Automated High-Throughput Quantification Reveals Morphological Heterogeneity in Mammalian Cell Populations

Abstract

Non-genetic heterogeneities such as transcriptional noise and epigenetic modifications can lead to genetically clonal cells exhibiting heterogeneous responses to a uniform input. Heterogeneities in cellular and organelle morphology represent an additional type of non-genetic heterogeneity that may be capable of producing phenotypic variability, given structure-function relationships demonstrated at the biochemical level as well as morphological dysregulations observed in disease contexts. Until now, morphological heterogeneity in genetically clonal populations has remained largely unexplored due to a lack of appropriate tools. In this study, we develop an automated high-throughput system to quantitatively map morphological heterogeneity in a population of genetically clonal mouse embryonic fibroblasts (MEFs). We show that the morphological landscape is non-smooth, with cells resting in distinct morphological subpopulations. We additionally demonstrate tools to determine the relationship between cell cycle stage and morphological space state, as well as tools to map changes to the morphology space upon drug treatment. In total, we demonstrate a robust, high-throughput automated system for mapping the morphological landscape of a population of mammalian cells as well as tools to probe changes to this landscape upon external perturbation. This work has potential applicability in a variety of settings ranging from morphology-based drug and siRNA screens to improved quantitative screening tools for cytopathological applications.