Specific flavonoids as interconnecting signals in the tripartite symbiosis formed by arbuscular mycorrhizal fungi, Bradyrhizobium japonicum (Kirchner) Jordan and soybean (Glycine max (L.) Merr.)

Many legume plants benefit from the tripartite symbiosis of arbuscular mycorrhizal fungi (AMF) and rhizobia. Beneficial effects for the plant have been assumed to rely on increased P supply through the mycorrhizas. Recently, we demonstrated that P does not regulate the establishment of the tripartite symbiosis. Flavonoids appear to play a role as early signals for both rhizobia and AMF. Four soybean lines known to express different concentrations of the isoflavones genistein, daidzein and glycitein in the seed were used to test three hypotheses: (i) The establishment of the tripartite symbiosis is not dependent of a nutrient mediated effect; (ii) There is a positive correlation between seed isoflavone concentrations of different soybean lines and the progress of the tripartite symbiosis; (iii) Specific flavonoids control the establishment of the tripartite symbiosis in that a change in flavonoid root accumulation resulting from the development of one microbial partner can stimulate colonization of soybean roots by the other. Disturbed versus undisturbed soil treatments were produced to vary the potential for indigenous AMF colonization of soybean. In contrast, the potential for Bradyrhizobium was kept identical in both soil disturbance treatments. The uptake of P and Zn and the concentration of flavonoids in mycorrhizal soybean roots at 10 d after emergence were analysed either separately of Bradyrhizobium or in context of the tripartite symbiosis. Zinc nutrition did not differ between AMF treatments which supports the first hypothesis. The concentration of daidzein was at least four times greater in the root than in the seed reaching 3958±249 μg g−1 dry across soybean lines. Coumestrol, which was absent in the seed, was synthesized to reach 2154±64 μg g−1 dry. Conversely, the concentration of genistein was approximately three times smaller in the root that in the seed (301±15 μg g−1 dry), while glycitein and formononetin were never detected. The establishment of the tripartite symbiosis was identical across soybean lines which does not support the second hypothesis. Concentrations of flavonoids were significantly greater in roots under disturbed soil, for which both symbioses were not as developed as in plants from undisturbed soil. This clearly supports the third hypothesis. This research provides the first data linking the function of different flavonoids to the establishment of the tripartite symbiosis, and suggests that these compounds are produced and released into the rhizosphere as a function of the colonization process.