Transforming growth factor-α (TGF-α) is a ligand for the EGF receptor (EGFR). EGFR activation is associated with fibroproliferative processes in human lung disease and animal models of pulmonary fibrosis. We determined the effects of EGFR tyrosine kinase inhibitors gefitinib (Iressa) and erlotinib (Tarceva) on the development and progression of TGF-α-induced pulmonary fibrosis. Using a doxycycline-regulatable transgenic mouse model of lung-specific TGF-α expression, we determined effects of treatment with gefitinib and erlotinib on changes in lung histology, total lung collagen, pulmonary mechanics, pulmonary hypertension, and expression of genes associated with synthesis of ECM and vascular remodeling. Induction in the lung of TGF-α caused progressive pulmonary fibrosis over an 8-wk period. Daily administration of gefitinib or erlotinib prevented development of fibrosis, reduced accumulation of total lung collagen, prevented weight loss, and prevented changes in pulmonary mechanics. Treatment of mice with gefitinib 4 wk after the induction of TGF-α prevented further increases in and partially reversed total collagen levels and changes in pulmonary mechanics and pulmonary hypertension. Increases in expression of genes associated with synthesis of ECM as well as decreases of genes associated with vascular remodeling were also prevented or partially reversed. Administration of gefitinib or erlotinib did not cause interstitial fibrosis or increases in lavage cell counts. Administration of small molecule EGFR tyrosine kinase inhibitors prevented further increases in and partially reversed pulmonary fibrosis induced directly by EGFR activation without inducing inflammatory cell influx or additional lung injury.
- pulmonary hypertension
interstitial lung diseases (ILD) are a heterogeneous group of disorders associated with over 130 distinct entities (6). Pulmonary fibrosis is characteristic of many forms of ILD, and idiopathic pulmonary fibrosis (IPF) affects over 90,000 people in the United States and more than 5 million patients worldwide (4, 27, 57). Currently, there are no proven therapies that prevent or reverse pulmonary fibrogenesis, emphasizing the need to identify new molecular targets.
A number of signaling pathways and molecules regulate matrix deposition and fibroblast proliferation in the lung including TNF, PDGF, basic FGF, IGF, and transforming growth factor-β1 (TGF-β1) (5, 29). TGF-α along with EGF and amphiregulin are ligands for the EGF receptor (EGFR). The EGFR is a membrane-bound receptor tyrosine kinase that belongs to a subfamily of four closely related receptors: HER1/EGFR/ERBB1, HER2/NEU/ERBB2, HER3/ERBB3, and HER4/ERBB4. Following ligand binding, these receptors form homo- and heterodimers leading to autophosphorylation of tyrosine residues in the cytosolic domains of the proteins. The phosphorylated tyrosine residues become docking sites for signaling molecules that activate cellular signaling pathways regulating a number of cellular processes, including proliferation and survival (38). A number of experimental studies support a role for EGFR activation in fibroproliferative processes. Madtes et al. (34) demonstrated increased TGF-α and EGFR after bleomycin injury in the rat with the highest expression noted at times of cellular proliferation and collagen deposition. Van Winkle et al. (55) demonstrated increased TGF-α and EGFR expression at sites of bronchiolar cell proliferation after naphthalene exposure. Similarly, EGFR and its ligands were increased in the lung following exposure to asbestosis and hyperoxia (32, 59).
We (28) previously generated transgenic mice wherein lung-specific expression of TGF-α was driven by the human surfactant protein C (SP-C) promoter. Expression of TGF-α in transgenic mouse lungs throughout pre- and postnatal lung development caused extensive peribronchial, perivascular, pleural, and interstitial fibrosis without inflammation (13, 18, 28). The extent of fibrosis was directly related to the level of TGF-α protein (9). Using a doxycycline (Dox)-regulatable transgenic system in the adult lung, expression of TGF-α caused progressive and extensive vascular adventitial, peribronchial, interstitial, and pleural fibrosis that was independent of inflammatory or developmental influences (17). Gene expression profiles observed after expression of TGF-α in the mouse lung were similar to those found in pulmonary fibrotic disease in humans (15).
Small molecule EGFR tyrosine kinase inhibitors (TKIs) act by competitive inhibition of ATP binding, which blocks intracellular autophosphorylation of EGFR tyrosine residues. EGFR TKIs are used in the treatment of tumors with activated EGFR including subtypes of non-small cell lung cancers. In a subgroup of lung cancer patients treated with the EGFR TKI, gefitinib, ILD was reported at a modestly increased frequency (22–24, 50, 54). Although placebo-controlled trials with the EGFR TKI erlotinib in lung cancer patients has not demonstrated pulmonary toxicity (2, 45), recent case reports have identified ILD in individual patients receiving erlotinib (1, 33, 36, 52).
Animal studies also suggest that EGFR TKIs may be useful in treating pulmonary fibrosis. Treatment with AG-1478 prevented vanadium pentoxide-induced peribronchial fibrosis in rats and prevented inflammation and ovalbumin-induced lung remodeling in mice (42, 56). Ishii et al. (25) reported that gefitinib prevented bleomycin-induced fibrosis in mice. In contrast, Suzuki et al. (49) reported that gefitinib augmented intratracheal bleomycin-induced fibrosis in mice. Considering the controversy over the efficacy of EGFR inhibition in bleomycin-induced fibrosis and the possibility of pulmonary toxicity due to the use of EGFR TKIs, the present study was undertaken to determine whether gefitinib and erlotinib prevented or reversed pulmonary fibrosis directly mediated through EGFR signaling.
CCSP-rtTA activator mice expressing the reverse tetracycline-responsive transactivator (rtTA) under control of the 2.3-kb rat Clara cell secretory protein (CCSP), also known as secretoglobin, family 1A, member 1 (Scgb1a1) gene promoter (53), were mated to conditional Dox-regulated transgenic mice containing the human TGF-α cDNA under the control of seven copies of the tetracycline operon [(TetO)7-cmv TGF-α] and a minimal cytomegalovirus (CMV) promoter (17). Single transgenic (CCSP-rtTA+/−) and bitransgenic CCSP-rtTA+/−/(TetO)7-cmv TGF-α+/− mice were produced within the same litter by mating homozygous CCSP-rtTA+/+ mice to hemizygous (TetO)7-cmv TGF-α+/− mice. All mice were derived from the FVB/NJ inbred strain. Mice were maintained in virus-free containment and handled in accordance with the Institutional Animal Use and Care Committee of the Cincinnati Children's Hospital Research Foundation. To induce TGF-α expression, Dox (Sigma, St. Louis, MO) was administered in the drinking water at a final concentration of 0.5 mg/ml and in food (62.5 mg/kg). Water was replaced three times per week. Mice were genotyped as previously described (17).
Administration of gefitinib and erlotinib.
Gefitinib (100 mg/kg; AstraZeneca, Macclesfield, England) was diluted in sterile, deionized water (0.0125 mg/μl) and stirred (4 h, 4°C) before administration to mice. Erlotinib powder (25 or 100 mg/kg; OSI Pharmaceuticals, Melville, NY) was suspended in 0.5% methylcellulose (0.015 mg/μl, 37°C; Colorcon, West Point, PA). Three hours before administration, food and water were removed from cages. Mice were then anesthetized (isoflurane; Abbott Laboratories, Chicago, IL), and 150–250 μl of sterile vehicle (water for gefitinib, methylcellulose for erlotinib) or drug was administered by gavage using a 20-gauge feeding catheter (Harvard Apparatus, Holliston, MA). Drug dosage was based on initial body weight and not adjusted with weight change during the study period. Mice were weighed at the beginning of the study and at weekly intervals. Statistical significance was determined by testing for normality and using a Kruskal-Wallis one-way ANOVA followed by an all pairwise Dunn's multiple-comparison test.
Lung histology, immunohistochemistry, and morphometrics.
Mice were killed with pentobarbital sodium (65 mg/ml) euthanasia solution (Fort Dodge Animal Health, Fort Dodge, IA), and lungs were inflation-fixed using 4% paraformaldehyde at 25 cmH2O of pressure and then allowed to fix overnight at 4°C. Fixed lungs were then washed with PBS, dehydrated through a graded series of ethanols, and processed for paraffin embedding. Sections (5 μm) were loaded onto polylysine slides for immunostaining, hematoxylin and eosin staining, or trichrome as previously described (10). A Nikon Microphot FXA EPI-FL3 microscope was used to acquire digital images, which were then analyzed in MetaMorph imaging software (v6.2; Universal Imaging). Pixel density for each ×10 digital image was 0.266 μm/pixel. As previously described (13, 14), pleural thickness was performed on digital images of five randomly selected fields per animal from distal gas exchange regions (n = 5 per group). Pleural thickness was measured using the measured distance function of MetaMorph.
Lung mechanics were assessed on mice with a computerized flexiVent system (SCIREQ, Montreal, Canada) as previously described (21, 44). Briefly, mice were anesthetized with ketamine and xylazine, tracheostomized, and then ventilated with a tidal volume of 8 ml/kg at a rate of 450 breaths/min and positive end-expiratory pressure (PEEP) of 2 cmH2O computerized by the SCIREQ system thereby permitting analysis of dynamic lung compliance. The ventilation mode was changed to forced oscillatory signal (0.5–19.6 Hz), and respiratory impedance was measured. Tissue resistance or damping and tissue elastance was obtained for mice at 2 cmH2O PEEP by fitting a model to each impedance spectrum. With this system, the calibration procedure removed the impedance of the equipment and tracheal tube.
Total lung collagen.
Total lung collagen was determined by quantifying total soluble collagen (Sircol Collagen Assay, Biocolor). The left lung was homogenized in 5 ml of 0.5 M acetic acid containing pepsin (1 mg/10 mg tissue; Sigma-Aldrich) and incubated (24 h, 24°C, with 240 rpm shaking). Sircol dye was added (1 ml/100 μl, 30 min), the sample was centrifuged (12,000 rpm for 12 min), and the pellet was suspended (1 ml of 0.5 M NaOH). The optical density was measured with a spectrophotometer (540 nm).
Right ventricular hypertrophy (RVH) was assessed as an index of pulmonary hypertension as previously described (7, 30). Briefly, hearts were removed and dissected to isolate the free wall of the RV from the left ventricle and septum (LV+S). The ratios of RV weight to LV+S weight (RV/LV+S) were used as an index of RVH, which develops as a result of pulmonary hypertension.
Quantitative real-time PCR.
Fourteen transcripts previously identified by microarray to be altered in human IPF and in murine models of pulmonary fibrosis were measured by quantitative real-time PCR (qRT-PCR; Ref. 15). Total RNA was isolated from mouse lung (n = 3–4 mice per group) with TRIzol reagent (Invitrogen, Carlsbad, CA), and quantity was assessed by absorbance at 260 and 280 nm (SmartSpec 3000; Bio-Rad, Hercules, CA). RNA (100 ng) was reverse-transcribed (High Capacity cDNA Archive Kit; Applied Biosystems, Foster City, CA), and cDNA (10 μg) was used in each PCR reaction with primers and TaqMan Universal PCR Master Mix (Applied Biosystems). Primers for procollagen type I-α1 (COL1A1; cat. no. Mm00801666_g1), procollagen type III-α1 (COL3A1; cat. no. Mm00802331_m1), endothelin receptor type B (EDNRB), frizzled homolog 1 (Drosophila) (FZD1; cat. no. Mm00445405_s1), midkine (MDK; cat. no. Mm00440279_m1), caveolin, caveolae protein 1 (CAV1; cat. no. Mm00483057_m1), platelet-endothelial cell adhesion molecule-1 (PECAM-1; cat. no. Mm00476702_m1), kinase insert domain protein receptor (KDR; cat. no. Mm00486524_m1), forkhead box F1a (FOXF1A; cat. no. Mm00487497_m1), resistin-like-α (RETNLA; cat. no. Mm00457862_m1), VEGF-A (cat. no. Mm00455174_m1), serine (or cysteine)/serpin peptidase inhibitor, clade F, member 1 (SERPINF1; cat. no. Mm00465988_m1), extracellular superoxide dismutase (SOD3; cat. no. Mm00448831_m1), and endothelial-specific receptor tyrosine kinase (TEK; cat. no. Mm00607939_s1). Analysis was performed with an Applied Biosystems 7900HT Fast Real-Time PCR System (95°C, 10 min; 40 cycles at 95°C, 15 s, and 60°C, 1 min). For relative quantization, 42 days on Dox samples were compared with nontreated controls samples using the comparative critical threshold (CT) method (ΔΔCT) normalizing each sample to β-actin.
TGF-α concentrations in mouse lung with and without Dox induction was determined by ELISA (Oncogene Research Products, Cambridge, MA) as previously described (17, 18). Briefly, lungs were homogenized in 2 ml of PBS (pH 7.4) containing protease inhibitors (Complete Protease Inhibitor Cocktail; Roche, Indianapolis, IN) and centrifuged (1,500 g, 15 min), and the supernatant was stored at −70°C.
EGFR Western blots.
EGFR activation was confirmed by assessing the levels of phosphorylated EGFR relative to total EGFR in control and bitransgenic mice after 1 day of Dox. Western blot analysis was performed in lung tissue from single-transgene controls (CCSP/-) and bitransgenic mice treated with and without 100 mg/kg erlotinib by using primary antibodies against phosphorylated EGFR (1:1,000, rabbit monoclonal, pY1086; Epitonic, Burlingame, CA) and total EGFR (1:5,000, rabbit polyclonal; kind gift from Dr. Brad Warner, Washington University). Goat anti-rabbit secondary antibodies were used, and chemiluminescence detection was performed using the ECL Plus system (Amersham Biosciences).
Bronchoalveolar lavage cell counts.
Following euthanization, the trachea was cannulated, and the lungs were lavaged three times with 1 ml of HBSS (137 mM NaCl, 5.4 mM KCl, 0.44 mM KH2PO4, 0.34 mM Na2HPO4, 4.2 mM NaHCO3, and 5.6 mM glucose) as previously described (17, 19). Differential cell counts were performed on Diff-Quick-stained (Baxter Diagnostics, McGraw Park, IL) cytospin (Cytospin 3, Shandon Scientific). Two hundred cells per slide were counted. The remaining bronchoalveolar lavage (BAL) fluid was then centrifuged (150 g, 10 min), cells were resuspended in 100 μl of PBS, and 10 μl of suspension was mixed with 10 μl of trypan blue (0.4%) and counted on a hemocytometer.
Data are means ± SE. Statistical comparisons were made using ANOVA followed by a Student-Newman-Keuls all pairwise comparison to identify significant differences except as described above for body weight measurements. P < 0.05 was considered statistically significant.
Prevention of TGF-α-mediated fibrosis with gefitinib.
CCSP-rtTA/TetO-TGF-α mice were treated with Dox to induce TGF-α expression and concomitantly gavaged daily with either sterile water or gefitinib for 4 wk. As previously reported, induction of TGF-α caused extensive fibrosis localized to the pleural surfaces, perivascular and peribronchial adventitia. Gefitinib-treated mice demonstrated reductions in pulmonary fibrosis (Fig. 1) and mean pleural thickness (Table 1) and improvements in altered lung mechanics (Fig. 2) caused by expression of TGF-α. Gefitinib prevented increases in lung collagen content in mice caused by expression of TGF-α (Fig. 3). Total and differential cell counts in BAL were not altered by gefitinib (Table 2).
Partial reversal of TGF-α-mediated fibrosis with gefitinib.
After 4 wk of TGF-α expression, lung histology revealed extensive fibrotic lesions associated with alterations in lung mechanics. To determine whether EGFR inhibition influences the progression of established fibrosis, mice were gavaged with gefitinib or water at the beginning of week 5 of Dox and treated daily during an additional 4 wk of Dox induction (8 wk total Dox). Gefitinib significantly improved pulmonary fibrosis assessed by histology, mean pleural thickness, and total lung collagen content and prevented further declines in lung mechanics (Table 1 and Figs. 2–4). Gefitinib prevented the progressive weight loss caused by expression of TGF-α in this model (Fig. 5). Expression of TGF-α for 8 wk caused RVH that was diminished by treatment with gefitinib (Fig. 6).
Effects of gefitinib on gene expression.
To assess the effects of gefitinib on expression of mRNAs previously associated with pulmonary fibrosis, qRT-PCR analysis for a selected group of mRNAs induced by TGF-α was determined (15). Gefitinib treatment reduced TGF-α-induced changes in expression of MDK, SERPINF1, and FZD1 (Fig. 7A) and a number of mRNAs associated with vascular development and remodeling (VEGF-A, KDR, TEK, EDNRB, FOXF1A, and PECAM-1; Fig. 7B). Gefitinib treatment did not alter the levels of TGF-α protein expressed in the lung of the Dox-treated mice (Table 1).
Erlotinib prevents TGF-α-induced fibrosis.
CCSP-rtTA/otet-TGF-α mice were placed on Dox and concomitantly gavaged with either sterile water or 25 or 100 mg/kg erlotinib daily for 4 wk. Erlotinib treatment at 100 mg·kg−1·day−1, but not at 25 mg·kg−1·day−1, reduced total lung collagen compared with vehicle-treated mice (Fig. 8A). Both doses of erlotinib prevented TGF-α-induced abnormalities in pulmonary mechanics (Fig. 8B) and prevented histological evidence of pulmonary fibrosis (data not shown). Erlotinib (100 mg/kg) prevented TGF-α-induced phosphorylation of EGFR as measured by Western blot of whole lung homogenates (Fig. 8C).
Effects of erlotinib on gene expression.
Treatment with erlotinib at 100 mg·kg−1·day−1 prevented TGF-α-mediated changes in expression of mRNAs (Fig. 9, A and B). Erlotinib at 25 mg/kg partially ameliorated TGF-α-induced changes in gene expression.
Using a transgenic mouse model of pulmonary fibrosis caused by lung-specific expression of the EGFR ligand, TGF-α, the present study demonstrates that administration of EGFR inhibitors prevents both the generation as well as the progression of established pulmonary fibrosis and associated alterations in lung mechanics. These findings, in the context of elevated TGF-α and EGFR detected in a number of human fibrotic lung diseases, support the concept of targeted EGFR inhibition for progressive fibrotic disease (12, 35, 39, 40, 47, 48, 51, 58).
In the present study, neither gefitinib nor erlotinib induced chronic lung injury nor did gefitinib exacerbate fibrosis or cause additional lung injury in mice with preexisting fibrosis. The potential use of EGFR inhibitors for treatment of pulmonary fibrosis is complicated by reports of ILD in patients with lung cancer who were treated with gefitinib and erlotinib (33, 50). In >185,000 cancer patients worldwide who have received gefitinib as of September 2004, the reported frequency of ILD is ∼1% (54). The frequency is higher in Japan (∼2%) than in other countries (0.3%). Two studies from Japan identified preexisting pulmonary fibrosis as among the highest risk factors for developing ILD with gefitinib treatment (22, 23). In contrast, in a study of lung cancer patients where erlotinib was used after failure of first or second line chemotherapy, the incidence of pulmonary fibrosis, pulmonary infiltrates, or pneumonitis did not differ between erlotinib-treated and placebo control patients (2, 45). However, in recent case reports, ILD was detected in lung cancer patients treated with erlotinib including a patient with IPF (1, 33, 36, 52). Currently, it is unclear what mechanisms lead to ILD in patients receiving gefitinib or erlotinib. Bleomycin injury is a frequently used experimental method of inducing pulmonary fibrosis in animal models. The benefit of pharmacological EGFR inhibition in reducing bleomycin-induced lung fibrosis is controversial. Suzuki et al. (49) reported that mice receiving intratracheal bleomycin treated with gefitinib had increased fibrosis compared with mice receiving bleomycin alone. In contrast, Ishii et al. (25) reported a protective effect of gefitinib. Differences in mouse strains and doses of gefitinib may have contributed to the different outcomes. Further studies are needed to determine if ILD detected in lung cancer trials is caused by the EGFR inhibition or complicated by radiation or chemotherapy. Considering the increased incidence in patients from Japan, ILD may also be a drug-specific response in a genetically susceptible subset of patients.
Induction of TGF-α caused elevated RVH associated with extensive adventitial thickening surrounding pulmonary vessels. Pulmonary hypertension is a frequent complication of advanced interstitial diseases and is associated with reduced survival (9, 11, 31). In the current study, TGF-α induction of 4 and 8 wk caused decreases in expression of several mRNA transcripts associated with vascular development and function, including VEGF-A, its receptor, KDR, and FOXF1, a transcription factor critical for pulmonary vascular genesis (26). TGF-α increased MDK, SERPINF1, and FZD1 gene expression. Elevated MDK expression in the lung epithelium of transgenic mice caused increased muscularization of small pulmonary arteries and increased smooth muscle actin deposition (41). SERPINF1 is increased in the fibroblastic foci of IPF patients in addition to decreased VEGF-A (3). FZD1 is an antagonist of the canonical Wnt/β-catenin signaling pathway that is critical to several developmental processes, including lung airway and vascular development (46, 60). Attenuation of these alterations in gene expression following gefitinib treatment in association with reductions in RVH (Fig. 6) suggests EGFR signaling induces a number of growth factors, receptors, and transcription factors directly mediating pulmonary fibrosis and pulmonary hypertension. Furthermore, dose-dependent normalization of both transcription changes and total lung collagen with erlotinib underscore that higher doses of EGFR TKI may be necessary to modulate the fibrotic response. This observation is supported by tumor models that demonstrate dose-related antitumor effects with erlotinib (20).
Mice expressing a lung-specific dominant-negative mutation in ERBB3 receptor demonstrate reduced lung fibrosis following intratracheal bleomycin associated with decreased ERBB2 phosphorylation (37). A recent study in mice treated with a monoclonal antibody directed against ERBB2 that blocks ERBB2/ERBB3 signaling also revealed reduced lung fibrosis following bleomycin (8). Studies demonstrate that epithelial tumor cell lines and human tumor xenograft models overexpressing ERBB2 are sensitive to gefitinib and erlotinib (43), and thus ERBB2 and ERBB3 may be involved in the induction of fibrosis in our model. However, we previously demonstrated that overexpression of a dominant-negative mutant EGFR in the distal lung epithelium prevented TGF-α-induced pulmonary fibrosis and pulmonary hypertension, suggesting that EGFR activation is the primary EGFR family member in the fibrotic response (14, 30). Nevertheless, signaling induced by other members of the Erb receptor family may contribute to lung fibrosis, and targeted inhibition of other Erb family members may also be useful in treating lung fibrosis.
In summary, the present study demonstrates that oral EGFR inhibitors prevent and inhibit progression of pulmonary fibrosis due to TGF-α induction of EGFR signaling. Moreover, the use of the inhibitors did not cause inflammatory cell influx or acute lung injury. These findings support further studies to determine the role of EGFR activation in human lung fibrotic disease, which could be amenable to targeted therapy.
This work was supported by National Institutes of Health Grants HL-086598 (W. D. Hardie), HL-077763 (G. D. Leikauf), ES-015675 (G. D. Leikauf), HL-061646 (M. Ikegami), HL-058795 (T. R. Korfhagen), HL-72894 (T. D. Le Cras), HL-90156 (J. A. Whitsett), and HL-61646 (J. A. Whitsett).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society