Project description:Randomly amplified polymorphic DNA (RAPD) analysis was used to investigate the diversity of 179 bean isolates recovered from six field sites in the Arcos de Valdevez region of northwestern Portugal. The isolates were divided into 6 groups based on the fingerprint patterns that were obtained. Representatives for each group were selected for sequence analysis of 4 chromosomal DNA regions. Five of the groups were placed within Rhizobium lusitanum, and the other group was placed within R. tropici type IIA. Therefore, the collection of Portuguese bean isolates was shown to include the two species R. lusitanum and R. tropici. In plant tests, the strains P1-7, P1-1, P1-2, and P1-16 of R. lusitanum nodulated and formed nitrogen-fixing symbioses both with Phaseolus vulgaris and Leucaena leucocephala. A methyltransferase-encoding nodS gene identical with the R. tropici locus that confers wide host range was detected in the strain P1-7 as well as 24 others identified as R. lusitanum. A methyltransferase-encoding nodS gene also was detected in the remaining isolates of R. lusitanum, but in this case the locus was that identified with the narrow-host-range R. etli. Representatives of isolates with the nodS of R. etli formed effective nitrogen-fixing symbioses with P. vulgaris and did not nodulate L. leucocephala. From sequence data of nodS, the R. lusitanum genes for symbiosis were placed within those of either R. tropici or R. etli. These results would support the suggestion that R. lusitanum was the recipient of the genes for symbiosis with beans from both R. tropici and R. etli.
Project description:Coevolutionary change requires reciprocal selection between interacting species, i.e., that the partner genotypes that are favored in one species depend on the genetic composition of the interacting species. Coevolutionary genetic variation is manifested as genotype ´ genotype (G ´ G) interactions for fitness from interspecific interactions. Although quantitative genetic approaches have revealed abundant evidence for G ´ G interactions in symbioses, the molecular basis of this variation remains unclear. Here we study the molecular basis of G ´ G interactions in a model legume-rhizobium mutualism using gene expression microarrays. We find that, like quantitative traits such as fitness, variation in the symbiotic transcriptome may be partitioned into additive and interactive genetic components. Our results suggest that plant genetic variation is the largest influence on nodule gene expression, and that plant genotype and the plant genotype ´ rhizobium genotype interaction determine global shifts in rhizobium gene expression that in turn feedback to influence plant fitness benefits. Moreover, the transcriptomic variation we uncover implicates regulatory changes in both species as drivers of symbiotic gene expression variation. Our study is the first to partition genetic variation in a symbiotic transcriptome, and illuminates potential molecular routes of coevolutionary change. We assayed gene expression using three biological replicates for each plant genotype × rhizobium genotype combination (4 combinations) for a total of 12 chips.