Abstract Body (Do not enter title and authors here): Background
Pulmonary arterial hypertension (PAH) and hereditary hemorrhagic telangiectasia (HHT) are congenital vascular syndromes arising from mutations in a common set of genes of the bone morphogenetic protein (BMP) and TGFβ signaling pathway. Sotatercept, a novel therapy for PAH targeting activin signaling, is associated with telangiectasias and arteriovenous malformations (AVM), suggesting perturbation of BMP or activin signaling in PAH can lead to HHT mimicry. To identify the ligands responsible for HHT-mimicry, we tested a panel of molecules with different activities against BMP and activin ligands in HHT-prone mice for the ability to induce AVMs.

Methods
Several recombinant proteins or antibodies with potential therapeutic effects in experimental or human pulmonary hypertension—anti-BMP9, anti-BMP10, ALK1-Fc (a BMP9/BMP10 ligand trap), ACTRIIA-Fc a.k.a. sotatercept (an activin/GDF/BMP10 and mild BMP9 ligand trap), ACTRIIB-Fc (an activin/GDF/BMP9/BMP10 ligand trap), or isotype control—were administered to HHT-prone 129X1/SvJ mice, at either 3- or 6-week of age for fifteen weeks (5-10 mg/kg twice weekly), with weekly paw and digital skin surveys, and blood counts taken at time of euthanasia.

Results
Treatment with ACTRIIB-Fc or ALK1-Fc led to prominent hindlimb digital AVMs and bleeding at high frequency, while anti-BMP10 induced tail AVMs. AVMs occurred within 30 days when treatment was initiated at 3 weeks of age, and 60 days when initiated at 6 weeks of age. Treatment with ACTRIIB-Fc promoted weight gain, consistent with known anabolic effects of this ligand trap on muscle and bone.

Conclusions
Combined antagonism of BMP9/BMP10 via ACTRIIB-Fc or ALK1-Fc exerted the strongest HHT mimicry in susceptible mice, while selective or weaker antagonism of BMP9 or BMP10 via neutralizing antibodies, ALK1-Fc, or ACTRIIA-Fc was less potent. HHT-susceptible strains of mice challenged with prospective PAH therapies can be used to reveal potential for HHT-mimicry due to BMP9/BMP10 antagonism in an age-dependent fashion. This pre-clinical model may be helpful to screen for the risk of AVMs in prospective therapies modulating BMP/activin signaling for PAH or other conditions.