Book of Abstracts: Albany 2005
DNA Deformability Energetic Changes Associated with Single Base Pair Mutants of CDE Binding Sites within S. cerevisiae Centromere DNAs Correlate with Measured Chromosomal Loss Rates
The centromeres in Saccharomyces cerevisiae are organized by short DNA sequences (125 bp) on each chromosome consisting of 3 elements, CDEI, CDEII, and CDEIII. CDEI is an 8 bp conserved element with a 6 bp palindrome. CDEII is an ∼78-86 bp A+T rich, length conserved sequence that occurs between the other two elements. CDEIII is a 26 bp conserved element with a 7 bp core palindrome. Following assembly with proteins, CDEI and CDEIII are positioned near each other on a specialized nucleosome and are critical sequence specific protein binding sites necessary for correct centromere formation. Hegemann et al. (1) produced single base DNA mutants within the critical CDEI and CDEIII binding sites on the centromere of chromosome 6 and measured loss rates for the different chromosome 6 mutants. Olson et al. (2) used protein-DNA crystallography data to produce a dinucleotide protein deformability energetic scale (PD-scale) that describes DNA deformability by sequence specific binding proteins. In the PD-scale each of the 10 unique dinucleotides is assigned a single number that represents the sum of the energetic dimensions found for that dinucleotide. Each mutant changes 2 PD-scale values at that changed base position relative to the wild type. In this study, we have utilized these single base pair chromosome 6 mutants to demonstrate a correlation between the change in DNA deformability of the CDEI and CDEIII core sites and the overall measured loss rates of the mutants.
The log(chromosome loss rate) values correlate to the sum of deviations (unsigned) in the PD-scale with respect to the wild type state. Correlations are stronger in the known conserved core protein binding centers as compared to strains with mutations in the less critical flanking areas. The CDEI mutant group has a correlation value of 55 and an associated P-value of .0057, while the entire 64 bp mutant group has a correlation value of 33 and an associated P-value of .0077. Notably, the intercept of both plots yield a value very close to the wild type chromosome 6 loss rate point, which is not included as a data point in the correlations. Position specific, 4 data-point correlations (as above), can be created using the wild type and the 3 associated alternate base mutants. The slopes of these correlations, shown in the Table, reflect the symmetry of the DNA binding sequences. Monte Carlo analysis of these slope patterns show that it is very unlikely for these to occur with random energetic scale values (p-value<.02).
These data suggest that the PD-scale representation of the energetics of a DNA sequence, is an important simple attribute of the DNA sequence that can be used to understand the consequence of alterations in the sequence upon its interaction with sequence specific protein.
References and Footnotes
Center for Intelligent Biomaterials