[Thesis Defence] Evolution of thermophily in the Alvinellidae (Annelida : Polychaeta)

Sarah SAMADI, MNHN Professor, Chair of the jury

Céline BROCHIER-ARMANET, Professor, Université Claude Bernard Lyon, Rapportrice

Bruno FRANZETTI, CNRS Research Director, IBS Grenoble, Rapporteur

Jean MARY, Senior Lecturer Sorbonne University, Thesis Director

Didier Jollivet, CNRS Research Director, SBR, Invité

Elodie Laine, Professor, Sorbonne University, Invitée

Abstract : The Alvinellidae (Annelida : Terebelliformiaterebellid) are a species family endemic to deep hydrothermal vents from the Pacific and Indian Oceans. Since the discovery of the emblematic species Alvinella pompejana, the Pompeii worm, these animals have aroused the interest of the scientific community. Although hydrothermal vents are extreme environments (strong temperature gradients, absence of photosynthesis, anoxy, presence of various metals and sulphides due to the percolation of hydrothermal fluid into the basaltic crust, acid pH), the Alvinellidae have managed to colonise a variety of ecological niches and show great morphological, physiological and genetic diversity, both between and within species. In this thesis, we were notably interested in the adaptations that enable these worms to cope with contrasting thermal regimes. A. pompejana, for example, is thermophilic, surviving at temperatures close to 50°C. Other species, however, such as Paralvinella grasslei, are psychrophilic, living further from hydrothermal chimneys at temperatures between 10 and 25°C. More specifically, we studied the acquisition of thermophilia/psychrophilia during the evolution of the lineage, in an attempt to characterize the thermal phenotype of the ancestor of the Alvinellidae. To this end, we have established the molecular phylogeny of the Alvinellidae, based on molecular transcriptomic data recovered for eleven of the fourteen species in the family during several scientific campaigns. This initial result points to an ancestor dating from the end of the Cretaceous (between 60 and 90 million years ago), already present in the hydrothermal vents of the eastern Pacific. The radiation of the Alvinellidae was a quick event, within a few million years, resulting in several species with high rates of incomplete lineage sorting and showing traces of high interspecific introgression. The results of this phylogeny enabled us to establish a model to construct statistical proposals of proteins belonging to the ancestors of the lineage. Three proteins were chosen, namley the cytosolic malate dehydrogenase, the Cu/Zn superoxide dismutase and an intracellular hemoglobin, for reconstruction, expression and experimental characterisation. For ectothermic organisms such as the Alvinellidae, proteins from thermophilic species are expected to be on average more stable at high temperatures compared to their counterparts from psychrophilic species. These ancestral reconstructions allowed us to conclude that the ancestor of the lineage was a worm that was already adapted to warm environments, and that psychrophily of modern-day alvinellid species is a derived character acquired more recently. Finally, I looked at the optimisation of models for reconstructing ancestral protein sequences. These models are based on the diversity of contemporary sequences and their phylogenetic relationships. I tried to implement these approaches using two types of additional information : those linked to sequence insertion/deletion events, and those regarding the evolution of secondary structures of proteins and temporal variability of the expected frequencies of residues at different protein positions. I show that the introduction of these last two types of parameters into ASR methods is beneficial and leads to models with better likelihoods. However, the optimisation of these models, which are necessarily probabilistic, does not guarantee a better result for the experimenter, and the limits of these models to estimate the uncertainty of the inferred ancestral sequences are discussed.