Recruitment of circulating natural killer (NK) cells into inflamed lymph nodes is known to provide a potent, IFN-γ–dependent boost for Th1-polarized immune responses in mouse models. Such NK cell recruitment into draining lymph nodes is induced by certain s.c. injected adjuvants, including mature vaccine dendritic cells (DC), and is mediated by a CXCR3-dependent pathway. Here, we show that monocyte-derived immature human DCs stimulated with polyinosinic acid:polycytidylic acid, IFN-α, tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and IFN-γ, α-type 1–polarized DC (αDC1), secrete profuse amounts of the CXCR3 ligand CXCL9/MIG and substantial amounts of CXCL10/IP-10 and CXCL11/I-TAC after withdrawal of maturation stimuli. In sharp contrast, no measurable production of these chemokines was found in DCs after maturation with the current gold standard maturation cocktail for human DC-based cancer vaccines consisting of TNF-α, IL-1β, IL-6, and prostaglandin-E2 (PGE2-DC). PGE2-DCs preferentially produced the Th2 and regulatory T-cell–attracting chemokines CCL17/TARC and CCL22/MDC, whereas only marginal levels of these chemokines were produced by αDC1s. Functional studies in vitro showed that supernatants from mature αDC1s actively recruited CD3CD56+ NK cells and that adding anti-CXCL9/MIG antibodies to the αDC1 supernatant substantially reduced this recruitment. Finally, αDC1s were able to induce IFN-γ production when cocultured with resting autologous NK cells, but only if concurrent CD40 ligation was provided. These novel findings indicate that injected human αDC1-based vaccines have the potential to recruit and activate NK cells during their arrival to draining lymph nodes and that this feature may be of relevance for efficient priming of Th1 cells and CTLs. [Cancer Res 2008;68(14):5965–71]

Dendritic cells (DC) play a central role in the initiation and regulation of innate and adaptive immune responses and have increasingly been applied as experimental vaccines for cancer patients (1, 2). On encountering certain pathogen-associated antigens and in response to proinflammatory cytokines, DCs become activated and start to secrete inflammatory mediators and up-regulate costimulatory molecules as well as MHC class I and class II molecules presenting processed antigens (3). Another functional consequence of activation in vivo is the up-regulation of the lymph node–directing chemokine receptor CCR7, enabling them to migrate in response to gradients of first CCL21 and then CCL19 to the T-cell zone of the draining lymph node (46). These unique features of DCs are increasingly exploited for the design of DC-based vaccines in immunotherapy.

Most commonly, DCs are obtained through in vitro differentiation of monocytes in the presence of granulocyte macrophage colony-stimulating factor (GM-CSF) and interleukin (IL-4), followed by exposure to inflammatory signals to induce final maturation (7). Additional signals provided by prostaglandin E2 (PGE2) are crucial for inducing a substantial chemotactic response in vitro to lymph node–derived chemokines (8, 9). DC-based vaccines, matured with a clinical grade, gold standard maturation cocktail consisting of tumor necrosis factor-α (TNF-α), IL-1β, IL-6, and PGE2, show impaired IL-12 production (10, 11) but has been believed to retain the property for priming Th1 CD4+ T cells and CD8+ CTLs as assessed by certain in vitro assays (11). However, recently the presence of IL-12 during CTL priming by DCs in vitro was found to be mandatory to enable direct recognition of tumor cells expressing the relevant MHC class I/peptide complex (12). In line with these findings, IL-12p70 secretion dramatically enhances the ability of DCs to induce tumor-specific CTLs and promote tumor rejection in therapeutic mouse models (13, 14). Taken together, these observations thus provide a reasonable explanation for the difficulty in consistently generating tumor regression in patients treated with antigen-loaded DCs that have been matured with the PGE2-containing DC-maturating protocols.

IFN-γ has been shown to facilitate the production of IL-12p70 by DCs primed by microbial products or inflammatory cytokines such as TNF-α and IL-1β (15), a feature that we recently tried to capitalize on in the generation of clinical grade DCs (16). However, addition of IFN-γ to the standard PGE2-containing maturation protocol inhibits PGE2-induced membrane expression of CCR7 and reduces DC migration toward lymph node chemokines (16). Recently, Mailliard and colleagues (17) reported that the inclusion of IFN-α and polyinosinic acid:polycytidylic acid (p-I:C) to the original IL-12p70–inducing cytokine cocktail, composed of TNF-α, IL-1β, and IFN-γ (15), allows for the generation of clinical grade DCs, α-type 1–polarized DC (αDC1) with high migratory function toward lymph node chemokines combined with a strong ability to produce IL-12p70.

During the nonspecific early phase of systemic infection, IFN-γ is thought to be primarily provided by natural killer (NK) cells and NK T cells in response to early danger signals derived from the invading pathogen (18, 19). Intriguing recent studies suggest that CD8+ T cells, including naïve CD8+ T cells, can also provide IFN-γ during the innate immune response (20, 21). However, recent data from mouse models indicate that NK cells play a major role in early IFN-γ production in lymph nodes, draining sites of infection or immunization (22, 23). The study of Martin-Fontecha and colleagues (22) specifically showed that circulating NK cells were recruited to lymph nodes on stimulation by mature DCs or certain adjuvants in a CXCR3-dependent manner and that these NK cells provide an early source of IFN-γ that was necessary for Th1 polarization. This unique function of certain adjuvants suggests that for vaccines that rely on Th1 responses, adjuvants could be selected according to their capacity to induce the recruitment of NK cells in antigen-stimulated lymph nodes. In light of this important finding, we investigated the ability of two different clinical grade DC-based vaccines to produce CXCR3 ligands and to recruit and activate human NK cells in vitro.

Our data show that αDCs secrete profuse amounts of the CXCR3 ligand CXCL9/MIG as well as substantial amounts of CXCL10/IP-10 and CXCL11/I-TAC after withdrawal of maturation stimuli, thus at a time point when they are ready to use for vaccination. In contrast, no measurable levels of the CXCR3 ligands were produced by DCs after maturation with the current PGE2-containing gold standard cocktail. Functional studies in vitro further show that αDC1 supernatants actively recruit NK cells in a CXCL9/MIG-dependent manner. Finally, αDC1s were able to induce IFN-γ production when cocultured with resting autologous NK cells, but only if CD40 ligation was provided. These novel findings indicate that fully mature and subsequently injected human αDC1-based clinical grade vaccines have the potential to recruit and activate NK cells on arrival at draining lymph nodes and that this feature may be relevant for efficient Th1 priming by DC-based vaccines.

Generation of monocyte-derived immature DCs. Peripheral blood mononuclear cells (PBMC) obtained from healthy blood donors were isolated on density gradients with Lymphoprep (Nycomed). The use of human blood donor cells was approved by the Human Research Ethics Committee at the Sahlgrenska Academy, Göteborg University. Isolated PBMCs were resuspended in AIM-V medium (Invitrogen), plated in 24-well plastic culture plates at 2.5 × 106 per well, and allowed to adhere for 2 h. Nonadherent cells were removed and the remaining adherent monocytes were cultured in AIM-V medium supplemented with recombinant human GM-CSF and IL-4 (R&D Systems; both at 1,000 units/mL) for 5 d, to obtain immature DCs. When indicated, DCs were generated by negative selection for CD14+ cells using magnetic cell separation (Miltenyi Biotec). The isolated CD14+ population had >85% purity as determined by flow cytometry, and contamination with CD56+CD3 NK cells was <1%. Purified monocytes were plated at 3 × 105 per well and differentiated to DCs as described above.

Maturation of DCs. Maturation of immature DCs was induced by supplementing the culture media with the gold standard maturation cocktail consisting of TNF-α (50 ng/mL), IL-1β (25 ng/mL), IL-6 (10 ng/mL; all from R&D Systems), and PGE2 (1 μg/mL; Sigma-Aldrich). Alternatively, DCs were matured by adding IFN-α (3,000 units/mL), IFN-γ (1,000 units/mL), TNF-α (50 ng/mL), IL-1β (25 ng/mL; all from R&D Systems), and p-I:C (20 μg/mL; Sigma-Aldrich) to obtain αDC1s. When indicated, different combinations of the components included in the αDC1 maturation cocktail were added. DCs cultivated without the addition of maturation cocktails for 24 h were used as controls.

Real-time PCR. RNA was prepared from purified DCs matured for 24 h, using the Qiagen Mini Kit. The concentration and quality of the RNA were evaluated with Agilent 2100 Bioanalyzer (Nano LabChip, Agilent Technologies, Dalco Chromtech AB). Reverse transcription of RNA and real-time PCR analysis were done as previously described (24). In brief, 500-ng RNA was transcribed into cDNA using avian myeloblastosis virus reverse transcriptase. The mRNA level of each target gene was quantified by real-time PCR on an ABI Prism 7900 Sequence Detection System (TaqMan, Applied Biosystems) using low-density array microfluidic cards that included the target genes and endogenous controls listed in Table 1. Samples (50 ng) were run in duplicates and the comparative ΔÄCT method of relative quantification was used to calculate the differences in gene expression between the control and the two treatment groups. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), β-actin, and hypoxanthine phosphoribosyl transferase–encoding gene (HPRT) were used as endogenous controls (to correct for variation in sample loading and efficiency of the amplification reaction).

Table 1.

Target genes analyzed by real-time PCR

GeneAccession no.*
CCR7 Hs00171054_ml 
CD40 Hs00386848_ml 
CD86 Hs00199349_ml 
CD83 Hs00188486_ml 
IL12A Hs00168405_ml 
IL12B Hs00233688_ml 
CXCL9 Hs00171065_ml 
CXCL10 Hs00171042_ml 
CXCL11 Hs00171138_ml 
CCL17 Hs00171074_ml 
CCL22 Hs00171080_ml 
GAPDH Hs99999905_ml 
β-Actin Hs99999903_ml 
HPRT1 Hs99999909_ml 
GeneAccession no.*
CCR7 Hs00171054_ml 
CD40 Hs00386848_ml 
CD86 Hs00199349_ml 
CD83 Hs00188486_ml 
IL12A Hs00168405_ml 
IL12B Hs00233688_ml 
CXCL9 Hs00171065_ml 
CXCL10 Hs00171042_ml 
CXCL11 Hs00171138_ml 
CCL17 Hs00171074_ml 
CCL22 Hs00171080_ml 
GAPDH Hs99999905_ml 
β-Actin Hs99999903_ml 
HPRT1 Hs99999909_ml 
*

Accession numbers refer to primer/probe sets on low-density array cards.

Chemokine determination. For measurement of chemokines, 24-h culture supernatants from previously washed mature DCs were collected and stored at −70°C until chemokine concentrations were determined by specific ELISAs. DCs used for this experiment were derived either from monocytes isolated by plastic adherence or by an additional purification step of CD14+ cells through magnetic cell separation. The commercially available ELISAs for CXCL9/MIG, CXCL10/IP-10, CXCL11/I-TAC, CCL17/TARC, and CCL22/MDC (R&D Systems) were done according to the manufacturer's instructions. To gain insight into the kinetics of the chemokine production, supernatants were collected after 12 h and cells were washed and replaced for a further 12 h before the final supernatant was collected. The lower limits of detection were 125 pg/mL for CXCL9/MIG, 62.5 pg/mL for CXCL10/IP-10, 15.6 pg/mL for CXCL11/I-TAC, 15.6 pg/mL for CCL17/TARC, and 15.6 pg/mL for CCL22/MDC.

NK cell isolation and migration assay. NK cells were purified with the NK cell isolation kit (Miltenyi Biotec) by negative immunomagnetic cell separation. Resulting NK cell populations were >95% CD56+CD3. Purified DCs matured with either the αDC1 or PGE2-DC maturation cocktail for 24 h were washed twice and replaced in fresh AIM-V medium. After 24 h, supernatants were collected and stored in −70°C for future use in migration assays. Chemotaxis of NK cells toward chemokines produced by mature DCs was tested by a transwell assay. Briefly, lower chambers of transwell plates (5.0 μm pore size, BD Biosciences) were filled with 500 μL of supernatant. Five hundred microliters of medium only were used as a control. Anti-CXCL9 or corresponding isotype (rabbit IgG) control antibody (Peprotech; both at 10 μg/mL) was added as indicated. Approximately 1 × 105 purified NK cells or 1 × 106 PBMCs were added in 200 μL of AIM-V medium in the upper chamber, and cells were incubated for 90 min. Cells that migrated to the lower chamber were harvested and stained with phycoerytrin-conjugated anti-CD56 and allophycocyanin-conjugated anti-CD3 (both from BD Biosciences). CD3CD56+ NK cells were subsequently defined and counted by flow cytometry. Specific migration was calculated as a percentage of control (spontaneously migrated cells, migrating to medium only).

Assessment of NK cell activation by flow cytometry. To evaluate DC-mediated IFN-γ production and CD69 expression by NK cells, nonadherent PBMCs were cultured with autologous or allogeneic monocyte-derived DCs (mononuclear cell/DC ratio, 5:1) in AIM-V medium supplemented with 5% autologous plasma in the presence of 10 μg/mL brefeldin A (GolgiPlug, Sigma). When indicated, cocultures of purified NK cells and purified DCs were done at a 1:1 DC/NK cell ratio. DCs used in this experiment were stimulated with either the gold standard or the αDC1 maturation cocktail for 24 h, washed twice, and replaced in the well and then cultured in fresh medium for a further 24 h before being harvested and added to the mononuclear cells. After 6 h of coculture, cells were harvested, fixed, permeabilized with the Cytofix/Cytoperm reagent (BD Biosciences), and stained with phycoerytrin-conjugated anti-CD56, allophycocyanin-conjugated anti-CD3, peridin chlorophyll protein–conjugated anti-CD69, and FITC-conjugated anti-IFN-γ (all from BD Biosciences). Cells were incubated at 4°C for 20 min and washed. All flow cytometry data were acquired on a FACSCalibur cytometer (BD Biosciences) with CellQuest software (BD Biosciences) and analyzed with Diva software (BD Biosciences).

IFN-γ production by NK cells. Purified DCs were stimulated with the αDC1 maturation cocktail for 24 h, washed twice, and replaced in the well and then cultured in fresh medium for a further 24 h before being harvested and plated in 96-well round-bottomed plates at 2 × 104 per well; then cells were stimulated with soluble, histidine-tagged, CD40 ligand (CD40L) protein (R&D systems; 200 ng/mL), followed by the addition of an anti-polyhistidine monoclonal antibody (R&D Systems; 4 μg/mL) 20 min later in the presence or absence of purified autologous peripheral blood NK cells (5 × 104). CD40 stimulation was omitted in control cultures. Supernatants were collected after 24 h and tested for the presence of IFN-γ by ELISA (R&D Systems). The lower limit of detection was 62.5 pg/mL.

Statistical analysis. The statistical significance of differences between experimental samples was determined using Student's t test for paired samples. Significance was accepted at the P < 0.05 (*) and P < 0.001 (**) levels.

αDC1s express high mRNA levels of CXCR3 ligands. Initial real-time PCR analysis revealed that mRNA expression of the CXCR3 ligands CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC was highly up-regulated in αDC1s after incubation with the maturation cocktail for 24 hours compared with PGE2-DCs (Fig. 1A). In contrast, levels of the Th2 and regulatory T-cell–attracting chemokines CCL17/TARC and CCL22/MDC were comparable (Fig. 1B). Furthermore, real-time PCR analysis was used to evaluate mRNA expression of different maturation-associated markers in monocyte-derived αDC1s and PGE2-DCs after 24 hours of maturation. Comparable mRNA levels of the maturation-associated CD40, CD80, CD86, CD83, and CCR7 were found (data not shown). In addition, αDCs were characterized by a significantly increased mRNA expression of both IL-12p35/IL-12A and IL-12p40/IL-12B, whereas only IL-12p40 was highly increased in PGE2-DCs (data not shown).

Figure 1.

αDC1s express high mRNA levels of CXCR3 ligands. Purified monocyte-derived immature DCs were matured for 24 h to induce αDC1s (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C) or PGE2-DCs (IL-1β, TNF-α, IL-6, and PGE2). Cells were subsequently washed and harvested for RNA preparation. Real time-PCR was done on cDNA from total RNA. mRNA levels of CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC (A) and CCL17/TARC and CCL22/MDC (B) are shown as fold change relative to the untreated control after normalization to the housekeeping genes GAPDH, β-actin, and HPRT. Columns, mean of four different donors (observe log scale); bars, SD.

Figure 1.

αDC1s express high mRNA levels of CXCR3 ligands. Purified monocyte-derived immature DCs were matured for 24 h to induce αDC1s (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C) or PGE2-DCs (IL-1β, TNF-α, IL-6, and PGE2). Cells were subsequently washed and harvested for RNA preparation. Real time-PCR was done on cDNA from total RNA. mRNA levels of CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC (A) and CCL17/TARC and CCL22/MDC (B) are shown as fold change relative to the untreated control after normalization to the housekeeping genes GAPDH, β-actin, and HPRT. Columns, mean of four different donors (observe log scale); bars, SD.

Close modal

αDC1s produce extremely high levels of CXCL9/MIG in a sustained fashion. To verify if protein production after maturation for 24 hours and onward correlated with mRNA data, standard ELISA was used to measure the amount of chemokine production in 24-hour culture supernatants of previously washed mature DCs. When compared with PGE2-DCs, αDC1s proved superior in the production of the CXCR3 ligands CXCL9/MIG (P = 0.0009), CXCL10/IP-10 (P = 0.0004), and CXCL11/I-TAC (P = 0.04; Fig. 2A). In contrast, PGE2-DCs preferentially produced the Th2 and regulatory T-cell–attracting chemokines CCL17/TARC (P = 0.03) and CCL22/MDC (P = 0.02; refs. 25, 26), whereas only marginal levels of these chemokines were produced by αDC1s (Fig. 2B). To confirm that the differential pattern of chemokine expression induced by maturation factors was resulting from DCs, and not from contaminating lymphocytes, these experiments were repeated with purified DCs isolated by an additional purification step of DC14+ monocytes before plastic adherence. We could clearly show that purified DCs responded to maturation factors in a similar way as those only purified by plastic adherence (Supplementary Fig. S1). Measurement of chemokine production during maturation was further conducted to get insight into the kinetics of chemokine production. The most remarkable finding was the extremely high and sustained production of CXCL9/MIG by αDC1s (Fig. 2C). To delineate factors of importance for CXCR3 ligand production, each component of the αDC1 maturation cocktail was abolished from the cocktail. As seen in Fig. 3, IFN-γ was the most critical factor in the maturation cocktail for the induction of high production of the CXCR3 ligands CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC. Maturation with one single factor in no case induced detectable amounts of CXCL9/MIG, CXCL10/IP-10, or CXCL11/I-TAC in stimulated DCs (data not shown).

Figure 2.

αDC1s produce high levels of CXCL9/MIG in a sustained fashion. Monocyte-derived immature DCs were matured for 24 h to induce αDC1s (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C) or PGE2-DCs (IL-1β, TNF-α, IL-6, and PGE2). Protein levels, as determined by ELISA, of CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC (A) and CCL17/TARC and CCL22/MDC (B) in 24-h culture supernatants of previously washed mature DCs. Immature DCs were used as a control. Columns, mean of three experiments with different donors; bars, SD. *, P < 0.05; **, P < 0.001. C, kinetics of CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC production after DC maturation was measured by collecting supernatants after 12 h, followed by washing and replacement of cells in fresh culture media for a further 12 h before the final supernatants were collected. Representative of three experiments.

Figure 2.

αDC1s produce high levels of CXCL9/MIG in a sustained fashion. Monocyte-derived immature DCs were matured for 24 h to induce αDC1s (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C) or PGE2-DCs (IL-1β, TNF-α, IL-6, and PGE2). Protein levels, as determined by ELISA, of CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC (A) and CCL17/TARC and CCL22/MDC (B) in 24-h culture supernatants of previously washed mature DCs. Immature DCs were used as a control. Columns, mean of three experiments with different donors; bars, SD. *, P < 0.05; **, P < 0.001. C, kinetics of CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC production after DC maturation was measured by collecting supernatants after 12 h, followed by washing and replacement of cells in fresh culture media for a further 12 h before the final supernatants were collected. Representative of three experiments.

Close modal
Figure 3.

IFN-γ is the most critical factor in the αDC1 maturation cocktail for a high CXCL9/MIG production. Immature DCs were cultured for 24 h in the presence of different combinations of the components included in the αDC1 maturation cocktail (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C). Each component was abolished from the cocktail one at a time. Cells were then washed and replaced in the well and cultured in fresh medium for a further 24 h before supernatants were collected. Results shown, as determined by ELISA, are from a representative of four experiments with different donors.

Figure 3.

IFN-γ is the most critical factor in the αDC1 maturation cocktail for a high CXCL9/MIG production. Immature DCs were cultured for 24 h in the presence of different combinations of the components included in the αDC1 maturation cocktail (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C). Each component was abolished from the cocktail one at a time. Cells were then washed and replaced in the well and cultured in fresh medium for a further 24 h before supernatants were collected. Results shown, as determined by ELISA, are from a representative of four experiments with different donors.

Close modal

Supernatants from αDC1s recruit NK cells in a CXCL9/MIG-dependent manner. Supernatants from purified αDC1s, but not from PGE2-DCs, induced a substantial recruitment of purified peripheral blood NK cells in transwell experiments. Because circulating human NK cells express functional CXCR3 (27), we tested the possibility that the extremely high level of CXCL9 in αDC1 supernatants was primarily responsible for the observed NK cell recruitment. Adding anti-CXCL9 antibodies to the αDC1 supernatants led to a marked reduction of active NK cell migration to the lower chamber whereas the corresponding isotype control antibody (rabbit-IgG) only marginally affected NK cell migration (Fig. 4). Additionally, a statistically significant increased recruitment of NK cells toward αDC1 supernatants compared with PGE2-DCs (P = 0.003) was obtained when total PBMCs, instead of purified NK cells, were added to the upper chamber (Supplementary Fig. S2).

Figure 4.

Supernatants from αDC1s recruit NK cells in a CXCL9/MIG-dependent manner. Chemotaxis of purified peripheral blood NK cells toward chemokines produced by DCs matured with either the αDC1 or PGE2-DC maturation cocktail was tested by using a transwell assay. Supernatants collected from purified DCs stimulated for 24 h, washed and incubated for a further 24 h in fresh media, were added to the lower chambers of transwell plates. Medium only was used as a control. Anti-CXCL9 or corresponding isotype (rabbit-IgG) control antibody was added as indicated. NK cells (1 × 105) were added to the upper chamber and cells were incubated for 90 min. Cells in the lower chambers were harvested and stained with phycoerytrin-conjugated anti-CD56 and allophycocyanin-conjugated anti-CD3. CD3CD56+ NK cells located in the FSC/SSC lymphocyte gate were subsequently defined and counted by flow cytometry. Specific migration was calculated as percent of control (spontaneously migrated cells, migrating to medium only). Columns, means of triplicate wells from a representative of three independent experiments; bars, SD.

Figure 4.

Supernatants from αDC1s recruit NK cells in a CXCL9/MIG-dependent manner. Chemotaxis of purified peripheral blood NK cells toward chemokines produced by DCs matured with either the αDC1 or PGE2-DC maturation cocktail was tested by using a transwell assay. Supernatants collected from purified DCs stimulated for 24 h, washed and incubated for a further 24 h in fresh media, were added to the lower chambers of transwell plates. Medium only was used as a control. Anti-CXCL9 or corresponding isotype (rabbit-IgG) control antibody was added as indicated. NK cells (1 × 105) were added to the upper chamber and cells were incubated for 90 min. Cells in the lower chambers were harvested and stained with phycoerytrin-conjugated anti-CD56 and allophycocyanin-conjugated anti-CD3. CD3CD56+ NK cells located in the FSC/SSC lymphocyte gate were subsequently defined and counted by flow cytometry. Specific migration was calculated as percent of control (spontaneously migrated cells, migrating to medium only). Columns, means of triplicate wells from a representative of three independent experiments; bars, SD.

Close modal

αDC1s activate cocultured NK cells as determined by CD69 expression and intracellular IFN-γ production. Because NK cells were actively recruited toward supernatants from αDC1s, we next investigated whether NK cells might become activated, as determined by expression of CD69 and intracellular IFN-γ, on interaction with αDC1s or PGE2-DCs. Before coculturing with autologous or allogeneic nonadherent PBMCs, previously washed mature DCs were cultured in fresh medium for a further 24 hours. Activation of NK cells was analyzed by flow cytometry after 6 hours of coculture. Only DCs matured with the αDC1 maturation cocktail were able to induce a substantial IFN-γ production (Fig. 5A). Furthermore, when compared with PGE2-DCs, αDC1s proved superior in the induction of CD69 up-regulation. Similar results were detected when allogeneic, nonadherent PBMCs were used (data not shown). To ensure that the observed NK cell activation was not merely reflecting contamination of activated NK cells from the unpurified αDC1 fraction and to ensure that NK cell activation was solely a result of αDC1/NK cell interaction, additional flow cytometry experiments were done. Fluorescence-activated cell sorting (FACS) studies on unpurified αDC1 cultures (without addition of responding NK cells) excluded any substantial IFN-γ production in contaminating NK cells within the unpurified αDC1 fraction (Supplementary Fig. S3). However, when purified DCs and purified NK cells were used, no substantial activation (as determined by intracellular IFN-γ and CD69 expression) of autologous or allogeneic NK cells was observed in any DC/NK cell coculture (data not shown).

Figure 5.

αDC1s activate cocultured NK cells. Immature DCs were matured for 24 h to induce αDC1s (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C) or PGE2-DCs (IL-1β, TNF-α, IL-6, and PGE2), washed twice, and cultured in fresh medium for a further 24 h before being harvested and used for cocultures. A, nonadherent autologous PBMCs were cultured for 6 h alone (medium) or cocultured with DCs matured with any of the two maturation cocktails. Numbers inside the dot plots indicate the percentage of CD3CD56+ NK cells located in the FSC/SSC lymphocyte gate producing IFN-γ and expressing CD69. It is noteworthy that cocultures of purified NK cells and αDC1s did not result in NK cell activation (not shown). B, purified αDC1s were stimulated with soluble histidine-tagged CD40L subsequently cross-linked with antipolyhistidine antibodies in the absence or presence of purified peripheral blood NK cells. IFN-γ was measured in the supernatants by ELISA at 24 h. Columns, mean of four replicate wells; bars, SD. Representative of three experiments (<<, below the detection limit).

Figure 5.

αDC1s activate cocultured NK cells. Immature DCs were matured for 24 h to induce αDC1s (IL-1β, TNF-α, IFN-γ, IFN-α, and p-I:C) or PGE2-DCs (IL-1β, TNF-α, IL-6, and PGE2), washed twice, and cultured in fresh medium for a further 24 h before being harvested and used for cocultures. A, nonadherent autologous PBMCs were cultured for 6 h alone (medium) or cocultured with DCs matured with any of the two maturation cocktails. Numbers inside the dot plots indicate the percentage of CD3CD56+ NK cells located in the FSC/SSC lymphocyte gate producing IFN-γ and expressing CD69. It is noteworthy that cocultures of purified NK cells and αDC1s did not result in NK cell activation (not shown). B, purified αDC1s were stimulated with soluble histidine-tagged CD40L subsequently cross-linked with antipolyhistidine antibodies in the absence or presence of purified peripheral blood NK cells. IFN-γ was measured in the supernatants by ELISA at 24 h. Columns, mean of four replicate wells; bars, SD. Representative of three experiments (<<, below the detection limit).

Close modal

CD40 ligation licences αDC1s to induce IFN-γ production by cocultured NK cells. Taken together, our data indicate that the observed activation of NK cells within bulk lymphocyte fractions, induced by cocultured αDC1s (Fig. 5A), was dependent on cofactors expressed by lymphocytes within the responding nonadherent PBMC population. One obvious cofactor candidate is CD40L, which recently was shown to play a prominent role in the DC-dependent activation of human NK cells primed with IL-18 (28). As shown in Fig. 5B, in the absence of CD40 stimulation by cross-linked soluble CD40L, no detectable levels of IFN-γ were found in any cell culture/coculture supernatant using purified NK cells and/or purified αDC1s. These ELISA results were thus in line with FACS data on intracellular expression of IFN-γ. However, in cocultures of autologous αDC1 and NK cells, concurrent CD40-mediated stimulation induced a substantial production of IFN-γ within 24 hours.

Our results assign an additional novel feature of mature clinical grade (propagated in serum-free media) αDC1s that is of potential importance for their efficacy as anticancer vaccines. In addition to recent data provided by Mailliard and colleagues (17), we have shown that αDC1s, in contrast to DCs matured with the “gold standard” cocktail, exhibit a profuse and sustained production of the CXCR3 ligand CXCL9/MIG as well as substantial amounts of CXCL10/IP-10 and CXCL11/I-TAC after withdrawal of maturation stimuli. Supernatants collected 24 hours after washing of mature αDC1s efficiently recruited NK cells in a CXCL9-dependent manner. Finally, cocultures of NK cells and αDC1s induced IFN-γ production by NK cells, but only when concurrent CD40 stimulation was provided.

NK cells control antitumor (2931), allogeneic (32), xenogeneic (33), and virus-specific CTL (34) responses, and emerging evidence from rodent models further shows that DC-based cancer vaccines seem to collaborate with NK cells in the activation of tumor-specific CTL responses (3537). In vitro studies have shown that one of the basic requirements for DC-NK cell interactions is proximity or direct cognate interactions (38, 39), but there has been little data showing NK cell-DC contact in vivo. However, two recent articles have focused on the interaction between DCs and NK cells in reactive lymph nodes (22, 23). S.c. injection of bone marrow−derived, lipopolysaccharide-matured DCs in mice induced a rapid increase in draining lymph node cellularity (22). Although the proportion of T cells and B cells was found to be similar in DC-draining and control lymph nodes, the frequency of NK cells was up to 10-fold higher in DC-draining than in control lymph nodes (22). The increase in NK cell numbers peaked within 2 days, was dependent on the DC dose, and was not due to cell proliferation but rather to enhanced cell recruitment (22). A similar increase of NK cell number in draining lymph node was shown after s.c. infection of mice with Leishmania major (23). Confocal microscopy of lymph node sections from L. major–infected mice revealed that NK cells from infected animals tended to accumulate in the T-cell area, which is the site where antigen-loaded DCs are known to settle to communicate with recruited T cells entering via high endothelial venules (40).

Of fundamental importance, Martin-Fontecha and colleagues (22) showed in mice that the chemokine receptor CXCR3 exhibited a prominent role in the NK cell recruitment induced by injecting mature syngeneic DCs. It has further been shown that CXCL9/MIG become detectable on the lumenal side of HEVs in the draining lymph nodes within a few hours of s.c. injection of mature DCs (41). Because arrival in the lymph node of s.c. injected DCs probably takes more than 16 to 24 hours, it could be argued that measurement of chemokine secretion from time of withdrawal of the maturation cocktail (a time point when mature αDC1s are proposed to be injected) until 24 hours is irrelevant. However, similar to monocyte chemoattractant protein 1 (42), it is likely that peripherally secreted CXCR3 ligands (produced by injected αDC1s before lymph node entry) may be transported via lymphatic vessels to draining lymph nodes, thus justifying measurement of accumulated chemokine secretion during the initial 24-hour period after full maturation. Although the murine-based study of Martin-Fontecha and colleagues indicates that CXCR3 is playing a prominent role in lymph node–homing properties of NK cells, we are not aware of any studies clearly showing that the same holds true in humans. However, about NK cell recruitment to other inflamed or immunologically active tissues, several studies in humans indicate that CXCR3-expressing NK cells may be recruited by CXCR3 ligands. For instance, CXCR3-expressing NK cells are actively recruited to psoriatic plaques (43), and ex vivo studies on those NK cells revealed that the CXCR3 ligand CXCL10 was the most potent chemoattractant for such CXCR3-expressing NK cells. In addition, decidual infiltration of human NK cells seems to be regulated by the local release of CXCL10 by the endometrium (44). The novel feature of high and sustained production of the NK cell–recruiting CXCR3 ligand CXCL9/MIG by mature αDC1s, as revealed in the present study, thus indicates a NK cell–recruiting capacity into draining lymph nodes when injected into human patients.

As expected (4547), the presence of IFN-γ in the maturation cocktail was of major importance for high CXCL9/MIG, CXCL10/IP-10, and CXCL11/I-TAC production. Rather surprisingly, IL-1β exhibited a major role in the induced CXCL9/MIG production whereas IFN-α, p-I:C, and TNF-α were of minor importance. However, single treatment of immature DCs with IFN-γ or IL-1β induced no detectable amounts of the CXCR3 ligands.

By using NK cell−depleted mice or Ifng−/− mice reconstituted with IFN-γ–sufficient or IFN-γ–deficient NK cells, Martin-Fontecha and colleagues (22) showed that recruited NK cells provide an early source of IFN-γ that was necessary for Th1 polarization. Bajenoff and colleagues (23) further showed that the recruited NK cells accumulated in the T-cell area, formed conjugates with DCs, and frequently expressed IFN-γ, whereas NK cells localized in the medulla were mainly IFN-γ negative. These observations thus indicate that proximity and/or direct cognate interactions between DCs and NK cells are required for the induction of IFN-γ production by NK cells. In line with these results, our in vitro data show that the observed activation of NK cells, induced by cocultured αDC1s, was dependent on cofactors expressed by lymphocytes within the responding nonadherent PBMC population. One obvious cofactor candidate is CD40L, which was recently shown to play a prominent role in the DC-dependent activation of human NK cells primed with IL-18 (28). Similar to these findings, we found that in cocultures of autologous αDC1s and NK cells, concurrent CD40-mediated stimulation induced a substantial production of IFN-γ within 24 hours. Furthermore, in the absence of CD40 stimulation, no detectable levels of IFN-γ were found in any cell culture/coculture supernatant using purified NK cells and/or purified αDC1s. Taken together, these data thus suggest that the observed discrepancy about IFN-γ production in αDC/NK cell cocultures using purified NK cells or NK cells contained within a total lymphocyte population depends on the absence or presence of potentially CD40L-expressing lymphocytes that may up-regulate CD40L on stimulation with autologous αDC1s. Because no nonself antigens, such as xenogneic proteins from FCS, were included in the present culture media, such cells could possibly be CD1d-restricted NK T cells recognizing endogenous glycolipids presented on mature DCs (48) or autoreactive CD4+ T cells that become activated during an autologous mixed lymphocyte reaction (49).

Collectively, the in vitro data recently presented by Mailliard and colleagues (17) in conjunction with our present data thus allow for the clinical use of DCs, combining a fully mature status with high migratory function, the ability to produce IL-12p70 on cognate interaction with T cells expressing CD40L (17), and the ability to recruit and activate IFN-γ production in NK cells in the presence of CD40L-expressing lymphocytes. All these important features predict a strong Th1-polarized antitumor immune response in patients receiving αDC1-based vaccines loaded with tumor antigens.

No potential conflicts of interest were disclosed.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Grant support: The MIVAC Swedish Foundation for Strategic Research Centre, the Health and Medical Care Committee of the Region Västra Götaland, and the Fondazione Cassa di Risparmio di Puglia.

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.

We thank Peter Eriksson for support, and Linda Paulson, Maurice Curtis, and Katarina Junevik for critical comments on the manuscript and for stimulating discussions.

1
Schuler G, Schuler-Thurner B, Steinman RM. The use of dendritic cells in cancer immunotherapy.
Curr Opin Immunol
2003
;
15
:
138
–47.
2
Brody JD, Engleman EG. DC-based cancer vaccines: lessons from clinical trials.
Cytotherapy
2004
;
6
:
122
–7.
3
Banchereau J, Steinman RM. Dendritic cells and the control of immunity.
Nature
1998
;
392
:
245
–52.
4
Saeki H, Moore AM, Brown MJ, Hwang ST. Cutting edge: secondary lymphoid-tissue chemokine (SLC) and CC chemokine receptor 7 (CCR7) participate in the emigration pathway of mature dendritic cells from the skin to regional lymph nodes.
J Immunol
1999
;
162
:
2472
–5.
5
Sallusto F, Lanzavecchia A. Understanding dendritic cell and T-lymphocyte traffic through the analysis of chemokine receptor expression.
Immunol Rev
2000
;
177
:
134
–40.
6
Cyster JG. Chemokines, sphingosine-1-phosphate, and cell migration in secondary lymphoid organs.
Annu Rev Immunol
2005
;
23
:
127
–59.
7
Sallusto F, Lanzavecchia A. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and down-regulated by tumor necrosis factor α.
J Exp Med
1994
;
179
:
1109
–18.
8
Luft T, Jefford M, Luetjens P, et al. Functionally distinct dendritic cell (DC) populations induced by physiologic stimuli: prostaglandin E2 regulates the migratory capacity of specific DC subsets.
Blood
2002
;
100
:
1362
–72.
9
Scandella E, Men Y, Legler DF, et al. CCL19/CCL21-triggered signal transduction and migration of dendritic cells requires prostaglandin E2.
Blood
2004
;
103
:
1595
–601.
10
Kalinski P, Hilkens CM, Wierenga EA, Kapsenberg ML. T-cell priming by type-1 and type-2 polarized dendritic cells: the concept of a third signal.
Immunol Today
1999
;
20
:
561
–7.
11
Lee AW, Truong T, Bickham K, et al. A clinical grade cocktail of cytokines and PGE2 results in uniform maturation of human monocyte-derived dendritic cells: implications for immunotherapy.
Vaccine
2002
;
20
Suppl 4:
A8
–22.
12
Xu S, Koski GK, Faries M, et al. Rapid high efficiency sensitization of CD8+ T cells to tumor antigens by dendritic cells leads to enhanced functional avidity and direct tumor recognition through an IL-12-dependent mechanism.
J Immunol
2003
;
171
:
2251
–61.
13
Zitvogel L, Couderc B, Mayordomo JI, Robbins PD, Lotze MT, Storkus WJ. IL-12-engineered dendritic cells serve as effective tumor vaccine adjuvants in vivo.
Ann N Y Acad Sci
1996
;
795
:
284
–93.
14
Nishioka Y, Hirao M, Robbins PD, Lotze MT, Tahara H. Induction of systemic and therapeutic antitumor immunity using intratumoral injection of dendritic cells genetically modified to express interleukin 12.
Cancer Res
1999
;
59
:
4035
–41.
15
Vieira PL, de Jong EC, Wierenga EA, Kapsenberg ML, Kalinski P. Development of Th1-inducing capacity in myeloid dendritic cells requires environmental instruction.
J Immunol
2000
;
164
:
4507
–12.
16
Alder J, Hahn-Zoric M, Andersson BA, Karlsson-Parra A. Interferon-γ dose-dependently inhibits prostaglandin E2-mediated dendritic-cell-migration towards secondary lymphoid organ chemokines.
Vaccine
2006
;
24
:
7087
–94.
17
Mailliard RB, Wankowicz-Kalinska A, Cai Q, et al. α-Type-1 polarized dendritic cells: a novel immunization tool with optimized CTL-inducing activity.
Cancer Res
2004
;
64
:
5934
–7.
18
Bancroft GJ. The role of natural killer cells in innate resistance to infection.
Curr Opin Immunol
1993
;
5
:
503
–10.
19
Andersson A, Dai WJ, Di Santo JP, Brombacher F. Early IFN-γ production and innate immunity during Listeria monocytogenes infection in the absence of NK cells.
J Immunol
1998
;
161
:
5600
–6.
20
Berg RE, Crossley E, Murray S, Forman J. Relative contributions of NK and CD8 T cells to IFN-γ mediated innate immune protection against Listeria monocytogenes.
J Immunol
2005
;
175
:
1751
–7.
21
Mailliard RB, Son YI, Redlinger R, et al. Dendritic cells mediate NK cell help for Th1 and CTL responses: two-signal requirement for the induction of NK cell helper function.
J Immunol
2003
;
171
:
2366
–73.
22
Martin-Fontecha A, Thomsen LL, Brett S, et al. Induced recruitment of NK cells to lymph nodes provides IFN-γ for T(H)1 priming.
Nat Immunol
2004
;
5
:
1260
–5.
23
Bajenoff M, Breart B, Huang AY, et al. Natural killer cell behavior in lymph nodes revealed by static and real-time imaging.
J Exp Med
2006
;
203
:
619
–31.
24
Jeansson M, Granqvist AB, Nystrom JS, Haraldsson B. Functional and molecular alterations of the glomerular barrier in long-term diabetes in mice.
Diabetologia
2006
;
49
:
2200
–9.
25
Bonecchi R, Bianchi G, Bordignon PP, et al. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s.
J Exp Med
1998
;
187
:
129
–34.
26
Iellem A, Mariani M, Lang R, et al. Unique chemotactic response profile and specific expression of chemokine receptors CCR4 and CCR8 by CD4+CD25+ regulatory T cells.
J Exp Med
2001
;
194
:
847
–53.
27
Colvin RA, Campanella GS, Manice LA, Luster AD. CXCR3 requires tyrosine sulfation for ligand binding and a second extracellular loop arginine residue for ligand-induced chemotaxis.
Mol Cell Biol
2006
;
26
:
5838
–49.
28
Mailliard RB, Alber SM, Shen H, et al. IL-18-induced CD83+CCR7+ NK helper cells.
J Exp Med
2005
;
202
:
941
–53.
29
Diefenbach A, Jensen ER, Jamieson AM, Raulet DH. Rae1 and H60 ligands of the NKG2D receptor stimulate tumour immunity.
Nature
2001
;
413
:
165
–71.
30
Kelly JM, Darcy PK, Markby JL, et al. Induction of tumor-specific T cell memory by NK cell-mediated tumor rejection.
Nat Immunol
2002
;
3
:
83
–90.
31
Mocikat R, Braumuller H, Gumy A, et al. Natural killer cells activated by MHC class I(low) targets prime dendritic cells to induce protective CD8 T cell responses.
Immunity
2003
;
19
:
561
–9.
32
Kos FJ, Engleman EG. Requirement for natural killer cells in the induction of cytotoxic T cells.
J Immunol
1995
;
155
:
578
–84.
33
Smyth MJ, Kelly JM. Accessory function for NK1.1+ natural killer cells producing interferon-γ in xenospecific cytotoxic T lymphocyte differentiation.
Transplantation
1999
;
68
:
840
–3.
34
Burlington DB, Djeu JY, Wells MA, Kiley SC, Quinnan GV, Jr. Large granular lymphocytes provide an accessory function in the in vitro development of influenza A virus-specific cytotoxic T cells.
J Immunol
1984
;
132
:
3154
–8.
35
Adam C, King S, Allgeier T, et al. DC-NK cell cross talk as a novel CD4+ T-cell-independent pathway for antitumor CTL induction.
Blood
2005
;
106
:
338
–44.
36
Dao T, Gomez-Nunez M, Antczak C, et al. Natural killer cells license dendritic cell cross-presentation of B lymphoma cell-associated antigens.
Clin Cancer Res
2005
;
11
:
8763
–72.
37
Kim A, Noh YW, Kim KD, Jang YS, Choe YK, Lim JS. Activated natural killer cell-mediated immunity is required for the inhibition of tumor metastasis by dendritic cell vaccination.
Exp Mol Med
2004
;
36
:
428
–43.
38
Walzer T, Dalod M, Robbins SH, Zitvogel L, Vivier E. Natural-killer cells and dendritic cells: “l'union fait la force”.
Blood
2005
;
106
:
2252
–8.
39
Cooper MA, Fehniger TA, Fuchs A, Colonna M, Caligiuri MA. NK cell and DC interactions.
Trends Immunol
2004
;
25
:
47
–52.
40
Bajenoff M, Granjeaud S, Guerder S. The strategy of T cell antigen-presenting cell encounter in antigen-draining lymph nodes revealed by imaging of initial T cell activation.
J Exp Med
2003
;
198
:
715
–24.
41
Guarda G, Hons M, Soriano SF, et al. L-selectin-negative CCR7 effector and memory CD8+ T cells enter reactive lymph nodes and kill dendritic cells.
Nat Immunol
2007
;
8
:
743
–52.
42
Palframan RT, Jung S, Cheng G, et al. Inflammatory chemokine transport and presentation in HEV: a remote control mechanism for monocyte recruitment to lymph nodes in inflamed tissues.
J Exp Med
2001
;
194
:
1361
–73.
43
Ottaviani C, Nasorri F, Bedini C, de Pita O, Girolomoni G, Cavani A. CD56brightCD16(−) NK cells accumulate in psoriatic skin in response to CXCL10 and CCL5 and exacerbate skin inflammation.
Eur J Immunol
2006
;
36
:
118
–28.
44
Sentman CL, Meadows SK, Wira CR, Eriksson M. Recruitment of uterine NK cells: induction of CXC chemokine ligands 10 and 11 in human endometrium by estradiol and progesterone.
J Immunol
2004
;
173
:
6760
–6.
45
Luster AD, Ravetch JV. Biochemical characterization of a γ interferon-inducible cytokine (IP-10).
J Exp Med
1987
;
166
:
1084
–97.
46
Farber JM. HuMig: a new human member of the chemokine family of cytokines.
Biochem Biophys Res Commun
1993
;
192
:
223
–30.
47
Cole KE, Strick CA, Paradis TJ, et al. Interferon-inducible T cell α chemoattractant (I-TAC): a novel non-ELR CXC chemokine with potent activity on activated T cells through selective high affinity binding to CXCR3.
J Exp Med
1998
;
187
:
2009
–21.
48
Mattner J, Debord KL, Ismail N, et al. Exogenous and endogenous glycolipid antigens activate NKT cells during microbial infections.
Nature
2005
;
434
:
525
–9.
49
Scheinecker C, Machold KP, Majdic O, Hocker P, Knapp W, Smolen JS. Initiation of the autologous mixed lymphocyte reaction requires the expression of costimulatory molecules B7-1 and B7-2 on human peripheral blood dendritic cells.
J Immunol
1998
;
161
:
3966
–73.