Kavli colloquium speaker Taekjip Ha gives a Hot Topics session on: Measuring DNA mechanics on the genome scale.

Host: Chirlmin Joo

Registration
Registration is now closed. The data will be used for organizational purposes only, we won't keep the registration list in our archives longer than necessary.

Abstract:

The physical properties of DNA are sequence-dependent and modulate genome-wide local DNA bending involved in nucleoprotein complexes such as nucleosomes. However, lack of a method to measure DNA bendability on a genome-wide scale has restricted our understanding of how variations in DNA bendability along genomes may influence biological processes. We have developed a high-throughput method to measure the cyclizabilities of 110 bp DNA fragments by enzymatically selecting fragments that cyclize rapidly. We report the cyclizabilities of ~ 235,000 DNA fragments which span the entire length of chromosome V in S. Cerevisiae at 7 bp resolution, span several other genomic regions, and also includes random sequences.  We show that chromosome wide, DNA sequences have evolved to be bendable around dyads and rigid around linkers. Further, DNA is unusually rigid in the Nucleosome Depleted Region (NDR) upstream of Transcription Start Sites (TSSs). We suggest how the nucleosome remodeler INO80 may be best suited to recognize the physical landscape of DNA around the NDR and accordingly position the +1 and TSS proximal nucleosomes, while local DNA bendability can directly promote the positioning of nucleosomes deeper in the gene body. We show that +1, +2 nucleosomes have highly asymmetric DNA bendability correlated with gene expression, which may influence polymerase translocation. By measuring the cyclizabilities of random sequences, we comprehensively tabulated the effect of nucleotide content, distribution and CpG methylation state on DNA cyclizability. These measurements suggest that similar DNA bendability maps exist near TSSs of diverse organisms, that sequence-dependent DNA bendability influences various processes such as Transcription Factor binding or DNA wrapping by Topoisomerases, and that methylation of CpG cites can dynamically modulate DNA flexibility in a manner independent of local sequence context. Notably, our measurements suggest that evolution of codon selection has accommodated for sequence-dependent DNA bendability by making the physical landscape of DNA more suitable for the positioning of gene body nucleosomes than would be possible from random codon selection. Our method and compilation of a genomic-scale dataset of sequence-dependent DNA bendability measurements is a starting point for understanding how physical forces involved in DNA deformations have influenced the evolution of genomes.