DNA Properties Specific for Protein/DNA Free Energy Changes Caused by Single Nucleotide Substitutions

Abstract

Single nucleotide polymorphism (SNP) analysis is the crossroad of genomics and drug-design. Currently, over 10000 SNPs of codons and splice sites and at most 200 SNPs within regulatory DNA are known. Thus, the alterations in codons and splice sites, and, hence, in protein structure and function are studied better than those of transcription factor binding sites (TF-sites). Among the recent reports focused at DNA/protein-interaction, only few describe all possible spectrum of single nucleotide substitutions causing free energy changes, ∆∆G’s, namely: the Inr [3], GCC [2], c-Myb site [4, 8], and the OR1-operator [6, 7]. To apply ∆∆G’s for site recognition, the linear additive approximation, ∆∆GSCORE(S) ≡ ∑ i ∆∆G(si), was introduced [1]. In this work, we have studied these ∆∆G-sets. For the OR1, Inr, c-Myb, GCC sites, three common regularities of the single substitutions were found. First, half of all substitutions with the lesser ∆∆G’s cluster together, whereas the rest substitutions with the higher ∆∆G’s cluster together too. This regularity we denote as dependence upon “∆∆G-location”. Second, for the sites with two ∆∆G sets measured in different measurement systems (e.g., in vivo and in vitro), the linear correlation between the ∆∆G’s was observed only for one half-set of substitutions, either with the lesser or with the higher ∆∆G’s. This regularity was denoted as “∆∆G-system” dependence. Third, we have revealed the conformational DNA properties such that their mean values averaged upon the site sequence with single substitutions were correlating either to lesser or to higher ∆∆Gs. This regularity was named as “∆∆G-conformation” dependence. Using these regularities, we have designed the tools predicting the ∆∆G’s for single and multiple substitutions.