Viruses and microorganisms associated with a wild community of Gambian Drosophilidae

Abstract

Drosophila melanogaster has its roots in the African continent and is estimated to have expanded from sub-Saharan Africa into Eurasia about 10,000 years. This was around the same time human commensalism of D. melanogaster is said to have occurred in the Matopos rainforest in Zimbabwe via the marula fruit, which was important in the diets of both the fruit flies and the hunter gatherers that occupied the region. Metagenomic sequencing of wild Drosophila has revealed over 130 naturally occurring viruses spanning over 25 families. These studies have also shown that some natural African Drosophila viruses have a deeper evolutionary history than their European and Northern American counterparts, suggesting the importance of studying African Drosophila viruses to understand insect-virus coevolution.

The main aim of my PhD is to use large scale metagenomic sequencing to understand host-virus interactions in Drosophila, both in laboratory culture and the wild. In Chapter 2, I investigated host-virus interactions using the transcriptome data of laboratory adult Drosophila that were serendipitously infected with viruses via a natural route. I found evidence that flies increase expression of some antiviral immune genes in response to viral infection. There was also marked sex differences in response to viral infection. In Chapter 3, I sought to discover the viruses and microorganisms associated with wild Drosophilidae I collected from The Gambia. I generated metagenomic libraries from individual Gambian flies and derived curated virus sequences from them. I discovered 27 new viruses associated with wild drosophilids including a DNA virus and both segmented and non-segmented RNA viruses, and their associated prevalences and host range. This work was extended in Chapter 4 where I found bacteria, fungi, microsporidia and trypanosomatids that are associated with Gambian Drosophilidae. I found that Gambian drosophilids were associated with many bacterial species that were isolated from West African fruits. Using a contig co-occurrence method combined with phylogenetic analysis, I was able to distinguish between the viruses that infect drosophilds from those that infect its microbial community. I discovered 27 new viruses that infect the trypanosomatids and fungi that are associated with drosophilids.

Overall, the work described in this thesis showed that viral infection induces detectable changes in the gene expression of laboratory flies which may form important background in laboratory experiments. Furthermore, it showed the importance of using individual flies in metagenomic virus discovery to accurately identify the host species and estimate virus prevalence. Finally, I was able to combine segment correlation and phylogenetic analysis to decipher the viruses infecting the drosophilds from its microbial community.