Article Images (0)
×![]()
featured_play_listSave article to folder
X
Create new folder
Full Text
PDF
Table of Contents
Abstract
ORIGINAL ARTICLE
Introduction
Results
Discussion
Materials And Methods
Mice
Cell Culture
Histology
Epithelial Lining Fluid
Chromatin Immunoprecipitation Assay
Statistical Analysis
Conflict Of Interest
Acknowledgements
Abstract
Laboratory of Cell and Tissue Biology, School of Medicine, Keio University, Tokyo, Japan; Research Institute for Molecular Pathology (IMP), Vienna, Austria; Cancer Cell Biology Programme, Spanish National Cancer Research Centre (CNIO), Madrid, Spain; Department of Pathology, School of Medicine, Keio University, Tokyo, Japan; Department of Surgery, School of Medicine, Keio University, Tokyo, Japan; Department of Medicine, School of Medicine, Keio University, Tokyo, Japan and Department of Orthopaedic Surgery, School of Medicine, Keio University, Tokyo, Japan
ORIGINAL ARTICLE
Interstitial lung disease induced by gefitinib and Toll-like receptor ligands is mediated by Fra-1 Y Takada1, L Gresh2,8, A Bozec3, E Ikeda4,9, K Kamiya5, MWatanabe5, K Kobayashi5, K Asano6, Y Toyama7, EF Wagner3 and K Matsuo1 1Laboratory of Cell and Tissue Biology, School of Medicine, Keio University, Tokyo, Japan; 2Research Institute for Molecular Pathology (IMP), Vienna, Austria; 3Cancer Cell Biology Programme, Spanish National Cancer Research Centre (CNIO), Madrid, Spain; 4Department of Pathology, School of Medicine, Keio University, Tokyo, Japan; 5Department of Surgery, School of Medicine, Keio University, Tokyo, Japan; 6Department of Medicine, School of Medicine, Keio University, Tokyo, Japan and 7Department of Orthopaedic Surgery, School of Medicine, Keio University, Tokyo, Japan The role of the AP-1 transcription factor Fra-1 (encoded by Fosl1) in inflammatory responses associated with lung disease is largely unknown. Here, we show that Fra-1 overexpression in mice reduced proinflammatory cytokine production in response to injection of lipopolysaccharide (LPS), a Toll-like receptor (TLR)-ligand. Unexpectedly, Fra-1 transgenic mice died rapidly following LPS treatment, showing severe interstitial lung disease and displaying massive accumulation of macrophages and overproduction of several chemokines, including macrophage chemoattractant protein-1 (MCP-1, encoded by Ccl2). To assess the clinical relevance of Fra-1 in lung pathology, mice were treated with the anticancer drug gefitinib (Iressa), which can lead to interstitial lung disease in patients. Gefitinib-treated mice showed increased Fosl1 and Ccl2 expression and developed interstitial lung disease in response to LPS, endogenous TLR ligands and chemotherapy. Moreover, deletion of Fra-1 or blocking MCP-1 receptor signaling in mice attenuated gefitinib-enhanced lethality in response to LPS. Importantly, human alveolar macrophages showed enhanced LPS-induced FOSL1 and CCL2 expression after gefitinib treatment. These results indicate that Fra-1 is an important mediator of interstitial lung disease following gefitinib treatment. Oncogene advance online publication, 4 April 2011; doi:10.1038/onc.2011.101 Keywords: Fra-1; gefitinib; interstitial lung disease; MCP-1; AP-1; TLR
Introduction
Inflammation is a host defense mechanism against injury and foreign molecules. Generally, inflammatory responses are self-regulated, although chronic inflammation may cause various diseases, including cancer (Han and Ulevitch, 2005; De Marzo et al., 2007). Activator protein-1 (AP-1) transcription factors exhibit a basic leucine zipper structure, which mediates heterodimerization and DNA binding. Cellular stress, such as bleeding, ultraviolet radiation and genotoxic agents, sequentially induce Fos family AP-1 components, namely, c-Fos, Fra-1, Fra-2 and FosB (Eferl and Wagner, 2003). Fos proteins heterodimerize with a partner Jun protein—c-Jun, JunB or JunD—and bind specific DNA sequences. Overall, AP-1 transcription factors appear to negatively regulate inflammation. For instance, inactivation of both JunB and c-Jun in keratinocytes triggers expression of inflammatory mediators, which recruit neutrophils and macrophages to the epidermis, thereby causing psoriasis in mice (Zenz et al., 2005; Guinea-Viniegra et al., 2009). Similarly, loss of c-Fos in mice results in elevated nuclear factor-kB (NF-kB) activity and increased production of proinflammatory cytokines in response to lipopolysaccharide (LPS) injection (Ray et al., 2006). Furthermore, Fra-1 overexpression in mice suppresses both bone fracture-induced inflammation and dextran sulfate sodium-induced colitis (Yamaguchi et al., 2009; Takada et al., 2010). In murine macrophages, Fra-1 is a negative regulator of proinflammatory cytokine production (Morishita et al., 2009). Collectively, these observations suggest that Fos and Jun proteins can contribute to the resolution of inflammation. The anticancer drug gefitinib (ZD1839, Iressa) was developed to inhibit the epidermal growth factor receptor. However, cellular targets of gefitinib other than epidermal growth factor receptor have been reported (Brehmer et al., 2005). Gefitinib is effective in treating patients with advanced non-small cell lung cancer (Fukuoka et al., 2003; Kris et al., 2003; Maemondo et al., 2010; Pircher et al., 2010). Its primary adverse effects include skin reactions, diarrhea and interstitial lung disease, the last of which may be fatalReceived 28 July 2010; revised 3 March 2011; accepted 4 March 2011 Correspondence: Professor K Matsuo, Laboratory of Cell and Tissue Biology, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo, Tokyo 160-8582, Japan. E-mail: matsuo@sc.itc.keio.ac.jp 8Current address: Biosciences Department, Council for Scientific and Industrial Research (CSIR), Pretoria 0001, South Africa. 9Current address: Department of Pathology, Yamaguchi University Graduate School of Medicine, Yamaguchi 755-8505, Japan. Oncogene (2011) 1–12 & 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc unless properly managed. The incidence of interstitial lung disease in gefitinib-treated patients is approximately 1% (2–4% in East Asia and 0.3% in the United States and Europe) (Cohen et al., 2003; Inoue et al., 2003; Mok et al., 2009; Lee et al., 2010). Currently, little is known about the molecular events underlying gefitinib-enhanced interstitial lung disease. Inflammation is frequently associated with acute and chronic lung disease. In animal models, intra-tracheal or intraperitoneal (i.p.) LPS administration is widely used to induce experimental acute lung injury (Ulich et al., 1991). LPS stimulates Toll-like receptor (TLR) 4, thereby inducing inflammatory mediators through activation of the transcription factor NF-kB (Kawai and Akira, 2010). Here, we report that Fra-1 is a pivotal mediator of interstitial lung disease in gefitinib-enhanced lung injury models. Transcriptional activation of Fosl1 (encoding Fra-1) and the Fra-1 target gene Ccl2 (encoding macrophage chemoattractant protein (MCP)-1) resulted in increased lung pathology and lethality in mice. Moreover, Fra-1 deletion or MCP-1 receptor inhibition attenuated gefitinib-enhanced interstitial lung disease. Our data suggest that combined activity of gefitinib and TLR ligands triggers Fra-1 and chemokine production, leading to interstitial lung disease in patients treated with gefitinib.
Results
Fra-1 enhances LPS-induced interstitial lung disease Fra-1 transgenic mice (Fra-1 mice) express Fra-1 under control of the major histocompatibility complex class I (H2-Kb) promoter in several tissues, including lung (Jochum et al., 2000). Given the reduced production of major proinflammatory cytokines seen in Fra-1 mice (Yamaguchi et al., 2009; Takada et al., 2010), we asked whether Fra-1 mice might be resistant to inflammatory responses induced by LPS. Surprisingly, when Fra-1 mice and littermate wild-type controls were injected i.p. with 50mg/kg LPS, all Fra-1 mice died within 38 h, whereas wild-type controls survived over a period of 110 h (Figure 1a). Histological analyses of vital organs, including heart, lung, liver, kidney and brain, 12 h after LPS injection revealed more severe LPS-induced lung injury (Figure 1b), a comparable number of Gr-1positive polymorphonuclear leukocytes (Figure 1c), and approximately fivefold increase (458±159%, Po0.05) in the number of CD11b-positive macrophages in the alveolar septa in Fra-1 mice compared with wild-type controls (Figure 1d). These dramatically increased CD11b-positive cells were also Fra-1-positive, and most Fra-1-expressing cells in the lung were macrophages in Fra-1 mice (Supplementary Figure S1a). The wet/dry weight ratio of lung tissue after LPS injection was significantly greater for Fra-1 mice compared with wild-type controls, indicating that Fra-1 overexpression worsens LPS-induced pneumonia (Figure 1e). The histopathology pattern was reminiscent of nonspecific interstitial pneumonia. The fact that abundant macrophages filled alveolar septa rather than alveolar spaces ruled out desquamous interstitial pneumonia (Katzenstein and Myers, 1998). Thus, LPS-induced changes in the lungs of Fra-1 mice share histological features with gefitinib-induced interstitial pneumonia in humans. As inflammatory responses, particularly macrophage recruitment, were more severe in lungs of Fra-1 mice compared with controls, we measured lung expression levels of cytokines and chemokines 12 h after LPS injection by quantitative reverse transcriptase–PCR. LPS-induced levels of proinflammatory cytokines, such as Tnfa (encoding tumor necrosis factor (TNF)-a), Il1b (encoding interleukin (IL)-1b), Il6 and Il10, were consistently lower in the lungs of Fra-1 mice than in those of wild-type controls, as were levels of chemokines Ccl3, Ccl4, Ccl7 and Ccl12, and of AP-1 components, including Jun (Figure 1f). By contrast, levels of LPSinduced transcripts of Ccl2 (encoding MCP-1) were higher in Fra-1 mice than in wild-type controls. Ccl6, Ccl8 (encoding MCP-2) and Cxcl5 transcript levels were also slightly elevated in Fra-1 mice versus wild-type mice (Figure 1f). Therefore, induction of chemokine genes such as Ccl2 by Fra-1 might be responsible for severe lung pathology seen in Fra-1 mice injected with LPS. Untreated Fra-1 and wild-type mice showed Fra-1/Wt ratios for various genes largely similar to those seen in the presence of LPS (Supplementary Figure S1b). These data suggest that LPS signaling acts primarily to exacerbate pre-existing lung pathology in untreated Fra-1 mice. As Fra-1 and other Fos proteins negatively regulate NF-kB (Takada et al., 2010), an important regulator of cell survival and inflammation (Aggarwal, 2003), we analyzed nuclear localization of NF-kB and apoptosis of lung cells. As expected, Fra-1 expression suppressed nuclear localization of the NF-kB subunit p65 (Supplementary Figure S1c) and enhanced apoptosis in the lung as demonstrated by terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) analysis (Supplementary Figure S1d). These data suggest that NF-kB activation on LPS injection is inhibited in Fra-1-expressing macrophages in the lung, which in turn enhances apoptosis. Gefitinib enhances LPS-induced interstitial lung disease We next asked whether the lung pathology seen in Fra-1 mice is related to interstitial lung disease observed in patients treated with gefitinib (Cohen et al., 2003; Inoue et al., 2003). To do so, we administered 250mg/kg gefitinib or water orally to wild-type mice 6 days a week for 2 weeks, and then injected LPS (50mg/kg) i.p. into both groups. Without LPS injection, gefitinib-treated mice are indistinguishable from mock-treated mice. After LPS injection, gefitinib-treated mice began to die at 6 h and all died within 20 h, while none of the mocktreated controls died before 15 h (Figure 2a). Histological analysis 12 h after LPS injection showed that compared with mock-treated mice, gefitinib-treated mice Oncogene developed more severe lung injury, which was characterized by thickened alveolar septa and accumulation of infiltrating cells (Figure 2b). Although accumulation of Gr-1-positive polymorphonuclear leukocytes was not significantly elevated, CD11b-positive macrophages more abundantly infiltrated alveolar septa of gefitinib-treated mice by approximately 1.5-fold (CD11b, 154±20%, Po0.05) compared with mocktreated mice following LPS treatment (Figures 2c and d). Therefore, LPS-induced lung injury in gefitinibtreated mice was similar to that seen in LPS-induced Fra-1 mice. We further quantified cytokine and chemokine mRNA levels in lung tissues of gefitinib- or mock-treated mice Oncogene 12h after LPS injection. Expression of proinflammatory cytokines, such as IL-12 p40, was significantly lower in gefitinib-treated than in mock-treated mice (Figure 2e, Supplementary Figure S2a). Furthermore, expression of Ccl2, Ccl6 and Ccl8, but not Ccl5 and other chemokines, was higher in gefitinib-treated mice (Figure 2e, Supplementary Figure S2b). Thus, the LPS-induced expression patterns of inflammatory cytokines and chemokines in gefitinib-treated wild-type mice were also similar to those seen in LPS-induced Fra-1 mice. Interestingly, LPS-induced expression of Fosl1 (encoding Fra-1) was significantly higher in gefitinib-treated mice than in the mock controls, while expression levels of the other six AP-1 components, including Jun, were unchanged or lower in gefitinib-treated mice (Figure 2e, Supplementary Figure S2c). Gefitinib treatment alone in the absence of LPS did not induce Fosl1 expression (Supplementary Figure S2c). These data suggest that Fra1 might mediate the effects of gefitinib on Tnfa and Ccl2 expression in response to LPS. Gefitinib modulates LPS-induced Tnfa and Ccl2 expression through Fra-1 To further dissect the molecular pathways downstream of gefitinib treatment or Fra-1, we used primary cell 0 20 40 60 80 100 0 Time after LPS injection (h) P er ce nt s ur vi va l Mock Gef + LPS Mock + LPS Gef Anti-CD11b Cont. M oc k G ef H&E Cont. M oc k G ef Anti-Gr-1 M oc k G ef Cont. LP S -in du ce d m R N A s in lu ng (G ef - / M oc k- tr ea te d ra tio ) T nf a Il1 b Il6 Il1 2 0 1 2 3 Cytokines ** Chemokines C cl 2 C cl 3 C cl 4 C cl 5 C cl 6 C cl 7 C cl 8 C cl 12 C xc l1 C xc l2 C xc l5 C xc l1 0 * * ** F os F os b F os l1 F os l2 Ju n Ju nb Ju nd 0 1 2 3 * * * AP-1 components § 10 20 30 40
LPS
LPS 0 1 2 3 * * Figure 2 Effects of gefitinib treatment on LPS-induced interstitial lung disease. (a) Survival curves of gefitinib-treated mice after LPS injection. C57BL/6J mice (n¼ 6) were pretreated with 250mg/kg gefitinib (Gef) or water as a control (Mock) orally once a day for 6 days a week for 2 weeks and then injected i.p. with 50mg/kg LPS. Survival of ‘GefþLPS’ mice was significantly worse than that of ‘MockþLPS’ mice (yP¼ 0.045). (b) Hematoxylin and eosin staining of lung tissue 12 h after 50mg/kg LPS i.p. injection. Scale bars, 100mm. (c, d) Detection of Gr-1- and CD11b-positive cells in lung 12 h after i.p. 50mg/kg LPS injection. Scale bars, 50 mm. Arrowheads indicate Gr-1-positive cells. (e) Quantitative reverse transcriptase (qRT)–PCR analysis of genes encoding inflammatory cytokines, chemokines and AP-1 components in lung. Data are presented as in Figure 1f. See also Supplementary Figure S2. *Po0.05, **Po0.01. Oncogene cultures. Primary mouse lung fibroblasts and alveolar macrophages were prepared and treated with gefitinib for 30min before LPS addition. Time-course analysis of lung fibroblasts revealed that LPS-induced Tnfa expression was suppressed, whereas Ccl2 and Fosl1 induction was enhanced in gefitinib-treated cultures compared with mock-treated cultures (Figure 3a). Alveolar macrophages treated with gefitinib also showed suppressed Tnfa expression, along with enhanced Ccl2 and Fosl1 expression in response to LPS (Figure 3b). We next determined whether gefitinib-treatment alters Tnfa, Ccl2 or Fosl1 promoter responses to LPS in Oncogene cultured lung fibroblasts. A Tnfa promoter-luciferase construct was activated following LPS addition, and gefitinib pretreatment of cells suppressed this activation (Figure 3c). By contrast, in similar assays using a Ccl2luciferase construct, gefitinib treatment dramatically enhanced LPS-induced Ccl2 promoter activity (Figure 3c). Furthermore, a promoter-luciferase construct containing the wild-type Fra-1 regulatory element found in Fosl1 intron 1 (Bergers et al., 1995) was upregulated by gefitinib treatment, while a similar construct with a mutated regulatory element showed no response (Figure 3c). These data demonstrate that promoter and regulatory elements of Tnfa, Ccl2 and Fosl1 mediate the gefitinib effect on LPS-induced gene expression. We next co-transfected primary lung fibroblasts with a Fra-1 expression vector and either the Tnfa or Ccl2 promoter-luciferase constructs. Exogenous Fra-1 suppressed and enhanced LPS-induced Tnfa and Ccl2 promoter activities, respectively (Figure 3d). Chromatin immunoprecipitation assays revealed a gefitinib-enhanced sequential recruitment of Fra-1 onto the B4, B2 and B1 putative AP-1-binding sites in the Ccl2 promoter, each with distinct time course, after LPS addition, suggesting that Fra-1 directly binds and activates the Ccl2 promoter in response to gefitinib (Figure 3e). MCP-1 is overproduced in human alveolar fluid or macrophages treated with gefitinib To determine whether MCP-1 (CCL2) induction occurs in human patients treated with gefitinib, we measured cytokine levels in the pulmonary epithelial lining fluid (ELF) of lung cancer patients before and after gefitinib treatment. ELFs were obtained by bronchoscopic microsampling. Although vascular endothelial growth factor and IL-8 levels did not change significantly after gefitinib administration, MCP-1 showed a 2.3-fold elevation (Figure 4a). We next isolated human primary alveolar macrophages from bronchoalveolar lavage fluids, and pretreated them with gefitinib or the solvent dimethylsulphoxide. These cells were further treated with LPS or water. Gefitinib treatment reduced TNFA and increased CCL2 and FOSL1 expression induced by LPS (Figure 4b). These data suggest that MCP-1 production can be enhanced by gefitinib treatment also in human cells. Induction of Fosl1 by combining gefitinib with other anticancer drugs As gefitinib-induced lung injury in cancer patients usually occurs in the absence of LPS, we asked what reagents or treatments could synergize with gefitinib to cause lung injury. We first tested the combination of gefitinib and bleomycin, since this combination has been tested in mouse models (Suzuki et al., 2003; Ishii et al., 2006). We found that levels of both Ccl2 and Fosl1 in the lungs of mice treated with gefitinib and bleomycin were upregulated (Figure 4c). We then analyzed the effect of gefitinib in human lung tumor-xenografted mice treated with the chemotherapeutic agent gemcitabine, which has been used clinically to treat lung tumors in combination with gefitinib (Giaccone et al., 2004). Ccl2 and Fosl1 expression was significantly elevated in lung tissues of tumor-xenografted mice treated with both gemcitabine and gefitinib (Figure 4d). Further histological study revealed that these tissues exhibited interstitial lung disease (Figure 4e). Therefore, there are anticancer drugs that can induce Ccl2 and Fosl1 expression in combination with gefitinib treatment. Tumors damaged during chemotherapy may release endogenous TLR-ligands such as heparan sulfate, lowmolecular weight hyaluronan and fibronectin, and activate TLRs under sterile conditions (Johnson et al., 2004; Jiang et al., 2005; Mollen et al., 2006). We therefore asked whether heparan sulfate could substitute for LPS in our mouse model. Similarly to LPS, heparan sulfate enhanced Ccl2 and Fosl1 and suppressed Tnfa expression in gefitinib-treated primary bone marrow macrophages (Supplementary Figure S3a). We also found that intravenous administration of low-molecular weight hyaluronan enhanced Ccl2 and Fosl1 expression in the lung and caused interstitial lung disease in gefitinibtreated mice (Supplementary Figure S3b and c). These data suggest that endogenous TLR ligands may contribute to gefitinib-induced interstitial lung disease in lung cancer patients. Attenuation of interstitial lung disease by Fra-1 deletion We next asked whether gefitinib-induced interstitial lung disease could be blocked in the absence of Fra-1. As Fra-1-knockout mice die during gestation from placental defects, we generated conditional Fra-1 knockout mice (Fosl1D/D) by crossing Fra-1-floxed mice (Fosl1flox/flox) with a deleter strain (MORE-Cre), which deletes Fosl1 in the embryo proper but not in placental tissues (Eferl et al., 2004). We administered 250mg/kg gefitinib orally to Fosl1D/D mice and littermate controls (Fosl1flox/flox) 6 days a week for 2 weeks, followed by LPS injection (50mg/kg). Both Fosl1D/D mice and littermate controls survived >12 days, but most controls died within 16 days of injection, a period during which most Fosl1D/D mice were unaffected (Figure 5a). Histological analysis revealed a total suppression of interstitial lung disease and a significant reduction of macrophage infiltration by >40% 12 h after LPS injection, without affecting number of polymorphonuclear leukocytes, in Fosl1D/D mice compared with littermate controls, both treated with gefitinib and LPS (Figures 5b–d). Consistently, gefitinib did not enhance LPS-induced Ccl2 expression in the lung of Fosl1D/D mice (Figure 5e). Importantly, in the absence of gefitinib, Fra-1 deletion did not alter cytokine and chemokine expression or lung histology in response to LPS injection (Supplementary Figure S4) indicating that Fra-1 mediates gefitinib effects. We performed additional loss-of-function experiments using lung fibroblasts. Fra-1 knockdown via small interfering RNA cancelled both gefitinib-induced suppression of Tnfa and gefitinib-induced induction of Ccl2 in response to LPS (Figure 5f). These data clearly indicate that Fra-1 is responsible for gefitinib-induced interstitial lung disease triggered by TLR signaling. Oncogene
Bleo
Oncogene Anti-Gr-1 H&E flox/flox + Gef Δ/Δ + Gef 0 20 40 60 80 100 0 Time after LPS injection (d) P er ce nt s ur vi va l Δ/Δ + Gef flox/flox + Gef § 0 0.02 0.04 0.06 F os l1 LPS Gef siFosl1 siFosl2 siCont. 0 0.02 0.04 0.06 C cl 2 0 1 2 3 T nf a 0 0.01 0.02 F os l2 * * * * 0 20 40 60 80 100 0 Mock Gef RS + Gef Time after LPS injection (h) P er ce nt s ur vi va l § LP S -in du ce d m R N A s in lu ng (Δ /Δ + G ef / flo x/ flo x + G ef r at io ) T nf a Il1 b Il6 Il1 2 C cl 2 C cl 3 C cl 4 C cl 5 C cl 6 C cl 7 C cl 8 C cl 12 C xc l1 C xc l2 C xc l5 C xc l1 0 F os F os b F os l1 F os l2 Ju n Ju nb Ju nd 0 2 4 6 8 1 ** ‡ Anti-CD11b flox/flox + Gef Δ/Δ + Gef flox/flox + Gef Δ/Δ + Gef 5 10 15 20 - + + + + + + + + + + + + - - - - - - - - - - -- - -- - - - #1 #1#2 #2 LPS Gef siFosl1 siFosl2 siCont. - + + + + + + + + + + + + - - - - - - - - - - -- - -- - - - #1 #1#2 #2 LPS Gef siFosl1 siFosl2 siCont. - + + + + + + + + + + + + - - - - - - - - - - -- - -- - - - #1 #1#2 #2 LPS Gef siFosl1 siFosl2 siCont. - + + + + + + + + + + + + - - - - - - - - - - -- - -- - - - #1 #1#2 #2 10 20 30 40 50 Figure 5 LPS responses in mice lacking Fra-1 or MCP-1 signaling in vivo. (a) Survival curves of gefitinib-pretreated Fosl1D/D (D/D, n¼ 6) and Fosl1flox/flox (flox/flox, n¼ 5) mice on a C57BL/6J 129 mixed background after LPS injection. Mice were treated with gefitinib and LPS as in Figure 2. Fosl1D/D mice survived significantly longer than control Fosl1flox/flox mice (yP¼ 0.036). (b–d) Lung histology 12 h after LPS injection. hematoxylin and eosin (H&E) (b), Gr-1 (c) and CD11b (d) stainings. Scale bars: 100mm for H&E, 50 mm for Gr-1 and CD11b staining. Arrowheads indicate Gr-1-positive cells. (e) Quantitative reverse transcriptase (qRT)–PCR analysis of genes encoding inflammatory cytokines, chemokines and AP-1 components in lung. Data are presented as in Figure 1f. **Po0.01 and zP¼ 0.0815. (f) Knockdown of Fosl1 and Fosl2 using small interfering RNA (siRNA) in primary lung fibroblasts. Cells were transfected overnight with each siRNA and stimulated with 30 mM gefitinib for 30min and 0.1 mg/ml LPS for another 12 h. Four different siRNAs were tested and Fosl1 #1 and Fosl2 #1 siRNAs most efficiently knocked down their respective target. *Po0.05. (g) Attenuation of LPS-induced lethality with the CCR2 antagonist RS 504393 (RS). Wild-type mice (n¼ 8 per group) were pretreated with gefitinib and injected i.p. with 2mg/kg RS. Survival of ‘RSþGef’ mice was significantly greater than that of ‘Gef’ mice (yP¼ 0.0086). Oncogene Unlike Fra-1 knockdown, Fra-2 knockdown did not alter gefitinib/LPS-induced levels of Tnfa and Ccl2 expression (Figure 5f). Furthermore, Fra-2 overexpression in lung fibroblasts did not suppress LPS-induced Tnfa expression or enhance Ccl2 expression to the levels mediated by Fra-1 overexpression (Supplementary Figure S5a). It is also important to note that gefitinib/ LPS treatment induced Fosl1 but not Fosl2 (Figure 2e). These data collectively suggest that Fra-1 but not Fra-2 mediates effects of gefitinib on Tnfa and Ccl2 expression in the lung. We also examined the possible role for c-Fos in gefitinib/LPS-regulated expression of Tnfa and Ccl2, and found that loss of c-Fos abolished reduction in Tnfa and increase in Ccl2 expression (Supplementary Figure S5b). Even in the absence of c-Fos, Fosl1 induction by gefitinib/LPS was intact (Supplementary Figure S5b). Therefore, although Fosl1 is the major target of c-Fos in the osteoclast lineage (Matsuo et al., 2000), Fosl1 is unlikely to be a major c-Fos target gene in the context of lung pathology. Attenuated lethality following MCP-1 inhibition MCP-1 is a chemoattractant for macrophages, lymphocytes and natural killer cells (Rollins, 1997). In the lung, MCP-1 is produced by type II alveolar epithelial cells, macrophages and fibroblasts (Paine et al., 1993). As high MCP-1 production may underlie Fra-1-mediated interstitial lung disease, we asked whether inhibiting MCP-1 receptor signaling could counteract LPSinduced lethality in gefitinib-treated mice. The chemokine (C-C motif) receptor 2 (CCR2) reportedly interacts with MCP-1 (Ccl2) and MCP-2 (Ccl8). We found that the CCR2 antagonist RS 504393 efficiently delayed gefitinib-induced death in LPS-injected wild-type mice (Figure 5g). Furthermore, i.p. injection of an antiMCP-1 antibody 2 h before LPS injection delayed LPSinduced death of Fra-1 mice (Supplementary Figure S6). These results show that overproduction of MCP-1 in response to TLR signaling is, at least in part, causative of interstitial lung disease in Fra-1 and gefitinibtreated mice.
Discussion
Interstitial pneumonia emerged as a serious and unexpected side effect when gefitinib was first introduced as an anticancer drug (Cohen et al., 2003; Inoue et al., 2003). We found that either Fra-1 overexpression or gefitinib-treatment in mice enhanced interstitial lung disease in response to TLR signaling. Histological examination of both Fra-1 mice and gefitinib-treated mice revealed massive infiltration of Fra-1- and CD11bpositive macrophages into alveolar septa after LPS injection. In both groups of mice, LPS-induced expression of specific chemokines, including Ccl2 (encoding MCP-1), Ccl6 and Ccl8, was elevated in lung. By contrast, LPS-induced expression levels of most inflammatory cytokines, such as Tnfa, were lower in Fra1 mice or gefitinib-treated mice than in non-transgenic or mock-treated control mice. Therefore, Fra-1 overexpression and gefitinib treatment have strikingly similar effects on lung histology and gene expression when combined with LPS. Furthermore, endogenous Fosl1 expression was prominently induced by gefitinib in the presence of TLR signaling, and deletion of Fosl1 in mice prevented gefitinib-enhanced lung pathology in a LPSinduced lung injury model. These data suggest that Fra-1 critically mediates gefitinib’s adverse effect on the lung. Fra-1 has been shown to be a negative regulator of proinflammatory cytokine production in a murine macrophage cell line and in mouse models of colitis and bone fracture, presumably through inhibiting NFkB activity (Morishita et al., 2009; Yamaguchi et al., 2009; Takada et al., 2010). Overexpression of the closely related Fos family member Fra-2 (encoded by Fosl2) induces pulmonary fibrosis (Eferl et al., 2008). Therefore, both Fra-1 and Fra-2 may enhance pulmonary fibrosis through similar mechanisms. However, it is noteworthy that we found that gefitinib together with LPS induced Fosl1 but not Fosl2 expression, suggesting that Fra-1 specifically mediates gefitinib’s adverse effects in the lung. The molecular basis for overlapping and specific functions of Fra-1 and Fra-2 is yet to be discovered. This also holds true for their roles in bone metabolism, where both Fra-1 and Fra-2 critically regulate bone cells such as osteoclasts and osteoblasts (Jochum et al., 2000; Matsuo et al., 2000; Bakiri et al., 2007; Bozec et al., 2008, 2010). This study reveals that proinflammatory cytokines such as Tnfa and a subset of chemokines such as Ccl2, Ccl6 and Ccl8 are differentially regulated by Fra-1/ AP-1. This differential regulation was unexpected because Ccl2 is in many cases co-regulated with Tnfa and Il12 (Salojin et al., 2006). Involvement of AP-1 family members including Fra-1 and c-Jun in Ccl2 transcriptional regulation was previously suggested (Finzer et al., 2000; Wolter et al., 2008). Here, using lung fibroblasts and macrophages, we show that Fra-1 upregulates the Ccl2 promoter in a reporter assay. Furthermore, chromatin immunoprecipitation experiments demonstrated that Fra-1 directly binds to the Ccl2 promoter, suggesting that Ccl2 is a direct Fra-1 target. Excess MCP-1 production accounts for massive infiltration of macrophages and lethal tissue damage in lung, and administration of a CCR2 antagonist RS 504393 or an anti-MCP-1 antibody attenuated gefitinibenhanced lung pathology in LPS-injected mice. Several other chemokines, such as Ccl6, Ccl8 and Cxcl5, were also elevated by Fra-1 or gefitinib in the presence of TLR signaling. Therefore, inhibition of all of these chemokines might protect mice more efficiently from lethality than inhibition of MCP-1 signaling alone. Since gefitinib treatment in human patients increased MCP-1 levels in lung ELF, monitoring MCP-1 levels in lung ELF might be clinically useful to evaluate patient sensitivity to gefitinib-enhanced interstitial lung disease, and subsequent administration of chemokine inhibitors might be of therapeutic relevance. Besides upregulation of Ccl2, Ccl6, Ccl8 and Cxcl5, we observed marked Oncogene downregulation of Il6 and Cxcl10 in LPS-treated Fra-1 mice. Conversely, in Fra-1-deficient mice, both Il6 and Cxcl10 tended to be upregulated in response to LPS. As Fra-1 negatively regulates NF-kB, both Il6 and Cxcl10 expression may be driven by enhanced NF-kB activity in Fra-1-deficient mice. In addition to LPS, endogenous ligands, such as heparan sulfate, hyaluronan and fibronectin, can be released from damaged tissue and activate TLR under sterile conditions (Johnson et al., 2004; Jiang et al., 2005; Mollen et al., 2006; Vogl et al., 2007). We found that not only injection of low-molecular weight hyaluronan, but also bleomycin administration, and tumor xenograft combined with gemcitabine treatment could substitute for the role of LPS in enhancing gefitinibinduced lung injury. Genetic deletion of TLR4 or genes encoding its adapter molecules should reveal whether TLR4 functions in chemotherapy-induced lung injury. It is also noteworthy that cigarette smoking reportedly induces Fra-1 expression (Zhang et al., 2006) and could also promote interstitial lung disease. An important finding of this study is that gefitinib treatment enhanced LPS-induced expression of Fosl1 and Ccl2. Brehmer et al. (2005) identified >20 gefitinib target proteins in vitro using affinity purification with resin-bound AX14596, a gefitinib derivative. AX14596 binds not only to epidermal growth factor receptor but also to other proteins, such as breast tumor kinase (also known as protein tyrosine kinase 6; a nonreceptor tyrosine kinase expressed in most breast cancers), c-Yes tyrosine kinase, receptor-interacting serine-threonine kinase, and cyclin G-associated kinase (Brehmer et al., 2005). We speculate that gefitinib-regulated kinases might induce Fosl1 in the presence of TLR signaling. We have attempted to identify them by affinity purification using gefitinib-bound nanobeads. Since the protocol requires a large volume of cell extracts, we have searched for immortalized cell lines that respond to LPS/gefitinib treatment. Further studies should explore how gefitinib induces and activates Fra-1. In conclusion, these data reveal a molecular mechanism underlying gefitinib-induced lung disease. Namely, Fra-1 induction and subsequent transcriptional activation of chemokines, including MCP-1, mediate the effect of gefitinib in the presence of TLR signaling, enhancing interstitial pneumonia. Inhibitors of Fra-1, MCP-1 or TLR signaling could be beneficial in reducing interstitial lung disease in lung cancer patients undergoing gefitinib treatment.
Materials and methods
Mice
Six-week-old female C57BL/6J, BALB/cnu/nu and ICR mice were purchased from Clea Japan (Tokyo, Japan). Fra-1 mice (Jochum et al., 2000) were backcrossed to C57BL/6J mice for more than seven generations, and wild-type, sex-matched littermates were used as controls. Fra-1-conditional knockout mice on a C57BL/6J 129 mixed background were produced by crossing Fosl1flox/flox mice with MORE-Cre mice (Tallquist and Soriano, 2000). All animal experiments were conducted in accordance with institutional review boardapproved protocols. Gefitinib- and LPS-induced lung disease model Gefitinib (AstraZeneca, London, UK) tablets were ground into powder and suspended in water. For 2 weeks before LPS injection, 250mg/kg gefitinib or water was orally administered once a day for 6 days a week using a feeding needle. were injected i.p. with LPS (S Minnesota Re595; Sigma, St Louis, MO, USA) immediately before the start of the dark period of their circadian rhythms for survival experiments. Two mg/kg CCR2 antagonist RS 504393 (TOCRIS Bioscience, Ellisville, MO, USA) was injected i.p. into gefitinib-pretreated mice four times at 12-h intervals before LPS injection. Chemotherapy-induced lung disease models ICR mice were anesthetized, administered 5mg/kg bleomycin (Nippon Kayaku, Tokyo, Japan) intratracheally, and treated with gefitinib as described (Suzuki et al., 2003). Briefly, 250mg/kg gefitinib was given orally 1 h before bleomycin administration, and 6 days a week for the following 3 weeks. To study the effect of gefitinib in a human lung tumor xenograft model, BALB/cnu/nu mice were injected via the tail vein with 1 106 cells per mouse of human lung tumor A549 cells (ATCC, Manassas, VA, USA). Four weeks later, mice were injected i.p. with 200mg/kg gemcitabine (Eli Lilly, Indianapolis, IN, USA). After a week, mice were treated with 250mg/kg gefitinib orally once a day for additional week. Quantitative reverse transcriptase–PCR and luciferase assay Quantitative reverse transcriptase–PCR was performed as described (Zhao et al., 2006). A Tnfa-luciferase reporter was constructed by inserting a PCR-amplified mouse Tnfa promoter fragment including the entire 50-untranslated region ( 577 to þ 3 relative to the ATG codon) into pGL3 (Promega, Madison, WI, USA). A reporter containing a 1.2-kb mouse Ccl2 promoter fragment was a gift from Dr A Dent (Toney et al., 2000). The Fosl1-luciferase reporter, wt-Luc (tk-luc-B) containing three tandem AP-1-binding sites of the mouse Fosl1 intron 1 fragment (57 bases), the mutant reporter mt-Luc (tk-luc-Bm), and the cytomegalovirus-Fra-1 expression vector pRK7-Fra-1 were all gifts from Dr M Busslinger (Bergers et al., 1995). Lung fibroblasts were transfected using FuGENE 6 (Roche, Basel, Switzerland) each with a firefly luciferase construct, pB-actin-RL (Promega) and, when indicated, cytomegalovirus-Fra-1 expression vector or empty vector. After 48 h, cells were treated with 30mM gefitinib for 30min, after which 0.1 mg/ml LPS was added to the medium, followed by incubation for an additional 48 h. Luciferase activity was measured using a dual-luciferase assay system and a luminometer (MicroLumatPlus LB96V; Berthold, Bad Wildbad, Germany). The values are firefly luciferase activities normalized to renilla luciferase activities.
Cell culture
Primary mouse alveolar and bone marrow macrophages were prepared as described (Peters-Golden and Thebert, 1987; Ray et al., 2006). Primary mouse lung fibroblasts were prepared as described (Kida et al., 2005). Human alveolar macrophages were obtained under approved ethical protocols from the School of Medicine, Keio University, by isolating adherent cells from bronchoalveolar lavage fluids after culturing cells overnight in the presence of 50mg/ml gentamicin Oncogene (Sigma). For small interfering RNA studies, small interfering RNAs for control (silencer select negative control), Fosl1 #1 (s66203), Fosl1 #2 (s201360), Fosl2 #1 (s66206) and Fosl2 #2 (s66208) were purchased (Ambion, Austin, TX, USA). Each 10 pmol small interfering RNA was transfected using Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA, USA) into 2 105 primary lung fibroblasts. Gefitinib was dissolved in dimethylsulphoxide and added to the medium.
Histology
Paraffin sections were soaked in 1% H2O2/methanol and treated with 100mg/ml proteinase K (Takara Bio, Shiga, Japan). After blocking with BlockAce (Dainihon Pharmaceutical, Tokyo, Japan), either rat anti-mouse Gr-1 (e-Bioscience, San Diego, CA, USA), rat anti-mouse CD11b-FITC (BD Biosciences, San Jose, CA, USA), mouse anti-mouse Fra-1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) or rabbit anti-mouse p65 (Abcam, Cambridge, MA, USA) antibodies were applied overnight at 4 1C. For Gr-1, anti-rat biotinylated antibody (BD Biosciences) served as a secondary antibody and was detected using a DAB kit (Dako, Glostrup, Denmark). For Fra-1 and p65, anti-mouse or rabbit Alexa Fluor 647 (BD Biosciences) served as a secondary antibody. For CD11b, Fra-1 and p65 analyses, slides were mounted with Vectashield containing DAPI (Vector Laboratories, Burlingame, CA, USA) and analyzed by confocal microscopy (LSM 510 META; Carl Zeiss, Oberkochen, Germany). Sections were also stained with hematoxylin and eosin. CD11b-positive cells were quantified using Image J software (NIH, Bethesda, MD, USA) using a fixed threshold in each experiment. The TUNEL assay was performed using a DeadEnd colorimetric TUNEL system kit (Promega).
Epithelial lining fluid
ELF of human lung was obtained from lung cancer patients using approved ethical protocols from the School of Medicine, Keio University, using a bronchoscopic microsampling method from both right and left sides before and after gefitinib treatment, as described (Ishizaka et al., 2001; Watanabe et al., 2003). Patients did not show apparent interstitial lung disease, and their tumors did not respond to gefitinib treatment. Protein levels of MCP-1, vascular endothelial growth factor and IL-8 were measured using enzyme-linked immunosorbent assay sets (MCP-1 and vascular endothelial growth factor: R&D Systems, Minneapolis, MN, USA, IL-8: Invitrogen). The institutional review board approved these studies, and informed consent was obtained from each patient.
Chromatin immunoprecipitation assay
Cells were fixed, and chromatin was precipitated using an anti-Fra-1 antibody (Santa Cruz Biotechnology) as described (Takada et al., 2004). Putative AP-1-binding sites were quantified by quantitative PCR using the following Ccl2 promoter primer pairs for putative AP-1-binding sites 1 to 4 (B1 to B4): 50-CGAGGGCTCTGCACTTACTC-30 and 50-TCTGGCTCT CTGCACTTCTG-30; 50-CCACCAAGTGGAGAGAATGC-30 and 50-GGCCAAGGAACCTAAAGTCC-30; 50-CATTCCAG TTGGCTCACTCA-30 and 50-TGGCTATCATCACATTACC TTCA-30; and 50-CCTTTCCCTTGGCTGCTC-30 and 50-GA AGTTCCCAGACCCTTCG-30, respectively.
Statistical analysis
Statistical significance was determined using an unpaired twotailed Student’s t-test or, where indicated, the two-sided Wilcoxon’s signed rank test. For survival curves, the two groups were compared using a log-rank test (http://bioinf. wehi.edu.au).
Conflict of interest
The authors declare no conflict of interest.
Acknowledgements
We thank A Oide, M Asakawa and S Fukuse for technical support; T Yamaguchi for Fra-1 mouse maintenance; A Dent, A Ueda and M Busslinger for plasmids; H Daub for AX14596; and H Handa, S Sakamoto, K Maruyama, T Mitsudomi and L Bakiri for helpful input. We thank the Core Instrumentation Facility, Keio University School of Medicine, for technical assistance. This work was funded by a grant-in-aid from the global COE program, a grant-in-aid for Scientific Research (B) from MEXT of Japan (21390425), and a grant-in-aid for Scientific Research on Priority Areas and the Naito Foundation (YT). LG was funded by an EMBO Long-Term Fellowship. IMP is funded by Boehringer Ingelheim.
Article Images (0)
×![]()
Bioz is the world’s most comprehensive AI search engine for scientific experimentation. The Bioz search engine offers researchers billions of data-driven product, technique, and protocol recommendations. Bioz taps into the latest advances in Artificial Intelligence (AI) – including Natural Language Processing (NLP), Machine Learning (ML) and Generative AI to mine and structure hundreds of millions of pages of complex and unstructured scientific papers. By harnessing the power of AI, Bioz provides researchers with an unprecedented amount of summarized scientific experimentation knowledge right at their fingertips, ultimately helping to speed up drug discovery and the rate of success in finding cures for diseases. Bioz is used by over 15 Million researchers from over 17,000 different universities and companies in 196 countries.
