Bradyrhizobia are common members of soil microbiomes and known as N
2-fixing symbionts of economically important legumes. Many are also denitrifiers, which can act as sinks or sources for N
2O. Inoculation with compatible rhizobia is often needed for optimal N
2-fixation, but the choice of inoculant may have consequences for N
2O emission. Here, we determined the phylogeny and denitrification capacity of
Bradyrhizobium strains, most of them isolated from peanut-nodules. Analyses of genomes and denitrification end-points showed that all were denitrifiers, but only ~1/3 could reduce N
2O. The N
2O-reducing isolates had strong preference for N
2O- over NO
3−-reduction. Such preference was also observed in a study of other bradyrhizobia and tentatively ascribed to competition between the electron pathways to Nap (periplasmic NO
3− reductase) and Nos (N
2O reductase). Another possible explanation is lower abundance of Nap than Nos. Here, proteomics revealed that Nap was instead more abundant than Nos, supporting the hypothesis that the electron pathway to Nos outcompetes that to Nap. In contrast,
Paracoccus denitrificans, which has membrane-bond NO
3− reductase (Nar), reduced N
2O and NO
3− simultaneously. We propose that the control at the metabolic level, favouring N
2O reduction over NO
3− reduction, applies also to other denitrifiers carrying Nos and Nap but lacking Nar.
相似文献