Defective mesenchymal Bmpr1a-mediated BMP signaling causes congenital pulmonary cysts

Abstract

Abnormal lung development can cause congenital pulmonary cysts, the mechanisms of which remain largely unknown. Although the cystic lesions are believed to result directly from disrupted airway epithelial cell growth, the extent to which developmental defects in lung mesenchymal cells contribute to abnormal airway epithelial cell growth and subsequent cystic lesions has not been thoroughly examined. In the present study, we dissected the roles of BMP receptor 1a (Bmpr1a)-mediated BMP signaling in lung mesenchyme during prenatal lung development and discovered that abrogation of mesenchymal Bmpr1a disrupted normal lung branching morphogenesis, leading to the formation of prenatal pulmonary cystic lesions. Severe deficiency of airway smooth muscle cells and subepithelial elastin fibers were found in the cystic airways of the mesenchymal Bmpr1a knockout lungs. In addition, ectopic mesenchymal expression of BMP ligands and airway epithelial perturbation of the Sox2-Sox9 proximal-distal axis were detected in the mesenchymal Bmpr1a knockout lungs. However, deletion of Smad1/5, two major BMP signaling downstream effectors, from the lung mesenchyme did not phenocopy the cystic abnormalities observed in the mesenchymal Bmpr1a knockout lungs, suggesting that a Smad-independent mechanism contributes to prenatal pulmonary cystic lesions. These findings reveal for the first time the role of mesenchymal BMP signaling in lung development and a potential pathogenic mechanism underlying congenital pulmonary cysts.

Keywords: Airway smooth muscle cells; BMP signaling; Bmpr1a; Lung branching morphogenesis; Lung development; Lung mesenchymal cells; Pulmonary cysts.

Fig.1.

Lung mesenchyme-specific deletion of Bmpr1a caused abnormal lung morphogenesis and prenatal airway cystic lesions beginning in mid-gestation. (A) Brightfield images of whole WT and Bmpr1a CKO mouse lungs at different embryonic stages. (B & C) Quantitative measurement and comparison of terminal airway branching numbers and sizes. (D) H&E-stained Bmpr1a CKO lungs at different embryonic stages. (E & F) EdU incorporation study for cell proliferation analysis in lung mesenchymal and cytokeratin-positive epithelial cells (n=6). (G) Apoptosis analysis by TUNEL assay. The positive control slides for apoptosis were generated by treating the tissue sections with DNase I. Pictures are representative of at least five samples in each condition.

Fig.2.

Airway smooth muscle was substantially reduced in Bmpr1a CKO lungs. (A) Scatter plots of RNA-Seq analysis for DEGs (n=3). (B) RPKM heatmap of Muscle System Process Genes (GO:0003012). Genes involved in the Muscle System Process with significant changes between Bmpr1a CKO and WT lungs are shown in the RNA-seq heatmap. The input data was RPKM values generated from the raw data using the R/Bioconductor program “edgeR”. (C & D) Whole mount immunostaining of Cdh1 and Acta2 in E14.5 and E15.5 lungs. Excessive dilation of airway terminals and extensive reduction of airway smooth muscle were observed in E15.5 Bmpr1a CKO lungs. The normal-looking airways are marked by arrows and the enlarged epithelial bud accompanied with compromised smooth muscle are marked by arrowheads. (E) E15.5 lung section stained with Cdh1& Acta2 or Myh11. *Vascular structures.

Fig.3.

Bmpr1a CKO resulted in a significant reduction in elastin expression underneath the airway epithelia of E15.5 lungs, but not in the pericytes, vasculature, and basement membrane. (A) Expression of Cspg4, Pecam1, Lama1, Lama2, Col3a1 and Eln at the mRNA level was measured by real-time PCR, *P < 0.05. (B) E15.5 lung section stained with Cspg4, Pecam1, Laminin and Col3a1. (C) E15.5 lung section stained with Cdh1, Acta2, and Elastin. (D) Reduced elastin expression of Bmpr1a CKO lungs at the protein level was detected by WB. (E-F): Elastin expression in E15.5 lung mesenchymal cells was upregulated by BMP4 and downregulated by BMP type 1 receptor-specific inhibitor LDN193189 (LDN), as detected by WB and immunofluorescence staining.

Fig.4.

The BMP pathway regulates the myogenesis of lung mesenchymal cells via the Smad independent pathway. (A) Activation of the intracellular downstream Smad1, p38, Jnk and Erk signaling pathways in WT and Bmpr1a CKO lung tissues was detected by the WB and quantified by densitometry. The levels of protein phosphorylation were normalized by the corresponding total protein and is presented as a relative change to the WT, *P < 0.05. (B) Gross view of whole lungs from Smad1/5 double conditional knockout mice (Smad1/5 CKO) and WT littermates showed that simultaneous deletion of Smad1 and Smad5 in lung mesenchyme completely disrupted lung development. (C) No airway dilation or cysts were observed in the H&E-stained tissue sections of the Smadl/5 CKO lungs at E17.5. (D) Expression of airway SMCs and elastin was not altered in E17.5 Smad1/5 CKO lungs, as shown by immunostaining of Cdh1, Acta2, and elastin. (E) Changes of intracellular signaling pathways in cultured fetal lung mesenchymal cells upon treatment with BMP4 (50ng/ml) and/or LDN193189 (200 nM) was detected by WB and quantified by densitometry. The relative change to the control condition is presented, *P < 0.05. (F-G) Altered expression of SMC genes at the protein level (Myh11) and the mRNA level (Myocd, Myh11 and Acta2) was respectively analyzed by immunostaining and real-time PCR for the primary culture of E15.5 WT lung mesenchymal cells treated with BMP4 (50ng/ml), LDN (200 nM) and SB (1 μM), *P < 0.05.

Fig.5.

Bmpr1a-mediated signaling played different roles in the differentiation of airway versus vascular SMCs. (A) YFP and DsRed expression patterns in the fetal lungs of Tagln-YFP/Cspg4-DsRed mice. a: airway, v: vessel. (B) Airway SMCs (YFP⁺) and vascular SMCs (YFP⁺/DsRed⁺) were isolated by FACS sorting. (C) The differential effect of Bmp4 treatment (50 ng/ml) on contractile protein Myh11 expression was analyzed in SMCs of non-vascular origin (YFP⁺) versus vascular origin (YFP⁺/DsRed⁺), and the role of Bmpr1a in mediating this effect was tested by adding its specific inhibitor LDN193189 (200 nM). The WB data of Myh11 expression was normalized to GAPDH (loading control) and is represented as a relative change to the control condition, *P < 0.05.

Fig.6.

Lung mesenchymal knockout of Myocd did not cause any branching abnormalities or lung cysts. (A) The expression of Myocd in Bmpr1a CKO lungs was substantially decreased, as measured by real-time PCR, **P < 0.01. (B) Deficiency in airway SMCs was observed in E15.5 mesenchyme-specific Myocd CKO lungs, as shown by co-immunofluorescence staining of Cdh1, Acta2, and elastin. (C) Comparison of Bmpr1a expression between E15.5 Myocd CKO and WT control lungs by immunofluorescence staining. (D) Comparison of the lungs between WT and Myocd CKO mice at the end of gestation (E18.5) did not reveal any significant morphological changes by gross view. No histological difference was found between the WT and the Myocd CKO lungs by examining their H&E-stained lung tissue sections.

Fig.7.

Mesenchymal Bmpr1a deletion disrupted airway epithelial proximal-distal differentiation and development. (A) Proximal epithelial cells, marked by Sox2 and Foxj1 staining, were significantly decreased in the proximal portion of the airways in E15.5 Bmpr1a CKO lungs. Ectopic distribution of distal epithelial cells marked by Sox9 and Spc staining was detected in the proximal airways of E15.5 Bmpr1a CKO lungs. (B) Heatmap of RNA-Seq data showing significant changes in the marker genes of proximal and distal epithelial cells. (C) Increased Bmp4 expression at the mRNA level was detected in E15.5 Bmpr1a CKO lung tissue by RT PCR. (D & E) Bmp4 expression in isolated fetal lung mesenchymal cells with genotypes of WT vs. Bmpr1a CKO was analyzed at both the mRNA and protein levels by RT PCR and WB respectively.