Background
The epidermal growth factor receptor (EGFR, ErbB1, HER1) is the prototypic member of the ErbB family of receptor tyrosine kinases (TKs), which further consists of ErbB2-4 (HER2-4). The ErbB receptors share a similar protein structure, consisting of an extracellular ligand binding domain, a single transmembrane domain and an intracellular C-terminal domain with tyrosine kinase activity [
1]. Upon specific binding of EGF-like ligands to the extracellular domain, ErbB receptors dimerize, either as homo- or heterodimers, and undergo autophosphorylation at specific tyrosine residues within the intracellular domain. The phosphorylated tyrosines serve as docking sites for adapter molecules, such as Grb2 and the p85 subunit of PI3K, which activate a complex downstream network. The activated signaling pathways, including the Ras/MAPK, Akt/mTOR kinase and STAT cascades, in turn regulate transcription factors and other proteins involved in cell proliferation, survival, motility and differentiation [
2]. Two main strategies targeting ErbB receptors have been developed: small-molecule inhibitors of the tyrosine kinase domain (EGFR tyrosine kinase inhibitors [TKIs], such as erlotinib and gefitinib), and monoclonal antibodies (such as cetuximab, anti-EGFR and trastuzumab, anti-HER2), directed against the extracellular domain, which inhibit phosphorylation/activation and promote internalization. EGFR and HER2 are overexpressed in 40-80% and 25-30%, respectively, of non-small cell lung cancer (NSCLC) patients and their overexpression has been frequently correlated with a poor prognosis [
3,
4].
Erlotinib is an effective treatment for NSCLC patients and has been registered as a second and third-line treatment of NSCLC regardless of EGFR mutation status [
5].
Gefitinib has been registered for the therapy of advanced NSCLC harbouring activating EGFR mutations in the tyrosine kinase domain, the most frequent being L858R in exon 21 and Del (746–750) in exon 19 [
6]. Although mutations in EGFR are useful predictors for the activity of EGFR-TKI, they cannot be used as the only criterion to determine who should receive anti-EGFR therapy and it is becoming increasingly clear that even patients with EGFR wild-type can benefit from EGFR-TKI [
5,
7,
8].
Cetuximab is a chimeric IgG1 monoclonal antibody (mAb) that blocks ligand binding to EGFR, leading to a decrease in receptor dimerization, autophosphorylation, and activation of signaling pathways [
9]. In addition the binding of cetuximab initiates EGFR internalization and degradation which leads to signal termination. Moreover, unlike EGFR-TKIs, cetuximab can induce antibody dependent cellular cytotoxicity (ADCC) activity, an important immunologic antitumour effect. Cetuximab in combination with chemotherapy has been approved by the FDA for the treatment of metastatic colorectal cancer and of locally advanced head and neck cancer.
Two randomized phase III trials in NSCLC patients, evaluating cetuximab in addition to first-line chemotherapy, showed a small benefit in overall survival for the experimental treatment, which was considered insufficient by the EMA for marketing approval [
10,
11]. However, a subgroup analysis of the FLEX phase III trial recently demonstrated a larger survival benefit from the experimental treatment in patients with high immunohistochemical EGFR expression [
12].
Trastuzumab, registered for the treatment of HER2 positive breast cancer, has also been tested in phase II trials as a single agent and in combination with cytotoxic chemotherapy for patients with NSCLC. These trials have not yet produced any convincing evidence of an improved antitumour activity by adding trastuzumab to standard chemotherapy in NSCLC [
13,
14].
Several preclinical studies on cell lines from different tumour types, indicated that the association between EGFR/HER2 mAbs with TKIs displays an increased efficacy [
15].
In this study we explored the potential of combining erlotinib with either cetuximab or trastuzumab in order to improve the efficacy of EGFR targeted therapy in EGFR wild-type sensitive NSCLC cell lines. Our results indicate that EGFR-TKI increases surface expression of EGFR and/or HER2 only in erlotinib sensitive NSCLC cell lines and, in turns, leads to increased susceptibility to ADCC both in vitro and in xenograft models.
Discussion
The potential for dual-agent molecular targeting of the ErbB family, has been clearly demonstrated in pre-clinical models and confirmed on the clinical setting for HER2-targeting agents in breast cancer. However, little is known about this therapeutic strategy for different targets in other tumour types.
In our current study we demonstrated that the combination of erlotinib with cetuximab or trastuzumab may enhance the antitumour activity of EGFR-TKI in NSCLC cell lines harbouring wild-type EGFR and in xenograft models.
The efficacy of the association between an EGFR/HER2 mAbs with TKIs has been documented in preclinical studies in several cell lines originating from different tumour types [
15]. In EGFR wild-type H292 and A549 NSCLC cell lines, the combination of either gefitinib or erlotinib with cetuximab was reported to enhance growth inhibition in comparison to single treatment, particularly in the H292 gefitinib sensitive cell line [
17]. In the A549 cell line, expressing both EGFR and HER2, the combination of gefitinib with trastuzumab significantly inhibited cell growth and proliferation [
18]. In Calu-3 xenograft models, the combined treatment of erlotinib and pertuzumab showed an enhanced antitumour activity [
19].
A correlation between cetuximab efficacy and EGFR expression has been reported in preclinical studies [
20] and recently confirmed in clinical trials. Thus, the phase III FLEX study involving patients with advanced NSCLC showed a strong correlation between high tumour EGFR overexpression and the efficacy of adding cetuximab to platinum based first-line chemotherapy [
12].
The combination of a TKI and a mAb was explored as a potential strategy to overcome acquired resistance to first-generation EGFR-TKIs. Kim and colleagues demonstrated that the combination of lapatinib with cetuximab overcame gefitinib resistance due to the secondary T790M mutation in NSCLC by inducing enhanced cytotoxicity both
in vitro and
in vivo [
21]. Furthermore, the association of cetuximab with afatinib has been shown to be effective to overcome T790M-mediated drug resistance [
22].
However, the combination of erlotinib with cetuximab did not lead to a significant radiological response in NSCLC patients with clinically defined acquired resistance to erlotinib indicating that such strategy is not sufficient to overcome acquired resistance to erlotinib [
23].
The mechanisms leading to an enhanced activity of combining a TKI with a monoclonal antibody have been ascribed, in other cancer cell models, either to a more efficient inhibition of TK receptors [
17] or to an increased targeted receptors on plasma membrane induced by TKIs [
24,
25]. Scaltriti et al. showed that lapatinib enhanced the effects of trastuzumab by inducing HER-2 stabilization and accumulation at the cell surface of breast cancer cell lines [
24], and Mimura et al. reported that lapatinib induced accumulation of HER-2 and EGFR on esophageal cancer cell lines evoking trastuzumab- and cetuximab- mediated ADCC [
25].
ADCC, one of the killing mechanism of the immune system mediated by Natural Killer cells, plays a pivotal role in the anti-cancer effects exerted by mAbs. Therefore, increasing the ADCC activity is an important objective in the development of novel therapeutic approaches.
It has been recently demonstrated that the EGFR inhibitors gefitinib and erlotinib enhance the susceptibility to NK cell mediated lysis of A549, NCI-H23 and SW-900 lung cancer cell lines [
26] by the induction of ULBP1 (a ligand of the NK cell activation receptor NKG2D). These data indicate that EGFR blockade could not be the only mechanism of action of EGFR inhibitors in vivo. The efficacy of these inhibitors in lung cancer may be at least in part mediated by increased susceptibility to NK activity. Moreover, cetuximab serves as a potent stimulus for NK functions including INF-gamma production [
27] and is also associated with a complement –mediated immune response [
28].
We here demonstrated that erlotinib induces an accumulation of EGFR and/or HER2 protein at the plasma membrane level only in TKI sensitive NSCLC cell lines whereas, in resistant cells (both, intrinsic or MET amplification-mediated acquired resistance), this enhancement was not observed. The anti-tumour effect of drug combination was more evident in ADCC experiments compared with cell viability experiments. In the Calu-3 xenograft model, the combined treatment resulted in a lower rate of tumour growth, suggesting the involvement of NK activity as a determinant factor to improve the efficacy of the combined treatment. Moreover, regressive phenomena and changes in size of neoplastic glands together with intense stromal reaction were observed in histologic samples of tumours from mice treated with cetuximab alone or the combination.
The reason why EGFR inhibitors such as gefitinib, erlotinib or lapatinib induce EGFR accumulation only in sensitive cells could be ascribed to their ability to inhibit signal transduction pathways downstream EGFR. The constitutive activation of signaling pathways downstream of EGFR (i.e. presence of RAS mutations) is indeed a recognized mechanism of resistance against reversible EGFR-TKIs [
29]. The inhibition of the MAPK pathway might represent a link between EGFR inhibition and EGFR accumulation since U0126, a well known MEK1/2 inhibitor, induced EGFR accumulation in Calu-3 cells, while none of PI3K/AKT/mTOR inhibitors tested was effective. A correlation between MAPK pathway and protein degradation by the ubiquitin system was described for the pro-apoptotic BH3-only protein BIM, indeed in the absence of MAPK activation, BIM protein accumulated in the cell promoting activation of apoptotic cell death [
30].
Considering that EGFR TKIs, in particular erlotinib, demonstrated to be effective only in a small percentage of NSCLC patients not harboring
EGFR mutations, our preclinical results could support clinical trials on the combinations of erlotinib and cetuximab or trastuzumab aiming to improve treatment efficacy. Although the addition of cetuximab to erlotinib is insufficient to overcome erlotinib resistance in
EGFR-driven lung adenocarcinoma [
23], the clinical potential of dual-agent molecular targeting of the EGFR in patients with
EGFR wild-type tumours remains to be elucidated and may represents an interesting research area to be pursued.
Methods
Cell culture
The human NSCLC cell lines H322, H292, Calu-3, H1299, A549, H1703 and Calu-1 were obtained from American Type Culture Collection (Manassas, VA, USA) and were cultured as recommended. The PC9, HCC827 and HCC827GR5 cell lines were kindly provided by Dr P. Jänne (Dana-Farber Cancer Institute, Boston MA, USA). All cells were maintained under standard cell culture conditions at 37°C in a water-saturated atmosphere of 5% CO
2 in air. As previously reported [
31] cells showing by proliferation assays IC
50 for erlotinib < 1 μM were considered sensitive (H322, H292, Calu-3, PC9, HCC827) while cell lines with IC
50 > 5 μM (H1299, A549, H1703, Calu-1, HCC827GR5) were considered resistant.
Drug treatment
Erlotinib, gefitinib, cetuximab, trastuzumab and rituximab were from inpatient pharmacy. RAD001, NVP-BKM-120 and NVP-BYL-719 were from Novartis.
Stock solutions of 20 mM drugs were prepared in dimethylsulfoxide (DMSO) (with the exception of mAbs), stored at −20°C and diluted in fresh medium for use. The final concentration of DMSO never exceeded 0.1% v/v.
Western blot analysis
Procedures for protein extraction, solubilization, and protein analysis by 1-D PAGE are described elsewhere [
32,
33]. Fifty μg of proteins from lysates were resolved by 7.5% SDS-PAGE and transferred to PVDF membranes. Membranes were incubated with: 1:1000 rabbit polyclonal anti-EGFR; 1:1000 rabbit mAb anti-HER2/ErbB2; 1:1000 rabbit mAb anti-Phospho-p70S6K (Thr421/Ser424); 1:1000 mouse mAb anti-Phospho-p44/42 MAPK (ERK1/2); 1:1000 rabbit mAb anti-p44/42 MAPK (ERK1/2) (Cell Signaling Technology, Beverly, MA, USA); 1:1000 mouse mAb anti- Transferrin Receptor (Invitrogen Corporation, Camarillo, CA, USA); 1:3000 mouse mAb anti-Actin (Sigma–Aldrich, St Louis, MO, USA). Blots were then washed and incubated with HRP-anti-mouse or HRP-anti-rabbit antibodies at 1:20000 dilution (Pierce, Rockford, IL, USA). Immunoreactive bands were visualized using an enhanced chemiluminescence system (Immobilion™ Western Cemiluminescent HRP Substrate, Millipore USA).
Cell surface protein isolation
Calu-3 cells were grown in T75 flasks and treated with 0.5 μM erlotinib for 24 h. Cells were incubated with EZ-LINK Sulfo-Biotin (Pierce) for 2 h at 4°C with gentle rotation. The reaction was stopped by washing twice with 25 mM Tris–HCl (pH 7.5) in PBS (phosphate-buffered saline) and cells were scraped into ice-cold lysis buffer (50 mmol/l HEPES, pH 7.0, 10% glycerol, 1% TritonX-100, 5 mmol/l EDTA (ethylenediaminetetraacetic acid), 1 mmol/l MgCl2, 25 mmol/l NaF, 50 μg/ml leupeptin, 50 μg/ml aprotinin, 0.5 mmol/l orthovanadate, and 1 mmol/l phenylmethylsulfonyl fluoride). Lysates were centrifuged at 15000 g for 20 min at 4°C, and supernatants were removed and assayed for protein concentration using the DC Protein assay (Bio-Rad, CA, USA). A volume of 500 μl of lysis buffer containing equal amount of proteins was incubated with UltraLink Immobilized NeutrAvidin protein (Pierce) for 2 h at 4°C with gentle rotation, washed three times with lysis buffer before suspension in SDS (sodium dodecyl sulfate)-loading buffer and then resolved by SDS-PAGE.
Flow cytometry
For the determination of EGFR and HER2 protein membrane levels, NSCLC cell lines H322, Calu-3 and H292 were treated with 1 μM erlotinib for 24 h. One million cells per condition were then incubated with Isotype control Monoclonal Mouse IgG1/R-PE (Ancell IRP, Bayport, MN, USA), PE mouse anti-Human EGFR (Calu-3 and H322 cells) (BD Biosciences, San Josè, CA, USA) or PE mouse anti-Human HER2 (H322 and H292) (BD Biosciences). After the incubation the analysis was performed with an EPICS-XL flow cytometer.
For the relative quantization of EGFR or HER2 binding sites, NSCLC cell lines H322, Calu-3, H292 were treated with 1 μM erlotinib for 24 h. One million cells were then dispensed for each condition and treated with either 20 μg/ml rituximab (Isotype control), cetuximab (Calu-3 and H322) or trastuzumab (H292) for 1 h. After the incubation with PE-anti-human-IgG (BD Biosciences), the analysis was performed with an EPICS-XL flow cytometer.
The values of mean fluorescence intensity (MFI) were converted in units of equivalent fluorochrome (MEF) using the FluoroSpheres 6-Peak Kit (Dako, CA, USA).
Quantitative real-time PCR
Total RNA was isolated by the TRIzol® reagent (Invitrogen, Carlsbad, CA, USA) and reverse transcribed as previously described [
33].
The transcript levels of EGFR gene were assessed by Real-Time qRT-PCR on an iCycler iQ Multicolor RealTime PCR Detection System (Bio-Rad, Hercules, CA, USA).
Primers and probes included: EGFR-F (5′-GCCTTGACTGAGGACAGCA-3′), EGFR-R (5-TTTGGGAACGGACTGGTTTA-3), EGFR-probe (5′-FAM CTTCCTCC3′DQ); PGK1-F (5′-GGAGAACCTCCGCTTTCAT-3′), PGK1-R (5′-CTGGCTCGGCTTTAACCTT-3′), PGK1-probe (5′-FAM GGAGGAAG 3′DQ); RPL13-F (5′-ACAGCTGCTCAGCTTCACCT-3′), RPL13-R (5′-TGGCAGCATGCCATAAATAG-3′), RPL13-probe (5′-FAMCAGTGGCA3′DQ); HPRT-F (5′-TGACCTTGATTTATTTTGCATACC-3′), HPRT-R (5′CGAGCAAGACGTTCAGTCCT-3′), HPRT-probe (5′-FAM GCTGAGGA 3′DQ).
The amplification protocol consisted of 15 min at 95°C followed by 40 cycles at 94°C for 20s and at 60°C for 1 min.
The relative transcript quantification was calculated using the geNorm algorithm for Microsoft Excel™ after normalization by expression of the control genes [phosphoglycerate kinase1 (PGK1), ribosomal protein L13 (RPL13) and hypoxanthine-guanine-phosphoribosyltransferase (HPRT)] and expressed in arbitrary units (a.u.).
MTT assay
The cells were seeded into 96-well plate in quadruplicate and were exposed to various treatments. After 96 h, 100 μl of 3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyltetrazolium bromide (MTT) solution (1 mg/ml, Sigma-Aldrich) was added to each well and incubated. After 4 h, crystalline formation was dissolved with DMSO and the absorbance at 570 nm was measured using the microplate-reader 550 (BioRad).
Isolation and culture of NK cells
Human PBMC were isolated from buffy coat of healthy donors by using a Lympholyte-H density gradient centrifugation (Cederlane Burlington, Ontario, Canada). Highly purified CD56+ natural killer (NK) cells were obtained by magnetic separation using the NK Cell Isolation Kit and the autoMACS Separator (Miltenyi Biotec, Cologne, Germany) according to the user manual.
Purified NK cells were resuspended in culture medium (RPMI 1640 without phenol red and supplemented with heat inactivated 10% FCS, 50 U/ml penicillin, 50 U/ml streptomycin, 2 mmol/l glutamine) plated and preincubated at 37°C for up to 18 h in the presence of human Interleukin-2 (IL-2, 100 U/ml, Miltenyi Biotec).
ADCC assay
Antibody-dependent cell-mediated cytotoxicity (ADCC) was measured with the CytoTox 96 non-radioactive cytotoxicity assay (Promega, Madison, WI, USA) according to manufacturer’s instructions. 2x10
3 Calu-3, H322, H292 or H1299 cells were treated for 24 h with 1 μM erlotinib, and then seeded with purified NK cells (ratio of 1:25 and 1:50) in a 96-well plate and incubated with 10 μg/ml cetuximab or trastuzumab. After 4 hours the lactate dehydrogenase (LDH) release was determined and the percentage of cytotoxicity was calculated after correcting for background absorbance values according to the following formula:
(1)
Tumour xenografts
All experiments involving animals and their care were performed with the approval of the Local Ethical Committee of University of Parma, in accordance with the institutional guidelines that are in compliance with national (DL116/92) and international (86/609/CEE) laws and policies. Twenty-four Balb/c-Nude female mice (Charles River Laboratories, Calco, Italy) were housed in a protected unit for immunodeficient animals with 12-hour light/dark cycles and provided with sterilized food and water ad libitum. At the time of xenograft establishment, mice were 8 weeks old and weighted ~20g. 200 μl of matrigel (BD Biosciences) and sterile PBS (1:1) containing 1x107 Calu-3 cells, were subcutaneously injected on the right flank of each mouse (using a 1 ml syringe, needle G25). When tumour volume reached an average size of 300 mm3, 14 days after injection, animals were randomized into four groups and the treatment started. After 4 weeks, mice were euthanized by cervical dislocation and tumours collected for immunohistochemistry and histological analysis.
Erlotinib (25 mg/Kg) was administered orally in 1% methylcellulose, 0.2% Tween 80 in sterilized water 5 days/week. Cetuximab (2 mg/Kg) was intraperitoneally injected in sterile saline solution 2 days/week. Control group received both oral gavage of 1% methylcellulose, 0.2% Tween 80 in sterilized water 5 days/week and i.p. injection of sterile saline solution (0.9%) 2 days/week.
Dosages of drugs were chosen halving the one used in a previous study in NSCLC-xenograft models, in order to avoid the complete inhibition of tumour growth by the single agent treatment and to better highlight the effect of erlotinib-cetuximab combination [
19,
34].
Tumour xenografts were measured twice a week, tumour volume was determined using the formula: (length x width2)/2. Final data are expressed as percent of volume increase: (tumour volume/pre-treatment tumour volume) x 100.
Morphometric and immunohistochemical analysis of tumour xenografts
Formalin fixed samples were embedded in paraffin. From each tumour serial sections of 5 μm thickness were obtained and stained with Haematoxylin and Eosin (H&E), Masson’s Trichrome and for immunohistochemistry. Morphometric analysis was performed in order to evaluate: (a) the numerical density of neoplastic cells, (b) the volume fraction of interstitial inflammatory cells, (c) the volume fraction of fibrosis and (d) the fraction of proliferating and apoptotic cells.
In particular, for each section stained with H&E, a quantitative evaluation of tissue composition was performed. To better define the fraction occupied by neoplastic and non neoplastic cells, sections were stained with pancytokeratin antibodies (monoclonal mouse, 1:500, o.n. 4°C, Dako) revealed through biotin-streptavidin-DAB system, as repeatedly described. The numerical density (n/mm2) of pancytokeratin positive neoplastic cells was computed.
In addition, cell proliferation and apoptotic death were investigated by fluorescence microscopy. Thus, Ki67 labeling (rabbit polyclonal antibody, Vector) and the Terminal deoxynucleotidyltransferase (TdT)–mediated dUTP nick end labeling (TUNEL) assay (Roche Diagnostics, Italy) on cytokeratinpos neoplastic cells were revealed by specific fuorescent probes.
The area occupied by interstitial cells was expressed as percentage of the total area explored. By the same approach, the volume fraction of fibrosis was calculated on Masson’s Trichrome stained sections. To define the volume fractions, the number of points overlying each tissue components was counted and expressed as percentage of the total number of points explored.
All these morphometric measurements were obtained with the aid of a grid defining a tissue area of 0.23 mm2 and containing 42 sampling points each covering an area of 0.0052 mm2.
All these evaluations were performed on the entire section of each tumour sample of each experimental group of animals using an optical microscope (250X final magnification).
Statistical analysis
Statistical analyses were carried out using GraphPad Prism version 5.0 software (GraphPad Software Inc., San Diego, CA, USA). Results are expressed as mean values ± standard deviations (SD) for the indicated number of independent measurements. Differences between the mean values recorded for different experimental conditions were evaluated by Student’s t-test, and P values are indicated where appropriate in the figures and in their legends. A P value <0.05 was considered as significant.
For in vivo studies comparison among groups was made using analysis of variance (two-way ANOVA repeated measures) followed by Bonferroni’s post-test. Analysis was performed using Prism 5.0 (GraphPad Software) and differences were considered significant when P value was below 0.05. The nature of the interaction between erlotinib and cetuximab was calculated using the Bliss interaction model [
35].
Competing interest
All authors declare that they have no competing interests.
Authors’ contribution
AC carried out ADCC experiments, interpreted the results and assisted with the draft of the manuscript; DC isolated and cultured NK cells and carried out flow cytometry analysis; FS, LA and PC performed the in vivo studies; PM carried out RT-PCR experiments; MG and MB carried out Western blot analysis; EG performed the statistical analysis, SLM and CF carried out cell growth experiments; FQ, GG and DM carried out immunohistochemical analysis; AM, MT, EB and PGP critically revised the manuscript; AA designed the project and assisted with the draft of the manuscript; RRA, analyzed the results and wrote the manuscript. All authors read and approved the final manuscript.