The Candidatus Glomeribacter gigasporarum genome sequencing project

Symbiotic associations between endocellular bacteria and eukaryotic cells are widespread in nature, but only a few examples have been described in the fungal kingdom. Arbuscular mycorrhizal species, belonging to the Gigasporaceae family, represent a specialized niche for a homogeneous population of rod shaped bacteria, named Candidatus Glomeribacter gigasporarum. Previous molecular studies on the Gigaspora margarita BEG34 obligate endosymbionts, closely related to the Burkholderiaceae beta-proteobacterial family, have estimated a genome size of ~1.4 Mb. A metagenomic library was constructed from G. margarita BEG34 spores into the fosmid vector pCC1Fos. The library contains ~36K primary clones with an average insert size of 35 kb. Fungal and bacterial clones were identified by database sequence homology searches using fosmid ends as query. Sixty eight bacterial clones, showing high similarity with Burkholderiaceae genomic sequences, have been validated by PCR and sequenced by classic Sanger method, using a transposon insertion strategy. Assembly of Sanger data produced 1.34 Mb, which represents more than 90% of the predicted Ca. Glomeribacter gigasporarum genome size. To integrate the Sanger data and to increase the genome coverage, we adopted the 454 technology. A new experimental procedure was set up to allow the isolation of Ca. G. gigasporarum cells from its host and the recovery of enriched bacterial DNA. The 454 sequencing technology was applied to the whole bacterial amplified genome, obtained through Strand Displacement Amplification. The sequencing run yielded ~440,000 reads with less than 7% estimated host fungus contamination. A blast based approach was used to remove contaminating sequences, and both Sanger and 454 data were incorporated into a hybrid assembly. This first attempt to sequence the endocellular symbiont of an AM fungus is expected to give genomic insights into such complex association.