Cytotoxic T lymphocytes (CTLs) play a major role in the rejection of immunogenic tumors.1 Classically, CTLs target tumors through recognition of a ligand consisting of the self major histocompatibility complex (MHC) class I molecule and peptides that are generally derived from tumor antigens synthesized within the tumor cells.2, 3 However, there is evidence for an exogenous pathway whereby antigens that are not expected to gain access to the cytoplasm are presented on MHC class I molecules.4, 5, 6 A most striking example of this is the in vivo phenomenon of cross-priming: antigens from donor cells are acquired by host antigen-presenting cells (APCs) and presented on MHC class I molecules in the appropriate context of co-stimulation. Delivery of exogenous antigen to the endogenous MHC class I-restricted processing pathway of APCs is a critical challenge in cancer vaccine design.
Dendritic cells (DCs) are one of the most potent APCs. They capture antigens in situ, and migrate to lymphoid organs to interact/activate naive T cells.7 DCs pulsed with synthetic tumor-derived MHC class I-restricted peptides or tumor lysates and tumor cell-derived RNA induced significant CTL-dependent antitumor immune responses not only in vitro but also after adoptive transfer in mice.8–13 Recently, Eggert et al.14 showed that DCs pulsed with B16 melanoma-specific mTRP2 peptide induced significant antitumor immunity. However, each of these methods has drawbacks.15 Foremost, the use of MHC class I binding peptides is associated with MHC restriction and the induced immune responses tend to be limited only to CD8+ T cells. Furthermore, the process of identifying MHC class I binding tumor peptides is labor intensive and time consuming, and only a small number of human tumor peptides have been identified.16, 17 In addition, the therapeutic efficiency of these DC vaccine strategies was limited in that they could only prevent the rechallenge of parental tumor cells with a low challenging dose and they inhibited the growth of established tumors but only in the early stages. The induction of stronger CTL responses has become a major goal of current cancer vaccine strategies.
Apoptotic cells are characterized by cell shrinkage, the collapse of the nucleus and cytoplasmic blebbing.18 It is now well known that apoptotic cells provide antigens that can effectively trigger recognition by the immune system.19, 20 Recently, it has been shown that (i) apoptosis increased the immunogenicity of a rat tumor cell line;21(ii) MHC class I and class II molecules of DCs better presented proteins from phagocytosed cellular fragments than pre-processed peptides;22, 23 and (iii) DCs that had acquired antigen from apoptotic bodies induced MHC class I-restricted CTLs and antitumor immunity.24–28 These results demonstrated that apoptotic tumor cells may be a good source of tumor antigens for presentation to DCs. However, to date, the phenotypic characteristics of DCs that have phagocytosed apoptotic tumor cells, the immune responses induced by these DCs and their vaccine efficiency against tumor in animal models have not been well studied.
To define an efficient DC vaccine strategy, this study investigated the efficiency of antitumor immunity derived from DCs that had phagocytosed apoptotic/necrotic tumor cells compared with that of DCs pulsed with the mTRP2 tumor peptide.
MATERIAL AND METHODS
Cell line, antibodies, chemokines, peptides and animals
BL6-10 is a poorly immunogenic and highly lung metastatic variant derived from the murine B16 melanoma cell line.29 EL4 is a murine T-cell lymphoma cell line of C57BL/6 mouse origin. These 2 cell lines were maintained in the complete medium that is MEM-alpha (GIBCO, Gaithersburg, MD) medium plus 10% fetal calf serum (FCS). BL6-10 tumor cells were detached from culture flasks using 0.5 mM EDTA in Ca++ and Mg++ free phosphate-buffered saline (PBS) containing 0.1% glucose (cPEG)29 and 0.03% trypsin/EDTA. Monoclonal antibodies including rat anti-mouse H-2Kb, Iab, CD3, CD4, CD8, CD11b, CD11c, CD40, CD80, CD86, intercellular adhesion molecule-1 (ICAM-1) and B220 antibodies were all purchased from PharMingen (San Diego, CA). The recombinant mouse interleukin (IL)-4 and granulocyte-macrophage colony-stimulating factor (GM-CSF) were purchased from Endogene (Woburn, MA). The recombinant chemokine MIP-3β was obtained from R&D System (Minneapolis, MN). The B16 melanoma-specific mTRP2 (VYDFFVWL) peptide14 was synthesized by Multiple Peptide Systems (San Diego, CA). Female C57BL/6 (H-2Kb) and BALB/c (H-2Kd) mice were obtained from Charles River and housed in the animal facility of Saskatoon Cancer Center.
Generation of apoptotic/necrotic cells
Apoptotic/necrotic tumor cells were prepared by using lovastatin, a competitive inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase that blocks cell cycling in the G1 phase and induces tumor cell apoptosis.30 Briefly, when BL6-10 tumor cells were grown to 60%–70% of confluence in flasks, media were aspirated and replenished with the complete medium containing 20 μM lovastatin (Merck, Rahway, NJ). The tumor cells were cultured for 1–2 days and harvested for apoptosis/necrosis analysis or kept frozen at −20°C until use for co-culturing with DCs.
Apoptosis and necrosis analysis
For flow cytometric analysis, the apoptotic/necrotic tumor cells were collected by centrifugation and washed in ice-cold PBS with 0.3 mM EDTA. The tumor cells were then fixed by gradual addition of ice-cold 100% ethanol to a final concentration of 80% while vortexing. After 1–3 days at 4°C, the fixed cells were pelleted and washed once with PBS, then resuspended in 1 ml of PBS containing 10 μg/ml RNase A (GIBCO, Gaithersburg, MD) and 5 μg/ml propidium iodide (PI; Sigma, St. Louis, MO). After incubation at 37°C for 30 min, the samples were analyzed by flow cytometry. Data were analyzed by the “overlapped peak” multicycle fitting option.30
For further confirmation of apoptosis and necrosis, the apoptotic/necrotic tumor cells were analyzed by using an Annexin V-FITC Apoptosis Detection kit (PharMingen) and APO-BRDU kit (Phoenix Flow Systems, San Diego, CA) according to the protocol of the manufacturers' manuals. When using the Annexin V-FITC Apoptosis Detection kit, apoptotic/necrotic tumor cells were incubated with Annexin V-FITC in a binding buffer containing PI and analyzed by flow cytometry. With APO-BRDU kit, apoptotic/necrotic tumor cells were fixed in 1% paraformaldehyde and stored in 70% ethanol at −20°C until use. For analysis, apoptotic tumor cells were washed with wash buffer and incubated in DNA labeling buffer at 37°C for 60 min to label the 3′-hydroxyl ends of the DNA fragments with bromolated deoxyuridine triphosphate nucleotides (Br-dUTP). The apoptotic/necrotic tumor cells were washed with rinse buffer and incubated with fluorescein-PRB-1 antibody solution at room temperature for 30 min and analyzed by flow cytometry.
Preparation of DCs pulsed with peptides and DCs that had phagocytosed apoptotic/necrotic cells
A previously described procedure was used for generation of DCs from bone marrow (BM) culture with some modification.31 Briefly, BM cells prepared from femurs and tibias of normal C57BL/6 mice were depleted of red blood cells with 0.84% ammonium chloride and plated in Dulbecco's modified Eagle's medium (DMEM) plus 10% FCS, GM-CSF (10 ng/ml) and IL-4 (10 ng/ml) on day 1. On day 3, non-adherent granulocytes and B and T lymphocytes were gently removed, and fresh media were added. On day 5, loosely adherent proliferating DC aggregates were dislodged and replated. On day 7 of culture, released, mature, non-adherent cells with the typical morphologic features of DCs were harvested and used for in vitro phenotypic analysis and for peptide pulsing or co-cultivation with apoptotic tumor cells.
For peptide pulsing, 1–2 × 106 DCs were resuspended in 1 ml of DMEM medium containing 50 mM 2-ME and 20 μM mTRP2 peptide. After a 3-hr incubation at 37°C with gentle shaking every 30 min, the peptide-pulsed DCs were washed twice with PBS and resuspended in PBS for in vitro characterization and in vivo vaccination of mice. For cultivation with apoptotic/necrotic tumor cells, DCs were incubated with apoptotic/necrotic tumor cells at a ratio of 3:1 in the culture medium containing GM-CSF (10 ng/ml) and IL-4 (10 ng/ml). To avoid adhesion of DCs to the plastic culture flasks, the plates were pre-coated with 10 mg/ml poly-2-hydroxy-ethymethacrylate (Sigma). After an 18-hr incubation, DCs were harvested, purified with Ficoll-Paque gradient (Pharmacia Biotech, Uppsala, Sweden) and washed twice with PBS. The DCs that had phagocytosed apoptotic/necrotic tumor cells were used for in vitro characterization and in vivo vaccination of mice. In addition, the DCs were also fixed with glutaraldehyde for electron microscopy.
Analysis of phagocytosis
The ability of DC phagocytosis was assessed using fluorescein isothiocyanate (FITC)-conjugated dextran (Molecular Probes, Eugene, OR). Briefly, 20 μl of dextran (0.05 mM) was incubated with 1 × 106 DCs in DMEM at 37°C. After 2 hr of incubation, cells were harvested and resuspended in medium containing Trypan blue, which quenches the fluorescence of extracellular particles. DCs were washed, resuspended in PBS and analyzed by flow cytometry.
The method for labeling apoptotic/necrotic tumor cells with TRITC (rhodamine, Sigma) was similar to that described previously.32 Briefly, apoptotic/necrotic tumor cells were resuspended in DMEM at 1 × 106 cells/ml and incubated with TRITC (0.5 μg/ml) at 37°C for 45 min. The labeled cells were washed 3 times with PBS. To evaluate DC phagocytosis of apoptotic/necrotic tumor cells, DCs were cultivated with apoptotic/necrotic tumor cells at a ratio of 3:1 in the culture medium containing GM-CSF (10 ng/ml) and IL-4 (10 ng/ml) at 37°C overnight. DCs were harvested, purified with Ficoll-Paque gradient and then analyzed by flow cytometry.
For phenotypic analysis, DCs from the BM culture and DCs that were pulsed with mTRP2 peptide or had phagocytosed apoptotic/necrotic tumor cells were stained with a panel of antibodies and then quantified by flow cytometry. The antibodies used were rat anti-mouse H-2Kb, Iab, CD3, B220, CD11b, CD11c, CD40, CD80, CD86 and ICAM-1 antibodies. Briefly, DCs were incubated with each of the above antibodies (5 μg/ml) on ice for 30 min. After 3 washes with PBS, cells were incubated with FITC-conjugated goat anti-rat IgG antibody (1:60) on ice for another 30 min. After 3 washes with PBS, cells were then analyzed by flow cytometry. Isotype-matched monoclonal antibodies were used as controls.
RNase protection assay
To examine the phenotypic changes of DCs on expression of cytokines, chemokines and chemokine receptors, DCs that were pulsed with mTRP2 peptide or had phagocytosed apoptotic/necrotic BL6-10 tumor cells were subjected to an RNase protection assay by using a RiboQuant Multi-Probe RNase Protection Assay System kit (PharMingen) according to the manufacturer's protocol. Briefly, RNA was extracted from DCs using the RNA Isolation kit (PharMingen). In vitro transcription of the Pharmingen Multi-Probe Template Sets (mCK2b, mCK5 and mCR5 with addition of CCR7 probe) with [α-32P]UTP (Amersham Canada Ltd, Oakville, Ontario, Canada) was carried out using T7 RNA polymerase followed by phenol-chloroform extraction and ethanol precipitation. The concentration of the probes was adjusted to 3 × 105 counts per minute (cpm)/μl. The hybridization of the sample RNA (5 μg) to 32P-labeled anti-sense RNA probe (6 × 105 cpm) in vitro transcribed from the mouse cytokine, chemokine and chemokine receptor template sets was carried out. Samples were digested with RNase followed by proteinase K treatment and phenol-chloroform extraction. After ethanol precipitation with 4 M ammonium acetate, protected samples were resuspended in 1× loading buffer and separated on a 5.7% acrylamide-bisacrylamide urea gel. The gel was absorbed onto filter paper, dried under vacuum, and exposed to Kodak X-AR film with intensifying screens at −80°C. The relative expression of cytokine, chemokine and chemokine receptor encoding mRNA was measured by scanning densitometry (Molecular Dynamics, Sunnyvale, CA) on subexposed autoradiograms, and further normalized using housekeeping gene value (GAPDH).
Mixed lymphocyte reaction (MLR)
Spleens were removed from BALB/c mice for preparation of the splenic lymphocyte suspension. Red cells were lysed by using 0.84% ammonium chloride. T cells were obtained from the splenic lymphocytes by nylon wool non-adherence to deplete residual APCs.33 The primary MLRs were performed as previously described.34 Briefly, irradiated DCs (3,000 rad, 1 × 104 cells/well) were incubated in graded doses with a constant amount (2 × 105) of allogeneic T cells of the BALB/c mouse in each well of 96-well culture plates, respectively. After 5 days, T-cell proliferation was measured by adding 1 μCi/well of [3H]-thymidine (1 mCi/ml, Amersham Canada Ltd) to cultures and subsequent liquid scintillation counting after an overnight incubation period.
In vitro chemotaxis assay
An in vitro chemotaxis assay was performed using a multimicro-well Boyden chamber (Neuroprobe, Gaithersburg, MD) and polyvinylpyrolidone-free polycarbonate membranes with 6.5-μm pores.35 Briefly, 25 μl of recombinant chemokine MIP-3β with concentrations ranging from 1 to 1,000 ng/ml were placed in triplicate to the lower wells of Boyden chambers. DCs (1 × 105 cells in 50 μl of DMEM plus 1% bovine serum albumin) were added to each of the top wells. After the plates were incubated at 37°C for 4 hr, the filters were removed, fixed in 70% methanol and stained using the Diff-Quik technique. DCs that had migrated through onto the lower surface of the filter were counted under the microscope.
In vivo migration
DCs that were pulsed with mTRP2 peptide and had phagocytosed apoptotic/necrotic tumor cells were radiolabeled with [51Cr]-chromate respectively (Amersham Canada Ltd). Radiolabeled DCs were prepared by culturing 10 × 106 DCs in 0.5 ml DMEM medium for 1 hr in the presence of 50 μl of sodium [51Cr]-chromate (36 mCi/ml) and washed 3 times with DMEM medium. Thereafter, 1 × 106 labeled cells were injected subcutaneously (s.c.) in 30 μl of PBS into hind footpads of mice. One day after injection, the mice were sacrificed. Mouse feet were amputated at a level 3 mm above the hairline, and the regional draining lymph nodes (RDLNs) were removed for measurement of radioactivity in a gamma counter. The relative migration index of DC was calculated as 100× cpm in RDLNs/cpm that remained in the footpad.
Quantitation of cytokine secretion
Vaccinations were performed by injection of 0.5 × 106 DCs that either were pulsed with mTRP2 peptide or had phagocytosed apoptotic/necrotic tumor cells into the mouse footpads. One week after vaccination, RDLNs were removed for harvesting lymphocytes. These lymphocytes were co-cultured with irradiated BL6-10 cells (20,000 rad) at 2:1 (i.e., 0.5 × 106 lymphocytes and 0.25 × 106 irradiated BL6-10 cells) per well in a 96-well plate. Co-culture was done in quadruplicate and their supernatants were harvested and pooled at 24 and 72 hr for interferon (IFN)-γ and IL-4 quantitation, respectively. Quantitation of secreted cytokines was done in an enzyme-linked immunosorbent assay (ELISA) using the respective cytokine kits for IFN-γ and IL-4 (Endogene). The results were normalized to the known standard curves.
One week after mice were vaccinated twice with DCs that were pulsed with mTRP2 peptide or had phagocytosed apoptotic/necrotic tumor cells, spleens were removed from these immunized mice for preparation of single-cell suspensions by pressing against fine nylon mesh. Red cells were lysed by using 0.84% ammonium chloride. Spleen lymphocytes were co-cultured with irradiated BL6-10 cells (20,000 rad) at 25:1 (i.e., 5 × 106 lymphocytes and 2 × 105 irradiated BL6-10 cells) in 2 ml of DMEM plus 10% FCS in each well of a 24-well plate, respectively. Five days later, T cells were harvested and analyzed using rat anti-mouse CD4 and CD8 antibodies by flow cytometry as described above. These T cells were also used as effector cells in a chromium-release assay. Target cells included BL6-10 cells and irrelevant EL4 cells. These target cells were radiolabeled with [51Cr]-chromate (Amersham Canada Ltd). Radiolabeled target cells were prepared by culturing target cells for 1 hr in the presence of 50 μl of sodium [51Cr]-chromate (36 mCi/ml) and washed twice with DMEM. Ten thousand labeled target cells per well were mixed with effector cells at various effector/target cell ratios in triplicate and incubated for 8 hr. Percentage of specific lysis was calculated as 100 × [(experimental cpm − spontaneous cpm)/(maximal cpm − spontaneous cpm)]. Spontaneous counts per minute released in the absence of effector cells were less than 10% of specific lysis. The maximal counts per minute was released by adding 1% Triton X-100 to wells in experiments.
For evaluation of tumor prevention, mice were vaccinated s.c. with 0.5 × 106 DCs that had phagocytosed apoptotic/necrotic tumor cells. For controls, mice were vaccinated with DCs alone and DCs pulsed with mTRP2 peptide, respectively. The vaccination was repeated once in 7 days. One week after the second vaccination, mice were injected intravenously (i.v.) with 0.3 × 106 BL6-10 tumor cells. Three weeks after tumor injection, mice were sacrificed and the lungs were removed. The extent of lung metastases was determined macroscopically. The lungs were fixed in 10% buffered formalin, embedded in paraffin and then examined histologically.
BL6-10 apoptosis/necrosis induced by lovastatin treatment
Lovastatin treatment induced significant morphologic changes. Two days after treatment of BL6-10 cells with lovastatin (20 μM), most of the cells became rounded and detached from the culture flasks (data not shown). To determine whether lovastatin is able to induce apoptosis, we analyzed cellular DNA fragmentation using flow cytometry. A significant amount of a subdiploid population, representing the apoptotic cells, appeared 1 day after and became dominant 2 days after treatment with lovastatin (Fig. 1A). The characteristics of apoptotic cells was further confirmed by using Annexin V-FITC Apoptosis Detection and APO-BRDU kits. As shown in Figure 1B, 1 and 2 days after treatment with lovastatin, 38% and 82% of tumor cells displayed positive Annexin V-FITC and PI staining, respectively, indicating that these tumor cells underwent apoptosis and necrosis,36 and thereby termed apoptotic/necrotic tumor cells. These apoptotic/necrotic tumor cells derived from 2 days of culture with lovastatin also displayed significant BrdU incorporation, thus confirming the formation of apoptotic DNA fragmentation (Fig. 1A). In addition, typical morphologic changes of these apoptotic cells such as cell shrinkage, nuclear collapse and cytoplasmic blebbing were shown by electron microscopy (Fig. 2a). These apoptotic/necrotic tumor cells were used for co-culture with DCs.
Enhanced maturation of DCs that had phagocytosed apoptotic tumor cells
DCs used in this study were derived from mouse BM cells cultivated in the complete medium plus IL-4 and GM-CSF. These immature dendritic cells showed (i) a typical morphologic characteristic of dendritic cells with numerous dendrites, (ii) functional phagocytosis of FITC-conjugated latex beads37 (Fig. 3a) and (iii) significant expression of MHC class I (H-2Kb) and II (Iab) antigens, co-stimulatory molecules (CD80 and CD86) and adhesion molecules (ICAM-1, CD11b, CD11c and CD40),38 but no expression of CD3 (T-cell marker) and B220 (B-cell marker) (data not shown).
After the physical contact with TRITC-labeled apoptotic/necrotic tumor cells in tissue culture, approximately 24% of DCs displayed the phagocytosed TRITC-apoptotic/necrotic tumor cells in their cytoplasm by flow cytometric analysis (Fig. 3c) and this finding was further confirmed by electron microscopic analysis (Fig. 2b). In RNase protection assays as shown in Figure 4, DCs that had phagocytosed apoptotic/necrotic tumor cells showed enhanced mRNA expression of (a) proinflammatory cytokines such as IL-1β (32.3-fold), IL-6 (10.6-fold), IFN-γ (9.8-fold), TNF-α (2.6-fold) and GM-CSF (5.5-fold); (b) chemokines such as MIP-1α (4.7-fold), MIP-1β (2.5-fold) and MIP-2 (12.5-fold); and (c) the CC chemokine receptor CCR7 (2.8-fold). These DCs also showed decreased mRNA expression of the CC chemokine receptors CCR2 (3.1-fold) and CCR5 (2.7-fold). In addition, these DCs displayed a similar amount of mRNA expression of RANTES (regulated upon activation, normal T cell expressed and secreted) compared with DCs pulsed with the tumor peptide (data not shown).
It has been reported that the proinflammatory cytokines are able to mature DCs.39–41 To verify whether the maturation of DCs is accompanied by proinflammatory cytokines, these DCs were further subjected to flow cytometric analysis by using a panel of antibodies against mouse MHC class I and II antigens, CD11b, CD11c, CD40, CD80, CD86 and ICAM-1. As shown in Figure 5, DCs that had phagocytosed apoptotic/necrotic BL6-10 cells displayed an up-regulated expression of Iab, CD11b, CD40 and CD86 by flow cytometric analysis in comparison with DCs pulsed with mTRP2 peptide. These results indicate that DC maturation was induced by phagocytosis of apoptotic/necrotic tumor cells. The amount of expression of MHC class I antigen, CD11c, CD80 and ICAM-1 on these DCs remained unchanged in comparison with DCs pulsed with the tumor peptide (data not shown).
Increased capability to migrate in vitro and in vivo
To address whether CCR7 expression is critical for DC migration, we performed DC migration assays in vitro and in vivo. An in vitro chemotaxis assay showed that the migration of DCs toward the CC chemokine MIP-3β was dose dependent (data not shown). DCs that had phagocytosed apoptotic/necrotic tumor cells had significantly greater migratory capability than DCs pulsed with the tumor peptide in response to MIP-3β (250 ng/ml) (Table I). Radioisotopes have been used to label T cells and DCs in their in vivo biodistribution studies.42–44 To examine the in vivo migratory capability, the DCs were labeled with [51Cr]-chromate, then injected s.c. into the hind footpads of mice. After 1 day, mouse feet and the RDLNs were removed. Their radioactivity was counted in a gamma counter. As shown in Table I, the relative migration index of DCs that had phagocytosed apoptotic/necrotic tumor cells was 7.23% compared with that of 2.56% for DCs pulsed with mTRP2 peptide (p < 0.01).
Table I. Enhanced Capability of DCs That Had Phagocytosed Apoptotic/Necrotic Tumor Cells to Migrate In Vitro and In Vivo
DC/mTRP2 and DC/apopt represent DCs pulsed with mTRP2 peptide and DCs that had phagocytosed apoptotic/necrotic BL6-10 cells, respectively.
In an in vitro chemotaxis assay, 1 × 105 DCs in 50 μl DMEM and 25 μl of MIP-3β (250 ng/ml) were added to each top and lower well of a Boyden chamber in triplicate, respectively. After incubation at 37°C for 4 hr, the filters were removed, fixed, and stained using the Diff-Quik technique. DCs that had migrated through onto the lower surface of the filter were counted under the microscope in a blinded fashion.
In an in vivo migration assay, 1 × 106 radiolabeled DCs in 30 μl of phosphate-buffered saline were injected into hind footpads of mice. One day after injection, the mice were killed. The feet and regional lymph nodes were removed from the mice (8 per group) and counted in a gamma counter. The relative migration index of DC was calculated as: 100 × CPM of regional draining lymph nodes (RDLNs)/CPM remained in the footpad.
The mean number of migrating DC/apopt to MIP-3β in vitro is significantly different (p < 0.05) from that of migrating DC/mTRP2 (Student t-test).
The mean migration index of DC/apopt to RDLNs is significantly different (p < 0.01) from that of DC/mTRP2 (Student t-test).
Immune response induced by DCs that had phagocytosed apoptotic/necrotic tumor cells
DCs are potent stimulators of primary MLRs45 and induce the proliferation of allogeneic CD8+ T cells in vitro.46 To characterize this function, we compared DCs that had phagocytosed apoptotic/necrotic tumor cells with immature DCs in terms of their effect on primary allogeneic MLRs. As shown in Figure 6, DCs that had phagocytosed apoptotic/necrotic tumor cells more strongly stimulated allogeneic T-cell proliferation than immature DCs.
The phenotype of T cells activated by DCs that had phagocytosed apoptotic/necrotic BL6-10 cells were evaluated by examining the cytokines they secreted in ELISA. As shown in Table II, in response to DC vaccination, T lymphocytes from RDLNs of mice vaccinated with DCs that had phagocytosed apoptotic/necrotic BL6-10 cells and DCs pulsed with mTRP2 peptide all secreted a higher level of IFN-γ than lymphocytes from mice vaccinated with DCs alone, whereas the secretion of IL-4 was low in all groups. These patterns are consistent with an enhanced Th1-dominant response in mice vaccinated with DCs that had phagocytosed apoptotic/necrotic BL6-10 tumor cells and DCs pulsed with mTRP2 peptide.
Table II. Cytokine Secretion in Response to Co-Culture With Irradiated BL6-10 Cells
DC, dendritic cell; IFN, interferon; IL, interleukin.
DC/mTRP2 and DC/Apopt represent DCs pulsed with mTRP2 peptide and DCs that had phagocytosed apoptotic/necrotic BL6-10 cells, respectively.
T cells from the draining lymph nodes harvested 7 days after DC vaccination were co-cultured with irradiated BL6-10 cells for 1 and 3 days. The supernatants were harvested for measurement of IFN-γ and IL-4 secretion, respectively, in an enzyme-linked immunosorbent assay.
Data represent the mean ± SD pg/ml/1 × 106 cells for IL-4 and IFN-γ of triplicate samples.
Not significant (p > 0.1) versus DC/mTRP2 group and significant (p < 0.05) versus DC group (Student t-test).
To examine the immune mechanisms involved in the protective immunity, we performed a cytotoxicity assay. Splenocytes from immunized mice were co-cultured with irradiated BL6-10 cells. After 5 days in culture, T lymphocytes were harvested and analyzed by flow cytometry. These T lymphocytes, which were mostly CD8+ T cells as analyzed by flow cytometry (data not shown), were used as effector cells in a chromium release assay. As shown in Figure 7, T lymphocytes derived from mice vaccinated with DCs that had phagocytosed apoptotic/necrotic tumor cells displayed enhanced cytotoxicity to BL6-10 cells (62% specific killing at an E:T cell ratio of 50) than those of mice vaccinated with DCs pulsed with mTRP2 peptide (40% specific killing at an E:T cell ratio of 50). However, the T lymphocytes derived from mice vaccinated with DCs that had phagocytosed apoptotic/necrotic tumor cells did not show any cytotoxic activity to EL4 tumor cells. Lymphocytes derived from naive mice also showed low killing activity to BL6-10 cells (4%). Our data indicate that vaccination of DCs that had phagocytosed apoptotic/necrotic BL6-10 cells was able to induce more efficient cytotoxic T-cell responses against the B16 melanoma than DCs pulsed with mTRP2 peptide.
Enhanced antitumor immunity induced by DC vaccination
To examine whether DCs that had phagocytosed apoptotic/necrotic BL6-10 cells were capable of induction of enhanced antitumor immunity, mice were vaccinated with the latter DCs or DCs pulsed with the tumor peptide, and challenged with BL6-10 tumor cell injection. As shown in Table III and Figure 8, mice vaccinated with DCs that had phagocytosed apoptotic/necrotic tumor cells showed no lung metastasis whereas mice vaccinated with DCs pulsed with the tumor peptide depicted an average of 16 lung metastases. The number of lung metastases in the control group of mice vaccinated with DCs alone was more than 200. Our data thus indicate that the vaccination of DCs that had phagocytosed apoptotic/necrotic tumor cells induces a more efficient antitumor immunity than that of DCs pulsed with the tumor peptide. In addition, the protective immunity against BL6-10 tumor challenge derived from vaccination of DCs that had phagocytosed apoptotic/necrotic tumor cells lasted at least 4 months (data not shown), indicating that long-term immunologic memory was induced after rechallenge of the animals.
Table III. Eradication of Lung Metastasis by Vaccination of DCs That Had Phagocytosed Apoptotic/Necrotic BL6-10 Tumor Cells
DC/mTRP2 and DC/apopt represent DCs pulsed with mTRP2 peptide and DCs that had phagocytosed apoptotic/necrotic BL6-10 cells, respectively.
Values represent the mean number of pulmonary metastases from 6 vaccinated mice of each group 3 weeks after intravenous injection of 3 × 105 BL6-10 tumor cells. Two independent experiments were performed. The results were consistent.
16 ± 5
The ability of DCs to present tumor antigen from phagocytosed apoptotic tumor cells and to stimulate tumor-specific CTL responses prompted us to evaluate the efficiency of vaccine strategies using these DCs and to study the immune mechanisms involved therein. In our animal tumor models, vaccination of mice using DCs pulsed with mTRP2 peptide also significantly reduced lung metastases with a mean number of metastasis of 16 compared with more than 200 in control mice vaccinated with DCs alone. Furthermore, we compared the efficiency of antitumor immunity derived from vaccination with DCs that had phagocytosed apoptotic/necrotic BL6-10 tumor cells with DCs pulsed with mTRP2 peptide. Our study showed that vaccination of DCs that had phagocytosed apoptotic/necrotic cells was able to induce stronger stimulation of allogeneic T-cell proliferation in vitro; stimulate Th1 immune response in vivo; promote more efficient tumor-specific CD8+ CTL-mediated immunity; and eradicate lung metastases in all 6 vaccinated mice, indicating a more robust antitumor immunity through vaccination of DCs pulsed with mTRP2 peptide. The reason for this robust antitumor immunity may have been because mice vaccinated with DCs that had phagocytosed apoptotic/necrotic BL6-10 tumor cells might generate an immune response not only against mTRP2- antigen but also against other tumor antigens presented by BL6-10 cells. Furthermore, contrary to the peptide-based approach, the use of DCs that have phagocytosed apoptotic/necrotic tumor cells may also provide both MHC class I and II epitopes and does not require the identification of tumor-associated antigens. Our data thus indicate that vaccines based on DCs that had phagocytosed apoptotic/necrotic tumor cells have additional advantages in tumor immunotherapy.
The degree of DC differentiation (immature vs. mature) determines subsequent function. In general, antigen processing is maximal in immature DCs whereas T-cell sensitization is more effective in mature DCs with enhanced expression of MHC class II, CD40, co-stimulatory and adhesion molecules. Some studies have shown that phagocytosis of apoptotic tumor cells by DCs induced DC maturation.23, 47 Recently, Sauter et al.48 reported different results—that the exposure of immature DCs to the necrotic, but not to the primary or apoptotic tumor cells induced DC maturation. In their studies, however, they also showed that phagocytosis of a mixture of necrotic/apoptotic tumor cells was able to induce DC maturation. In the present study, we demonstrated that phagocytosis of apoptotic/necrotic tumor cells by immature DCs resulted in maturation of DCs with up-regulated expression of cytokines (IL-1β, IL-6, TNF-α, IFN-γ and GM-CSF) and chemokines (MIP-1α, MIP-1β and MIP-2). The proinflammatory cytokines such as IL-1β, IL-6, TNF-α and IFN-γ are able to stimulate DCs into more mature stages with strong T-cell stimulatory potential.40, 41,  We subsequently determined whether the secretion of proinflammatory cytokines induced by DC phagocytosis of apoptotic/necrotic tumor cells was accompanied by DC maturation. Our flow cytometry data demonstrated that DCs that had phagocytosed apoptotic/necrotic tumor cells displayed up-regulated expression of cell surface molecules such as MHC class II antigen, CD11b, CD40 and CD86. These data indicate that phagocytosis of apoptotic/necrotic tumor cells induced DC maturation.
CC chemokines MIP-1α and MIP-1β are chemotactic for macrophages and T cells,50, 51 respectively, whereas CXC chemokine MIP-2 is for neutrophils.52 It has been reported that infiltration of macrophages, T cells and neutrophils into tumors resulted in inhibition of tumor growth and antitumor immune responses.51, 53, 54 GM-CSF is a growth factor for hematopoietic progenitor cells. GM-CSF-secreting tumor cell vaccines have been shown to elicit tumoricidal antitumor immune responses by recruiting dendritic cells to immunization sites in animal models and in human clinical trials.55–57 It has recently been reported that transfection of DCs with the GM-CSF gene potently enhances their in vivo antigen-presenting capacity.34 Therefore, up-regulation of MIP-1α, MIP-1β, MIP-2 and GM-CSF may play some role in enhanced antitumor immunity of DCs that had phagocytosed apoptotic/necrotic tumor cells.
The capacity of mature DC to migrate into T-cell areas of LNs for induction of a primary immune response is a key factor in initiating immunity.58 Recent studies have demonstrated that chemokines play a critical role in DC migration. The migratory capability of DCs is dictated by the change of responsiveness of DCs to various chemokines during their development and maturation.59–63 Immature DCs respond to MIP-3α, RANTES and MIP-1α via chemokine receptor CCR1, CCR5 and CCR6, whereas mature DCs respond to MIP-3β and SLC via CC chemokine receptor CCR7 receptor.59 The down-regulation of receptors for the inflammatory cytokines and up-regulation of CCR7 receptor for MIP-3β that is expressed in secondary lymphoid organs such as LNs allow mature DCs to leave the sites of inflammation and to migrate to LNs for activation of T lymphocytes. In the present study, we showed that DCs that had phagocytosed apoptotic/necrotic tumor cells displayed down-regulated expression of CCR2 and CCR5 receptors and up-regulated expression of the CCR7 receptor, respectively, and demonstrated enhanced migration toward the CC chemokine MIP-3β using a chemotaxis assay in vitro. DCs that had phagocytosed apoptotic/necrotic tumor cells showed a relative migration index of 7.23% compared with 2.56% (p < 0.01) for DCs pulsed with tumor peptide in an animal model in vivo. This finding indicates that DC maturation promotes the migration of DCs into regional LNs in vivo, which is consistent with 2 recent reports showing enhanced migration of mature DCs expressing the CCR7 receptor toward LNs in vivo.37, 64
In conclusion, this study demonstrates that vaccination using DCs that had phagocytosed apoptotic/necrotic tumor cells induces stronger antitumor immunity, even against poorly immunogenic tumor cells, compared with that of DCs pulsed with MHC class I-restricted tumor peptide alone. The principle of this study could be applied in the clinical setting, namely, inducing apoptosis/necrosis of autologous tumor cells by reagents such as lovastatin in vitro followed by vaccinating the patients with their respective DCs that have phagocytosed these apoptotic/necrotic tumor cells. This method would allow patients to benefit from cancer immunotherapy designed specifically for each patient.