The problem of water suppression is exacerbated by the same for exchangeable and nonexchangeable species. The effects of radiation damping, particularly at high magnetic ''splicing'' of sophisticated water-suppression pulse subfields in high-sensitivity NMR probes. A new method of units into pulse sequences complicates both the phase cywater suppression is presented in which homospoil pulses cling and the timing and perhaps introduces other side effects are used to manipulate the effects of radiation damping on such as gradient echoes. the water resonance and thereby selectively alter the effec-
The efficiency and selectivity of normal pulse-sequence tive longitudinal relaxation time of the water resonance with approaches for water suppression are plagued by a combinarespect to the protein resonances. The difference in the relaxtion of five factors: (1) selective pulses are not perfectly ation time of the water and biological macromolecules can selective and generally do not have a pure phase, (2) inhothen be used in an inversion-recovery experiment to selecmogeneity of the RF field, (3) spin diffusion, (4) the water tively suppress the water as in the WEFT method of water and protein resonances have similar relaxation times, and suppression (1-3). This method allows the observation of ( ) the insidious effects of radiation damping. The first facprotein resonances which resonate near or under the water tor means that it is impossible to selectively excite either resonance. In contradistinction to the WEFT method, this the water resonance or the protein resonances and in any method is not troubled by the effects of radiation damping case some a protons resonate beneath the water resonance. and can be used even if pure H 2 O is used as the solvent.
The use of selective pulses in water suppression [e.g., WA-Water is the anathema of biological NMR studies. It is TERGATE (8, 9), flipback (11), and RAW (12)] compliboth a necessary component of biological NMR studies and cates pulse sequences through phase cycling and temporal a source of huge problems in spectral acquisition. In protein difficulties, and, as they are almost invariably not of pure NMR studies the protein is typically at a concentration in phase and as they require finite amounts of time, the effithe mM range while the water-proton concentration may be ciency of the pulse sequence is compromised. In any case, up to five orders of magnitude higher. This large difference the applied RF field is never completely homogeneous. in concentrations hampers spectral acquisition in two ways.
An elegant method for suppressing the water resonance First, the large water-proton concentration precludes the setis to use either selective (3) or nonselective inversion (1, 2) ting of the receiver gain to a level well suited to acquiring of the water resonance followed by a delay such that the the protein resonances, and second, the water resonance may water resonance is nulled but the protein resonances, by cover the resonances of some protein resonances (especially virtue of their shorter relaxation times, are already fully a protons).
relaxed. This method is virtually unique in allowing observa-Many methods exist for handling the problems of the tion of protein resonances at or near the water resonance water resonance ranging from solvent exchange to radiofrefrequency. However, with increasingly higher field strengths quency pulse sequences of varying degrees of sophistication and probe sensitivity, the effects of radiation damping be-(4-6), including gradient-based methods (7-10). Howcome more intense and the observed water relaxation time ever, all of these methods have one or more weak points. becomes very close to that of the protein resonances. As far Presaturation can be used if one is prepared to forgo the as protein studies are concerned, the salient feature of radiaobservation of resonances beneath and exchanging with the tion damping is that it affects only the water resonance. water resonance and within the effective spectral width of Radiation damping results from mutual coupling between the decoupling field. The gradient-based methods generally the receiver coil and an intense NMR resonance. The water result in diffusional losses in intensity which may not be the resonance in biological samples is particularly susceptible to the effects of radiation damping. It has the effect of forcing the resonance back to its equilibrium position at a rate that