MSc defence by David Essenbæk

Symbioses between animals and microbes can tremendously affect the evolutionary trajectory of both. However, particularly in the context of complex and promiscuous vertebrate-microbiome holobionts, the evolutionary processes remain poorly understood. Recently, several phylogenetically unrelated bacterial lineages which convergently exhibit extensively reduced metabolic capacities have been observed in vertebrate microbiomes. In this thesis, lineages of such genome reduced animal-associated bacteria (GRABs) were systematically identified across a wide range of vertebrates. Four large clades of GRABs, considered to represent independent genome reduction events, exhibited similar patterns in terms of loss and retention of functional capacities compared to phylogenetically related non-GRAB lineages. While GRABs exhibited extensive depletion of common metabolic capacities including nucleic- and amino-acid biosynthesis, all four clades exhibited relatively conserved capacity for short chain fatty acid biosynthesis. Remarkably, GRABs also exhibited similar patterns in fullness of the underlying pathways of several traits. Additionally, capacities for structural traits were relatively well conserved between individual GRAB lineages and adjacent non-GRAB lineages. Despite having a more restricted pool of metabolic capacities, GRAB lineages were not associated with a narrower host range and closely related GRABs were found to be associated with hosts across several different vertebrate classes including amphibians, birds, reptiles and mammals. However, structural traits including endospore formation and flagella were significantly associated with the host range of individual GRAB lineages. These findings highlight extensive genome reduction as a widespread phenomenon in vertebrate hologenomic evolution and imply that common principles governing loss and retention of functions and host range can be identified. Finally, this thesis also highlighted potential pitfalls in certain methodological approaches which might be specifically prone to bias studies of genome reduced animal-associated bacteria.