Dendritic cells (DC) are increasingly applied as a cellular adjuvant in immunotherapy of cancer. Two major myeloid DC subsets are recognized: interstitial DC (IDC) that infiltrate connective tissues and Langerhans cells (LC) that line epithelial surfaces. Yet, functional differences between IDC and LC remain to be defined. We recently showed that the CD34+ acute myeloid leukemia cell line MUTZ-3 supports differentiation of both DC-SIGN+ IDC and Langerin-positive Birbeck granule-expressing LC. By comparative functional characterization of MUTZ-3 IDC and MUTZ-3 LC, we aimed to elucidate the relative abilities of these two DC subsets to induce a specific T cell response and reveal the more suitable candidate for use as a clinical vehicle of tumor vaccines. Although mature LC and IDC displayed comparable lymph node-homing potential, mature LC showed higher allogeneic T cell stimulatory capacity. Nevertheless, IDC supported the induction of tumor Ag-specific CD8+ T cells at an overall higher efficiency. This might be related to the observed inability of LC to release T cell stimulatory cytokines such as IL-12p70, IL-23, and IL-15. Although this inability did not result in a detectable deviation in the cytokine expression profile of primed T cells, transduction with IL-12p70 significantly improved priming efficiency of LC, and ensured a functional equivalence with IDC in this regard. In conclusion, except for the inability of LC to release distinct type 1 T cell stimulatory cytokines, in vitro function of LC and IDC suggests comparable abilities of both subsets for the in vivo induction of antitumor T cells.
Dendritic cells (DC)2 are professional APC with the unique ability to initiate primary T cell responses (1, 2, 3). Myeloid DC differentiate from CD34+ hematopoietic progenitor cells and can develop into two recognized subsets, Langerhans cells (LC), which are mostly found in the epidermis, and interstitial DC (IDC), which are mostly located in connective tissues (4, 5). Because of their critical role in orchestrating the immune response, DC have been used as a cellular adjuvant in immunotherapy of cancer. Numerous animal studies have shown that administration of tumor Ag-loaded DC could induce protective immunity in naive animals (6, 7, 8), as well as tumor regression and tumor-free survival in mice bearing established tumors (9). The immunogenicity of DC has also been shown in human clinical trials. Initially, DC vaccination studies were performed with DC directly isolated from peripheral blood (10, 11). Major disadvantages of this approach were the low yields of DC (11), the variety of peripheral blood DC subsets (12) and the varying percentages of these DC subsets in cancer patients. The development of methods to generate DC from CD34+ hematopoietic progenitor cell progenitors (CD34-DC) or CD14+ monocytes (monocyte-derived DC (MoDC)) in vitro, have facilitated DC vaccination studies tremendously. Numerous clinical trials, using either MoDC or CD34-DC, have demonstrated the feasibility and safety of DC-based vaccination approaches, of which some reported encouraging clinical responses (13, 14, 15, 16).
So far, in the vast majority of clinical trials MoDC have been used for vaccination, while only few trials have made use of CD34-DC (as reviewed by Davis et al. (17)). However, it has been suggested that CD34-DC are more effective as compared with MoDC in clinical DC vaccination studies (15), as well as in in vitro Ag-specific T cell stimulation (18). The beneficial effect of these CD34-DC might be due to the presence of “contaminating” LC in the CD34-DC preparations. This hypothesis was supported by Ratzinger and coworkers (19), demonstrating that CD34-LC exhibit a superior capacity to induce influenza virus- and tumor-specific CTL in vitro, compared with MoDC and CD34-DC (both belonging to the IDC subset). For clinical DC vaccination purposes, however, functional differences between IDC and LC remain to be characterized more extensively, not only at the level of T cell priming, but also for their capacity to migrate toward lymph nodes (LN) and the ability of IDC and LC to release T cell stimulatory cytokines and skew Th1/Th2 cytokine profiles. Such extensive comparative studies are hampered by difficulties in the generation of large amounts of Birbeck granule-containing Langerin-positive LC. We have previously shown that DC-SIGN+ IDC and bona fide Langerin/Birbeck granule-positive LC can be generated from the MUTZ-3 cell line, a human CD34+ acute myeloid leukemia cell line (20, 21). In addition, we observed equivalent in vitro T cell priming efficiency for autologous MoDC and MUTZ-3 IDC, matched for HLA-A2 or HLA-A3 expression, both generating functional tumor-specific CTL (22). These observations indicate the validity of the MUTZ-3 cell line model, which facilitates extensive head-to-head comparative functional studies of IDC vs LC.
In the current study we characterized functional properties crucial for the in vivo generation of CTL-mediated immunity of MUTZ-3-derived IDC and LC. Our data show that both DC subsets are capable of tumor-specific CTL induction. However, T cell priming efficiency of IDC was superior compared with LC, most likely due to the inability of LC to release type 1 T cell stimulatory cytokines. Of note, study of the corresponding DC subsets from human skin (i.e., primary LC and IDC/dermal DC (DDC)) similarly revealed superior CD8+ T cell priming efficiency for DDC over LC. Transduction with IL-12p70 significantly enhanced the CTL priming efficiency of MUTZ-3-derived LC and ensured a functional equivalence with IDC in this regard. We conclude that the MUTZ-3 cell line is a valid human DDC/LC cell line model and that both DC subsets are viable vaccine vehicles for tumor immunotherapy.
Materials and Methods
The human CD34+ acute myeloid leukemia cell line MUTZ-3 (Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany) was cultured as previously described (20, 22). The EBV-transformed B cell line JY and the TAP-deficient cell line T2 (both HLA-A2+) were cultured in IMDM (BioWhittaker) supplemented with 10% FBS (Perbio), 100 IU/ml sodium penicillin (Yamanouchi Pharma), 100 μg/ml streptomycin sulfate (Radiumfarma-Fisiopharma), 2.0 mM l-glutamine (Invitrogen Life Technologies), and 0.01 mM 2-ME (Merck), i.e., complete medium. The prostate cancer cell line PC-3, the colon carcinoma cell line SW620, the glioblastoma cell line U251, and the melanoma cell lines Mel-JKO and Mel-AKR were cultured in DMEM complete medium.
The HLA-A2-restricted peptides carcinoembryonic Ag (CEA)571–579 (YLSGANLNL), prostate-specific Ag (PSA)146–154 (KLQCVDLHV), Her-2/neu369–377 (KIFGSLAFL), MART-126–35L (ALGIGILTV), and Bcr-abl926–934 (SSKALQRPV) were synthesized by solid-phase strategies on an automated multiple peptide synthesizer (Syro II; MultiSyntech) using F-moc chemistry. Peptides were analyzed by reversed-phase HPLC, dissolved in DMSO (Merck), and stored at −20°C.
Retroviral IL-12p70 transduction of MUTZ-3
The IL-12elasti cassette containing the p35 and p40 subunits of IL-12 joined together by a flexible linker (InvivoGen) was introduced into MUTZ-3 cells by retroviral transduction. The truncated, signaling defective form of the nerve growth factor receptor (ΔNGFR) was used as a marker gene for transduced cells. The retroviral vector LZRS-IL-12-IRES-ΔNGFR, constructed by replacing the human TERT open reading frame from the LZRS-hTERT-IRES-ΔNGFR construct (23) with the IL-12elasti cassette, was used to produce the retroviral supernatant. Retroviral transduction of MUTZ-3 was performed as previously described (24). Briefly, 5 × 105 MUTZ-3 cells, were resuspended in retroviral supernatant supplemented with 10% 5367 conditioned medium and transferred to a fibronectin-coated (RetroNectin; Takara Shuzo) well of a nontissue culture-treated 24-well plate (BD Biosciences). Plates were centrifuged, followed by 5-h incubation at 37°C. The next day, retroviral transduction was repeated. A NGFR-specific Ab (Chromoprobe) was used to analyze transduction efficiency and isolate transduced cells by flow sorting.
In vitro generation of MoDC and MUTZ-3-derived IDC and LC
MoDC, MUTZ-3 IDC, MUTZ-3 LC, and MUTZ-3 IL-12-derived IDC and LC (referred to throughout this study as MUTZ-3-IL-12 IDC and LC) were generated as described (20, 21, 25). Briefly, PBMC of normal human volunteers were allowed to adhere to the bottom of plastic culture flasks for 1–2 h (Greiner Bio-One) at 37°C, nonadherent cells were removed and the adherent cells were cultured for 5–7 days in IMDM complete medium supplemented with 100 ng/ml GM-CSF (Schering-Plough) and 10 ng/ml IL-4 (R&D Systems). MUTZ-3 progenitors were cultured in 12-well tissue-culture plates at a concentration of 1 × 105/ml in the presence of 100 ng/ml GM-CSF, 20 ng/ml IL-4, and 2.5 ng/ml TNF-α (Strathmann Biotec) for 7 days for MUTZ-3 IDC differentiation and in the presence of 100 ng/ml GM-CSF, 10 ng/ml TGF-β (BioVision), and 2.5 ng/ml TNF-α for 10 days for MUTZ-3 LC differentiation. Every 3 days new cytokines were added. At day 7 or 10, maturation of MoDC, MUTZ-3 IDC, and MUTZ-3 LC was induced by adding a cytokine mixture consisting of 50 ng/ml TNF-α, 100 ng/ml IL-6, 25 ng/ml IL-1β (referred to throughout this study as MCM mimic; Strathmann Biotec), with or without 1 μg/ml PGE2 (Sigma-Aldrich) or by adding a type 1-polarizing cytokine mixture containing 50 ng/ml TNF-α, 25 ng/ml IL-1β, 1000 U/ml IFN-γ, 20 μg/ml polyinosinic-polycytidylic acid (poly(I:C); Sigma-Aldrich), and 3000 U/ml IFN-α (PeproTech), referred to throughout this study as IFN-α/poly(I:C) type 1 cytokine mixture (26).
Abs, tetramers, and flow cytometry
PE- or FITC-labeled Abs directed against human CD34 (Strathmann Biotec), CD40, CD83, CD8β, Langerin, αβ TCR, γδ TCR (Immunotech), CD1a, CD8α, CD14, CD28, CD27, CD45RA, CD54RO, CD54, CD62 ligand, CD80, CD86, CCR7 (2H4), DC-SIGN, HLA-DR (all from BD Biosciences) were used for flow cytometric analysis. PE- or allophycocyanin-labeled HLA-A2 tetramers presenting the CEA571, PSA146, MART-126L, and Her-2/neu369 epitopes were prepared as previously described (27). Ab or tetramer staining was performed in PBS supplemented with 0.1% BSA and 0.02% natrium azide for 30 min at 4°C and for 15 min at 37°C, respectively. Stained cells were analyzed on a FACSCalibur (BD Biosciences) using CellQuest software. To exclude dead cells in flow cytometric tetramer analysis, 0.5 μg/ml propidium iodide (ICN Biomedicals) was used.
Primary CTL induction in vitro
Ag-specific CD8+ T cells were generated as described, with some minor modifications (22, 23). In brief, mature MUTZ-3 IDC and MUTZ-3 LC, either derived from wild-type MUTZ-3 or IL-12-transfected MUTZ-3, were loaded with 1 μg/ml peptide in the presence of 3 μg/ml β2-microglobulin (Sigma-Aldrich) for 2–4 h at room temperature and irradiated (40 Gy). A total of 1 × 105 peptide-loaded DC were cultured for 10 days with 1 × 106 CD8β+CD8+ T precursors and 1 × 106 irradiated (80 Gy) CD8β− autologous PBMC in Yssel medium (28) supplemented with 1% human AB serum (ICN Biochemicals), 10 ng/ml IL-6, and 10 ng/ml IL-12 (in experiments with wild-type MUTZ-3 IDC and MUTZ-3 LC) in a 24-well tissue-culture plate. At day 1, 10 ng/ml IL-10 (R&D Systems) was added. After 10 days, CD8+ T cell cultures were restimulated with 1 × 105 fresh MUTZ-3 IDC or MUTZ-3 LC, loaded with 10 ng/ml peptide, in the presence of 5 ng/ml IL-7 (Strathmann Biotec). From day 17 onward, CD8+ T cell cultures were stimulated weekly with 1 × 105 peptide-loaded JY cells, unless indicated otherwise. Two days after each restimulation, 10 U/ml IL-2 (Strathmann Biotec) was added. One day before each restimulation, a sample was taken and analyzed by flow cytometry using both PE- and allophycocyanin-labeled tetramers presenting the relevant epitope. Tetramer-positive CD8+ T cells were isolated by tetramer-positive magnetic cell sorting and subsequently expanded. For this purpose, CD8+ T cells were weekly stimulated with irradiated feeder-mix consisting of allogeneic PBMC and JY cells in Yssel medium supplemented with 100 ng/ml PHA (Murex Biotech) and 20 U/ml IL-2.
Intracellular IFN-γ detection
To determine the capacity of the CD8+ T cells to produce IFN-γ upon recognition of a specific target, intracellular IFN-γ staining was performed. CD8+ T cells were cultured with target cells at an E:T ratio of 2:1 in 96-well round-bottom plate in the presence of 0.5 μl of GolgiPlug (BD Biosciences). After 5 h, cells were harvested, washed, stained with allophycocyanin-labeled tetramer, and PE-labeled anti-CD8 mAb. After fixation with Cytofix/Cytoperm solution and permeabilization with Perm/Wash solution (both from BD Biosciences), cells were labeled with FITC-conjugated anti-IFN-γ Ab (BD Biosciences). Stained cells were analyzed on a FACSCalibur.
Chromium release assay
Cytotoxic activity of CD8+ T cell lines was determined by standard chromium release assay as described (23). Target cells (0.5 × 106) were labeled with 100 μCi of Na2[51Cr]O4 (Amersham Biosciences) for 1 h at 37°C, washed extensively, and loaded with 1 μg/ml peptide (when indicated) for 1 h at 37°C. Effector CTL clones were added to 2 × 103 target cells at the indicated E:T ratio in triplicate wells of a round-bottom 96-well plate. After a 4-h incubation at 37°C, 50 μl of the supernatant was harvested, and its radioactive content was measured. The percentage-specific lysis was defined as follows: (experimental release − spontaneous release)/(maximum release − spontaneous release) × 100.
The T cell stimulatory capacity of MUTZ-3 IDC and MUTZ-3 LC was determined by allogeneic MLR as described (21). In brief, graded doses of immature and mature MUTZ-3 DC and MUTZ-3 LC were irradiated (80 Gy) and added as stimulator cells to round-bottom, 96-well plates (Greiner Bio-One). Nonadherent PBMC or CD8β+ T cells were used as a source for responder cells, and 1 × 105 cells/well were added to the allogeneic DC. The cells were cultured for 4–5 days in complete medium, and during the last 18 h, [3H]thymidine (Amersham Biosciences) was added (0.4 μCi/well). Subsequently, the cells were harvested onto fiberglass filters, and [3H]thymidine incorporation was determined using a flatbed liquid scintillation counter (Wallac).
Transwell migration assay
Chemotaxis of mature MoDC, MUTZ-3 IDC, and MUTZ-3 LC was measured by migration through a polycarbonate filter of 5-μm pore size. For this purpose, 100,000 mature MoDC, MUTZ-3 IDC, or MUTZ-3 LC were seeded in serum-free medium in the upper well of a 24-well Transwell chamber and allowed to migrate for 4 h toward serum-free medium alone or serum-free medium containing 250 ng/ml of the chemokines CCL19 (PeproTech) or CCL21 (BioSource International). After 4 h, the migrated cells were harvested and counted by flow cytometry by adding a fixed number of fluorescent flow count beads (Beckman Coulter). The absolute number of migrated cells was expressed as follows: (total number of cells counted × total number of added beads)/total number of counted beads.
Cytokine release by DC
Immature MoDC, MUTZ-3 IDC, and MUTZ-3 LC were analyzed for the release of IL-12p70, IL-10, and IL-15 as previously described (26, 29). Briefly, for IL-12 and IL-10 production, the different DC subsets were stimulated for 8 h in a 96-well round-bottom culture plate in IMDM plus 10% FCS with either MCM mimic with or without PGE2 or with the IFN-α/poly(I:C) type 1-polarizing cytokine mixture, followed by 16 h of stimulation with the CD40L-transfected J558 cell line in the presence or absence of 1000 U/ml IFN-γ (R&D Systems). IL-12p70 production by MUTZ-3-IL-12 IDC and LC was determined in the absence of CD40 ligation. To do so, MUTZ-3-IL-12 IDC and LC were stimulated for 48 h with MCM mimic in the presence of TNF-α. After cytokine removal, cells were cultured for an additional 24 h, and supernatants were collected. For IL-15 production, the different DC subsets were stimulated for 20 h with either MCM mimic with or without PGE2 or the IFN-α/poly(I:C) type 1-polarizing cytokine mixture. Next day, the supernatants were harvested and analyzed by ELISA to detect IL-10 (IL-10 ELISA kit; Sanquin), IL-12p70, or IL-15 (QuantiGlo ELISA kit; R&D Systems).
Cytokine production by MUTZ-3 IDC and MUTZ-3 LC-stimulated CD8+ and CD4+ T cells
A total of 1 × 106 CD8β+ T cells and CD8β− PBMC were cultured with either medium alone (unstimulated) or medium containing 1 × 105 mature MUTZ-3 IDC or MUTZ-3 LC in the presence of 5 ng/ml IL-7 in a 24-well plate. On day 7, the T cells were restimulated with 25 ng/ml PMA and 1 μg/ml ionomycin (Sigma-Aldrich) in the presence of GolgiPlug (BD Biosciences) to detect intracellular production of IFN-γ and IL-4. After 4 h, cells were harvested, washed, and stained with allophycocyanin-labeled anti-CD8 or anti-CD4 and PerCP-Cy-5.5-labeled anti-CD3 mAbs. After fixation and permeabilization, cells were labeled with FITC-labeled anti-IFN-γ mAb (BD Biosciences) and PE-labeled anti-IL-4 mAb. Stained cells were analyzed on a FACSCalibur.
Isolation of DDC and LC from skin
Human HLA-A2+ skin specimens were obtained from healthy donors undergoing corrective breast or abdominal plastic surgery after informed consent. The 3-mm thick slices of skin containing both the epidermis and the dermis were cut by use of a dermatome. Slices of skin were cut in pieces of 1 cm2 and incubated with 2.4 U/ml Dispase II (Roche Diagnostics) for 30 min at 37°C. The epidermis and dermis were separated with tweezers and subsequently (separately) cultured in IMDM complete medium supplemented with 1000 U/ml GM-CSF and 10 ng/ml IL-4 for 48 h, after which the epidermal and dermal sheets were removed. After 48–72 h, epidermis- and dermis-migrated cells were harvested and counted, and the percentage of CD1a+ cells within these populations was determined by flow cytometry.
Phenotypic and functional analysis of skin-emigrated DDC and LC
Phenotype of epidermis- and dermis-emigrated DDC and LC was analyzed by flow cytometry. For further functional analysis, the amount of DDC and LC used was calculated based on the expression of the pan-skin DC marker CD1a. Allogeneic T cell stimulatory capacity was determined by MLR as already described. LN-homing capabilities of skin-emigrated DDC and LC was determined as described, with some minor modifications. In brief, 25,000 skin-emigrated DDC and LC were allowed to migrate toward serum-free medium alone or serum-free medium containing 250 ng/ml of the chemokines CCL19 or CCL21. After 18 h, the percentage of migrated CD1a+ DDC and LC was determined by fluorescent flow count bead-mediated counting on a flow cytometer. In vitro tumor-specific CTL-inducing capabilities of skin-emigrated DDC and LC was determined as described, with some minor modifications. In brief, multiple bulk cultures of 1 × 106 CD8β+ T precursors were cultured with 5 × 104 MART-126L peptide-loaded CD1a+ skin DDC or LC and with 1 × 106 irradiated (80 Gy) CD8β− autologous PBMC in Yssel medium supplemented with 1% human AB serum (ICN Biochemicals), 10 ng/ml IL-6, and 10 ng/ml IL-12 in a 24-well tissue-culture plate. At day 1, 10 ng/ml IL-10 (R&D Systems) was added. After 10 days, CD8+ T cell cultures were analyzed for the presence of MART-126L-specific CD8+ T cells by tetramer staining.
Differences between mean percentage of migrated DC, between mean levels of T cell stimulatory capacity, and between mean values of cytokine production of the different DC subsets were compared using a paired two-sided t test. Differences between median percentage of tetramer-positive CD8+ T cells were compared using a two-sided Mann-Whitney U test. Differences were considered significant when p < 0.05.
Phenotype and allogeneic T cell stimulatory capacity of MUTZ-3-derived IDC and LC
As shown in Fig. 1 and previously described (21), MUTZ-3 progenitors are able to differentiate into either IDC, expressing CD1a and DC-SIGN but lacking Langerin, or into LC, expressing CD1a and Langerin, but lacking DC-SIGN (Fig. 1,A). Upon maturation, up-regulation of costimulatory molecules, de novo expression of CD83, and enhanced allogeneic T cell stimulation was observed for both MUTZ-3 IDC and MUTZ-3 LC (21). Of note, expression levels of costimulatory and adhesion molecules were generally higher on MUTZ-3 LC (Fig. 1,A), which translated into a significantly higher proliferation of total PBL (p < 0.03) in an allogeneic MLR (Fig. 1,B). Importantly, this increased proliferation was also observed (p < 0.023) for isolated CD8β+ CTL precursors (Fig. 1 C), revealing a superior allogeneic (CD8+) T cell stimulatory capacity for MUTZ-3 LC as compared with MUTZ-3 IDC.
LN migration capacity of MoDC, MUTZ-3 IDC, and MUTZ-3 LC
DC are professional APC with the capacity to take up Ag in the periphery and present this Ag in the secondary lymphoid organs. As described, the migration of MoDC is dependent on PGE2 (30, 31, 32). To determine whether mature MUTZ-3 IDC and LC are capable of migrating and whether this migration is also dependent on PGE2, MUTZ-3 IDC and MUTZ-3 LC were analyzed for the expression of the chemokine receptor CCR7 by flow cytometry and for their migratory capacity toward the LN-homing chemokines CCL19 and CCL21 in Transwell migration assays. To this end, MUTZ-3 IDC, MUTZ-3 LC, and MoDC matured with a mix of cytokines (MCM mimic) with or without PGE2 were tested as such. Flow cytometric analysis revealed up-regulation of CCR7 on matured MoDC, MUTZ-3 IDC, and MUTZ-3 LC (data not shown), which was accompanied by migration toward LN-homing chemokines CCL19 and CCL21 (Fig. 2). No significant difference in the percentage of migrating cells of all three MCM mimic plus PGE2-matured DC subtype combinations was observed. However, in contrast to MoDC and MUTZ-3 LC, the migration of MUTZ-3 IDC was not dependent on the presence of PGE2 in the maturation mix (Fig. 2). Although PGE2 has been described to be a critical component for DC migration, it has been demonstrated recently that immunostimulatory DC with LN migration and IL-12p70 release capabilities can be generated in the absence of PGE2, by making use of an IFN-α/poly(I:C) type 1 cytokine mixture (26). To this end, IFN-α/poly(I:C) type 1-matured MUTZ-3 IDC, MUTZ-3 LC, and MoDC were also tested for their responsiveness toward CCL19 and CCL21. As shown in Fig. 2, MUTZ-3 IDC were more responsive to CCL19 and CCL21 upon IFN-α/poly(I:C) type 1-induced maturation as compared with MoDC and MUTZ-3 LC, underlining the observation that MUTZ-3 IDC are more capable of PGE2-independent migration.
Cytokine secretion by MUTZ-3 IDC, MUTZ-3 LC, and MoDC
Another important function of fully mature, immunostimulatory DC is the ability to produce cytokines important for helper T cell and cytotoxic T cell activation, like IL-12, IL-23, and IL-15 (33, 34), whereas cytokines that play a role in the induction of T cell tolerance, such as IL-10, should preferably not be produced (35). For that reason, CD40L-stimulated MUTZ-3 IDC, MUTZ-3 LC, and MoDC were analyzed for the release of IL-12p70, IL-15, IL-23, and IL-10. Whereas MoDC released considerable amounts of IL-10 (ranging from 2 to 5 ng/ml) with the maturation methods used in this study, no IL-10 was secreted by either MUTZ-3 IDC or MUTZ-3 LC (Fig. 3,A). MoDC also secreted higher amounts of IL-12p70 as compared with MUTZ-3 IDC, whereas MUTZ-3 LC did not secrete any detectable IL-12 (Fig. 3,B). The IL-12p70 secretion by MoDC was impaired by the addition of PGE2 in the maturation mix, although this amount was not significant. MUTZ-3 IDC, MUTZ-3 LC, and MoDC were also analyzed for the capacity to produce IL-15 and IL-23. As shown in Fig. 3 C, immature and MCM mimic with or without PGE2-matured MUTZ-3 IDC, MUTZ-3 LC, and MoDC hardly produced IL-15, whereas IFN-α/poly(I:C) type 1-matured MoDC (p < 0.028) and MUTZ-3 IDC (p < 0.0033) but not MUTZ-3 LC, did produce significantly higher amounts of IL-15. IL-23 was not released at detectable levels by either IDC or LC (data not shown). In summary, CD40L-stimulated MUTZ-3 IDC was shown to release T cell stimulatory cytokines, albeit at relatively low levels compared with MoDC, whereas MUTZ-3 LC did not do so at all.
Cytokine profile of MUTZ-3 IDC- and MUTZ-3 LC-stimulated CD8+ and CD4+ T cells
To study whether the low level or absence of IL-12p70 released by MUTZ-3 IDC and MUTZ-3 LC, respectively, led to a change in cytokine production by T cells, thereby shifting T cells toward a more Th2 cytokine profile, we analyzed MUTZ-3 IDC- and MUTZ-3 LC-stimulated CD8β+ T cells and CD8β-depleted PBL for the production of IL-4 and IFN-γ. As shown in Fig. 3D, both MUTZ-3 IDC- and MUTZ-3 LC-stimulated CD8+ and CD4+ T cells mainly produced IFN-γ, but not IL-4. Moreover, the production of IFN-γ by CD8+ and CD4+ T cells was significantly increased as compared with unstimulated CD8+ and CD4+ T cells. Importantly, in all conditions, no IFN-γ or IL-4 double positive T cells could be detected (data not shown), suggesting that MUTZ-3 IDC and MUTZ-3 LC primarily expanded type 1 cytokine-producing T cells, despite the low level or complete absence of IL-12p70 during stimulation.
Induction of tumor-specific CD8+ T cells in vitro using peptide-pulsed HLA-A2-matched MUTZ-3 IDC or MUTZ-3 LC
DC exhibit the unique capacity to induce and activate tumor-specific CTL both in vivo and in vitro. We previously described that MUTZ-3 IDC can be used for in vitro priming and expansion of functional TAA-specific effector CTL and that the efficiency of this induction was comparable to autologous peptide-loaded MoDC (22). To determine whether MUTZ-3 LC were also capable of inducing functional tumor-specific CTL and whether the CTL induction efficiency is comparable to that of MUTZ-3 IDC, both IDC and LC were used as stimulator cells in an in vitro CTL induction protocol. To this end, HLA-A2+ CD8β+ CTL precursors were stimulated with HLA-A2-matched and peptide-loaded mature MUTZ-3 IDC and MUTZ-3 LC in multiple parallel cultures starting with 1 × 106 CD8β+ CTL precursors in each culture. Epitopes selected were the immunodominant HLA-A2-restricted MART-1-derived peptide MART-126L, the CEA-derived peptide CEA571, the Her-2/neu-derived peptide Her-2/neu369, and the PSA-derived epitope PSA146. From the second round of stimulation onward, the presence of tumor-specific CD8+ T cells was monitored by tetramer staining. When stimulated with MUTZ-3 IDC, significantly higher levels of MART-126L tetramer-positive T cells could be detected after two events of in vitro stimulation, when compared with stimulation of MUTZ-3 LC, in two of three donors analyzed (Fig. 4,A). In addition, a similar trend in the number of tetramer-positive cultures was also observed for the adenocarcinoma-associated peptides CEA571, PSA146, and Her-2/neu369 (17/48 vs 9/48 individual cultures for MUTZ-3 IDC and MUTZ-3 LC, respectively). A higher detectable maximum tetramer-positive T cell frequency was also observed (Fig. 4 B), although these differences did not reach statistical significance.
Although functionally active, functional avidity of the MUTZ-3 IDC- and MUTZ-3 LC-generated MART-126L- and Her-2/neu369-specific CD8+ T cells were relatively low. Both MUTZ-3 IDC- and MUTZ-3 LC-generated MART-126L- and Her-2/neu369-specific CD8+ T cells were able to recognize exogenously loaded MART-126L- and Her-2/neu369 target cells, respectively. However, recognition of endogenously processed and presented MART-126L and Her-2/neu369 on tumor cells was low, as analyzed by intracellular IFN-γ (Fig. 4,C) and 51Cr release assay (Fig. 4 D).
As described, type 1 T cell stimulatory cytokines such as IL-12 are important in functional T cell priming. To determine whether the reduced CD8+ T cell priming efficiency of MUTZ-3 LC might be related to the inability of LC to produce type 1 stimulatory cytokines, IL-12p70 was introduced into MUTZ-3 cells by retroviral transduction, and IDC and LC derived from these MUTZ-3-IL-12 cells were analyzed for their CD8+ T cell priming capacity. Il-12p70 production by MUTZ-3-IL-12 IDC and LC was confirmed and ranged from 1.0 to 1.7 ng/ml per 200,000 cells per 24 h (data not shown). As shown in Fig. 5,A, CD8+ T cell priming efficiency of MUTZ-3-IL-12 IDC and LC was significantly increased compared with MUTZ-3 IDC and LC, as indicated by the generation of enhanced levels of MART-126L-specific CD8+ T cells. Moreover, introducing IL-12p70 also enhanced the functional avidity and tumor cell recognition abilities of the generated CD8+ T cells. Whereas MART-126L-specific CD8+ T cells primed with MUTZ-3 IDC and MUTZ-3 LC exhibited only intermediate functional avidity (Fig. 5,B; open symbols), MUTZ-3-IL-12 IDC- and LC-generated MART-126L-specific CD8+ T cells exhibited high functional avidity (Fig. 5,B, closed symbols). The increased functional avidity of the MUTZ-3-IL-12 IDC-generated and MUTZ-3-IL-12 LC-generated CD8+ T cells resulted in an increased capacity to recognize endogenously expressed MART-126 on tumor cells as measured by intracellular IFN-γ (Fig. 5,C) and 51Cr release assay (Fig. 5,D). Indeed, MUTZ-3-IL-12 IDC-generated (data not shown) and MUTZ-3-IL-12 LC-generated MART-126L-specific CD8+ T cells (Fig. 5,D; right) were both able to specifically recognize and kill the HLA-A2+/MART-1+ tumor cell line Mel-AKR, whereas MUTZ-3 IDC-generated (data not shown) and MUTZ-3 LC-generated MART-126L-specific CD8+ T cells (Fig. 5 D; left) were not capable. In addition, MART-126L-specific CD8+ T cells induced with MUTZ-3-IL-12 IDC and MUTZ-3-IL-12 LC also showed a reduced expression of CD27, CD28, CD62 ligand, and CCR7 compared with MUTZ-3 IDC- or MUTZ-3 LC-stimulated MART-126L-specific CD8+ T cells, which is indicative of a more advanced effector T cell differentiation (data not shown).
Analysis of T cell stimulatory properties of human skin-derived DDC and LC
To verify our findings in a more physiological setting, DDC and LC derived from healthy human skin were analyzed for the expression of costimulatory and adhesion molecules, for their allogeneic and tumor Ag-specific stimulatory capacity, and for their LN migratory potential. DDC and LC were obtained from human skin samples by active migration from dermal and epidermal sheets. As we previously described (36) and in our recent observations (S. Santegoets, R. Scheper, and T. de Gruijl, submitted for publication), LC migrating from epidermis represent “true” LC as demonstrated by the expression of Langerin and typical high levels of CD1a, whereas DDC migrated from dermis represent “true” DDC, displaying only intermediate levels of CD1a and no Langerin at the cell membrane and expressing DC-SIGN mRNA. Purity of CD1a+ skin-emigrated DDC and LC in migrated skin cells was between 15 and 45% (data not shown). Furthermore, skin-emigrated DDC and LC displayed a mature phenotype, expressing intermediate-to-high levels of costimulatory and adhesion molecules, as well as the maturation marker CD83 (Fig. 6,A), and exhibited allogeneic T cell stimulatory capabilities, as demonstrated by allogeneic MLR (Fig. 6 B). Of note, different from MUTZ-3-derived IDC and LC, no difference in activation status and allogeneic T cell stimulation was observed between the two skin-derived DC subsets.
Skin DDC and LC were also analyzed for their migratory capacity toward LN-homing chemokines. As demonstrated in Fig. 6,C, mature skin-emigrated DDC were able to migrate toward MIP-3β, whereas migration of mature skin-emigrated LC toward this chemokine was only low. Interestingly, although skin-emigrated DDC and LC were able to migrate toward 6Ckine, percentages were rather low. Of note, moderate to low expression levels of CCR7 were detected by flow cytometry, as shown in Fig. 6 A.
Finally, tumor Ag-specific CTL induction capabilities of both skin DDC and LC were also analyzed. To this end, skin DDC and LC were used as stimulator cells in our in vitro CTL induction protocol. As demonstrated in Fig. 6,D, both skin-emigrated DDC and LC were capable of inducing MART-126L-specific CD8+ T cells in vitro. Similar as for MUTZ-3-derived IDC and LC (Fig. 4,A), significantly higher levels of MART-126L-specific CD8+ T cells could be detected when stimulated with skin DDC in two of three donors analyzed (Fig. 6,D), indicating superior CD8+ T cell priming efficiency of skin DDC compared with skin LC. Functional avidity analysis revealed no significant difference in functional activity between the skin DDC- and LC-generated MART-126L-specific CD8+ T cells (Fig. 6 E).
In summary, the relative CD8+ T cell induction capacities of skin-derived primary DDC and LC accurately reflect our findings for MUTZ-3 IDC and MUTZ-3 LC, further validating the MUTZ-3 as a human DDC- and LC-equivalent cell line model and demonstrating that both DC subsets can serve as valid vaccine vehicles for tumor immunotherapy.
Because of their critical role in orchestrating the immune response, DC are increasingly applied as vaccines for the treatment of cancer (17). So far, in the majority of clinical trials, MoDC have been used for vaccination. However, it has been described recently that CD34-derived LC exhibit superior T cell priming capacity over CD34-derived IDC and MoDC in vitro, suggesting a potential benefit of the use of LC in clinical DC vaccination. In this study, extensive characterization of the functional differences between IDC and LC was performed by making use of the MUTZ-3 cell line as a model system. We show in this study that both DC subsets are capable of inducing antitumor T cell immunity and provide valid vaccine vehicles for tumor immunotherapy, albeit that provision of IL-12p70 is required to ensure optimal T cell stimulation.
As previously described, using allogeneic MUTZ-3 IDC as stimulator cells in an in vitro induction protocol, functional tumor-specific CTL could be generated (22). In this study, by using the same method, we confirmed previous findings of the applicability of allogeneic HLA-A2-matched MUTZ-3 IDC. Although mature MUTZ-3 LC showed significantly higher allogeneic T cell stimulatory capacity, MUTZ-3 IDC supported the induction of tumor Ag-specific CD8+ T cells at an overall higher efficiency as compared with LC. Similarly, skin DDC also displayed superior CTL priming capacity over skin LC. This display is not in line with findings from other groups, reporting superior CTL priming capacity of CD34-derived LC over CD34-derived IDC or MoDC (15, 19). Whereas for MUTZ-3 IDC and LC this discrepancy could be explained by the observed enhanced allogeneic T cell stimulatory capacity of MUTZ-3 LC as compared with MUTZ-3 IDC (Fig. 1, B and C), possibly resulting in preferential outgrowth of alloreactive T cells, our data clearly show that this cannot explain the observed differences in CTL priming efficiency by primary skin-derived DDC and LC, which showed equivalent allostimulatory capacities (Fig. 6 B). Furthermore, comparing the induction efficiency of autologous MoDC and allogeneic MUTZ-3 IDC (22), we previously did not find evidence for allogeneic response interfering with the Ag-specific CTL response.
The superior tumor-Ag-specific CD8+ T cell induction of IDC might also be related to the observed inability of LC to release T cell stimulatory cytokines such as IL-12p70, IL-15, and IL-23, as also described for CD34-derived and primary skin-derived LC (19, 37, 38). Indeed, although no differences were observed in the cytokine expression profile of IDC and LC primed T cells, transduction with IL-12p70 significantly improved the priming efficiency of both IDC and LC and abrogated the difference in tumor-specific CD8+ priming efficiency between the DC subsets. Moreover, introducing IL-12p70 also resulted in a more advanced effector T cell differentiation with enhanced functional avidity and tumor cell recognition abilities of the IDC- and LC-generated CD8+ T cells, as we also recently described using mRNA-transfected MoDC (39).
The low level or absence of proinflammatory cytokine production by MUTZ-3 IDC and LC as compared with MoDC might be explained by the leukemic origin of the MUTZ-3 cell line because AML blast-derived DC are described to produce only low amounts of IL-12 (40). However, the observation that MUTZ-3 IDC and MUTZ-3 LC do not produce any IL-10 with the maturation mixes used in this study, argues against a possible immunosuppressive phenotype. Moreover, the low level or absence of IL-12p70 production is consistent with recently described findings that skin-migrated IDC and LC, as well as CD34-derived LC, also do not produce IL-12p70 (19, 41, 42).
A pivotal role of fully mature, immunostimulatory DC is the ability of DC to migrate to the draining LN. Upon maturation, CCR7 expression was induced on MUTZ-3 IDC and MUTZ-3 LC, and this expression was accompanied by responsiveness to LN-homing chemokines CCL19 and CCL21. Interestingly, and unlike migration of MoDC (30, 31), migration of MUTZ-3 IDC and MUTZ-3 LC was independent (IDC) or less dependent (LC) on PGE2. Importantly, so far it is not clear whether CD34-derived IDC or LC migration is dependent on PGE2. Of note, PGE2 dependence for MoDC migration could be confirmed in our experiments, thereby validating our migration assay. Furthermore, the particularly high migration of MUTZ-3 IDC in response to CCL19 and CCL21 upon maturation with the IFN-α/poly(I:C) type 1 mixture also underlines this PGE2 independence (26). Because PGE2 has been described as an inflammatory mediator with a Th2 driving role (43, 44), the ability of MUTZ-3 IDC and MUTZ-3 LC to migrate independent of PGE2 toward LN-homing chemokines might be beneficiary in clinical DC vaccination setting. Furthermore, the relatively low migration of human skin-emigrated DDC and LC in response to CCL19 and CCL21 may be explained by the absence of exogenously added PGE2 both during maturation induction (over 48 h of culture of the epidermal and dermal sheets in the presence of GM-CSF and IL-4) and subsequently during migration in the Transwell setup.
Taken together, comparative functional analysis of MUTZ-3-derived IDC and LC and primary skin-derived DDC and LC revealed that, except for the inability of LC to release distinct type 1 T cell stimulatory cytokines, both DC subsets exhibit functional properties that are essential for the in vivo generation of CTL-mediated immunity, further validating the MUTZ-3 as a human DDC- or LC-equivalent cell line model and demonstrating that both DC subsets can serve as valid vaccine vehicles for tumor immunotherapy.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Abbreviations used in this paper used: DC, dendritic cell; IDC, interstitial DC; MoDC, monocyte-derived DC; DDC, dermal DC; LC, Langerhans cell; CEA, carcinoembryonic Ag; PSA, prostate-specific Ag; poly(I:C), polyinosinic-polycytidylic acid; NGFR, nerve growth factor receptor; LN, lymph node.