Morphological Profiling of Imaging Flow CytometryData Uncovers Heterogeneity in Infected Gephyrocapsahuxleyi Cultures

Abstract

Phytoplankton, such as the coccolitophore Gephyrocapsa huxleyi (G. huxleyi), has a major ecological impact through photosynthesis—the production of oxygen and organic material. A significant threat to G. huxleyi populations is viral infection with the specific Gephyrocapsa huxleyi virus (GhV). Previous research has provided important insight into the infection cycle of G. huxleyi. However, research including quantitative morphological information on infected cells is lacking, potentially masking heterogeneity in the infection cycle. In this study, we propose a machine learning (ML) pipeline to incorporate morphological profiling into the analysis of spatially resolved single-molecule mRNA fluorescence in situ hybridization (smFISH)—imaging flow cytometry (IFC) data acquired on infected G. huxleyi populations. First, we propose to simplify infection monitoring by using a classification model that does not rely on mRNA staining. Second, we propose an exploratory data analysis pipeline to disentangle two modes of cell death in infected cultures and a subpopulation of healthy cells that potentially will not die from infection, but from programmed cell death (PCD). Overall, we show that morphological profiling of smFISH–IFC data is highly suited for studying microbial interactions in phytoplankton populations.