Mutant analysis, gene sequencing, and expression assays have helped characterize plant genes involved in the symbiosis [12,24,25,26,53,73,119]. Gene expression assays have identified plant genes common to both the AM symbiosis and the rhizobial association between legumes and nitrogen-fixing mutualistic rhizobacteria [59,121], which indicates the two mutualisms share some mechanisms of defense suppression [12,55,92,84]. In contrast, the obligate nature of the mycobiont has made understanding this group of soil-borne mycelial fungi a challenge [12,55,56], though some fungal genes have been cloned and sequenced [56,92,104].
As a complementary approach to analyzing sequences one at a time, availability of large numbers of expressed sequence fragments enables a global, comparative view of many transcripts expressed in a library, and comparisons among several libraries derived from related tissues [2,21,39,45,58,60,66,67,82,83,93].
Recently developed culture techniques have made it possible to extract RNA from mycorrhizal tissues in axenic culture, in the presence of a plant host [104], but without host cell tissues present in the mycelial culture. Comparative analysis of sequences from extraradical mycelia with sequences expressed in other tissues, such as germinating fungal spores and mycorrhizal plant roots, helps to identify genes involved in Glomales growth and reproduction, and development of the AM symbiosis [12,55,56,92,104].
Comparing transcript libraries prepared at varied developmental stages, in pure and mixed culture, allows mixed samples to be used and deconvoluted. The spirit of this effort is similar to that of preparing a library with subtractive hybridization [18], but does not reduce library diversity in the process of isolating those transcripts specific to a particular tissue type.
This study applies several computational techniques for analyzing many symbiotic transcripts at once, described in Chapters 2 and 3. Related approaches to comparing transcript libraries with computational methods are described in [43] and [20] for human transcript libraries, though our methods differ in the implementation details. Further, this study considers attributes of mixed interaction cultures. The intent here is to answer specific questions about transcripts from a mutualistic system with computational means.
Many questions can be asked of a symbiotic interaction such as this. This chapter will focus on addressing the following questions: How many genes that have plant-like composition are present in a library? How many are fungal or bacterial? What is the total observed sequence diversity in a library? What is the estimated sequence diversity? How does the composition of one fungal library relate to another and to axenic and mixed-culture plant libraries? What do these insights reveal about molecular mechanisms of mutualism?