Soils and Crops Research and Development Center, Agriculture and Agri-Food Canada, 2560 Hochelaga Blvd, Sainte-Foy, PQ, G1V 2J3 Canada
This session brings significant new findings on the evolution of legume nodulating bacteria from a phylogenetic point of view as well as on the genome evolution in relation to ecological factors. During the last decade, a large number of reports on the diversity of nodulating bacteria revealed that they evolved and adapted in response to various environmental factors, host legume species and ecological niches. Evolutionary relationships among bacteria are usually estimated by sequence comparisons of the highly conserved 16S rRNA genes. However, phylogenetic analysis of genes involved in the establishment and the function of the symbiosis and of those associated in the adaptation to selective pressure increases our knowledge on their transfer and their origin. Furthermore, the use of elegant molecular techniques is leading to a more comprehensive understanding of the evolutionary changes that confer fitness to the environment and of the factors responsible for this evolution in the nodule forming bacteria.
2. From the Evolution of Nodulation Genes to That of 16S rRNA Genes
A very interesting finding presented by C. Boivin-Masson came out from an extensive phylogenetic analysis of the nodulation genes (nodA), for which a strong correlation was found between the NodA protein and the Nod factor (NF). The phylogenetic analysis of strains harboring novel NodA sequences showed that they belong to the genus Burkholderia. This genus (and a new Ralstonia sp. isolated from Mimosa spp. in Taiwan) is in the /3-subclass of Proteobacteria. All other genera of nodulating bacteria ("Rhizobium" named species and Methylobacterium) reported up to now belong to the a-subclass. It has been already suggested that the nodulating capacity is distributed laterally among distant taxa. The present finding supports this hypothesis and suggests that nod genes from both a and p nodulating bacteria were acquired through horizontal gene transfer. Furthermore, a correlation was observed between nodA and 16S rRNA phylogenies of the strains studied. Other recent reports showed that the close relationship among symbiotic genes (nodC and nifll) of Phaseolus symbionts (different genera) was associated with their host range, but independent of their classification based on the 16S rRNA gene (Laguerre et al. 2000). However, with the different genera of rhizobia isolated from Astragalus, Oxytropis and Onobrychis spp., there was a phylogenetic congruence at the genus level between symbiotic (nodC and nifH) and 16S rRNA genes, but no relation was found with the host nodulation range, indicating a specific evolution pattern for these rhizobia (Prévost et al. 2000). Since bacterial symbionts of more than 90% of legumes genera have not been studied, it is evident that there are still exciting discoveries ahead.
While there have been major developments in our understanding of the evolution, we also appreciate that the methods used for phylogenetic analysis are in constant progress. The interpretation of the divergence estimated in the sequences of the 16S rRNA genes depends on the overall comprehension of bacterial evolution. In this session, P. van Berkum presented a critical examination of the use of 16S rRNA gene sequence for reconstruction of evolutionary histories. A major concern stems from the observation that there is genetic evidence that the evolution of 16S rRNA genes could be reticulate instead of hierarchical. The topologies of trees constructed from 16S rRNA genes differ depending on the taxa selected in the analysis, also the trees obtained with 16S rRNA gene, 23S rRNA gene and ITS region sequences are statistically not congruent. This calls for caution in the conclusions that can be drawn from the use of 16S rRNA gene alone, and constitutes a serious argument in favor of the standardization of the approaches. This could be important especially when the analysis is used for decisions in taxonomy. One example is the distinction of Sinorhizobium from Rhizobium, based from the analysis of 16S rRNA gene, which was not supported by the analysis of the 23 S rRNA gene (Martinez-Romero et al. 2000).
3. The Acquisition of Genes for Adaptation and the Identification of Stress-induced Genes
The recent discovery of symbiosis islands (chromosomal elements) in Mesorhizobium strains from Lotus and the demonstration of their lateral transfer to non-symbiotic mesorhizobia present in soils is strong evidence for the evolution of new symbiotic strains in the field. C. Ronson and his group pursued their investigation with the aim to learn more about the contribution of these genomic islands to the evolution and niche adaptation. In this session, C. Ronson reported on the presence of genes downstream of the symbiosis island insertion sites (phe-tRNA locus). By grouping a population of 35 strains of mesorhizobia according to sequence similarity downstream of the symbiosis island, there was evidence for 15 different acquired DNA regions. The genomic structure determined in a few strains showed that some of these regions encode traits such as iron acquisition and adhesins involved in seed colonization. These acquisitions may be advantageous for ecological adaptation, and competition and provide additional evidence of the adaptive value of horizontal gene transfer.
The identification of genes that are expressed under specific environmental conditions has considerably increased with the use of molecular tools, such as the marker-reporter genes. F. de Bruijn and his team already got important data on genes induced in S. meliloti. They used a Tn5-luxAB reporter system to monitor gene expression in response to N, C, 02 limitations and dessication. Many tagged genes were identified and studied for their regulation and their role in persistence and competition. In this session, F. de Bruijn demonstrated the complementarity and the power of using comparative and functional genomics (micro-arrays) to study stress-induced rhizobial loci. Another promising approach, the IVET (in vitro expression technology) is in progress to identify genes induced in the rhizosphere (Izallalen et al. 2000). These long-term studies will help to elucidate the processes involved in competition and persistence of introduced strains in soils and in the rhizosphere of legumes.
The progress in our understanding of the evolution and ecology will certainly be applicable to the development of new microbial technologies. For instance, the knowledge on lateral transfer of symbiotic genes and the identification of the genetic background for plant infection would allow the development of more efficient symbioses. Moreover, the understanding of the processes involved under environmental conditions for adaptation, persistence and competition will be useful in inoculant technology.
Izallalen et al. (2000) In Program and Abstracts, 17th NACSNF, pp. 69, Québec, Canada Laguerreetal. (2001) Microbiol. 147, 981-993
Martinez-Romero et al. (2000) In Program and Abstracts, 17th NACSNF, pp. 33, Québec, Canada Prévost D et al. (2000) In Pedrosa F et al. (ed) Nitrogen Fixation: From Molecules to Crop Productivity, pp. 205, Kluwer Academic Publishers, Dordrecht, The Netherlands
RHIZOBIA: THE FAMILY IS EXPANDING
L. Moulin1, W.M. Chen2, G. Bena1, B. Dreyfus1, C. Boivin-Masson1
'LSTM, IRD-INRA-CIRAD-ENSAM, TA 10/J, Baillarguet, 34 398 Montpellier Cedex 5, France
2Tajen Institute of Technology, No. 20, Wei-Shin Road, Shin-Erh Village, Yen-Pu, Ping-Tung 90703, Taiwan
Whereas nitrogen fixation is widespread in bacteria and archae, nitrogen fixation in symbiosis with legumes is performed by a group of bacteria, known as rhizobia, that until recently belonged exclusively to the a-subclass of Proteobacteria (Young, Haukka 1996; Sy et al. 2001). The ability of rhizobia to nodulate legumes is determined by a set of genes, the nodulation genes, essential to trigger nodule induction (Perret et al. 2000). Nodulation genes are involved in the production of lipochito-oligosaccharides (Nod factors) that act as signaling molecules for nodulating specific legume hosts (Lerouge et al. 1990; Spaink et al. 1991). Rhizobia are distributed in four distinct branches of the a-Proteobacteria, each containing many bacterial species that are not rhizobia. It is now widely accepted that nod gene transfers have occurred among members of the four rhizobial clades and account for the polyphyletic origin of rhizobia (Suominene et al. 2001). Gene transfer leading to an efficient symbiosis was indeed demonstrated both in the field and in the laboratory (Sullivan et al. 1995, 1998). The clustering of rhizobia within a same bacterial phylum suggested that only a-Proteobacteria possessed the genetic background for legume symbiosis. Our recent results (Moulin et al. 2001) show that the ability to establish a symbiosis is more widespread in bacteria than anticipated to date since we found nodulating bacteria within the P-Proteobacteria.
We discuss how this finding may open the way for the discovery of new rhizobia and may contribute to a better understanding of the origin and evolution of rhizobium-legume symbioses.
2. Extension of Rhizobia from a- to P-Proteobacteria: Identification of Nodulating Burkholderia
The nodA gene is involved in the synthesis of the core Nod Factor by specifying the transfer of an acyl chain to the acceptor chitooligosaccharide. In the course of the phylogenetic analysis of the NodA protein from a collection of rhizobia (Moulin et al. this volume), we were intrigued by two sequences that did not group with other NodA sequences. One nodA was amplified from STM678, a strain isolated from Aspalathus carnosa in South Africa. The other nodA belonged to STM815, a strain isolated from Machaerium lunatum in French Guyana. By sequencing the 16S rDNA of these strains we found that they belonged to the Burkholderia genus within the P-subclass of Proteobacteria and are thus phylogenetically distant from known rhizobia (Moulin et al. 2001) (Figure 1). The phylogenetic position of STM678 was confirmed by 23S rDNA and dnaK partial sequencing. Recent investigations indicate that they correspond to two distinct species, close to B. kururiensis (P. Vandamme, unpublished results). The nodulation ability of strains STM678 and STM815 was confirmed by inoculation of Macroptilium atropurpureum, a broad host range legume, and by re-isolation and characterization of the bacteria isolated from the induced nodules. Nodules induced on M. atropurpureum were ineffective in terms of nitrogen fixation, probably because M. atropurpureum is not the original symbiotic partner. By screening among bacteria isolated from root nodules collected from various legumes in Senegal we found a third Burkholderia strain,
ORS1827 (Figure 1), isolated from Alysicarpus glumaceiis and fixing nitrogen in symbiosis with many tropical legumes.
3. Identification of a Second Rhizobial Genus, Ralstonia, within P-Proteobacteria
The taxonomic characterization of a collection of root isolates (about 180) from Mimosa pudica and M. diplotricha in Taiwan revealed that most of the strains (94% of the isolates) also belong to the p-Proteobacteria. 16S rDNA sequencing positioned all these strains in the Ralstonia genus, close to R. eutropha (Figure 1). These strains represent a novel Ralstonia species for which the name R. taiwanensis was proposed (Chen et al. in press). LMG19424 is the type strain. We checked the ability of 4 isolates to re-nodulate their host plants. All formed nitrogen-fixing nodules on Mimosa pudica. Characterization of the re-isolates confirmed the rhizobial status of these Ralstonia strains.
We propose to use the terms a-rhizobia and p-rhizobia to distinguish rhizobia belonging to a-Proteobacteria from rhizobia of P-Proteobacteria. This nomenclature will be used in the following text.
Was this article helpful?