Form and fate: The morphological landscape of zebrafish lateral line neuromast cells

Abstract

The shape of a cell both influences and is influenced by its fate. However, quantitative studies of cell shape in living, developing organisms have been limited in dimension and scope. To establish the utility and feasibility of cell shape analysis in vivo, cell shape analysis could first be applied to well-studied model systems that are amenable to high resolution live imaging. One such model is the lateral line, a sensory system in aquatic vertebrates that detects changes in water flow. Organs of the lateral line, called neuromasts, are composed of rosette-like clusters of mechanosensory hair cells surrounded by nonsensory support cells. Lateral line neuromasts have been studied by microscopic imaging for over a century, owing to their accessible location (just beneath the surface of the skin). Much is also known about the molecular regulation and cellular composition of zebrafish lateral line neuromasts. However, prior to my work, the morphologies of neuromast cells and nuclei had yet to be thoroughly described. In this work, I first introduce the utility of shape analysis in biology and review what was previously known about hair cell regeneration in nonmammalian vertebrates. I then describe a new study in which I performed data-driven, three-dimensional shape analysis of zebrafish lateral line neuromast cells using spherical harmonics, principal components analysis, and unsupervised clustering. I find that hair cells have a discretely distinct shape from support cells and successfully train a classifier to identify hair cells from shape features. I also find that support cell subpopulations have identifiable shape phenotypes, although the distinctions between support cell subpopulations were more nuanced compared to the difference between hair cells and support cells as a whole. To show the effects of loss of a cell type on shape, I characterized cell shape atoh1a mutants, which lack hair cells; I find that atoh1a mutants lack the shape phenotype associated with hair cells but do not appear to exhibit a mutant-specific shape. I also describe the coding pipeline I developed to process and analyze my dataset. Finally, I reflect on the conclusions and potential future applications of my work, which lays a foundation for further studies of the relationships between cell shape, cell fate, and organ shape in sensory epithelia and beyond.