Accurate determinations of one-bond 13C-13C couplings in 13C-labeled carbohydrates

Carbon plays a central role in the molecular architecture of carbohydrates, yet the availability of accurate methods for (1)D(CC) determination has not been sufficiently explored, despite the importance that such data could play in structural studies of oligo- and polysaccharides. Existing methods require fitting intensity ratios of cross- to diagonal-peaks as a function of the constant-time (CT) in CT-COSY experiments, while other methods utilize measurement of peak separation. The former strategies suffer from complications due to peak overlap, primarily in regions close to the diagonal, while the latter strategies are negatively impacted by the common occurrence of strong coupling in sugars, which requires a reliable assessment of their influence in the context of RDC determination. We detail a (13)C-(13)C CT-COSY method that combines a variation in the CT processed with diagonal filtering to yield (1)J(CC) and RDCs. The strategy, which relies solely on cross-peak intensity modulation, is inspired in the cross-peak nulling method used for J(HH) determinations, but adapted and extended to applications where, like in sugars, large one-bond (13)C-(13)C couplings coexist with relatively small long-range couplings. Because diagonal peaks are not utilized, overlap problems are greatly alleviated. Thus, one-bond couplings can be determined from different cross-peaks as either active or passive coupling. This results in increased accuracy when more than one determination is available, and in more opportunities to measure a specific coupling in the presence of severe overlap. In addition, we evaluate the influence of strong couplings on the determination of RDCs by computer simulations. We show that individual scalar couplings are notably affected by the presence of strong couplings but, at least for the simple cases studied, the obtained RDC values for use in structural calculations were not, because the errors introduced by strong couplings for the isotropic and oriented phases are very similar and therefore cancel when calculating the difference to determine (1)D(CC) values.