Mesenchymal stem cell-mediated reversal of bronchopulmonary dysplasia and associated pulmonary hypertension

Abstract

Clinical trials have failed to demonstrate an effective preventative or therapeutic strategy for bronchopulmonary dysplasia (BPD), a multifactorial chronic lung disease in preterm infants frequently complicated by pulmonary hypertension (PH). Mesenchymal stem cells (MSCs) and their secreted components have been shown to prevent BPD and pulmonary fibrosis in rodent models. We hypothesized that treatment with conditioned media (CM) from cultured mouse bone marrow-derived MSCs could reverse hyperoxia-induced BPD and PH. Newborn mice were exposed to hyperoxia (FiO(2)=0.75) for two weeks, were then treated with one intravenous dose of CM from either MSCs or primary mouse lung fibroblasts (MLFs), and placed in room air for two to four weeks. Histological analysis of lungs harvested at four weeks of age was performed to determine the degree of alveolar injury, blood vessel number, and vascular remodeling. At age six weeks, pulmonary artery pressure (PA acceleration time) and right ventricular hypertrophy (RVH; RV wall thickness) were assessed by echocardiography, and pulmonary function tests were conducted. When compared to MLF-CM, a single dose of MSC-CM-treatment (1) reversed the hyperoxia-induced parenchymal fibrosis and peripheral PA devascularization (pruning), (2) partially reversed alveolar injury, (3) normalized lung function (airway resistance, dynamic lung compliance), (4) fully reversed the moderate PH and RVH, and (5) attenuated peripheral PA muscularization associated with hyperoxia-induced BPD. Reversal of key features of hyperoxia-induced BPD and its long-term adverse effects on lung function can be achieved by a single intravenous dose of MSC-CM, thereby pointing toward a new therapeutic intervention for chronic lung diseases.

Keywords: airway hyperresponsiveness; chronic lung disease of infancy; hyperoxia; lung injury; lung vascular pruning.

Conflict of interest statement

Conflict of Interest: None declared.

Figures

Figure 1

MSC-CM-treatment partially reverses neonatal hyperoxia-induced alveolar injury, septal thickening, collagen accumulation, myofibroblast infiltration, and inflammation. Upper panel: Experimental Design. Newborn mice were either left in room air, or exposed to hyperoxia (FiO2=0.75) for 2 weeks (P1-14), and then intravenously injected once with CM from either bone marrow-derived MSCs or primary MLFs, and placed in room air for 2 additional weeks, followed by harvest, inflation and fixation of lungs. Lung sections were stained for hemotoxylin & eosin (A, C, E; 100×), and Mason Trichrome (B, D, F; 400×). Inserts were taken at 200× (H&E) and 400× (Mason Trichrome) magnification to illustrate interstitial inflammation (C, arrows), inflammation and fibrosis (D). Severe destruction of the alveolar architecture with overall widened airspaces and interstitial infiltration of inflammatory cells (macrophages, leukocytes) and myofibroblasts was seen in the hyperoxia-exposed/MLF-CM-treated mice (C, D) when compared to normoxic controls (A, B). Enhanced collagen deposition was seen in myofibroblasts, alveolar septa, and perivascular spaces in hyperoxia-exposed/MLF-CM-treated mouse lungs (D, arrows and inset). These changes were absent or greatly abrogated in MSC-CM-treated mice that had honeycomb-like alveoli (E, F) similar to normoxic controls, with residual emphysema. Quantification of mean linear intercept (Lm) as a surrogate of average air space diameter (G) and collagen content (H) are shown in the lower panels. Mean±SEM, n=4-7 per group, ***P<0.001, **P<0.01, *P<0.05. Scale bar = 100 μm (A, C, E) and 50 μm (B, D, F).

Figure 2

MSC-CM improve lung function after hyperoxia-induced lung injury. Pulmonary function testing was performed in six-week-old mice four weeks after the end of hyperoxia, and in age-matched normoxic control mice (see experimental design shown in Figure 1, and Methods section). Airway resistance (A) was measured at baseline and under escalating doses of intratracheal methacholine in order to quantify bronchial hyperreactivity. At baseline, airway resistance was not different between the three groups. MSC-CM-treatment fully reversed the abnormal increase in airway resistance seen in the MLF-CM group after administration of low to high intratracheal methacholine doses (5, 16, 50 mg/ml), to levels not different from normoxic controls (A). Dynamic lung compliance (Cdyn) was remarkably impaired in the hyperoxia-exposed/MLF-CM-treated mice at a methacholine dose of 50 mg/ml but normal in the MSC-CM-treated mice when compared with normoxic controls, indicating normalized compliance under metacholine stress (B). The corresponding histological findings are shown in Figure 1. Mean±SEM, n=3–4 per group, ** P<0.01, * P<0.05.

Figure 3

MSC-CM-treatment reverses pulmonary hypertension and RV hypertrophy in hyperoxia-induced lung injury. Pulmonary artery acceleration time (PAAT; syn. PAT), as a surrogate of mean PA pressure, was echocardiographically measured by PW-Doppler (A, C, E) and found to be shortened in hyperoxia-exposed/MLF-CM-treated animals (C), but normal in hyperoxia-exposed/MSC-CM-treated mice (blue line, E), when compared to normoxic controls (A, graphically summarized in G). Similar results were obtained for the PAAT/PAET ratio, where PAET is the pulmonary artery ejection time (syn. PET; data not shown). The end-diastolic diameter of the RV free wall (RVWT, RV wall thickness) was measured by M-mode echocardiography (B, D, F) and found to be increased in the hyperoxia-exposed/MLF-CM-treated mice but normalized in the MSC-CM-treated animals (H). Thus, a single dose of MSC-CM reversed the moderate pulmonary hypertension (G) and RV hypertrophy (H) that was associated with hyperoxia-induced BPD. See Figure 1 for experimental design. Mean±SEM, n=4 per group, * P<0.05.

Figure 4

MSC-CM attenuate hyperoxia-induced peripheral pulmonary arteriole remodeling. Lung sections were stained for the smooth muscle marker, α-SMA, and peripheral PA muscularization (P<0.05. Scale bar=25 μm.

Figure 5

MSC-CM rescue hyperoxia-induced loss of peripheral pulmonary blood vessels. Lung sections were stained for the endothelial cell marker vWF. vWF-positive vessels between 25 and 200 μm outer diameter were counted at 200× magnification in 10–15 random views as described under Methods. Compared to normoxic controls (A), hyperoxia exposure led to significant loss of small vessels P=0.0534). Quantification of pulmonary blood vessels of less than 50 and 50–100 μm diameter in vWF-stained lung sections (D) was performed as described under Methods. There was no significant difference in numbers of larger vessels (100-200 μm) between the groups. See Figure 1 for experimental design. Mean±SEM, n=4–7 per group, ** P<0.01, * P<0.05. Scale bar = 50 μm.

Figure 6

MSC-CM reverse pulmonary artery pruning in hyperoxia-induced lung injury. Pulmonary artery barium injections and subsequent computed tomography angiograms with 3D reconstruction of the PA vascularization of left lungs were performed ex vivo, as described in the Methods section. When compared to normoxic controls (A), severe rarefication (pruning) of peripheral PAs (arrows) was evident on CT-angiograms of hyperoxia-exposed/MLF-CM-treated mice at 4 weeks recovery in room air (B). Such hyperoxia-induced PA pruning was completely absent in lungs from mice at 4 weeks recovery in room air injected with a single intravenous dose of MSC-CM at the end of hyperoxia exposure (C). For experimental design, see Figure 1.