Periostin-expressing cell-specific transforming growth factor-β inhibition in pulmonary artery prevents pulmonary arterial hypertension

Figures

• 

  Fig 1. Schema of our hypoxia-induced PAH model in Pn-Cre mice. In this model, we assume that hypoxia-induced Pn-mediated Cre-loxP activation will reduce Tgfbr1 expression in remodeled pulmonary artery and attenuate TGF-β signaling, resulting that Pn-Cre mice being available for conditional gene-knockdown under hypoxia.

• 

  Fig 2. Fluorescent immunostaining with lung tissues of wild-type mice (A-C). Immunostaining showing that Periostin (Pn) was mainly expressed in airway epithelial cell in normoxia control mice (A). Hypoxia-induced Pn was highly expressed in the perivascular area (B) and in neointima (C). Smooth muscle α actin, αSMA. Pn mRNA levels were significantly increased under hypoxia compared to normoxia control mice (a 4.0-fold increase) in lung (n = 4 in each group) (D). Group comparisons were performed by unpaired 2-tailed Student’s t test. Immunostaining with lung tissues of Pn-Cre/lacZ reporter mice showing LacZ-positive Pn-Cre-expressing cells (E-G). Dotted circles show overlap of two proteins. Pn-Cre is partially expressed with vimentin (E) and NG2 (F). Pn-Cre is partially expressed with α-SMA (G). Size bar: 20μm.

• 

  Fig 3. TGF-β signaling inhibition in Pn-Cre/Tgfbr1fl/fl mice. Real-time quantitative RT-PCR of lung tissues in control (n = 4), Tgfbr1fl/fl (n = 3), and Pn-Cre/Tgfbr1fl/fl (n = 3) mice (A and B). TGF-β receptor type I (A), but not type II (B) of Pn-Cre/Tgfbr1fl/fl was decreased in lung tissue under hypoxic condition. Western blots for Smads of lung tissue in control, Tgfbr1fl/fl, and Pn-Cre/Tgfbr1fl/fl with or without hypoxia exposure (C-F). Phosphorylation levels of Smad3 and Smad1 were estimated by the ratio of phospho-Smad ant total Smad band intensities. A representative blot for Smad3 and the summary data (N = 6-8/each) is shown in C and D, respectively. A representative blot for Smad1 and the summary data (N = 6-8/each) is shown in E and F, respectively. Group comparisons were performed by 1-way ANOVA.

• 

  Fig 4. Co-immunostaining with Phosho-Smad2/3, Pn, α-SMA and DAPI in lung tissues of Tgfbr1fl/fl mice (A) and Pn-Cre/Tgfbr1fl/fl mice (B) under hypoxia. Immunostaining shows Smad2/3 activation (red) under hypoxic condition in Tgfbr1fl/fl mice (A). Smad2/3 activation was reduced in Pn-Cre/Tgfbr1fl/fl mice especially in Pn-positive area indicated by dotted circles (B). Size bar: 20μm. Bar charts show RVESP, RVdP/dtmax, and indices of RV weight (RVW/LVW) in normoxia control (n = 10), hypoxia Tgfbr1fl/fl mice (n = 6), and hypoxia Pn-Cre/Tgfbr1fl/fl mice (n = 6) (C-E). RV pressure was significantly decreased in Pn-Cre/Tgfbr1fl/fl mice compare to Tgfbr1fl/fl mice (C). There was no significant difference either in RV dP/dt max or RVW/LVW between the two groups (D and E). Bar chart F shows PA muscularization in normoxia control (n = 9), hypoxia Tgfbr1fl/fl mice (n = 4), and hypoxia Pn-Cre/Tgfbr1fl/fl mice (n = 4), showing that PA muscularization was significantly attenuated in Pn-Cre/Tgfbr1fl/fl mice. Group comparisons were performed by 1-way ANOVA.