Background
Dendritic cells (DC) are specialized in antigen presentation which plays a key role in the initiation of primary immune responses. Immature DC phagocyte and process antigens and after maturation they stimulate antigen specific T cells. This is the prerequisite for orchestrating the cellular and humoral immune response [
1].
This unique role of DC in the activation of host defense has made them a promising candidate for vaccination against a wide range of infectious agents and tumor antigens. DC can be generated by culturing monocytes
in vitro with medium containing interleukin (IL)-4 and granulocyte-macrophage colony-stimulating factor (GM-CSF). TNF-α or a mixture of different proinflammatory molecules are needed to generate mature DC [
2,
3]. So far, the therapeutic results observed in patients with malignancies following vaccination with IL-4-DC are encouraging at best [
4,
5]. Therefore, there is a particular need for culture conditions facilitating the generation of more efficacious DC.
Recently, several groups generated DC by culturing monocytes in the presence of IFN-α and GM-CSF (IFN-DC) for three days [
6‐
11]. IFN-α is released in large amounts during antiviral immune responses and is involved in the activation of cells of the innate and adaptive immune system [
12]. In particular, IFN-α enhances the cytotoxic capacity of NK cells. IFN-α has also been successfully used for the treatment of patients with chronic myeloid leukemia (CML) [
13] and Non-Hodgkin lymphoma (NHL) [
14]. The therapeutical effects could be related to IFN-α stimulated NK cells and DC. Therefore, it is conceiving that IFN-DC would be more efficient for vaccination of patients with NHL or CML.
In order to examine the differences between IFN-DC and conventional IL-4/TNF-DC, we compared the morphology, immunophenotype, functional efficacy and gene expression profiles of these cell preparations with regard to their usefullness in anti-tumor vaccination strategies.
Methods
Isolation and culture of cells
Mononuclear cells (PBMC) were obtained from buffy coats of healthy individuals. Monocytes were isolated by negative selection using a RosetteSep antibody cocktail (Stemcell Technologies, Vancouver, Canada), according to the manufacturer's protocol. The resulting cell population after this procedure had a median purity of 72% CD14+ monocytes.
IFN-DC were generated by culturing monocytes in plastic flasks (BD Falcon, UK) for 3 days in serumfree X-VIVO 20 medium (BioWhitaker Europe, Belgium), supplemented with 1000 U/ml IFN-α (IntronA, Griffith Micro Science, Rantigny, France) and 1000 U/ml GM-CSF (Immunex, Seattle, US). For the generation of IL-4/TNF-DC, monocytes were cultured in serumfree medium containing 500 U/ml IL-4 (Promocell, Heidelberg, Germany) and 800 U/ml GM-CSF for 5 days. The resulting immature DC were further treated by a 2 day culture step with fresh medium containing 1000 U/ml TNF-α (Sigma) and 800 U/ml GM-CSF. For all experiments, IFN-DC and IL-4/TNF-DC were used after a culture period of 3 and 7 days, respectively. If not mentioned otherwise, preparations of both groups were derived from different individuals. The viability of cells was determined by Trypan blue exclusion.
Immunophenotypic analysis
Flow cytometry was performed on a FACScan flow cytometer (BD Biosciences, San Jose, US). The following FITC or PE labeled mouse antibodies were used: CD45, CD1a, CD3, CD11c, CD14, CD19, CD40, CD49b, CD56, CD80, CD83, CD86, CD123, HLA-DR, TRAIL, NKG2D, nonspecific IgG1, IgG2a, a mixture of IgG1 and IgG2a (BD Biosciences, San Jose, US), CD209 (Beckman Coulter, Marseille, France) and GZMB (Hölzel Diagnostika, Köln, Germany). For intracellular staining, cells were permeabilized with BD Cytofix/Cytoperm (BD Bioscience, San Jose, US) according to the manufacturer's guideline.
Analysis of DC functions
The allostimulatory capacity of DC was measured in an allogeneic mixed leukocyte reaction (MLR). DC were resuspended in RPMI 1640 medium (Biochrome, Berlin, Germany), supplemented with 10% FCS (PAA, Pasching, Austria), 2 mM Glutamine, 100 U/ml Penicillin and 100 μg/ml Streptomycin (Sigma) and irradiated with 30 Gy. Different DC numbers were cultivated with 1 × 105 allogeneic PBMC of a healthy donor in a round-bottomed 96-well plate (Corning, NY, US). Antibodies against CD28 and CD49d (BD Biosciences, San Jose, US) were added at 1 μg/ml. The cells were incubated for 4 days at 37°C. For the last 20 h of the culture, 1 μCi/well of 3 [H]-Thymidin (Amersham, Braunschweig, Germany) was added. Finally, 3 [H]-Thymidin uptake was measured on a β-scintillation counter (Perkin Elmer, Shelton, CT, US). The stimulatory capacity was expressed by the stimulation index SI = cpm of stimulated PBMC/cpm of unstimulated PBMC. Each experiment was done in triplicates.
Induction of cytokine production in T cells by DC was determined by intracellular staining. As described above, 1 × 105 freshly isolated PBMC were cocultured with 5 × 104 DC per well in a 96-well plate for 3 days. To block protein secretion, 10 μg/ml BrefeldinA (Sigma) was added for the last 4 hours of the culture period. Cells were harvested and incubated with a FITC labelled anti CD3 antibody (BD Bioscience, San Jose, US). Cells were then permeabilized with BD Cytofix/Cytoperm solution, stained with PE labelled anti IFN-γ or IL-4 antibodies (BD Biosciences, San Jose, US) and analyzed by flow cytometry.
Migration of DC was measured in 24-well transwell culture chambers (Costar, Cambridge, MA, US) as previously described [
8,
15]. The 8 μm-pore transwell filters were briefly coated with 10 μg/cm
2 fibronectin (Sigma). The upper chamber compartment was loaded with 2.5 × 10
5 IFN-DC or IL-4/TNF-DC in 150 μl X-VIVO 20 medium. The lower chamber compartment was filled with 500 μl medium supplemented with 100 ng/ml recombinant Mip-3β (Promocell, Heidelberg, Germany). After 2 h incubation at 37°C, cells were harvested from the lower chamber compartment and counted by FACS analysis using BD calibration beads.
Cytotoxicity assay
Freshly prepared DC preparations derived from the same healthy individuals for both methods were tested for their cytolytic activity against K562 target cells by flow cytometry. Before coculture, 1 × 106 K562 cells were labeled with 0.5 μM carboxyfluorescein diacetate succinimidyl ester (CFDA-SE, Molecular probes, Paisley, UK) for 15 min. at 37°C. Different numbers of effector cells were cultured with 1 × 105 labelled K562 cells in RPMI 1640 medium containing 10% FCS. NK cells used as a positive control were isolated from peripheral blood of a healthy donor with a MACS NK cell separation kit (Miltenyi, Bergisch Gladbach, Germany), and stimulated with 1000 U/ml IL-2 (Chiron). The human lymphoma B cell line MHH-PREB-1 (German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany) served as a negative control. After 4 h of culture dead cells were stained with propidium iodide (PI, Becton Dickinson) and analyzed by flow cytometry. Specific lysis was determined by the formula: % specific lysis = experimental % of PI+ CFDA-SE labeled cells - spontanous % of PI+ CFDA-SE labeled cells.
Identification of differential gene expression
Total RNA from DC preparations was isolated with the RNeasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer's recommendations. cRNA labelling, hybridization to HG-Focus GeneChips (Affymetrix, UK, Ltd.) and processing of the microarray data was performed as described elsewhere [
16]. Differential gene expression was defined by a
false discovery rate (FDR) of 5%, as indicated by a q-value ≤ 5% for individual genes. To reduce the number of genes, we concentrated on genes with a fold change ≥ 2 comparing both groups. For a specific analysis of NK cell markers, all differentially expressed genes with q-values ≤ 5% were included.
RT-PCR analysis
cDNA was synthesized from the same RNA preparations used for the microarray hybridizations as described [
17], and amplified using the Assays-on-Demand Gene Expression products on the ABI PRISM 7900HT sequence detection system instrument (Applied Biosystems, Applera Deutschland GmbH, Darmstadt, Germany). For relative quantification, β2-microglobulin mRNA served as an external standard. The following genes were detected by Assays-on-Demand gene expression products: CCL8, BCL2A1, GZMB, CCR7, LAMP3, PKR, ADAMDEC1, FCGR1A, PDHA1, CCND2 and B2M. Relative gene expression levels were expressed as the difference in Ct values (deltaCt) of the target gene and B2M.
Statistical analysis
The results were analysed for statistical significance by a two-sided, unpaired student's t-test for experiments with independent samples and a two-sided, paired student's t-test for paired samples. p-values < 0.05 indicate significant differences.
Discussion
DC based vaccines for patients with malignant diseases generated under different culture conditions have been investigated for more than a decade. Despite these efforts, clinical results of DC vaccination studies showed therapeutic efficacy only in a limited number of patients so far [
4]. In search of an alternative way for DC generation we examined the molecular and functional characteristics of IFN-DC in comparison to IL-4/TNF-DC.
We could show that both, IFN-DC and IL-4/TNF-DC, display typical DC characteristics, but also have distinct molecular and functional phenotypes, as a reflection of the distinct transcriptional signature of IFN-α in comparison to other cytokines as recently described [
19]. Our results from gene expression analysis confirm previous reports that IFN-DC have signs of a pronounced maturation state and an increased migratory capacity to the lymph nodes in comparison to IL-4/TNF-DC [
8,
10]. Strikingly, IFN-DC showed a more plasmacytoid phenotype associated with NK cell characteristics on molecular and protein level as well as a functional cytotoxic activity against tumor cells. Therefore, the use of IFN-DC in vaccination trials may result in a better clinical antitumor immune response.
As others have shown before [
6‐
11], IFN-DC in our study had a DC morphology and immunophenotype with high levels of CD11c, CD86 and HLA-DR as well as functional DC characteristics like the capacity to stimulate T cells. The expression of costimulatory molecules was in accordance to other studies using serumfree culture conditions [
37]. Nevertheless, we found a stronger upregulation of the costimulatory molecules CD40 and CD80 on IFN-DC than on IL-4/TNF-DC, although IFN-DC did not mediate an increased allostimulatory reaction. Further, IFN-DC triggered a balanced Th1/Th2 response, whereas IL-4/TNF-DC were strongly biased to evoke a Th1 response. This is in line with Lapenta
et al., 2003 and 2006, who showed that IFN-DC could induce a massive humoral and cellular immune response [
9,
38].
In addition, the greater RNA levels for cytokines and chemokines in IFN-DC like IFN-β, MCPs, MIP2A and MIP2B as well as the IL-1β converting enzyme (CASP1) that catalyzes the secretion of active forms of IL-1β and IL-18, suggest that IFN-DC may also recruit other innate cytotoxic effectors like NK cells [
12,
28,
29,
31,
32] and neutrophils [
29,
30].
It is well accepted that a pronounced DC maturation status is important for the induction of efficient immune responses by DC immunotherapy [
39]. In our study, both DC preparations showed only marginal upregulation of the maturation marker CD83, which is a result from culture conditions using serum free medium [
37]. Gene expression analysis revealed that IL-4/TNF-DC have more immature DC characteristics. This was indicated by the higher expression of several genes envolved in phagocytosis such as Fc and complement receptors as well as genes envolved in epithelial adhesion structure formation including the genes for vinculin or the integrin αE chain. In contrast to this finding, IFN-DC showed a higher expression of several alternative DC maturation markers than IL-4/TNF-DC that are involved in antigen processing (DCLAMP [
22]), migration to and localization in the lymph nodes (CCR7 [
23], integrin α4 [
24] and decysin [
26]) as well as survival (BCL2A1 [
27]). Therefore, on molecular level, IFN-DC show the prerequisite to initiate an adaptive immune response in the lymph node [
1,
39]. Importantly, this capacity of IFN-DC could be demonstrated functionally by a higher migratory capacity of IFN-DC
in vitro compared to IL-4/TNF-DC as shown by transwell experiments. This is also in line with Parlato
et al., 2001, who had shown before that IFN-DC have a higher migratory potency than IL-4-DC not stimulated by TNF-α [
8].
Interestingly, the higher expression of genes of the IFN pathway like the transcription factors STAT1 and IRF7 as well as PKR, and IFN-β in IFN-DC resembles the expression pattern of plasmacytoid DC [
40‐
43] that are the major type I IFN producers during viral infections [
44]. We found high surface levels of the plasmacytoid DC marker CD123 and low levels of the myeloid DC marker CD209 on IFN-DC, which is in line with Mohty
et. al., 2003, who also described other plasmacytoid DC markers like TLR7 on IFN-DC [
45].
The most important new finding of our study was the significant upregulation of 32 genes strongly related to NK cell functions in IFN-DC compared to IL-4/TNF-DC. These include NK cell receptors NKp80, NKp44, NKp46 and NKG2D that are synergizing the cytotoxic activity of NK cells [
46,
47], as well as CD56 and cytotoxic effector molecules such as granzymes and TRAIL. Indeed, on protein level, we could detect intracellular pools of TRAIL and granzyme B in IFN-DC. Finally, as a further corroboration of the suggested cytotoxic capacity, IFN-DC but not IL-4/TNF-DC were able to kill K562 cells
in vitro. This is in concordance with the previously made observation that DC can kill tumor cells by TRAIL mediated lysis [
7,
48]. Still, the expression of granzymes might further argue for a perforin mediated killing mechanism by IFN-DC.
These findings are of particular interest, as a new murine DC cell population has been recently described, termed interferon-producing killer dendritic cells (IKDC), that express molecular markers of plasmacytoid DC and NK cells [
49,
50]. IKDC exhibit specific cytolytic activity upon contact with tumor cells or activation with CpG oligonucleotides and subsequently upregulate costimulatory molecules, migrate to the lymph nodes and present antigen to T cells. Indeed, 9 of the genes specifically expressed by IKDC including granzymes, NKG2D, NKp46, and CD49b as detemined by microarray analsysis [
49], were also differentially expressed by IFN-DC in comparison to IL-4/TNF-DC. Together with the pronounced migratory potential and the cytotoxic capacity of IFN-DC, the similarities between IFN-DC and mouse IKDC suggest that also in humans a molecular and functional relationship exists between DC and NK cells.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MK, RK, RH, GK, RF designed the experiments.
MK, NS, DM, CP, EDB, MW, RF performed experiments and data analysis.
MK, RK, MS, RH, GK, RF performed interpretation of the data.
MK, RK, MS, AC, RH, GK, RF were involved in drafting the manuscript.
All authors were involved in revising the manuscript critically for important intellectual content and have given their approval to the final version of the manuscript.