Genome variation in the symbiotic nitrogen-fixing bacterium Sinorhizobium meliloti

Differences in genome size and gene content are among the most important signatures of microbial adaptation and genome evolution. Here, we investigated the patterns of genome variation among 10 natural strains of the symbiotic nitrogen-fixing bacterium Sinorhizobium meliloti, using pulse field gel electrophoresis (PFGE) and whole-genome microarray hybridizations. Our PFGE analysis showed a genome size range of 6.45-7.01 Mbp, with the greatest variation arising from the pSymA replicon, followed by pSymB; no size difference was evident among the chromosomes. Consistent with this pattern of size differences, 41.2% of open reading frames (ORFs) on pSymA were variably absent/present, followed by 12.7% on pSymB and 3.7% on the chromosome. However, the ORFs that were variably duplicated were more evenly distributed among the three replicons: 11.0%, 16.5%, and 15.3% of ORFs on pSymA, pSymB, and the chromosome, respectively. Among the 10 strains, the percentage of ORFs that were absent ranged from 1.51% to 6.35%, and the percentage of ORFs that were duplicated ranged from 0.27% to 8.56%. Our analyses showed that host plants, geographic origins, multilocus enzyme electrophoretic types, and replicon sizes had little influence on the distribution patterns of absent or duplicated ORFs. The proportions of ORFs that were either variably absent/present or variably duplicated differed greatly among the functional categories, for each of the three replicons as well as for the whole genome. Interestingly, we observed positive correlations among the three replicons in the number of absent ORFs as well as the number of duplicated ORFs, consistent with coordinated gene gains and losses in this important bacterium in nature.