As molecular interactions of plants with N-2-fixing endophytes are largely uncharacterized, we investigated whether the common signaling pathway (CSP) shared by root nodule symbioses (RNS) and arbuscular mycorrhizal (AM) symbioses may have been recruited for the endophytic Azoarcus sp.-rice (Oryza sativa) interaction, and combined this investigation with global approaches to characterize rice root responses to endophytic colonization.Putative homologs of genes required for the CSP were analyzed for their putative role in endophytic colonization. Proteomic and suppressive subtractive hybridization (SSH) approaches were also applied, and a comparison of defense-related processes was carried out by setting up a pathosystem for flooded roots with Xanthomonas oryzae pv. oryzae strain PXO99 (Xoo).All tested genes were expressed in rice roots seedlings but not induced upon Azoarcus sp. inoculation, and the oscyclops and oscastor mutants were not impaired in endophytic colonization. Global approaches highlighted changes in rice metabolic activity and Ca2+-dependent signaling in roots colonized by endophytes, including some stress proteins. Marker genes for defense responses were induced to a lesser extent by the endophytes than by the pathogen, indicating a more compatible interaction.Our results thus suggest that rice roots respond to endophytic colonization by inducing metabolic shifts and signaling events, for which the CSP is not essential.

Azoarcus sp. strain BH72 is an endophytic betaproteobacterium able to colonize rice roots without induction of visible disease symptoms. BH72 possesses one polar flagellum. The genome harbors three copies of putative fliC genes, generally encoding the major structural protein flagellin. It is not clear whether, in endophytic interactions, flagella mediate endophytic competence or act as MAMPs (microbe-asscociated molecular patterns) inducing plant defense responses. Therefore, possible functions of the three FliC proteins were investigated. Only fliC3 was found to be highly expressed in pure culture and in association with rice roots and to be required for bacterial motility, suggesting that it encodes the major flagellin. Endophytic colonization of rice roots was significantly reduced in the in-frame deletion mutant, while the establishment of microcolonies on the root surface was not affected. Moreover, an elicitation of defense responses related to FliC3 was not observed. In conclusion, our data support the hypothesis that FliC3 does not play a major role as a MAMP but is required for endophytic colonization in the Azoarcus-rice interaction, most likely for spreading inside the plant.