Clin Exp Metastasis DOI 10.1007/s10585-014-9653-6

RESEARCH PAPER

FN14 expression correlates with MET in NSCLC and promotes MET-driven cell invasion Timothy G. Whitsett • Shannon P. Fortin Ensign • Harshil D. Dhruv • Landon J. Inge • Paul Kurywchak • Kerri K. Wolf • Janine LoBello • Christopher B. Kingsley • Jeffrey W. Allen • Glen J. Weiss • Nhan L. Tran

Received: 15 October 2013 / Accepted: 27 March 2014 Ó Springer Science+Business Media Dordrecht 2014

Abstract The five-year survival rate in advanced nonsmall cell lung cancer (NSCLC) remains below ten percent. The invasive and metastatic nature of NSCLC tumor cells contributes to the high mortality rate, and as such the mechanisms that govern NSCLC metastasis is an active area of investigation. Two surface receptors that influence NSCLC invasion and metastasis are the hepatocyte growth factor receptor (HGFR/MET) and fibroblast growth factorinducible 14 (FN14). MET protein is over-expressed in NSCLC tumors and associated with poor clinical outcome and metastasis. FN14 protein is also elevated in NSCLC tumors and positively correlates with tumor cell migration and invasion. In this report, we show that MET and FN14 protein expressions are significantly correlated in human primary NSCLC tumors, and the protein levels of MET and FN14 are elevated in metastatic lesions relative to patientmatched primary tumors. In vitro, HGF/MET activation significantly enhances FN14 mRNA and protein

expression. Importantly, depletion of FN14 is sufficient to inhibit MET-driven NSCLC tumor cell migration and invasion in vitro. This work suggests that MET and FN14 protein expressions are associated with the invasive and metastatic potential of NSCLC. Receptor-targeted therapeutics for both MET and FN14 are in clinical development, the use of which may mitigate the metastatic potential of NSCLC.

T. G. Whitsett (&)  S. P. Fortin Ensign  H. D. Dhruv  P. Kurywchak  K. K. Wolf  G. J. Weiss  N. L. Tran (&) Cancer and Cell Biology Division, The Translational Genomics Research Institute (TGen), 445 N. Fifth St., Suite 400, Phoenix, AZ 85004, USA e-mail: [email protected]

J. LoBello Integrated Cancer Genomics Division, The Translational Genomics Research Institute (TGen), Phoenix, AZ, USA

N. L. Tran e-mail: [email protected] S. P. Fortin Ensign Cancer Biology Graduate Interdisciplinary Program, University of Arizona, Tucson, AZ, USA

Keywords Metastasis

Non-small cell lung cancer  MET  FN14 

Introduction Lung cancer remains the leading cause of cancer mortality in the US and throughout the world [1], with most patients presenting with advanced stage disease. The five-year survival rate for advanced stage non-small cell lung cancer

C. B. Kingsley Diabetes, Cardiovascular and Metabolic Diseases Division, The Translational Genomics Research Institute (TGen), Phoenix, AZ, USA J. W. Allen Humboldt Medical Specialists, Eureka, CA, USA G. J. Weiss Cancer Treatment Centers of America, Goodyear, AZ, USA

L. J. Inge Center for Thoracic Disease and Transplantation, Heart and Lung Institute, St. Joseph’s Hospital and Medical Center, Phoenix, AZ, USA

123

Clin Exp Metastasis

(NSCLC) remains below 10 %, necessitating the need for novel therapeutic strategies against advanced disease [2]. A significant clinical hurdle to reduce mortality in NSCLC is the propensity for tumor cell invasiveness and metastasis. Despite pharmacological advances for NSCLC, current treatments have limited efficacy in metastatic disease, and the majority of patients succumb to the overwhelming tumor burden resulting from tumor spread. Even molecularly targeted therapeutics such as the epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) fail due to tumor resistance, followed by disease progression. Two targets associated with NSCLC metastasis are the hepatocyte growth factor receptor (HGFR/MET) and tumor necrosis factor receptor superfamily, member 12A (TNFRSF12A, also known as fibroblast growth factorinducible 14 (FN14)). MET and its cognate ligand, hepatocyte growth factor (HGF), are associated with tumor progression and metastasis in many solid tumor types [3]. Within adenocarcinomas, the most common subtype of NSCLC, the protein expression of MET was detected in *30 % of cases and MET gene amplification was seen in 10 % of cases [4]. Increased MET activity occurs secondary to oncogenic activation of KRAS [5], while gene amplification of MET is often a mechanism of resistance to EGFR-TKIs [6]. Furthermore, the expression of MET was more common in poorly differentiated adenocarcinomas compared to welldifferentiated tumors [4]. Among all types of NSCLC, MET expression significantly correlated with brain metastasis [7]. Ectopic expression of both HGF and MET in the NSCLC cell line H460 induced spontaneous metastasis to distant organs [8]. Thus, an improved understanding of HGF/MET signaling may offer novel therapeutic opportunities against metastatic NSCLC. The TNF-like weak inducer of apoptosis (TWEAK)FN14 signaling axis modulates tumor growth, cell survival, and tumor invasion. Increased expression of FN14 has been observed in a number of solid tumors, including hepatocellular carcinoma [9], glioblastoma (GB) [10, 11], esophageal adenocarcinoma [12], and HER2 ? breast cancer [13]. FN14 is also highly expressed in NSCLC and FN14 protein expression significantly correlates with activated EGFR signaling [14]. Additionally, FN14 expression was maintained at a high level in NSCLC cells resistant to EGFR-TKIs [14]. In GB, FN14 signaling modulated cell survival through regulation of NF-jB, BCL2L1/Bcl-xL, and BCL2 expression and AKT2 activation [15, 16]. FN14 signaling also promoted glioma and breast cell invasion through activation of RAC1 and NF-jB [11, 13]. In NSCLC, FN14 induced cell migration and invasion in vitro, and ectopic expression of FN14 enhanced metastasis in vivo [14]. Thus, FN14 may play a critical role in cancer cell invasion and metastasis and represent a

123

potential therapeutic vulnerability in advanced stage NSCLC. Efforts to directly target the TWEAK/FN14 signaling pathway are currently under investigation [17], and clinical trials are ongoing (http://clinicaltrials.gov/ct2/ show/NCT01383733). In this study, we demonstrate that FN14 and MET protein expressions are significantly correlated by immunohistochemistry (IHC) in primary NSCLC tumors, and that FN14 and MET protein levels are elevated in metastatic tumors relative to patient-matched primary tumors. In vitro, activation of MET signaling significantly induces FN14 mRNA and protein levels, while pharmacologic suppression of MET signaling diminishes FN14 mRNA and protein levels. Lastly, we show that depletion of FN14 diminishes MET-driven cell migration and invasion in vitro. These data suggest that MET and FN14 contribute to an invasive phenotype in NSCLC, and that therapeutic targeting of these cell surface receptors may provide clinical benefit in advanced metastatic NSCLC.

Materials and methods Tumor tissue microarray The construction and use of tissue microarrays (TMA) from primary patient NSCLC tumor specimens has been previously described [14]. IHC analysis for FN14 was done using the FN14 monoclonal antibody P4A8 (Biogen Idec, Inc.) as previously described [11, 14]. MET analysis was performed using an antibody specific for MET (kindly provided by Dr. George Vande Woude, Van Andel Research Institute) [18]. A scoring system for each chromophore comprised of staining intensity and extensiveness captured the outcome: 0, negative; 1, weak; 2 moderate; 3, strong. Tests for correlation using the rank based Kendall’s tau statistic were calculated using the cor.test function in the R statistical package. p values were adjusted for multiple testing and a p \ 0.05 was considered significant. Metastatic lung cancer mouse model KRASG12D/LKB1lox/lox were generated by selective breeding of Lox-Stop-Lox KRASG12D mice [19] purchased from Jackson Laboratories (Bar-Harbor, ME) with LKB1lox/lox mice obtained from Mouse Repository at the NCI and described previously [20], and were inbred on a FVB background. Induction of lung tumors was performed as described previously [21]. Briefly, six-week-old KRASG12D/LKB1lox/lox mice were transiently infected with 5 9 106 p.f.u. of Cre adenovirus (University of Iowa adenoviral core) via intranasal infection. KRASG12D/ LKB1lox/lox mice were monitored post-infection for clinical

Clin Exp Metastasis

signs of disease (labored breathing, severe weight loss), at which time mice were euthanized, necropsied and grossly evaluated for metastases. Primary and suspected metastatic lung tumors were collected and prepared for immunohistological analysis using routine procedures. Following confirmation of status of metastatic lesion, matching primary and metastatic tumors were sectioned and stained for FN14 (#ab109365) and MET (#ab101539) using antibodies purchased from Abcam (Cambridge, MA) using previously described procedures [14]. Cell culture Human NSCLC cell lines H2073 and H1993 (American Type Culture Collection, Manassas, VA) and mouse cell lines CMT64 and CMT167 (Sigma-Aldrich, St. Louis, MO) were maintained in RPMI 1640 (Invitrogen, Carlsbad, CA) and DMEM (Invitrogen), respectively, supplemented with 10 % heat-inactivated fetal bovine serum (FBS) in a 37 °C, 5 % CO2 atmosphere. For the HGF stimulation, cells were placed in RPMI 1640 without FBS for 18 h prior to growth factor exposure. Reagents, antibodies and immunoblot analysis HGF was purchased from Millipore (Billerica, MA). SU11274 was purchased from Selleck Chemicals (Houston, TX) and used at a final concentration of 2 lM. Antibodies for immunoblot analysis against FN14, MET, phosphorylated MET and GAPDH were purchased from Cell Signaling Technology (Danvers, MA), while tubulin (TUBA1A) was purchased from Millipore. Immunoblot analysis was performed as previously described [14]. Densitometry was performed using the publically available imaging software program, Image J (http://rsbweb.nih.gov/ ij/). Fluorescence activated cell sorting (FACS) FN14 cell surface protein expression was determined by Fluorescence activated cell sorting (FACS) analysis. Briefly, H2073, H1993, CMT64, and CMT167 cells were treated with 10 mM EDTA in PBS solution to obtain single cell suspension. For HGF exposure, H2073 cells were serum starved overnight prior to HGF (20 ng/mL) exposure and EDTA treatment. Single cell suspensions were incubated with ITEM-4 antibody (20 lg/mL, the ITEM-4 antibody was a generous gift from Dr. Hideo Yagita [22]) or isotype control antibody (20 lg/mL, Mouse IgG2b, Thermo Scientific, Rockford, IL) for 1 h on ice, followed by 30 min incubation with Alexafluor 488 labeled antimouse antibody (1:200, Invitrogen) on ice. After washing with PBS, the stained cells (gated on the basis of forward

and side scatter profiles) were analyzed on a FACSCaliber (Becton–Dickinson, San Jose, CA, USA) and data were processed using the FlowJo program (Tree Star Inc.). Quantitative, real-time PCR (qPCR) mRNA expression was determined by qPCR as previously described [23]. Briefly, total RNA was extracted from cell lines using the mirVana isolation kit (Ambion, Austin, TX) according to the manufacturer’s directions. cDNA was synthesized from total RNA using SuperScript III FirstStrand Synthesis SuperMix (Life Technology, Grand Island, NY) according to the manufacturer’s protocol. Quantitative RT-PCR analyses of FN14 (forward: 50 -CC AAGCTCCAACCACAA-30 ; reverse: 50 -TGGGGCCTAG TGTCAAGTCT-300 ) and histone H3.3 (forward: 50 -CCA CTAACTTCTGATTCGC-30 ; reverse: 50 -GCGTGCTAGC TGGATGTCTT-30 ) were carried out in triplicate in a 384-well plate using a LightCycler 480 (Roche Applied Sciences, Indianapolis, IN) and analyzed as previously described [10]. Lentiviral constructs and transduction Lentiviral constructs (pGIPZ) containing non-silencing shRNA or shRNA targeting the FN14 transcript (Fn14shRNA156, clone ID V3LHS_380156) were purchased from Open Biosystems (Huntsville, AL). A nonsilencing shRNAmir vector was used as a control in knockdown experiments. Pseudo-typed lentiviruses encoding shRNAs were produced by co-transfection 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 h posttransfection, concentrated by PEG precipitation and centrifugation, and added to subconfluent cultures of cells together with 8 lg/ml polybrene. Small-interfering RNA preparation and transfection Small interfering RNA (siRNA) oligonucleotides specific for GL2 Luciferase (siRNA-Luci) were previously described [24]. siRNA sequences for FN14 full-length transcripts (FN14-3: CTCAGATGTCCTGAAATTCCA; FN14-4: CG CCCACTCATCATTCATTCA; and FN14-5: ATCATTC ATTCATCCATTCTA) were purchased from Qiagen. Transient siRNA transfection was carried out as previously described [24]. All siRNA transfections were done at 20 nM siRNA using Lipofectamine RNAi MAX reagent (Invitrogen) and no cytotoxicity were observed 24 h post-

123

Clin Exp Metastasis

transfection. Maximum inhibition of protein levels was achieved approximately 48–72 h post-transfection. Migration and invasion Migration was assessed using a 25 9 80 mm polycarbonate membrane (8 lM pore) and 12-well Chemotaxis Chamber (Neuro Probe Inc., Gaithersburg, MD). Cells (5 9 104) were seeded into each of 12 wells and allowed to migrate through the membrane for 5 h. Cells were maintained in RPMI 1640 (Invitrogen) without FBS throughout the experiment. 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 5 random fields using fluorescent microscopy. A p value B0.05 was considered statistically significant as determined using the two sample Student’s t test. For HGF chemoattractant migration, serum starved H2073 cells were seeded in the upper chamber and HGF (20 ng/mL) was placed in the lower chamber. For the invasion assays, cells (1 9 105) suspended in RPMI 1640 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 and 10 % FBS. Cells were allowed to invade through the membrane for 20 h. All treatments were performed in triplicate. Cells were fixed and stained with DAPI. 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 five random fields using light microscopy. Statistical analyses were done as described for the migration assays.

Results FN14 protein is over-expressed in primary human NSCLC tumors and correlates with MET protein levels Elevated protein expression of MET has been observed in metastatic lesions compared to primary lung tumors [7]. Thus, we examined a NSCLC tissue microarray by IHC to determine whether those tumors with elevated FN14 protein expressed MET. Representative staining patterns of a FN14 positive/MET positive patient are shown in Fig. 1a. To determine whether FN14 protein correlated with MET expression, tumors were scored for both FN14 and MET staining intensity. There was a statistically significant positive correlation (p value \0.01) between FN14 and MET protein expression in both adenocarcinoma and squamous cell carcinoma subtypes as determined by the Kendall’s tau rank correlation test (Fig. 1b). In addition, significant positive correlation was observed between the

123

mRNA expression levels of FN14 and MET in both lung adenocarcinomas (Fig. 1c) and squamous cell lung carcinoma (Fig. 1d) samples in a publically available dataset (www.genesapiens.org). FN14 and MET proteins are highly expressed in lung cancer metastases Next, we examined the protein levels of FN14 and MET by IHC staining of human primary lung tumors and patientmatched metastases. In fourteen cases of patient-matched primary lung and metastatic tumors from NSCLC patients, FN14 and MET protein levels were more highly expressed in the metastatic tumors (Fig. 2a). In the metastatic tumor samples, 12/13 displayed elevated (IHC score of 2–3) MET protein expression, and 12/14 displayed elevated FN14 protein expression. The combined, averaged IHC score for both FN14 and MET were significantly elevated in the metastatic tumors relative to patient-matched primary lung tumors (Fig. 2b). Using an in vivo mouse model of metastatic lung cancer we further investigated the protein expressions of FN14 and MET by IHC. Transient expression of Cre and subsequent expression of KRASG12D and homozygous deletion of LKB1 within the lungs of the KRASG12D/LKB1lox/lox mice results in multi-focal, highly aggressive NSCLC with both local and distant metastases [25, 26]. Consistent with previous findings, we found that KRASG12D/LKB1lox/lox mice with significant signs of clinical disease (labored breathing, severe weight loss) 8-10 weeks post-infection with Cre adenovirus harbored distant metastases to several sites (data not shown). A board-certified pathologist (J.L.) determined that both primary and metastatic lesions were primarily lung adenocarcinoma, although small regions of squamous cell carcinoma were observable as previously described [25]. Following histological confirmation of metastatic NSCLC at these distant sites, we stained for FN14 and MET protein expressions in matched primary and metastatic NSCLC tumors from these mice. Although both the primary and matched metastatic tumors displayed FN14 staining, intensity of FN14 staining within the primary tumor appeared more heterogenous, whereas staining intensity of Fn14 within metastatic NSCLC lesions appeared more uniform (Fig. 3). Interestingly, primary NSCLC lesions from KRASG12D/LKB1lox/lox mice displayed low or no staining of MET (Fig. 3), while the matched metastatic lesions displayed increased MET staining (Fig. 3). MET-driven NSCLC cell lines express elevated FN14 protein and message levels We next investigated the relationship between FN14 and MET in an in vitro model of patient-matched primary lung

Clin Exp Metastasis

A

C AC n=107 r=0.38 p=4.5-5

MET

FN14

D SCC n=44 r=0.76 p=1.5-9

B Correlation between Fn14 and c-Met by Kendall’s Tau NSCLC Specimens

Antigen 1

Antigen 2

p value

Adenocarcinoma

FN14

MET

0.00002

Squamous Cell

FN14

MET

0.01

Fig. 1 FN14 protein expression in NSCLC tumors correlates with MET protein levels. a FN14 and MET staining on representative samples from one patient (10 9 objective (left column) and 40 9 objective (right column), Aperio GL Scanner). b Tumor-cell specific FN14 and MET staining in each of the tumor punches was scored by a board-certified pathologist (J.L.); a score of zero indicates staining level equal to adjacent non-tumor cells. A non-zero score indicates increased staining (1 = minimum, 2 = moderate,

3 = strong positive). A total of 290 samples were scored for FN14 and MET expression and the correlation between the two stains was analyzed using Kendall’s tau rank correlation test, adjusted for multiple testing. The mRNA expressions of FN14 and MET were examined in (c) lung adenocarcinomas (AC) and (d) squamous cell lung carcinomas (SCC) using publically available gene expression data (www.genesapiens.org). Statistical correlations are performed on the website

cancer and lymph node metastasis. The H2073 and H1993 cells lines represent a primary lung adenocarcinoma and lymph node metastasis derived from the same patient; respectively, and H1993 harbors a MET receptor amplification [7]. H1993 displayed elevated total and activated MET protein as previously described. In conjunction, H1993 displayed elevated Fn14 protein expression compared to H2073 (Fig. 4b). Densitometry revealed a statistically significant, 3.5-fold protein induction in H1993 cells compared to H2073 cells. Figure 4b further confirmed that H1993 cells express elevated surface expression of FN14 protein compared to H2073 as determined by FACS analysis compared to H2073 cells. To further confirm the elevated protein expression of MET and FN14 in metastatic tumors, we examined a mouse cell line model of matched primary and metastatic tumors. The CMT64 cell line was established from a spontaneous lung carcinoma in a mouse [27], while the CMT167 is a highly metastatic subline of CMT64 [28]. We examined the protein levels of phosphorylated and total MET, FN14 and TUBA1A as a loading control in these cell lines. The metastatic CMT167 cells displayed elevated protein levels of

both total and activated MET with concomitant elevated protein expression of FN14 compared to CMT64 cells (Fig. 4c). Figure 4d further confirmed that CMT167 cells express elevated surface expression of FN14 protein compared to CMT64 cells as determined by FACS analysis. Pharmacologic suppression of MET activation in NSCLC cells decreases FN14 mRNA and protein levels in vitro We next investigated whether the elevated expression of FN14 in H1993 was dependent on MET activation. H1993 cells treated with the MET inhibitor, SU11274, showed inhibition of the Fn14 protein expression (Fig. 5a). Densitometry revealed a statistically significant (p value \0.05), 65 % reduction in FN14 protein levels after SU11274 exposure. To further confirm the suppression of FN14 by pharmacologic inhibition of MET activation, we examined the mRNA levels of FN14 after exposure of H1993 cells to the MET inhibitor. Figure 5b demonstrates a significant reduction in FN14 mRNA levels after exposure to SU11274.

123

Clin Exp Metastasis

A Primary NSCLC

Patient-Matched Metastasis

Patient

Type

MET IHC

FN14 IHC

MET IHC

FN14 IHC

1

S

1

1

3

2

2

A

1

3

3

3

3.5

3

A

1

2

3

2

3.0

4

S

3

2

2

2

5

A

0

3

NA

3

6

A

3

2

3

3

7

A

3

3

3

3

8

A

0

1

3

3

9

A

1

2

2

1

10

A

1

1

3

3

11

A

2

2

2

2

Average IHC Score

B *

2.5 2.0 1.5 1.0 0.5 0.0

12

S

0

1

2

3

13

S

1

1

0

1

14

A

3

3

2

2

36%

64%

92%

86%

IHC Score 2/3

Fig. 2 FN14 and MET protein levels are elevated in metastatic lesions compared to patient-match primary tumors. a IHC scoring of FN14 and MET in patient-matched primary and metastatic lung

Fig. 3 Metastatic NSCLC lesions express both FN14 and MET protein in a transgenic model of NSCLC. Images of matched primary and metastatic NSCLC tumors present in KRASG12D/LKB1lox/lox mice stained for FN14 and MET. Representative images were taken with a 20 9 objective

Primary Lung Adenocarcinoma

Pancreatic

Matched Metastasis

tumors. b Combined FN14 and MET IHC scores were compared by Student’s t test. Asterisk represents a p value \0.05

FN14

Metastasis

123

Primary Tumor

MET

A

TUBA1A

Relative FN14 Densitometry

Clin Exp Metastasis

CMT64

CMT167

H2073 H1993

WB: FN14 P-MET

MET

C

4.5

*

4

B

3.5 3 2.5 2 1.5 1 0.5 0 H2073

WB:

H1993

D

P-MET MET FN14 TUBA1A

WB: FN14 TUBA1A

+ -

+

DMSO SU11274

\0.05 by Student’s t test. b Surface analysis of FN14 protein levels was determined for H2073 and H1993 by FACS analysis, with isotype antibody used as a negative control. c Total cell lysates were prepared from CMT64 and CMT167 cell lines and immunoblotted with the indicated antibodies. Tubulin was used as a loading control. d Surface analysis of FN14 protein levels was determined for CMT64 and CMT167 by FACS, with isotype antibody used as a negative control

B

1.2

1.2

1 0.8 0.6

*

0.4 0.2 0 H1993

H1993+SU11274

Relative Fold Change

A

Relative FN14 Densitometry

Fig. 4 MET amplification enhances FN14 expression levels in metastasis-derived NSCLC cell lines. a Total cell lysates were prepared from H2073 (primary adenocarcinoma NSCLC cell line) and H1993 (patient-matched metastatic cell line with known MET amplification) and immunoblotted with the indicated antibodies. TUBA1A was used as a loading control. Densitometry was performed using Image J on biologic triplicates. Bars represent average ratio of FN14 protein to TUBA1A protein ? SEM. Asterisk denotes a p value

1 0.8 0.6

* 0.4 0.2 0 H1993

H1993 + SU11274

Fig. 5 Pharmacologic inhibition of MET activation decreases FN14 mRNA and protein levels in vitro. a H1993 cells were treated with vehicle or SU11274 for 24 h. Cells were harvested, total cell lysates were prepared and immunoblotted with an antibody against FN14. TUBA1A was used as a loading control. Densitometry was performed using Image J on biologic triplicates. Bars represent average ratio of

FN14/Tublulin protein ? SEM. Asterisk denotes a p value \0.05 by Student’s t test. b mRNA expression of FN14 was determined in H1993 cells treated with vehicle or SU11274 by qPCR. Bars represent the average fold change ? SEM compared to vehicle. Asterisk denotes a p value \0.05 as determined by Student’s t test

Activation of MET signaling through HGF induces FN14 mRNA and protein levels in vitro

levels of FN14 as early as 3 h post-HGF, and sustained higher mRNA levels of FN14 up to 24 h. The induction of FN14 through HGF/MET activation was further confirmed at the protein level. H2073 cells exposed to HGF for 3 h displayed elevated FN14 protein expression compared to untreated H2073 cells (Fig. 6b). Figure 6c further confirmed that H2073 cells exposed to HGF express elevated surface expression of FN14 protein as determined by FACS analysis.

As MET signaling in NSCLC cells is activated by exposure to its cognate ligand HGF, we sought to examine FN14 protein and mRNA levels after MET activation via HGF. H2073 cells were exposed to 20 ng/mL HGF over the course of 24 h. Figure 6a demonstrates that HGF exposure induced significant mRNA

123

Clin Exp Metastasis

A

B

FN14

p < 0.001

9

TUBA1A Relative Fold Change

8

H2073

7

H2073 +HGF

6

C

5 4 3 2 1 0 0

0.5

1

3

6

24

HGF Exposure Time (hours)

Fig. 6 Activation of MET signaling though HGF induces FN14 mRNA and protein expression in vitro. a Total RNA was collected from serum starved H2073 cells treated with 20 ng/mL HGF for the indicated time points. mRNA expression of FN14 was determined by qPCR with histone H3.3 mRNA levels used as endogenous control. Bars represent the average fold change ± standard error of triplicate qPCR reactions. A p value \0.05 compared to no treatment was

considered statistically significant by Student’s t test. b Serum starved H2073 cells were exposed to HGF (20 ng/mL) for 3 h. Cells were harvested, total cell lysates were prepared and immunoblotted with antibodies to FN14. TUBA1A was used as a loading control. c Surface analysis of FN14 protein levels was determined for H2073 exposed to 20 ng/mL for 3 h by FACS, with isotype antibody used as a negative control

FN14 depletion reduces MET-driven cell migration and invasive capacity in vitro

sufficient to induce cell migration at 5 h in H2073 cells (data not shown). Thus MET-driven motility, either through receptor amplification (H1993) or HGF exposure (H2073), is significantly suppressed when FN14 is depleted.

We next investigated whether depletion of FN14 expression by shRNA could suppress MET-driven NSCLC cell migration and invasion. H1993 (metastasis derived) cells are highly motile compared to H2073 (primary tumor derived) cells [7]. H1993 cells stably expressing shRNA targeting the FN14 transcript showed *90–95 % decreased FN14 (Fig. 7a), and were found to be significantly less migratory and invasive in comparison to empty vector (shCont) containing cells. Specifically, the knockdown of FN14 by shRNA significantly suppressed H1993 cell migration by 60 % (Fig. 7b) and cell invasion through Matrigel by 55 % (Fig. 7c). The reduction observed with FN14 depletion is similar to that observed between H1993 and patient-matched H2073 cells [7]. To further define the role of FN14 in MET-driven cell migration, H2073 cell migration was induced by HGF exposure. Figure 7d shows that transient transfection of siRNA constructs targeting FN14 suppressed FN14 protein expression in the presence of HGF exposure by western blot analysis. As the FN14 target sequences siRNA-FN14-4 and -5 resulted in the most robust depletion of FN14 protein compared to cells treated with siRNA-Luci (control), these siRNAs were used in a migration assay. Figure 7e demonstrates a statistically significant reduction (*60 %) of HGF-induced H2073 cell migration when FN14 is depleted by siRNA constructs compared to control (siRNA-Luci). We have previously determined that HGF is

123

Discussion In this study, we showed that FN14 and MET proteins and mRNA message levels are significantly correlated in primary NSCLC lung tumor specimens. Furthermore, we demonstrated that the protein expression of FN14 and MET are increased in metastatic tumors compared to patientmatched primary lung tumors. Employing an in vitro model system of patient-matched primary and metastatic cell lines, we found that FN14 protein expression was significantly enhanced in a metastatic NSCLC line, which harbors MET amplification, compared to the patient-matched primary lung-derived NSCLC line. Suppression of MET activation reduced FN14 protein levels. Lastly, depletion of FN14 was sufficient to significantly inhibit cell migration and invasion driven by MET activation through receptor amplification of HGF exposure. The HGF/MET signaling axis is established as a driver of NSCLC progression, therapeutic resistance, and metastasis [29, 30]. Here we confirm that metastatic lesions express elevated MET protein compared to patient-match primary lung tumors. We also observed enhanced MET staining in a metastatic mouse model driven by mutant

Clin Exp Metastasis

B

A

C

TUBA1A

120

100 80 p < 0.005

60 40 20

Avg Cells Invaded per Field

FN14

Avg. Cells Migrated per Field

120

100 80 p < 0.05 60 40 20 0

0 shCont

E Fn14

GAPDH

Avg. Cells Migrated per Field

D

shCont

shFn14

shFn14

70 60 50 40 30

p < 0.01

p < 0.01

20 10 0 siRNA-Luci

siRNA-FN14-4

siRNA-FN14-5

Fig. 7 Depletion of FN14 expression reduces MET-driven cell migration in vitro. a H1993 cells were infected with lentiviruses expressing FN14 shRNA or control non-targeting shRNA. Stable cell lines were isolated, cell lysates prepared, and immunoblotting was conducted using the indicated antibodies to FN14 or TUBA1A (loading control). b Transwell migration assay of H1993 cell lines expressing control shRNA or FN14 shRNA. c Matrigel invasion assay of H1993 cell lines expressing control shRNA or FN14 shRNA. For migration and invasion assays, the values shown are mean ± SEM of triplicate chambers, and all experiments were duplicated. A p value \0.05, and was considered statistically significant by Student’s t test.

d H2073 cells were transfected with siRNA constructs targeting FN14 (siRNA-3, -4, or -5 or a control siRNA targeting luciferase (siRNALuci). 48 h post-transfection, cell lysates prepared, and immunoblotting was conducted using the indicated antibodies to FN14 or GAPDH (loading control). e Transwell migration assay of H2073 cells transfected as described in (d). Cells were serum-starved overnight and seeded in triplicate. HGF (20 ng/mL) was loaded in the bottom chamber as a chemoattractant. The bars represent the mean ± SEM of triplicate chambers. A p value \0.05, and was considered statistically significant by Student’s t test

kRas. Webb et al. demonstrated that kRas-driven cell transformation could lead to MET expression and enhanced cell invasion [31]. Interestingly, nuclear staining of MET was observed in tumors from the KRASG12D/ LKB1lox/lox mice, known to produce aggressive NSCLC tumors. Matteucci et al. [32] observed nuclear localization of activated MET in a metastatic breast cancer cell line. We further determined that in adenocarcinoma and squamous cell carcinoma NSCLC, FN14 protein and mRNA expression was significantly correlated to MET. Moreover, the correlation of FN14 to MET in both metastatic tumors and metastasis-derived cell lines further confirms a role for MET in lung cancer metastasis and suggests a role for FN14 in this invasive phenotype. The pharmacologic suppression of MET activation in H1993 cells was sufficient to reduce FN14 protein levels, and either MET inhibition [7] or stable FN14 depletion significantly decreased cell motility, thus highlighting the inter-dependence of FN14induced NSCLC cell invasion on MET activity in some

lung tumors. The observed correlation between FN14 and MET suggests that the therapeutic targeting of FN14 in NSCLC tumors that are driven by MET, in particular, those advanced/metastatic tumors for which no current therapeutic is effective, could be of clinical benefit for NSCLC patients. FN14 protein expression is highly expressed in primary human NSCLC tumors [14]. Several reports have demonstrated elevated protein levels of FN14 in primary human tumors across a wide array of solid malignancies [33]. In NSCLC, FN14 expression in primary NSCLC tumors correlated with activated EGFR [14], a known driver of tumor growth and progression, and is associated with poor prognosis. In this report, we showed that FN14 protein levels are enhanced in metastatic lesions compared to patient-matched, primary lung tumors. Our data are consistent with reports demonstrating elevated FN14 levels in advanced GB tumors [10, 11] and metastatic breast tumors [34, 35]. In fact, ectopic expression of human FN14 in

123

Clin Exp Metastasis

A549 NSCLC cells enhanced experimental metastasis in vivo [14]. Interestingly, FN14 was observed to be elevated in both the primary and metastatic lesions from the KRASG12D/LKB1lox/lox mice (Fig. 3). The protein expression of FN14 is elevated by mutant KRAS expression in lung cancer cell lines and normal rat lung cells and may explain the expression of FN14 in the primary tumors [14]. Thus, FN14 detection in primary tumors may be exploited as a biomarker for metastatic risk, and the elevated protein expression of FN14 in metastatic tumors represents a therapeutic opportunity for FN14 inhibition. Our report is the first to described the correlation between MET and FN14 expression in both primary patient NSCLC tumors as well as in metastatic NSCLC lesions. This is both in keeping with other labs who have also demonstrated the association of these cell surface receptors within a metastatic phenotype in NSCLC [7, 14, 29] and importantly further adds support for the early targeting of these receptors in NSCLC patient clinical management. Moreover, each target represents a therapeutic opportunity in advanced tumor stages, which are known for dismal fiveyear survival rates [8]. Several inhibitors of MET are in clinical development, including onartuzumab and MGCD265 [3, 36]. Inhibitors to the TWEAK/FN14 pathway are also in pre-clinical and clinical development. Two recent articles have highlighted the potential of a toxic immunoconjugate targeting FN14 to have anti-tumor effects against solid tumors [17] and malignant melanoma expressing FN14 [37]. Thus, the inhibition of MET and/or FN14 may be readily feasible for the therapeutic targeting of invasive or metastatic tumors cells, which have a high propensity for the coincident elevated expression of these receptors. Acknowledgments We would like to thank Serdar Tuncali, Andrew Nelson, Irene Cherni, and Guy Raz for technical assistance and Drs. Hideo Yagita and Jennifer Michaelson (Biogen Idec) for providing the FN14 monoclonal antibody. This work was supported in part by the NIH grant, R01 CA130940 (N.L.T.), and grants from the St, Joseph’s Foundation and American Lung Association, RG-224607-N (L.J.I.).

References 1. Jemal A et al (2011) Global cancer statistics. CA Cancer J Clin 61(2):69–90 2. Heist RS, Engelman JA (2012) SnapShot: non-small cell lung cancer. Cancer Cell 21(3):448 e2 3. Stella GM, Benvenuti S, Comoglio PM (2010) Targeting the MET oncogene in cancer and metastases. Expert Opin Investig Drug 19(11):1381–1394 4. Tsuta K et al (2012) c-MET/phospho-MET protein expression and MET gene copy number in non-small cell lung carcinomas. J Thorac Oncol 7(2):331–339 5. Yang Y et al (2008) A selective small molecule inhibitor of c-Met, PHA-665752, reverses lung premalignancy induced by mutant K-ras. Mol Cancer Ther 7(4):952–960

123

6. Bean J et al (2007) MET amplification occurs with or without T790 M mutations in EGFR mutant lung tumors with acquired resistance to gefitinib or erlotinib. Proc Natl Acad Sci U S A 104(52):20932–20937 7. Benedettini E et al (2010) Met activation in non-small cell lung cancer is associated with de novo resistance to EGFR inhibitors and the development of brain metastasis. Am J Pathol 177(1):415–423 8. Navab R et al (2009) Co-overexpression of Met and hepatocyte growth factor promotes systemic metastasis in NCI-H460 nonsmall cell lung carcinoma cells. Neoplasia 11(12):1292–1300 9. Feng SL et al (2000) The Fn14 immediate-early response gene is induced during liver regeneration and highly expressed in both human and murine hepatocellular carcinomas. Am J Pathol 156(4):1253–1261 10. Tran NL et al (2003) The human Fn14 receptor gene is up-regulated in migrating glioma cells in vitro and overexpressed in advanced glial tumors. Am J Pathol 162(4):1313–1321 11. Tran NL et al (2006) Increased fibroblast growth factor-inducible 14 expression levels promote glioma cell invasion via Rac1 and nuclear factor-kappaB and correlate with poor patient outcome. Cancer Res 66(19):9535–9542 12. Watts GS et al (2007) Identification of Fn14/TWEAK receptor as a potential therapeutic target in esophageal adenocarcinoma. Int J Cancer 121(10):2132–2139 13. Willis AL et al (2008) The fibroblast growth factor-inducible 14 receptor is highly expressed in HER2-positive breast tumors and regulates breast cancer cell invasive capacity. Mol Cancer Res 6(5):725–734 14. Whitsett TG et al (2012) Elevated expression of Fn14 in nonsmall cell lung cancer correlates with activated EGFR and promotes tumor cell migration and invasion. Am J Pathol 181(1):111–120 15. Fortin SP et al (2009) Tumor necrosis factor-like weak inducer of apoptosis stimulation of glioma cell survival is dependent on Akt2 function. Mol Cancer Res 7(11):1871–1881 16. Tran NL et al (2005) The tumor necrosis factor-like weak inducer of apoptosis (TWEAK)-fibroblast growth factor-inducible 14 (Fn14) signaling system regulates glioma cell survival via NFkappaB pathway activation and BCL-XL/BCL-W expression. J Biol Chem 280(5):3483–3492 17. Zhou H et al (2011) Development and characterization of a potent immunoconjugate targeting the Fn14 receptor on solid tumor cells. Mol Cancer Ther 10(7):1276–1288 18. Knudsen BS et al (2009) A novel multipurpose monoclonal antibody for evaluating human c-Met expression in preclinical and clinical settings. Appl Immunohistochem Mol Morphol 17(1):57–67 19. Jackson EL et al (2001) Analysis of lung tumor initiation and progression using conditional expression of oncogenic K-ras. Genes Dev 15(24):3243–3248 20. Bardeesy N et al (2002) Loss of the Lkb1 tumour suppressor provokes intestinal polyposis but resistance to transformation. Nature 419(6903):162–167 21. DuPage M, Dooley AL, Jacks T (2009) Conditional mouse lung cancer models using adenoviral or lentiviral delivery of Cre recombinase. Nat Protoc 4(7):1064–1072 22. Nakayama M et al (2003) Fibroblast growth factor-inducible 14 mediates multiple pathways of TWEAK-induced cell death. J Immunol 170(1):341–348 23. Fortin Ensign SP et al (2013) The Src homology 3 domain-containing guanine nucleotide exchange factor is overexpressed in high-grade gliomas and promotes tumor necrosis factor-like weak inducer of apoptosis-fibroblast growth factor-inducible 14-induced cell migration and invasion via tumor necrosis factor receptor-associated factor 2. J Biol Chem 288(30):21887–21897

Clin Exp Metastasis 24. Chuang YY et al (2004) Role of synaptojanin 2 in glioma cell migration and invasion. Cancer Res 64(22):8271–8275 25. Ji H et al (2007) LKB1 modulates lung cancer differentiation and metastasis. Nature 448(7155):807–810 26. Chen Z et al (2012) A murine lung cancer co-clinical trial identifies genetic modifiers of therapeutic response. Nature 483(7391):613–617 27. Franks LM et al (1976) Metastasizing tumors from serum-supplemented and serum-free cell lines from a C57BL mouse lung tumor. Cancer Res 36(3):1049–1055 28. Layton MG, Franks LM (1984) Heterogeneity in a spontaneous mouse lung carcinoma: selection and characterisation of stable metastatic variants. Br J Cancer 49(4):415–421 29. Birchmeier C et al (2003) Met, metastasis, motility and more. Nat Rev Mol Cell Biol 4(12):915–925 30. Gherardi E et al (2012) Targeting MET in cancer: rationale and progress. Nat Rev Cancer 12(2):89–103 31. Webb CP et al (1998) Evidence for a role of Met-HGF/SF during Ras-mediated tumorigenesis/metastasis. Oncogene 17(16):2019– 2025

32. Matteucci E, Bendinelli P, Desiderio MA (2009) Nuclear localization of active HGF receptor Met in aggressive MDA-MB231 breast carcinoma cells. Carcinogenesis 30(6):937–945 33. Winkles JA (2008) The TWEAK-Fn14 cytokine-receptor axis: discovery, biology and therapeutic targeting. Nat Rev Drug Discov 7(5):411–425 34. Sanz-Pamplona R et al (2011) Expression of endoplasmic reticulum stress proteins is a candidate marker of brain metastasis in both ErbB-2 ? and ErbB-2- primary breast tumors. Am J Pathol 179(2):564–579 35. Wang J et al (2013) Clinical correlations and prognostic relevance of Fn14 expression in breast carcinoma. Histol Histopathol 28(7):859–864 36. Eder JP et al (2009) Novel therapeutic inhibitors of the c-Met signaling pathway in cancer. Clin Cancer Res 15(7):2207–2214 37. Zhou H et al (2013) The TWEAK receptor Fn14 is a therapeutic target in melanoma: immunotoxins targeting Fn14 receptor for malignant melanoma treatment. J Invest Dermatol 133(4): 1052–1062

123