Molecular and cellular mechanisms of epithelial-mesenchymal transition (EMT), crucial in development and pathogenesis, are still poorly understood. Here we provide evidence that distinct cellular steps of EMT occur sequentially during gastrulation. Basement membrane (BM) breakdown is the first recognizable step and is controlled by loss of basally localized RhoA activity and its activator neuroepithelial-transforming-protein-1 (Net1). Failure of RhoA downregulation during EMT leads to BM retention and reduction of its activity in normal epithelium leads to BM breakdown. We also show that this is in part mediated by RhoA-regulated basal microtubule stability. Microtubule disruption causes BM breakdown and its stabilization results in BM retention. We propose that loss of Net1 before EMT reduces basal RhoA activity and destabilizes basal microtubules, causing disruption of epithelial cell-BM interaction and subsequently, breakdown of the BM.

Disruption of the CLASP- and Dystroglycan-mediated cortical microtubule anchoring reduces epiblast–basement membrane interactions and initiates gastrulation.

Regulated disruption of the basement membrane (BM) is a critical step in many epithelial-mesenchymal transition (EMT) processes. Molecular mechanisms controlling the interaction between the BM and the basal membrane of epithelial cells and its subsequent disruption during EMT are poorly understood. Using chick embryos as a model, we analyzed the molecular complexity of this interaction during gastrulation EMT. Transcriptome data indicated that the BM of the gastrulation stage chick epiblast contains a full range of BM component proteins with unique subtype combinations. Integrins and dystroglycan are 2 major groups of basal membrane proteins involved in BM interaction. We provide evidence that dystroglycan gene expression is restricted to the epiblast during early development and its expression is downregulated in cells undergoing gastrulation EMT. The β-dystroglycan protein is localized to the basolateral membrane in epiblast cells and the basal localization is lost in cells undergoing EMT. Disruption of actin filaments leads to a decrease in the lateral membrane localization of β-dystroglycan and a relative increase in basal membrane localization, whereas disruption of microtubules leads to the loss of BM/basal membrane interaction and basal membrane β-dystroglycan localization. Overall, these data suggest an involvement of dystroglycan, especially the regulation of its expression and localization, in gastrulation EMT.

Hemangioblasts are bi-potential precursors for blood and endothelial cells (BCs and ECs). Existence of the hemangioblast in vivo by its strict definition, i.e. a clonal precursor giving rise to these two cell types after division, is still debated. Using a combination of mitotic figure analysis, cell labeling and long-term cell tracing, we show that, in chicken, cell division does not play a major role during the entire ventral mesoderm differentiation process after gastrulation. One eighth of cells do undergo at least one round of division, but mainly give rise to daughter cells contributing to the same lineage. Approximately 7% of the dividing cells that contribute to either the BC or EC lineage meet the criteria of true hemangioblasts, with one daughter cell becoming a BC and the other an EC. Our data suggest that hemangioblast-type generation of BC/EC occurs, but is not used as a major mechanism during early chicken development. It remains unclear, however, whether hemangioblast-like progenitor cells play a more prominent role in later development.

No abstract