can be """"""""resolved"""""""" using the resolution of the adjacent 15NH resonance. One solution to the problem of proton resolution is to use triple-resonance three-dimensional NMR spectroscopy in combination with heteronuclear (double resonance) 3D experiments: thus, the request for the """"""""Third Channel"""""""" upgrade. These methods take advantage of the large coupling constants between 13C and 15N, and between those nuclei and their attached protons. For molecules less than 25 kDa, these couplings in the backbone resonances are large enough to allow efficient magnetization transfer and good sensitivity. The additional resolution in the 15N and 13C dimensions eliminates the spectral overlap in the proton dimension, as each proton resonance is now correlated with at least one heteronuclear frequency. For example, a 3D experiment TOCSY-HMQC (total correlation spectroscopy- heteronuclear multiple quantum correlation), correlates 15N, NH, and Halpha chemical shifts for backbone assignments in proteins. After assignment is completed, the coupling constants measured from these through-bond experiments are used in addition to distance information obtained from analogous through-space experiments to provide the constraints for structural determinations. These 3D triple-resonance experiments place many requirements on the spectrometer: three channels (with amplifiers and frequency synthesizers) are needed to generate the (at least) three frequencies (1H, 15N, 13C); likewise, triple resonance probes with the receiving and transmitting coils tuned for these frequencies are necessary. For some of the more complicated pulse sequences, a pulse generator may be necessary to provide sophisticated pulse shaping capabilities. While these features are expensive, it is clear that these techniques are successful and indeed necessary to allow structural determination of larger proteins and complexes.
