Department of Chemistry, Rutgers, the State University of New Jersey, New Brunswick, New Jersey 08903, U.S.A.
The idea that DNA base sequence is more than a carrier of the genetic blueprint first arose from statistical analyses of residue occurrences in natural nucleotide sequences. The bending, twisting, and stretching observed at individual residues in nucleic acid crystal structures have shed important clues for deciphering the spatial code hidden within the double helix. This sequence-dependent structural heterogeneity, while less impressive on a local structural scale than that observed in proteins, plays a crucial role for all processes involving DNA recognition. As part of a program to understand how local base sequence effects reveal themselves in the structure of the long threadlike polymer, we have been developing knowledge-based "energy" functions of the local angular and translational movements of adjacent bases pairs and new computational tools for optimizing the configurations of long closed circular and looped chain molecules based on these data. The energies are based on harmonic analyses of the computed base pair parameters of dimer steps in A-DNA, B-DNA, and protein-DNA structures stored in the Nucleic Acid Database. The polymer computations are a simple application the theory of anisotropic elastic rods. The long DNA molecule is divided into a finite number of elements, the physical properties and shapes of which can differ from one unit to the next. The three-dimensional arrangements and associated energies of the system as a whole are expressed in terms of the natural bending, twisting, and stretching of the elements. The effects of external forces, such as those which might be associated with binding proteins and long-range self-contacts, can also be included. The energetically preferred configurations of the system are then obtained by numerical solution of a set of non-linear algebraic equations. The local conformational mechanics of neighboring base pairs not only determines the secondary structure of successive residues but also mediates the folding and interactions of long chain fragments. In other words, bound proteins and/or native sequence help to determine which parts of the DNA will influence one another; without such guidance, widely separated parts of the long chain molecule might have little or no contact. Thus, the specific geometry of a long sequence of base pairs in one part of the DNA polymer can potentially affect actions at other sites.