Abstract
Background/Aim: We compared pulmonary irradiation-induced whole-lung, gene transcripts over 200 days after 20 Gy thoracic irradiation in female fibrosis-prone C57BL/6NHsd mice with fibrosis-resistant C3H/HeNHsd mice. Materials and Methods: Lung specimens were analyzed by real time polymerase chain reaction (rt-PCR) and changes over time in representative gene transcript levels were correlated with protein levels using western blot. Results: C3H/HeNHsd mice showed a significantly longer duration of elevation of gene transcripts for stress-response genes nuclear factor kappa-light-chain-enhancer of activated B cells (Nfkb), nuclear factor (erythroid-derived 2)-like 2 (Nrf2), transcription factor SP1 (SP1), activator protein 1 (AP1), radioprotection gene manganese superoxide dismutase (Sod2), and endothelial cell-associated genes von Willebrand factor (Vwf) and vascular endothelial growth factor (Vegf). C57BL/6NHsd mice showed acute elevation then down-regulation and a second elevation in gene transcripts for Nfkb, connective tissue growth factor (Ctgf), insulin-like growth factor-binding protein 7 (Igfbp7), tumor necrosis factor-alpha (Tnfa) Ctgf, Igfbp7, Tnfa, collagen 1a, and toll like receptor 4 (Tlr4). There were reciprocal patterns of elevation and decrease in levels of transcripts for epigenetic reader proteins bromodomain coding protein 1 (Brd1)Brd2,-3, and -4 between mouse strains. Conclusion: Regulatory pathways linked to radiation pulmonary fibrosis may identify new targets for mitigators of radiation-induced fibrosis.
A major dose-limiting complication of thoracic radiotherapy is lung damage (1-15). Acute radiation pneumonitis is characterized by endothelial cell swelling, alveolar transudates, and local pulmonary, as well as circulatory, elevation of inflammatory cytokines, prominently Interleukin-1 (Il1), Il6, Il10, Transforming growth factor beta (Tgfb), and Tnfa (4, 7-10, 16). Treatment with non-steroidal, or steroid anti-inflammatory agents may ameliorate symptoms of radiation pneumonitis, which is dependent upon the volume of irradiated lung, fraction size, and total dose (1, 2). Not all patients who suffer radiation pneumonitis go on to develop late pulmonary fibrosis, suggesting distinct differences in the pathophysiology between these two types of lesions (4, 12, 15, 17).
A valuable model system in which to dissect the mechanisms of late pulmonary fibrosis is the fibrosis-prone C57BL/6NHsd mouse compared to the pneumonitis-prone, but fibrosis-resistant C3H/HeNHsd mouse (17, 18-21). C3H/HeNHsd mice are intrinsically radiosensitive to total-body irradiation (TBI), and display radiation dose-dependent life shortening (18-22) and radiation pneumonitis. In contrast, C57BL/6NHsd mice are relatively radioresistant to TBI, demonstrating a brief interval of acute radiation pneumonitis followed by a period during which pulmonary histopathology is indistinguishable from unirradiated mice, then develop distinct organizing alveolitis (fibrosis) (18), involving proliferation of both intrinsic lung fibroblasts and bone marrow origin fibroblast progenitors, which migrate to sites of fibrosis (15, 18, 23, 24).
Recent studies with C57BL/6NHsd mice (24, 25) demonstrated an initial increase in expression of genes for promoter-binding proteins, stress response, and inflammatory cytokine genes, which returned to normal levels during a subsequent latent period, followed by elevation of many of the same gene transcripts and proteins, at the time of first detection of histopathological fibrosis (24). Transcripts for endothelial cell-related genes including Vwf and Vegf remained elevated in irradiated lungs of C57BL/6NHsd mice, indicating a persistent irradiation damage response (24, 25). The genetic and molecular biological determinants that initiate pulmonary fibrosis in C57BL/6NHsd, but not C3H/HeNHsd mice are not known.
In the present study, we quantitated post-thoracic irradiation levels of 25 representative gene transcripts in the irradiated lungs of C57BL/6NHsd, compared to C3H/HeNHsd mice. These transcripts were organized into six groups based on either their published involvement in radiobiological responses of cells and tissues to ionizing irradiation or data showing their association with lung fibrosis that was attributable to other causes. The first group consisted of representative inflammatory response genes for proteins known to be acutely elevated after thoracic irradiation of C57BL/6J/NHsd mice including: Nfkb, Nrf2, Sp1, and Ap1. A second group of gene transcripts was chosen based on endothelial cells, which have been implicated in early irradiation responses in several organs, including the intestine; these included Vwf, Vegf, Ctgf, and Il6 (26). A third group of transcripts included those for gene products known to be associated with initiation of fibrosis or found elevated in fibrotic areas of the lungs of patients with lung transplant rejection or those having lung resection for scleroderma lung including maganese superoxide dismutase (Sod2), Il1, Tnfα, Lysl Ox, Igfbp7, and Tgfb (27-28). As a fourth indicator of the fibrosis response, we measured levels of RNA transcripts for collagen 1a, known to be a dominant part of the fibrotic lung in C57BL/6NHsd mice (24-25). A fifth group included Toll-like receptors (Tlr) 1-7 known to be up-regulated during the inflammatory response (29, 30). An initial inflammatory response has been reported to occur in lungs of both fibrosis-prone and fibrosis-resistant mice (24, 31). Finally, a sixth group of transcripts included bromodomain epigenetic reader protein Brd1-4 (32, 33).
Materials and Methods
Thoracic radiation of mice. C57BL/6NHsd and C3H/HeNHsd mice were obtained from Harlan Laboratories (Indianapolis, IN, USA) and housed five per cage according to University of Pittsburgh Institutional Animal Care and Use Committee (IACUC) protocols. Mice were irradiated to the thoracic cavity with shielding of the head and neck, abdomen, and lower body according to published methods (34). Female C57BL/6NHsd mice received 20 Gy and groups of female C3H/HeNHsd mice received 14, 16.5, or 20 Gy irradiation to the thoracic field and were then maintained according to IACUC-directed laboratory conditions. Mice were sacrificed at serial time points after thoracic irradiation including pre-irradiation, days 2, 7, 14, 28, 60, 100, 125, 150 and 200 post-irradiation. A log-rank test was used to statistically analyze the survival curves after in vivo irradiation.
Measurement of levels of gene transcripts for irradiation-inducible transcription factors, growth factors, inflammatory cytokines, adhesion molecules, and radiation-protective enzymes by real-time polymerase chain reaction (RT-PCR). RNA was extracted from mouse lung using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions, quantified using a spectrophotometer, and stored at −80°C (34). Reverse transcription of 2 μg of total RNA to complementary DNA (cDNA) was accomplished using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's protocol.
In subsequent steps, expression of specific RNA moieties included: Gpdh (Gen-Bank: NM_008084.2), Gusb (Gen-Bank: NM_010368.1), Nfkb (Gen-Bank: NM_199267.2), Tnfa (Gen-Bank: NM_013693.2), Nrf2 (Gen-bank: NM_010902.3) (21), Nfkb (Gen-Bank: NM_008689.2), Jun (Gen-Bank: NM_010591.2), Sp1 (Gen-Bank: NM_013672.2), Ap1 (Gen-Bank: NM_001243043.1), Lysl Ox (Gen-Bank: NM_001178102.1), Tgfb1 (Gen-Bank: NM_011577.1), Vegfa (Gen-Bank: NM_001025250.3), Il1a (Gen-Bank: NM_010554.4), Fgf1 (Gen-Bank: NM_010197.3), Ifng (Gen-Bank: NM_008337.3), Il6 (Gen-Bank: NM_031168.1), Fap (Gen-Bank: NM_007986.2), Vwf (Gen-Bank: NM_011708.3), Ctgf (Gen-Bank: NM_010217.2), Sod2 (Gen-Bank: NM_013671.3), Igfbp7 (Gen-Bank: NM_001159518.1) and epigenetic reader proteins Brd1 (Gen-Bank: AK149714.1), Brd2 (Gen-Bank: AB010246.1), Brd3 (Gen-Bank: AB206708.2), and Brd4 (Gen-Bank: AF273217.1). collagen 1a (Gen-Bank: AK132180.1). Each was quantitated by RT-PCR. Ninety-six-well plates were prepared with 10 μl of Taqman Gene Expression Master mix, 5 μl of RNase-free water, 1 μl of the corresponding Taqman Gene Expression probe, and 4 μl of cDNA (totaling 2 μg cDNA) using the Eppendorf epMotion 5070 automated pipetting system (Eppendorf, Westbury, NY). The cDNA was amplified with 40 cycles of 95°C (denaturation) for 15 s and 60°C (annealing and elongation) for 1 min using the Eppendorf Realplex2 Mastercycler (17, 35).
Data for each gene transcript were normalized by calculating the differences (ΔCt) from the Ct of Gusb and Ct of target genes. The relative increase or decrease in expression was calculated by comparing the reference gene with the target gene (ΔΔCt) and using the formula for relative expression (=2ΔΔCt). Subsequently, ΔΔCt levels were compared and p-values were calculated using one-way ANOVA followed by Tukey's multiple comparison tests. The results were presented as the percentage increase in RNA above baseline levels which were adjusted to that of unirradiated C57BL/6NHsd and C3H/HeNHsd mice (34). Baseline transcript levels were standardized to that of Gpdh.
Western analysis for protein expression in irradiated mouse lungs. To determine levels of representative proteins Sod2, Nfkb, Brd4, and collagen-1 in C3H/HeNHsd lung tissue post 20 Gy thoracic irradiation, lung tissue was taken at acute (day 2), latent (day 60), and late (day 150) times and lysed in NP-400 buffer [50 mM Tris, pH 7.8, 10 mM ethylenediaminetetaacetic acid (EDTA), 150 mM NaC1, 1 mM phenylmethylsulfonyl fluoride (PMSF), 1% NP-40, and a protease inhibitor cocktail tablet (Roche Diagnostics, Indianapolis, IN, USA)]. Protein samples were separated in 15% polyacrylamide gels by electrophoresis and transferred to nitrocellulose membranes. Primary antibody to MnSOD (Novus Biologicals, Littleton, CO, USA) or α-tubulin (Sigma Aldrich, St. Louis, MO, USA) antibody were used. Horseradish peroxidase anti-rabbit or anti-mouse secondary antibody (Promega, Pittsburgh, PA, USA) was then applied and membranes developed with Super Signal West Dura ECL (Thermo Scientific, Rockford, IL, USA). Antibodies were obtained from Santa Cruz Biochemical Laboratories, Santa Cruz, CA, USA. Antibodies used were anti-Sod2 (ab13533), anti-Nfkβ p65(ab16502), anti-collagen 1 (ab34710), and anti-Brd4 [EPR5150(2)] (ab128874) from Abcam, Cambridge, MA, USA. For quantification of levels of proteins, band densities were quantified with Image J (National Institutes of Health, www.rsbweb.nih.gov/ij), previously as published (34).
In vivo imaging of coat changes post-thoracic irradiation. Groups of C3H/HeNHsd and C57BL/6NHsd mice were irradiated to the thorax to 20 Gy. At 47 days post-irradiation, when C57BL/6NHsd mice typically show graying of fur in irradiated fields (24), two isoflurane-anesthetized mice from each strain were imaged using a Xenogen IVIS 200 Imaging System (Advanced Molecular Vision Ltd, Lincolnshire, United Kingdom).
Serial imaging of luc+ bone marrow stromal cells. C3H/HeNHsd mice were injected intraperitoneally (i.p.) at each of several time points after 20 Gy thoracic irradiation with 1×106 luc+ bone marrow stromal cells from a C3H/HeNHsd luc+ stromal cell line. Following injection with D-luciferin (Gold Biotechnology, St. Louis, MO, USA), mice were imaged at serial timepoints using a Xenogen IVIS 200 Imaging System and the bioluminescent signal for each mouse was quantitated. As controls for pulmonary migration of luc+ bone marrow stromal cells in C3H/HeNHsd mice, C57BL/6NHsd mice received 20 Gy thoracic irradiation and were injected i.p. with 1×106 C57BL/6NHsd luc+ bone marrow stromal cells from a C57BL/6 luc+ stromal cell line (24). Mice were imaged and the bioluminescent signal for each mouse was quantified. Late-phase bioluminescence in C3H/HeNHsd mice was compared to late-phase bioluminescence C57BL/6NHsd mice using Student's t-test (24).
Pulmonary histopathology. Lungs from irradiated and unirradiated control C3H/HeNHsd and C57BL/6NHsd mice were removed, embedded in OCT and frozen. Frozen sections were stained with hematoxlyn and eosin (H&E), Masson's trichrome (for collagen), and antibody to leukocyte common antigen (CD45) (28). For CD45 immunostaining, sections were incubated with a monoclonal rat primary antibody to mouse CD45 (BD Pharmingen, San Jose, CA, USA), followed by a goat anti-rat Alexa Fluor 555 secondary antibody (Invitrogen, Grand Island, NY, USA).
Measurement of CpG promoter methylation. DNA was directly extracted from lung samples according to the manufacturer's instructions using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). The percentage CpG promoter methylation was then measured using the EpiTect Methyl II PCR Assay (Qiagen). Briefly, 250 ng of isolated DNA was incubated in 26 μl of 5× restriction digestion buffer and methylation-sensitive/-dependent/null enzymes overnight at 37°C, and then heat-inactivated for 20 min at 65°C. Ninety-six-well plates were prepared with 12.5 μl of SYBR Green qPCR Master mix, 6.5 μl of RNase-free water, 1 μl of the corresponding EpiTect PCR Primer, and 5 μl of DNA digest. The DNA was then amplified with 40 cycles of 97°C (denaturation) for 15 s and 72°C (annealing and elongation) for 1 min using the Eppendorf Realplex2 Mastercycler (Eppendorf, Westbury, NY, USA). The fraction of methylated DNA for each gene promoter was calculated by normalizing the DNA amount to the amount of digestible DNA. The amount of digestible DNA was equal to the total amount of DNA (determined from the mock digest) minus the amount of DNA resistant to DNA digestion (determined from the double digest).
Statistics. Survival of thoracic-irradiated C3H/HeNHsd and C57BL/6NHsd mice was compared pairwise with the two-sided log-rank test.
For the comparison of gene transcript expression between the two mouse strains, data are summarized as mean±standard deviation in each group. For lungs from each mouse strain, expression of each of the 25 genes was compared against that at day 0 for each day after irradiation. We also compared the two mouse strains for each gene transcript level at each day.
For the analysis of mouse lung protein levels by western blot data, data were summarized as mean±standard deviation for the densitometry of each protein for each group. For each mouse strain and each of four representative proteins, values on each day were compared to day-0 values, which were set to 1, using the two-sided one-sample t-test. We also compared mouse strains for each gene product protein at each day, using the two-sided two-sample t-test. In all the above tests, p-values less than 0.05 were regarded as significant. As these were exploratory studies, p-values were not adjusted for multiple comparisons.
Results
C3H/HeNHsd mice are relatively sensitive to 20 Gy thoracic irradiation compared to C57BL/6NHsd mice. After 20 Gy thoracic irradiation, C57BL/6NHsd mice all survived to 125 days. In contrast, 20% of C3H/HeNHsd mice irradiated to 20 Gy died within seven days from acute pneumonitis (p=0.0013) (Figure 1). Following lower doses of 14.5 Gy or 16 Gy thoracic irradiation, all C3H/HeNHsd mice survived to 100 days (Figure 1).
Distinct gene expression patterns in irradiated C3H/HeNHsd compared to C57BL/6NHsd mouse lung. Levels of baseline expression of each of the 25 gene transcripts were first standardized using glucose phosphate dehydrogenase (GPDH) as a control (Table I). For 13 transcript levels, there was no significant difference between mouse strains. The levels of transcripts for nine genes: vWF, VEGFa, TGFβ, TNFα, COL1a, TLR1, TLR4, Brd2, and Brd3 were higher in C57BL/6NHsd mouse lungs, and levels of three other gene transcripts: CTGF, Lysl Ox, and TLR7 were higher in C3H/HeNHsd mouse lungs. After 14.5 Gy or 16 Gy thoracic irradiation to C3H/HeNHsd mice, transcript responses were similar to those of the 20-Gy irradiation group, and there was no detectable late histopathological evidence of pulmonary fibrosis in any of the groups of irradiated mice (data not shown). Despite early death of some C3H/HeNHsd mice irradiated with 20 Gy, we compared lung tissue responses to the known fibrosis-inducing dose of 20 Gy thoracic irradiation in C57BL/6NHsd mice. Analysis of 25 representative pulmonary gene transcripts in C3H/HeNHsd mice following 20 Gy thoracic irradiation revealed an overall similar pattern to C57BL/6NHsd mice irradiated to 20 Gy to the thorax, but there were also distinct differences. In the irradiated C57BL/6NHsd mouse lung, there was an acute increase, a second interval decrease, and a clearly distinct late increase in expression of gene transcripts for NFkβ, Nrf2, SOD2, TGF-β, SP1, AP1, TNFα, IL-6, CTGF, and collagen 1a (Figure 2A, C and D). There was a persistent increase in expression throughout 200 days after irradiation of vWF, and VEGFa in C57BL/6NHsd mouse lungs (Figure 2B). While fibrosis-resistant C3H/HeNHsd mice also demonstrated thoracic irradiation-induced acute pulmonary increase in transcripts for NFkβ, Nrf2, Ap1, and Sp1, there was no comparable early increase in Sod2 or collagen 1a (Figure 2C and D), and there were significantly lower levels of endothelial cell-associated gene transcripts (Figure 2B). Therefore, while pulmonary fibrosis-resistant C3H/HeNHsd mice exhibited some gene transcript elevations in common with similarly 20 Gy thoracic-irradiated C57BL/6NHsd mice; the time course of elevation of several transcripts including Sod2, Tgfb, collagen 1a, Vwf, and Vegf differed. C3H/HeNHsd mice irradiated to lower doses of 14.5 Gy or 16 Gy showed the same temporal pattern of gene expression as did the 20-Gy irradiated mice (data not shown).
Distinct expression patterns of Tlr4 transcripts in the irradiated lungs of C57BL/6NHsd compared to C3H/HeNHsd mice. Levels of expression of Tlr family receptor RNA transcripts were measured in irradiated lungs of each mouse strain. In C57BL/6NHsd mouse lung, there was an acute phase and latent period decrease, followed by late increase at the time of detectable pulmonary fibrosis in transcripts for Tlr4 (Figure 2E) and Igfbp7 (Figure 2C). A prominent difference was the elevated levels of Tlr4 in C57BL/6NHsd mouse lungs during the late fibrotic phase, which was absent in lungs of C3H/HeNHsd mice (Figure 2E).
Reciprocal patterns of elevated bromodomain epigenetic reader protein gene transcripts in C3H/HeNHsd compared to C57BL/6NHsd mouse lung. A major difference between mouse strains following 20 Gy thoracic irradiation was observed with pulmonary levels of transcripts for bromodomain epigenetic reader proteins (Brd1-4). In C57BL/6NHsd mouse lungs, there were initial low levels of Brd1, -3 and -4, then increases during the latent period (days 50-125) in expression of Brd1-4, followed by clear decreases at day 150 when pulmonary fibrosis was detected (Figure 2F), while levels of Brd2 rose earlier at day 2 and persisted to day 50, levels were also low at day 200 in C57BL/6NHsd mouse lungs (Table II). In contrast, C3H/HeNHsd mouse lung showed increased expression of transcripts for Brd1-4 during the acute phase that persisted over 200 days (Figure 2F). Thus, the irradiated C57BL/6NHsd mouse lung showed a distinct pattern of late decrease in levels of Brd1-4 transcripts beginning with the onset of fibrosis (Figure 2F). While pre-irradiation pulmonary levels of Brd2 and Brd3 were elevated in C57BL/6NHsd mice, there were no significant differences in levels of Brd1 or Brd4 between strains (Table I). A summary of the comparative differences between the mouse strains in all 25 pulmonary gene transcript levels over time after lung irradiation is shown in Table II and statistical analysis in Tables III and IV.
Pulmonary protein levels are concordant with gene transcript levels in thoracic-irradiated C3H/HeNHsd and C57BL/ 6NHsd mice. We next determined whether increased gene transcript levels correlated with increased levels of protein. Representative proteins tested included two that have been associated with the acute pulmonary radiation reaction between days 1 and 14, a promoter-binding protein associated with the oxidative stress response (NFkβ), a radiation-protective antioxidant enzyme, MnSOD, and a protein associated with pulmonary fibrosis (collagen 1A). We also measured levels of Brd4 protein. Each protein was compared at times that corresponded to transcript levels in irradiated C57BL/6NHsd compared to C3H/HeNHsd mouse lungs (Figure 3) (Table V). In some cases, elevated RNA transcript levels were not concordant with elevated protein levels (NF-kβ, MnSOD) at the times tested, but in other cases, there was concordance. In C3H/HeNHsd mouse lung tissue at day 150, collagen 1a protein (elevated in pulmonary fibrosis) was significantly lower than levels in lungs of irradiated C57BL/6NHsd mice, while Brd4 protein level was significantly higher than the levels in C57BL/6NHsd mouse lung (Figure 3) (Table V). The data for collagen 1a and Brd4 were concordant with RNA transcript levels at day 150 (Figure 2D and F, respectively).
Distinct lung histopathology in thoracic-irradiated C3H/HeNHsd compared to C57BL/6NHsd mice. C3H/HeNHsd mice showed no physical evidence of late coat greying (Figure 4) and no histopathological evidence of pulmonary fibrosis after any of the three radiation doses, and dying mice showed no detectable pulmonary fibrosis (Figure 5A-D). By 150 days post-irradiation, C57BL/6NHsd mice displayed pulmonary organizing alveolitis (fibrosis) (Figure 5E). These results confirm and extend those in a prior publication (24).
The histopathology of C57BL/6NHsd lung tissue showed an acute inflammatory reaction, a stable latent period, followed by a late fibrosis phase 150 days post-irradiation. In contrast, lung tissue from C3H/HeNHsd mice irradiated with 20 Gy showed a robust acute inflammatory reaction following thoracic irradiation, but no late-phase (day 150) fibrosis reaction (Figure 5D).
At acute, latent period, and late time points, (days 2, 100, and 150 post-irradiation), C57BL/6NHsd mice lungs were quantitated for fibrosis, collagen deposition, and inflammatory cell accumulation in the lungs, and results were compared against those of C3H/HeNHsd mice irradiated to 14.5,16, or 20 Gy. Light microscopic quantitation of inflammation and percentage fibrosis in H&E-stained lung sections from the acute-phase, latent period, and late-fibrotic time points was performed. As shown in Figure 5, there was a pulmonary inflammatory infiltrate during the acute phase in lungs of C57BL/6NHsd mice irradiated with 20 Gy, and an inflammatory infiltrate in the acute phase in lungs of C3H/HeNHsd mice after doses of 14.5, 16, and 20 Gy. At day 100, the lungs of C3H/HeNHsd and C57BL/6NHsd mice showed little histopathological change. Histopathological sections of lung from C57BL/6NHsd but not C3H/HeNHsd mice showed significant fibrosis at day 150 (Figure 5E).
Lack of detectable bone marrow stromal cell homing to the lungs of thoracic-irradiated C3H/HeNHsd mice. C57BL/6NHsd mice chimeric for luc+ bone marrow have been reported to display fibrosis at 150-200 days after 20 Gy thoracic irradiation, and coincident homing of luc+ bone marrow stromal cells to the lungs (24).
To determine whether luc+ marrow stromal cell homing occurred in C3H/HeNHsd mice in the absence of a detectable pulmonary fibrosis, mice were injected intraperitoneally with 1×106 luc+ bone marrow stromal cells prepared from C3H/HeNHsd mouse marrow according to published methods (24) using luciferase mouse gene transduced stromal cell lines derived from long term bone marrow cultures. Mice were injected at serial times for live imaging of lung migration of luc+ cells using published methods (24). At each of the three time points after 20-Gy thoracic irradiation: immediately, 67 days, or 134 days post- thoracic irradiation, C3H/HeNHsd demonstrated no detectable luc+ bone marrow stromal cell homing to the lungs (Figure 6A-D). In contrast, C57BL/6NHsd mice injected i.p. at day 129 after 20 Gy thoracic irradiation showed significant homing of luc+ stromal cells derived from luciferase gene-transduced cells harvested from C57BL/6NHsd long-term marrow cultures (Figure 6E). The present results with C57BL/6NHsd mice confirm and extend other results from prior studies (24).
Changes in levels of CpG methylation in the gene promoter(s) for TGFβ, MnSOD (SOD2), FGF1, and, IGFbp7 correlate with levels of gene transcripts in irradiated mouse lungs. We next determined if increases in specific gene transcript levels correlated with de-methylation of the promoters for those genes. In lung tissue from C57BL/6NHsd mice irradiated to 20 Gy, inflammation-associated genes (Tgfb and Sod2) showed significant promoter de-methylation (<3% methylation) during both the acute and late stages, and increased methylation (25.67% and 20.67%, respectively) during the latent period (Figure 7). De-methylation of the Tgfb promoter in C57BL/6NHsd mouse lung was consistent with elevated levels of gene transcripts for Tgfb. In contrast, in lung tissue from C3H/HeNHsd mice, the promoter for the Tgfb gene showed significantly greater methylation during the late phase (days 150 and 200), concordant with lower levels of Tgfb transcript and absence of histopathological evidence of pulmonary fibrosis (Figure 7) (Table VI).
The MnSOD CpG gene promoter methylation levels were similar in the lung tissues of both C57BL/6NHsd and C3H/HeNHsd mice irradiated to 20 Gy, concordant with the observed similar levels of MnSOD transcripts (Figure 7B).
Levels of de-methylation of the CPG promoters for Fgf-1 (Figure 7C) and Igfbp7 (Figure 7D) were also concordant with the relative levels of transcripts. In C57BL/6NHsd mouse lung, the CpG promoter for the Igfbp7 gene showed stable methylation during the acute and latent period (19% and 17%, respectively), followed by significant de-methylation (<2% methylation) (Figure 7D) during the late fibrosis stage consistent with increased transcript levels (Figure 7C) (Table VI).
In C57BL/6NHsd mice, CpG promoters for the Vegf and Ctgf gene showed significant de-methylaton (<4% methylation) at day 150 following irradiation, consistent with levels of transcripts (Table VI). In contrast, lung tissue of C3H/HeNHsd mice irradiated to 20 Gy showed VEGF and CTGF gene promoters had lower levels of de-methylation consistent with the low levels of transcripts in these lung tissues. Therefore, levels of de-methylation of specific gene promoters in lung tissue correlated with increased gene transcript levels for those same genes over 200 days after 20 Gy irradiation in both pulmonary fibrosis-prone C57BL/6NHsd and fibrosis-resistant C3H/HeNHsd mice.
Discussion
In the present study, we measured levels of RNA transcripts for 25 genes in the lungs of irradiated C3H/HeNHsd mice compared to C57BL/6NHsd mice. C57BL/6NHsd mice showed bi-phasic early and late increase in expression of NFkβ, Nrf2, Sod2, Tgfb, Fgf-1, Sp-1 and Ap-1 separated by a time period when levels were reduced. While C3H/HeNHsd mice also demonstrated irradiation-induced elevation of NFkβ, Sp1, and Ap1, levels of MnSOD, Tgfb, and Fgf-1 were lower. Prior studies showed that relative levels of irradiated whole-lung RNA transcripts correlated with levels in separated endothelial and epithelial cells (25); therefore, we did not separate these cell populations for the comparison of irradiation-induced lung transcripts between these two mouse strains. Twenty percent of C3H/HeNHsd mice irradiated to 20 Gy to the thoracic cavity died rapidly, suggesting esophagitis or liver damage; however, no histopathological evidence of injury to these organs was detected. Since gene transcript elevations in lower-irradiation dose groups of C3H/HeNHsd mice were similar to the 20-Gy irradiation group, we compared the 80% lung irradiation survivors after 20 Gy in an attempt to detect differences in the response of the lung in this strain, which did not lead to the pulmonary fibrosis observed in 20 Gy irradiated C57BL/6NHsd mice.
There was prominent late elevation of RNA transcripts for Tlr4 and Igfbp7 in the lungs from irradiated C57BL/6NHsd mice compared to low levels at the same late time points in C3H/HeNHsd mice. Macrophage TLR4 elevation occurs during lung inflammation (36-42), and promotes hepatic (43), and renal fibrosis (37). The low level of TLR4 expression in irradiated C3H/HeNHsd mouse lung is consistent with prior data showing that low TLR4 levels ameliorate pulmonary (41) and renal fibrosis (37, 44). However, other data report different patterns of TLR4 expression. One report showed that both TLR4 and TLR2 were required to induce pulmonary fibrosis (38). Another report showed that TLR4 elevation prevented bleomycin fibrosis (41). A mutation on exon 3 of the TLR4 gene in C3H/HeNHsd mice was associated with infection and aberrant graft versus host disease responses, but these mice were still fibrosis-resistant (42). Since fibrosis resistance in C3H/HeNHsd mice appears to be independent of levels of TLR4, it is unlikely that TLR4 is the single critical regulator of irradiation pulmonary fibrosis.
Igfbp7 transcript levels were increased in the irradiation-induced fibrotic lungs of C57BL/6NHsd mice. This cell adhesion molecule promotes inflammation (45, 46), is elevated in human sclerodermatous lungs and in lungs of patients with lung transplant rejection (28), and in idiopathic pulmonary fibrosis (47), and abnormal wound repair. IGFbp7 down-modulates insulin-like growth factor 1 receptor, and inhibits cell growth and angiogenesis (8). IGFbp7 levels have been shown to correlate with slowed tumor progression suggesting that proliferating myofibroblasts in tumors and in fibrotic lungs of C57BL/6NHsd mice may also be examples of tissue responses in that strain (45-48).
A major observed difference between mouse strains was in the pulmonary levels of transcripts for bromodomain epigenetic reader proteins Brd1-4. Bromodomain epigenetic reader proteins bind to the acetylated lysine group of histone 4 (49) and modulate expression of genes including NFkb (33, 50-56). These gene products can induce fibrosis (33) and alter cell-cycle progression (57). Elevation of Brd1-4 transcripts in the lungs of C57BL/6NHsd mice during the period between the acute and fibrosis phases correlated with the time of decreased Tgfb, Tnfa, Nfkb, and Nrf2. Decrease in bromodomain gene transcripts in C57BL/6NHsd mice at days 150-200 correlated with elevated levels of collagen 1a, Tgfb, and Tlr4, and in other studies (24) coincided with the time of both histopathological lung fibrosis and bone marrow stromal cell homing to lungs. In contrast, stably-elevated levels of Brd1-4 over 200 days post-irradiation in C3H/HeNHsd mice correlated with decreased levels of transcripts for Vwf and Vegf.
BET bromodomain proteins, Brd2-4, are known to be associated with histones through mitotic divisions (32, 57). Our observation of low levels of Brd1-4 during fibrosis is inconsistent with prior publications showing that elevation of Brd1-4 causes fibrosis (32-33). Our data are also inconsistent with those showing that low levels of Brd1-4 down-regulate TLR4 (50-52, 58), which was elevated in irradiated C57BL/6NHsd mouse lungs at 150 days. Brd4 binds to and up-regulates the promoter for the gene for the fibrogenic cytokine Il6 (52, 55), and activates NFkb, which is also up-regulated during radiation fibrosis. Our data are consistent with a publication showing that Brd4 elevation inhibits bleomycin lung fibrosis in C57BL/6NHsd mice (33). This data is also consistent with our demonstration that Brd1-4 levels are elevated in irradiated fibrosis-resistant C3H/HeNHsd mice. Studies with a small-molecule inhibitor of BRD proteins, should allow for testing of the hypothesis that irradiation-induced lung fibrosis can be initiated in C3H/HeNHsd mice by reducing Brd1-4 levels (59).
Irradiation pulmonary fibrosis has been linked to elevated levels of TGFβ1 (7, 60-61), IL6 (9, 10) increased binding of NFκB and AP1 to the chemokine ligand 2 promoter (16), IGFBP-3 (13), and to the induction of extracellular matrix protein Tenascin (14, 16). Irradiated C57BL/6NHsd mice are known to display increased lung levels of TGFβ (17, 24), but also showed greater levels of irradiation-induced apoptosis in thymocytes and intestinal crypts (62). C57BL/6NHsd mice are prone to bleomycin-induced (18, 20, 22, 63), as well as irradiation-induced, lung (31) and intestinal (62) fibrosis. In contrast, C3H/HeNHsd mice are known to resist not only irradiation-induced pulmonary fibrosis (31), but also fibrosis caused by bleomycin (38), silica (64), ozone (65-66), and hyperoxia (67). Mouse strain-specific propensity for radiation fibrosis was not correlated with in vitro fibroblast radiosensitivity, suggesting that a paracrine or indirect mechanism of cell signaling is involved in vivo (31). We did not detect differences in the histopathology of cardiac tissues between irradiated mouse strains, and injected luc+ stromal cell progenitors of fibrosis-associated areas of the lung did not home to the heart (25); however, differences in physiological cardiac response to thoracic irradiation may be involved in the etiology of both rapid death of C3H/HeNHsd mice and fibrosis in lungs of C57BL/6NHsd mice.
While some molecular biological pathways are common to multiple etiologies of pulmonary fibrosis (38, 52, 57), the present data revealed unique signatures associated with the onset of radiation pulmonary fibrosis in C57BL/6NHsd mice: prominently elevated Tlr4 and a drop in transcripts and protein for the Brd4 epigenetic reader protein. Further studies are required to elucidate the potential interactive role of these elevated or decreased gene transcripts in the mechanism of irradiation pulmonary fibrosis.
Acknowledgements
This study was supported by NIH grants CA-R01-CA119927 and NIAID U19-A1068021. This project used the UPCI animal facility that is supported in part by award P30CA047904.
Footnotes
-
↵† Presented at the ASTRO 55th Annual Meeting, Atlanta, Georgia 22-25/9/2013.
- Received December 12, 2013.
- Revision received February 5, 2014.
- Accepted February 6, 2014.
- Copyright © 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved