Deep Se(a)quencing : a study of deep sea ectosymbioses using next generation sequencing

Deep-sea hydrothermal vent fields and cold seeps are oases for deep-sea life in an otherwise nutrient-poor environment. They release energy-rich inorganic compounds that sustain rich microbial and invertebrate communities on the basis of bacterial chemosynthesis. Many endemic invertebrate species have established symbiotic relationships with chemosynthetic bacteria and thrive in these habitats. Symbioses occur in many forms: in endosymbioses the bacteria are located within a host cell or tissue, whereas in ectosymbioses the bacteria colonize their hosta s body surfaces such as epithelia. Deep-sea research is challenging with isolated and remote study sites requiring extensive logistical operations for sample collection. However, the rise of next-generation sequencing has allowed the gathering of large datasets from small samples and thus allows in-depth exploration of symbiotic systems. This Ph.D. thesis was focused on the investigation of two deep-sea epibiotic systems using next-generation sequencing methods. The first project investigates Epsilonproteobacteria that occur on several deep sea mussels of the subfamily Bathymodiolinae. Previous work with 16S rRNA clone libraries had suggested that the mussels may host epsilonproteobacterial symbionts in addition to the well-known gammaproteobacterial endosymbionts. First I analyzed the localization of the epsilonproteobacterial sequences within the mussela s gill tissue using microscopy and investigated their diversity and phylogeny using 16S rRNA sequencing methods. I was able to show an epibiotic association of the Epsilonproteobacteria and determined that seven out of the twelve mussel species studied were associated with closely-related Epsilonproteobacteria. The phylogenetic reconstruction of the 16S rRNA sequences suggested that the epibionts belong to a new family of Epsilonproteobacteria. In the second part of this project, I aimed to determine the nature of the association and metabolic potential of the epibionts, using metagenome and metatranscriptome analysis of two different bathymodiolin species. Based on genomic data, I was able to reconstruct their inorganic carbon fixation pathway, which was unexpectedly predicted to occur through the Calvin Benson Bassham (CBB) cycle. To date every other chemoautotrophic Epsilonproteobacteria has been described to fix inorganic carbon using the reverse tricarboxylic acid (rTCA) cycle. These epibionts acquired the CBB cycle from two separate horizontal gene transfer (HGT) events and lost the rTCA cycle. The key gene of the CBB, coding for 1,5-ribulose bisphosphate carboxylase, may have been acquired from a relative of the bathymodiolin gammaproteobacterial endosymbionts, whereas all the other CBB genes originate from an unknown Betaproteobacteria. I then discussed the implication of such HGTs and hypothesized that the epibionts are commensal or mutualistic, because most pathogens are not autotrophic. The third part of this project was a comparative analysis of the genomic data of the two epsilonproteobacterial epibionts. My phylogenomic analysis using multigene phylogeny showed that these two epibionts were two different species. I described the genetic potential of these epibionts and presented their reconstructed metabolism. This shows small metabolic differences between both bacterial draft genomes, and an overall array of different metabolic and genetic tools available that gives them a metabolic versatility to adapt to the environment. My second project investigated the ectosymbiotic bacterial populations associated with the deep-sea shrimp Rimicaris hybisae. The R. hybisae shrimp was discovered in 2010 along with two hydrothermal vent fields, Von Damm and Piccard, located on the Mid-Cayman Spreading Center. These two hydrothermal vents had very different environmental conditions and offered a unique setting to study the influence of the environment on ectosymbiotic populations. Von Damm, an ultramafic vent with high concentration of methane and low concentration of hydrogen sulfide, is located at 2500 m depth. Piccard, a basaltic vent field with low concentration of methane but high concentration of hydrogen sulfide, is located at 5000 m depth and is the deepest hydrothermal vent field found so far. I compared the different symbiotic and free-living populations using amplicon libraries of a variable region of the 16S rRNA sequence. I showed that the ectosymbiotic populations associated with R. hybisae are significantly different between the two hydrothermal vent fields and are more similar to their respective free-living bacterial communities. I hypothesize that the R. hybisae shrimp are taking up ectosymbionts from their environment, because they are probably the best-adapted to local environmental conditions.