Doctorat / PhD — Procedure no. DSMBBC/2027/08: DNA twist and topology in target search by DNA-binding proteins

Doctorat / PhD Profil chercheur : First Stage Researcher (R1) Type de contrat : Temporary Date limite de candidature : 30 Nov 2026 - 23:59 (Europe/Warsaw) Date de début : 1 Mar 2027 Programme UE : Not funded by a EU programme Domaine : Biological sciences

Background. Proteins that target specific genomic sites - repair enzymes, recombinases, transcription factors - must locate them among millions of near-identical sequences. It is well established that they do this by facilitated diffusion, alternating between 3D diffusion in solution and 1D sliding along DNA. Most of these observations were made in vitro on topologically relaxed DNA. In cells, however, DNA is kept slightly underwound, and transcription, replication and loop extrusion constantly add local twist. Live-cell imaging, in turn, preserves this native topological state but cannot control or measure it locally. Recent single-molecule work shows the emerging supercoiling matters - for example underwinding of the DNA makes Cas9 bind and cut at multiple off-target sites (Newton et al., 2023), and RecA filaments only engage with homologous sequences when that target is negatively supercoiled (Tišma et al., 2026). Methodological challenge. Existing methods to supercoil DNA in single-molecule imaging rely mainly on intercalators, which alter twist and helix stability. The main alternative avoids dyes but requires pulling the DNA far beyond its normal length first, which damages the duplex. In this project, we will establish a new method for generating defined negative supercoiling in optical tweezers, free of intercalators and without mechanical overstretching, reaching physiological supercoiling densities. This method will be independently validated on magnetic tweezers, where twist can be applied directly. Research question. Does underwinding change how a protein moves along DNA, or how it recognises its target site? Model system. The project will start with Fpg, a bacterial glycosylase that finds and removes 8-oxoguanine, the most common oxidative lesion in DNA. It is a well characterised system: its sliding along DNA has been described in detail, but mainly on relaxed DNA, and not under controlled torsion. Fpg will give a direct mechanical readout - after removing the damaged base it cuts the DNA backbone, and a single cut releases all the torsional strain at once. In parallel, we will develop a strategy to purify a fluorescently labelled protein to monitor its sliding on torsionally constrained DNA using fluorescence-coupled optical tweezers. Once established, the same assay can be applied to other DNA-binding proteins which differ in how much they need to distort the double helix to recognize their target: from flipping out a single base to opening a full RNA–DNA hybrid. Outcome. The project will establish whether DNA topology determines how repair enzymes locate damage and therefore whether repair is intrinsically biased towards transcribed and replicated regions, where negative…