Société Française de Bioinformatique

Mots-Clés

DNA repair, Nucleosome, Molecular Dynamics,

Description

Light, oxidative stress or exogenous molecules can modify the well-designed structure of DNA by
inducing nucleobases lesions. The accumulation of these damages can hinder the DNA transcription
or replication and lead to mutations, cell apoptosis or cancers. As a consequence, numerous studies
focus on the elucidation of the mechanisms of damages formations or their repair by dedicated proteins.
Because of the complexity of the DNA molecule in its biological context, the problem becomes rapidly
combinatorial, involving sequence, structural and dynamical effects. Indeed, beyond the double strand
structure, the DNA polymer is wrapped around a core of eight proteins call histones to form
nucleosomes.[1] This specific and dynamical environment mechanically constrains the DNA
conformation and creates an heterogeneous electrostatic field, which impacts the physicochemical
properties of the nucleobases, their reactivity and their accessibility to protein interacting with DNA.
In this project, we focus on the 8-oxoguanine damage which results from the oxidation of a guanine.
This damage is known to modify the DNA conformation, and also to evolve to secondary damages
and abasic sites because of its redox properties. On the other hand, it is recognized by several proteins
such glycosylases for its repair. We thus plan in this project to use classical molecular dynamics
simulations to explore the conformational behavior of the 8-oxoguanine in the context of nucleosomal
DNA its possible interaction with repair proteins. These methods have been successfully used in our
group for simulations of damaged nucleosomal DNA [2-4] and of glycosylase-damage DNA
interaction [5]. We aim to go further in the analysis of the combinatorial mechanisms at play in the
damage behavior in the nucleosomal DNA environment by inclusion of dedicated machine learning
approaches. This project is part of a collaboration with an experimental group at IAB, Grenoble, to
propose a repair mechanism hypothesis involving different proteins. The simulations performed during
this master will be helpful to decide which damaged sequences will be selected for the experimental
and computational studies of protein and nucleosome interplay.

The candidate must have a background and/or interest in physical chemistry, biochemistry or
molecular biology. Competences in programming, computational chemistry or bioinformatics are also
welcomed. PhD application is open in the continuation of this master project.