Natural variation in macrophage polarization and function impact pneumocyte senescence and susceptibility to fibrosis

Abstract

Radiation-induced pulmonary fibrosis (RIPF), a late adverse event of radiation therapy, is characterized by infiltration of inflammatory cells, progressive loss of alveolar structure, secondary to the loss of pneumocytes and accumulation of collagenous extracellular matrix, and senescence of alveolar stem cells. Differential susceptibility to lung injury from radiation and other toxic insults across mouse strains is well described but poorly understood. The accumulation of alternatively activated macrophages (M2) has previously been implicated in the progression of lung fibrosis. Using fibrosis prone strain (C57L), a fibrosis-resistant strain (C3H/HeN), and a strain with intermediate susceptibility (C57BL6/J), we demonstrate that the accumulation of M2 macrophages correlates with the manifestation of fibrosis. A comparison of primary macrophages derived from each strain identified phenotypic and functional differences, including differential expression of NADPH Oxidase 2 and production of superoxide in response to M2 polarization and activation. Further, the sensitivity of primary AECII to senescence after coculture with M2 macrophages was strain dependent and correlated to observations of sensitivity to fibrosis and senescence in vivo. Taken together, these data support that the relative susceptibility of different strains to RIPF is closely related to distinct senescence responses induced through pulmonary M2 macrophages after thoracic irradiation.

Keywords: alveolar epithelial cell Type II; macrophage; senescence; strain.

Figure 1

Varying susceptibility to radiation induced pulmonary fibrosis and pneumocyte senescence among three strains of mice. C57L, C57BL6/J, and C3H/HeN mice were exposed to 5 daily fractions of 6 Gy (5x6 Gy) of thoracic irradiation. At 15, 22, 32 and 57 weeks after irradiation, lung tissue was collected (n=5 mice per timepoint and condition). (A) Kaplan–Meier plot of freedom from euthanasia of irradiated mice (n≥10 mice per group) with comparison of curves using log rank test. (B) Hydroxyproline content was assessed in lung tissue at the indicated time point (in weeks) after irradiation. (C) Senescence associated-β-Galactosidase activity was assessed in lung samples collected at the indicated time points after irradiation, followed by immunocytochemical localization of pro-surfactant C. The percent of senescent AECII was scored. Columns: mean, error bars: +SD, ^*p<0.05 for comparison to 0 Gy for the corresponding strain and timepoint by ANOVA with Tukey’s correction. ^§p<0.05 for comparison to C57L lungs exposed to 5x6 Gy by ANOVA with Tukey’s correction. (D) Representative images of costaining of tissue sections for senescence associated-β-Galactosidase activity and pro-surfactant C in the C57L strain.

Figure 2

Impact of thoracic irradiation on gene expression in lung tissue from three mouse strains with varying susceptibility to fibrosis. Mice were exposed to 5x6 Gy thoracic IR or no IR (0Gy). Samples of lung tissue (n=4 per dose) were collected at 15 weeks after IR, RNA was isolated, and further evaluated with RNA sequencing. (A) Principal component analysis of differentially expressed genes for all evaluated groups (n=4 per condition). (B) Volcano plot of the ratio of gene expression (irradiated: unirradiated) and p value for each observed gene. (C) Unsupervised hierarchical clustering of evaluated samples based on different expression of genes after irradiation. Irradiated samples were evaluated relative to a paired unirradiated sample of the same strain, thus expression depicted is the ratio (irradiated: unirradiated). (D) Hierarchical clustering of samples based on different expression of senescence and aging genes after irradiation. Irradiated samples were evaluated relative to a paired unirradiated sample of the same strain, thus expression depicted is the ratio (irradiated: unirradiated).

Figure 3

Identifying the genes belonging to ROS and NO production and the types of inflammatory cells in mouse lungs after thoracic radiation. (A) Reactome enrichment analysis of differentially expressed genes in the lungs of each strain (irradiated: unirradiated) at 15 weeks after irradiation. (B) The expression of genes in the ROS and NO production pathway were compared between the three strains (irradiated: unirradiated). The y-axis corresponds to the mean fold-change of each gene comparing irradiated to unirradiated within a strain (at 15 weeks after irradiation).

Figure 4

Accumulation of macrophages in irradiated lungs from three strains of mice. Immunohistochemical assays for (F4/80+), M1 macrophages (CD86+), and M2 macrophages (Agr-1+) were performed on lung tissue sections collected at the indicated timepoints after thoracic irradiation (5 x 6Gy). The numbers of cells were scored in whole lung (A) or within perifibrotic regions (within 400 μm of fibrotic regions) (B) in n=5 mice per time point and condition. Columns: mean, error bars: + SD, ^*p<0.05 for comparison to the corresponding 0 Gy by ANOVA with Tukey’s correction. ^§p<0.05 for comparison to lungs exposed to 5x6 Gy at 15 weeks by ANOVA with Tukey’s correction HPF: high power field (20X), IR: irradiation.

Figure 5

Characterization of macrophage phenotype across mouse strains after exposure to polarizing stimuli. Bone marrow derived macrophages from each strain were polarized with vehicle (PBS), LPS (1 ng/ml) or IL-13 (10 ng/ml). After 3 days of exposure to each stimulus, total RNA was isolated for further assays. (A) Polarization in response to LPS (M1) or IL-13 (M2) was evaluated by assessing the level of Cd86 and Arg1 mRNA with quantitative PCR (QPCR) normalized to β-actin mRNA. (B) The expression of genes related to M2 polarization was evaluated in macrophages treated with vehicle or IL-13 using the NanoString nCounter Gene Expression Assay and a custom code set. Unsupervised hierarchical clustering of M2 related genes was performed (left panel). (C) mRNA expression of Mrc1, Il10, Ccl2 and Ccl17 in polarized macrophages was confirmed by QPCR. (D) The concentrations of IL-10, IL-13, CCL2 and CCL17 in supernatants collected from polarized macrophages were determined with ELISA. Veh: PBS, BL6J: C57BL6/J, HeN: C3H/HeN. Columns: mean, error bars: +SD, ^*p<0.05 for comparison to the corresponding macrophages treated with vehicle. ^§p<0.05 for comparison to C57L macrophages exposed to IL-13 by ANOVA with Tukey’s correction.

Figure 6

Increased expression of components of NOX complexes and ROS production in M2 macrophages from fibrosis sensitive strains. The expression of ROS producing enzymes (NOX1, NOX2, NOX4) and subunits (NCF1, NCF2, NCF4) of the NOX2 complex was assessed in M2 macrophages polarized by IL-13 treatment. (A) mRNA expression of components of NOX complexes in M2 macrophages was quantified with QPCR and normalized to the expression of β-actin mRNA. (B) The expression of ROS producing enzymes (NOX1, NOX2) and subunits (NCF1, NCF2, NCF4) of the NOX2 complex was determined with western blot. Densitometry was performed for each protein and normalized to actin. Densitometry values are noted below each band. Columns: mean, error bars: +SD, ^*p<0.05 for comparison to the corresponding macrophages treated with vehicle. ^§p<0.05 for comparison to C57L macrophages exposed to IL-13 by ANOVA with Tukey’s correction.

Figure 7

NOX2-mediated macrophage superoxide production and senescence inducing capacity varies by strain. (A) Superoxide production was measured by luminol-amplified chemiluminescence in M2 macrophages (IL-13 treated) stimulated with PMA (100 ng/ml) in the presence or absence of a NOX2 inhibitor, GSK2795039 (100 nM). Each symbol: mean, error bars: +SD, ^*p<0.05 for comparison between macrophages with vehicle and IL-13, ^§p<0.05 for comparison between macrophages with vehicle +/- GSK2795039, ^ǂp<0.05 for comparison between macrophages with L-13 +/- GSK2795039 by multiple T Test. (B) Area under the curve analysis for Superoxide production after IL-13 and PMA treatment. (C) Enriched primary AECII from the three strains of mice were seeded on transwell inserts (pore size: 0.4 μm) and cocultured for 3 days with syngeneic M2 macrophages polarized with IL13. AECII were fixed after 3 days, and senescence associated β-gal activity was assessed followed by confirmatory immunocytochemical localization of pro-surfactant C (pro-SPC). The percent of senescent AECII was scored. Columns: mean, error bars: +SD, ^*p<0.05 for comparison to macrophages with vehicle. ^§p<0.05 for comparison to C57L macrophages exposed to IL13 by ANOVA with Tukey’s correction.

Figure 8

Increased numbers of macrophages expressing NOX2 in fibrotic lungs. C57L, C57BL6/J and C3H/HeN mice were exposed to 5 daily fractions of 6 Gy (5x6 Gy) of thoracic irradiation. (A) At 15 weeks after radiation, samples of frozen lung tissue were collected and dihydroethidium (DHE) oxidation was assessed. The level of cellular superoxide anion was quantified as the percentage of DHE-positive cell in each HPF (20x) region. (B) The expression of NOX2 was examined by QPCR and immunohistochemical assays (NOX2: brown, Nucleus: blue) in mouse lungs from the three strains 15 weeks after irradiation. (C) NOX2 expressing cells in lung were identified as a macrophage with co-labeling for F4/80. Columns: mean, error bars: +SD, ^*p<0.05 for comparison to each lung with 0 Gy. ^§p<0.05 for comparison to C57L lungs exposed to 5x6 Gy by ANOVA with Tukey’s correction.