Accessing the ecology of uncultured picoeukaryotes through a high-throughput automatic cell enumeration approach

Abstract

Small planktonic protists (i.e. picoeukaryotes, cells of 0.8-3 µm) have fundamental roles in the functioning of marine ecosystems, both as primary producers and as microbial grazers, and are likely the most abundant eukaryotes on Earth. Over the last fifteen years, technological progresses in molecular ecology and environmental sequencing have substantially boosted our understanding of these marine microbes, unveiling an unexpected diversity and notably the existence of new uncultured clades. These methodological improvements have notably offered the possibility to analyse simultaneously a substantial number of samples, allowing a more accurate description of their diversity in time and space. Faced with this growing amount of data available on microbial diversity, it is now challenging to design parallel molecular tools combined with microscopy to acccess their quantititive importance in environmental systems. However, this quantitative approach is relatively expensive, time-consuming and observant-dependant. Furthermore, their tiny size, their lack of specific morphological traits and the changing number of 18S rDNA copies between taxa among the picosize protistan community do not allow the use of classical microscopy and SSU-based molecular approaches (qPCR) to access their abundance in a large numbers of samples. Within this study we tested a newly developed automatic image acquisition and subsequently cell enumeration system of picoeukaryotic organisms. Two uncultured groups of picoeukaryotes differing by their cell size, MAST-4 (2 µm) and MAST-1C (5 µm), have been considered here to optimize this high-throughput quantitative approach for picoeukarytic cells. After targeting these groups by TSA-FISH using specific oligonucleotide probes, microscopic images were acquired fully automatically and cells enumerated using the program ACMEtool 2.0. In general, direct microscopic cell counts were in agreement with our automatic cell counting approach. The automated method can process a larger number of fields of view (FOVs) and consequently analyzes more cells, so it provides an estimate that is closer to true cell abundance. We were further able to test this automatic microscope and cell enumeration system on a spatial survey with a wide geographical coverage, about 100 stations from a circumglobal expedition, by depicting the spatial distribution of the two above-mentioned MAST groups