Genomics and metabolomics in the North Atlantic deep-sea sponge Geodia barretti

Abstract: Sponges are among the earliest diverging taxa in the animal tree of life. They are sessile, filter-feeding animals found in marine and freshwater habitats. Many species are characterized by a close, specific and consistent association with microbes, mainly Bacteria and Archaea. This feature has been known for a long time and is suggested to be a factor contributing to the rich and diverse chemical output of the sponges. This thesis explored the effect of the habitat, specifically water mass or depth on sponges, their associated microbes, and their combined chemical output. The focal species of this thesis was the North Atlantic deep-sea high microbial abundance (HMA) demosponge Geodia barretti.In Paper I, 16S rRNA gene amplicon sequencing and untargeted metabolomics were used to quantify variation in prokaryotic community composition and chemical output in three sponge species. Water masses structured the prokaryotic community composition in the HMA species G. barretti and Stryphnus fortis. The community composition of the low microbial abundance (LMA) sponge Weberella bursa was unaffected by depth. Untargeted metabolomic data was modelled by depth. This allowed for identification of individual compounds varying with depth. Among those compounds were many putative osmolytes as well as diketopiperazines. Bioactive peptides and brominated tryptophan derivatives were unaffected by depth.In Paper II the diversity of the barrettide peptide family was explored in DNA sequencing data and chemical profiles across a wide selection of sponge species and G. barretti in particular. Five new barrettides were predicted and one sequence, barrettide C, was confirmed by solid phase peptide synthesis and co-elution with a native extract, antifouling bioassays and NMR structure elucidation. The confidence gained from sequence analysis and validating predictions lead us to suggest barrettides are a family of antifouling peptides in G. barretti.In Paper III, a reduced representation sequencing approach was used to evaluate the Stacks de novo pipeline in HMA sponges with the help of a whole genome assembled for this purpose. With this data, gene flow and connectivity were investigated in G. barretti populations sampled across the North Atlantic. The de novo pipeline was found to assemble and retain many putatively microbial loci and should thus only be used with reservations in HMA sponges. However, regarding biological inferences, strong population structure was recovered despite the apparent contamination.