Root nodule symbioses may have evolved from a pre-existing pathway for the formation of arbuscular mycorrhiza
There are parallels between the establishing of arbuscular mycorrhiza and of root nodule symbiosis. In both cases, receptor-like kinases (RLK,) appear to be involved, linked to signal cascades, which induce the synthesis of the proteins required for the controlled infection. These sig-nal cascades probably involve G-proteins, MAP-kinases, and Ca++ ions as messengers . For several legume species, mutants are known that have lost the ability to establish both root nodule symbiosis and arbus-cular mycorrhiza. One of the genes that causes such a defect in different legume species has been identified as encoding an RLK, indicating that this RLK has an essential function in the formation of both arbuscular mycor-rhiza and root nodule symbiosis. Fungi and bacteria, despite their different natures, apparently induce similar genetic programs upon infection.
Molecular phylogenetic studies have shown that all plants with the abil-ity to enter root nodule symbiosis, rhizobial or actinorhizal, belong to a single clade ( =branch of phylogenetic tree, named Eurosid I). This implies that these species go back to a common ancestor, although not all descend-ants of this ancestor are symbiotic. Obviously, this ancestor has acquired a property on the basis of which a bacterial symbiosis could develop. Based on this property, root nodule symbiosis evolved about 50 million years ago, not as a single evolutionary event, but reoccurred about eight times. In order to transfer the ability to enter a root nodule symbiosis to agricul-turally important monocots, such as rice, maize, and wheat by genetic engi-neering, it will be necessary to find out which properties of the Eurosid I clade plants allowed the evolution of such symbiosis.