Book of Abstracts: Albany 2009
June 16-20 2009
© Adenine Press (2008)
Evolutionary Constraint on DNA Structure in the Human Genome
Computational algorithms that assess evolutionary constraint on the sequence of the human genome do not account for the possibility that some nucleotide substitutions have little or no effect on the three-dimensional structure of the DNA molecule. Since DNA-binding proteins recognize structural features of DNA as well as nucleotide sequence, we suggest that natural selection may act to preserve the local shape and structure of DNA without maintaining the primary order of nucleotides. To investigate this hypothesis we developed a new computational algorithm, called Chai, to detect evolutionary constraint on DNA structure. Chai uses hydroxyl radical cleavage patterns as a measure of DNA structure, and compares cleavage patterns among genomes of different species to detect evolutionary constraint on structure.
We applied Chai to multi-species sequence alignments from the ENCODE pilot project regions of the human genome, and identified 12% of the bases as constrained?nearly twice as much constrained genomic territory as is found by nucleotide sequence-based constraint algorithms. We found that Chai regions correlate better with experimentally-determined non-coding functional elements. We used reporter assays in cultured cells to experimentally test the function of genomic regions that are uniquely identified by Chai, and found that some structure-constrained regions in the human genome act as transcriptional enhancers.
Our results support the hypothesis that the three-dimensional structure of DNA can be a substrate for natural selection. To understand genome evolution and function we suggest that it is critical to consider DNA structure as well as nucleotide sequence.
Thomas D. Tullius1, 2
1Department of Chemistry