Delphine Capela

Our team is interested in the molecular and evolutionary mechanisms involved in the establishment of the rhizobium-legume symbiosis, mainly focussing on the bacterial partner.


Rhizobia do not form a homogeneous taxonomic group but belong to more than ten genera and hundreds of different species among the α- and β-proteobacteria. Phylogenetic, genetic and genomic data predict that rhizobia evolved through lateral transfer of key symbiotic functions in various soil bacteria, followed by adaptation of the host genome to the legume niche by natural selection. The post-transfer adaptation mechanisms and genetic predispositions required, if any, for the conversion of these soil bacteria into effective legume symbionts are not known.

Our work seeks to understand the emergence and evolution of these mutualistic interactions by addressing two questions:

i) how the symbiotic capacity was disseminated to very different genera of bacteria

ii) what molecular events have accompanied the genesis of this symbiosis on the plant side?

For this, we are studying the Cupriavidus taiwanensis-Mimosa pudica model, a symbiosis predicted to have  emerged relatively recently (around 12 to 16 million years ago) and representative of the symbioses of the Mimosoideae subclade. We use molecular genetics, genomics and experimental evolution approaches.



Experimental evolution of the phytopathogenic bacterium Ralstonia solanacearum into legume symbionts

Over 10 years ago, the team launched an evolutionary experiment, pioneering internationally, aimed at reproducing the evolution of a new genus of rhizobia under laboratory conditions, based on the natural evolutionary history of these bacteria. For this, we introduced the symbiotic plasmid of the rhizobium Cupriavidus taiwanensis into a non-rhizobium bacterium of a related genus, Ralstonia solanacearum.


The resulting chimeric Ralstonia was evolved under selection pressure from the host plant of C. taiwanensis, Mimosa pudica, with the aim of activating and / or successively improving the various symbiotic stages, namely nodulation, infection, and persistence of bacteria in plant cells and possibly mutualistic nitrogen fixation. This experience and its analysis have already enabled major advances in the understanding of the evolution of rhizobia. The team has indeed shown that: i) the first two stages of the symbiosis, nodulation and intracellular infection, were rapidly acquired and optimized, the clones isolated from most of the final populations exhibiting similar infection capacities. those of the natural rhizobium C. taiwanensis after 17 cycles of evolution (approximately 400 generations) (Marchetti et al. 2017), ii) the extracellular pathogeny-intracellular symbiosis transition was based on modifications of the regulatory network of the recipient bacterium (Marchetti et al. 2010, Guan et al. 2013, Capela et al. 2017, Tang et al. 2020), iii) the co-transfer of the symbiotic genes with mutability genes present on the symbiotic plasmid accelerates the evolution of rhizobia in the laboratory and probably also in nature (Remigi et al. 2014), iv) nodulation cycles longer than 21 days are favorable to the emergence of nitrogen fixation (Daubech et al. 2017) and finally v) the flight The natural use and experimental evolution of Mimosa symbionts show notable similarities despite differences in ancestors, evolution times and conditions (Clerissi et al. 2018).

This experiment continues in the team with the idea of ​​continuing the adaptation of bacteria to the symbiosis potentially until mutualism is obtained. The experiment is now analyzed in a more dynamic and exhaustive way by combining approaches of sequencing populations of bacteria, microbial genetics, phenotyping and modeling. The aim of this project is to identify not only the genetic bases for the adaptation of bacteria to symbiosis and the pathogenesis-symbiosis transition, but also the selection forces that have driven the evolution of legume symbionts.

Shaping new bacterial symbionts

Analysis of plant transcriptomic responses to the adaptation of bacteria to endosymbiosis

The objective of this project is to exploit the unique biological material generated by the evolutionary experiment to analyze the plant responses to bacterial evolution. A collection of quasi-isogenic Ralstonia mutants is constructed as the adaptive mutations responsible for the acquisition of symbiotic traits are identified.


The transcriptomes of Mimosa pudica obtained with the various mutants of Ralstonia progressively adapted to the symbiosis are compared with the transcriptomes of Mimosa obtained with the natural symbiont C. taiwanensis.


This project aims to identify the plant molecular mechanisms that control these different processes (nodulation, infection, persistence and mutualism), and potentially to predict the sequence of plant molecular events that accompanied the genesis of the rhizobium-legume symbiosis.


Our main collaborators are PM Delaux (LRSV, Toulouse), JB Ferdy (EDB, Toulouse), C. Pouzet and A. Leru (Imaging platform Toulouse), E. Rocha (Pasteur Institute, Paris) and D. Roche (Genoscope-CEA , Evry).

Ongoing funding

SPE INRAE ​​EPIMODE (2021-2022) Impact of epigenetic modifications in bacterial adaptation to a new way of life. Partner: A. Guidot (LIPME), 15 k €.

FRAIB DYNAMIC (2021-2022) Evolutionary dynamics of bacterial adaptation to symbiosis with legumes. Partner: JB Ferdy (EDB, Toulouse), € 15k.

ANR REPLAY (2017-2021) Replaying the evolution of rhizobia: towards a conceptual and practical framework for the design of new nitrogen-fixing plant symbionts. Partners: Genoscope (D. Roche), IP Paris (E. Rocha), 460 k €.