Solution Structures of the Huntington's Disease DNA Triplets, (Cag)n.
SVS Mariappan,LA Silks,X Chen,PA Springer,RL Wu,RK Moyzis,EM Bradbury,AE Garcia,G Gupta
DOI: https://doi.org/10.1080/07391102.1998.10508988
1998-01-01
Abstract:Highly polymorphic DNA triplet repeats, (CAG)n, are located inside the first exon of the Huntington's disease gene. Inordinate expansion of this repeat is correlated with the onset and progression of the disease. NMR spectroscopy, gel electrophoresis, digestion by single-strand specific P1 enzyme, and in vitro replication assay have been used to investigate the structural basis of (CAG)n expansion. Nondenaturing gel electrophoresis and 1D 1H NMR studies of (CAG)5 and (CAG)6 reveal the presence of hairpins and mismatched duplexes as the major and minor populations respectively. However, at high DNA concentrations (i.e., 1.0-2.0 mM that is typically required for 2D NMR experiments) both (CAG)5 and (CAG)6 exist predominantly in mismatched duplex forms. Mismatched duplex structures of (CAG)5 and (CAG)6 are useful, because they adequately model the stem of the biologically relevant hairpins formed by (CAG)n. We, therefore, performed detailed NMR spectroscopic studies on the duplexes of (CAG)5 and (CAG)6. We also studied a model duplex, (CGCAGCG)2 that contains the underlined building block of the duplex. This duplex shows the following structural characteristics: (i) all the nucleotides are in (C2'-endo, anti) conformations, (ii) mismatched A x A base pairs are flanked by two Watson-Crick G x C base pairs and (iii) A x A base pairs are stably stacked (and intra-helical) and are formed by a single N6-H--N1 hydrogen bond. The nature of A x A pairing is confirmed by temperature-dependent HMQC and HMQC-NOESY experiments on the [(CA*G)5]2 duplex where the adenines are 15N-labeled at N6. Temperature- and pH-dependent imino proton spectra, nondenaturing electrophoresis, and P1 digestion data demonstrate that under a wide range of solution conditions longer (CAG)n repeats (n> or =10) exist exclusively in hairpin conformation with two single-stranded loops. Finally, an in vitro replication assay with (CAG)8,21 inserts in the M13 single-stranded DNA templates shows a replication bypass for the (CAG)21 insert but not for the (CAG)8 insert in the template. This demonstrates that for a sufficiently long insert (n=21 in this case), a hairpin is formed by the (CAG)n even in presence of its complementary strand. This observation implies that the formation of hairpin by the (CAG)n may cause slippage during replication and thus may explain the observed length polymorphism.