Background
Immunotherapy is a promising strategy for increasing survival in cancer patients, and has been an active area research for decades. Amongst various types of immunotherapy, the induction of anti-tumor CD8 cytotoxic T lymphocyte (CTL) responses via vaccination with peptide epitopes has been widely applied in the clinical setting [
1]. Unfortunately, CTL vaccines have not yet yielded clear favorable clinical results for treating cancer, possibly due to a combination of suboptimal immune responses and to tumor-derived immune suppressive activities.
Many strategies have been applied to enhance antigen-specific anti-tumor immunity, including the activation of natural killer (NK) cells, conversion of macrophage phenotype, and immune-modulating adjuvants [
2]-[
4]. Among these, the blockade of immunological checkpoints such as CTLA-4/B7 and PD-1/PD-L1 is quite advanced and has yielded promising clinical results [
5]. It is predicted that the use of non-specific anti-cancer immunity targeted therapy may be a valuable complement to tumor antigen-specific immunity against cancer.
CD4
+ helper T lymphocytes (HTLs) play a critical role in anti-cancer immunity by promoting the induction and survival of CD8
+ CTLs. In addition, in some instances HTLs can also exhibit direct anti-tumor cytotoxic activity. In view of this, our laboratories have focused on the identification of peptide epitopes capable of inducing cytotoxic HTL responses against tumor cells that express surface MHC class II molecules [
6]. Recently, long-peptide vaccines have been used with the purpose of inducing both CTL and HTL anti-tumor responses, with promising clinical results [
7].
The disruption of the antigen-processing machinery is one of the mechanisms utilized by tumor cells to evade T cell recognition. To overcome this problem, we and other groups have recently proposed that the increase of MHC class II protein expression on tumor cells obtained with EGFR inhibitors could be implemented to enhance HTL anti-tumor responses [
8],[
9]. Although EGFR inhibitors have been widely used to treat many types of cancer, the usefulness of these therapies is limited due to the appearance of drug resistance [
10],[
11]. Immune regulatory cytokine TGF-β has been reported to mediate the resistance to EGFR inhibition, however, direct activity of EGFR mediated TGF-β production from tumor toward antitumor immune cells has remained largely unknown [
12].
In this study, we discovered that EGFR inhibition although increased MHC-II expression, paradoxically it attenuated HTL responses against some head and neck squamous cell carcinoma (HNSCC) cells. We observed that secretion of TGF-β and PGE2 by the HNSCC cells was increased following EGFR inhibition, despite a lack of evident changes in immune costimulatory molecules or EGFR expression in tumor cells. Inhibition of TGF-β or COX-2 salvaged HTL responses against EGFR inhibitor-treated HNSCC cells, suggesting that these pathways played a crucial role in immunosuppression. Taken together, our results demonstrate that in some cases, EGFR inhibitors may skew the immune response towards T cell suppression, and that concomitant blockade of EGFR and TGF-β/COX-2 may be promising combinatorial therapeutic approaches for patients with EGFR-expressing tumors.
Materials and methods
Cell lines
HNSCC cell lines HSC-3, HSC-4 (tongue SCC, DR1/4) and Sa-3 (gingival SCC, DR9/10) were provided by the RIKEN Bio-Resource Center (Tsukuba, Japan). CA9-22 (gingival SCC) and HPC-92Y (hypopharyngeal SCC) were kindly provided by Dr. Yasuharu Nishimura (Dep. of Immunogenetics, Kumamoto University, Kumamoto, Japan) and Dr. Syunsuke Yanoma (Yokohama Tsurugamine Hospital, Yokohama, Japan), respectively. SAS (tongue SCC), Calu-1 (non-small cell lung carcinoma) and 5637 (bladder cancer) were purchased from American Type Culture Collection (Manassas, VA). All cell lines were maintained in RPMI 1640 (nacalai tesque, Kyoto, Japan) supplemented with 10% fetal bovine serum.
Western blotting
Cells (1 × 106) were washed in phosphate-buffered saline (PBS) and lysed in NuPAGE sample buffer (Invitrogen, CA). The lysates were subjected to electrophoresis (NuPAGE bis-Tris SDS-PAGE gel (Invitrogen, CA)) and transferred to Immobilon-P membrane (Millipore, Bedford, MA). The membrane was soaked in blocking buffer (PBS containing 5% non-fat dry milk and 0.01% Tween 20, 1 h) at room temperature. Blots were then incubated with polyclonal rabbit anti-human EGFR (sc-03; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), polyclonal rabbit anti-human phospho-EGFR (Tyr1068; Cell Signaling Technology, Denver, MA), polyclonal rabbit anti-human heat shock protein 70 (HSP70) (Enzo Life Sciences, Inc., Farmingdale, NY), or monoclonal rat anti-human heat shock protein 90 (HSP90) (Enzo Life Sciences) diluted 1:500 in blocking buffer, or anti β-actin mAb (Santa Cruz Biotechnology) diluted 1:1,000 in blocking buffer, for 18 h at 4°C. The membrane was incubated with HRP-labeled sheep anti rabbit or anti mouse IgG after washing, and made visible by an enhanced chemiluminescence (ECL) system (Amersham, Buckinghamshire, UK).
Synthetic peptides
The synthetic peptide used throughout this study was EGFR
875–889 (KVPIKWMALESILHR) [
8]. The peptide epitope was synthesized by solid-phase organic chemistry and purified by high performance liquid chromatography. The purity (80%<) and identification of peptides were assessed by high performance liquid chromatography and mass spectrometry, respectively.
Measurement of antigen-specific responses with antigen reactive CD4+ T-cell clones
Antigen-specific CD4
+ T cells were induced by peptide stimulation of fresh peripheral blood mononuclear cells (PBMCs) from healthy volunteers [
13]. EGFR
875–889-reactive CD4
+ T cell clones T8 (from an HLA-DR 9/12 individual) and M8 (from an HLA-DR 9/13 individual) were used. These clones were restricted by HLA-DR53 as recently described [
8]. CD4
+ T-cells (3 × 10
4 cells/well) were mixed in 96 well culture plates with irradiated antigen-presenting cells (APCs) that consisted of either autologous PBMCs (1 × 10
5 cells/well) or tumor cell lines (3 × 10
4 cells/well). HNSCC cells were pretreated with interferon gamma (IFN-γ, 500 U/ml, 48 h) to increase HLA-DR expression prior to the assay. To examine the role of EGFR inhibitor in augmenting the MHC class II molecules expression, HNSCC cells were preincubated with reversible EGFR tyrosine kinase inhibitor (TKI) erlotinib (1 μM; Selleck Chemicals, Houston, TX), for 2 h at 37°C before addition of IFN-γ. DMSO was used as control. Tumor cells were washed twice with PBS to eliminate the residual chemicals. Expression of the HLA-DR and B7-H1 on tumors was evaluated by flow cytometry using anti HLA-DR mAb conjugated with fluorescein isothiocyanate (FITC), anti B7-H1 mAb (eBioscience, Minneapolis, MN), and anti-mouse immunoglobulin conjugated with FITC (Dako Denmark A/S, Glostrup, Denmark). Detection of surface CD80 and CD86 was carried out using unconjugated mouse anti-human CD80 IgG1 (MAB104,Immunotech, Marseille, France) and unconjugated mouse anti-human CD86 IgG2b (HA5.2B7, Immunotech, Marseille, France) followed by FITC-conjugated rabbit anti-mouse immunoglobulin antibody (1:100; Dako, Denmark A/S, Glostrup, Denmark).
Anti-TGF-β Ab (10 μg/ml; Abcam, Tokyo, Japan), celecoxib (10 μM; Sigma-Aldrich Japan, Tokyo, Japan), recombinant PGE2 (1 μM; Sigma-Aldrich Japan) or supernatant of tumor culture were added to the co-culture medium for functional studies. CD4+ T-cells culture supernatants were collected after 48 h to quantify antigen-induced IL-4, IL-10 or IFN-γ production using ELISA assays (BD Pharmingen, San Diego, CA). Culture supernatants from erlotinib-treated tumor cells were collected for quantification of TGF-β and PGE2 using ELISA kits (TGF-β eBioscience, San Diego, CA; PGE2: R&D Systems, Inc., Minneapolis, MN).
Statistical analysis
All data are presented as mean ± standard deviation. In all experiments, group differences were analyzed by using the two-tailed Student's t test and p <0.05 was considered as statistically significant.
Discussion
We recently reported the capacity of EGFR inhibitors to augment HLA-DR surface expression on tumor cells [
8]. After HLA-DR up-regulation, most EGFR inhibitor-treated HNSCC cell lines became more susceptible to antitumor responses mediated by CD4
+ T cells. However, in the present study, we report that despite HLA-DR augmentation on the tumor cells, in some instances EGFR inhibition suppressed antitumor T cell responses by inducing the production of TGF-β and PGE2. Suppression of anti-tumor immunity was reversed by the addition of anti-TGF-β antibody or COX-2 inhibitor, supporting the rationale for inhibition of TGF-β and COX-2 pathways to overcome potential immunosuppressive effects due to EGFR inhibition. Since EGFR inhibitor plays a detrimental role in tumor proliferation, it would be assumed that the decreased number of tumor cell could affect the immune reaction because of the lesser antigen. Pretreatment with erlotinib definitely decreased tumor cell survival in our study (data not shown), however, we did not conclude this antigen reduction is the main reason of attenuated T cell responses against tumor cells. Firstly, we washed erlotinib-pretreated tumor cells before cocultured with T cells to remove residual erlotinib. Second, we found that anti TGF-? antibody clearly recovered the function of T cells. Lastly, supernatant of tumors treated with erlotinib attenuated the T cell responses suggesting that humoral factors from tumor cells affect the T cell responses in our assay. Thus, TGF-? might at least play a harmful role in antitumor T cell responses against tumor treated with EGFR inhibitor.
In this study, we found the high variability between cell lines of expressing TGF-? and PGE2 after EGFR inhibition. Because substantial heterogeneity within tumors has been elucidated [
20], it is difficult to determine a single factor that induces diversity of tumors. For example, tumor cell uses an alternative signaling such as HER-3 when EGFR signaling is inhibited [
21], suggesting that tumor cells can transform their function to adapt to the surrounding microenvironment. Thus, it is speculated that TGF-? or PGE2 producing tumor cells are established from tumors that survive under immune surveillance and further studies elucidating the biomarker to distinguish the TGF-? or PGE2 producing and non-producing tumors with EGFR inhibition may help us to better treat the patients with anti TGF-? antibody or COX inhibitor.
Experimental evidences that the tumor microenvironment plays a significant role in resistance of EGFR inhibitor have been reported [
12],[
22]-[
25]. EGFR inhibition induces tumor cells to the mesenchymal phenotype, which cell type show resistant to EGFR inhibitor, via cytokines such as IL-6 and TGF-? [
22]. Both exogenous IL-6 and TGF-? induced EGFR inhibitor resistance [
12],[
23],[
24] and endogenous TGF-? was produced from EGFR inhibitor resistant tumor cells [
22]. Strikingly, TGF-? receptor inhibitor abrogated motility of erlotinib-resistant tumor cells suggesting that cytokines might be promising target to overcome EGFR inhibitor resistance [
25]. The TGF-? pathway is known as an important immune suppressor pathway affecting tumor microenvironment. TGF-? can regulate both the innate and acquired immune systems by inducing regulatory CD4+/FoxP3+ T cells, which represent on of the main barriers to antigen-specific antitumor immunity, and by skewing NK T cells or macrophages to regulatory phenotypes [
3],[
4],[
26]. TGF-? functions not only as an immune regulator, but also as an oncogenesis promoter by inducing epithelial-to-mesenchymal transition through both Smad-dependent and independent pathways [
27] or transitioning constitutive cells to a tumor-associated phenotype that facilitates tumor progression [
28]. The present study revealed that TGF-? production or reduction by EGFR inhibition is depended on individual tumors. Further studies will be needed to determine whether TGF-? can function as a biomarker to assess the effectiveness of EGFR targeted therapy with and without concurrent immunotherapy.
Prostaglandins and leukotrienes are produced by the COX pathway and function as potent immune regulators. Both the tumor and surrounding stroma are capable of producing PGE2 [
29], which can increase regulatory T cell activity [
18]. Furthermore, PGE2 induces myeloid-derived suppressor cells, which inhibit effector T cells [
30]. Accumulating evidence suggests that the application of COX inhibitors may be useful for cancer treatment both in colorectal cancer and in HNSCC [
31]. In this study, we showed that EGFR inhibition augmented PGE2 production by Sa-3 tumor cell, and that COX-2 inhibitor could restore the suppression of antigen-specific CD4
+ T cell responses. Thus, the COX-2/PGE2 pathway is partially responsible for immunosuppressive effects of tumor cells through EGFR blockade. While reduction of PGE2 by erlotinib has been reported [
19], we could not detect PGE2 reduction by EGFR inhibition suggesting that the fluctuation of PGE2 production by EGFR blockade is affected by cancer heterogeneity. Recently, COX-2 inhibitor with erlotinib has been reported to inhibit the proliferation of head and neck squamous cell carcinoma in patients [
32]. Although we could show the increase of PGE2 production by EGFR inhibition only in tumor cell Sa-3, we believe our result may partly elucidate the mechanism of positive effects of COX-2 inhibitor with erlotinib.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
TK carried out and participated in all of the studies. TK, KO, NA, SK and YH made substantial contributions to acquisition of the results. HK and EC designed, supervised, and coordinated the study, and drafted the manuscript. All authors read and approved the final manuscript.