Selective enhancement of endothelial BMPR-II with BMP9 reverses pulmonary arterial hypertension

Abstract

Genetic evidence implicates the loss of bone morphogenetic protein type II receptor (BMPR-II) signaling in the endothelium as an initiating factor in pulmonary arterial hypertension (PAH). However, selective targeting of this signaling pathway using BMP ligands has not yet been explored as a therapeutic strategy. Here, we identify BMP9 as the preferred ligand for preventing apoptosis and enhancing monolayer integrity in both pulmonary arterial endothelial cells and blood outgrowth endothelial cells from subjects with PAH who bear mutations in the gene encoding BMPR-II, BMPR2. Mice bearing a heterozygous knock-in allele of a human BMPR2 mutation, R899X, which we generated as an animal model of PAH caused by BMPR-II deficiency, spontaneously developed PAH. Administration of BMP9 reversed established PAH in these mice, as well as in two other experimental PAH models, in which PAH develops in response to either monocrotaline or VEGF receptor inhibition combined with chronic hypoxia. These results demonstrate the promise of direct enhancement of endothelial BMP signaling as a new therapeutic strategy for PAH.

Figure 1. BMP9 preferentially stimulates hPAECs

Volcano plots of differentially expressed genes in PAECs treated with (a) 1 ng/mL BMP9, (b) 10 ng/mL BMP6 or (c) 10 ng/mL BMP2 versus control after fitting linear models and adjusting P values for multiple testing (FDR). Differentially expressed genes of the TGF-β pathway are highlighted in grey. Dashed lines represent an adjusted P value of 0.05 and a fold change of +/− 1x. (d) Heat map showing whole gene-set perturbation (test statistics) of regulated pathways in PAECs following BMP9 treatment. (e) Signaling pathway impact analysis (SPIA) to detect alterations in common signaling pathways in PAECs in response to BMP9 treatment. Corrected for false discovery rate. (f) Immunoblotting for phosphorylated Smad1/5/8, Id1, Id3, BMPR-II and β-actin in PAECs cultured with or without varying concentrations of BMP9, BMP4 and BMP6 ligand for 2 hours. Representative of 3 experiments. (g) Expression of BMPR2 in PAECs with or without treatment with 10 ng/mL BMP9 for 24 hours (n=3; Student’s t test). (h) Representative agarose gels for products from chromatin immunoprecipitation on lysates from HMEC-1 cells (top) without treatment, (middle) 72 hours after siRNA knockdown of SMAD1 or (bottom) following a 24 hour treatment with 1 ng/mL BMP9. (i) Luciferase activity in HMEC-1 lysates transfected with a luciferase reporter construct with or without an upstream 5kb portion of the BMPR2 promoter with or without mutation of the putative Smad binding region. Cells bearing the reporter construct were treated with 10 ng/mL BMP9 for 24 hours (n=3; 1-way ANOVA, Tukey’s post test). ***P<0.001, **P<0.01, *P<0.05. Mean +/− SEM.

Figure 2. BMP9 prevents apoptosis in human PAECs via BMPR-II

Representative immunoblots and densitometric analysis for (a) phosphorylated JNK (n=4) and (b) cleaved caspase 3 (n=3) in human PAECs cultured with or without BMP9 (5 ng/mL) for 16 hours prior to an apoptotic stimulus with TNFα (10 ng/mL) and cyclohexamide (20 μg/mL) (1-way ANOVA, Tukey’s post test). (c) Representative flow cytometry plots of human PAECs stained with Annexin-V and propidium iodide (PI) with or without BMP9 pre-treatment and apoptotic stimulus. (d) Quantification of apoptotic (Annexin-V⁺/PI⁻) PAECs (n=5); 1-way ANOVA, Tukey’s post test). (e) Validation of siRNA knockdown in PAECs by immunoblotting for BMPR-II following treatment with either Dharmafect 1 transfection reagent (DH1), siRNA for BMPR2 (siBMPR2) or a pooled siRNA control (siCP). (f) Cytometric quantification of apoptosis by staining for Annexin-V and PI in PAECs treated with DH1 alone, siBMPR2 or siCP and cultured with or without BMP9 pre-treatment and apoptotic stimulus (n=6 for DH1 and siBMPR-II, n=5 for siCP; 1-way ANOVA for each siRNA group, Tukey’s post test). (g) Representative immunoblot and densitometric analysis of cleaved caspase-3 in PAECs following siRNA transfection with or without BMP9 pre-treatment and apoptotic stimulus. (n=3; 1-way ANOVA for each siRNA group, Tukey’s post test). All blots were re-probed for α-tubulin as a loading control. ***P<0.001, **P<0.01, * P<0.05. Mean +/− SEM.

Figure 3. BMP9 prevents apoptosis and promotes monolayer integrity in BOECs with and without BMPR-II mutations

(a) Quantification of apoptotic (Annexin-V⁺/PI⁻) control (n=5 individuals) and BMPR2 mutation-bearing BOECs (n=6 individuals) after culture with or without BMP9 (5 ng/mL) for 16 hours prior to the addition of TNFα (10 ng/mL) and cyclohexamide (20 μg/mL) for 6 hours. (b–c) Immunoblotting for (b) phosphorylated JNK and (c) cleaved caspase-3 in control and BMPR2 mutation-bearing BOECs with or without BMP9 pre-treatment and apoptotic stimulus. Representative of five immunoblots. All blots were re-probed for α-tubulin as a loading control. (d–f) Permeability of (d) control or (e) BMPR2 mutation-bearing BOEC monolayers to HRP, assessed as a measure of colorimetric absorbance after incubation periods ranging from 30 minutes to 2 hours with or without BMP9 (20 ng/mL) and/or LPS (400 ng/mL) (f) Quantification of monolayer permeability at 2 hours post-LPS. (n=4 individuals per group). AU: Arbitraty units. ***P<0.001, **P<0.01, *P<0.05, 1-way ANOVA, repeated measures Tukey’s post test for Control or BMPR2 Mutation BOECs; ###P<0.001, ##P<0.01, 1-way ANOVA, Tukey’s post test for all groups. Mean +/− SEM. (g) Representative immunofluorescence images of PAECs stained for VE-cadherin following 24 hour culture with or without BMP9 (10 ng/mL) and/or LPS (400 ng/mL). Scale bars 10 μm.

Figure 4. Pulmonary hypertension in Bmpr2^+/R899X knock-in mice is reversed by BMP9

(a) Assessment of right ventricular systolic pressure (RVSP) in 6-month-old, naive Bmpr2^+/R899X mice (n=21, 11 male, 10 female) and WT littermate controls (n=23, 13 male, 10 female) or with 4 weeks of BMP9 treatment (75 ng/day, i.p., n=10 (9 male, 1 female) for WT and n=11 (9 male, 2 female) for Bmpr2^+/R899X; 1-way ANOVA, Tukey’s post test). (b) Right ventricular hypertrophy (Fulton index, ratio of RV weight over LV and septal weight) in the same animals as a (not significant). (c) Quantitative assessment of pulmonary arterial muscularization in the same groups as a. Displayed as a non-, partially- and fully-muscularized arteries as a percentage of total alveolar wall and duct arteries (n=12 (6 male, 6 female) for naïve WT and Bmpr2^+/R899X mice; with BMP9 treatment n=3 (2 male, 1 female) for WT and n=5 (3 male, 2 female) for Bmpr2^+/R899X; 1-way ANOVA, Tukey’s post test). (d) Representative images of immunohistochemical staining for smooth muscle α-actin in lung sections from WT and Bmpr2^+/R899X mice with or without BMP9 treatment. (e) Assessment of RVSP and (f) RV hypertrophy in male 6 month-old WT (n=16), Smad1^+/− (n=9), Bmpr2^+/R899X (n=10) and compound Bmpr2^+/R899X/Smad1^+/− heterozygotes (n=10; 1-way ANOVA, Tukey’s post test). (g) Quantification of monolayer permeability at 2 hours post-LPS. (n=3; 1-way ANOVA, repeated measures Tukey’s post test for WT or Bmpr2^+/R899X cells; #P<0.05, 1-way ANOVA, Tukey’s post test for all groups). AU: Arbitrary units. (h) Images of lungs isolated from mice injected with 2 mg/kg LPS or vehicle and either 36.79 ng/25g BMP9 or vehicle (all i.p.) 22 hours prior to the i.p. injection of Evans blue dye, which was delivered 2 hours prior to sacrifice. (i) Quantitative assessment of extravascular Evans blue dye in the lungs of the mice described in h (Control: n=7 (5 male, 2 female) for WT, n=9 (4 male, 5 female) for Bmpr2^+/R899X; LPS: n=8 (5 male, 3 female) for WT, n=10 (6 male, 4 female) for Bmpr2^+/R899X; LPS + BMP9: n=8 (4 male, 4 female) for WT, n=9 (5 male, 4 female) for Bmpr2^+/R899X; 1-way ANOVA, Tukey’s post test for WT or Bmpr2^+/R899X mice; #P<0.05, 1-way ANOVA, Tukey’s post test for all groups). ***P<0.001, **P<0.01, *P<0.05. Mean +/− SEM.

Figure 5. BMP9 reverses established monocrotaline-induced pulmonary hypertension and prevents endothelial cell apoptosis in rats

(a) Assessment of right ventricular systolic pressure (RVSP) and (b) RV hypertrophy (Fulton index) in rats on a prevention protocol given vehicle (n=6) or monocrotaline (60 mg/kg, i.p.) and treated with BMP9 (n=7, 600 ng/day, i.p.) or vehicle (n=8) from days 0 to 21 post-MCT (1-way ANOVA, Tukey’s post test). (c) Quantitative assessment of pulmonary arterial muscularization in control rats or rats given monocrotaline, followed by daily injections with saline or BMP9 in a prevention (days 0–21 post-MCT) or reversal (days 21–35 post-MCT) protocol. Displayed as non-, partially- and fully-muscularized arteries as a percentage of total alveolar wall and duct arteries (n=5 for prevention, n=6 for reversal; 1-way ANOVA, Tukey’s post test for fully muscularized vessels). (d) Immunohistochemical staining for smooth muscle α-actin in lung sections from the rats as described in c. Scale bars 100 μm. (e) Assessment of right ventricular systolic pressure (RVSP) and (f) RV hypertrophy (Fulton index) in rats on a reversal protocol given vehicle (n=4) or monocrotaline (40 mg/kg, i.p.) and treated with BMP9 (n=10, 600 ng/day) or vehicle (n=9) from days 21 to 35 post-MCT (t test on 35 day groups treated with saline or BMP9). (g) Representative images of immunohistochemical staining for von Willebrand Factor (upper) and cleaved caspase-3 (lower) in lung sections from rats given vehicle control or monocrotaline (60 mg/kg, i.p.), treated with BMP9 (600 ng/day, i.p.) or saline from day 0 and sacrificed on day 5 post-MCT. Scale bars 50 μm. (h) Quantification of cleaved caspase-3 positive endothelial cells in lung sections from the groups described in a, b and g and sacrificed on days 2, 5 or 21 post-MCT (n=11 for Control, n=6 for days 2 and 5 post-MCT, n=5 for day 21 post-MCT; 1-way ANOVA, Tukey’s post test). ***P<0.001, **P<0.01, *P<0.05. Mean +/− SEM.

Figure 6. BMP9 reverses established pulmonary hypertension in rats in response to treatment with the VEGF receptor inhibitior SU-5416 in combination with chronic hypoxia

(a) Rats were given vehicle injections and maintained in normoxia (n=4) or challenged with SU-5416 (20 mg/kg, i.p.) and 3 weeks hypoxia (10% O₂) prior to 5 weeks normoxia and assessment at 8 weeks (n=7) or at 11 weeks following daily treatment with saline vehicle (n=11) or BMP9 (n=11; 600 ng/day, i.p.). (b) Assessment of right ventricular systolic pressure (RVSP) and (c) RV hypertrophy (Fulton index) for the rats described in a (1-way ANOVA, Tukey’s post test). (d) Quantification of non-, partially- and fully-muscularized arteries as a percentage of total alveolar wall and duct arteries (n=4 for control, n=6 for all other groups; 1-way ANOVA, Tukey’s post test for fully muscularized vessels). (e) Assessment of pulmonary arterial wall thickness as a percentage of luminal diameter (n=4 for control, n=6 for all other groups; 1-way ANOVA, Tukey’s post test). (f) Quantification of neointimal lesion frequency in the lungs of the rats described in a (n=3 for control, n=6 for all other groups; 1-way ANOVA, Tukey’s post test). (g) Quantification of cleaved caspase-3 positive endothelial cells in lung sections from the groups described in a (n=3 for control, n=6 for all other groups; 1-way ANOVA, Tukey’s post test). ***P<0.001, **P<0.01, *P<0.05. Mean +/− SEM. (h) Neointima formation in the lungs of the rats described in a. Lung sections were stained for smooth muscle α-actin (SMA) or with elastic van Gieson (EVG) stain. Scale bar = 50μm.