Systematic and Applied Microbiology 34 (2011) 524–530 Contents lists available at ScienceDirect Systematic and Applied Microbiology j ourna l ho mepage: www.elsevier.de/syapm Diversity of nodule-endophytic agrobacteria-like strains associated with different grain legumes in Tunisia Sabrine Saïdi, Bacem Mnasri, Ridha Mhamdi ∗ Laboratory of Legumes, Centre of Biotechnology of Borj-Cédria, University of Tunis El Manar, Tunisia a r t i c l e i n f o Article history: Received 19 October 2010 Keywords: Agrobacterium/Rhizobium atpD PCR-RFLP Phaseolus vulgaris Phylogeny recA REP-PCR Vicia faba a b s t r a c t This study represents the first report describing the genetic diversity of nodule-endophytic agrobacteria isolated from diverse legumes and their phylogenetic relationships with the valid species of agrobacte- ria, as well as the non-recognized genomospecies of the former Agrobacterium tumefaciens (Rhizobium radiobacter). The genetic diversity of a collection of 18 non-nodulating agrobacteria-like strains, pre- viously isolated from root nodules of Vicia faba, Cicer arietinum and Phaseolus vulgaris from different geographical regions of Tunisia, was studied by REP-PCR and PCR-RFLP of the 16S-23S rDNA IGS, as well as by sequence analysis of the 16S rDNA and the housekeeping genes recA and atpD. The aim of the work was to study the genetic diversity of the different isolates and to check for any host-specificity. The results from the different techniques were congruent and suggested a specific interaction for P. vulgaris, whereas no specific endophytic interaction was observed for V. faba and C. arietinum. The phylogenetic analysis clearly indicated that some isolates were affiliated to R. radiobacter or to its non-recognized genomic species (genomovars G2, G4 and G9). However, the other isolates probably constitute new species within Rhizobium (Agrobacterium) and Shinella. © 2011 Elsevier GmbH. All rights reserved. Introduction The nomenclature and taxonomy of agrobacteria are still being debated and there is a large disagreement concerning the classi- fication of Agrobacterium (Rhizobium) species and the genus itself [8,45–47]. Whereas Agrobacterium rubi and Agrobacterium vitis are well defined separate taxa, Agrobacterium tumefaciens (Agrobac- terium radiobacter) and Agrobacterium rhizogenes are not. In fact, studies based on comparative molecular techniques have indi- cated that the systematic classification based firstly on pathogenic capacity is not stable at the species level [45], since this char- acter is carried on Ti (tumourigenic) or Ri (rhizogenic) plasmids [12,13,23,41,43]. Consequently, A. tumefaciens and A. rhizogenes are represented by strains which may be either tumourigenic, rhizo- genic, or non-pathogenic according to whether they lack either type of plasmid or not, as is the case for A. radiobacter. Further- more, the hypothesis indicating that pathogenic reactions would be supported by physiological reactions [3] has long been shown to be false [48]. Moreover, Sawada et al. [32] proposed the applica- tion of the epithet A. radiobacter in place of A. tumefaciens because it was the first one proposed (Rule 38 of the International Code ∗ Corresponding author at: Laboratory of Legumes, Centre of Biotechnology of Borj-Cedria, BP 901, Hammam-Lif 2050, Tunisia. Tel.: +216 79325948; fax: +216 79325948. E-mail address: ridha.mhamdi@cbbc.rnrt.tn (R. Mhamdi). of Nomenclature of Bacteria). Young et al. [45] also proposed that the correct name should be A. radiobacter, with the use of its type strain ATCC 19358 T , in order to avoid the confusion between the principle of nomenclatural priority and the principle of the appli- cation of types. Another taxonomic problem is related to the genus itself. Although Young et al. [45] proposed the incorporation of all the members of the Agrobacterium genus into the Rhizobium genus, this reclassification was debated because it was not supported by enough arguments [8,37]. This may be correct for A. rhizogenes (bio- var 2), which is closely related to Rhizobium [20,36,38], but not for biovar 1 and biovar 3 that were reported as incontestably differ- ent from the members of the genus Rhizobium [8]. For commodity reasons and to avoid confusion with rhizobia, we will therefore use the epithet ‘agrobacteria’ in this paper to designate the endo- phytic nodule-isolates affiliated to the former genus Agrobacterium (Rhizobium). Endophytic agrobacteria have been frequently isolated from surface sterilized nodules, roots and stems of several legume species [5,9,10,21,27,33,40], although they failed to re-nodulate their original hosts. Actually, several studies demonstrated the co- existence of agrobacteria with symbiotic rhizobial strains inside root nodules [22,40]. These non-nodulating strains are able to invade mature nodules even after 24 h incubation [22]. Some authors found that the recovery of agrobacteria from root-nodules of chickpea was induced by water deficiency [2]. However, the mechanism by which these non-symbiotic strains entered nod- ules remains unclear. Surprisingly, non-nodulating agrobacteria 0723-2020/$ – see front matter © 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.syapm.2011.01.009