cyclic AMP signalling and the control of infection in the Medicago symbiont Sinorhizobium meliloti – RhizocAMP

Task 1 Project coordination
Task 2 Genetic and phenotypic caracterization of the cascade
Task3: Biochemical characterization of the plant signal.
Task 4 Mechanism of plant infection control ( Genetics, biochemistry and imaging).

We have shown that rhizobia contribute to the control of their own infection level. This is mediated by an original mechanism based on cAMP signalling.

The project fits into two different perspectives:
- The mechanisms that keep nodulation and infection under control in the rhizobium legume symbiosis are presently the subject of intensive efforts worldwide.
- cAMP signalling is a so far poorly documented mode of signalling in mutualistic symbioses.

Tian CF, Garnerone AM, Mathieu-Demazière C, Masson-Boivin C and J. Batut (2012) Proc. Natl. Acad. Sci.USA 109(17):6751. Communiqué de presse de l’INRA 18.05.2012
2. 10th European Nitrogen Fixation Conference Munich (2-5 sept. 2012): talk (J. Batut)
3. Mathieu-Demazière et al. 10è Rencontres Plantes-Bactéries, Aussois (France), 30 Janvier - 3 Février 2012. Poster
4. Gastebois A,et al. 10è Rencontres Plantes-Bactéries, Aussois (France), 30 Janvier - 3 Février 2012. Poster

Plant legumes and bacteria collectively known as rhizobia have evolved a nitrogen-fixing symbiosis of major ecological importance that accounts for a fourth of the nitrogen fixed annually on Earth. Rhizobia elicit on legume plants the formation of specialized organs, called nodules, that they colonize intracellularly and within which they fix nitrogen to the benefit of the plant. The initial stage of infection generally involves the formation in root hairs of tubular structures called infection threads (ITs) that subsequently elongate into emerging nodules before delivering bacteria into plant cells.
Tight control of infection is crucial to maintain the symbiosis in a mutualistic state, ie to avoid excessive IT proliferation and root infection. Our recent work on the Sinorhizobium meliloti-Medicago sativa symbiosis unexpectedly revealed that IT formation is partly under negative bacterial control. Indeed, we have identified in S. meliloti three homologous receptor-like adenylate cyclases (CyaD1, CyaD2 and CyaK) involved in cyclic AMP (cAMP) biosynthesis and a cAMP-binding transcriptional regulator (Csr for Crp-like symbiotic regulator) whose inactivation led to a hyper-infective phenotype on M. sativa roots, despite the formation of a normal number of nitrogen-fixing nodules. Preliminary evidence indicates that activation of these adenylate cyclases depends upon a plant signal thus suggesting the existence of a hitherto unknown molecular dialogue between the two symbionts that contributes to keeping IT formation under control.
The present project aims at elucidating this cAMP-mediated mechanism of control of infection in the S. meliloti/Medicago symbiosis.
In a first task, the S. meliloti cAMP-cascade will be characterized genetically and phenotypically. Additional cascade components will be identified both through a candidate gene approach and a transcriptome approach and their role in cAMP signalling and Medicago infection will be evaluated by ex planta and in planta characterization of the corresponding mutants.
In a second task, the plant signal(s) activating the cyclases will be biochemically characterized by combining plant extracts-chemical fractioning and analysis and an activity bioassay (target gene fused to lacZ). To evaluate the organ-, species-and legume-specific character of the signal, its distribution in several tissues and leguminous and non-leguminous plants will be analyzed. To get insight in signal biogenesis, the impact of a collection of S. meliloti and Medicago truncatula symbiotic gene mutations on signal formation will be assessed. The mechanism of signal perception by the adenylate cyclases will be precised using E. coli as a heterologous reconstruction system.
In a last task, we will attempt to elucidate the mechanism(s) by which the cAMP cascade affects Medicago infection. The possibility that a csr-target gene(s) mediates infection control will be tested by detailed characterization of the csr targets whose inactivation confers a hyper-infection phenotype (including cellular localisation, impact on Nod Factors and exopolysaccharide production). Alternatively, the possibility that bacterial cAMP itself controls infection will be investigated through cAMP measurements in different ex planta and in planta conditions.