La faune sauvage a été depuis longtemps incriminée dans la survenue de zoonoses et joue le rôle de réservoir d'agents pathogènes (virus Nipah, Hendra, Ebola, Hantaan etc.) pour l'homme. Les îles tropicales et subtropicales du Sud-Ouest de l'Océan Indien (SOOI) constituent l'une des 34 régions reconnues comme « hotspot » de biodiversité au niveau mondial. Elles sont caractérisées par un très fort endémisme de la faune sauvage surtout sur l'Ile de Madagascar. Le caractère multi-insulaire de la région du SOOI, la diversité de ses biotopes et ses disparités biogéographiques et humaines offrent un champ d'investigation unique pour explorer « in natura » la dynamique évolutive des agents infectieux et les relations hôtes-virus. Nos travaux de recherche ont porté sur deux modèles de virus à ARN de polarité négative, les paramyxovirus et les lyssavirus. Le premier modèle viral nous a permis d'aborder les questions relatives à la dynamique de transmission virale au sein de communauté d'hôtes, plus particulièrement, les chauves-souris et les petits mammifères terrestres de Madagascar et d'identifier les facteurs agissant sur cette dynamique de transmission et de diversification virale, en particulier les facteurs bio-écologiques associés à leurs hôtes. Le second modèle viral, les lyssavirus, nous a permis de décrire sur l'ensemble des îles du SOOI échantillonés, la circulation virale dans ce système multi-insulaire diversifié, au sein des chauves-souris dont la plupart des espèces sont endémiques à cette région. Dans l'ensemble, nos investigations ont permis de mettre en évidence des échanges viraux (« host-switch ») importants entre chauves-souris, petits mammifères terrestres endémiques de Madagascar et les rongeurs introduits, le rôle de ces mammifères en tant que réservoir viral majeur et souligner le rôle disséminateur de Rattus rattus. Par ailleurs, nous avons pu identifier ce phénomène de « host-switch » comme étant le mécanisme macro-évolutif prépondérant et l'importance des facteurs biotiques et abiotiques à l'origine de la dynamique de transmission et de la diversification virale observée chez les paramyxovirus de chauves-souris de Madagascar. / Since many decades, the wild fauna has been incriminated as an important reservoir of many zoonotic pathogens (Nipah, Hendra, Ebola, Hantaan viruses etc.) at risk for humans. Tropical and subtropical islands of the South West Indian Ocean (SWIO) are part of the 34 areas of the world recognized as "hotspot" of biodiversity. They are characterized by a strong wildlife endemism especially on Madagascar. The multi-island structure of the SWIO region, the diversity of its biotopes and its biogeographical human disparities, offer a unique opportunity to investigate "in natura" the evolutionary dynamics of infectious agents and the host-virus relationships. Our research has focused on two models of negative RNA viruses, paramyxoviruses and lyssaviruses. The first virus model allowed us to address issues related to the dynamics of viral transmission within a host community, in particular, bats and small terrestrial mammals of Madagascar and to identify the driving factors, especially bio-ecological factors associated with their hosts, affecting the dynamic of transmission and of viral diversification. The second model allowed us to describe on the islands of the SWIO, the intense circulation of bats lyssaviruses in this multi-island system which bats are endemic to this region. Overall, our investigations highlighted (i) intense viral exchanges ("host-switch") between bats, endemic terrestrial small mammals and introduced rodents from Madagascar, (ii) the role of these mammals as major viral reservoir and (iii) the key role played by Rattus rattus as viral spreader. Furthermore, we identified both the phenomenon of "host-switch" as the major macro-evolutionary mechanism among bat paramyxoviruses from Madagascar and the importance of biotic and abiotic factors in shaping the transmission dynamics and viral diversification.

Insects are host to a diverse range of vertically transmitted micro-organisms, but while their bacterial symbionts are well-studied, little is known about their vertically transmitted viruses. The sigma virus (DMelSV) is currently the only natural hostspecific pathogen to be described in Drosophila melanogaster. In this thesis I have examined; the diversity and evolution of sigma viruses in Drosophila, their transmission and population dynamics, and their ability to host shift. I have described six new rhabdoviruses in five Drosophila species — D. affinis, D. obscura, D. tristis, D. immigrans and D. ananassae — and one in a member of the Muscidae, Muscina stabulans (Chapters two and four). These viruses have been tentatively named as DAffSV, DObsSV, DTriSV, DImmSV, DAnaSV and MStaSV respectively. I sequenced the complete genomes of DObsSV and DMelSV, the L gene from DAffSV and partial L gene sequences from the other viruses. Using this new sequence data I created a phylogeny of the rhabdoviruses (Chapter two). The sigma viruses form a distinct clade which is closely related to the Dimarhabdovirus supergroup, and the high levels of divergence between these viruses suggest that they may deserve to be recognised as a new genus. Furthermore, this analysis produced the most robustly supported phylogeny of the Rhabdoviridae to date, allowing me to reconstruct the major transitions that have occurred during the evolution of the family. This data suggests that the bias towards research into plants and vertebrates means that much of the diversity of rhabdoviruses has been missed, and rhabdoviruses may be common pathogens of insects. In Chapter three I examined whether the new sigma viruses in Drosophila affinis and Drosophila obscura are both vertically transmitted. As is the case for DMelSV, both males and females can transmit these viruses to their offspring. Males transmit lower viral titres through sperm than females transmit through eggs, and a lower proportion of their offspring become infected. I then examined natural populations of D. obscura in the UK; 39% of flies were infected and the viral population shows clear evidence of a recent expansion, with extremely low genetic diversity and a large excess of rare polymorphisms. Using sequence data I estimate that the virus has swept across the UK within the last ~11 years, during which time the viral population size doubled approximately every 9 months. Using simulations based on lab estimates of transmission rates, I show that the biparental mode of transmission allows the virus to invade and rapidly spread through populations, at rates consistent with those measured in the field. Therefore, as predicted by the simulations, the virus has undergone an extremely rapid and recent increase in population size. In Chapter four I investigated for the first time whether vertically transmitted viruses undergo host shifts or cospeciate with their hosts. Using a phylogenetic approach I show that sigma viruses have switched between hosts during their evolutionary history. These results suggest that sigma virus infections may be short-lived in a given host lineage, so that their long-term persistence relies on rare horizontal transmission events between hosts. In Chapter five I examined the ability of three Drosophila sigma viruses to persist and replicate in 51 hosts sampled across the Drosophilidae phylogeny. I used a phylogenetic mixed model to account for the non-independence of host taxa due to common ancestry, which additionally allows integration over the uncertainty in the host phylogeny. In two out of the three viruses there was a negative correlation between viral titre and genetic distance from the natural host. Additionally the host phylogeny explains an extremely high proportion of the variation (after considering genetic distance from the natural host) in the ability of these viruses to replicate in novel hosts (>0.8 for all viruses). There were strong phylogenetic correlations between all the viruses (>0.65 for all pairs), suggesting a given species’ level of resistance to one virus is strongly correlated with its resistance to other viruses. This suggests the host phylogeny, and genetic distance from the natural host, may be important in determining viruses ability to host switch. This work has aimed to address fundamental questions relating to host-parasite coevolution and pathogen emergence. The data presented suggests that sigma viruses are likely to be widespread vertically transmitted insect viruses, which have dynamic interactions with their hosts. These viruses appear to have switched between hosts during their evolutionary history and it is likely the host phylogeny is a determinant of such host shifts.