Multi–scale study of chromatin organisation and function: DNA topology, epigenetics and chromatin compaction

Abstract

Understanding chromatin organisation at different length scales is still one of the most puzzling challenges in biophysics. Nowadays, it is clear that DNA or chromatin conformational changes can profoundly affect gene expression. Yet, the mechanisms underlying such conformational changes remain elusive. Several factors can intervene in gene regulation: supercoiling (SC), the extent of over– or under– twist of DNA double helix, can compact DNA in both bacteria and eukaryotes, yielding transcriptional over–expression or repression. Post-translational modifications of histone tails demarcate the “epigenetic” domains, which are therefore vital to establish the correct chromatin environment. Chromatin–binding proteins can form biological “condensates” via phase separation mechanisms. Recently, liquid–liquid phase separation (LLPS) has much been touted to motivate the formation of protein clusters in vivo, often referred to as ‘nuclear bodies’. In addition, the so-called bridging-induced phase separation (BIPS), explains how protein aggregation can be mediated by chromatin only, even in the absence of protein-protein interaction. By using a multi-technique approach, in this thesis’ work I investigate the structural and dynamical properties of DNA and chromatin at different length scales. Monte Carlo algorithms were implemented to simulate SC dynamics in a stochastic model for bacterial transcription. Similar techniques were used to show that an infection–like model can entail epigenetic bistability. Molecular dynamics simulations were employed to study the static and dynamical properties of model protein aggregates; the interplay between LLPS and BIPS was explored, showing properties which go far beyond the liquid state. Depending on the parameters, solid–like, glassy and fractal protein condensates can co–localise with chromatin.