Smad1 and Smad5 differentially transduce BMP signaling during in vitro differentiation of mouse embryonic stem cells into dorsal interneurons

Abstract

A central unresolved question in development biology is how systems of overwhelming complexity arise from relatively few families of growth factors. Compounding this issue, signaling pathways often show signal convergence, where many ligands interact with fewer receptors, which then signal through a single second messenger complex. Here we investigate this question in the context of bone morphogenetic protein (BMP) signaling and its role directing dorsal spinal cord development, focusing on two receptor-regulated (R) Smads, Smad1 and Smad5. Multiple models have been proposed for their mode of action from acting redundantly through combined signal strength, to having distinct activities that drive different fate outcomes. We sought to distinguish between these models by generating CRISPR-edited Smad1 and Smad5 null mouse embryonic stem cell (ESC) lines to dissect the cell fate of activities of individual R-Smads, with a resolution not possible in vivo. Using a directed differentiation protocol for dorsal interneurons (dI), together with bioinformatic analyses, we have defined the roles of the R-Smads at key decision points along the dI specification timeline. Together, these findings support a model in which Smad1 and Smad5 play largely distinct roles in dorsal spinal cord development. While both R-Smads can activate canonical BMP-responsive transcriptional targets, they asymmetrically contribute to cell fate specification. Smad1 plays a restricted role, while Smad5 has a dominant role, regulating dorsal progenitor transcriptional dynamics and reiteratively directing the dorsal-most dI fates.