CartoCell, a high-content pipeline for 3D image analysis, unveils cell morphology patterns in epithelia

Abstract

Decades of research have not yet fully explained the mechanisms of epithelial self organization and 3D packing. Single-cell analysis of large 3D epithelial libraries is crucial for understanding the assembly and function of whole tissues. Combining 3D epithelial imaging with advanced deep learning segmentation methods is essential for enabling this high-throughput analysis. We introduce CartoCell, a deep learning-based pipeline that uses small datasets to generate accurate labels for hundreds of whole 3D epithelial cysts. Our method detects the realistic morphology of epithelial cells and their contacts in the 3D structure of the tissue. CartoCell enables the quantification of geometric and packing features at the cellular level. Our Single-cell Cartography approach then maps the distribution of these features on 2D plots and 3D surface maps, revealing cell morphology patterns in epithelial cysts.

Developmental Pattern Formation: Spanish Contributions from a Biophysical Perspective

“…This is also related to other relevant problems that need to be addressed. On the one hand, while progress has been made during the last few years [115], better computational tools are needed to extract 3D information from microscopy data (i.e., segmentation and tracking). On the other hand, the effects of the geometry of the 3D embedding space of the developing tissue, in particular the curvature, is becoming recognized as a major driving force in developmental patterning, where more research is required [91,116].…”

Local and global changes in cell density induce reorganisation of 3D packing in a proliferating epithelium

“…For the automatic segmentation of 3D embryo stacks, we have followed a specific workflow pipeline adapted from a previously performed procedure called CartoCell (Andrés-San Román et al, 2023)…”

“…For the automatic segmentation of 3D embryo stacks, we have followed a specific workflow pipeline adapted from a previously performed procedure called CartoCell (Andrés-San Román et al, 2023) Training dataset was established from 3D Voronoi diagrams which was obtained combining the centroids of the cell nuclei of the sea star embryo as seeds and making masks from the cellular membranes to define the space to be filled. Custom Matlab scripts were used to calculate the nuclei centroids and the application VolumeSegmenter from Matlab were used to curate the membrane regions of the cells.…”

“…The method for extracting the proportion of scutoids from the Voronoi constructions was adapted from CartoCell (Andrés-San Román et al 2023). After extracting the outer and inner layers, we quantified the number of neighbours for each cell on both surfaces.…”