DeXtrusion: automatic recognition of epithelial cell extrusion through machine learning in vivo.

Accurately counting and localising cellular events from movies is an important bottleneck of high-content tissue/embryo live imaging. Here, we propose a new methodology based on deep learning that allows automatic detection of cellular events and their precise xyt localisation on live fluorescent imaging movies without segmentation. We focused on the detection of cell extrusion, the expulsion of dying cells from the epithelial layer, and devised DeXtrusion: a pipeline based on recurrent neural networks for automatic detection of cell extrusion/cell death events in large movies of epithelia marked with cell contour. The pipeline, initially trained on movies of the Drosophila pupal notum marked with fluorescent E-cadherin, is easily trainable, provides fast and accurate extrusion predictions in a large range of imaging conditions, and can also detect other cellular events, such as cell division or cell differentiation. It also performs well on other epithelial tissues with reasonable re-training. Our methodology could easily be applied for other cellular events detected by live fluorescent microscopy and could help to democratise the use of deep learning for automatic event detections in developing tissues.

Collective effects in epithelial cell death and cell extrusion.

Programmed cell death, notably apoptosis, is an essential guardian of tissue homeostasis and an active contributor of organ shaping. While the regulation of apoptosis has been mostly analysed in the framework of a cell autonomous process, recent works highlighted important collective effects which can tune cell elimination. This is particularly relevant for epithelial cell death, which requires fine coordination with the neighbours in order to maintain tissue sealing during cell expulsion. In this review, we will focus on the recent advances which outline the complex multicellular communications at play during epithelial cell death and cell extrusion. We will first focus on the new unanticipated functions of neighbouring cells during extrusion, discuss the contribution of distant neighbours, and finally highlight the complex feedbacks generated by cell elimination on neighbouring cell death.

Robustness of epithelial sealing is an emerging property of local ERK feedback driven by cell elimination.

What regulates the spatiotemporal distribution of cell elimination in tissues remains largely unknown. This is particularly relevant for epithelia with high rates of cell elimination where simultaneous death of neighboring cells could impair epithelial sealing. Here, using the Drosophila pupal notum (a single-layer epithelium) and a new optogenetic tool to trigger caspase activation and cell extrusion, we first showed that death of clusters of at least three cells impaired epithelial sealing; yet, such clusters were almost never observed in vivo. Accordingly, statistical analysis and simulations of cell death distribution highlighted a transient and local protective phase occurring near every cell death. This protection is driven by a transient activation of ERK in cells neighboring extruding cells, which inhibits caspase activation and prevents elimination of cells in clusters. This suggests that the robustness of epithelia with high rates of cell elimination is an emerging property of local ERK feedback.

Cell Extrusion: Crowd Pushing and Sticky Neighbours.

Cell extrusion is a highly coordinated process allowing the removal of an epithelial cell from the tissue layer without disrupting its integrity. Two new studies shed new light on the complexity of cell-cell coordination at play during cell extrusion.