An ancient example of cooperation, the AM fungal association with plant roots has existed for at least 400 million years, as evidenced by early vascular plant microfossils from the Devonian age [90,91,98,99]. The partnership, or one resembling it, is thought to have facilitated the transition of plants from aquatic to terrestrial habitats [9,65,107]. In the mutualistic interaction, fungal mycelia grow among cells of the root cortex and form intracellular arbuscular interfaces to plant cortical cell membranes, into which phosphate and other nutrients are actively transported from extraradical mycelia in return for carbohydrates [3,63,123]. Because plants are autotrophic and lack of phosphate can limit their growth, the partners exchange valued resources [77,86,105]. The enduring coevolutionary history shared by these symbiotic partners has resulted in a major evolutionary transition: a symbiont's loss of independent growth, and possibly also the sexual phase of its life cycle, in the absence of its obligatory host [12,78,90].
Development of the AM symbiosis is a complex process regulated by the plant host at multiple checkpoints. Isolation of plant mutants that arrest development of the mycorrhizal symbiont at different stages of the mutualism indicates that multiple distinct, yet unidentified, genes regulate the endosymbiosis [12,54,55]. In M. truncatula and Pisum sativum, mycorrhizal incompatible mutants incapable of defense suppression are also unable to associate with nitrogen-fixing rhizobacteria [12]. The correlated loss of function suggests the existence of common genes essential to both mutualistic symbioses [12,55,59,121]. Neither members of the Chenopodiaceae family nor the Brassicaceae, including Arabidopsis thaliana, are able to support the AM association. Discovery of a mycorrhizal capable mutant A. thaliana phenotype is conceivable because the association is ancestral to angiosperms but has been lost in exceptional lineages [12].
Sequencing from tissue-specific cDNA has helped to identify candidate transcripts likely to be involved in endosymbiosis. Most of these have been obtained from root hairs [34] or root nodules [51], but some have been characterized from the AM symbiosis in either Lotus japonicus or M. truncatula. Several plant defense genes suppressed during the AM symbiosis have been isolated, as have a handful of non-defense related cDNAs that are elicited during the mycorrhizal association [53,56,73,119], and in some instances also during phosphate depletion [24,25,26]. Discovery of genes expressed in both arbuscular mycorrhizal and rhizobacterial associations hints that signal transduction pathways in early stages of both symbioses may be conserved [59,121], perhaps due to the Rhizobium association having co-opted mechanisms of the mycorrhizal symbiosis during its evolution [12,84]. Though the stages of development for the association have been characterized by cytological and genetic studies, many details of the molecular mechanisms of the AM symbiosis remain to be discovered and understood [12,54,55]. Driven by these advances, investigators continue to search for genes responsible for development and regulation of endosymbiosis.
The following chapters describe the development and validation of novel computational methods that increase the success of high-throughput screening for candidate symbiotic genes. Intended to elucidate mechanisms of interaction during development of the AM symbiosis, these methods complement ongoing high-throughput EST sequencing projects. Such methods may equally be applied to understand other symbiotic associations, such as the mutualism between nodule-forming, nitrogen-fixing bacterial endosymbionts and their leguminous plant hosts, and the antagonistic interactions between pathogens and their hosts.