Applications of 1 H a – 1 H d correlated 2h J (N,N) spectroscopy for identifying H a —C—N a ···H d —N d hydrogen bonds in nucleic acids

Trans‐hydrogen bond coupling constants across 15 N d —H··· 15 N a hydrogen bonds [ 2h J (N,N)] have been immensely beneficial in providing direct evidence for these hydrogen bonds in nucleic acids and their complexes. A very commonly occurring hydrogen bonding topology is the H a —C a —N a ···H d —N d alignment, which is usually identified via 1 H d – 15 N a correlated spectra (HNN‐COSY, soft HNN‐COSY) or 1 H a – 15 N d correlated spectra [H(C)NN(H)]. In these experiments, the donor/acceptor nitrogens need to be assigned independently, which is often compromised by degeneracy problems among the 15 N nuclei. Recently, we have designed NMR techniques for alleviating these problems by providing direct 1 H a – 1 H d correlations across H a —C a —N a ···H d —N d hydrogen bonds in G·G·G·G tetrads and sheared G·A mismatches. Here, we present further demonstrations of these H(NN)H sequences as useful tools for identifying H a —C a —N a ···H d —N d alignments in G·G and A·A mismatches, and reverse A–T Watson–Crick base pairs, on two uniformly 13 C/ 15 N‐labeled dimeric DNA molecules, d(GGAGGAT) and d(G 2 A 2 G 2 T 3 G 3 AT). Copyright © 2001 John Wiley & Sons, Ltd.