The solution conformation of gramicidin S in deuterated dimethyl sulfoxide was investigated by using the intramolecular nuclear Overhauser effect experiment. Experimental Overhauser enhancements were compared with predicted values for each ofthe nine most-stable conformations (M1-M9) calculated by Dygert et al on the basis of energy-minimization procedures [Dygert, M., Go, N. & Scheraga, H. A. (1975) Macromolecules 8,[750][751][752][753][754][755][756][757][758][759][760][761]. By using statistical hypothesis testing, the three lowestenergy conformations (MI, M2, and M3) were shown to give the best fit with the experimental data. All other conformations (M4-M9) were found to be inconsistent with the experimental data.Intramolecular nuclear Overhauser effects (NOEs) are very sensitive to interproton distances when dipole-dipole interaction is the predominant mechanism ofproton relaxation (1). Because of this distance dependence, NOE measurements have been used extensively in conformational studies of small molecules (1); however, their application to peptides is relatively recent (2-7).The cyclic decapeptide antibiotic gramicidin S has been the subject ofmany experimental [ref. 8 perature was 297 ± 1 K. In the NOE experiment, the resonance signal of interest was selectively presaturated for 4 sec before the observation pulse. The decoupler power was set such that the irradiation was highly selective; the decoupler power spillover to the neighboring resonance signals was less than 2% in general, as determined from the difference spectra. For the control spectrum, the irradiation frequency was set away from the spectral region by more than 2 kHz. The degree of saturation of the irradiated lines was 80-90%, and the observed NOEs were finally corrected for 100% saturation of the irradiated peak by dividing by the degree of saturation. One thousand scans were accumulated for each spectrum. To minimize the effects associated with spectrometer drift and to obtain clean difference spectra, the accumulation was alternated every eight scans between the control and the NOE experiments. A 0.5-Hz exponential line broadening was used in Fourier transforming the free induction decays. The NOE difference spectra were obtained by subtracting the control spectrum from the double resonance spectra. The NOE enhancements were calculated as fractional intensity changes for each resonance signal. For overlapping signals and for the C'H2 multiplet of D-phenylalanine, integrated intensities were used in computing the NOEs.
RESULTS AND DISCUSSIONThe nuclear Overhauser effect experiment measures the steadystate effect of saturation of a specific spin transition or group of transitions on the intensities of other resonances in the NMR spectrum. The fractional change in the intensity of spin i when a spin s (or a group of such spins) is saturated can be computed by solving a set ofsimultaneous equations of the form (2) oijff(s) + I oyfj(s)where the summation on the right-hand side involves only the saturated spins. The oii and (ri...