Regenerative clustering of Enteroblasts in the Drosophila midgut revealed by a morphometric analysis

Abstract

Enteroblasts (EBs) are the cells responsible for the maintenance of the epithelium that lines the adult midgut in Drosophila. In response to cell death and damage, EBs undergo a Mesenchymal-Epithelial-Transition (MET) as they incorporate into the epithelium and differentiate into enterocytes (ECs). The morphogenetic mechanisms driving this MET process are not well understood. To improve phenotypic analysis of EBs, we established an analysis pipeline that uses machine learning segmentation to produce a reliable and automated quantification of cellular morphology and spatial distributions. EB morphology and fate is visualised using the Repressible Dual Differential stability cell Marker (ReDDM) approach. We show that wildtype EB cells exhibit a bimodal distribution pattern in which midguts fall into two categories: "quiescent" guts, in which EBs are evenly spaced out and newly formed ECs are uncommon, and "regenerative" guts, in which EBs are clustered and new ECs are prevalent. Using this system we first show that RNAi knockdown of Septate Junction proteins disrupts normal EB morphology and spatial distribution. With time-lapse imaging, we have also established that EBs are motile in nature, and when artificial tissue damage was introduced, exhibited increased cytoplasmic movements, and formed distinct clusters. We then demonstrate the utility of our pipeline in a small candidate screen for genes that might mediate EB clustering. Based on our results, we propose a working model that links the dynamic behaviour of EBs with midgut regeneration.