Stochastic models of nucleosome dynamics reveal regulatory rules of stimulus-induced epigenome remodeling

The genomic positions of nucleosomes are a defining feature of the epigenomic state and hence of cell identity. Signal-dependent transcription factors (SDTFs), upon activation, modify the positioning of nucleosomes and cause epigenome remodeling. Here, we developed Markov models of nucleosome wrapping and unwrapping, and fit them to high resolution deep sequencing data of DNA accessibility to reveal biophysical principles of nucleosome dynamics. We found that 1) the dynamics of DNA unwrapping are significantly slower in vivo than reported from in vitro experimental data, 2) there is clear evidence for cooperativity in wrapping and unwrapping, 3) SDTF activity produced highest eviction probability when its binding site is close to but not on top of the nucleosome dyad, and 4) oscillatory SDTFs produce more variability than constant SDTF activities. Our work uncovers the regulatory rules governing nucleosome dynamics in vivo, which can predict epigenomic alterations during inflammation at single nucleosome resolution. Paired-end ATAC-sequencing of mutant macrophage cells at 0hrs and treated with TNF and LPS at 4hrs.