Journal Club

HomeMembersAssemblies and SectionsAssembliesPulmonary CirculationJournal Club ▶ Selective enhancement of endothelial BMPR-II with BMP9 reverses pulmonary arterial hypertension
Selective enhancement of endothelial BMPR-II with BMP9 reverses pulmonary arterial hypertension

Long L, Ormiston ML, Yang X, Southwood M, Gräf S, Machado RD, Mueller M, Kinzel B, Yung LM, Wilkinson MJ, Moore SD, Drake KM, Aldred MA, Yu PB, Upton PD, Morrell NW. Nature Medicine. 21(7):777-785.

Article Summary:

Background: Bone morphogenetic protein type II receptor (BMPR2) mutations are found in the majority of familial and up to 40% of idiopathic cases of pulmonary arterial hypertension (PAH). Reduced BMPR2 expression is a common feature of both human PAH and animal models, making it a potential novel therapeutic target in PAH. However, the complexity of the BMP receptor family, including 9 receptor types and over 20 ligands, has thus far hindered this approach. Of the known ligands, circulating BMP9 is thought to act as a vascular quiescence factor, effectively inhibiting the pulmonary artery endothelial cell (PAEC) proliferation that is characteristic of PAH. The authors demonstrate the specificity of BMP9 for the PAEC, followed by the use of BMP9 in 3 animal models of PAH to both prevent and reverse established disease.

Methods and Results: As an initial screen, human PAECs were stimulated with BMP9, BMP2, or BMP6, and gene expression analyzed by microarray. BMP9 alone induced regulation of multiple key genes involved in canonical BMP signaling, as well as down regulation of pro-apoptotic signaling. The ability of BMP9 to induce BMPR2 gene expression in human PAECs was validated by immunoblotting and qPCR demonstrating increased Smad1/5/8, Id1, Id3, and BMPR2. Similar results were produced using human blood outgrowth endothelial cells (BOECs).

In vivo treatment of both human PAECs and BOECs from controls and BMPR2 mutation-bearing subjects with BMP9 prevented TNFa/cycloheximide induced apoptosis. Furthermore, BMP9 reversed lipopolysaccharide-induced increases in monolayer permeability in both control and BMPR2 mutation-bearing human BOECs.

A mouse model of genetic PAH was developed using a knock-in allele of a common disease-causing mutation in humans, R899X, characterized by a premature stop mutation in the BMPR2 locus. Phenotypically, these mice spontaneously develop an elevated right ventricular systolic pressure (RVSP) and increased pulmonary artery (PA) muscularization without significant right ventricular hypertrophy (RVH) at 6 mo, which was reversed by administration of intraperitoneal (IP) BMP9 for 4 weeks. Furthermore, treatment with BMP9 was associated with enhanced monolayer integrity and decreased pulmonary vascular leak in wild type and mutant pulmonary endothelial cells following LPS challenge.

Two additional non-genetic models of PAH, monocrotaline and Sugen hypoxia, were used to further validate the findings. Treatment with IP BMP9 starting immediately after monocrotaline exposure prevented the elevated RVSP, RVH, and PA muscularization, with restoration of BMPR2 levels and downstream signaling. In a treatment protocol where BMP9 was given 3 weeks after monocrotaline exposure, BMP9 halted further disease progression. In the Sugen hypoxia rat model, daily BMP9 treatment initiated after the development of PAH resulted in reversal of established disease, including decreased RVSP, RVH, PA muscularization, frequency of neointimal lesions, and presence of pro-apoptotic and pro-proliferative pulmonary endothelial cells. Preventative treatment with BMP9 in a Sugen hypoxia mouse model averted the development of increased RVSP, RVH and PA muscularization. There was no evidence of off-target effects, specifically heterotopic ossification, following intraperitoneal or intramuscular injection.

Conclusions: These studies identify BMP9 as a BMPR2 ligand that preferentially targets the pulmonary endothelium, and at doses markedly lower than what has previously been described for alternative BMP ligands (BMP2, BMP4, BMP6). This degree of specificity for the pulmonary vasculature promotes canonical BMP signaling in PAECs without inducing signaling in PASMCs or promoting extra-pulmonary ossification. Furthermore, the efficacy in not just preventing but reversing established pulmonary vascular disease in 3 animal models of PAH highlights the therapeutic potential of BMP9.

Expert Commentary:

Suppression of the BMP pathway lies at the heart of pulmonary arterial hypertension (PAH), with BMPR2 mutations found in the large majority of heritable cases and a significant fraction of idiopathic cases and even of PAH assumed to be secondary to other causes.1 Moreover, even when it is not mutated, the BMP pathway is suppressed, and this suppression is widely thought to be required for the progression of disease. Although induction of the BMP pathway by adding ligand may seem to be an obvious treatment modality, it was generally assumed to be impractical for at least two reasons. First, the BMP pathway has important roles in just about every tissue type, and in specific, excess BMP leads to heterotopic ossification - creation of bone in soft tissue. Chronic over activation of the BMP pathway is the molecular cause of fibrodysplasia ossificans progressiva. Second, most BMP ligands are relatively short range, and so delivery is impractical.

The clever part of the current manuscript is identifying a ligand for which these two problems seem to disappear: BMP9. BMP9, also known as GDF2, is primarily produced in the liver, and is not one of the BMP ligands normally expressed at high levels in the lungs (those are BMP2, 4, and 7). In the present manuscript BMP9 is shown to have exquisitely higher activity (at least an order of magnitude) in pulmonary artery endothelial cells than the ligands normally expressed in lungs. Moreover, unlike other BMP ligands, it can be delivered through the circulation - in this manuscript, via I.P. injection. They demonstrate efficacy in preventing and reversing PAH in not only a BMPR2 mutant model (R899X knock-in), but also in two models in which BMP signaling is merely suppressed - Sugen/hypoxia and monocrotaline. Moreover, they show that using the same dose directly injected into muscle does not induce heterotopic ossification. This manuscript thus has high significance in that it may indicate the possibility of therapeutically attacking the problem of suppressed BMP signaling in an extremely direct manner - addition of ligand. 

Some caution is warranted, however - although it did not cause heterotopic ossification in the present study, it has been published to be the most osteogenic of the BMPs.2 It has also been associated with some liver pathologies, and so the therapeutic balance between efficacy and safety is likely to need fine tuning. It would also be useful to know what is happening with circulating BMP9 in human PAH patients. Despite these potential risks, the main reason for optimism, and the central finding of the manuscript, is the possibility of avoiding the off-target risks due to the extremely high sensitivity of pulmonary endothelial cells to BMP9, allowing therapeutic benefit with doses low enough to avoid causing problems elsewhere.


  1. Cogan JD, Pauciulo MW, Batchman AP, et al. High frequency of BMPR2 exonic deletions/duplications in familial pulmonary arterial hypertension. Am J Respir Crit Care Med 2006;174:590-8.
  2. Kang Q, Sun MH, Cheng H, et al. Characterization of the distinct orthotopic bone-forming activity of 14 BMPs using recombinant adenovirus-mediated gene delivery. Gene therapy 2004;11:1312-20.

Article summary by: Kara Goss, MD, Assistant Professor of Medicine, University of Wisconsin Madison

Expert commentary by: James West, PhD, Associate Professor of Medicine, Vanderbilt University School of Medicine

Last Reviewed: September 2016