Plasmacytoid dendritic cells (pDC) are capable of producing high levels of type I IFNs upon viral stimulation, and play a central role in modulating innate and adaptive immunity against viral infections. Whereas many studies have assessed myeloid dendritic cells (mDC) in the induction of antitumor immune responses, the role of pDC in antitumor immunity has not been addressed. Moreover, the interaction of pDC with other dendritic cell subsets has not been evaluated. In this study, we analyzed the capacity of pDC in stimulating an Ag-specific T cell response. Immunization of mice with Ag-pulsed, activated pDC significantly augmented Ag-specific CD8+ CTL responses, and protected mice from a subsequent tumor challenge. Immunization with a mixture of activated pDC plus mDC resulted in increased levels of Ag-specific CD8+ T cells and an enhanced antitumor response compared with immunization with either dendritic cell subset alone. Synergy between pDC and mDC in their ability to activate T cells was dependent on MHC I expression by mDC, but not pDC, suggesting that pDC enhanced the ability of mDC to present Ag to T cells. Our results demonstrate that pDC and mDC can interact synergistically to induce an Ag-specific antitumor immune response in vivo.
As the major producer of type I IFNs, plasmacytoid dendritic cells (pDC)4 represent key effector cells in both innate and adaptive immunity (1, 2, 3, 4, 5). Upon invasion by pathogens, pDC promptly produce large amounts of type I IFN, which can activate a variety of immune cells such as B cells, NK cells, and macrophages (1, 4, 6, 7, 8). Although most studies have focused on the role of type I IFN in antiviral immunity, several lines of evidence suggest that pDC are also involved in tumor immunity, as evidenced by identification of pDC in several human and murine cancers, including melanoma, head and neck cancer, ovarian cancer, breast cancer, and acute myelogenous leukemia (9, 10, 11, 12, 13, 14). However, there is little direct evidence to show whether pDC are capable of inducing antitumor immune responses in vivo.
Moreover, it has been suggested that dendritic cell (DC) subsets may interact in the development of antiviral immunity and autoimmune disease (15, 16). For example, pDC have been implicated in myeloid DC (mDC) differentiation in systemic lupus erythematosus patients (15). In addition, pDC may interact with lymph node DC in the generation of anti-HSV CTL (16). However, the direct interaction between purified pDC and mDC subsets has not been clearly characterized, and this interplay has not been studied in the development of antitumor immunity.
In this study, we characterize the ability of pDC to stimulate an Ag-specific immune response in a murine tumor model. We evaluated the ability of peptide-pulsed pDC to induce Ag-specific T cell responses and protect against tumor challenge, and analyzed the interactions between pDC and mDC in their ability to activate adaptive immunity. We demonstrate that not only can pDC directly induce an Ag-specific CD8+ T cell immune response, but they can also enhance the priming ability of mDC in a cell-to-cell contact-dependent fashion. Our results suggest that interactions between DC subsets may play an important role in the generation of effective T cell immune responses against tumor.
Materials and Methods
Female C57BL/6 and β2-microglobulin-targeted mutation (MHC I-deficient) mice, both on C57BL/6 background, were purchased from National Cancer Institute and The Jackson Laboratory, respectively, and maintained in a pathogen-free animal facility at the M.D. Anderson Cancer Center. Mice were used at 8–10 wk of age, and all animal work was performed according to National Institutes of Health and institutional guidelines.
Peptides, tetramer, cell lines, and reagents
The H-2b-restricted CTL epitope from OVA257–264 (SIINFEKL) was used as the immunogen. The H-2b-restricted epitope of the influenza nucleoprotein (NP)366–374 (ASNENMETM) was used as an irrelevant control peptide. All of the peptides were made by standard 9-fluorenylmethoxycarbonyl chemistry and purified by reverse-phase HPLC (Peptides International). OVA257–264-H-2-Kb tetramer was purchased from Beckman Coulter. The EL-4 and E.G7 thymoma were purchased from the American Type Culture Collection and maintained according to manufacturer’s instructions. Neutralizing Ab against mouse IFN-α and mouse rIFN-α were purchased from PBL Biomedical Laboratories. Neutralizing Abs against mouse TNF-α and IL-12 were purchased from R&D Systems. All other Abs were obtained from BD Biosciences.
The expression vectors encoding a full-length murine Flt3 ligand (Flt3L) cDNA designated as pORF-mFlt3L, or the control vector designated as pORF-mcs, were purchased from InvivoGen. Injection of plasmid DNA encoding Flt3L or control vector pORF-mcs was performed using the hydrodynamics-based gene delivery technique, as described before (17, 18). The expression level of Flt3L in mouse serum was confirmed by ELISA (R&D Systems).
DC preparation and activation
Mice were injected with a single dose of 10 μg of plasmid DNA encoding Flt3L or control vector on day 0. Ten days later, mice were sacrificed and bone marrow cells were isolated from femurs and tibiae. Bone marrow cells were incubated with rat anti-CD16/32 mAbs (2.4G2) to block nonspecific binding, and then stained with the following mAbs: anti-CD11c PE, anti-B220 PerCP-Cy5.5, and anti-CD11b allophycocyanin (all from BD Biosciences). Cells were then sorted into mDC (CD11chighCD11b+B220−) and pDC (CD11cintCD11b−B220+) populations using BD FACSAria, and the cell purity was >98%. Purified pDC or mDC were activated by incubation with 10 μg/ml CpG oligodeoxynucleotide (CpG1826: TCCATGACGTTCCTGACGTT, Operon Biotechnologies, or CpG2216: GGG GGA CGA TCG TCG GGG GG; Invitrogen Life Technologies) for 4 h or overnight. The expression levels of costimulatory and MHC II molecules were analyzed using FACSCalibur. Cytokine production was determined in culture supernatant after overnight stimulation by either ELISA kits (IFN-α, PBL Biomedical Laboratories; IL-12, R&D Systems) or mouse multiple cytokines Luminex assay (Linco Research). Samples were analyzed in duplicate according to manufacturer’s instruction.
Analysis of Ag-specific immune response
PDC alone, mDC alone, or a mixture of pDC and mDC (at a 1:1 ratio) was pulsed for 4 h at 37°C with 10 μM appropriate peptides in OPTI-MEM medium (Invitrogen Life Technologies) in the presence or absence of CpG1826. DC were then washed three times with PBS and used for immunization. In some experiments, DC were isolated from MHC I-deficient mice. Otherwise, without indication, DCs were isolated from wild-type C57BL/6 mice. Mice (three mice/group) received 4–5 × 105 DC by s.c. injection (when a combination of pDC and mDC was administered, the total cell number was maintained to be equivalent to those groups receiving either subset alone). PBL were isolated from the tail vein at indicated time after immunization, and erythrocytes were depleted using ACK lysing buffer. Cells were incubated with rat anti-CD16/32 mAbs (2.4G2) to block nonspecific binding, and then stained with anti-CD8α and OVA257–264-H-2-Kb tetramer. Cells were then washed and analyzed using FACSCalibur. For ELISPOT assay, 2 × 105 PBL isolated 7 days after immunization were plated in a 96-well plate in duplicate or triplicate in the presence or absence of 1 μM OVA257–264 peptide for 24 h. Enumeration of Ag-specific mouse IFN-γ-secreting cells was performed according to the manufacturer’s instructions (Cellular Technology).
Tumor immunity study
Mice were immunized as above with pDC, mDC, or a combination of pDC and mDC. Seven days later, immunized mice were challenged s.c. with 3 × 106 E.G7 tumor cells or the parental, non-OVA-expressing tumor cells EL-4. In some experiments, mice (five to eight mice per group) were first inoculated s.c. with 3 × 106 E.G7 tumor cells, and then treated with 4–5 × 105 DC via s.c. injection on day 4 when tumor was measurable. Tumor growth was monitored by measuring the perpendicular diameters of tumors. Mice were sacrificed when tumor size reached 20 mm in diameter. All of the experiments were conducted in a blinded, randomized fashion.
PDC were first stimulated with or without CpG1826 for 4 h, then washed three times with PBS. MDC were then cocultured for 16 h with nonactivated or activated pDC in the presence or absence of a series of blocking Abs, as indicated, or plated in a 0.4-μm Transwell with mDC in the lower chamber and activated pDC in the upper chamber. The expression of costimulatory molecules and intracellular cytokines by mDC was analyzed by flow cytometry. For in vivo studies, mice were immunized s.c. with 5 × 105 CpG-activated, OVA peptide-pulsed, mDC, pDC, or a combination of mDC plus pDC that were either activated together or separated in a Transwell. Seven days after immunization, PBL were isolated and OVA-specific cells were analyzed by tetramer staining, as described above.
The statistical analyses to compare tumor size and mouse survival rate were determined using Mann-Whitney nonparametric U test and Kaplan-Meier test, respectively.
Characterization of in vivo-generated murine pDC
The relatively low frequency of pDC in vivo has hampered the study of this cell type. To date, most studies characterizing the phenotype and function of murine pDC have used in vitro-generated pDC that required a lengthy culture of bone marrow cells in the presence of cytokines and growth factors (19, 20). However, these in vitro bone marrow-derived pDC often displayed poor viability, and yet, may not represent normal physiology due to the extensive in vitro manipulation (21, 22). In this study, we developed an in vivo system to easily generate large numbers of pDC by hydrodynamic injection of plasmid DNA encoding murine Flt3L. As shown in Fig. 1,A, a single injection of Flt3L DNA resulted in an extensive expansion of both pDC and mDC in bone marrow, defined as CD11cintCD11b−B220+ for pDC and CD11chighCD11b+B220− for mDC, as compared with DC subsets isolated from control mice. The absolute number of pDC and mDC in bone marrow was significantly increased ∼10-fold after Flt3L DNA injection (Fig. 1 B). A more profound increase in the percentage and total number of pDC and mDC was also observed in mouse spleens (data not shown).
It has been shown that freshly isolated murine pDC display an immature DC phenotype expressing very low levels of MHC and almost negligible levels of costimulatory molecules, and therefore, are poor T cell stimulators (20, 21, 23). However, following in vitro stimulation with virus, CD40L, or CpG, pDC rapidly up-regulate the expression of MHC and costimulatory molecules, and develop the ability to induce T cell proliferation, drive Th1 polarization, and produce type I IFN (21, 24, 25). To characterize the phenotype and function of in vivo Flt3L-generated pDC, we isolated highly purified pDC (>98% purity) from the bone marrow of Flt3L-treated mice by cell sorting. Consistent with previous studies, freshly isolated, purified pDC displayed undetectable level of CD80, CD86, and CD40, and very low levels of MHC II (20, 21, 22, 23, 24, 25). However, upon stimulation with CpG, the expression levels of all of these molecules were significantly up-regulated, especially CD40, CD86, and MHC II (Fig. 1,C). In addition, CpG stimulation induced pDC to secrete IFN-α, IL-12, TNF-α, and IL-6, further confirming the identity of pDC (Fig. 1 D). By contrast, when stimulated with CpG under the same conditions, mDC produced high levels of IL-12, TNF-α, and IL-6, but no detectable IFN-α production. As a comparison, nonstimulated pDC or mDC secreted little, if any, cytokines. In addition, in vivo Flt3L-generated pDC displayed enhanced viability, survival, and costimulatory molecule expression compared with classic in vitro-generated pDC (data not shown).
Immunization with activated, peptide-pulsed pDC induces Ag-specific CTL responses against tumor
Although mature pDC have been demonstrated to be capable of stimulating T cells in vitro, the ability of pDC to stimulate naive, nontransgenic T cells in vivo remains unclear (19, 20, 21, 22, 25, 26, 27, 28). To investigate this question, naive mice were immunized s.c. with CpG-activated pDC or mDC pulsed with either MHC I-restricted OVA257–264 or control NP peptides. As shown in Fig. 2,A, CpG-activated pDC elicited an Ag-specific CTL response at a similar level as mDC, whereas immunization with nonactivated pDC or control peptide-pulsed, CpG-activated pDC generated little Ag-specific T cells. Consistent with the tetramer staining, ELISPOT assay confirmed that the number of IFN-γ-secreting PBL was significantly increased in mice immunized with CpG-activated pDC compared with mice immunized with nonactivated pDC or activated pDC pulsed with control NP peptides (Fig. 2 B). These results demonstrate that CpG activation endows pDC with the capability of priming Ag-specific CD8+ T cell responses in vivo.
We next sought to determine whether the induction of an Ag-specific immune response by pDC could protect mice from subsequent tumor challenge. Naive mice were first immunized with CpG-activated or nonactivated pDC or mDC pulsed with OVA or NP peptides. Seven days later, the immunized mice were challenged with a lethal dose of E.G7 tumor cells. As shown in Fig. 2,C, mice receiving CpG-activated, OVA peptide-pulsed pDC or mDC were protected from tumor challenge as compared with mice immunized with nonactivated pDC or mDC pulsed with OVA peptide, or mice receiving CpG-activated pDC pulsed with NP peptide (p < 0.01). Four of five mice in both CpG-activated, OVA peptide-pulsed pDC and mDC groups remained tumor free. To determine whether immunization with CpG-activated, peptide-pulsed pDC could generate tumor Ag-specific memory T cells, mice immunized with activated pDC that had rejected an initial tumor challenge were rechallenged with either E.G7 tumor cells or the parental, non-OVA-expressing tumor cells EL-4. Four of five pDC-immunized mice were completely protected from E.G7 rechallenge (Fig. 2 D). By contrast, all five mice rechallenged with EL-4 succumbed to tumor growth at an early stage and eventually died of a significant tumor burden. Furthermore, the antitumor activity was mainly mediated by Ag-specific CD8+ T cells because the antitumor activity was completely abolished in CD8 knockout mice, but not in CD4 knockout mice (data not shown). Taken together, these results demonstrate that immunization with CpG-activated, peptide-pulsed pDC can generate long-lasting Ag-specific memory CD8+ T cell responses against tumor in vivo.
pDC synergize with mDC in the induction of Ag-specific CTL responses in vivo
Upon pathogen stimulation, pDC produce varied cytokines and proinflammatory chemokines, which in turn profoundly impact surrounding immune cells both positively and negatively (3, 29). For example, murine CMV-stimulated pDC have been reported to induce CD8α+ DC maturation in vivo and to promote CD8+ T cell IFN-γ production through type I IFN (30, 31, 32). However, the inhibitory effect of pDC on synthesis of IL-12 from mDC has also been reported (33). In light of these complex interactions between pDC and other immune cells, we next evaluated the interplay between pDC and mDC on the induction of an antitumor response. Mice were immunized with either CpG-activated, OVA peptide-pulsed pDC or mDC alone, or in combination, using the same total cell number for each group. Similar to the results shown in Fig. 2,B, immunization with CpG-activated, OVA peptide-pulsed pDC or mDC alone led to Ag-specific immune responses, evidenced by significantly increased IFN-γ-secreting cells in PBL upon Ag exposure (Fig. 3,A). However, when mice were immunized with a mixture of pDC and mDC (1:1) that were activated by CpG concurrently, the Ag-specific CTL response was dramatically increased compared with mice receiving pDC or mDC alone. Importantly, coincubation of pDC and mDC during activation before in vivo administration was crucial for the observed enhancement because immunization with a mixture of pDC and mDC that were activated separately did not result in enhanced levels of an Ag-specific T cell response (data not shown). Thus, these results suggest that immunization with a mixture of concurrently CpG-activated, peptide-pulsed pDC and mDC synergistically enhances Ag-specific immune responses in vivo. Furthermore, the responses induced by immunization of mice with pDC or mDC alone reached a peak on day 5 after primary immunization and then decreased gradually and returned to the baseline 2 wk after immunization (Fig. 3 B). By contrast, immunization of mice with mDC and pDC that were activated together in vitro elicited a stronger response with the enhanced CD8+ CTL peaked on day 8 after immunization.
pDC enhance the Ag-priming ability of mDC
Although our results suggest that a synergistic effect exists between pDC and mDC in the induction of an Ag-specific T cell response, the underlying mechanism is unclear. Theoretically, this synergy could be due to improved T cell stimulatory activity by both pDC and mDC or from one dominant DC subset presenting Ag to T cells with the other subset providing an adjuvant signal. To elucidate the mechanism, we next repeated the synergy experiment using combinations of pDC and mDC from wild-type and/or MHC I-deficient mice. As shown in Fig. 4, immunization with a mixture of pDC from MHC I-deficient mice plus mDC from wild-type mice induced a similarly high level of immune response to that induced by a mixture of wild-type pDC plus wild-type mDC. However, immunization with a mixture of pDC from wild-type mice plus mDC from MHC I-deficient mice resulted in a markedly decreased immune response. As expected, no response was observed in mice immunized with NP peptide-pulsed pDC plus mDC (Fig. 4). Thus, these data suggest that the synergy observed between the DC subsets in their capacity to activate T cells is mainly dependent on direct Ag presentation by mDC, but not pDC. PDC appear to enhance the ability of mDC to present Ag to T cells.
Cell-to-cell contact is important in the synergy between pDC and mDC
The critical role of coincubation of pDC and mDC during activation before in vivo administration suggested that direct cell-to-cell contact and/or soluble factors secreted by DCs during coincubation may play an important role in the observed synergistic effect. To study this mechanism, we cocultured pDC with mDC together or separated in Transwell plates in the presence or absence of a series of blocking Abs against cytokines produced by DC. As shown in Fig. 5,A, coculture with CpG-activated pDC, but not nonactivated pDC, significantly enhanced the expression of costimulatory molecules CD40, CD86, and CD80 on mDC, as well as secretion of cytokines IL-12 and TNF-α by mDC (Fig. 5,A and data not shown). Surprisingly, activation of mDC by pDC was mainly dependent on cell-to-cell contact, as mDC activation was completely diminished when mDC and pDC were separated by a Transwell. Cytokines IFN-α, IL-12, and TNF-α did not play a role in pDC-mediated mDC activation because the addition of saturating amounts of neutralizing Abs specific for these cytokines had no effects on the expression of costimulatory molecules and cytokines. Moreover, incubation of mDC with IFN-α at a variety of concentrations did not result in significant mDC activation. To confirm these results in vivo, we immunized mice with pDC and mDC that were either cocultured together or separated in a Transwell during activation. Again, cell-to-cell contact before infusion was essential to acquire synergy between pDC and mDC in the generation of enhanced Ag-specific T cell responses as determined by tetramer staining (Fig. 5 B). Thus, these results suggest that physical contact between pDC and mDC plays a major role in the observed synergy between these two DC subsets.
Immunization with a combination of pDC plus mDC improves therapeutic efficacy against tumor
Our observation that immunization with CpG-activated, peptide-pulsed pDC plus mDC led to an enhanced Ag-specific immune response against tumor prompted us to further study the therapeutic potential in established tumor. Mice were first s.c. inoculated with E.G7 tumor cells and then immunized with different DC regimens when the tumor was palpable. As shown in two representative experiments, Fig. 6, A and C, treatment of mice bearing s.c. E.G7 tumor with a single injection of CpG-activated, OVA peptide-pulsed pDC or mDC resulted in inhibition of tumor growth compared with no treatment. Significantly, treatment with a mixture of activated pDC plus mDC resulted in an enhanced antitumor response compared with either DC subset alone (p < 0.05). By contrast, treatment with a mixture of activated pDC plus mDC, pulsed with control peptide, failed to inhibit tumor growth, demonstrating that the therapeutic effect induced by pDC plus mDC was Ag specific. Furthermore, this improved therapeutic efficacy against tumor by a mixture of pDC plus mDC also correlated well with improved survival (Fig. 6, B and D; p < 0.05). Taken together, these results suggest that immunization with a combination of pDC and mDC, activated by CpG concurrently, not only induces a stronger Ag-specific immune response, but also can lead to better therapeutic efficacy.
Although the participation of pDC in immune responses against viral infection through type I IFN production has been well established, the capability of pDC to present tumor Ag to T cells and the interactions between pDC and other DC subsets in the induction of antitumor immune responses have been unclear. In this study, we demonstrate that not only can pDC directly induce an Ag-specific, antitumor immune response, but they can also enhance the ability of mDC to stimulate a CD8+ T cell response, resulting in an improved antitumor effect.
Whether or not pDC can act as a professional APC to prime T cell responses has become an important question regarding pDC function (34). Although pDC have been categorized as an APC ever since their discovery, the proper study of such a function on murine pDC, especially in a tumor model, is needed. Our data reveal that immunization with CpG-activated, peptide-pulsed pDC is not only capable of priming nontransgenic, naive CD8+ T cells in vivo, but is also capable of protecting mice from an Ag-specific tumor challenge. In these studies, the possibility that other DC subsets are presenting Ag taken from peptide-pulsed pDC to initiate a T cell response is low because immunization with CpG-activated splenocytes pulsed with Ag failed to induce a T cell response in a similar experimental setting (data not shown). Our study also demonstrates that CpG-activated pDC not only can prime a CD8+ T cell response in vivo, but also are able to generate Ag-specific CD8+ memory T cells with recall function against a tumor rechallenge.
Interestingly, immunization of mice with a mixture of CpG-stimulated pDC plus mDC resulted in stronger T cell stimulatory capacity and therapeutic efficacy against tumor than that induced by each DC subset alone, indicating that a synergistic effect exists between these DC subsets in the generation of an immune response. Moreover, the lack of MHC I expression on the mDC subset, but not the pDC subset, significantly impaired the induction of the synergistic effect in generating a CTL response, suggesting that mDC may take the major role in Ag presentation in this combined DC vaccination, even though both mDC and pDC could individually induce Ag-specific T cell responses. However, it is clear that pDC contributed to the immune response because equal numbers of mDC alone induced a much smaller response than that induced by pDC plus mDC. Thus, the enhanced Ag-specific T cell response derived from the mixture of activated pDC and mDC is due to an effect of pDC on mDC, but not from enhanced direct stimulation of T cells by pDC.
Although IFN-α production by pDC contributes to its effects on a number of immune cells (31, 35), surprisingly, the synergy between pDC and mDC that we observed appeared to be largely due to cell-to-cell contact. Addition of exogenous IFN-α did not cause mDC activation in vitro, and neutralizing Abs against IFN-α did not block pDC-dependent mDC activation. Although there are many type I IFNs that have been described to be produced by pDC, it is unlikely that these other type I IFNs are mediating the pDC-mDC synergy because the separation of mDC and pDC by a Transwell system diminished the ability of pDC to activate mDC in vitro and to enhance T cell stimulatory capacity by mDC in vivo. Recently, several lines of evidence suggest that there is cooperation between different DC subsets in orchestrating adaptive immune responses against infectious diseases (16, 36). Consistent with our findings, physical contact has been suggested to be crucial for pDC to help other DC subsets to induce appropriate adaptive immunity. Engagement of CD2-CD2L and CD40L-CD40 interaction between pDC and lymph node DCs appear to be important mechanisms for pDC to license lymph node DCs to acquire enhanced Ag presentation activity and thereafter to prime anti-HSV CTL immunity (16). Importantly, interaction between CD40 on conventional DC and CD40L on pDC has also been recently suggested to be essential for CpG-induced immune activation against bacterial infection through IL-15-mediated mechanisms (36). Our study uses a tumor model to contribute additional evidence to the emerging concept that there is potential cross-talk between pDC and other DC types in the establishment of an Ag-specific immune response. Although CD40L-CD40 interaction has been implicated in the interplay between pDC and other DC subsets in these infectious disease models (16, 36), we have not found these pathways to explain the synergy in our tumor model (data not shown). Thus, further studies are needed to characterize the receptor-ligand interactions between these DC subsets in our system. This molecular characterization may in turn lead to novel strategies to directly activate mDC in vivo. In addition, it will be important to evaluate interactions between pDC and other cells involved in the innate immune response, most notably NK cells. Although our study uses an OVA system, which models a tumor neoantigen, other classes of tumor Ags, such as self Ags, have been found to be important in the development of antitumor immunity. Therefore, further studies with self Ags such as the melanoma Ag gp100 will be important to provide additional insights in the development of antitumor immunity.
Despite the characterization of numerous tumor Ags recognized by T cells, cancer vaccines using these Ags have resulted in limited success (37, 38). Recent studies have shown that strong innate immunity is required to develop a potent adaptive immune response (39, 40, 41, 42). In addition, increasing evidence suggests that the innate immune system is complex, and maximum activation is a result of interactions between multiple cell types (29, 39, 40, 42, 43, 44). Current cancer vaccines have focused on Ag-loaded mDC, but we show that optimal immunization may occur with the appropriate interplay between multiple cell types. Our study represents clear evidence that peptide-pulsed pDC cannot only induce an Ag-specific T cell response and antitumor activity, but are also able to enhance the ability of mDC to activate T cells, which may lead to new approaches for the immunotherapy of cancer.
We thank Dr. Greg Lizee for helpful review of this manuscript.
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: DC, dendritic cell; pDC, plasmacytoid DC; Flt3L, Flt3 ligand; int, intermediate; mDC, myeloid DC; NP, nucleoprotein.