To explore the role of
rhizobitoxine in Bradyrhizobium-legume symbiosis, 11
rhizobitoxine mutants of B. japonicum USDA61 were isolated on the basis of their inability to synthesize the toxin in culture. Each mutant is prototrophic and symbiotically effective on soybean, cowpea, siratro, and
Glycine soja. The
rhizobitoxine mutants differ in their
chlorosis phenotypes and
rhizobitoxine production in planta. As expected, one group of mutant fail to make toxin in planta, resulting in the absence of
chlorosis. Another group of mutants causes severe
chlorosis on all cultivars of soybean tested. Surprisingly, this group of mutants makes more
rhizobitoxine in soybean nodules than the wild-type strain does. This phenotype is only observed on soybean and not on other hosts such as cowpea, siratro, or G. soja. The remaining mutants all produce
rhizobitoxine in planta but vary in the amount of toxin they produce and the severity of
chlorosis they induce in soybean plants. Biochemical analysis of mutants demonstrates that one mutant is unable to synthesize
serinol, a molecule hypothesized to be an intermediate in
rhizobitoxine biosynthesis. By using these mutants, it was found that
rhizobitoxine plays no apparent role in the nodulation of rj1 soybeans. Recently, it was found that inhibition of
ethylene biosynthesis allows Rhizobium meliloti to overcome
nitrate inhibition of nodule formation on alfalfa. Because
rhizobitoxine also inhibits
ethylene biosynthesis, we tested the ability of mutants which accumulate high levels of toxin in planta to overcome
nitrate inhibition of nodule formation on soybean plants and found that the nodule formation induced by the wild type and that induced by mutant strains were equally suppressed in the presence of
nitrate.