Investigating the population dynamics of phyllosphere bacteria using a novel tagging system

Abstract

Bacteria are ubiquitous, colonizing a vast array of habitats, forming complex communities, called microbiomes. Some of these microbiomes are associated with living hosts, which can be any organism larger than the bacteria themselves, from single-celled algae over animals to plants. Such host-associated microbiomes are now widely recognized to affect host fitness in a range of ways that can be classified into three main categories of symbiosis: Mutualists that benefit their host, pathogens harming the host, and commensals, whose presence appear to have no influence on the host. Such commensals can nonetheless be important, for example, to prevent or limit the invasion of pathogenic species. Microbiota in any given host establish reproducibly, implying the existence of underlying assembly principles yet to be elucidated. To effectively manage microbiota assembly, it is imperative to gain a comprehensive understanding of the rules governing this process. The ability to alter the species composition could be employed to engineer desired microbiomes that lead, for example, to improved host health. In this thesis, the focus lies on the phyllosphere microbiome and the development of tools that can be used to unravel the dynamics of its community assembly across spatial and temporal dimensions. The phyllosphere comprises all parts of the plant above the soil, including fruits, flowers, stems, and leaves. Microbial life in the phyllosphere is exposed to rapid changes in environmental conditions such as solar radiation, humidity, and temperature. The phyllosphere is also an oligotrophic habitat, as the plant limits the availability of nutrients. The microbiota that occupies this habitat is adapted to these conditions, with bacteria, the most abundant organisms found in the phyllosphere, not just persisting but multiplying. Thale cress, more commonly known by its scientific name Arabidopsis thaliana (Arabidopsis), has emerged as one of the most used model plants to study bacterial life in the phyllosphere on account of its ease of handling, small size, and short generation time. In the research presented in this thesis, synthetic communities (SynCom) were used to explore characteristics of phyllosphere microbiota. To reconstitute SynComs under controlled conditions the At LSPHERE was used, which represents a collection of 224 bacteria that were isolated from environmental Arabidopsis plants. In this work, a novel set of genomic barcodes was established to enable the investigation of intra-species dynamics. These barcodes were termed WISH-tags for Wild-type standardized hybrid tags. The name reflects specific attributes that were considered in the design of the barcodes. Wild-type isogenic means that there is no difference between the tagged bacteria and the wild-type genetically, except for the integrated barcodes. They follow the same standard design and only vary by a 40 bp DNA stretch, uniquely identifying them. Finally, the barcodes can be quantified by qPCR or next generation sequencing (NGS), allowing hybrid approaches of both readouts. The performance of the WISH-tags was validated for several bacterial species from the phyllosphere and for strains that colonize the mouse gut. The tags were then used to investigate intra-strain priority effects in the phyllosphere and the mouse gut in a community context. We found that intra-species priority effects were pronounced in the mouse gut, effectively preventing the late arrivals from establishing a population, whereas in the phyllosphere, no disadvantage was observed for the late arriving isogens. While the presence of a microbial community reduced the population of the focal strains of interest in both hosts, it did not impact the ratio of the late and early arriving strains compared to the mono association. Investigating the spatial component of phyllosphere colonization, the dispersal of bacteria across Arabidopsis was examined, excluding external disturbances like wind, rain, or animals. To resolve the spatial component of this dispersal, an approach of spatially segmented harvest was developed. We investigated the distribution after growth on the plant for five different bacterial species. We found that one of them was dispersed less often to other parts of the plant than the other four species. This result suggests that species-specific properties like biofilm formation might influence the likelihood of dispersal events to occur. In summary, the tools introduced in this work enable the resolution of processes of microbiota assembly, furthering our mechanistic understanding. We applied the WISH-tags to show that intra-species effects can lead to different outcomes in distinct model systems. We also showed that the inherent dispersal of a bacterial species is influenced by its characteristics. The experiments described in this thesis demonstrate the value of barcoding for microbiome research in general and the study of the phyllosphere in particular.