Microorganisms are the most dominant forms of life in Earth's history. Interactions between prokaryotic and eukaryotic microorganisms play key roles in ecosystem functioning and can take many forms. Traditionally, such interactions have been studied either by isolating individual microeukaryotes and observing them and their symbionts under a microscope, or by identifying all microeukaryotes and prokaryotes in a community and analysing correlation patterns of their co-occurrence. The former methods fail to scale to the numbers and complexity typical of natural systems, while the latter fail to capture individual interactions and their functional context. This thesis aims to add single-cell transcriptomics as a high-throughput high-resolution tool to the microbial ecology toolbox to characterize natural microeukaryote function and their interactions with prokaryotes.

First, the thesis presents a study using amplicon sequencing for identifying microeukaryotes and prokaryotes. Analyses of their compositions across time and environmental conditions revealed that changes in the prokaryotic community had a causal impact on the eukaryotic community and the environment. However, neither functional activity nor individual- or species-level interactions could be elucidated from amplicon sequences. Thus, the thesis presents subsequent studies using single-cell transcriptomics to isolate individual microeukaryotes and sequence their transcriptomes to infer population-level taxonomic identity, sub-population-level functional profiles and metabolic states, and individual-level prokaryotic associations. Gene expression and associated prokaryote compositions enabled the identification of cells potentially close to dormancy or under bacterial colonisation.

Most environmental microeukaryotes, including many in this thesis, are thought to be uncultivable in laboratory conditions and are absent from reference genome or transcriptome databases. This thesis aggregates computational methods—de novo assembly of microeukaryote transcriptomes, transcript annotation using sequence and protein structure homology, differential expression and abundance analyses, and metabolic mapping—for such environmental microeukaryotes. By demonstrating single-cell transcriptomics' and these computational methods' efficacy among environmental mixotrophic microeukaryotes without cultures or references, I have set the stage to scale this reference-free culture-free method to the natural microeukaryotic diversity and study microeukaryote-prokaryote functions and interactions in the stressful, heterogeneous, and ephemeral conditions that they occur in.