Alk1 ligand requirements for AVM prevention in HHT - Bone morphogenetic protein 9 and 10 (BMP9 and BMP10) are plasma proteins that bind with high affinity to endothelial cell (EC) receptors endoglin (ENG, a non-signaling receptor) and activin receptor-like kinase 1 (ALK1, a serine/threonine kinase encoded by ACVRL1). Decreased signaling flux through this pathway due to mutations in ENG and ACVRL1 causes hereditary hemorrhagic telangiectasia (HHT), an autosomal dominant genetic disease characterized by development or arteriovenous malformations (AVMs). BMPs are generated as proproteins, with an N-terminal prodomain and a C-terminal growth factor domain (GFD), separated by a furin cleavage site. Cleavage releases the active covalent GFD dimer. Paradoxically, although our team has demonstrated that BMP10 is the only required ALK1 ligand for AVM prevention in zebrafish and mice, unprocessed pro-BMP10, which is assumed to be latent, predominates in human plasma. In this application, we present new data demonstrating that the most abundant ALK1 ligand in human and mouse plasma is a novel heterodimer consisting of unprocessed pro-BMP10 and processed BMP9, or “half-processed” pro- BMP10/BMP9. Moreover, we demonstrate that cultured hepatic stellate cells, which transcribe both BMP9 and BMP10 in vivo, secrete processed BMP9 but only unprocessed pro-BMP10, whereas cultured cardiomyocytes, which transcribe BMP10 in vivo, secrete processed forms of both ligands. Accordingly, we postulate that hepatic stellate cells are the source of circulating pro-BMP10/BMP9. Additionally, we hypothesize that BMP9 GFD dimer is insufficient to maintain normal vessels in the absence of BMP10 due to a critical role for the BMP10 prodomain in ligand stability and/or homing to endothelial cells. In Aim 1, we will apply cell biological and biochemical techniques to probe the mechanistic basis for differential processing of BMP9 and BMP10. Additionally, using mouse models, we will test the hypothesis that hepatic stellate cell-derived BMP10 is required for AVM prevention. In Aim 2, we will use X-ray crystallography and single-particle cryogenic election microscopy to gain insight into structural factors that drive BMP9/BMP10 heterodimer formation and limit processing of pro-BMP10. In Aim 3, we will assess pro-BMP10-containing ligand activity in human endothelial cells, zebrafish, and mice, probing requirements for endothelial ENG, proteolytic cleavage of pro-BMP10, and shear stress/blood flow. The short-term impact of this work will be a profound deepening of our mechanistic understanding of BMP biogenesis and regulation of activity. The long-term impact of this work will be to provide critical foundational knowledge of ALK1 ligand synthesis and activation that will inform development of safe, stable, and targeted biotherapeutics to prevent AVMs in HHT.
