If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Division of Cancer and Cell Biology, The Translational Genomics Research Institute, Phoenix, ArizonaVirginia G. Piper Cancer Center Clinical Trials at Scottsdale Healthcare, Mayo Clinic Arizona, Scottsdale, Arizona
Lung cancer is the leading cause of cancer deaths worldwide; approximately 85% of these cancers are non-small cell lung cancer (NSCLC). Patients with NSCLC frequently have tumors harboring somatic mutations in the epidermal growth factor receptor (EGFR) gene that cause constitutive receptor activation. These patients have the best clinical response to EGFR tyrosine kinase inhibitors (TKIs). Herein, we show that fibroblast growth factor–inducible 14 (Fn14; TNFRSF12A) is frequently overexpressed in NSCLC tumors, and Fn14 levels correlate with p-EGFR expression. We also report that NSCLC cell lines that contain EGFR-activating mutations show high levels of Fn14 protein expression. EGFR TKI treatment of EGFR-mutant HCC827 cells decreased Fn14 protein levels, whereas EGF stimulation of EGFR wild-type A549 cells transiently increased Fn14 expression. Furthermore, Fn14 is highly expressed in EGFR-mutant H1975 cells that also contain an EGFR TKI-resistance mutation, and high TKI doses are necessary to reduce Fn14 levels. Constructs encoding EGFRs with activating mutations induced Fn14 expression when expressed in rat lung epithelial cells. We also report that short hairpin RNA–mediated Fn14 knockdown reduced NSCLC cell migration and invasion in vitro. Finally, Fn14 overexpression enhanced NSCLC cell migration and invasion in vitro and increased experimental lung metastases in vivo. Thus, Fn14 may be a novel therapeutic target for patients with NSCLC, in particular for those with EGFR-driven tumors who have either primary or acquired resistance to EGFR TKIs.
Lung cancer is the leading cause of cancer deaths worldwide. In the United States alone, there were 222,520 new cases and 157,300 deaths in 2010.
Of lung cancers, 85% are classified as non-small cell lung cancer (NSCLC), which is further divided into adenocarcinoma, squamous cell carcinoma, and large-cell carcinoma. More recently, molecular subsets of adenocarcinomas have been identified, with specific genomic alterations initiating and maintaining the malignancy. One such genomic alteration is the overexpression and/or mutation of the epidermal growth factor receptor (EGFR).
EGFR is overexpressed in 40% to 80% of patients with NSCLC and is correlated with disease progression and overall poor survival.
Thus, EGFR tyrosine kinase inhibitors (TKIs), such as gefitinib and erlotinib, became promising therapies for patients with NSCLC. The clinical response rates for these drugs were unexpectedly low (<10% as single agents) and prompted investigation into patient subgroups that might benefit from EGFR-targeted drugs.
Patients with EGFR mutations were more sensitive to gefitinib and showed improved progression-free survival. Approximately 10% to 15% of the patients with NSCLC in the Western hemisphere and 30% to 50% of patients of Asian ethnicity have tumors harboring somatic mutations in the EGFR that cause constitutive activation of this receptor.
Despite the success of TKIs in patients with lung tumors harboring EGFR-activating mutations, many patients develop resistance to gefitinib and erlotinib and show disease progression. Mechanisms of acquired resistance to EGFR TKI therapy, such as the secondary mutation in exon 20 (T790M),
Thus, the molecular mechanisms that govern the progression of these lung tumors with EGFR mutations and resistance to anti-EGFR therapies remain to be elucidated.
Fibroblast growth factor–inducible 14 (Fn14; gene TNFRSF12A) is the smallest known member of the tumor necrosis factor superfamily of receptors, and its only known ligand is the multifunctional cytokine tumor necrosis factor–like weak inducer of apoptosis (TWEAK).
but little is known about the role of Fn14 in this tumor type.
In this study, we show that Fn14 is frequently highly expressed in NSCLCs and that, in NSCLC cell lines, Fn14 expression often occurs concurrently with activating EGFR mutations. We further show that Fn14 expression is induced by EGF and repressed by erlotinib, implicating the importance of EGFR signaling in regulating Fn14 expression. In an NSCLC cell line that is resistant to erlotinib because of an EGFR T790M mutation, Fn14 expression remains elevated unless a relatively high EGFR TKI dose is used. We also demonstrate that the expression of Fn14 modulates cell migration and invasion in NSCLC cells in vitro and that ectopic expression of Fn14 augments NSCLC tumor formation in an experimental metastasis assay. Together, these data suggest that Fn14 signaling contributes to NSCLC cell motility and invasion and that Fn14 may be a new potential target for NSCLC treatment.
Materials and Methods
Lung cancer samples were obtained from patients who underwent total tumor resection. Specimen blocks chosen for the TMA met the criteria of nonnecrotic, nonirradiated, or chemo-treated lung cancer tissue. NSCLC subtypes included adenocarcinoma (n = 179) and squamous cell carcinoma (n = 111). Samples were double punched (0.6 μm diameter) using an indexed manual arrayer with an attached stereomicroscope under the direction of one of the authors (G.H.), who also reviewed and verified the tumor content. IHC analysis for Fn14 was performed using the Fn14 monoclonal antibody P4A8 (Biogen Idec, Inc., Weston, MA), as previously described.
p-EGFR analysis was performed using an antibody specific for EGFR-Y1068 (Cell Signaling Technologies, Beverly, CA). A scoring system for each chromophore, composed of staining intensity and extensiveness, captured the outcome: 0, negative; 1, weak; 2 moderate; and 3, strong.
Human NSCLC cell lines H520, H2122, A549, H1703, H358, H3255, H1975, HCC2279, and HCC827 (ATCC, Manassas, VA) were maintained in RPMI 1640 medium (Invitrogen, Carlsbad, CA), supplemented with 10% heat-inactivated fetal bovine serum (FBS) in a 37°C, 5% CO2 atmosphere. For the EGF stimulation and erlotinib treatment experiments, cells were placed in RPMI 1640 medium supplemented with 0.5% FBS for 18 hours before growth factor or drug addition.
Reagents, Antibodies, and Immunoblot Analysis
Erlotinib was obtained from BioVision (Mountain View, CA). EGF was obtained from Millipore (Billerica, MA) or R&D Systems (Minneapolis, MN). Polyclonal Fn14 antibodies were either generated by us
or obtained from Cell Signaling Technologies. Antibodies specific to p-EGFR (Y-1068), total EGFR, EGFR L858R mutant, and the EGFR E746-A750 deletion mutant were obtained from Cell Signaling Technologies. The α-tubulin antibody was obtained from Millipore or eBioscience (San Diego, CA), and the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and hemagglutinin epitope antibodies were obtained from Cell Signaling Technologies. Immunoblot analysis was performed as previously described.
The pBABE retroviral constructs of wild-type EGFR (Addgene 11011) and EGFR mutants (L858R-11012, L747-E749 del-11015, D770-N771 ins-11016, and D837-11014) were obtained from Addgene (Cambridge, MA) and were previously described.
The K-ras V12 pBABE construct (9052) and the empty pBABE vector (1764) were also obtained from Addgene.org. Replication-incompetent retroviruses were produced from the EGFR constructs by transfection into the Phoenix 293T packaging cell line (Allele Biotech, San Diego, CA) using Lipofectamine 2000 (Invitrogen). RL-65 cells were infected with these retroviruses in the presence of 5 g/mL polybrene. At 24 hours after infection, 2 g/mL puromycin was added to media, and cells were maintained for 5 days before lysis and immunoblotting for Fn14, p-EGFR, EGFR, and GAPDH.
Lentiviral Constructs and Transduction
Lentiviral constructs (pGIPZ) containing nonsilencing short hairpin RNA (shRNA) or shRNAs targeting two different regions of the Fn14 transcript (Fn14shRNA154, clone ID V3LHS_380154; Fn14shRNA156, clone ID V3LHS_380156) were obtained from Open Biosystems (Huntsville, AL). To generate the Fn14 overexpression construct, the coding sequence for Fn14 was amplified by PCR and ligated in-frame upstream of a 3XHA epitope in pcDNA3. For stable transduction, the HA epitope–tagged Fn14 fragment was excised from pcDNA3 and ligated into the lentiviral transfer vector pCDH (System Biosciences, Mountain View, CA) that contains a second transcriptional cassette for the expression of green fluorescent protein (GFP). An empty pCDH vector expressing only GFP or a nonsilencing shRNAmir vector expressing GFP was used as a control in an overexpression or knockdown experiment, respectively. Vesicular stomatitis virus-G-pseudotyped recombinant lentiviruses encoding Fn14 were produced by cotransfection of 293 packaging cells with the pCDH-Fn14 HA construct and the pPACK packaging mix (System Biosciences), according to the manufacturer's directions. Pseudotyped lentiviruses encoding shRNAs were produced by cotransfection of packaging cells with the appropriate shRNA construct and the Trans-Lentiviral Packaging Extract (Open Biosystems), according to the manufacturer's protocol. For lentiviral transduction, medium containing recombinant lentiviruses was harvested from the packaging cells at 48 hours after transfection, concentrated by polyethylene glycol precipitation and centrifugation, and added to subconfluent cultures of cells together with 8 μg/mL polybrene. Positively transduced cells were enriched by mass sorting the GFP-positive cells on a Vantage flow cytometer (BD Biosciences, San Jose, CA).
Migration and Invasion
Migration was assessed using a 25 × 80-mm polycarbonate membrane (8 μmol/L pore) and 12-well Chemotaxis Chamber (Neuro Probe Inc., Gaithersburg, MD). Cells (5 × 104) were seeded into each of 12 wells and allowed to migrate through the membrane for 5 hours. Cells were maintained in RPMI 1640 medium (Invitrogen) without FBS throughout the experiment. For EGF-induced migration, EGF (50 ng/mL) was placed in the bottom chamber. All treatments were performed in triplicate. Stationary cells were removed from the top of the membrane by scraping, and the membrane was fixed in 70% methanol. The membrane was stained with DAPI and mounted on a glass slide. Cells were counted from five random fields using fluorescent microscopy.
For the invasion assays, 1 × 105 cells suspended in RPMI 1640 medium with 0.5% serum were seeded into growth factor–reduced Matrigel invasion chambers (BD Biosciences). Chambers were placed in a 24-well plate containing RPMI 1640 medium and 10% FBS. Cells were allowed to invade through the membrane for 20 hours. For EGF-induced invasion, the bottom chamber contained RPMI 1640 medium with 0.5% serum and 50 ng/mL EGF. All treatments were performed in triplicate. Cells were fixed and stained in 0.1% crystal violet in 20% methanol. Stationary cells were removed from the top of the membrane by scraping, and the membrane was mounted on a glass slide. Cells were counted from six random fields using light microscopy.
Tail Vein Injection of A549 Cells
A549 lung adenocarcinoma cells (5 × 105) stably infected with either lentiviral vector or human Fn14 with an HA epitope tag (hFn14-HA) lentivirus (n = 9 per group) were injected i.v. (tail vein) into beige mice with severe combined immunodeficiency (Taconic, Hudson, NY), as previously described.
Six weeks after injection, mice were sacrificed and lungs were fixed in 10% neutral-buffered formalin. Lungs were evaluated for metastatic surface nodules by gross visualization. Tumor lesions were counted in a blinded manner using a dissecting microscope. Tissue was then embedded in paraffin and divided into sections (4 μm thick). Sections were stained with H&E using standard procedures, and immunohistochemical (IHC) staining was performed using an anti-HA antibody (Cell Signaling Technologies).
For TMA, tests for correlation using the rank-based Kendall's τ statistic were calculated using the cor. test function in the R statistical package (http://www.r-project.org, last accessed February 11, 2011). Migration and invasion assay results were assessed using the two-sample Student's t-test. P < 0.05 was considered significant.
Fn14 Is Overexpressed in Primary Human NSCLC Tumors, and High Fn14 Levels Correlate with EGFR Phosphorylation
We examined, by IHC staining of a lung TMA, whether Fn14 was expressed in primary human NSCLC tumors and, if so, whether those tumors with elevated Fn14 expression had activated EGFR, as assessed by staining with a p-EGFR (Y-1068) antibody. Representative staining patterns of an Fn14-negative/p-EGFR–negative patient and an Fn14-positive/p-EGFR–positive patient are shown in Figure 1A. Fn14 showed moderate to strong staining (scores 2 and 3) in 48.6% of adenocarcinoma specimens and 31.5% of squamous cell carcinoma specimens (Figure 1B). To determine whether Fn14 protein staining correlated with activated EGFR expression, tumors were scored for both Fn14 and p-EGFR protein staining. There was a statistically significant positive correlation (P < 0.01) between Fn14 and p-EGFR expression in both adenocarcinoma and squamous cell carcinoma subtypes, as determined by the Kendall's τ rank correlation test (Figure 1C).
Fn14 Is Highly Expressed in EGFR- or K-ras–Mutated NSCLC Cell Lines
We assessed the protein expression of Fn14 across nine NSCLC cell lines by immunoblot analysis. Fn14 expression was detected in five of six cell lines classified as adenocarcinoma, and in an adenocarcinoma in situ cell line, but expression was much lower in the two squamous cell carcinoma cell lines (Figure 2). Relatively high levels of Fn14 often coincided with p-EGFR expression (Y-1068), driven by activating EGFR mutations (L858R and ΔE746-A750), including the H1975 cell line, which also contained a secondary mutation (T790M) known to confer resistance to EGFR TKIs. In addition, Fn14 protein was also highly expressed in two of the three cell lines (A549 and H358) with activating K-ras mutations.
Erlotinib Suppresses Fn14 Expression
To assess whether constitutive EGFR signaling was directly triggering Fn14 expression, we used the small-molecule EGFR kinase inhibitor, erlotinib. Treatment of HCC827 cells with 1 μmol/L erlotinib inhibited EGFR activation, as detected by decreased p-EGFR protein expression (Figure 3A). In addition, erlotinib treatment decreased the level of Fn14 protein expression in a time-dependent (Figure 3A) and dose-dependent (Figure 3B) manner. Erlotinib treatment also decreased Fn14 expression in two other cell lines with activating EGFR mutations: H3255 and HCC2279 (data not shown). Treatment of the H1975 NSCLC cell line, which contains both an EGFR-activating mutation and an EGFR-TKI resistance mutation, required a 1000-fold higher erlotinib dose to decrease p-EGFR and Fn14 levels compared with HCC827 cells (Figure 3B).
Activating EGFR Mutations Increase Fn14 Expression in Lung Epithelial Cells
To further investigate the relationship between activated EGFR and Fn14 expression, we expressed both activating and nonactivating EGFR constructs in a nontransformed, immortalized rat lung epithelial cell line (RL-65). Expression of wild-type EGFR and mutant EGFR receptors increased p-EGFR levels compared with vector control or mutant K-ras retrovirus-infected cells (Figure 4). It is likely that phosphorylation of the kinase-dead EGFR mutant (D837A) occurred either via homodimerization with endogenous EGFR or by the activity of other tyrosine kinases in the RL-65 cells. Fn14 protein expression was induced in both the wild-type and activating EGFR mutation cells; in contrast, expression of the kinase-dead EGFR mutant failed to enhance Fn14 expression. Notably, expression of a K-ras mutant (V12) in RL-65 cells also resulted in increased Fn14 expression.
Fn14 Depletion Reduces NSCLC Cell Migration and Invasive Capacity in Vitro
We next investigated whether depletion of Fn14 expression by shRNA could suppress NSCLC cell migration and invasion. Cells stably expressing Fn14 shRNA 154 or 156, targeting the Fn14 transcript, showed a suppression of Fn14 levels in both HCC827 (Figure 5A) and H1975 (Figure 5B) cells. These cell lines were less migratory and invasive compared with control cells. Specifically, the knockdown of Fn14 by shRNA suppressed NSCLC cell migration by 64% (shRNA 154) or 81% (shRNA 156) in HCC827 cells and 50% (shRNA 156) in H1975 cells (Figure 5, C and D). The depletion of Fn14 by shRNA suppressed cell invasion by 50% (shRNA 154) or 66% (shRNA 156) in HCC827 cells and 37% (shRNA 156) in H1975 cells (Figure 5, E and F).
Fn14 Depletion Attenuates EGF-Induced NSCLC Cell Migration and Invasion in Vitro
Because Fn14 expression was modulated by EGFR signaling, and Fn14 influenced baseline NSCLC cell migration and invasion, we tested whether Fn14 depletion also reduced EGF-driven cell migration and invasion. We found that EGF treatment of A549 cells increased Fn14 protein expression, with maximal induction at 4 hours (Figure 6A). Next, we generated an Fn14-deficient A549 cell line using shRNA lentiviral infection, and cells infected with the shRNA 156 virus expressed low levels of Fn14 (Figure 6B). Although EGF treatment enhanced A549-control shRNA cell migration by approximately threefold, EGF treatment of the A549 Fn14 shRNA cells increased migration by only approximately twofold (Figure 6C). Similarly, EGF treatment also increased A549-control shRNA cell invasion by approximately 10-fold, whereas the knockdown of Fn14 suppressed EGF-driven cell invasion by 30% compared with A549-control shRNA cells (Figure 6D).
Fn14 Overexpression Enhances NSCLC Cell Motility and Invasion in Vitro
We next investigated whether Fn14 overexpression would modulate NSCLC cell motility and invasion. We infected A549 cells with control lentivirus or lentivirus encoding full-length, hFn14-HA, and stable cell lines were isolated. Fn14-HA expression in the A549 cell line was confirmed by immunoblotting (Figure 7A). Compared with the vector-infected A549 cells (control), ectopic expression of Fn14 enhanced cell migration approximately twofold (Figure 7B). The expression of Fn14 also enhanced A549 cell invasion by approximately 1.6-fold (Figure 7C).
Fn14 Overexpression Enhances Experimental Metastasis in Vivo
Because our in vitro data indicated involvement of Fn14 in migration and invasion of NSCLC cells, we next tested whether overexpression of Fn14 could enhance NSCLC tumor formation in an in vivo model of experimental metastasis. Control or hFn14-HA–overexpressing A549 cells were injected into the tail vein of Beige mice with severe combined immunodeficiency. Mice were evaluated grossly for lung tumors 6 weeks after injection. H&E staining of formalin-fixed, paraffin-embedded sections confirmed the presence of lung tumors (Figure 8A). The number of gross lung tumors was 40% higher in mice injected with A549-hFn14-HA cells compared with A549 control cells (Figure 8B).
In this study, we demonstrate that Fn14 was highly expressed in human NSCLCs and correlates with the expression of phosphorylated EGFR. We also show that activation of EGFR signaling, either through EGF:EGFR binding or mutations in the EGFR kinase domain, enhance Fn14 protein expression, and that pharmacological inhibition of EGFR signaling decreases Fn14 expression. In NSCLC cells harboring an EGFR-activating mutation and the T790M EGFR mutation that reduces EGFR TKI sensitivity, Fn14 levels are elevated and only decrease after exposure to a high EGFR TKI dose. We further demonstrate that Fn14 expression levels could modulate NSCLC cell migration and invasion in vitro and experimental lung metastasis in vivo.
The Fn14 protein is highly expressed in primary human NSCLC tumors. This follows reports showing that Fn14 expression is enhanced in several human malignancies,
EGFR activation is achieved through ligand binding, receptor overexpression, or receptor mutation. EGF and many other growth factors and cytokines, including fibroblast growth factors 1 and 2, platelet-derived growth factor-BB, transforming growth factor-β1, and TWEAK, can induce Fn14 gene expression.
Thus, we determined whether Fn14 was preferentially overexpressed in NSCLC primary tumors with activated EGFR. In both adenocarcinoma and squamous cell carcinoma subtypes, Fn14 expression was significantly correlated with tyrosine phosphorylated EGFR, a marker of EGFR activation. Fn14 is also highly expressed in HER2+ breast tumors.
The observed association between Fn14 expression and ErbB family member activation suggests that therapeutic targeting of Fn14 in tumors driven by oncogenic ErbB family members could benefit patients.
Along with receptor overexpression and EGF:EGFR engagement, activation of EGFR can occur via mutations in the EGFR kinase domain. EGFR-TK mutations occur in 10% to 15% of patients with NSCLC in the Western hemisphere and in 30% to 50% of patients of Asian ethnicity.
In this study, we demonstrate that NSCLC cell lines harboring EGFRs with activating mutations have increased Fn14 expression. We further show that expressing EGFRs with TK domain mutations in immortalized rat lung epithelial cells increased Fn14 expression. We found that erlotinib exposure inhibited p-EGFR levels in three NSCLC cell lines harboring EGFR-activating mutations, with a concomitant reduction in Fn14. Although patients with activating EGFR mutations initially respond to EGFR-TKIs, subsequent resistance and relapse are common. Mechanisms of resistance include secondary mutations in EGFR
We also showed that H1975 NSCLC cells harboring an activating EGFR mutation and the T790M EGFR mutation, which reduces sensitivity to EGFR-TKIs, sustain Fn14 protein expression after exposure to erlotinib, except at high erlotinib concentrations. Thus, targeting of Fn14 may be an attractive approach to treat patients with NSCLC who are resistant to EGFR-TKI therapy.
EGF, through binding and activation of EGFR, is a well-characterized inducer of cell migration and is frequently used to stimulate lung cancer cell migration in vitro.
Herein, we showed that Fn14 depletion through RNA interference reduced NSCLC cell migration and invasion driven by EGF:EGFR binding or by activating mutations in EGFR. The suppression of Fn14 does not completely abrogate EGF-induced lung cell migration or invasion, indicating that other motility mechanisms are still active in these cells. We also demonstrated that overexpression of Fn14 enhanced NSCLC cell migration and invasion in vitro and enhanced experimental lung metastasis in vivo. Thus, our data suggest that Fn14 levels can modulate NSCLC cell migration and invasion. This supports growing evidence that Fn14 is involved in control of motility and the invasive properties of cancer cells. For example, Watts and colleagues
Our laboratory has previously demonstrated that Fn14 is expressed at high levels in invading glioma cells in vivo and that ectopic expression of Fn14 stimulated glioma cell invasion and migration through an Rac1-dependent mechanism in vitro.
described an Fn14 monoclonal antibody with recombinant gelonin chemical conjugate, that suppressed the growth of several cancer cell lines in vitro and inhibited tumor growth in an in vivo model of bladder cancer. In some cases, administration of anti-Fn14 monoclonal antibodies alone can inhibit tumor growth in xenograft assays.
Thus, tumor suppression through the targeting of Fn14 may prove to be a therapeutic intervention in NSCLC and other tumor types.
In conclusion, Fn14 is highly expressed in primary human NSCLCs and correlated with activation of EGFR. Consistent with this observation, most NSCLC cell lines with activating EGFR mutations express high levels of Fn14. Suppression of Fn14 reduced NSCLC migration and invasion in vitro, whereas the overexpression of Fn14 enhanced migration and invasion in vitro and experimental metastasis in vivo. Thus, Fn14 may be a therapeutic target for patients with NSCLC, in particular for those with EGFR-driven tumors who have either primary or acquired resistance to EGFR TKIs.
We thank Serdar Tuncali and Allison Cheng for technical assistance, Dr. Jennifer Michaelson (Biogen Idec, Inc.) for providing the Fn14 monoclonal antibody, and the Collaborative Bioinformatics Center at the Translational Genomics Research Institute for their efforts in statistical analysis.
This work was supported in part by NIH grants R01 NS055126 (J.A.W), R01 CA130940 (N.L.T.), R01 CA103956 (J.C.L.); by the Translational Genomics Research Institute Foundation and Scottsdale Healthcare Foundation (G.J.W.), St. Joseph's Foundation (Phoenix, AZ) for the Heart and Lung Institute Research Initiative (L.J.I. and R.M.B.), and T32 HL007698 (University of Maryland School of Medicine to E.C.).
N.L.T. and J.A.W. contributed equally to this work.
CME Disclosure: The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interest to disclose.
In the article entitled, “Elevated Expression of Fn14 in Non-Small Cell Lung Cancer Correlates with Activated EGFR and Promotes Tumor Cell Migration and Invasion” (Volume 181, pages 111–120 of the July 2012 issue of The American Journal of Pathology), the author footnotes contained errors. The support footnote and author contribution footnote should have read as follows: