Experimental frameworks for deciphering bacterial adaptation in the plant root microbiota

Abstract

The plant root microbiome consists of phylogenetically diverse bacteria that play a critical role in host health and development. Recent research has demonstrated that these microbial communities exhibit a strong degree of host preference, preferentially colonizing plant species with which they share a co-evolutionary history. This indicates that bacterial fitness in the rhizosphere is dependent on the host species. To date, research on bacterial adaptation within the rhizosphere has relied heavily on metagenomic surveys investigating the correlation between the abundance of specific functional genes and their corresponding strains. While experimental evolution approaches have been employed to shed light on rhizosphere adaptation in a time-resolved manner, most studies utilized single-strain approaches, thereby omitting the complexities of inter-species dynamics in communities. Consequently, the specific genetic determinants of bacterial host preference, and the evolutionary trajectories of these genes under imposed selective pressure, remain elusive. This thesis aims to bridge that knowledge gap by tracing bacterial adaptation to the roots of a novel host plant within a community context. To address this question, we conducted a long-term evolution experiment utilizing bacterial communities composed of diverse phylogenetic clades. These synthetic communities (SynComs), consisting of 17 different strains, were successively propagated on the roots of their original and a novel host over a period of nearly two years. Analysis of these evolving communities through metagenomic shotgun sequencing revealed a remarkable level of parallel evolution. Cross-inoculation experiments demonstrated a significant increase in bacterial fitness over the course of the experiment. At the final time point, the fitness of the populations which evolved on the novel host plant was indistinguishable from the colonization fitness on the roots of the original host. Notably, bacteria evolving on the new host plant displayed a higher number of fixed mutations, suggesting a higher adaptive potential and stronger selective pressure under these conditions. During the evolution experiment, we established a comprehensive library of approximately 23,000 individual evolved isolates collected from different time points. Because the scale of this culture collection rendered sequencing-based strain verification impractical, we developed a high-throughput method for automated strain identification. To this end, we present an autofluorescence-based framework for the taxonomic classification of known bacterial isolates. By applying this method to our collection of 23,000 evolved isolates, we were able to assign taxonomic labels to every individual isolate. This automatic indexing enabled the design of SynComs composed of evolved bacteria, allowing for a direct comparison of their root colonization capacity against their ancestral counterparts. Through these competition assays, we identified significant increases in fitness across several key taxa, demonstrating successful adaptation to the new host. Whole-genome sequencing of these isolates identified mutations in metabolic genes, regulators of biofilm formation and motility, and genes affecting cell-surface proteins. These findings suggest that bacterial adaptation in the rhizosphere is a substrate-driven process where the ability to adhere to the root surface is a primary selective advantage along with modulation of immunogenic cell-surface proteins. Our results demonstrate that bacterial communities can be conditioned to achieve improved fitness on a novel host plant within reasonably relevant timescales. Consequently, the experimental framework presented in this thesis provides a robust foundation for the enhancement of bioinoculants. By utilizing this experimental evolution approach, researchers can optimize the rhizosphere competence of agriculturally relevant strains across diverse crop species, soil types, and fertilization regimes.

Abstract from the paper’s OpenAlex record, shown for indexing and research. Copyright remains with the authors or the publisher.

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.