details of the approach have been published (15), and an In principle, NMR spectroscopy can provide a wealth of application to the study of the side-chain dynamics of an information about molecular dynamics extending over a SH2 domain from phospholipase C-g1 in the presence and wide range of motional time scales. Studies to date have absence of a 12-residue phosphopeptide has been reported focused to a large extent on backbone dynamics through . The goal of the present Communication is to examine measurement of 15 N relaxation properties in molecules uniin detail the effects of cross correlation and cross relaxation formly 15 N labeled (1-4). An attractive feature of backbone on the fidelity of the 2 H relaxation methods presented pre-15 N relaxation studies is that the relaxation of an 15 N-NH viously. We show that accurate 2 H T 1 and T 1r values can be spin pair can simply be described by the 15 N-1 H dipolar easily obtained. interaction and to a smaller extent by 15 N chemical-shift For completeness, a brief description of the experiments anisotropy (5). While cross-correlation effects do exist bethat have been developed is provided here. The magnetizatween these relaxation mechanisms, methods have been detion-transfer pathway in either the T 1 or T 1r experiment can veloped for their efficient removal (6-8). Extension of the be described as methods developed for probing backbone dynamics to the study of side-chain motions has proven to be difficult. One approach involves either uniform (9, 10) or fractional 1 H J CH 13 C (t 1 ) J CD 2 D (T ) J CD labeling of the molecule with 13 C; in the case of side chains, however, the majority of spin systems are of the AX 2 , AMX 13 C J CH 1 H (t 2 ), [1] (methylene), or AX 3 (methyl) variety, and cross-correlation effects between the various dipolar spin pairs can significantly complicate the interpretation of the relaxation data where the active couplings involved in each transfer process . Unfortunately, such effects are difficult or imposare indicated above the arrows and t 1 , t 2 denote acquisition sible to remove, and the detailed experiments necessary to times. During the delay T, the magnetization of interest is exploit the information content of such cross-correlation of the form I z C z D z or I z C z D y , depending on whether a T 1 terms are often not possible for complex molecules such as (I z C z D z ) or T 1r (I z C z D y ) experiment is recorded. The decay proteins.
of these triple-spin terms proceeds at this point in the se-With these problems in mind, we have recently developed quence. At the completion of the delay T, the magnetization a novel 2 H-based relaxation approach for studying dynamics is returned to protons for detection by reversing the first of methyl-containing side chains. The method makes use of half of the transfer pathway. Fourier transformation of the a uniformly 13 C-, fractionally 2 H-labeled sample and selects resultant data set generates a ( 13 C, 1 H) correlation map. A set specifically for 13 CH 2 D methyl groups. Because of the poor of two-dimensional spectra is obtained, with each member of resolution and sensitivity of direct-observe 2 H spectroscopy, the set recorded with a different value of T. The intensities the experiments measure the relaxation properties of the deuof ( 13 C, 1 H) correlations as a function of T provide a direct teron indirectly by recording a series of constant-time twomeasure of the relaxation rates of the triple-spin terms, dimensional 1 H-13 C correlation spectra. Excellent resolution I z C z D z or I z C z D y . In order to obtain the relaxation of pure is therefore afforded, and high-quality spectra are obtained deuterium spin order, D z or D y , it is necessary to record an in only 2-3 hours of measuring time for sample concentraadditional experiment which measures the decay of longitudinal order I z C z and subtract the decay rate of the double-tions on the order of 1 mM. The experimental and theoretical 213