Computational studies of RNA and DNA / Challenges and Advances in Computational Chemistry and Physics Bd.2 (PDF)

chain of proteins. This simplicity led early researchers to initially reject DNA as the potential carrier of genetic information. While the beautiful double helical structure proposed by Watson and Crick1,2 and the subsequent discovery of the triplet genetic code,3,4 explained how DNA could stock enormous amounts of information, it again suggested that structurally there was not much to study. At the core of the double helical structure was the observation that the spiral phosphodiester backbones could accommodate any Watson-Crick base pair sequence without deformation.

The first step to refining this viewpoint comes from realizing that DNA must be packed quite densely to fit into a cell. This is easily illustrated in the case of human cells which contain around 1 m of DNA (corresponding to 4 x 109 base pairs) in a nucleus within a diameter of only a few microns. Within sperm heads, the packing density is even higher. A partial explanation of how this is achieved comes from modeling DNA as a flexible rod, which naturally forms a random coil to increase its conformational entropy. But this factor alone only is not enough to account for the packing that occurs within the nucleus. As we now know, the remainder is due to protein-induced superhelical compaction leading to the complex and hierarchical structure of chromatin.

A second type of deformation was detected early in the study of DNA and concerns its overall helical form. Fiber diffraction studies already showed that the double helical structure could be modified as a function of its solvent and counterion environment. The A and B forms of the double helix first named by Rosalind Franklin5 are now structurally well-characterized and they have been joined by many other conformational families which go even further in tampering with DNA structure, by modifying its helical chirality, changing its number of strands, its base pairing and its relative strand orientations. In recent years, structural studies have been joined by single molecule manipulation experiments which offer us a new way to directly probe the mechanical properties of DNA.6 These experiments have again showed that DNA is more complex than initially expected and that, when pulled or twisted, it can undergo transitions to new and unexpected conformations.