Smad1/5 is acetylated in the dorsal aortae of the mouse embryo before the onset of blood flow, driving early arterial gene expression.

Abstract

During embryonic development, arteriovenous (AV) differentiation ensures proper blood vessel formation and maturation. Defects in arterial or venous identity cause inappropriate fusion of vessels, resulting in atypical shunts, so-called arteriovenous malformations (AVM). Currently, the mechanism behind AVM formation remains unclear and treatment options are fairly limited. Mammalian AV differentiation is initiated before the onset of blood flow in the embryo; however, this pre-flow mechanism is poorly understood. Here, we aimed to unravel the role of Smad1/5 signalling in pre-flow arterial identity, and in the process uncovered an unexpected control mechanism of Smad1/5 signalling. We establish that despite Notch1 being expressed in the pre-flow mouse embryo, it is not activated, nor is it necessary for the expression of the earliest arterial genes in the dorsal aortae (i.e., Hey1 and Gja4). Furthermore, interrupting blood flow by using the Ncx1 KO model completely prevents the activation of Notch1 signalling, suggesting a strong role of shear stress in maintaining arterial identity. We demonstrate that early expression of Hey1 and Gja4 requires SMAD1/5 signalling. Using embryo cultures, we show that Smad1/5 signalling is activated through the Alk1/Alk5/TGFβR2 receptor complex, with TGFβ1 as a necessary ligand. Furthermore, our findings demonstrate that early arterial gene expression requires the acetylation of Smad1/5 proteins, rendering them more sensitive to TGFβ1 stimulation. Blocking acetyl-CoA production prevents pre-flow arterial expression of Hey1 and Gja4, while stabilizing acetylation rescues their expression. Our findings highlight the importance of the acetyl-CoA production in the cell and provide a novel control mechanism of Smad1/5 signalling involving protein acetylation. As disturbed canonical Smad1/5 signalling is involved in several vascular conditions, our results offer new insights in treatment options for circumventing canonical Smad1/5 signalling.