Thèse Variabilité Épigénétique Interindividuelle et Stress Abiotique Précoce chez Arabidopsis H/F - 75

Établissement : Université Paris-Saclay GS Biosphera - Biologie, Société, Ecologie & Environnement, Ressources, Agriculture & Alimentation
École doctorale : Sciences du Végétal : du gène à l'écosystème
Laboratoire de recherche : IJPB - Institut Jean-Pierre Bourgin-Sciences du Végétal
Direction de la thèse : Vincent COUSTHAM ORCID 0000000253992723
Début de la thèse : 2026-10-01
Date limite de candidature : 2026-05-06T23:59:59Chez les animaux et l'Homme, certaines régions du génome ont un niveau de méthylation de l'ADN fortement variable entre individus de manière indépendante de la séquence d'ADN. Les VMR (pour Variable Méthylation Régions) sont distribuées de manière non-aléatoires dans les régions régulatrices de gènes dont les fonctions sont notamment associées à la plasticité du développement. Le degré de méthylation d'une partie des VMR serait déterminé lors des premières étapes du développement embryonnaire, pré-différenciation, et pourrait ainsi être un chainon manquant du mécanisme de mémoire des expositions précoces connu pour contribuer aux origines développementales de la santé et des maladies. Chez les plantes, la variabilité épigénétique est fortement structurée par la géographie et le climat, et contribuerait aux processus évolutifs à court et moyen terme. Cependant, le degré de variation interindividuelle n'a jamais étudié de manière directe, ce qui fait que ce mécanisme potentiel de mémorisation des expositions précoces reste inexploré chez les végétaux.

Le projet de thèse a pour objectif de mettre en évidence les VMR chez la plante modèle Arabidopsis par l'étude du méthylome foliaire d'un grand nombre d'individus isogéniques cultivés en condition homogène. Une caractérisation approfondie des VMR (contexte de méthylation, distribution génomique, fonction des gènes associés...) sera réalisée. Grâce à deux protocoles de stress sur la graine et la plantule en développement, le projet visera à vérifier l'influence de l'environnement précoce sur le statut de méthylation des VMR. La relation entre les VMR, l'expression génique et la structure de la chromatine sera étudiée à l'aide d'approches tout-génome. Le ou la doctorant.e
aura à sa disposition une collection de mutants, d'écotypes et de lignées recombinantes épigénétiques (epiRIL) pour explorer le statut des VMR par des approches ciblées. L'édition d'épigénome ciblée pourra être utilisée pour valider la fonction des VMR d'intérêt. Enfin, la transmission ou non du statut de méthylation sera vérifiée dans la descendance, afin d'évaluer la nature stochastique ou héritable du phénomène.

The term epigenetics refers to the molecular mechanisms that regulate gene expression, are reversible and transmissible during development, and sometimes between generations, without altering the DNA sequence. The epigenome acts as an interface between the genome and the environment and contributes to phenotypic variation. However, even in the absence of genetic or environmental variation, it has been shown in human and other vertebrates that some regions exhibit high stochastic variability in DNA methylation levels (Feinberg and Irizarry, 2010; Gu et al., 2016; Lefort et al., 2026). These regions, called Variable Methylation Regions (VMRs), are not distributed randomly but are preferentially located in regulatory sequences of genes involved in developmental plasticity. It has been proposed that VMRs, under the combined influence of genetics and environment, contribute to phenotypic plasticity (Garg et al., 2018), although the cause-and-effect relationship remains to be clarified. It has also been shown that a significant number of VMRs are common between tissues with different functions and have correlated methylation levels (Derakhshan et al., 2022; Lefort et al., 2026). This suggests that the methylation status of VMRs can be determined early before differentiation, at the beginning of embryonic development, a period of development that is particularly sensitive to environmental fluctuations (Junien et al., 2016).

In plants, variation in DNA methylation has been extensively documented in epiRILs (Furci et al., 2019; Johannes et al., 2009), multigenerational lines (Becker et al., 2011; van der Graaf et al., 2015), natural accessions (Baduel et al., 2025; Dubin et al., 2015; Kawakatsu et al., 2016; Vaughn et al., 2007) and under stress conditions (Van Dooren et al., 2020). Plants exhibit stochastic methylation changes at individual CG sites at a high rate, although average levels across genes remain stable during cell divisions (Goeldel and Johannes, 2023). The frequency of the stochastic events varies considerably across the genome, mainly affecting transposable elements (Baduel et al., 2025), but also gene bodies (Zhang et al., 2020). Variability in DNA methylation at the gene body (gbM) level has been proposed as a result of genetic and/or environmental factors and is thought to influence DNA methylation homeostasis. In addition, the 1001 epigenomes' sequencing project in A. thaliana demonstrated that methylation variability correlates with geography and climate of origin, and that genes linked to variable DNA methylation regions showed particularly variable expression levels depending on tissue or environment (Kawakatsu et al., 2016). The most variable genes were enriched in functions related to biotic responses and temperature, which probably reflects adaptation to their natural environment. However, inter-individual variation in DNA methylation in the form of VMR within a homogeneous population and its potential structural, gene and phenotypic consequences have never been directly investigated in plants.

The aim of the thesis is to identify hypervariable regions of DNA methylation at the interindividual level in Arabidopsis and to carry out an in-depth functional characterisation of these regions. By combining abiotic stress protocols, the project will aim to highlight the effect of early environmental exposures on these regions. The intergenerational stability of the methylation level of these regions will also be investigated.

Task 1: Mapping and in sillico analysis of VMRs
All plants were derived from a single Col0 founder to minimize genetic variation. F1 progeny were either cultivated in standard control conditions or exposed to elevated temperature during seed production. F2 sister plants were grown under homogeneous conditions on the IJPB Phénoscope, a high-throughput phenotyping platform that handle and monitor hundreds of individual pots. Half of F2 plants were grown in control conditions while the other half were subjected to a drought stress, generating four experimental conditions to study variably methylated regions (VMRs) across different environments. Rosette leaves from a largen umber of plants were sampled for ONT long-read sequencing and sequencing results will be available at the start of the project, enabling simultaneous methylome and genome analysis. VMRs will be identified across CG, CHG, and CHH contexts and analyzed in relation to transposable elements, regulatory regions, and TADs, integrating public Arabidopsis methylome datasets. Stress-induced hyper- and hypomethylation, as well as novel VMR emergence, will be assessed, alongside phenotyping of growth, seed production, and germination, to link methylation variation with functional outcomes.

Task 2: Functional characterization of VMRs
RNA-seq will be performed on the same samples to assess whether VMR status correlates with gene expression. ATAC-seq will be performed to evaluate chromatin accessibility at these regions. If VMRs are identified at loci with relevant biological functions, particularly related to temperature or drought responses, epigenome editing will be used to manipulate VMR methylation status to investigate the effects on gene expression and associated phenotypes.

Task 3: Intergenerational stability of VMR methylation
To study the reversible nature of VMRs across generations, F3 seedlings derived from F2 plants will be grown on the Phénoscope under identical environmental conditions, with or without stress. ONT methylomes will be analyzed to determine whether the same regions vary and whether methylation levels correlate across generations, complemented by targeted assays and RT-qPCR of nearby gene expression.