Microbial communities are found everywhere in nature: in the soil and water, on plants, or in the organs of animais. These communities play important raies in shaping their specific environments. The assembly and composition of microbial communities is driven by many ecological factors such as physiochemical conditions and interactions with other community members. Studying interactions between microbial community members and their environment is challenging in dynamic, difficult to access environments with diverse microbial species. Therefore, amenable, ecologically relevant models are important. The honey bee gut microbiota is an example of an easy to manipulate, simple, and conserved system which offers many advantages to disentangle microbial interactions. ln tnis thesis, 1 focused on understanding how metabolic interactions shape the coexistence between two honey bee bacterial symbionts, and their honey bee host.
ln Chapter 1, we investigated how the divergent species of the genus Gilliamella and the species Snodgrassella alvi (S. alvi) influence each other's colonization in microbiota-depleted (MD) bees fed a complex or simple diet. To our surprise, we found that S. alvi colonizes the gut by using host-derived metabolites. Using 13C tracers coupled with metabolomics and nano scale ion mass spectrometry (NanoSIMS}, we demonstrated that S. alvi used host-secreted 3-hydroxy-3- methylglutarate, citrate, and glycerate among other carboxylic acids. However, we also showed that, as expected, S. alvi utilized organic acids produ ed by Gilliamella in the gut. We further demonstrated that both the host- and Gilliamella- derived metabolites were utilized by S. alvi in vitro. Finally, using amplicon sequencing we found that Gilliamella apis was the most abundant of the Gilliamella species to colonize the gut. Our results show the first evidence for direct host-compound foraging in the honey bee gut microbiota, which can be supplemented by microbial cross-feeding. However, the importance of cross-feeding remains elusive as neither species enhanced the colonization of the other.
ln Chapter 11, we followed up on the metabolic interactions between both species with in vitro growth experiments. We used.co-culture assays in minimal medium to investigate whether the two bacteria affect each other's growth and whether they exchange the same metabolites as the ones detected in vivo. We found that G. apis positively affects the growth of 5. alvi when grown on carbon sources that cannot be utilized by 5. alvi (glucose and N-acetyl-glucosamine). ln both conditions G. apis produced pyruvate, lactate and succinate which were consumed by 5. alvi. Growth of G. apis on N-acetyl-glucosamine, a major host derived compound, led to an
increase of acetate in the medium and was partially depleted by 5. alvi.
To further understand how the two bacteria influence each other's activity in the gut, we also attempted a transcriptome analysis in vivo. We colonized MD bees with the two bacteria separately or together and extracted bacterial mRNA enriched RNA samples for sequencing. However, due to low recovery of bacterial reads, especially for G. apis, these experiments did not allow us to carry out the transcriptomic analysis.
These results show that in vitro time-course co-culture assays cou Id recapitulate in vivo phenotypes and to accurately follow the dynamics of metabolic interactions. We also found that both a better understanding of factors regulating Gilliamella colonization as well as further method development were necessary to study gene expression of relatively low abundance bacteria in the gut.
The genus Gilliamella includes three distinct species, each composed of multiple strains, all of which can coexist in the honey bee gut. ln Chapter Ill, we hypothesized that the genomic diversity of Gilliamella translates into different functional capacities, gut colonization patterns, and interactions with 5. alvi. We selected a set of ten divergent strains belonging to the three different Gilliamella species and
colonized MD bees with each of them alone or in the presence of a defined community composed of the predominant species of the bee gut microbiota. Our findings revealed substantial variation in colonization success and loads both at the species- and strain-level. A single strain of Gilliamella apicola (ESL0309) exceptionally colonized almost all bees at high loads in the ileum. Furthermore, the presence of the community generally reduced the bacterial loads of all strains of Gilliamella in the gut except for strain ESL0309. Surprisingly, we also found that most strains colonized the posterior hindgut (i.e., rectum) to higher levels than the ileum, where Gilliamella forms a biofilm with 5. alvi on the hast surface. To identify possible traits which could explain the different colonization patterns we investigated the carbohydrate degradation genes, carbon source utilization profiles, and biofilm formation ability of ail·ten strains in vitro. lnterestingly, wë found a significant correlation between bacterial loads in vivo and the ability to forma biofilm in vitro. These results show that different species and strains within the genus do not colonize equally well, ·suggesting that strain-level diversity is a key aspect of the tri­ partite interaction between S. alvi, Gilliamella and the hast.
ln summary, our results show that host-metabolite foraging supports colonization of symbionts. Taken together with previous work, these results suggest that host­ foraging might be more widespread in the gut than initially anticipated. Our results also indicate that multipartite interactions between gut bacteria and the host are key for gut community assembly. ln addition, we find that genetic variation of symbionts likely is determinant for community assembly and niche occupation. Thus, our findings confirm the importance of studying species- and strain-level divergence in natural bacterial communities.