Visualization of an unwound DNA duplex

Abstract

Certain dyes^1,2 and drugs^3,4 with planar aromatic components can intercalate these into stacks of base pairs and thereby bind tightly to DNA duplexes. Intercalation at one site usually precludes intercalation between the base pairs immediately adjacent⁵. This exclusion implies that two distinct nucleoside conformations are needed in the dinucleoside phosphates which include the intercalation site. The simplest distinction would involve no more than quantitative differences in the (usually anti) conformations at the glycosidic bonds⁶. This could be reinforced by additional, qualitative differences in the furanose ring puckerings (C-2′-endo and C-3′-endo)⁷. For the most pronounced difference there could be qualitative differences (syn and anti) in the conformations of the glycosidic bonds as well as in the conformations of the sugar rings. The model discussed here is an example of this most emphatic distinctiveness, as the nucleosides at the 5′ ends of the intercalation sites are C-3′-endo and syn and at the 3′ ends are C-2′-endo and anti. X-ray diffraction analysis suggests that a completely unwound allomorph of the DNA duplex can persist in oriented fibres when stabilized by certain platinum-containing intercalators. In the untwisting of (usually) right-handed DNA double helices, unwound duplexes are presumably fleeting intermediates.