Interactions between the gut symbiont Frischella perrara and its host the honey bee (Apis mellifera)

Résumé 0

Not all bacterial gut symbionts are necessarily beneficial to the host. Some of them may be neutral while others can even have detrimental effects. Determining the impact of individual gut symbionts can be challenging because the borders between being beneficial and detrimental are often fuzzy, and gut bacteria typically live in complex and highly variable multispecies communities. The honey bee possesses a relatively simple gut microbiota, providing a trackable model to study the effects of individual species. Among the few members of the honey bee gut microbiota, Frischella perrara is a gammaproteobacterium that colonizes a specific gut region where it causes the so-called “scab” phenotype, a dark colored band that appears on the luminal side of the epithelial surface. The scab has been hypothesized to result from melanization, a common insect immune response typically elicited after wounding or pathogen exposure. Despite inducing this putative immune response, there is currently no evidence that F. perrara is pathogenic for bees. In fact, F. perrara is highly prevalent among adult worker bees in healthy colonies across the world. This raises a number of interesting questions about the symbiosis between F. perrara and the host. Is the scab really a melanization response? Does F. perrara impact bee health? What genes from F. perrara are responsible for gut colonization and scab formation? Are there seasonal patterns of F. perrara prevalence along the year or interactions with other microbiota members or pathogens? The present thesis tackles these questions while investigating the symbiosis between F. perrara and the honey bee from three perspectives: the host side (chapter 1), the symbiont side (chapter 2) and in the context of the hive along seasons (chapter 3). In order to understand how F. perrara affects the gut homeostasis and immune status of the host, I used RNA-Seq to determine changes in host gene expression in the gut in response to experimental colonization with F. perrara. This showed that colonization with F. perrara led to the specific upregulation of many genes involved in the host immune response. In particular, multiple genes of the melanization cascade were upregulated by F. perrara, supporting the idea that the scab corresponds to a host melanization response. Despite this strong immune response, experimental colonization with F. perrara did not reduce the lifespan of bees relative to non-colonized bees or bees colonized with another symbiont not causing the scab. To identify F. perrara genes involved in colonization, persistence or scab formation, I investigated gene expression changes with RNA-Seq in F. perrara during host colonization relative to growth on agar plates, in collaboration with another PhD student. We found a number of interesting differentially expressed genes, with many genes upregulated in vivo involved in tryptophan biosynthesis, carbohydrate or ion transport, and some genes involved in tolerance to oxidative stress. Downregulated genes included genes coding for cell motility and sulfur metabolism. Finally, to identify specific conditions in the bee gut that impact colonization by F. perrara, we monitored the microbiota of individual bees from a hive through time. While we did not find significant correlations between F. perrara and other gut microbiota members or pathogens, we found that winter bees had a distinct microbiota structure than foragers that may be dictated at least in part by diet. In particular, F. perrara was the only species to be at significantly lower levels in winter bees relative to foragers. Overall, we can conclude from this PhD thesis that the scab phenotype is very likely the result of a melanization response upon F. perrara colonization. The absence of any detectable detrimental effect of F. perrara on the host is in line with its wide distribution across space and time. However, other pathogens are also highly prevalent in thriving honey bee colonies. Hence it is possible that the negative effect of F. perrara is small enough so that this gut symbiont is tolerated in the bee gut. The immune response mounted by the host may play an important role for the tolerance of the host. Rather than eliminating F. perrara, the specific immune response may keep the bacterium in check. However, further experiments need to be performed to test this hypothesis. On the contrary, we cannot exclude either that F. perrara has a beneficial role for the host. In particular, host immune activation by F. perrara may protect against subsequent pathogen assaults and the biosynthesis of the essential amino acid tryptophan or other chemical compounds by F. perrara may be used by the host. In summary, F. perrara is a clear example of a gut symbiont that cannot be easily classified according to the three classical categories encompassing mutualists, pathogens and commensals. This highlights the need to think about symbiosis as a continuum between pathogenicity and mutualism, and to find precise measures to quantify the costs and benefits for the involved partners.