Effects of nintedanib on the microvascular architecture in a lung fibrosis model

Abstract

Nintedanib, a tyrosine kinase inhibitor approved for the treatment of idiopathic pulmonary fibrosis, has anti-fibrotic, anti-inflammatory, and anti-angiogenic activity. We explored the impact of nintedanib on microvascular architecture in a pulmonary fibrosis model. Lung fibrosis was induced in C57Bl/6 mice by intratracheal bleomycin (0.5 mg/kg). Nintedanib was started after the onset of lung pathology (50 mg/kg twice daily, orally). Micro-computed tomography was performed via volumetric assessment. Static lung compliance and forced vital capacity were determined by invasive measurements. Mice were subjected to bronchoalveolar lavage and histologic analyses, or perfused with a casting resin. Microvascular corrosion casts were imaged by scanning electron microscopy and synchrotron radiation tomographic microscopy, and quantified morphometrically. Bleomycin administration resulted in a significant increase in higher-density areas in the lungs detected by micro-computed tomography, which was significantly attenuated by nintedanib. Nintedanib significantly reduced lung fibrosis and vascular proliferation, normalized the distorted microvascular architecture, and was associated with a trend toward improvement in lung function and inflammation. Nintedanib resulted in a prominent improvement in pulmonary microvascular architecture, which outperformed the effect of nintedanib on lung function and inflammation. These findings uncover a potential new mode of action of nintedanib that may contribute to its efficacy in idiopathic pulmonary fibrosis.

Keywords: Angiogenesis inhibitors; Idiopathic pulmonary fibrosis; Intussusceptive angiogenesis; Microvascular corrosion casting; Synchrotron radiation tomographic microscopy.

Fig. 1

Functional lung testing showed that nintedanib was associated with a trend toward improved lung function and a reduced proliferative activity. a Functional lung testing. Pressure volume loops were conducted with an inflation pressure of up to 30 cm H₂O. Data are mean ± standard error of the mean. Vehicle group without bleomycin stimulation (-Bleo), n = 13, positive control animals stimulated with bleomycin 0.5 mg/kg (+Bleo), n = 24, bleomycin-stimulated animals treated with nintedanib 50 mg/kg twice daily from day 7 until day 19 (+Bleo +Nint), n = 23. b Tissue density assessed volumetrically by μCT. The ratio between the volume of the dense fibrotic tissue and the total lung volume (F/L) is presented. Vehicle group without bleomycin stimulation (-Bleo), n = 18, positive control animals stimulated with bleomycin (+Bleo), n = 24, bleomycin-stimulated animals treated with nintedanib 50 mg/kg twice daily (+Bleo +Nint), n = 24; ***p < 0.001, ****p < 0.0001. c Quantitative assessment of interstitial pulmonary fibrosis was carried out based on the standardized histopathologic quantification according to Ashcroft. Treatment with nintedanib (?Bleo ?Nint) resulted in a significant reduction of the fibrosis grade compared with bleomycin-stimulated lungs (+Bleo); ***p < 0.001. d Quantification of anti-Ki-67-positive cells showed that treatment with nintedanib significantly reduced proliferation rate. Data are mean ± standard error of the mean (-Bleo), n = 12, (+Bleo), n = 12, (+Bleo +Nint), n = 12; ***p < 0.001

Fig. 2

Increased tissue density and fibrotic foci was readily recognized in lower-resolution analysis of Azan- and Sirius-red stained sections in positive controls (+Bleo). The grade of fibrosis in nintedanib-treated animals was reduced (+Bleo +Nint). Bars 200 μm

Fig. 3

The expression of pro-fibrogenic, fibrolytic and pro-inflammatory transcripts in bleomycin-treated lungs (+Bleo). Relative lung mRNA transcript levels of a procollagen α1(I), b TGFβ1 (transforming growth factor-β1), c TIMP-1 (tissue inhibitor of metalloproteinase-1), and d MMP-2 (matrix metalloproteinase isoform 2) were determined by quantitative real-time polymerase chain reaction. Data are mean ± standard deviation. The x-fold increase in expression normalized to the expression in the vehicle group (-Bleo) is presented. *p < 0.05, **p < 0.01, ***p < 0.001

Fig. 4

Flow cytometric evaluation of bronchoalveolar lavage fluid. Cytometry of BALF showed that treatment with nintedanib had striking effects on the amount of inflammatory cells. Treatment with nintedanib resulted in a trend toward reduced cell counts. Total cell count: The total cell count (a) was reduced by 46%. Inflammatory cells: The monocyte (b) and lymphocyte (c) counts were halved, whereas neutrophils (d) were not affected by nintedanib. Data are mean ± standard error of the mean. Vehicle group without bleomycin stimulation (-Bleo), n = 9, positive control animals stimulated with bleomycin (+Bleo), n = 12, bleomycin-stimulated animals treated with nintedanib 50 mg/kg twice daily (+Bleo +Nint), n = 12; **p < 0.01

Fig. 5

Nintedanib restored a “normal” pulmonary vascular architecture. Scanning electron micrographs of microvascular corrosion casts illustrate the striking architectural differences between the positive control animals stimulated with bleomycin (+Bleo) and the bleomycin-stimulated animals treated with nintedanib (+Bleo +−Nint). The microvascular architecture of the lungs of animals treated with nintedanib was characterized by a recurrence of alveolar basket structure, whereas the lungs of the positive controls showed huge areas with densely packed vessel bulks constricting the airway system. Bars 200 μm

Fig. 6

Nintedanib normalized microvascular architecture as assessed by 3D-morphometry. Intervascular distances (a) and vessel diameters (b) in fibrotic foci assessed tridimensionally in microvascular corrosion casts shown as cumulative frequencies and box-whisker plots with the median, 5th, 10th, 25th, 75th, 90th, and 95th percentiles; mean (red). All p values were significant (p < 0.001). c Scanning electron micrographs of a fibrotic bleomycin lung (+Bleo). Vascular casts showing a chaotic tumor-like vasculature with sprouting angiogenesis (blue arrows). Adjacent to the bronchi are intussusceptive pillars (red arrowheads), hallmarks of intussusceptive angiogenesis. Lungs from positive controls stimulated with bleomycin (+Bleo), n = 12, bleomycin-stimulated animals treated with nintedanib 50 mg/kg twice daily (+Bleo +Nint), n = 14. Bars 20 μm.

Fig. 7

Sprouting and intussusceptive angiogenesis play a pivotal role in fibrogenesis shown in SRXTM. a Three-dimensional evaluation of microvascular corrosion casts by SRXTM illustrating the regular alveolar duct structure accompanied by the limiting alveolar entrance ring vessels. Bars 75 lm. b A typical example of vasculature in the basket-shaped alveoli. Bars 30 μm. c Analysis of the fibrotic lungs of bleomycin-stimulated animals revealed an increased, irregular vascularity with loss of tissue integrity and double-layered vessels. Bars 60 μm. d In fibrotic lungs intussusceptive holes (yellow circles) were found, indicative of the occurrence of intussusceptive angiogenesis around larger vessel structures. Several angiogenic sprouts are highlighted with a red square. e In the lungs of animals treated with nintedanib, the vascular density was decreased and regular alveolar patterns were observed. Bars 50 μm. f After nintedanib treatment, the alveolar entrance ring vessels (dashed red line) defining the alveolar opening were predominantly found in the remodeled tissue reenabling the blood-gas exchange. Bars 30 μm (see movies in supplemental material).

Fig. 8

Lung ultrastructure revealed therapeutic effects of nintedanib. a Semithin sections of a bleomycin-treated alveolar tissue showing loss of normal alveolar architecture with thickened interalveolar septa and an increased number of type II pneumocytes (yellow asterisks), bars 50 μm. b Transmission electron micrographs illustrating the close interaction of collagen deposition, fibroblasts, and inflammatory cells in bleomycin-treated mice. fibr fibroblasts, bars 10 μm. c In bleomycin-treated mice, a bulk of densely packed, thin capillaries were observed constricting alveolar space, bars 5 μm. d Nintedanib-treated animals did not display this huge mass of differentiated collagen fibers, bars 10 μm. e Collagen fibers appeared looser and more degraded in nintedanib-treated mice, bars 2 μm. Capillaries, c.