Murine Dendritic Cells Pulsed with an Anti-Idiotype Antibody Induce Antigen-Specific Protective Antitumor Immunity1

[CANCER RESEARCH 63, 2844–2854, June 1, 2003] Murine Dendritic Cells Pulsed with an Anti-Idiotype Antibody Induce Antigen-specific Protective Antitumor Immunity1

Asim Saha, Sunil K. Chatterjee, Kenneth A. Foon,2 F. James Primus, and Malaya Bhattacharya-Chatterjee3 Department of Internal Medicine and the Barrett Cancer Center, University of Cincinnati, Cincinnati, Ohio 45267 [A. S., S. K. C., M. B-C.], and Department of Pathology, Vanderbilt University Medical Center, Nashville, Tennessee 37232 [F. J. P.]

ABSTRACT upon reinjection into mice, a protective humoral response (10). Be- cause of these properties, much attention has been directed toward the In this study, using the carcinoembryonic antigen (CEA)-expressing use of DCs in vaccine strategies for the treatment of cancer. In this C15 murine colon carcinoma system in syngeneic C57BL/6 mice, we have regard, DCs pulsed with tumor-associated antigens in various forms, evaluated the efficacy of bone marrow-derived dendritic cells (DCs) pulsed with the murine anti-idiotype antibody 3H1 as a tumor vaccine. including whole cell lysate (11–14), peptides (15, 16), proteins (17, Anti-idiotype 3H1 mimics a distinct and specific epitope of CEA and can 18), RNA (19), DNA (20, 21), or DCs fused with tumor cells (22), generate anti-CEA immunity in mice, rabbits, monkeys, and humans have been studied for antitumor effects in experimental tumor models. when used with a conventional immune adjuvant. Our goal was to deter- Whole tumor cells have also been used for DC priming (23). In all of mine whether the use of DC as direct antigen-presenting cells would these approaches, there is a requirement for in vitro manipulation of improve the potency of 3H1 as vaccine. Bone marrow-DC pulsed with 3H1 the DCs to acquire and present tumor-specific antigens. and injected into naõ¬ve mice induced both humoral and cellular anti-3H1, A prerequisite for successful vaccination is the establishment of a as well as anti-CEA immunity. Specific killing of C15 cells in in vitro long-lasting protective immune response, and in this regard, immuno- antibody-dependent cellular cytotoxicity has been observed by immune therapy is an attractive approach to cancer therapy. The aim of sera. Immune-splenic lymphocytes when stimulated in vitro with 3H1 or ؉ immunotherapy is to induce or increase the ability of tumor-reactive CEA, showed increased proliferative CD4 Th1 type T-cell response and secreted significantly high levels of Th1 cytokines [IFN-␥, interleukin T cells to mediate antitumor immune responses in vivo. One area of (IL)-2] and low levels of Th2 cytokines (IL-4, IL-10). This vaccine also active immunotherapy involves the use of anti-Id antibodies. This idea -induced MHC class I antigen-restricted CD8؉ T-cell responses. The up- is based on Jerne’s network concept (24). According to this hypoth regulation of activation markers CD69 and CD25 on CD8؉ CTLs corre- esis, antibody of the type Ab2␤ represents essentially the internal lated with antigen-specific strong CTL responses in vitro. The immunity image of the target antigen, and this idiotopic antibody can be used as induced in mice resulted in a complete rejection of CEA-expressing C15 a surrogate TAA. We have generated a murine monoclonal Ab2␤ tumor cells in 100% of experimental mice, whereas no protection was designated 3H1 (25), which mimics a distinct and specific epitope of observed when 3H1-pulsed DC-vaccinated mice were challenged with the TAA, CEA. CEA is present at high density on tumor cells of CEA-negative MC-38 cells. The tumor rejection in 3H1-pulsed DC-treated Ͼ95% of colorectal, 70% of lung adenocarcinoma, and 50% of breast mice was associated with the induction of a memory response that helped cancer patients (26–28). We have shown previously that 3H1 can those mice to survive a second challenge with a lethal dose of C15 cells. induce anti-CEA antibody in small animals (25), nonhuman primates (29), and in humans (30–32). In these studies, although induction of INTRODUCTION anti-CEA humoral immunity by the surrogate antigen could be demonstrated, a definitive CTL response could not be shown. The objec- 4 DCs form a network of sentinel cells in the periphery to facilitate tive of the current study was to assess the ability of bone marrow- antigen delivery to secondary lymphoid organs (1). Transport of generated DCs pulsed with 3H1 to induce long-term antitumor antigen by DC is probably of key importance to initiate an immune immunity. response, thus permitting establishment of immunological memory (1–4). In addition, because of their excellent T-cell-stimulatory properties, small numbers of DCs are sufficient to induce an immune MATERIALS AND METHODS response in vivo (4). In different settings, these specialized antigen- Animals presenting cells can induce both the generation and proliferation of specific CTLs and Th cells via antigen presentation by MHC class I Female (6–8 weeks old) C57BL/6 (H-2b) mice were used in all experiments and class II molecules, respectively. In addition to activating naı¨ve T and were purchased from Harlan Laboratories (Indianapolis, IN). Mice were cells in the extrafollicular areas of secondary lymphoid organs, DCs housed in a specific pathogen-free environment and treated in accordance with may directly modulate B cell growth and differentiation (5). The role the guidelines established by the Animal Care and Use Committee of the of DCs in humoral responses has been documented in vitro (6) and in University of Cincinnati Medical Center. vivo (7–10). Notably, DCs incubated in vitro with antigen can induce, Generation, Culture, and Flow Cytometry of BmDCs

Received 1/7/03; accepted 3/28/03. BmDCs were generated as previously described (33) with some modifica- The costs of publication of this article were defrayed in part by the payment of page tions. Briefly, bone marrow was collected from tibias and femurs of female charges. This article must therefore be hereby marked advertisement in accordance with C57BL/6 mice, passed through a nylon mesh to remove small pieces of bone 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by NIH Grant RO1 CA86025. and debris, resuspended in complete medium [CM: RPMI 1640 containing 2 Present address: The University of Pittsburgh Cancer Institute, Pittsburgh, PA 15232. 10% heat-inactivated fetal bovine serum, 2 mM glutamine, 100 ␮g/ml strep- 3 To whom requests for reprints should be addressed, at Vontz Center for Molecular tomycin, 100 units/ml penicillin (all from Life Technologies, Inc., Grand Studies, Room 1316, University of Cincinnati, 3125 Eden Avenue, Cincinnati, OH Island, NY), and 50 ␮M 2-mercaptoethanol (Sigma, St. Louis, MO)], and 45267-0509. Phone: (513) 558-0425; Fax: (513) 558-1096; E-mail: malaya.chatterjee@ uc.edu. cultured in tissue culture dishes (Becton Dickinson, Franklin Lakes, NJ) for 4 The abbreviations used are: DC, dendritic cell; CEA, carcinoembryonic antigen; BmDC, 2 h. Nonadherent cells were collected and aliquots of 3 ϫ 106 cells were placed bone marrow-derived dendritic cells; Th, T helper; Id, idiotype; TAA, tumor-associated in a 60-mm dish (Becton Dickinson) containing 5 ml of CM with 10 ng/ml rm antigen; CM, complete medium; IL, interleukin; rhIL, recombinant human IL; mAb, mono- granulocyte macrophage colony-stimulating factor and 5 ng/ml rmIL-4 clonal antibody; PE, phycoerythrin; Ab3, anti-3H1 antibody; ADCC, antibody-dependent cellular cytotoxicity; ConA, concanavalin A; nm, recombinant murine; mIL, murine interleu- (PharMingen, San Diego, CA). Cells were cultured for 9 days. On day 9 of kin; mIFN, murine interferon. culture, most of the nonadherent cells had acquired typical DC morphology, 2844

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. ANTI-IDIOTYPE ANTIBODY-PULSED DC AND TUMOR IMMUNITY and these cells were used as the source of DCs in subsequent experiments. Briefly, spleen cells were isolated from immunized mice for use as effector Escherichia coli-lipopolysaccharide (0.2 ␮g/ml; Sigma) was added for the cells. Assays were performed in triplicate wells. Preimmune sera as well as final 2 days of cell culture. sera from mice immunized with 1A7-pulsed DCs were used as controls. To Expression of surface molecules was quantitated by flow cytometry using calculate lysis specifically attributable to ADCC, the percentage of lysis the following antibodies against: MHC molecules: H2K (FITC-conjugated), attributable to effector cells in the absence of the antibody was subtracted from I-A/I-E (FITC-conjugated); CD11c (FITC-conjugated); CD40 (PE-conjugat- each value obtained. ed); CD54 (FITC-conjugated); CD80 (FITC-conjugated); CD86 (PE-conjugated); and CD14 (PE-conjugated), all obtained from PharMingen. Species and T-Cell Proliferation Assays isotype-matched mAb were used as controls. For flow cytometry, aliquots of Spleens were harvested from mice 7–10 days after the last immunization or 2 ϫ 105 BmDCs were incubated with the mAbs for 30 min at 4°C. The cells harvested after 5 weeks of posttumor challenge. These cells (2 ϫ 105/well) were washed twice in fluorescence-activated cell sorting buffer [PBS ϩ 0.1% were cultured in absence and presence of different stimulants (0.5–1 ␮g/well). fetal bovine serum (v/v) ϩ 0.02% NaN ] and subsequently analyzed in a flow 3 In select experiments, peptides were used as stimulants at a concentration of 5 cytometer (EPICS XL/MCL, Beckman-Coulter Inc., Hialeah, FL). Data were ␮g/well. T-cell proliferation was measured by [3H]thymidine incorporation as analyzed using WinList software (version 5.0). Live cells were selected for described previously (36). Assays were performed in triplicate wells. In analysis using forward versus side scatter gating. parallel experiments, cell-free supernatants were harvested after 48 and 72 h of culture for quantitation of IL-2, IL-4, IL-10, and IFN-␥, respectively. Anti-Id Vaccine and Peptide Cytokine Analysis Generation, purification, and characterization of anti-Id antibody 3H1 designated as CeaVac have been described earlier (25). One control anti-Id Supernatant from T cell cultures were analyzed for the presence of IL-2, antibody, 1A7 (34), which mimics a melanoma-associated antigen, ganglioside IL-4, IL-10, and IFN-␥ by ELISA using commercially available kits (R&D GD2, was also used in this study. Peptides were designed based on the amino Systems). All samples were tested in triplicate. The lower limit of detection for acid sequence homology of 3H1 and CEA and were synthesized as described cytokines was 5 pg/ml each. previously (35). The purity of the peptides was Ͼ90% by high-performance liquid chromatography analysis. Antigen-specific CTLs

Antigen Pulsing of DCs Splenocytes were harvested and pooled from 3 mice/group 7–10 days after the last immunization or after 5 weeks of posttumor challenge. These cells For antigen pulsing, bone marrow-derived day 8 DCs were incubated with (2 ϫ 106/ml) were restimulated by culturing with 3H1 (25 ␮g/ml) and 20 anti-Id mAb, 3H1, or isotype-matched control anti-Id mAb 1A7 for overnight units/ml rhIL-2 (Sigma). After 5 days of culture, the in vitro restimulated at 37°C, in presence of 120 ␮g/ml antigen. The nonadherent cells were then viable splenocytes were examined for specific cell lysis using standard 6 or collected and were mainly DCs (referred to as antigen-pulsed DCs). At this 18-h 51Cr release assays (15–17) against a variety of target cell lines. Assays time point supernatants from unpulsed or antigen-pulsed DC cultures were were performed in triplicate wells. The spontaneous release of all assays was harvested for measurement of IL-12p70 production by standard ELISA kit Ͻ25% of the maximum release. (R&D Systems, Minneapolis, MN). Antibody-blocking experiments with anti-CD8 (53–6.7) or anti-CD4 (GK1.5) mAb were performed to establish the effector cell phenotype respon- Immunization Protocol sible for specific killing of C15 cells and were carried out by preincubating effector cells for 60 min at 37°C. Antibody-blocking experiments with anti- Before immunization, unpulsed or antigen-pulsed 9-day-cultured BmDCs H-2Kb/H-2Db or anti-I-Ab mAb were performed to demonstrate MHC class I were washed three times in PBS. A total of 2 ϫ 105 DCs in 0.1 ml of PBS were antigen restriction of CTL response against C15 cells and were carried out by injected s.c. in the lower right flank of syngeneic C57BL/6 mice. Each mouse preincubating labeled target cells for 30 min at 37°C. Isotype-matched mAbs received three immunizations every other week. One group of mice received were used as control. Effector or labeled target cells were preincubated with 5 immunizations with 3H1-pulsed DCs; the other group received immunizations ␮g/ml antibody before the addition of untreated labeled target or effector cells with unrelated anti-Id 1A7-pulsed DCs. A third group received immunizations respectively. All antibodies used in blocking experiments were obtained from with unpulsed DCs, and a mock vaccination with PBS was also performed in PharMingen. another group of mice for comparison. All mice were bled 7–10 days after each ϩ ϩ In select experiments, purified CD8 and CD4 T cells were obtained from immunization and sera were stored at Ϫ20°C. antigen-specific CTL culture by magnetic bead separation of T cells using either anti-CD8 (53–6.7) or anti-CD4 (GK1.5) magnetic beads, according to Determination of Antigen-specific Antibody Levels the manufacturer’s specifications (Mini MACS; Miltenyi Biotec, Auburn, CA). RIA for detection of Ab3 was performed as described previously (36). All samples were passed through the column twice and were 95% pure as ϩ ϩ Briefly, microtiter plates coated with 3H1 were incubated with diluted mice determined by flow cytometry. These purified CD8 and CD4 T cells were sera. After washings, the antigen-antibody reaction was tagged using 125I- used as effector cells in CTL assay. labeled 3H1. Assays were performed in triplicate. Sera obtained from mice Flow Cytometry of CTL Culture immunized with 1A7-pulsed DCs were used as control. ELISA for detection of anti-CEA antibodies was performed as described Mice immunized with 3H1-pulsed DCs or 1A7-pulsed DCs and challenged previously (36). Assays were performed at least in triplicate for each sample. with C15 tumor cells were sacrificed after 5 weeks of posttumor inoculation, In select experiments, alkaline phosphatase-conjugated anti-IgG1, anti-IgG2a, and splenocytes were harvested and restimulated by culture with 3H1 and anti-IgG2b, anti-IgA, or anti-IgM (1:1500; Southern Biotechnology, Birming- rhIL-2 as described above. After 5 days of culture, viable splenocytes were ham, AL) antibodies were used for subclass determination. used for staining. Antibodies (PharMingen) used to phenotype the cells were Idiotope and epitope analysis of Ab3. Experiments were performed as anti-CD28-PE (hamster antimouse), anti-CD4-PE (rat antimouse), anti- described previously (36). The serum from each mouse was checked for the CD8-PE (rat antimouse), anti-CD25-FITC (rat antimouse), and anti-CD69- 125 ability to inhibit the binding of I-labeled 8019 (Ab1) to 3H1 bound to FITC (hamster antimouse). For two-color immunofluorescence, cells were microtiter plates. Also, the inhibition of 8019 binding to CEA by murine sera stained simultaneously with FITC-and PE-labeled antibodies for 30 min at was tested by RIA. Pooled serum from normal mice was used as control in 4°C, washed twice, and analyzed. Before analysis, compensation for two-color these experiments. flow cytometry has been carried out.

ADCC Murine Model for Determining the Efficacy of 3H1 Vaccine

The ability of Ab3 to lyse CEA-positive tumor cells in conjunction with A murine tumor model expressing human CEA was developed by Dr. F. effector cells was tested by standard ADCC as described previously (36). James Primus at Vanderbilt University (Nashville, TN). Murine colorectal 2845

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. ANTI-IDIOTYPE ANTIBODY-PULSED DC AND TUMOR IMMUNITY carcinoma cells, MC-38, were transduced with human CEA (37). The trans- s.c. tumors, 3H1-specific CTLs were generated from tumor-free mice as duced cell line, C15, constitutively express CEA in culture. When the stably described above and then i.v. infused into naı¨ve mice (3 ϫ 106 cells/mouse). transfected CEA-transduced cells (1 ϫ 106 cells/mouse) or nontransduced cells One day later, these mice were s.c. challenged with 1 ϫ 106 C15 cells and (5 ϫ 105 cells/mouse) are injected s.c. into syngeneic C57BL/6 (H-2b) mice, monitored for tumor growth and survival. tumors develop in 100% of the mice within 10–15 days. Experiments included 10 mice/group. Statistical Analysis

Cell Lines Student’s t test was applied to perform statistical analysis using SigmaStat software (Jandel, San Rafael, CA). P Ͻ 0.05 was considered to indicate EL-4 is a T thymoma derived from the C57BL/6 mouse. YAC-1 cells are statistical significance. sensitive to natural killer cells. MC-38 is a chemically induced murine colon carcinoma cell line, and the CEA-expressing (clone C15-4.3) MC-38 transfectant has been described previously (37). RESULTS Survival of Mice Immunized with 3H1-pulsed DCs and Cell Surface Phenotype of and IL-12 Production by BmDCs. Control Vaccines The culture conditions resulted in the generation of BmDCs with Mice were immunized with 3H1-pulsed DCs or control vaccines (1A7- typical morphology and fine membrane projections. DCs were stained pulsed DCs or unpulsed DCs) or PBS as described above. Each group of and analyzed by flow cytometry for surface expression of MHC immunized mice was additionally divided into two equal groups for tumor molecules, costimulatory molecules, and antigens associated with challenge. Tumor challenge was performed s.c. with 1 ϫ 106 of CEA-positive maturation. BmDCs expressed high levels of the DC-associated C15 cells or 5 ϫ 105 of CEA-negative MC-38 cells in the lower left flank. marker CD11c, the costimulatory molecules CD80, CD86, CD40, and Growth of tumor and survival were monitored daily. Tumor size was assessed CD54, and MHC class I/II (Fig. 1). To assess the percentage of 2 once a week and recorded as a tumor area (in mm ) by measuring the largest contaminating macrophages in the culture conditions, the expression perpendicular diameters with vernier calipers. All experiments were performed of the macrophage-related surface molecule CD14 was determined. two to three times using individual groups of 12–24 mice. Mice were sacrificed Ͻ when tumors became ulcerated or when they reached a size Ͼ250 mm2. The percentage of cells expressing CD14 was 5%, indicating low numbers of contaminating macrophages in the BmDC cultures. Secondary Tumor Challenge in Tumor-free Mice We have evaluated the production of a Th1-polarizing cytokine, IL-12, by 1 ϫ 106 DCs pulsed for overnight with anti-Id antibody as a source of To determine the persistence of tumor-specific immunity in the mice im- antigen. As shown by a representative experiment (of three performed) in munized with 3H1-pulsed DCs at day 90 after first tumor inoculation, mice that had rejected C15 tumors were given a second s.c. tumor challenge. These mice Table 1, DCs cultured with anti-Id antibody, 3H1, produced higher were divided into two equal groups for tumor challenge. Tumor challenge was amounts of IL-12p70 as compared with unpulsed DCs. performed with C15 or MC-38 cells (tumor cells in 100 ␮l of PBS in the lower Antigen-pulsed DCs Can Induce Antigen-specific Humoral flank) as described above. As a control, naı¨ve mice were challenged in a Response in Vivo. To evaluate the ability of DCs to induce an in vivo parallel fashion. The mice were followed for survival and measurement of humoral response, we have measured the titer of anti-anti-Id antitumor size. bodies in the sera of immunized mice. The animals were bled 7–10 days after each immunization, as indicated in the figure legend 2. Ab3 Adoptive CTL Transfer for Tumor Prevention produced in the sera of these mice was estimated by a RIA. Data in To determine whether the tumor-specific CTLs induced by BmDCs via s.c. Fig. 2 show that levels of Ab3 in 3H1-pulsed DC-immunized mice injections acquiring tumor antigens in vivo were responsible for rejection of sera are low but detectable after first immunization. However, signif-

Fig. 1. Surface phenotype of BmDCs generated under culture conditions as described in “Materials and Methods.” Cells were analyzed by flow cytometry. ⅐⅐⅐⅐ shows the fluorescence-activated cell sorting profiles after staining with isotype control mAb. The data are representative of three different experiments.

2846

Table 1 IL-12 production by BmDCs generated under culture conditionsa serum as described in “Material and Methods.” As presented in IL-12p70 secretion Table 2, Ͼ40% inhibition of radiolabeled Ab1 binding to Ab2 was Group (pg/ml Ϯ SE) observed by the Ab3 sera at a 1:20 dilution. We additionally analyzed 3H1-DCs 1824 Ϯ 52.0b the binding inhibition of radiolabeled 8019 to CEA in presence of Ϯ Unpulsed DCs 1069 43.3 mouse Ab3 serum. Sera from 5 of 5 mice inhibited this binding, and a 6 BmDCs (1 ϫ 10 /ml) were cultured in medium as described in “Materials and the inhibition was significant at 20-fold dilution of the serum. These Methods.” Day 8 BmDCs were left unpulsed or were cocultured with anti-Id antibody, 3H1. Protein level of cytokine secreted in supernatants during the last 18 h of the 9-day results suggested that the Ab3 induced in mice immunized with culture period were determined by standard ELISA. These data are representative of three 3H1-pulsed DCs share the same Id as Ab1 and may also contain Ab1Ј independent experiments with similar results. b P ϭ 0.1 with respect to unpulsed DCs. antibodies. ADCC by Immune Sera. To determine whether the Ab1Ј generated by 3H1-pulsed DC immunization is cytolytic for CEA-positive tumor cells such as C15, ADCC was determined in vitro. As shown in Fig. 4, significant ADCC was observed in mice immunized with 3H1-pulsed DCs, whereas ADCC with preimmune sera or sera from mice immunized with unpulsed DCs or control 1A7-pulsed DCs was negligible. MC-38 cells lacking CEA expression as targets resulted in a background lysis, suggesting that tumor cell lysis was dependent on antigen expression by target cells. Proliferation of T Cells in Presence of Proteins and Peptides. Assuming that antigen-pulsed DC prime naı¨ve T cells toward antigen- specific T cells and that these cells predominantly proliferate when

Fig. 2. Induction of anti-anti-Id antibody (Ab3) in mice by immunization with 3H1- pulsed DC. Ab3 generation was assayed by a sandwich RIA with 3H1-coated plates. Sera were diluted 1:20 with PBS. Data included 5 individual mice/group, and the experiment was repeated at least three times. icant Ab3 was generated after three immunizations. For control, sera from mice immunized with 1A7-pulsed DCs were used for 3H1 binding assay. Radioactive 3H1 bound by control sera were equal to or less than the preimmune sera (data not shown). Anti-CEA antibody developed by mice immunized with 3H1- pulsed DCs was determined by ELISA. Ab3 antibody induced by Ab2␤ is polyclonal, and a subgroup of the Ab3 has a similar binding site (paratope) as Ab1. This type of Ab3 is also known as Ab1Ј to indicate that it might differ in its other idiotopes from Ab1. Ab1Ј recognizes the cognate TAA and is the only antibody in the Ab3 population that is likely to have the antitumor property. As shown in Fig. 3A, the level of Ab1Ј was detectable after first immunization in sera obtained from mice immunized with 3H1-pulsed DCs. Signifi- cant level of Ab1Ј appeared after the third immunization. Sera from mice after each immunization with 1A7-pulsed DCs were used as control in this assay. Preimmune serum (1:20 dilution) had an absorbance of 0.20 at 405 nm. The selective activation of cloned murine CD4ϩ cells representing Th1 and Th2 cells depends on unique cofactors produced by special- Fig. 3. Induction of Ab1Ј antibody in mice immunized with 3H1-pulsed DCs. Ab1Ј was assayed by ELISA with CEA-coated, 96-well plates as described in “Materials and ized antigen-presenting cell populations (38). Because the regulation Methods.” A, absorbance at 405 nm of the sera diluted 1:20 with PBS after each of isotype switching in vivo is dependent upon the activation of immunization; B, absorbance at 405 nm of the sera drawn after three immunizations for detection of isotypes after 1:20 dilution with PBS. Data represent means Ϯ SE of 5 different types of T-helper cells (39), we investigated the class dis- individual mice/group. One of two experiments is shown. tribution of the specific anti-CEA antibodies. Fig. 3B summarizes the isotypes of the CEA-specific antibodies from mice tested in two independent experiments. Specific IgG2a antibodies mostly were de- Table 2 Idiotype analysis of Ab3 induced in mice immunized with 3H1-pulsed DCsa tected in this group of 3H1-pulsed DC-immunized mice, suggesting Percentage Ab1 (8019) Percentage Ab1 (8019) that this isotype may be relevant for the in vivo resistance to tumor binding inhibition to binding inhibition to Ab2 (3H1) CEA growth. Enhanced production of IgG1 reflects a Th2 response in mice, whereas predominant IgG2a production indicates a Th1 response (40). 3H1- 1A7- 3H1- 1A7- Serum pulsed pulsed pulsed pulsed Murine IgG2a is more effective than IgG1 at mediating antibody- dilution DC DC DC DC dependent cellular cytotoxicity (41). Indeed, among the antibodies, 1:20 44 Ϯ 55Ϯ 225Ϯ 34Ϯ 2 murine IgG2a have been reported to be more effective than other 1:40 30 Ϯ 43Ϯ 216Ϯ 33Ϯ 2 isotypes in the suppression of tumor growth in a number of animal a Procedures for analysis of Ab3 induced in mice immunized with 3H1-pulsed DCs or models (42, 43). 1A7-pulsed DCs vaccine are described in “Materials and Methods.” Serum from 5 mice were separately assayed to obtain the mean and SE. Preimmune sera at 1:20 dilution were Ab3 induced in mice by 3H1-pulsed DC vaccine was analyzed by used as control, which showed Ͻ5% binding inhibition. In CEA binding inhibition assay, the inhibition of binding of Ab1 (8019) to Ab2 (3H1) by the mouse preimmune sera at 1:20 dilution showed no binding inhibition. 2847

of the Th2-associated cytokines IL-4 and IL-10 obtained from the cultures were much lower than that of the Th1-associated cytokines. Induction of CEA-specific Cytolytic T-Cell Response in Mice Immunized with 3H1-pulsed BmDCs. To assess whether DCs pulsed with soluble 3H1 protein are capable of inducing cytolytic response in vivo, spleen cells from immunized mice were harvested 7–10 days after the last immunization and restimulated in vitro in presence of 3H1 and rhIL-2 for 5 days, and these cells were used as effector cells. Effector cells obtained from mice immunized with 3H1-pulsed DCs lysed the CEA-expressing murine colon carcinoma cells, C15 significantly (Fig. 6A). Tumor cell lysis was antigen specific because the use of MC-38 parental tumor cells lacking CEA expression as targets resulted in a background lysis. The lysis of C15 tumor cells was significantly greater (P Ͻ 0.005) than that obtained Fig. 4. ADCC by serum from immunized mice. Data represent means Ϯ SE of specific with MC-38 targets at all E:T cell ratios tested. The use of nonspecific lysis by each vaccine group consisting of three mice. Sera were obtained from mice EL-4 thymoma cells lacking CEA expression as targets also resulted immunized three times with either 3H1-pulsed DCs or 1A7-pulsed DCs or unpulsed DCs. Specific cell lysis was determined by 6-h 51Cr release assay using C15 and MC-38 as target cells. Sera were diluted 1:5 and 1:10 with PBS. A representative experiment of three is shown.

exposed to antigen, we assessed the proliferation of spleen cells from immunized mice in presence of different stimulants. For positive control, 1.0 ␮g of ConA was used in these experiments. Spleens from at least 3 mice in each group, selected at random, were pooled for use in this assay. Significant proliferation was obtained in presence of 3H1 or CEA from spleen cells isolated from mice immunized with 3H1-pulsed DCs (Fig. 5A). Spleen cells from 1A7-pulsed DCs, unpulsed DCs, or PBS-vaccinated mice showed negligible stimulation with either 3H1 or CEA. The proliferation of T cells from mice after three immunizations with 3H1-pulsed DCs, in the presence of 1A7 as the stimulant in the assay, was significantly lower than the value obtained with 3H1 (P Ͻ 0.005) as the stimulant. However, spleen cells harvested from 1A7-pulsed DC-immunized mice and placed in secondary culture in vitro in presence of 1A7 resulted in 1A7-specific proliferative response. To analyze the subsets of T-cell populations that become respon- sive by 3H1-DC vaccination, CD4ϩ and CD8ϩ T cells were separated from immunized mice splenocytes, and their proliferation was analyzed in presence of different stimulants. From our previous study (35), we have also considered two peptides for analysis, one from 3H1 (LCD-2) and the other from CEA (CEA-B), containing common sequence LIDG. Besides LCD-2 and CEA-B, two control peptides, HFW-1 and CAP-1, were also included for analysis. As illustrated in Fig. 5. Proliferation of splenocytes from mice after three immunizations in the presence ϩ Fig. 5B, the bulk splenocyte populations and purified CD4 cells of different stimulants. T-cell proliferation was assayed by determining [3H]thymidine responded to 3H1 and peptides (LCD-2, CEA-B), whereas CD8ϩ incorporation, and proliferation in presence of media has been deducted from values obtained in presence of different stimulants. Results are the mean Ϯ SE of one repre- T-cell populations were unresponsive. No population of lymphocytes sentative experiment of three performed. A, stimulants were used at 0.5–1 ␮g/well. B, proliferated in presence of control anti-Id antibody 1A7 or control analysis of subsets of T cells in splenocytes after stimulation with anti-Id antibodies and peptides. Splenocytes were obtained from mice immunized three times with 3H1-pulsed peptides HFW-1 and CAP-1 (Fig. 5B). It may be possible that purified ϩ ϩ ϩ ϩ DCs. CD4 and CD8 T cells were separated by microbeads as described in “Materials CD8 T cells were relatively unresponsive in vitro as CD4 T-cell and Methods.” Macrophages isolated from splenocytes were used as antigen-presenting help was not provided. cells in this assay. Cytokine Production by Activated T Cells. Because cellular immunity appears to be critical in mediating antitumor effects, we Table 3 Th1-associated immune response in splenocytes from immunized micea analyzed the pattern of cytokines released in vitro by stimulated pg/ml CD4ϩ T cells of immunized mice. Th1 subset of CD4ϩ T-helper cells secretes IL-2 and IFN-␥, whereas Th2 subset secretes IL-4, IL-5, and Immunized with mIL-2 mIFN-␥ mIL-4 mIL-10 IL-10. To determine whether the stimulated CD4ϩ cells constitute 3H1-DCs 960 Ϯ 32 5340 Ϯ 80 234 Ϯ 16 104 Ϯ 12 1A7-DCs 204 Ϯ 12 480 Ϯ 15 170 Ϯ 12 96 Ϯ 8 mainly Th1 or Th2 helper cells, freshly isolated splenocytes were Unpulsed DCs 16 Ϯ 430Ϯ 516Ϯ 48Ϯ 4 cultured in vitro in presence of 3H1, and the culture supernatant was 5 a Splenocytes (2 ϫ 10 /well) obtained from different groups of immunized mice were then analyzed by ELISA for levels of mIL-2, mIFN-␥, mIL-4, and cocultured in presence of 3H1 (1.0 ␮g) in 96-well flat-bottomed microtiter plates in a final mIL-10 production. A Th1-associated response was observed, with a volume of 200 ␮l. Cell-free supernatants were harvested after 48 h (for quantitation of ␥ IL-2, IL-4, IL-10) and 72 h (for quantitation of IFN-␥) of culture and cytokine levels were significant enhancement of mIL-2 and mIFN- production in the measured by standard ELISA kits. Data represent means Ϯ SE of triplicate cultures. One group of mice immunized with 3H1-pulsed DCs (Table 3). The levels representative experiment of two is shown. 2848

Fig. 6. Induction of CEA-specific lytic activity against tumor cells by immunization with DCs pulsed with 3H1. Spleens were harvested from mice immunized three times with 3H1-pulsed DCs (A), or 1A7-pulsed DCs (B), or unpulsed DCs (C), or mock vaccinated with PBS (D). Lympho- cytes were restimulated in vitro in the presence of 3H1 and rhIL-2. Five days later, cytotoxic activity was tested by 6-h 51Cr-release assay using the targets indicated in each panel. Experiments were performed using 6–9 mice/group. These data are representative of three experiments performed. In E and F, 3H1-pulsed DC-immunized mice splenocytes were restimulated in vitro in presence of either 3H1 or hCEA or 1A7 and rhIL-2. Five days later, cytotoxic activity was measured by 6-h 51Cr release assay using C15 (E) and MC-38 (F) as target cells. In G, 3H1-pulsed DC-immunized mice splenocytes were restimulated in vitro in presence of 3H1 and rhIL-2. Five days later, stimulated cells were tested for their ability to lyse the DC target cells pulsed with either 3H1 or hCEA or 1A7 as indicated in the panel. Each experiment included 6 mice and was repeated at least twice.

in a background lysis. The nonspecific killing activity was measured Splenocytes from 3H1-DC-vaccinated mice were also stimulated in by lysis of natural killer-sensitive YAC-1 cells. However, the lysis of in vitro culture with either 3H1 or CEA or 1A7 and tested 5 days later C15 cells with spleen cells from all 3H1-DC immune mice was for lytic activity on CEA transfected C15 or parental MC-38 cells. significantly greater (P Ͻ 0.005) than that obtained with control Splenocytes stimulated in vitro with either 3H1 or CEA-lysed C15 targets. Spleen cells from an additional 6 immunized mice produced cells specifically (Fig. 6E) but not MC-38 cells (Fig. 6F). However, similar results. There was no difference in lysis between the control splenocytes stimulated in vitro with control anti-Id antibody, 1A7, did targets (P Ͼ 0.1). Similar results were obtained when additional not display any significant in vitro cytotoxicity (Fig. 6, E and F). The immunized mice were studied. DCs pulsed with unrelated anti-Id lytic activity against C15 cells was significantly higher with CTLs antibody, 1A7, did not induce CEA-specific cytolytic response in derived from both splenocytes stimulated with either 3H1 (P Ͻ 0.005) vivo. Only background lysis was observed with splenocytes obtained or CEA (P Ͻ 0.003) than that obtained with control target cells at all from this group of immunized mice (Fig. 6B). Similar results were E:T cell ratios tested. obtained from other control animals immunized either with DCs alone Moreover, that DCs exposed to 3H1 can induce antigen-specific (Fig. 6C) or mock vaccinated with PBS (Fig. 6D). Note, however, that CTL response was shown by the experiment in which DCs were used immunization of mice with DCs generated some nonspecific lytic as targets. 3H1-DC-vaccinated mice splenocytes stimulated in vitro activity, which is typically seen with BmDCs as evidenced by previ- with 3H1 were used as effector cells. As expected, stimulated cells ous studies (16, 44). could not lyse DCs pulsed with 1A7 (Fig. 6G) effectively, but DCs 2849

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. ANTI-IDIOTYPE ANTIBODY-PULSED DC AND TUMOR IMMUNITY were lysed if pulsed with 3H1 (P Ͻ 0.04) or CEA (P Ͻ 0.009). This result indicates that DCs can process the soluble protein for antigen- specific MHC class I presentation. DCs Pulsed with 3H1 Induces Specific MHC Class I Antigen- restricted CD8؉ T-Cell-mediated Cytotoxicity. To determine whether the antitumor cytolytic activity was associated with the presence of tumor-specific CD8ϩ CTLs and/or CD4ϩ helper T lymphocytes, anti-CD8 or anti-CD4 mAb was incorporated as blocking reagents in cytotoxicity assay. The specific killing of C15 cells was significantly inhibited by preincubation of effector cells with anti- CD8 mAb (Fig. 7A), whereas the antibody against CD4 was relatively less effective. This inhibition ranged from 80 to 92% at all E:T cell ratios tested compared with the isotype-matched control antibody Fig. 8. DC-3H1 vaccination protects C57BL/6 mice from a subsequent challenge with Ͻ syngeneic tumor cells. Groups of mice were immunized as described in “Materials and (P 0.001). The inhibition of CTL activity with anti-CD4 mAb Methods.” Two weeks after last immunization, all mice were challenged s.c. with 1 ϫ 106 ranged from 8 to 30% at all E:T cell ratios tested (P Ͼ 0.1). This result viable C15 cells. Tumor growth was measured over time. All experiments included 6–12 indicates that antitumor response observed was primarily mediated by mice/group. These data are representative of three experiments with similar results. CD8ϩ CTLs. Antibody-blocking experiments with anti-MHC class I or anti- MHC class II mAb were also performed to determine the MHC class CTL activity was obtained by treatment with anti-H-2Kb/H-2Db mAb I antigen-restricted CTL specificity for C15 target cells. Inhibition of (Fig. 7B). This inhibition ranged from 78 to 83% at all E:T cell ratios tested compared with isotype-matched control antibody (P Ͻ 0.02). In contrast, the inhibition of CTL activity with anti-I-Ab mAb ranged from 0 to 25% (P Ͼ 0.1), which indicated MHC class I-restricted tumor cell lysis. Additional experiments were performed to establish that the CD8ϩ effector cell population was responsible for specific killing of C15 cells. Immunized mice splenocytes were restimulated in vitro in presence of 3H1 and rhIL-2. On day 5, CD8ϩ and CD4ϩ T cells were purified from this stimulated culture using microbeads and used separately for cytotoxicity assay. The assay was performed at lower E:T cell ratios because of scarcity of purified effector cells. As shown in Fig. 7C, CD4ϩ T cells did not display any significant in vitro cytotoxicity, whereas CD8ϩ T cells had significant antigen-specific lytic activity (P Ͻ 0.001), indicating that CD8ϩ CTLs were primarily responsible for antigen-specific lysis induced by BmDCs. Vaccination with 3H1-pulsed DCs Protects Mice against a Challenge with CEA-transfected Tumor Cells. To determine whether the immune response induced by protein-pulsed DCs can induce resistance against a challenge with live tumor cells 2 weeks after the last immunization, each group of immunized mice were divided into two equal subgroups, and one group was challenged with CEA-expressing C15 cells, whereas the second group was challenged with CEA-negative MC-38 cells. The group of mice immunized with 3H1-pulsed DCs were completely protected from C15 tumor growth (Fig. 8) and from death (Fig. 9A). Tumors in control mice (immunized with unpulsed DCs or PBS) grew progressively (Fig. 8) and were lethal (Fig. 9A). Mice immunized with DCs pulsed with 1A7 were not protected (Figs. 8 and 9A), suggesting that s.c.-injected DCs do not induce tumor immunity by antigen-independent mechanisms in this model. It is also unlikely that protection was the result of carryover of free 3H1 because the 3H1-pulsed DCs were extensively washed before injection. Also, 3H1 alone does not induce any immune responses in mice (unpublished data). Furthermore, mice immunized with 3H1-pulsed DCs were not protected from challenge with nontransfected parent MC-38 cells (Fig. 9B), indicating that protective Fig. 7. MHC class I antigen-restricted lysis of tumor cells by CTLs. Mice were immunity was antigen specific, depending on CEA expression by the immunized three times with 3H1-pulsed DCs, and spleens were harvested 7–10 days after last immunization. Lymphocytes were restimulated in vitro in presence of 3H1 and rhIL-2 tumor target. ؉ for 5 days. Cytotoxic activity was then measured by 6-h 51Cr-release assay using C15 as Th1-associated Cytokines and CD8 T Cells Are Involved in target cells. A, blocking of cytotoxicity was performed in presence of anti-CD4, anti-CD8, the DC-mediated Antitumor Effects. To additionally understand or isotype-matched control mAb. B, blocking of cytotoxicity was performed in presence of anti-MHC class II mAb (I-Ab), anti-MHC class I mAb (H-2Kb/H-2Db), or isotype the mechanism responsible for complete protection of s.c. tumors, matched control mAb. C, CD4ϩ and CD8ϩ T cells were purified from restimulated tumor-protected mice were selected at random after 5 weeks of lymphocytes using microbeads and used as effector cells. Restimulated total lymphocytes were also used. Each experiment included 6 mice. Data are representative of two posttumor challenge, and cellular immune responses were measured experiments performed. by (a) T-cell proliferation assay, (b) in vitro cytokine release pattern, 2850

unpulsed or control protein-pulsed DCs to elicit a detectable level of protective immunity in mice inoculated with either C15 or MC-38 tumor cells (Figs. 9, A and B). The role of CD8ϩ T cells in antitumor immunity was additionally suggested by up-regulation of the activation markers such as CD69 (46), CD25, and CD28 on T cells. The data depicted in Fig. 11A indicate that 1.4% of CD8ϩ and 1.5% of CD4ϩ cells expressed CD69, whereas 4.2% of CD8ϩ and 1.5% of CD4ϩ cells expressed CD25 when effector cells obtained from tumor-bearing 1A7-DC-immunized control mice were used for analysis. On the other hand, expression of both CD69 and CD25 on CD8ϩ cells was significantly high when effector cells obtained from tumor-protected mice were used for analysis. A total of 33.3% of CD8ϩ and 2.6% of CD4ϩ cells expressed CD69, whereas 42.7% of CD8ϩ and 2.4% of CD4ϩ cells expressed CD25 (Fig. 11A). The expression of CD28 was negligible in tumor-bearing control mice, whereas 93.0% cells expressed CD28 when effector cells of tumor-protected mice were used for analysis (Fig. 11B). These results, therefore, are consistent with the idea that CD8ϩ T cells are the primary effector arm in CTL response and that CD4ϩ T cells are necessary for the induction of an antitumor CD8ϩ T-cell-dependent response but may not participate as an effector population directly. Immunized Mice that Had Rejected Tumor-generated Anti- tumor Memory Response. Immunized mice that had rejected C15 tumor cells were additionally divided into two subgroups. One group was challenged with C15 cells, whereas the other group was chal- Fig. 9. Tumor immunity induced by 3H1-pulsed DCs is antigen specific. Groups of lenged with nontransfected parent MC-38 cells. The results show that mice were immunized as described in “Materials and Methods.” Two weeks after final these mice resisted lethal doses of C15 cells (Fig. 12A), indicating that immunization, mice were challenged s.c. with CEA-transfected C15 (A) or nontransfected parental MC-38 cells (B). Mice were sacrificed when any tumor size exceeded 250 mm2, the tumor protection was mediated by persistent antitumor immunity and survival was recorded as the percentage of surviving animals of total animals on a specific for antigens relevant to C15 tumor. It is of interest that the given day. All experiments included 6–12 mice/group and were repeated at least three tumor-free mice also had protective immunity against subsequent times. and (c) cytotoxic activity of CTLs in vitro stimulated with 3H1. The proliferation of splenocytes in presence of 3H1 or CEA was significantly high. However, the proliferation in presence of 1A7 in this assay was negligible compared with that of 3H1 or CEA as stimulant (Fig. 10A). The cytokine release pattern by spleen cells was analyzed in vitro. Splenocytes were cultured in presence of 3H1, and the culture supernatant was then analyzed by ELISA. The Th1-associated cytokine response remained almost the same in tumor-protected mice before and after tumor challenge (mIL-2, 880 Ϯ 20 pg/ml; mIFN-␥, 4680 Ϯ 60 pg/ml; mIL-4, 146 Ϯ 12 pg/ml; and mIL-10, 32 Ϯ 4 pg/ml). It has been reported (45) that IL-4 can synergize with IL-12 in enhancing the proliferation of antigen-specific Th1 clones. In our model, IL-4 might have been cooperated with coinduced Th1- associated cytokines in promoting antitumor immunity. The in vivo protection results were reflected in in vitro studies quantifying the specific cytolytic activity of in vivo primed cells present in the spleens of mice immunized three times with 3H1-pulsed DCs. Splenocytes obtained from tumor-protected mice were restimulated in vitro for 5 days with 3H1 and then measured for specific cell lysis against different target cells in a standard 51Cr release assay. In vitro-stimulated cells lysed CEA-expressing C15 target cells significantly, and tumor cell lysis was specific because the use of MC-38 tumor cells as targets resulted in a background lysis (Fig. 10B). The lysis of C15 tumor cells was significantly greater (P Ͻ 0.001) than that obtained with MC-38 targets at all E:T cell ratios tested, indicating that the cellular immune response was specific against CEA and ϩ Fig. 10. Cellular immune response induced in mice protected from tumor challenge. A, CD8 T cells were primarily responsible for antigen-specific lysis splenocytes were harvested after 5 weeks of posttumor challenge, and T-cell proliferation induced by BmDCs as measured in a standard 51Cr release assay (data was measured as described (Fig. 5A), and spleens from at least three mice were pooled to perform the experiment. B, splenocytes were restimulated in vitro (5 days) in the presence not shown). We have found that nonspecific CTLs induced by of 3H1 and rhIL-2. Cytolytic activity was measured in an 18-h 51Cr release assay using BmDCs have no physiological role in vivo, as judged by the failure of different target cells as indicated. Data are representative of two experiments performed. 2851

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. ANTI-IDIOTYPE ANTIBODY-PULSED DC AND TUMOR IMMUNITY challenge with the nontransfected parental tumor line MC-38 (Fig. 12B), suggesting that mice rejecting CEA-transfected carcinoma developed immunity to other antigens expressed on C15 and shared with the nontransfected parent carcinoma cell line MC-38, resulting in long-lasting memory against these tumors. Induction of immunity to shared tumor antigens has also been observed in other tumor model during tumor rejection after particular antigen immunization (47, 48). Of note, in different sets of experiments, 3H1-DC-immunized mice that had rejected C15 tumor cells remained tumor free, and after 8 months of first tumor inoculation, these mice were additionally challenged with either C15 or MC-38 cells and followed for tumor growth and survival. Interestingly, they also resisted the second tumor challenge (data not shown). To additionally evaluate the effects of in vivo priming of T cells, adoptive transfer experiments were performed. 3H1-pulsed DC- immunized mice that had rejected C15 tumors were sacrificed after 5 weeks of posttumor challenge, and splenocytes were restimulated in vitro in presence of 3H1 and rhIL-2 for 5 days. The resulting stimulated cells were composed primarily of CD8ϩ T cells (Ͼ75%), and the remaining cells were mostly CD4ϩ (10%) as determined by flow cytometry (data not shown). These cells were transferred to naı¨ve mice, and the mice were challenged with C15 tumor cells. Naı¨ve mice without transfer of stimulated cells and challenged with C15 cells were used as control. As shown in Fig. 13, mice receiving the in vitro-stimulated immune splenocytes were protected from a subsequent lethal tumor challenge, but no protection was observed in Fig. 12. Vaccination with 3H1-pulsed DCs and challenge with C15 tumor cells induces long-lasting protective immunity to subsequent tumor challenge. Surviving mice that had control mice. The detail mechanisms of this protection is under study. been immunized with 3H1-pulsed DCs and challenged as described (Fig. 8) were rechal- lenged with either CEA-transfected C15 tumor cells (A) or parental MC-38 tumor cells (B) s.c. 13 weeks after first tumor inoculation. A set of age-matched syngeneic mice was also challenged with either C15 or MC-38 cells and used as controls. Survival is recorded as described (Fig. 9). Each group contained 6 mice. Experiments were repeated twice, and a representative experiment is shown.

Fig. 13. Adoptive transfer of immune splenocytes to prevent the growth of a subsequent tumor challenge. 3H1-specific lymphocytes were generated from C15 tumor-free mice as described in Fig. 8. Stimulated cells (3 ϫ 106) were infused i.v. into naı¨ve mice. One day later, the mice were challenged s.c. with 1 ϫ 106 C15 cells and considered as the experimental group. Naı¨ve mice without transfer of stimulated cells and challenged with C15 cells were used as control. The mice were followed for tumor growth as in Fig. 8. The number of tumor-bearing animals compared with the total number of mice in each group is indicated. A representative experiment of two is shown.

DISCUSSION In this study, we have shown that BmDCs pulsed with 3H1 induce a protective immune response against a challenge with tumor cells expressing CEA antigen. The tumor protection is antigen specific Fig. 11. Expression of activation markers by CTLs obtained from tumor-protected because tumor development was not prevented when the immunized mice. Splenocytes harvested from groups of mice after 5 weeks of posttumor challenge were restimulated with 3H1 and rhIL-2 for 5 days. At the end of the culture, cells were mice were challenged with the same tumor cells that were CEA analyzed for the expression of CD4, CD8, CD69, and CD25 by two-color flow cytometry negative. DCs can express a number of Fc receptors, which bind the (A). Expression of CD28 by total lymphocyte populations was also analyzed (B). Middle panel represents CTLs obtained from tumor-bearing control mice, and right panel repre- constant chain of immunoglobulins and mediate the endocytosis of sents CTLs obtained from tumor-free mice. A representative experiment of three is shown. immune complexes (49). Several studies have shown that DCs pulsed 2852

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. ANTI-IDIOTYPE ANTIBODY-PULSED DC AND TUMOR IMMUNITY with soluble proteins can present peptide epitopes derived from these CD4ϩ T cells. Cytokines such as IL-2 and IFN-␥ may allow for exogenous antigens to MHC class I molecules and induce an antigen- dissemination and expansion of tumor-specific T cells elicited by this specific CTL response. It has been reported (18) that DCs use mac- DC vaccine. Therefore, the use of 3H1-pulsed DCs may be advanta- ropinocytosis to capture soluble antigens that are then presented on geous because it would contribute MHC class II-restricted epitopes, as MHC class I molecules. These exogenous antigens access the cytosol well as additional MHC class I restricted epitopes. of DCs and are processed for presentation via the same pathway The immunity induced in mice immunized with 3H1-pulsed DC described for conventional MHC class I-restricted cytosolic antigens. vaccine resulted in complete protection in 100% of experimental mice 3H1 protein is likely to be internalized and degraded to peptides by against live tumor challenge expressing CEA (Figs. 8 and 9A). Our DCs. The degraded peptides bound to the MHC molecules are pre- data correlates with the results performed by Mayordomo et al. (15), sented to T cells by DCs. T cells with appropriate receptors are which indicate that BmDCs grown in medium containing granulocyte expanded and expected to constitute the anti-3H1 cytotoxic, helper, macrophage colony-stimulating factor and IL-4 are capable of gener- and memory cells. T cells induced by immunizations with 3H1-pulsed ating a complete protective antitumor immune response. To addition- DCs will recognize CEA-positive tumor cells only when amino acid ally evaluate the role of CD4ϩ and CD8ϩ T cells in mice, which were sequence of 3H1 has linear homology to CEA. We have identified protected from tumor growth, cellular immune responses were mea- a number of peptides in 3H1 having linear sequence homology to sured after 5 weeks of posttumor challenge. CD4ϩ T-helper cell CEA (35). response was detected by proliferative T-cell response (Fig. 10A) and The role for DCs in the activation of T cells is well established, it cytokine release pattern in tumor-protected mice. Interestingly, the is less clear to what extent DCs regulate other cells. It is known that amount of Th1-biasing cytokines such as IL-2 and IFN-␥ production DC are essential for the development of antibody responses (5–10, remained almost unchanged before and after tumor challenge in mice 50), but this has been thought to be simply a secondary consequence vaccinated with 3H1-pulsed DCs. The increases in the expression of of the requirement for DCs to activate naı¨ve helper T cells. Several the activation markers such as CD69 and CD25 on CD8ϩ T cells but findings suggest that DCs may directly regulate B cell maturation (5). not on CD4ϩ T cells correlated with antigen-specific CTL responses Our results demonstrate that a syngeneic DC that has been pulsed in vitro (Fig. 11). The increase over control in the expression of CD25 in vitro with native protein induces a strong specific B cell response and CD69 indicates that T-cell activation took place in secondary in vivo in mice after immunizations. Ab3 induced in mice by 3H1- lymphoid tissues after vaccination and tumor cell challenge. The pulsed DC vaccine inhibited the Ab1-Ab2 binding in this system, and marked increase in the expression of CD28 on T cells is particularly binding of Ab1 to CEA was also inhibited (Table 2). These results significant because ligation of CD28 with B7.1 and B7.2 initiates suggested that the relevant Ab1Ј had been induced by 3H1 immuni- T-cell responses and the production of armed effector T cells. Impor- zations. The development of CEA-specific IgG2a antibody correlates tantly, the up-regulation of these activation markers correlated com- with ADCC invoked in mice by immunizations with 3H1-pulsed DCs. pletely with the increase in tumor-protective immunity induced by ADCC may be an additional important mechanism for tumor protec- 3H1-based DC vaccine. tion in this model. Our results are in accordance with previous studies In this model, DC vaccinations have also induced antitumor mem- showing that murine IgG2a is effective for prevention of tumor ory responses. The presented results show that anti-Id antibody-pulsed growth (43). DCs induce long-lived CTL memory responses that are capable of The proliferative response of splenocytes in presence of 3H1 and rapidly responding to and protecting against a subsequent tumor CEA reflects that mice immunized with 3H1-pulsed DCs induced challenge (Figs. 12 and 13). anti-3H1, as well as anti-CEA immune responses. The phenotyping of In summary, our results have demonstrated that anti-Id 3H1-pulsed proliferating splenocytes indicated that they are predominantly CD4ϩ DCs can induce both humoral and cellular immune responses in mice T cells. Specific stimulation of the subsets of T cells in the presence with protective CEA-specific antitumor immunity. We have shown of anti-Id antibodies, and 3H1-derived peptides confirmed this obser- convincing evidence that CEA-specific CTL response can be gener- vation (Fig. 5). The cytokine profile in cocultures of mononuclear ated after immunization with 3H1-pulsed DCs in the wild-type cells with 3H1 indicated the induction of a Th1 immune response with C57BL/6 mice in the prophylactic setting. The use of protein-pulsed high levels of IL-2 and IFN-␥ and low levels of IL-4 and IL-10. The DCs may be advantageous because it can activate both MHC class enhanced IFN-␥ production correlates with CEA-specific IgG2a titers II-restricted CD4ϩ T cells, as well as MHC class I-restricted CD8ϩ T observed and is likely to be mediated by IL-12 produced by DCs (51, cells. The implications of these findings for the use of 3H1-pulsed 52). As s.c. vaccination leads to a rapid accumulation of DCs in lymph DCs as vaccine in CEA-positive human cancer patients or in CEA- nodes, we speculate that BmDCs pulsed with antigen migrate to transgenic mice need to be investigated. Currently, we are engaged to lymph nodes and secrete or induce the secretion of IL-12, thereby evaluate the potency of this vaccine in a therapeutic model with driving a CD4ϩ Th1-associated immune response. established tumor, which is transgenic for human CEA. We will also CD8ϩ CTLs are an important effector arm in antitumor immunity. determine the role of CD8ϩ and or CD4ϩ T cells directly in this To investigate the in vivo priming of a CD8-mediated T-cell response model after selective depletion of particular T cells. after prophylactic vaccination with soluble protein-loaded DCs, antigen-specific CTL response was observed in vitro. The effector cells obtained from mice immunized with 3H1-pulsed DCs could lyse the ACKNOWLEDGMENTS CEA-expressing murine carcinoma cells, C15, but not the CEA- We thank Dr. Jeffrey Schlom (NIH) for providing the MC-38 cell line and negative parent tumor cells, MC-38. This is not surprising because a Dr. John Yannelli (University of Kentucky, Lexington, KY) for helpful dis- number of peptides from 3H1 show considerable amino acid sequence cussions. We also thank Mary B. Palascak and Peter Ciraolo for flow cytom- homology to CEA (35). HLA motif analysis of these cross-reacting etry. We thank Audrey Morrison for typing the manuscript. peptides show low but positive scores (53) with MHC class I Kb molecules. The CTL activity was mostly mediated by CD8ϩ T cells and MHC class I restricted as was evidenced by blocking experiments REFERENCES ϩ (Fig. 7). DCs may present class I peptides directly to CD8 T cells, 1. Banchereau, J., and Steinman, R. M. Dendritic cells and the control of immunity. which may also be activated after help provided by the DC-primed Nature (Lond.), 392: 245–252, 1998. 2853

2. Bell, D., Young, J. W., and Banchereau, J. Dendritic cells. Adv. Immunol., 72: 30. Foon, K. A., Chakraborty, M., John, W. J., Sherratt, A., Kohler, H., and Bhattacharya- 255–324, 1999. Chatterjee, M. Immune response to the carcinoembryonic antigen in patients treated 3. Hart, D. N. Dendritic cells: unique leukocyte populations, which control the primary with an anti-idiotype antibody vaccine. J. Clin. Investig., 96: 334–342, 1995. immune response. Blood, 90: 3245–3287, 1997. 31. Foon, K. A., John, W. J., Chakraborty, M., Sherratt, A., Garrison, J., Flett, M., and 4. Steinman, R. M. The dendritic cell system and its role in immunogenicity. Annu. Rev. Bhattacharya-Chatterjee, M. Clinical and immune responses in advanced colorectal Immunol., 9: 271–296, 1991. cancer patients treated with anti-idiotype monoclonal antibody vaccine that mimics 5. Dubois, B., Vanbervliet, B., Fayette, J., Massacrier, C., Van Kooten, C., Briere, F., the carcinoembryonic antigen. Clin. Cancer Res., 3: 1267–1276, 1997. Banchereau, J., and Caux, C. Dendritic cells enhance growth and differentiation of 32. Foon, K. A., John, W. J., Chakraborty, M., Das, R., Teitelbaum, A., Garrison, J., CD40-activated B lymphocytes. J. Exp. Med., 185: 941–951, 1997. Kashala, O., Chatterjee, S. K., and Bhattacharya-Chatterjee, M. Clinical and immune 6. Inaba, K., Witmer, M. D., and Steinman, R. M. Clustering of dendritic cells, helper responses in resected colon cancer patients treated with anti-idiotype monoclonal T lymphocytes, and histocompatible B cells, during primary antibody responses in antibody vaccine that mimics the carcinoembryonic antigen. J. Clin. Oncol., 17: vitro. J. Exp. Med., 160: 858–876, 1984. 2889–2895, 1999. 7. Inaba, K., Granelli-Piperno, A., and Steinman, R. M. Dendritic cells are critical 33. Inaba, K., Inaba, M., Romani, N., Aya, H., Deguchi, M., Ikehara, S., Muramatsu, S., accessory cells for thymus-dependent antibody responses in mouse and man. Proc. and Steinman, R. M. Generation of large number of dendritic cells from mouse bone Natl. Acad. Sci. USA, 80: 6041–6045, 1983. marrow cultures supplemented with granulocyte/macrophage colony-stimulating fac- 8. Inaba, K., Metlay, J. P., Crowley, M. T., and Steinman, R. M. Dendritic cells pulsed with protein antigen in vitro can prime antigen-specific, MHC-restricted T cells in tor. J. Exp. Med., 176: 1693–1702, 1992. situ. J. Exp. Med., 172: 631–640, 1990. 34. Sen, G., Chakraborty, M., Foon, K. A., Reisfeld, R. A., and Bhattacharya-Chatterjee, 9. Sornasse, T., Flamand, V., deBecker, G., Bazin, H., Tielemans, F., Thielemans, K., M. Preclinical evaluation in nonhuman primates of an anti-idiotype antibody that Urbain, J., Oberdan, L., and Moser, M. Antigen-pulsed dendritic cells can efficiently mimics the disialoganglioside GD2. Clin. Cancer Res., 3: 1969–1976, 1997. induce an antibody response in vivo. J. Exp. Med., 175: 15–21, 1992. 35. Chatterjee, S. K., Tripathi, P. K., Chakraborty, M., Yannelli, J., Wang, H., Foon, 10. Flamand, V., Sornasse, T., Thielemans, K., Demanet, C., Bakkus, M., Bazin, H., K. A., Maier, C. C., Blalock, J. E., and Bhattacharya-Chatterjee, M. Molecular Tielemans, F., Leo, O., Urbain, J., and Moser, M. Murine dendritic cells pulsed in mimicry of carcinoembryonic antigen by peptides derived from the structure of an vitro with tumor antigen induce tumor resistance in vivo. Eur. J. Immunol., 24: anti-idiotype antibody. Cancer Res., 58: 1217–1224, 1998. 605–610, 1994. 36. Pervin, S., Chakraborty, M., Bhattacharya-Chatterjee, M., Zeytin, H., Foon, K. A., 11. Cohen, P. J., Cohen, P. A., Rosenberg, S. A., Katz, S. I., and Mule, J. J. Murine and Chatterjee, S. K. Induction of antitumor immunity by an anti-idiotype antibody epidermal Langerhans cells and splenic dendritic cells present tumor-associated mimicking carcinoembryonic antigen. Cancer Res., 57: 728–734, 1997. antigens to primed T cells. Eur. J. Immunol., 24: 315–319, 1994. 37. Clarke, P., Mann, J., Simpson, J. F., Richard-Dickson, K., and Primus, F. J. Mice ϩ 12. Cohen, P. A., Cohen, P. J., Rosenberg, S. A., and Mule, J. J. CD4 T cells from mice transgenic for human carcinoembryonic antigen as a model for immunotherapy. immunized to syngeneic sarcomas recognize distinct, non-shared tumor antigen. Cancer Res., 58: 1469–1477, 1998. Cancer Res., 54: 1055–1058, 1994. 38. Weaver, C. T., Hawrylowicz, C. M., and Unanue, E. R. T helper cell subsets require 13. Fields, R. C., Shimizu, K., and Mule, J. J. Murine dendritic cells pulsed with whole the expression of distinct costimulatory signals by antigen-presenting cells. Proc. tumor lysates mediate potent antitumor immune response in vitro and in vivo. Proc. Natl. Acad. Sci. USA, 85: 8181–8185, 1988. Natl. Acad. Sci. USA, 95: 9482–9487, 1998. 39. Mosmann, T. R., and Coffman, R. L. TH1 and TH2 cells: different patterns of 14. Shimizu, K., Fields, R. C., Giedlin, M., and Mule, J. J. Systemic administration of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol., 7: interleukin 2 enhances the therapeutic efficacy of dendritic cell-based tumor vaccine. 145–173, 1989. Proc. Natl. Acad. Sci. USA, 96: 2268–2273, 1999. 40. Stevens, T. L., Bossie, A., Sanders, V. M., Fernandez-Botran, R., Coffman, R. L., 15. Mayordomo, J. I., Zorina, T., Storkus, W. J., Zitvogel, L., Celluzzi, C., Falo, L. D., Mosmann, T. R., and Vitetta, E. S. Regulation of antibody isotype secretion by Jr., Melief, C. J., Ildstad, S. T., Kast, W. M., Deleo, A. B., and Lotze, M. T. Bone subsets of antigen-specific helper T cells. Nature (Lond.), 334: 255–258, 1988. marrow-derived dendritic cells pulsed with synthetic tumor peptides elicit protective 41. Imai, K., Pellegrino, M. A., Wilson, B. S., and Ferrone, S. Higher cytolytic efficiency and therapeutic antitumor immunity. Nat. Med., 1: 1297–1302, 1995. of an IgG2 ␣ than of an IgG1 monoclonal antibody reacting with the same (or 16. Porgador, A., Snyder, D., and Gilboa, E. Induction of antitumor immunity using bone marrow-generated dendritic cells. J. Immunol., 156: 2918–2926, 1996. spatially close) determinant on a human high molecular-weight melanoma-associated 17. Paglia, P., Chiodoni, C., Rodolfo, M., and Colombo, M. P. Murine dendritic cells antigen. Cell. Immunol., 72: 239–247, 1982. loaded in vitro with soluble protein prime cytotoxic T lymphocytes against tumor 42. Kipps, T. J., Parham, P., Punt, J., and Herzenberg, L. A. Importance of immunoglob- antigen in vivo. J. Exp. Med., 183: 317–322, 1996. ulin isotype in human antibody-dependent, cell-mediated cytotoxicity directed by 18. Brossart, P., and Beven, M. J. Presentation of exogenous protein antigens on major murine monoclonal antibodies. J. Exp. Med., 161: 1–17, 1985. histocompatability complex class I molecules by dendritic cells: pathway of presen- 43. Kaminski, M. S., Kitamura, K., Maloney, D. G., Campbell, M. J., and Levy, R. Impor- tation and regulation by cytokines. Blood, 90: 1594–1599, 1997. tance of antibody isotype in monoclonal anti-idiotype therapy of a murine B cell lym- 19. Boczkowski, D., Nair, S. K., Snyder, D., and Gilboa, E. Dendritic cells pulsed with phoma. A study of hybridoma class switch variants. J. Immunol., 136: 1123–1130, 1986. RNA are potent antigen-presenting cells in vitro and in vivo. J. Exp. Med., 184: 44. Porgador, A., and Gilboa, E. Bone marrow-generated dendritic cells pulsed with a 465–472, 1996. class I-restricted peptide are potent inducers of cytotoxic T lymphocytes. J. Exp. 20. Condon, C., Watkins, S. C., Celluzzi, C. M., Thompson, K., and Falo, L. D., Jr. Med., 182: 255–260, 1995. DNA-based immunization by in vivo transfection of dendritic cells. Nat. Med., 2: 45. Kennedy, M. K., Picha, K. S., Shanebeck, K. D., Anderson, D. M., and Grabstein, 1122–1128, 1996. K. H. IL-12 regulates the proliferation of Th1, but not Th2 or Th0 clones. Eur. 21. Manickan, E., Kanangat, S., Rouse, R. J., Yu, Z., and Rouse, B. T. Enhancement of J. Immunol., 24: 2271–2276, 1994. immune response to naked DNA vaccine by immunization with transfected dendritic 46. Testi, R., D’Ambrosio, D., De Maria, R., and Santoni, A. The CD69 receptor: a cells. J. Leukoc. Biol., 61: 125–132, 1997. multi-purpose cell-surface trigger for hematopoietic cells. Immunol. Today, 15: 22. Gong, J., Chen, D., Kashiwaba, M., and Kufe, D. Induction of antitumor activity by 479–483, 1994. immunization with fusions of dendritic and carcinoma cells. Nat. Med., 3: 558–561, 47. Falo, L. D., Jr., Kovacsovics-Bankowski, M., Thompson, K., and Rock, K. L. 1997. Targeting antigen into the phagocytic pathway in vivo induces protective tumor 23. Strome, S., Voss, S., Wilcox, R., Wakefield, T., Tamada, K., Flies, D., Chapoval, A., immunity. Nat. Med., 1: 649–653, 1995. Lu, J., Kasperbauer, J., Padley, D., Vile, R., Gastineau, D., Wettstein, P., and Chen, 48. Celluzzi, C. M., Mayordoma, J. I., Storkus, W. J., Lotze, M. T., and Falo, L. D., Jr. L. Strategies for antigen loading of dendritic cells to enhance the antitumor immune Peptide-pulsed dendritic cells induce antigen-specific, CTL-mediated protective tu- response. Cancer Res., 62: 1884–1889, 2002. mor immunity. J. Exp. Med., 183: 283–287, 1996. 24. Jerne, N. K. Towards a network theory of the immune system. Ann. Immunol. (Inst. 49. Cella, M., Sallusto, F., and Lanzavecchia, A. Origin, maturation and antigen present- Pasteur), 125C: 373–389, 1974. ing function of dendritic cells. Curr. Opin. Immunol., 9: 10–16, 1997. 25. Bhattacharya-Chatterjee, M., Mukherjee, S., Biddle, W., Foon, K. A., and Kohler, H. 50. Pulendran, B., Smith, J. L., Jenkins, M., Schoenborn, M., Maraskovsky, E., and Murine monoclonal anti-idiotype antibody as a potential network antigen for human carcinoembryonic antigen. J. Immunol., 145: 2758–2765, 1990. Maliszewski, C. R. Prevention of peripheral tolerance by a dendritic cell growth 26. Steward, A. M., Nixon, D., Zamcheck, N., and Aisenberg, A. Carcinoembryonic factor: flt3 ligand as an adjuvant. J. Exp. Med., 188: 2075–2082, 1998. antigen in breast cancer patients: serum levels and disease progress. Cancer (Phila.), 51. Koch, F., Stanzl, U., Jennewein, P., Janke, K., Heufler, C., Kampgen, E., Romani, N., 33: 1246–1252, 1974. and Shuler, G. High level IL-12 production by murine dendritic cells: up-regulation 27. Vincent, R. G., and Chu, T. M. Carcinoembryonic antigen in patients with carcinoma via MHC class II and CD40 molecules and down-regulations by IL-4 and IL-10. of the lung. J. Thorac. Cardiovasc. Surg., 66: 320–328, 1978. J. Exp. Med., 184: 741–746, 1996. 28. Muraro, R., Wunderlich, D., and Thor, A. Definition by monoclonal antibodies of 52. Pulendran, B., Lingappa, J., Kennedy, M. K., Smith, J., Teepe, M., Rudensky, A., repertoire of epitopes on carcinoembryonic antigen differentially expressed in human Maliszewski, C. R., and Maraskovsky, E. Developmental pathways of dendritic cells colon carcinomas versus normal adult tissues. Cancer Res., 45: 5769–5780, 1985. in vivo. J. Immunol., 159: 2222–2231, 1997. 29. Chakraborty, M., Foon, K. A., Kohler, H., and Bhattacharya-Chatterjee, M. Preclin- 53. Parker, K. C., Bednarek, M. A., and Coligan, J. E. Scheme for ranking potential ical evaluation in non-human primates of an anti-idiotypic antibody that mimics HLA-A2 binding peptides based on independent binding of individual peptide side- carcinoembryonic antigen. J. Immunother., 18: 95–103, 1995. chains. J. Immunol., 152: 163–175, 1994.

2854

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2003 American Association for Cancer Research. Murine Dendritic Cells Pulsed with an Anti-Idiotype Antibody Induce Antigen-specific Protective Antitumor Immunity

Asim Saha, Sunil K. Chatterjee, Kenneth A. Foon, et al.

Cancer Res 2003;63:2844-2854.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/63/11/2844

Cited articles This article cites 52 articles, 32 of which you can access for free at: http://cancerres.aacrjournals.org/content/63/11/2844.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/63/11/2844.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/63/11/2844. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.