Microbial ecosystems represent a gold mine for the exploration of new functions of biotechnological interest. However, most of the species that make up these complex communities are not cultivated and constitute a black box whose functioning is difficult, if not impossible to study. To address these challenges, the TBI and TWB, in collaboration with the INSERM I2MC unit, have integrated microbiomics and microfluidic droplet technology to develop a new ultra-high-throughput screening strategy for culturomics, functional metagenomics, and the engineering of enzymes, strains and microbial consortia. Two breakthrough technologies, which do not rely on fluorogenic substrates or fluorescent libraries, have been developed and patented by INRAE. They are currently used in the framework of the ANR CAZIBD project to decipher the host-microbiota-food dialogue, and the European BLUETOOLS project dedicated to the study and biotechnological exploitation of marine ecosystems.
Microfluidics allows coupling ultra-high throughput isolation and culture of microorganisms in drops with phenotypic, taxonomic and genomic characterization. However, existing technologies require either labeled recombinant cell banks, cell survival markers, or fluorogenic or chromogenic probes or substrates. They are therefore not adapted to the culture and screening of native microorganisms, nor to the use of non-chemically modified substrates naturally present in microbial ecosystems. These barriers limit the understanding of the functioning of ecosystems to their biotechnological exploitation.
Two revolutionary technologies, based on different principles, have been developed. The first integrates drop microfluidics, imaging, artificial intelligence and flow cytometry for automated detection and sorting of positive drops, at a rate of about 105 drops per hour. The second technology, dedicated to culturomics and strain and consortia engineering, directly couples drop microfluidics and flow cytometry for the selection of growing microorganisms, at a rate of 106 drops per hour. With a picoliter scale, a million times smaller than the sample volumes used with conventional phenotypic screening techniques, these technologies are compatible with any type of substrate at any cost, and the screening of any phenotype, as long as it can be detected by confocal microscopy or linked to microbial cell growth.
The technologies developed can be exploited for functional metagenomics, culturomics, enzyme engineering even using cell-free systems, engineering of strains or consortia, e.g. for the production of antimicrobials or for the degradation of polluting synthetic polymers such as plastics.
Légende et copyright image : Apprentissage pour la détection automatisée et le tri de gouttes pour le criblage à ultra-haut débit de fonctions microbiennes (Sophie Lajus et Remy Flores-Flores)