A short abstract of the doctoral research

High-throughput functional information is essential to understand microbial communities. Multi-omics techniques are routinely used for this purpose, but they measure the combined signal of many cells at once, making inference of functionality for single-cells more difficult. In contrast, traditional microbiological cultivation-based methods can reveal direct functional information, but they are low in throughput. Increasing resolution and throughput of functionality screening is therefore crucial for better understanding how microbial communities function.

This dissertation investigated the use of droplet microfluidics to address this technological gap. This technique can generate millions of tiny droplets, each acting as an individual microreactor, allowing an entire microbiome to be screened in parallel. We examined its potential both for (i) bacterial isolation, to answer “which species can or cannot grow in droplets?”, for (ii) functional screening, to answer “who does which function?”, and for (iii) interaction screening, to answer “who does what with whom?” in microbial communities.
We found that droplet microfluidics offers evident advantages in throughput and scalability, but that it also introduces specific biases affecting which taxa can be isolated. When throughput advantages were controlled for, conventional cultivation techniques performed comparably to droplet cultivation, with each method showing a unique set of taxa isolated. In functional assays, droplet microfluidics enabled rapid screening of bacterial function, revealing patterns not predicted by taxonomic functional inference. Moreover, we obtained preliminary information on microbial interactions within the studied communities, although further research is needed to refine these data.

In summary, this dissertation demonstrates that droplet microfluidics offers a promising route toward high-throughput, function-based microbiome research, bridging the gap between bulk -omics and actual microbe behavior.