The first two authors contributed equally to this article.
Immunotherapy of murine colon cancer using receptor tyrosine kinase EphA2-derived peptide-pulsed dendritic cell vaccines
Article first published online: 8 AUG 2007
Copyright © 2007 American Cancer Society
Volume 110, Issue 7, pages 1469–1477, 1 October 2007
How to Cite
Yamaguchi, S., Tatsumi, T., Takehara, T., Sakamori, R., Uemura, A., Mizushima, T., Ohkawa, K., Storkus, W. J. and Hayashi, N. (2007), Immunotherapy of murine colon cancer using receptor tyrosine kinase EphA2-derived peptide-pulsed dendritic cell vaccines. Cancer, 110: 1469–1477. doi: 10.1002/cncr.22958
- Issue published online: 14 SEP 2007
- Article first published online: 8 AUG 2007
- Manuscript Accepted: 6 JUN 2007
- Manuscript Revised: 8 MAY 2007
- Manuscript Received: 9 JAN 2007
- Grant-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan
- Grant-in-Aid for Research on Hepatitis and Bovine Spongiform Encephalopathy from the Ministry of Health, Labor, and Welfare of Japan
- dendritic cells;
- colorectal cancer;
- cancer immunotherapy
Further optimization of dendritic cell (DC)-based vaccines is required clinically against advanced stage cancer. Given the broad range of expression levels observed in the recently defined tumor antigen EphA2 in a diverse types of cancers, especially in advanced stage or metastatic cancers, the authors evaluated the effectiveness of vaccination using DCs pulsed with EphA2-derived peptides (Eph-DCs) in a murine colon cancer model.
EphA2 protein expression levels were evaluated in advanced colorectal carcinoma tissues from 10 patients by Western blot analysis. C57BL/6 mice were immunized with Eph-DCs twice weekly. Interferon γ (IFN-γ) ELISPOT assays were used for the analysis of CD8-positive T cells that were specific for EphA2-derived peptide. Immunized mice were challenged subcutaneously with EphA2-positive murine colorectal adenocarcinoma (MC38) mouse colon tumors or with EphA2-negative BL6 melanoma tumors. In some experiments, mice were injected with anti-CD8, anti-CD4, or antiasialo GM1 antibody to deplete corresponding lymphocyte subsets.
Among 10 samples of advanced colorectal carcinoma, 6 samples (60%) overexpressed EphA2. IFN-γ ELISPOT assays revealed that EphA2-derived peptide-specific CD8-positive T cells were generated by immunization with Eph-DCs. Immunization with Eph-DCs inhibited MC38 tumor growth compared with immunization using unpulsed DCs or phosphate-buffered saline. In contrast, Eph-DC vaccination had no effect on BL6 growth. Antibody depletion studies revealed that both CD8-positive T cells and CD4-positive T cells, but not natural killer cells, played critical roles in the efficacy observed for immunizations with Eph-DCs. Eph-DC vaccines resulted in long-term antitumor immunity against a rechallenge with MC38 tumor cells.
The current results demonstrated that Eph-DC vaccines may represent a promising preventative/therapeutic modality in the cancer setting. Cancer 2007. © 2007 American Cancer Society.
Dendritic cell (DC)-based vaccines are attractive cancer modalities, because DCs can induce both tumor antigen-specific cytotoxic T lymphocytes (CTLs) and helper-T cells. In this regard, DCs pulsed with tumor-associated antigens in various forms, including whole cell lysates,1 proteins,2 peptides,3 RNA,4 or DNA,5 have been proven effective in eliciting protective and therapeutic antitumor immunity in murine models. The results of several DC-based tumor vaccine trials also recently have been reported for patients who had late-stage B-cell lymphoma, melanoma, prostate cancer, and renal cell carcinoma.6–9 In colorectal carcinomas, DC-based vaccines using synthetic peptides derived from known tumor antigens, such as carcinoembryonic antigen (CEA), also have been reported; however, to our knowledge to date, objective clinical responses have been observed only in a minority of treated patients with colon cancer.10–12 Thus, a new strategy for tumor antigen-derived peptide-DC vaccines is expected to improve the clinical efficacy in patients with advanced colon cancer.
The Eph family constitutes the largest family of receptor tyrosine kinases, consisting of 2 Eph classes (EphA and EphB) and 2 classes of corresponding ligands, ephrin A and ephrin B, respectively. Because they are known largely for their role in neuronal development and tissue remodeling,13–15 it has been suggested recently that Eph receptors play a role in oncogenesis16–18 and tumor angiogenesis.19, 20 EphA2 is overexpressed in numerous cancer types, including melanoma21 and prostate,8 breast,22 lung,23 renal cell,24 and colorectal25 carcinomas; and it is altered functionally to promote the development of disseminated disease in a large number of different cancers. Indeed, the highest degree of EphA2 expression among tumors is observed most commonly in metastatic lesions, suggesting that EphA2 may represent a high-priority target for immunotherapy, especially in patients with advanced stage or metastatic cancer. We previously demonstrated that some patients with renal cell carcinoma exhibited both CD8-positive and CD4-positive T-cell responses to novel, EphA2-derived epitopes and that EphA2-derived epitopes were useful for predicting disease status and outcome as immunomonitoring targets.26, 27 These results also supported the therapeutic potential of EphA2 peptide-pulsed DC (Eph-DC)-based vaccines, although they have not been evaluated to date.
In the current study, we have demonstrated that vaccination with Eph-DCs elicits EphA2-specific CTL responses that are protective against EphA2-positive tumors, but not against EphA2-negative tumors. These results support the translational development of Eph-DC vaccines for patients with EphA2-positive colon cancer.
MATERIALS AND METHODS
Female C57BL/6 mice were purchased from Clea Japan, Inc. (Tokyo, Japan) and were used at ages 6 weeks to 8 weeks. They were housed under conditions of controlled temperature and light with free access to food and water at the Institute of Experimental Animal Science, Osaka University Graduate School of Medicine. All animals received humane care, and the study protocol complied with the institution's guidelines.
MC38, a mouse colon carcinoma cell derived from C57BL6/J mice, was generously provided by Dr. Kazumasa Hiroishi (Showa University School of Medicine, Tokyo, Japan). BL6, a melanoma cell line, and YAC-1, a sensitive cell line to natural killer (NK) cells, were purchased from American Type Culture Collection (ATCC) (Rockville, Md). These cell lines were maintained in complete medium (CM) (RPMI medium supplemented with 10% fetal bovine serum, 100 U/mL, penicillin and 100 μg/mL streptomycin) at 37°C in 5% carbon dioxide.
The protein sequences of mouse EphA2 were obtained from Genbank and were analyzed for H-2Kb binding motifs using BioInformatics and Molecular Analysis Section and a proteosomal cleavage site-prediction system. The H-2Kb-binding murine (m)EphA2682-689 epitope (VVSKYKPM) was synthesized using an automated, solid-phase peptide synthesizer in the protein synthesis facility at the University of Pittsburgh Cancer Institute and was purified (to >95%) by using reverse-phase high-performance liquid chromotography.28
Western Blot Analyses
The proteins in samples from 10 patients with advanced colorectal carcinoma (stage III or IV) and the lysates from mouse tumor cell lines were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and were analyzed for expression of EphA2 using EphA2 monoclonal antibody (C20 Ab; Santa Cruz Biotechnology, Inc., Santa Cruz, Calif). Blots were imaged on Hyperfilm (Amersham Bioscience, Buckinghamshire, U.K.) after using horseradish peroxidase-conjugated goat-antirabbit immunoglobulin G (Bio-Rad, Hercules, Calif) and the Super Signal West Pico Luminol Enhance Solution kit (Pierce, Rockford, Ill). Expression of β-actin served as a loading control.
Generation of DCs From Bone Marrow and DC-based Peptide Vaccines
With minor modifications, the procedure that we used in this study was described previously.29 Briefly, C57BL/6 bone marrow cells were cultured in CM supplemented with 500 U/mL of recombinant murine granulocyte–macrophage-colony-stimulating factor (R&D Systems Inc., Minneapolis, Minn) and recombinant interleukin 4 (IL-4) (R&D Systems Inc.) for 9 days. DCs were separated by magnetic cell sorting using CD11c microbeads (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) and typically represented >90% of the harvested population of cells based on morphology and expression of the CD40, CD80, CD86, and major histocompatibility (MHC) Class II cells (data not shown). DCs were incubated with mouse Eph682-689 peptide at a concentration of 10 μg/mL per 106 DC/mL CM for 2 hours at 37 °C. The cells were harvested and washed 3 times with phosphate-buffered saline (PBS) before use.30
IFN-γ ELISPOT Assays for Peptide-reactive CD8-positive T-cell Response
Splenocytes were harvested 5 days after subcutaneous immunization with 1 × 106 Eph-DCs twice over a 1-week interval. CD8-positive T cells were isolated selectively from splenocytes by magnetic cell sorting using CD8 microbeads (Miltenyi Biotec). Mouse IFN-γ ELISPOT assays were performed using a mouse IFN-γ ELISPOT kit (R&D Systems Inc.) according to the manufacture's instructions. IFN-γ-secreting cells appeared as blue spots. The data are represented as the mean number (±standard deviation [SD]) of IFN-γ spots per 100,000 CD8-positive T cells analyzed.
C57BL/6 mice were immunized subcutaneously into the left flank with 1 × 106 Eph-DCs or unpulsed DCs in a total volume of 100 μL of PBS twice a week. On Day 0, for the second Eph-DC immunization, 2 × 105 MC38 cells as Eph A2-positive cells or 5 × 104 BL6 cells as EphA2-negative cells were injected subcutaneously into the right flank. To assess the impact of systemic immunity from subcutaneous injection of Eph-DCs, tumor size was assessed every week and was recorded in mm2 by determining the product of the greatest perpendicular dimensions measured by Vernier calipers. Data are reported as the average tumor area ±SD.
Splenocytes were harvested 14 days after tumor inoculation. Responder cells (5 × 106 per well) were restimulated in vitro with 1 × 106 MC38 cells that had been treated with 0.5 mg/mL mitomycin C (Kyowa-Hakko, Tokyo, Japan) in the presence of 30 IU/mL recombinant murine IL-2 (Strathmann Biotech, Hannover, Germany). After 5 days of in vitro restimulation, lymphocytes were subjected to 4-hour 51Cr release assays against the MC38 or BL6 targets, as described previously.29 CD4-positive and CD8-positive T cells were depleted selectively from whole splenocytes by magnetic cell sorting using CD4 and CD8 microbeads (Miltenyi Biotec), respectively. In some experiments, splenocytes were harvested 1 day after tumor inoculation and were subjected directly to 4-hour 51Cr release assays against YAC-1 targets (NK cell-sensitive cells).
T-Cell and NK Cell In Vivo Depletion Experiments
To deplete T cells in vivo, anti-CD4 (GK1.5 hybridoma; ATCC) or anti-CD8 antibody (53–6.72 hybridoma; ATCC) were administered intraperitoneally 4 days and 1 day before every tumor inoculation and then every 5 days after tumor inoculation. For depletion of NK cells in vivo, we used antiasialo GM1 antibody (Wako, Osaka, Japan), which was administered intraperitoneally 1 day before tumor inoculation and then every 5 days after tumor inoculation. The efficiency of specific subset depletions was validated by flow cytometry analysis of splenocytes. In all samples, 99% of the targeted cell subset was depleted specifically (data not shown).
C57BL/6 mice were immunized subcutaneously with 1 × 106 Eph-DCs twice weekly and were challenged subcutaneously with 2 × 105 MC38 cells at the second Eph-DC immunization. Forty-two days after tumor inoculation, 2 × 105 MC38 cells were injected subcutaneously into the contralateral flank of the initial MC38 tumor on Day 0. For control experiments, 2 × 105 MC38 cells were injected subcutaneously into naive C57BL/6 mice on Day 0, at the same time of MC38 rechallenge. Tumor size was assessed every week after the second tumor inoculation.
The statistical significance of differences between the groups was determined by applying a Student t test with Welch correction or a 1-way analysis of variance after each group had been tested with equal variance and Fisher exact probability test. Statistical significance was defined as P < .05.
Expression of EphA2 in Human Colorectal Cancer Tissues and Murine Tumor Cells
We evaluated the expression of EphA2 in samples from 10 patients with colorectal carcinoma and H-2b-syngeneic murine tumor cell lines from C57BL/6 mice by Western blot analysis using a species cross-reactive monoclonal antibody. Figure 1A shows that 6 samples (60%) of advanced-stage disease (stage III or IV) overexpressed EphA2 compared with normal colon tissues. In addition, the murine colon cancer cell line (MC38) expressed EphA2 protein, but the murine melanoma cell line (BL6) did not (Fig. 1B).
Detection of EphA2-derived Peptide-specific CD8-positive T Cells Secreting IFN-γ After Vaccination With Eph-DCs
We performed IFN-γ ELISPOT assays to examine whether subcutaneous injection of Eph-DCs could generate CD8-positive T cells that were specific for EphA2-derived peptide in vivo. Figure 2 shows that the frequency of specific CD8-positive T cells secreting IFN-γ in mice treated with Eph-DCs was significantly higher than the frequency observed in naive mice or in mice treated with unpulsed DCs. These results demonstrate that EphA2-specific, type 1, CD8-positive T cells effectively are generated by in vivo immunization with Eph-DCs.
Immunization With Eph-DCs Prevents Progression of EphA2-positive Tumors, but Not EphA2-negative Tumors, In Vivo
Next, we examined whether immunization with Eph-DCs (on Days −7 and 0) would induce protective antitumor effects against EphA2-positive MC38 colon cancer. Figure 3A shows that MC38 tumor growth in mice that were immunized with Eph-DCs was inhibited strongly compared with tumor growth in mice that were immunized with unpulsed DCs (P < .05 on Days 21 and 28) or with PBS (P < .05 on Days 14, 21, and 28), whereas tumor growth in mice that were immunized with unpulsed DCs was inhibited slightly compared with tumor growth in mice that received PBS (P < .05 on Days 21 and 28). Conversely, in vivo growth of EphA2-negative BL6 tumors in mice that were immunized with Eph-DCs or unpulsed DCs was inhibited slightly compared with the tumor growth in mice that received PBS (P < .05 on Day 28). However, it is noteworthy that there was no significant difference in BL6 tumor growth between the Eph-DC group and the unpulsed DC group (Fig. 3B). These results indicate that vaccination with Eph-DCs provides specific antitumor effects against relevant EphA2-positive MC38 tumors.
Induction of Specific CTLs Against MC38 Cells After Immunization With Eph-DCs
For the next experiment, we examined whether our Eph-DC regimen could induce specific cytolytic reactivity against MC38 or BL6 cells. Splenocytes were harvested from the various treatment groups of mice that were killed 14 days after tumor inoculation. Figure 4A shows that splenocytes from mice treated with unpulsed DCs displayed weak cytolytic reactivity against MC38 targets, whereas splenocytes from mice treated with PBS failed to exhibit detectable reactivity against this cell line. In contrast, splenocytes harvested from mice treated with Eph-DCs displayed far stronger anti-MC38 cytolytic reactivity than any control treatment group. CD8-positive, T cell-depleted splenocytes (harvested from mice treated with Eph-DCs) displayed significantly weaker anti-MC38 cytolytic reactivity than whole splenocytes; however, CD4-positive T cell-depleted splenocytes did not (Fig. 4B). Conversely, cytolytic activity was not observed against EphA2-negative cells (BL6) in any of the control/treatment arms, as shown in Figure 4A. We also harvested splenocytes 1 day after the second immunization (ie, Day 1 after tumor inoculation) to examine the potential early activation of NK cells by immunization with Eph-DCs. Figure 4C shows that no cytolytic activity was observed against the NK-sensitive YAC-1 cells in any treatment arm. These results suggest that principal antitumor effector cells in vaccinated mice are CD8-positive CTLs.
Requirement of Both CD4-positive T Cells and CD8-positive T Cells, but not NK Cells, for the Antitumor Effect of Immunization With Eph-DCs
To examine which lymphocyte subsets contributed to Eph-DC or unpulsed DC treatment, we performed depletion studies on a subset of CD4-positive T cells, CD8-positive T cells, and NK cells. Figure 5A shows that the therapeutic efficacy of Eph-DC therapy was reduced strongly in CD8-positive, T cell-depleted mice and was reduced partially in CD4-positive, T cell-depleted mice. In contrast, tumor growth still was suppressed in vaccinated mice that had been depleted of NK cells. In addition, the therapeutic efficacy of unpulsed DC therapy in CD4-positive or CD8-positive, T cell-depleted mice also was reduced, whereas that in NK cell-depleted mice was not reduced (Fig. 5B). These results suggest that both CD8-positive T cells and, to a lesser degree, CD4-positive T cells are required for the observed antitumor effects noted for Eph-DC vaccination. Both CD8-positive and CD4-positive T cells also were involved in unpulsed DC vaccination in our model.
We then sought to determine whether prior Eph-DC treatment would have a durable effect on a subcutaneous rechallenge with MC38 tumor cells. C57BL/6 mice were immunized with subcutaneous injections of Eph-DCs and challenged with subcutaneous MC38 tumors. Forty-two days after the primary tumor inoculation, 2 × 105 MC38 cells were injected subcutaneously into the contralateral flank of these mice on Day 0. Figure 6 shows that rechallenged tumors in mice that received the Eph-DC treatment regimen were inhibited significantly in their progression through the chosen endpoint of these experiments on Day 28 (P < .05 vs naive mice on Days 14, 21, and 28).
Modified, DC-based vaccines using synthetic peptides derived from known tumor antigens, such as CEA, have been reported for colon cancer; although, to date, objective clinical responses have been observed only in a minority of patients who received treatment with these modalities.10–12 This may be explained in part by the application of DC-based vaccines toimmunosuppressed patients with advanced colon cancer and/or to the modest immunogenicity of tumor antigens (ie, CEA) that have been applied to date in this setting. Recently, a novel tumor antigen, EphA2, has been identified that has specific characteristics and that frequently is overexpressed in advanced cancers, suggesting that this antigen may have great potential as a target for immunotherapy, especially in patients with advanced-stage or metastatic cancer. In the current study, 60% of colon cancer tissue samples overexpressed EphA2, consistent with a recent report by Saito et al.25 We demonstrated that Eph-DC vaccines effectively promoted antitumor effects in a colon cancer model, suggesting that EphA2-derived CTL epitopes have the potential to serve as relevant components of novel DC-based vaccines for colon cancer.
IFN-γ ELISPOT assays revealed that immunization with Eph-DCs in normal mice resulted in the induction of specific CD8-positive T cells. Based on these results, we examined the antitumor effectiveness of Eph-DC vaccines in a syngenic, EphA2-positive MC38 colon cancer model. The Eph-DC vaccines induced antitumor effects against EphA2-positive MC38 colon carcinoma, but not against EphA2-negative BL6 melanoma, suggesting that EphA2-specific antitumor immunity was generated by Eph-DC vaccines, consistent with the results from our earlier IFN-γ ELISPOT assays.
In vitro assays revealed that the main antitumor effector cells for killing MC38 colon cancer cells were CD8-positive T cells and, possibly, CTLs. This cytolytic activity was specific for MC38 cells, because splenocytes did not kill BL6 cells. These results suggested that Eph-DC vaccines could efficiently generate specific CTLs that recognize and kill relevant EphA2-positive (but not irrelevant EphA2-negative) tumor targets.
Our in vivo lymphocyte-depletion studies demonstrated that CD8-positive T cells contributed to the inhibition of tumor growth in Eph-DC immunization and that CD4-positive T cells contributed to a lesser extent. Moreover, our tumor experimental data demonstrated that immunization with Eph-DCs maintains the antitumor effect against MC38 tumor over an extended period of time, despite the use of a single EphA2-derived CD8-positive T cell epitope in the DC-based vaccine. Typically, effector CD8-positive T cells induced by minimal CTL epitope peptides do not persist for a long time, and the induction of durable-memory CD8-positive T cells requires the support of CD4-positive T cells.31–33 Therefore, these results suggest that Eph-DC vaccines may activate CD8-positive T cells (that recognize EphA2-derived CTL epitopes) and CD4-positive T cells (that recognize tumor antigens related to MC38 colon cancer cells), which are taken up by specifically dedicated antigen-presenting cells, and that the activated CD4-positive T cells likely contribute to the generation and maintenance of memory in EphA2-specific, CD8-positive T cells.
Immunization with control, unpulsed DCs was inhibited both EphA2-positive MC38 tumor growth compared with PBS. Generally, unpulsed DC vaccines are not expected to generate CTLs. However, our lymphocyte-depletion studies in the MC38 tumor model demonstrated that the therapeutic efficacy of unpulsed DC therapy in CD4-positive or CD8-positive T cell-depleted mice was reduced equally. These results suggest that unpulsed DCs can induce protective antitumor effects in mice through the presentation of “self” peptides in MHC complexes to specific autoreactive CTLs, which are capable of recognizing tumor cells that also present these peptides, consistent with the previous report by Dworacki et al.34 Moreover, unexpectedly, Eph-DC or unpulsed DC vaccines had weak antitumor effects against EphA2-negative BL6 tumors compared with PBS treatment. BL6 cells do not express EphA2, and EphA2-specific CTLs do not have cytolytic activity against BL6 tumors. Dworacki et al. reported that immunization with unpulsed DCs inhibited a variety of syngeneic tumors through the activation of both CD4-positive T cells and CD8-positive T cells,34 suggesting that Eph-DC or unpulsed DC vaccines may activate CD4-positive and CD8-positive T cells weakly and that these cells may play a role in weakly inhibiting BL6 tumor growth.
Recent research in DC biology has revealed that DCs also contribute to innate immune responses by activating NK cells through IL-12 secretion and direct cellular interaction.35 However, our current data demonstrated that NK cells were not involved in generating antitumor effect of Eph-DC or unpulsed DC vaccination in our lymphocyte-depletion studies. We speculate that subcutaneous, local NK cells may not be activated efficiently by immunization of Eph-DCs or unpulsed DCs, because NK cells are not so abundant in the subcutaneous tissue. Instead, if we were to apply this DC vaccine in sites rich in NK cells (ie, the liver), then this strategy may prove to be more effective in treating NK cell-sensitive liver cancers, for instance. Currently, we are evaluating these possibilities and performing histopathologic evaluations to confirm that treated animals do not exhibit autoimmune pathology in organs (such as lung, kidney, etc) that constitutively express low levels of EphA2.
Despite the recent progress and early successes reported for DC-based cancer immunotherapies, there is significant room for improvement in these regimens, especially with respect to advanced colon cancer. In this study, we demonstrated that Eph-DC vaccines revealed antitumor effects against colon cancers. In addition to CEA-based vaccines, EphA2-derived peptide-pulsed DC vaccines may represent a promising therapeutic modality against advanced colon cancers. Recently, it has been reported that some clinical trials using peptide cocktail-pulsed DCs may be useful strategies for treating patients with malignant tumors.36, 37 Therefore, DCs pulsed with multiple peptides derived from various tumor-associated antigens, including both EphA2 and CEA, may improve the therapeutic effects against advanced colon cancers.
We thank Ms. Kyoko Iwase (Osaka University) for her excellent technical support.
- 10Colorectal cancer vaccines: principles, results, and perspectives. Gastroenterology. 2004; 27: 1821–1837., , , .
- 15A role for the EphA family in the topographic targeting of vomeronasal axons. Development. 2000; 128: 895–906., , , .