Data-Driven 3D Shape Analysis Reveals Cell Shape-Fate Relationships in Zebrafish Lateral Line Neuromasts

Abstract

AbstractCell shape is a powerful readout of cell state, fate, and function. With the advent of sophisticated microscopes, image segmentation algorithms, and numerical shape representations, it is becoming more feasible to study cell shape in developing tissues. However, few studies have analyzed cell shape in three dimensions in living, intact organisms. Here, we took advantage of the favorable imaging qualities of zebrafish lateral line neuromasts to generate a dataset of high resolution images with labeled cells and nuclei. Using a custom Python-based workflow, we performed semi-automated, 3D cell and nucleus segmentation. We then used spherical harmonics and principal components analysis to distill neuromast cell and nuclear shape variation into several interpretable, biologically meaningful parameters. We found that neuromast cell and nuclear shapes vary with cell location and identity. The distinction between hair cells and support cells was discrete and accounted for much of the variation in neuromast cell and nucleus shape, which allowed us to train classifiers to predict hair cell identity from cell and nucleus shape features. Using markers for support cell subpopulations, we found that support cell subtypes also had different shapes from each other; however, shape features did not distinguish as sharply between support cell subtypes, suggesting that support cells vary continuously in shape. To investigate the effects of genetic perturbation that results in loss of a cell type on neuromast cell shape, we examinedatoh1amutants that lack hair cells. We found that neuromasts fromatoh1amutants lacked the cell shape phenotype associated with hair cells, but did not exhibit a mutant-specific cell shape. Our results demonstrate the utility of using 3D cell shape features to characterize, compare, and classify cells in a living, developing organism.