Stimulatory effects of CpG oligodeoxynucleotide on dendritic cell-based immunotherapy of colon cancer in CEA/HLA-A2 transgenic mice†
Article first published online: 19 SEP 2008
Copyright © 2008 Wiley-Liss, Inc.
International Journal of Cancer
Volume 124, Issue 4, pages 877–888, 15 February 2009
How to Cite
Saha, A., Bhattacharya-Chatterjee, M., Foon, K. A., Celis, E. and Chatterjee, S. K. (2009), Stimulatory effects of CpG oligodeoxynucleotide on dendritic cell-based immunotherapy of colon cancer in CEA/HLA-A2 transgenic mice. Int. J. Cancer, 124: 877–888. doi: 10.1002/ijc.24009
This study was presented as an abstract at the 94th Annual Meeting of the AAI, May 18–22, 2007, Miami Beach, FL.
- Issue published online: 11 DEC 2008
- Article first published online: 19 SEP 2008
- Accepted manuscript online: 19 SEP 2008 12:00AM EST
- Manuscript Accepted: 28 AUG 2008
- Manuscript Received: 13 JUN 2008
- NIH. Grant Number: RO1 CA104804
- dendritic cells;
- CpG ODN;
- CTL epitope;
- anti-idiotype antibody;
- CEA-A2Kb double transgenic mice;
- antigen presenting cells
Immunostimulatory DNA containing unmethylated cytosine-guanine (CpG) motifs have been successfully used as adjuvants to enhance the immunity of vaccines designed to trigger antitumor T-cell responses. We examined the effect of a CpG oligodeoxynucleotide (CpG ODN) for its ability to potentiate the activity of tumor antigen-pulsed dendritic cells (DC) in a clinically relevant mouse model, which is transgenic for both carcinoembryonic antigen (CEA) and HLA-A2 for the treatment of colon carcinoma in a therapeutic setting. The systemic administration of CpG ODN 1826 alone had modest effect on tumor growth when tumors were palpable and had no effect with larger tumor burden. However, coadministration of CpG ODN 1826 with the vaccine provided significant increase in tumor-free survival compared with mice immunized with DC-based vaccines alone. The DC/CpG combined vaccination strategy resulted in increased secretion of Th1 cytokines and HLA-A2-restricted CEA-specific CTL responses were also enhanced. Both tumor regression and extended tumor-free survival resulting from DC/CpG combination therapy required the participation of T cells. Tumor-free mice were resistant to tumor rechallenge and immunity conferred by the vaccine was transferable in athymic nude mice. These results provide evidence that vaccination with antigen-pulsed DC with CpG ODN as adjuvant can lead to effective tumor regression and long-term survival in a murine model of colon carcinoma. © 2008 Wiley-Liss, Inc.
Colorectal cancer is the third most common cancer diagnosed (excluding skin cancer) and the second most common cause of cancer death in the United States, accounting for an estimated 50,000 deaths each year.1 Despite advances in chemotherapy regimens, radiation and surgical treatments, the overall long-term survival is ∼60–70%.1 Based on encouraging results in several clinical trials,2–4 the currently recommended first-line therapy for metastatic colorectal cancer includes 5-fluorouracil and leucovorin with irinotecan or oxaliplatin.5 Despite these advances, no established treatment regimens are available when patients fail to improve with these agents. Therefore new treatment approaches are needed to improve the survival of these patients. Immunotherapy offers a promising approach to cancer treatment that may someday overcome the shortcomings of traditional cancer management.
The identification of tumor antigens to which autologous T cells respond has stimulated efforts to develop vaccines incorporating these antigens for immunotherapy.6 Cytotoxic T lymphocytes (CTLs) are a major focus of this effort because of their well-documented ability to eradicate tumor cells in vivo.7 The use of dendritic cells (DC) in immunotherapy offer the potential to sustain a more durable T-cell response, and can function as important initiators and modulators of a specific and lasting immune response against tumor antigens.8 Human as well as animal studies have documented that these extremely potent antigen presenting cells (APCs) can induce immunity to otherwise weakly immunogenic tumor antigens and clinical trials of DC-based vaccines have shown promise in patients with a variety of tumors.9–11 However, questions continue to arise about the optimal maturation status of DC, the appropriate antigenic stimulus, and importantly, the optimal immunoadjuvant that should be used.
To enhance an antitumor immune response to solid tumors, strategies to augment APC or T-cell number and/or function have used cytokines such as Flt3L, granulocyte-macrophage colony-stimulating factor, IL-2 and IL-12.12–15 Since recombinant cytokine administration suffers from several drawbacks, therapeutics that can “naturally” induce activation of various immune cell subsets and production of various cytokines, which participate in the development of an active response, is thought to be more effective and less toxic than immunization using external cytokine as adjuvant. To this end, synthetic ODNs containing specific bacterial unmethylated CpG motif, which are one of so-called pathogen-associated molecular patterns, have attracted a great deal of attention as a novel and safe adjuvant.16, 17 CpG ODNs are recognized by cells of innate immune system of vertebrates, such as B cells, macrophages, NK cells and DC, and activate these cells.16, 17 CpG ODNs preferentially induce Th1 immune response through its receptor, TLR9, with the production of cytokines, such as TNF-α, IL-12 and IFN-γ, appropriate for the development of antitumor immunity.18 In cancer immunotherapy, CpG ODNs have been used as monotherapy,19, 20 adjuvant for vaccination,21 and in combination with chemotherapy,22 radiation treatment,23 antibody therapy,24 cellular vaccines22 and even other “danger signals” such as double-stranded RNA.25 These antitumor effects are observed both with local and systemic injection of CpG ODNs. The predominant effector cells responsible for the antitumor effect have varied and include NK cells, T cells and/or macrophages, depending on the model system studied. The effect of CpG ODN on human cancers is currently being evaluated in clinical trials.26
In a previous study, we assessed the ability of DC-based vaccines consisting of CTL peptide of carcinoembryonic antigen (CEA) along with 3H1, an anti-idiotype (Id) antibody mimicking CEA as a potential vaccine candidate to induce antitumor immunity in a clinically relevant mouse model that expresses both CEA and HLA-A2. We have found that, in a therapeutic setting with 7-day-old established tumors, mice immunized with agonist peptide for CEA691 (YMIGMLVGV) or mice immunized with agonist peptide for CAP-1 (YLSGADLNL, CAP1-6D) along with 3H1 induced significant increase in tumor-free survival compared with mice immunized with peptide-pulsed DC alone.27 In the present study, we wished to extend these observations to study the effect of a select CpG ODN 1826 in DC-based vaccines containing CTL epitope of CEA and anti-Id antibody 3H1. We found that systemic injections of CpG ODN combined with DC-based vaccines resulted in potent antitumor effects with substantial increase in tumor-free survival, even when therapy was initiated in a setting of bulk disease. CpG ODN 1826 induced strong activation of CEA-specific T-cell responses and in particular CTL responses, and promoted long-term survival of antigen-specific T-cell population. Based on these data we propose that an immunotherapy approach combining DC with CpG ODN would be more effective in inducing a robust antitumor immune response for the treatment of colorectal carcinoma.
Material and methods
Mice and cell lines
C57BL/6J-CEA-A2Kb double transgenic (Tg) mice were generated by crossing homozygous male C57BL/6J-CEA-Tg with homozygous female C57BL/6J-A2Kb-Tg mice.27 PCR products indicated that F1 offspring, derived from several breeding pairs, expressed both CEA and HLA-A2 transgenes (data not shown). Male (5–6 weeks old) athymic nude mice were purchased from Harlan Laboratories (Indianapolis, IN). For in vivo studies, all mice were used at 6–8 weeks of age. Mice were maintained in a pathogen-free environment and all procedures were approved by the Institutional Animal Care and Use Committee guidelines. MC-38-CEA-A2Kb cells,27 which expressed CEA and HLA-A2, were generated by transfecting murine colon carcinoma cell line MC-38-CEA28 with a plasmid encoding A2Kb. T2, a human cell line (TAP-deficient, HLA-A2+), was kindly provided by Dr. Walter J. Storkus (University of Pittsburgh School of Medicine, Pittsburgh, PA). LS 174T (CEA+, HLA-A2+), a human colon carcinoma cell line was purchased from ATCC (Manassas, VA).
Synthetic peptides, CpG ODN
Two different HLA-A2-restricted CTL epitopes of CEA were selected for our study: CEA691-69929 and CEA605-613 (designated CAP-130). HIV-1gp160120-128 (KLTPLCVTL) peptide, agonist peptide for CEA691 (YMIGMLVGV, peptide 1) and agonist peptide for CEA605 designated as CAP1-6D (YLSGADLNL, peptide 2) were synthesized by Synthetic Biomolecules (San Diego, CA) with purity of >90%. HLA-A2-restricted HIV-1gp160120-128 peptide was used as a control. The immunostimulatory synthetic CpG ODN 1826 (5′-TCCATGACGTTCCTGACGTT-3′), previously shown to be a potent stimulator of mouse TLR9 and innate and adaptive immunity, was used in our experiments. The backbone for these ODN had phosphorothioate linkage throughout their sequence and was obtained from Coley Pharmaceutical Group (Ontario, Canada).
Tumor therapy experiments
C57BL/6J-CEA-A2Kb mice were transplanted with 5 × 105 of MC-38-CEA-A2Kb colon carcinoma cells s.c. in the lower left flank. This dose of tumor cells develop palpable tumors (4–5 mm in diameter) in 100% of mice within 1 week of tumor transplant and is lethal within 5 to 6 weeks of transplant if left untreated. Tumor-bearing mice were randomly divided into several groups (8–10 mice/group) for immunizations and therapy was initiated either on day 7, day 10 or day 14 after tumor transplant. For vaccination with ODN as adjuvant, 50 μg CpG ODN 1826 was injected s.c. in the lower right flank on days 7, 12 and 17 (day 7 therapy), or on days 10, 15 and 20 (day 10 therapy), or on days 14, 19 and 24 (day 14 therapy). DC were generated from bone marrow of syngeneic C57BL/6J-CEA-A2Kb mice and were pulsed with peptide or 3H1 as described.27 These peptide-pulsed DC (2–3 × 105 cells) and 3H1-pulsed DC (2–3 × 105 cells) were also injected s.c. in the lower right flank on days 8, 10, 13, 15, 18 and 20 (day 7 therapy), or on days 11, 13, 16, 18, 21 and 23 (day 10 therapy), or on days 15, 17, 20, 22, 25 and 27 (day 14 therapy), respectively. Mice were monitored twice weekly for tumor growth and survival, and were sacrificed when tumors reached a diameter >20 mm; survival was recorded accordingly.
Assessment of in vitro cytotoxic activity
Experiments were performed as described previously27 and according to the standard protocols.31 Briefly, splenocytes were isolated from 3 mice per group 5 days after the final vaccination, and were stimulated by coculture with DC pulsed with immunizing peptide and DC pulsed with 3H1 along with recombinant human IL-2 (Sigma, St. Louis, MO). On day 5, these in vitro stimulated cells were used as CTL effector cells, and the CTL activity was determined by a standard 6 hr 51Cr-release cytotoxicity assay using a variety of target cell lines. For inhibition experiments, CTL activity was tested in the presence of anti-mouse CD8 monoclonal antibody (mAb), anti-mouse CD4 mAb, anti-mouse H-2Kb/H-2Db mAb, anti-mouse I-Ab mAb or anti-human HLA-A2 mAb (BD Biosciences, San Diego, CA). Isotype-matched mAbs were used as control.
BLT-esterase release assay
CTLs generated from immunized mice splenocytes were cultured with MC-38 cells or MC-38-CEA-A2Kb cells for 6 hr. Supernatant was harvested and the BLT (N-a-benzyl-oxycarbonyl-L-lysine-thiobenzyl ester hydrochloride) esterase activity of the supernatant was measured as described previously.32
For the detection of intracellular perforin, CTLs generated from immunized mice splenocytes were first cultured with media alone, MC-38 cells, or MC-38-CEA-A2Kb cells for 6 hr. Splenocytes were then stained with surface antibodies, fixed and permeabilized. Next, cells were stained with anti-perforin Ab (clone KM585; Kamiya Biomedical, Seattle, WA) and then incubated with a secondary Ab, FITC-conjugated goat anti-rat IgG (Southern Biotechnology, Birmingham, AL). Cells were analyzed using a flow cytometer (Beckman-Coulter, Hialeah, FL), and results are presented as percentage of positive cells. For the detection of surface expression of Ly-6C and CD44 on CD4+ and CD8+ T cells, immunized mice were sacrificed at different time points as indicated in the figure and splenocytes were stimulated in vitro for 2 days in the presence of antigen-pulsed DC. Cells were then stained and analyzed.
A2Kb/CAP1-6D pentamer-PE was obtained from ProImmune (Oxford, United Kingdom). Single cell suspensions were prepared from immunized mice splenocytes and CD8+ T cells were purified using magnetic micro beads by positive selection methods according to manufacturer's specifications (Miltenyi Biotec, Auburn, CA). The purified CD8+ T cells (1 × 106) were incubated for 10 min at 22°C with 10 μl PE-labeled pentamer; then, FITC-labeled anti-mouse CD8 mAb was added at the recommended concentration for an additional 20 min at 4°C. After washing, cells were analyzed by flow cytometry.
ELISPOT assays for IFN-γ and TNF-α production
CEA-specific and peptide-specific immune responses were also evaluated in ELISPOT assays. Briefly, splenocytes were isolated from different groups of immunized mice 5 days after the final vaccination and these cells (1–2 × 105) were cultured in medium with or without stimulator cells (1–2 × 104) in 96-well ELISPOT plates for 18–20 hr. The assay was developed according to the manufacturer's protocol. Immunized mice splenocytes (1 × 105) were also evaluated for their cellular avidity to the respective peptides by measuring IFN-γ spot in ELISPOT assay in the presence of titrated peptide pulsed on T2 cells (1 × 104).
Cytokine and chemokine analyses
In vitro cytokine and chemokine production was assessed by culturing immunized mice splenocytes (2 × 105) and DC (2 × 104) pulsed with immunizing peptide or DC pulsed with purified human CEA (Fitzgerald Industries, Concord, MA) in a 96-well tissue culture plate for 72 hr. Supernatants were harvested and analyzed by Multiplex ELISA (Quansys Biosciences, Logan, UT). Immunized mice sera were also analyzed for the detection of selected cytokine by ELISA. Intracellular cytokine analysis was performed using a cytokine detection kit according to the manufacturer's protocol (BD Biosciences). Briefly, immunized mice splenocytes were cocultured with CEA-pulsed DC for 2–3 days. Cells were then stained with surface antibodies, fixed and permeabilized. Next, cells were stained for intracellular IL-2, IFN-γ or TNF-α and analyzed by flow cytometry.
In vivo depletion of effector cells
Depletion experiments were performed as described previously.27 Briefly, mAb against CD4, mAb against CD8 and mAb against NK-1.1 were used in these experiments. C57BL/6J-CEA-A2Kb mice were transplanted with MC-38-CEA-A2Kb tumor cells on day 0 and therapy was started on day 7 as described above. Mice were depleted during immunization and then once weekly until the end of the study. The efficiency of depletion was always >90% as determined by flow cytometry (data not shown).
Adoptive transfer experiments in nude mice
Adoptive transfer experiments were performed in prophylactic setting as well as in therapeutic setting in athymic nude mice. C57BL/6J-CEA-A2Kb mice cured of MC-38-CEA-A2Kb tumors and those survived longer than 90 days (Fig. 1b) served as donors of lymphocytes. Splenocytes obtained from tumor-free mice were stimulated in vitro with peptide-2-pulsed DC, 3H1-pulsed DC and rhIL-2 for 5 days as described above. In prophylactic setting, stimulated splenocytes (1 × 107 per mouse) were injected i.v. into nude mice and after 1 day, each group of mice was challenged with 5 × 106 of LS 174T tumor cells s.c. in the left flank. In therapeutic setting, nude mice were challenged with 5 × 106 of LS 174T tumor cells s.c. in the left flank on day zero. On day 7, when tumors reached 6 to 7 mm in diameter, stimulated splenocytes (1 × 107 per mouse) were injected i.v. into tumor-bearing mice and adoptive transfers were repeated on days 11 and 15. Nude mice adoptively transferred with in vitro stimulated CEA-A2Kb naïve mice splenocytes and challenged with LS 174T cells served as control.
Statistical analyses were performed using Student's unpaired t-test or nonparametric Mann-Whitney rank-sum test using SigmaStat software (Jandel, San Rafael, CA). Survival data were plotted using the method of Kaplan-Meier and were analyzed by t-test or using Fisher exact test at specific time point. A value of p < 0.05 was considered significant.
CpG ODN improves the therapeutic potential of antigen-pulsed mature DC in vivo
Our previous studies suggested that in a therapeutic setting with 7-day-old established tumors, a DC-based vaccination protocol consisting of CTL peptides of CEA along with 3H1 was effective in increasing tumor-free survival compared with mice immunized with peptide-pulsed DC or 3H1-pulsed DC alone.27 In the present study, our goal was to evaluate the use of a select CpG ODN 1826 as an adjuvant for DC-based vaccination in a murine model of colon carcinoma. Based on our previous findings,27 here we have used 2 different HLA-A2-restricted CTL epitopes of CEA. Agonist peptide for CEA691 (YMIGMLVGV, peptide 1) or agonist peptide for CEA605 (YLSGADLNL, peptide 2) was used along with 3H1 through out the experiments.
To determine whether CpG ODN 1826 could enhance the antitumor immunity of DC-based therapy of colon carcinoma in mice with established tumor burden, we first determined the efficacy of CpG ODN as a single agent. MC-38-CEA-A2Kb tumor cells were injected s.c. into CEA-A2Kb-Tg mice on day 0. Seven days after tumor transplant, when tumors were palpable (4–5 mm in diameter) in 100% of mice, CpG therapy was initiated. CpG ODN was administered s.c. into the flank opposite of the tumor every 5 days for a total of 3 doses. Treatment with CpG ODN alone inhibited the tumor growth initially (Fig. 1a), but 2 weeks later tumors grew progressively and had a slight improvement in survival compared with concurrent controls (13.0% vs. 0%, respectively; Fig. 1b). Next, we wanted to evaluate whether inclusion of CpG ODN as an adjuvant in the DC-based therapy would improve CEA-specific immune response that would increase the overall survival of the immunized mice. To this end, using the same vaccination schedule, tumor-bearing mice were immunized with CpG ODN and/or peptide-pulsed DC plus 3H1-pulsed DC. Compared with a relatively weak effect on tumor growth inhibition when used as a single agent, CpG ODN demonstrated a strong antitumor effect on DC-based therapy (Figs. 1a and 1b). Tumor growth delay after the DC/CpG combined treatment was more than the sum of tumor growth delays caused by either DC-based vaccination or CpG ODN. Accumulating data from several experiments suggest that the systemic administration of CpG ODN improved the tumor free survival of mice when given in combination with peptide-1-pulsed DC plus 3H1-pulsed DC compared with mice receiving peptide-1-pulsed DC plus 3H1-pulsed DC alone (67% vs. 48%, respectively, p = 0.25), or in mice when given in combination with peptide-2-pulsed DC plus 3H1-pulsed DC compared with mice receiving peptide-2-pulsed DC plus 3H1-pulsed DC alone (76% vs. 56%, respectively, p = 0.08). This antitumor effect was antigen-specific, since groups of mice transplanted with nontransfected parental MC-38 cells and immunized with CpG ODN and/or peptide-pulsed DC plus 3H1-pulsed DC were not protected from tumor growth (Fig. 1c) and all mice died within 42 days (Fig. 1d).
The use of immunotherapy strategies to treat tumors has been most successful when the tumor burden has been minimal. However, it is not always possible to have a state of minimal residual disease. Therefore, we evaluated the effect that combining CpG ODN with DC-based therapy would have in the setting of large tumors. In the first set of experiments, therapy was initiated on day 10 after tumor transplant. The DC/CpG combination therapy improved the tumor free survival from 25 to 50% (p = 0.24) or from 33 to 50% (p = 0.44) for the groups of mice treated with peptide-1-pulsed DC plus 3H1-pulsed DC or peptide-2-pulsed DC plus 3H1-pulsed DC, respectively (Fig. 1e). The fact that combined therapy with systemic CpG ODN and DC-based vaccines lead to improved survival in mice with larger tumors is further evidenced by the regression of tumor in some mice when therapy was initiated on day 14 after tumor transplant. The combination therapy improved the tumor free survival from 0 to 42% (p < 0.04) or from 8 to 33% (p = 0.31) for the groups of mice treated with peptide-1-pulsed DC plus 3H1-pulsed DC or peptide-2-pulsed DC plus 3H1-pulsed DC, respectively (Fig. 1f). These data suggest that there may be an additive or synergistic effect of the systemically administered CpG ODN, even in the setting of large tumors.
More efficient induction of CEA-specific immune responses of DC-based therapy by coadministration with CpG ODN in vivo
The functions of CD8+ CTLs and CD4+ T helper lymphocytes in providing survival advantage of mice to CEA+ colon carcinoma by DC-based therapy along with CpG ODN as an adjuvant were examined. To detect the induction of antigen-specific cytolytic potential of the activated T-cell response in vivo, immune spleen cells were stimulated in vitro with the immunizing peptide and 3H1, and CTL activity was measured against different target cells in standard 51Cr-release assays. Antigen-specific CTL responses were induced in groups of mice immunized with CpG ODN and/or peptide-pulsed DC plus 3H1-pulsed DC, and exhibited lysis against MC-38-CEA-A2Kb cells (Fig. 2b) and T2 cells pulsed with immunizing peptide (Fig. 2c). Whereas, the use of nontransfected parental MC-38 cells (Fig. 2a) and T2 cells pulsed with an irrelevant peptide HIV-1 gp160120-128 or unpulsed T2 cells as targets resulted in background lysis (data not shown). The above results suggested that these CTLs were CEA specific and HLA-A2 restricted. Also, the extent of lysis of target cells by CTLs induced by peptide-2 proved to be more effective than peptide-1 at all E/T ratios tested (Figs. 2b and 2c). Furthermore, the cytotoxicity of CTLs induced in mice immunized with peptide-pulsed DC plus 3H1-pulsed DC along with CpG ODN against target cells was more potent than that of CTLs induced in mice immunized with peptide-pulsed DC plus 3H1-pulsed DC alone (p < 0.05, Figs. 2b and 2c), confirming the adjuvant activity of CpG ODN. Whereas, no significant lysis was observed when CTLs obtained from mice immunized with CpG ODN alone was used for analysis (Fig. 2b). As expected, the lysis of MC-38-CEA-A2Kb cells (Fig. 2d) was mediated by CD8+ T cells (p < 0.001) and was HLA-A2 restricted (p < 0.001). Next, we confirmed that CTL mediated lysis of target cells was mostly inhibited by concanamycin A, a potent inhibitor of perforin-mediated lysis,33 suggesting a major effector role for granule-mediated lysis in vitro (data not shown). Therefore, we wanted to examine whether there was any direct correlation between granule exocytosis and tumor killing. Degranulation of CTL was estimated by the release of BLT esterase activity (Fig. 2e). BLT esterase is stored in the granules together with perforin or granzymes, and its secretion correlates with the exocytosis of lytic granules. Supernatants from the coculture of CTLs and MC-38-CEA-A2Kb cells showed higher levels of BLT esterase activity in the groups of mice coadministered with CpG ODN and also the exocytosis of lytic granules was maximum where immunizations were performed with peptide-2-pulsed DC. The analysis of intracellular perforin expression from CTLs also provided the evidence that the use of CpG ODN as an adjuvant resulted in higher activation of CD8+ T cells (Fig. 2f). We further examined the specificity of CTLs by using peptide and MHC pentamer technology. Using A2Kb/CAP1-6D peptide pentamers to stain T cells in splenocytes, we have found significant expansion of CAP1-6D specific CD8+ T cells in mice immunized with peptide-2-pulsed DC plus 3H1-pulsed DC along with CpG ODN compared with mice immunized with peptide-2-pulsed DC plus 3H1-pulsed DC alone (Fig. 2g). These results suggest that CEA-peptide pulsed DC along with 3H1-pulsed DC are effective immunogen and possess a high efficiency of antigen transfer to lead to the induction of CEA-specific CTL response in vivo. We also confirmed that induction of CD8+ T cells by DC-based therapy along with CpG ODN as an adjuvant is stronger than that by DC-based therapy alone, further demonstrating the potent adjuvant effect of CpG ODN in the combined vaccination protocol.
Because effector CD8+ T cells are known to produce IFN-γ in an antigen-specific manner, which has been regarded as a reliable indicator for a Th1 response, we measured this cytokine after in vitro stimulation with peptide-pulsed T2 cells by ELISPOT assay. The results presented in Figure 3a show that HLA-A2-restricted, peptide-specific, IFN-γ-secreting cells were present in immunized mice, and the use of CpG ODN as an adjuvant in the DC-based therapy resulted in increased number of IFN-γ-secreting cells in immunized mice. The use of peptide-pulsed DC or CEA-pulsed DC as stimulant resulted in significant increase of IFN-γ-secreting cells and the numbers increased further in groups of mice treated with DC/CpG combined vaccines (Fig. 3a). However, IFN-γ expression by CD8+ T cells was not detectable in mice immunized with CpG ODN alone (data not shown). Next, the functional avidity of T cells was evaluated by the detection of IFN-γ spot in response to stimulation with T2 cells pulsed with diminishing amounts of immunizing peptide. The use of CpG ODN as an adjuvant in DC-based therapy resulted in increased numbers of IFN-γ spot by CD8+ T cells in response to low concentration of peptide (4 μg/ml, p < 0.04) and the response was relatively better in the group of mice immunized with peptide-2-pulsed DC compared with mice immunized with peptide-1-pulsed DC (Fig. 3c). These data suggest that inclusion of CpG ODN might have increased the overall avidity of CD8+ T cells for peptide, which resulted in better protection against CEA-expressing tumors. The induction of another proinflammatory cytokine, TNF-α could be detected by ELISPOT assay after antigen-specific stimulation in vitro. The data presented in Figure 3b suggest that the use of CpG ODN as an adjuvant also resulted in increased production of TNF-α. The enhanced cytolytic activity by antigen-specific CTLs could also be explained by tumor cell specific TNF-α production in immunized mice (Fig. 3d). Taken together, ELISPOT and cytotoxicity assay data clearly suggest that the use of CpG ODN as an adjuvant in DC-based therapy resulted in substantial increase in CEA-specific HLA-A2 restricted CTL responses in vivo, and the lytic activities correlated with survival of immunized mice.
CpG ODNs have been shown to activate B cells and plasmacytoid DC characterized by expression of costimulatory molecules, enhanced antigen presentation to T cells and secretion of T helper 1 (Th1)-promoting chemokines and cytokines.34 We also observed that the use of CpG ODN as an adjuvant in vivo resulted in increase in IL-2 (p < 0.04), IFN-γ (p < 0.03) and TNF-α secretion (p = 0.08), and decrease in IL-10 secretion (p < 0.05) in vitro (Figs. 4a and 4b). Intracellular cytokine analysis data also suggested that both CD4+ and CD8+ T cells were positive for cytokine secretion and were activated further by CpG ODN (Fig. 4d). The use of CpG ODN also resulted in increased secretion of monocyte chemotactic protein 1 (MCP-1, p < 0.01) but had little effect on the secretion of other 2 inflammatory chemokines, regulated upon activation, normal T cell expressed and secreted (RANTES) and macrophage inflammatory protein 1α (MIP-1α, Fig. 4c). The proof-of-principle of T-cell activation in vivo by CpG ODN treatment was also provided by flow cytometric analysis of decreased expression of CD62L on both CD4+ and CD8+ T cells (Fig. 4e).
T cells are involved in the enhanced antitumor immunity induced in mice after CpG ODN administration
CpG ODNs are known to have effects on stimulating NK cells and APCs, the latter leading to T-cell activation. Therefore, we undertook depletion studies using mAbs to T cells and/or NK cells to determine the effector cells responsible for the antitumor effects observed in DC-based therapy along with CpG ODN. The analysis of the data on day 28 post tumor challenge suggested that depletion of CD8+ T cells resulted in rapid tumor growth in mice and tumor growth rate was even faster when both CD4+ and CD8+ T cells were depleted (p < 0.05, Fig. 5b). All those mice were sacrificed within 7 to 8 weeks of tumor challenge and the difference in survival was significant compared with mice depleted with control antibody (p < 0.01, Fig. 5e). The depletion of CD4+ T cells or NK cells alone had partial effect (p = 0.50 and 0.46, respectively) on tumor growth and survival, and the depletion of CD8+ T cells along with NK cells also did not alter the antitumor effect of the DC/CpG combined vaccine compared with mice depleted of CD8+ T cells alone (p = 0.89). Thus, CD8+ T cells are the predominant effector cells responsible for the tumor rejection induced in mice that received systemic CpG and DC-based vaccines. Nevertheless, these results did not exclude the participation of the innate response in the antitumor effect of CpG ODN. A similar trend was observed in the groups of mice treated with peptide-2-pulsed DC plus 3H1-pulsed DC and depleted of T cells and/or NK cells (Figs. 5a and 5d).
Even though we did not find any significant role of NK cells in the antitumor effect of DC-based therapy, we could not rule out the possible involvement of other effector cells of the innate arm of the immune response that were activated by CpG ODN. IL-12, a key inducer of cell-mediated immunity is produced by activated DC and macrophages. The use of CpG ODN as an adjuvant increased the serum IL-12 in the treated mice and the depletion of T-cell subsets had no effect on IL-12 production compared with mice treated with DC-based vaccines alone (Fig. 5c). Therefore, we speculate that the increased amount of IL-12 in the serum was contributed by APCs activated by CpG ODN that could stimulate IFN-γ production by T cells as well as by NK cells.
CpG ODN enhances long-lasting immune response of DC-based therapy
One important hallmark of adaptive immune responses is the generation of antigen-specific memory. Our results indicated that CpG ODN strongly enhanced tumor-specific immune responses of DC-based therapy. Therefore we sought to investigate the impact of this adjuvant on the expression of Ly-6C and CD44, which are associated with activated/memory T-cell populations.35, 36 Results presented in Figures 6a and 6b show that the use of CpG ODN as an adjuvant increased the expression of Ly-6C and CD44 on both CD4+ and CD8+ T cells. When looking at Ly-6C, we found that expression of this surface marker was maximal at 60 days post tumor challenge and there was a 3-fold increase in expression. However, that expression diminished at later time point. The level of CD44 was also increased following immunization with CpG ODN and significant expression was observed at 90 days post tumor challenge. These results suggest that mice cured of MC-38-CEA-A2Kb tumors developed antigen-specific memory T cells and the use of CpG ODN as an adjuvant resulted in higher expression of Ly-6C and CD44.
The MC-38-CEA-A2Kb tumor growth was completely eradicated in mice, which received vaccination with CpG ODN and peptide-1-pulsed DC plus 3H1-pulsed DC (67% of the mice) or CpG ODN and peptide-2-pulsed DC plus 3H1-pulsed DC (76% of the mice) and survived for more than 90 days without any relapse of tumor (Fig. 1b). To test whether immunologic memory was developed in these tumor-free mice, they were subsequently rechallenged with the same tumor cells used for the first challenge or with MC-38 cells. Mice cured of MC-38-CEA-A2Kb tumors rejected subsequent challenges with the same tumor or nontransfected parental MC-38 cells and remained tumor-free until the end of the experiment (data not shown and Ref. 27). These results suggest that mice rejecting CEA-transfected colon carcinoma developed immunity to antigens relevant to MC-38-CEA-A2Kb tumor and also other antigens expressed on MC-38-CEA-A2Kb cells and “shared” with parental MC-38 cells, resulting in memory response against these tumors.
To observe the antitumor effects of splenocytes from MC-38-CEA-A2Kb tumor free mice which were immunized with CpG ODN and/or peptide-2-pulsed DC plus 3H1-pulsed DC, we further tested whether those cells could delay or abolish the outgrowth of CEA+HLA-A2+ LS 174T tumor cells in nude mice after the adoptive splenocyte transfer. Adoptive transfer experiments were performed in both prophylactic setting as well as in therapeutic setting. In prophylactic setting, the adoptive transfer of bulk splenocytes from peptide-2-pulsed DC plus 3H1-pulsed DC (Group 1) or CpG ODN and peptide-2-pulsed DC plus 3H1-pulsed DC (Group 2) immunized mice delayed the tumor growth in nude mice (Fig. 6c) and also improved the overall survival of these mice (Fig. 6d). However, the difference in survival was not significant compared with control group of mice (p = 0.30). Whereas in therapeutic setting, the adoptive transfer of bulk splenocytes resulted in substantial delay in tumor growth (Fig. 6c) and significant increase in survival compared with control group of mice (Fig. 6d). The adoptive transfer of bulk splenocytes from peptide-2-pulsed DC plus 3H1-pulsed DC (Group 3) or CpG ODN and peptide-2-pulsed DC plus 3H1-pulsed DC (Group 4) immunized mice resulted in 2 weeks (p = 0.12) and 3 weeks (p < 0.03) of survival advantage compared with control group of mice, respectively. These results intensively suggest that CEA peptide (CTL epitope)-pulsed DC and 3H1-pulsed DC along with CpG ODN as an adjuvant is an effective immunotherapeutic approach for colon cancer expressing CEA and HLA-A2.
DC-based vaccines are a promising approach to cancer treatment and a number of positive and negative studies evaluating their efficacy have been published. It now seems clear that the development of effective cancer vaccine protocols will require further optimization and refinement, a process greatly facilitated by appropriate animal models. Previously, we described that in a therapeutic setting with 7-day-old established tumors, DC-based vaccines stimulated an effective immune response against colon carcinoma.27 Since it is necessary to trigger innate immunity for subsequent effective acquired immunity,18 we attempted to maximize the antitumor activity induced by immunization with peptide-pulsed DC plus 3H1-pulsed DC by using CpG ODN as an adjuvant.
CpG ODN in general stimulates the innate immune system and antitumor studies have focused on developing CpG ODN as an adjuvant for tumor-associated antigens or cellular vaccinations.37 The published literature using CpG ODN as a single agent in models of cancer indicates that systemic or local CpG ODN administration alone can induce tumor regression in established tumors.19, 20, 38 Different CpG ODNs have distinct functions on the immune system, depending on the exact ODN sequence and modifications of the ODN backbone. Since the use of CpG ODN as a single agent was not sufficient for complete tumor regression in several model systems but enhanced the antitumor effects when used as an adjuvant with antibodies, tumor vaccines or peptides, it is possible that CpG ODN will function best as an adjuvant with other therapies for the treatment of malignancies. Recent advances in the understanding of the functional mechanisms involved in the TLR pathway activation by CpG ODN have resulted in several possible applications of its use as an immune adjuvant when combined in a multimodal fashion.
In our studies, CpG ODN 1826 had a very modest antitumor effect as a single agent when administered systemically after colon tumor was established (day 7 therapy). When administered with DC-based vaccines, the antitumor response with systemic CpG ODN was greatly enhanced. The exact mechanism for this enhanced antitumor response with DC/CpG combination therapy has not been fully elucidated, although the combined effect is mostly T-cell dependent. We have found a clear beneficial effect in mice immunized with CpG ODN along with DC-based vaccines. This trend toward better protection was correlated with increased in vitro cytotoxic activity, significant expansion of peptide-specific CD8+ T cells as measured by pentamer-staining, larger numbers of IFN-γ- or TNF-α-producing cells in ELISPOT assay, increased levels of CC chemokine (MCP-1) and type 1 cytokine secretion, and activation of both CD4+ and CD8+ T cells as evidenced by downregulation of CD62L expression. CpG ODNs are known to activate APCs and enhance the capacity of APCs to stimulate both CD4+ and CD8+ T-cell responses.39, 40 Immature DC treated with CpG ODN and cocultured with irradiated tumor cells have been shown to provide protection against tumor challenge in vaccinated mice.41 Although, we have not used CpG ODN for maturation of DC in vitro, it is likely that in vivo APC stimulation induced by CpG ODN contributed to the antitumor effects in our model of colon carcinoma. The immunization of tumor bearing mice with peptide-pulsed DC along with 3H1-pulsed DC induced T-cell activation in vivo and activated CTLs caused lysis and apoptosis of tumor cells. DC activated in vivo by CpG ODN could have taken up these tumor fragments and apoptotic tumor cells and presented the tumor antigens within the context of MHC molecules to CTLs, which then could eliminate remaining tumor cells.
In mice, CpG ODN trigger stronger Th1 responses than other known adjuvants in direct comparison, including CFA and BCG.37 In addition, this adjuvant induces large populations of antigen-specific effector CD8 T cells capable of expressing IFN-γ and TNF-α and displaying cytotoxic activity.40, 42 In support of its strong immunostimulatory qualities, several investigators have shown marked enhancement of the therapeutic responses by CpG ODN to other immunotherapies in mice, such as donor lymphocytes infusions, peptide-based vaccines and irradiated tumor cells.43, 44 Even in the absence of the immunologic stimulus associated with the addition of autologous tumor cells, CpG ODN has been shown to activate CD8+ T cells to induce regression of established tumors with a durable memory response in mouse tumor models.42, 19 Recently, using a murine B16 melanoma model, it has been reported that therapeutic efficacy of peptide-pulsed DC is enhanced by combined vaccination of CpG ODN that lead to tumor regression and long-term survival, which correlated with an enhanced antigen-specific T-cell response.21
NK cells can also be activated by CpG ODN, and activated NK cells are directly cytotoxic to tumor cells. Activated DC and NK cells produce IL-12 and IFN-γ, which cause further activation of T cells and NK cells for tumor cell lysis.45, 46 IFN-γ may also upregulate MHC class I expression on tumor cell surface, making them a better target for T-cell mediated lysis. In general, CpG ODNs with phosphorothioate-modified backbones are not as efficient in inducing NK-cell function but are more efficient in inducing IL-12 production than those with unmodified backbones.45 The expression of MHC class I molecules on APCs negatively regulates NK cell effector functions.47 MC-38-CEA-A2Kb tumor cells are highly positive for surface expression of MHC class I molecules and therefore it is speculated that in vivo tumor regression was not due to direct cytotoxic effect of NK cells in our model system.
We observed a significant increase of IL-12p40 secretion in mice immunized with CpG ODN along with DC-based vaccines. The production of IL-12p40, synthesized by macrophages, DC and granulocytes, directly leads to high levels of bioactive IL-12 secretion by DC when induced primarily by Th1-promoting innate stimulus48 and engages p19 subunit to form a cytokine IL-23, which triggers IFN-γ production of both naïve and memory human T cells.49 In our study, high quantity of IL-12p40 production by DC/CpG combined vaccination suggest the induction of both naïve T lymphocytes and proliferation of memory T cells, which are increasingly important in the vaccine approach to evoke strong and continued antitumor immune response.
We demonstrated that CpG ODN 1826 provided long-term maintenance of antigen-specific T cells induced by DC-based therapy leading to better protection against CEA+ colon carcinoma in adoptive transfer experiments. Although the precise mechanisms by which CpG ODN is capable of maintaining the antitumor effect generated by DC-based therapy are unknown, several possibilities can be explored. It has been reported that CpG ODN not only enhance but also maintain CD8+ effector CTL response through the expansion, inhibition of apoptosis and subsequent promotion of long-term survival of CD8+ effector and memory T cells.50, 51 Especially the expansion and survival of memory CD8+ T cells have been reported to be mediated by IL-15,52 which is produced by DC in response to type I IFN53 and CpG ODN stimulates DC to produce type I IFN.54
Taken together, the results presented in this article show the use of CpG ODN 1826 to activate and potentiate the effect of DC-based therapy in a murine tumor model. We postulate that CpG ODN is activating the DC in a way that overcomes the immunosuppressive effects of the colon tumor. The DC/CpG combined vaccination approach helped to mount strong immune responses that were effective in increasing the tumor-free survival in therapeutic setting, even in larger tumor burden. This vaccination strategy proved helpful in allowing the immune system to develop long-term memory response because splenic T cells isolated from DC/CpG immunized mice were able to transfer tumor specific immunity in naïve athymic nude mice. Thus, by optimizing DC immunization schedules, choosing the appropriate CpG ODN adjuvant and coupled with additional immunologic manipulations, such as removal of regulatory T cells, it is reasonable to continue to design clinical trials for the treatment of CEA+ colon carcinoma using a DC-based vaccine approach that incorporate existing technologies.
The authors would like to thank Dr. F. James Primus (Vanderbilt University Medical Center, Nashville, TN) for providing MC-38-CEA-A2Kb tumor cell line, and C57BL/6J-CEA and C56BL/6J-A2Kb transgenic breeder mice. They also wish to thank Mr. David Ginsburg for excellent assistance in animal studies and Ms. Mary B. Palascak for help in flow cytometry.
- 1American Cancer Society. 2008. Available at:http://www.cancer.org/docroot/CRI/CRI_2_3x.asp?dt=10. Accessed on June 4, 2008.
- 5The NCCN colon cancer clinical practice guidelines in oncology. JNCCN 2003; 1: 40–53., , .
- 15Improved efficacy of dendritic cell vaccines and successful immunization with tumor antigen peptide-pulsed peripheral blood mononuclear cells by coadministration of recombinant murine interleukin-12. Int J Cancer 1999; 80: 324–33., , , .