Microevolution And Dynamics Of Symbiotic Bradyrhizobium Populations Associated With Soybeans In North America

Soybean (Glycine max) is the most important grain legume on a production basis in the world and is a major source of oil and protein. The genus Bradyrhizobium includes species of economically important soil bacteria that fix atmospheric nitrogen in symbiotic association with soybeans and thereby minimize the requirement for nitrogen fertilizer inputs in crop production. Populations of Bradyrhizobium were widely studied for its valuable genetic resources and economic values of soybean. Canada represents the northern-most limit for soybean cultivation and ranks as the world’s seventh largest producer. Despite the importance of soybeans to Canada, no information is available on the genetic resource represented by populations of symbiotic Bradyrhizobium associated with this legume with regard to their diversity, dynamics, structure or micro-evolutionary history. The research on biodiversity of Bradyrhizobium will benefit exploitation of new Rhizobium resources and novel gene resources for the sustainable agricultural production, provide further insight into taxonomy and evolutionary relationship within Bradyrhizobium genus and assist in selection of bacteria with desirable traits as inoculants for soybean production.This study presented a systematic investigation of Bradyrhizobium populations that were isolated from two field sites (A and B, about280km apart in Eastern Canada) with contrasting histories of Bradyrhizobium inoculation and soybean cultivation. The results are listed below.1. A collection of symbiotic bacteria (220isolates) was made by plant trap. At the same time, we assembled a reference collection of Bradyrhizobium strains known to have been used in inoculants for soybeans in Canada. All isolates were slow growing on YEM agar (colony diameter＜1mm after7to14days at28℃) and elicited nodules on soybeans.2. Fragments of6housekeeping genes were obtained by PCR amplifications and sequenced. The length of individual gene sequences ranged from369bp (dnaK) to771bp (rpoB). The genetic diversity (π) of individual loci ranged from0.0251(atpD) to0.0338(recA) with the number of allelic types ranging from9(dnaK) to17(rpoB). The mean values of dN/dS were all＜1indicating that these loci were subject to stabilizing or purifying selection. The phi test on sequences of individual loci detected significant recombination at gyrB and recA loci, but not at the remaining loci.3. The concatenated sequences (six housekeeping genes) of220isolates were classified into35unique STs. Isolates representing totals of15and27STs were recovered from sites A and B whereas isolates of7STs were common to both sites. Isolates of ST21(53isolates) was most abundant and accounted for24.1%of all isolates. Isolates of ST15(36isolates) accounted for16.4%of the isolates.4. Among reference strains, USD A136was found to have the same multilocus genotype (ST4) as USDA122;532C and USDA138also have an identical multilocus genotype (ST24) to B. japonicum USDA6T. A minority of all isolates were identified as putative inoculant strains. At site A, only one isolate was a putative inoculant strain. At site B, about31%of the isolates were putative inoculant strains. Isolates corresponding to reference inoculant strain61A124were not recovered at either site.5. Multilocus sequence typing and microevolutionary analyses indicated that the Bradyrhizobium populations associated with soybeans were highly structured and genetically diverse. All analyses supported the division of220isolates into five lineages corresponding either to B. japonicum groups1and1a or to one of three novel lineages within the genus Bradyrhizobium.6. STRUCTURE analysis indicated that220isolates originated from five ancestral lineages. Evidence for spatial structuring was provided by lineage Ⅰ and Ⅲ isolates that were encountered only at one of two sites and had100%of their genetic material derived from the respective ancestral lineage.20isolates in lineages Ⅱ, Ⅳ and Ⅴ showed evidence of mixed ancestry indicating that inter-lineage gene flow had infrequently taken place. Of these STs identified as putative inoculant strains, only ST24(532C) exhibited mixed ancestry with genetic material inherited from ancestral lineages Ⅱ, Ⅲ, Ⅳ and Ⅴ. ST24was the only ST from site B to possess genetic material derived from ancestral lineage III.7. The network graph showed well defined clusters of isolates corresponding to the five lineages inferred by STRUCTURE. Clusters corresponding to lineages Ⅰ, Ⅱ and have no type strain indicating that they are novel evolutionary lineages within the genus Bradyrhizobium. Moderate reticulation was evident in Lineage Ⅳ and Ⅴ. Reticulation reflects extent of phylogenetic uncertainty and indicates recombination.8. ClonalFrame analysis indicated that recombination has significant effect on phylogenetic inference. A total of21recombination events were found and twelve events were found at the gyrB locus. Analyses were done to infer ancestors representing internal nodes and did not identify any putative inoculant strains as MRCAs of external nodes. The ratio of rates of recombination relative to mutation (ρ/θ) for all five lineages was0.05indicating that recombination was about twenty times less frequent than mutation. The effect of recombination relative to mutation (r/m) was0.59indicating that recombination introduced almost two times fewer substitutions than mutation. Lineage Ⅱ may have been subject to a recent expansion in population size based on external to internal branch length ratio tests.9. Unifrac analysis showed that there were overall differences in the frequency of lineages between sites A and B (P＜=0.01) and between cultivars M and O (P＜=0.01). It was also found that significant differences in frequencies of isolate representing lineages and individual STs between field sites and soybean cultivars. All data indicated spatial structuring of the populations.10. Sequences of the symbiotic nodC gene were divided into7allelic types with genetic diversity (π) of0.0167and GC content of57.3%; the mean value of dnlds was0.076suggesting that this gene is subject to stabilizing or purifying selection. The phi test did not detect significant recombination at the nodC locus. nodC sequences were placed in two well defined groups (nod Ⅰ and nod Ⅱ) in the consensus network. It was almost complete absence of reticulation, suggesting that recombination was rare at the nodC locus. Chi-square tests of independence indicate that the frequency of nod Ⅰ isolates differed between sites whereas the relative frequencies of nod Ⅱ isolates group were not significantly different. The frequency of isolates representing nod Ⅰ and nod Ⅱ groups did not differ significantly between soybean cultivars.11. Relative effectiveness test showed that there was considerable variation in the effectiveness of isolates tested. Several putative inoculant strains were poorly effective in nitrogen-fixation suggesting that these isolates may have undergone genetic exchange of nitrogen fixation genes or were soil residents that co-incidentally possessed the same multilocus genotype (ST) as an inoculant.