Both authors contributed equally to this work.
CD40 ligation during dendritic cell maturation reduces cell death and prevents interleukin-10-induced regression to macrophage-like monocytes
Version of Record online: 13 DEC 2007
© 2007 The Authors. Journal compilation © 2007 Blackwell Munksgaard
Volume 17, Issue 3, pages 177–187, March 2008
How to Cite
Haenssle, H., Buhl, T., Knudsen, S., Krueger, U., Rosenberger, A., Reich, K. and Neumann, C. (2008), CD40 ligation during dendritic cell maturation reduces cell death and prevents interleukin-10-induced regression to macrophage-like monocytes. Experimental Dermatology, 17: 177–187. doi: 10.1111/j.1600-0625.2007.00668.x
Conflict of interest statement: All authors declare no conflict of interest. The corresponding author has full access to all the data in the study and had final responsibility for the decision to submit the manuscript for publication.
- Issue online: 13 DEC 2007
- Version of Record online: 13 DEC 2007
- Accepted for publication 16 November 2007
- bcl-XL expression;
- cell survival;
- cytokine cocktail;
- dendritic cells;
Abstract: Dendritic cells (DCs) have become popular candidates in cancer vaccination because of their crucial role in inducing T-cell responses. However, clinical studies greatly differ in their protocols for generating DCs and the efficacy in treating established tumors needs to be improved. We systematically analyzed DCs maturated by five different protocols for surface markers, the alloproliferative T-cell response, the DC survival after cytokine deprivation, the stability of surface markers under the influence of interleukin-10 (IL-10) and the DC cytokine secretion pattern. Monocyte-derived DCs were maturated by CD40-ligand (CD40-L), unmethylated cytosine–guanosine dinucleotides-oligodinucleotides (CpG-ODN), an inflammatory cytokine cocktail (ICC), a combination of ICC and CD40-L, or ICC, CD40-L and CpG-ODN. A high co-expression of DC maturation and costimulation markers was found after treatment with ICC plus CD40-L (69.3 ± 9.6% CD83/CD80 double positive staining) and correlated with a significantly increased cell survival, a high expression of the antiapoptotic factor bcl-XL, a stable CD83high/CD14low expression under the influence of IL-10, and a strong alloproliferative T-cell response. In conclusion, our data support the use of maturation protocols containing ICC plus CD40-L in order to generate highly mature, phenotypically stable, cell-death resistant, and T-cell stimulatory DCs for clinical application in cancer patients.
plamacytoid dendritic cells
inflammatory cytokine cocktail
In a normal functioning immune environment, dendritic cells (DCs) acquire, process and present antigen in the context of costimulation, resulting in the generation of antigen-specific T-cell responses (1). DCs that migrate from tissue to lymph nodes in vivo have a life expectancy of 2–3 days and can, therefore, be considered as disposable packets, each carrying a given amount of peptide-major histocompatibility complex (MHC), costimulatory molecules and cytokines (2,3). Antigen presentation by functional DC induces both humoral and cell-mediated responses, antitumor activity being particularly dependent on the latter (4). Besides the hope to successfully apply DC-based immunotherapy for cancer, it has become clear now that DCs also play a critical role in the induction of peripheral immunological tolerance (5). The heterogeneous functions of DCs in immune regulation depend on the diversity of DC subsets and lineages, and on their functional plasticity at different maturational stages (6,7). For any successful immunisation strategy in cancer, it will, therefore, be critical to apply effective DC subsets mediating appropriate costimulation and cytokine profiles.
Several in vitro and in vivo studies have established a crucial role for CD40-ligand [CD40-L (CD154)] during the priming phase of adaptive immune responses. The function of CD40-L expressing T helper (Th) cells in the stimulatory loop between DCs, cytotoxic T lymphocytes and Th-cells can be replaced in vitro by preactivating DCs with anti-CD40 antibodies or soluble CD40-L (8). Ligation of CD40 on DCs was shown to stimulate their differentiation and activation. Authors of a first human clinical trial utilising CD40-L-maturated DCs for immunotherapy of patients with resected melanoma reported successful induction of antigen-specific T-cell responses (9). Synthetic oligonucleotides containing unmethylated cytosine–guanosine dinucleotides (CpG-ODN) mimic the presence of bacterial DNA and promote maturation and activation of DC subsets (10). Toll-like-receptor 9 (TLR9) seems to be the primary target for CpG-ODN and is highly expressed on human plasmacytoid dendritic cells (PDC), which constitute a cell fraction after monocyte enrichment by plastic adherence. With the addition of CpG sequences, it might be possible to recruit this PDC fraction into the maturation process. A defined inflammatory cytokine cocktail (ICC), inducing maturation of monocyte-derived DCs, was first described by Jonuleit et al. (11). This cocktail, containing Tumor Necrosis Factor-α (TNF-α), interleukin-1β (IL-1β), IL-6 and Prostaglandin E2 (PGE2), was employed in many recent clinical cancer trials using DCs loaded with tumor cell preparations (12–14).
Aim of the present study was to compare five maturation protocols with regard to DC qualities that may have a strong impact on the outcome of vaccination therapy in cancer patients. Maturation stimuli consisted of immunostimulatory CpG-ODN (protocol 1), CD40-L (protocol 2), ICC as a gold standard (protocol 3), the combination of ICC and CD40-L (protocol 4), and the triple combination of ICC, CD40-L and CpG-ODN (protocol 5).
Material and methods
Antibodies and reagents
The following monoclonal antibodies (mAb) were used for flow cytometry experiments: Fluorescein-isothiocyanat (FITC) or phycoerythrin (PE)-conjugated mAbs to human CD3, CD4, CD14, CD16, CD19 (Caltag; Hamburg, Germany), CD58, CD80, CD86, CD154 (BD Pharmingen; Heidelberg, Germany), CD83 (Beckmann Coulter Immunotech; Krefeld, Germany), BDCA2 (R&D; Wiesbaden, Germany), and Human Leukocyte Antigen-DR (HLA-DR) (Leinco Technologies; St Louis, MO, USA). For staining of necrotic cells, propidium iodide (Invitrogen; Karlsruhe, Germany) was purchased. For intracellular flow cytometry of bcl-XL and bcl-2, unconjugated primary antibodies (Biomol GmbH; Hamburg, Germany) and FITC or PE-labelled secondary antibodies (Biosource; Camarillo, CA, USA) were used. For magnetic cell sorting (MACS) of lymphocytes, anti-CD4 and anti-CD8 beads were purchased (Miltenyi Biotech; Bergisch Gladbach, Germany).
DC generation and maturation from leukaphereses products
Leukaphereses of healthy donors were obtained by a Cobe Spectra Auto PBSC device (Gambro BCT Inc.; Lakewood, CA, USA) after approval by local ethics committees and written informed consent according to institutional guidelines. Peripheral blood mononuclear cells (PBMC) were prepared by density centrifugation using a Ficoll-Hypaque gradient (PAA Laboratories GmbH; Coelbe, Germany). PBMC were resuspended at 2 × 107 cells/ml in RPMI-1640 medium (Bio-Whittaker; Walkersville, MD, USA) supplemented with 5% autologous serum and brought to culture dishes (Greiner; Solingen, Germany) at 1.2 × 108 cells/well. After 1 h, non-adherent cells were removed at 37°C, whereas adherent cells were fed with RPMI-1640 containing 2.5% autologous serum, 300 IU/ml Granulocyte Macrophage Colony-Stimulation Factor (GM-CSF) and 300 IU/ml IL-4 (R&D) and cultured for 6 days. Cells were substituted with fresh medium and cytokines on day 2 and 4 of culture. On day 6, immature DCs (iDCs) were put on ice and harvested by gentle mechanical detachment, washed, counted and characterised by flow cytometry.
Five aliquots of iDCs were cultured for another 2 days (day 6–8) applying a different maturation protocol to each culture as described below. As a control, iDCs were cultured for another 2 days in RPMI-1640 supplemented with 2.5% autologous serum, 300 IU/ml GM-CSF and 300 IU/ml IL-4.
For maturation by CpG sequences (protocol 1), the previously described sequence of CpG-ODN 2006 (24-mer), 5′-TCGTCGTTTTGTCGTTTTGTCGTT-3′ (10,15) and the corresponding non-CpG control oligonucleotide (24-mer), 5′-TGCTGCTTTTGTGCTTTTGTGCTT-3′ of unmethylated and PTO-stabilised single-stranded DNA (MWG-Biotech AG; Ebersberg, Germany) were used. Day 6 cultures were supplemented with 2 μg/ml CpG-ODN for 2 days.
For maturation by CD40 ligation (protocol 2), running cultures of day 6 iDC were supplemented with soluble CD40-L (Alexis; Gruenberg, Germany) at 0.05 μg/ml for the last 24 h of culture. CD40-ligation was always performed in the presence of enhancer (Alexis) at a final concentration of 1 μg/ml.
For maturation by ICC (protocol 3), day 6 cultures were supplemented with IL-1β (10 ng/ml), IL-6 (5 ng/ml), TNF-α (10 ng/ml) (Strathmann Biotech; Hamburg, Germany) and PGE2 (1 μg/ml) (ICN Biomedicals; Eschwege, Germany) for 2 days (11).
For maturation by a combination of ICC and CD40-L (protocol 4), day 6 cultures were supplemented with ICC (as described above). Soluble CD40-L was added at 0.05 μg/ml for the last 24 h.
In protocol 5, day 6 cultures were supplemented with ICC and CD40-L (as described in protocol 4) in combination with CpG-ODN (2 μg/ml) for 2 days.
Cell surface immunophenotyping by flow cytometry
For one- and two-colour flow cytometry, a total of at least 1 × 105 cells were labelled with either FITC or PE-conjugated mAbs to CD3, CD4, CD14, CD16, CD19, CD58, CD80, CD83, CD86, CD154, BDCA2, HLA-DR or the respective isotype-matched control immunoglobulins (Igs) for 30 min at 4°C (all mAbs at 5–10 μg/ml). Unconjugated bcl-XL and bcl-2 primary antibodies and FITC or PE-labelled secondary mAbs were used after permeabilisation of cells according to manufacturer’s instructions. Cells were washed twice with PBS and immediately analyzed on a PAS-III flow cytometer (Partec GmbH; Munster, Germany) equipped with a 488-nm argon laser. Dead cells and debris were gated out on the basis of their light scatter properties. At least 5.000–10.000 gated events were aquired using a winlist software (version 3.0, Varity Software House; Topsham, ME, USA).
RNA extraction and cDNA synthesis
RNA was extracted from approximately 1 × 106 DC at day 6 for iDC and at day 8 of culture for mDC according to the manufacturer’s instructions (RNeasy, Qiagen; Hilden, Germany). RNA was eluted in 30 μl RNase-free water (0.1% diethylpyrocarbonate; Sigma, St. Louis, Missouri, USA). RNA content was determined spectrophotometrically. Subsequent to a DNase digestion step (FCLP-pure DNase I; Pharmacia; Freiburg, Germany), 1 μg RNA was reverse transcribed in a total volume of 50 μl RNase-free water including 1.5 μm p(dT)12–18 (Pharmacia), 0.4 mm of each dNTP (MBI Fermentas; St Leon Rot, Germany) and 200 U reverse transcriptase (Superscript II, Life Technologies; Eggenstein, Germany).
Quantitative real-time RT-PCR for immunoregulatory and chemotactic cytokines
Quantitative real-time Reverse Transcription Polymerase Chain Reaction (RT-PCR) was performed using the LightCycler® PCR system (Roche diagnostics; Mannheim, Germany) for 40 cycles with external standards as described (16,17). To exclude primer–dimer artifacts, fluorescence emitted by the intercalating SYBR Green dye (Roche diagnostics) was measured at the end of a postextension detection step (2 s) at a cytokine-specific temperature (as obtained from melting curve analyses) above the melting point of primer–dimers and below the melting point of the specific PCR product. Cytokine values were standardised for β-actin values obtained in the same samples. Primers and real-time PCR conditions for TNF-α, lymphotactin, IL-12p35 and -p40, IL-10, TGF-β and IL-16 were applied as described earlier (16–19).
Values (attomoles/μl) obtained for each cytokine were standardised for β-actin values. Quantitation was performed in duplicate (variation typically < 10%), and negative control reactions without template were always included.
Monitoring of DC survival and bcl-XL/bcl-2 expression after deprivation of cytokines
Bcl-XL and bcl-2 expression of day 8 DCs was measured by intracellular flow cytometry immediately after completion of maturation procedures and again 24 h thereafter. For the assessment of the cell survival, DCs were cultured in RPMI-1640 supplemented with 2.5% autologous serum only (no cytokines added) from day 8 to 11. The DC viability was analysed by trypan blue exclusion before (mature day 8 DCs) and after 3 days of cytokine deprivation at day 11. The extent of necrosis of day 11 DCs was additionally measured by flow cytometry after propidium iodide staining to confirm results from trypan blue exclusion experiments.
Stability of the DC phenotype under the adverse influence of IL-10
Day 8 DCs, either immature or maturated by the five different protocols, were cultured for another 2 days in medium containing GM-CSF and IL-4 with or without the supplementation of 40 ng/ml IL-10 (Peprotech; London, UK). The cells were analyzed on days 8 and 11 by flow cytometry for their CD83 and CD14 expression.
Mixed leukocyte reaction
Allogenic T-cell-enriched fractions were obtained as 1-h non-adherent PBMC from leukaphereses. Further, lymphocyte enrichment (purity 89.7 ± 3.1%) was achieved by rosetting with sheep erythrocytes (Dade Behring; Liederbach, Germany). For additional experiments, MACS with anti-CD4 and anti-CD8 beads was used to separate both T-cell subpopulations with >95% purity. Cells were cultured in triplicate in flat-bottom 96-well microtitre plates (Costar; High Wycombe, UK) in a final volume of 200 μl/well for 4 days and an additional 18 h in the presence of 0.2 μCi/well [3H]thymidine (Amersham Buchler; Freiburg, Germany). T cells were stimulated at 2 × 105 cells with 2 × 104, 1 × 104, 5 × 103 and 2.5 × 103 DCs (DC/T-cell ratios 1:10, 1:20, 1:40 and 1:80). Cells were harvested and [3H]thymidine incorporation was measured by a Matrix 96 direct β-counter (Hewlett–Packard; Meriden, CT, USA) in counts per minute (cpm). DCs or T-cells cultured alone served as controls.
For all outcome measurements, we performed the non-parametric normal rank test of van der Waerden recommended for samples of low size (20). All observed values were ranked and then transformed to an inverse normal score before statistical testing. For each outcome measure, a global test for any difference between results after the different protocols, a test contrasting iDC (=control) with all ICC-containing groups, and finally tests for values after each protocol in contrast to values of iDC were performed.
All results in figures are presented as mean ± SEM or as mean and the observed values. The level of significance was set to α = 0.05. The statistical analysis was performed using statistika© version 7.1 software (Statsoft; Tulsa, OK, USA) or sas 9.1 (SAS Institute Inc.; Cary, NC, USA).
Maturation protocols containing the ICC induce the highest level of DC-costimulation and DC-maturation markers
Monocytes were enriched from PBMCs by plastic adherence (82 ± 5.2% positive for CD14, <1% expressed CD3, CD19, CD16 and CD56) and were subsequently differentiated to iDCs by culturing with GM-CSF and IL-4 for 6 days. Day 6 iDCs showed down-regulation of CD14 to 3.7 ± 3.6% and high levels of HLA-DR expression (94.7 ± 1.6%) which is consistent with the development of an immature DC phenotype. A median fraction of 3.3 ± 1.8% of cells at day 6 expressed the surface marker BDCA2 characteristic for PDC. After having performed the five different maturation-protocols from day 6 to day 8 of culture (immature day 8 DCs served as a control), we documented the stages of DC maturation by analyzing representative surface markers (HLA-DR, CD83, CD86, CD80, CD40 and CD14) by flow cytometry. According to the coexpression of maturation (CD83) and costimulation (CD80) markers, the five protocols could be categorised into two groups (Fig. 1). Cells treated by the control protocol (GMCSF and IL-4) kept their immature phenotype (day 8 iDCs) with the lowest coexpression of CD83/CD80. An average twofold increase of maturation/costimulation markers was induced in cells treated by protocols 1 (CpG-ODN) and 2 (CD40-L). In contrast, cells which had been maturated by protocols containing ICC (protocol 3, 4 and 5) yielded an approximately sevenfold increase in CD83/CD80 coexpression. Day 8 DCs of all protocols showed a comparable median expression of HLA-DR (>91.7%), CD86 (>98.2%), CD40 (>92.8%) and CD14 (<5.6%).
DCs maturated by ICC induce a strong proliferation of allogenic T-cells
In order to analyze the allostimulatory capacity of the generated DCs, we cocultured DCs and allogenic lymphocytes (purity 89.7% ± 3.1%) derived from mismatched leukapheresis donors. The T-cell proliferative response (Fig. 2a and b) clearly correlated with the expression of maturation and costimulation markers induced by the DC maturation protocols. Neither CpG-ODN nor CD40-L could influence the low rate of allogenic T-cell proliferation measured for iDCs. DCs maturated by ICC-containing protocols induced a 2.5- to 3-fold increase in T-cell proliferation. Although statistically not significant, DCs maturated by ICC plus CD40-L induced higher proliferation rates than DCs from other protocols (highest mean proliferation rate). The separation of CD4+ and CD8+ T-cell fractions by MACS sorting enabled a more defined analysis of the proliferative response (Fig. 2c). While both cell fractions showed responses in agreement to unsorted allogenic T-cells (Fig. 2a), CD4+ T-cells proliferation exceeded that of CD8+ cells by up to 40%.
Cytokine production by dendritic cells
Various maturation agents can affect the capacity of DCs to secrete immunomodulatory cytokines (21). With the measurements of cytokine profiles on the mRNA level, we intended to answer the question which cytokines could potentially be secreted after application in the setting of a clinical study. We, therefore, chose DCs at day 8 of culture for the preparation of cDNA and RT-PCR. In a clinical trial, cells would be injected at this time. The cytokine profile of DCs was measured on the mRNA level by quantitative real-time RT-PCR before and immediately after maturation by the different protocols (Fig. 3a and b). Proinflammatory T helper 1 (Th1)-driving cytokines (IL-12p35 and -p40, TNF-α), chemoattractants for T cells and DCs (lymphotactin, IL-16), and cytokines considered to suppress Th1 responses (IL-10, TGF-β) were analyzed.
All protocols (including iDCs) did not induce relevant levels of IL-12p35 mRNA (data not shown). Maturation of DCs with protocols containing ICC (protocols 3–5) clearly downregulated mRNA levels of IL-16, IL-10 and TGF-β in comparison to expression levels in iDCs. TNF-α expression was downregulated to a lesser extent (Fig. 3a). Significant mRNA levels of the immunosuppressive cytokines TGF-β and IL-10 were only induced with CpG-ODN alone.
Interleukin-12p40 expression was strongly upregulated on the mRNA level up to 70-fold by maturation protocols containing ICC. The mean values of lymphotactin mRNA levels also showed an up to 10-fold increase in comparison to iDCs, but, on a whole, the differences were not significant (Fig. 3b).
Dendritic cell maturation with ICC-containing protocols resulted in a cytokine secretion pattern adequate for induction of Th1 immune responses. With these protocols, IL-10 and TGF-β expressions (both suppressive for Th1 responses) were extremely low (Fig. 3a).
DCs maturated by ICC plus CD40-L are most resistant to cell death induced by deprivation of cytokines
The survival and the phenotypic stability of DCs is crucial for maintaining their immunostimulatory capacity after injection into the patient. In vivo the cytokine milieu may be suboptimal and we, therefore, assessed the percentage of DCs surviving an in vitro cytokine wash out interval of 3 days (Fig. 4a).
After deprivation of cytokines, iDCs, DCs maturated by CD40-L and DCs maturated by CpG-ODN showed survival rates of only 2–8%. All DCs obtained with protocols containing ICC showed significantly superior survival rates. Importantly, the maturation by a combination of ICC plus CD40-L resulted in a DC population with a twofold higher death resistance than achieved by ICC alone (Fig. 4a). Moreover, already after completion of the maturation process on day 8 the same protocol had yielded the highest numbers of total cells (results not shown). Additional data for the extent of cell necrosis on day 11 was obtained by flow cytometry after propidium iodide staining and confirmed the differences in cell survival between DCs generated by the different protocols (Fig. 4c). While the differences in numbers between viable cells on day 8 and 11 reach significance because of accumulating cell death over 3 days (Fig. 4a), propidium iodide staining on day 11 gives a snap-shot image of cell necrosis at that moment.
Bcl-XL but not bcl-2 expression correlates with cell survival and is highest after maturation of DCs with ICC plus CD40-L.
Previous work has implicated regulation of DC survival by antiapoptotic molecules such as bcl-XL and the proto-oncogene bcl-2 (22,23). We measured the intracytoplasmic bcl-XL and bcl-2 expression by flow cytometry in order to detect a possible correlation with the observed survival benefit by maturation protocols (Fig. 4b). While neither CD40-L nor ICC had a significant effect on the bcl-XL and bcl-2 expression when measured immediately after completion of maturation procedures at day 8, the expression of bcl-XL was significantly upregulated after another 24 h interval (day 9 of culture) in ICC-maturated DCs. The most significant upregulation of bcl-XL was found after maturation with ICC plus CD40-L and the bcl-XL expression profile closely correlated with the observed death resistance to a cytokine wash out period. In contrast to bcl-XL, bcl-2 expression remained unchanged (results not shown).
Maturation protocols containing ICC induce a stable CD83high/CD14low expression in cultures supplemented with IL-10
Tumor-induced cytokines such as IL-10 have been found to be involved in suppressing effective antitumor functions of DCs (24). Therefore, the stability of immature and mature DCs under the influence of IL-10 was analyzed by measuring the expression of CD14, as a marker of DCs regressing to macrophage-like monocytes (data not shown), and CD83 on DCs maturated by the different protocols before and after exposure to 40 ng/ml IL-10 for 3 days (day 8–11) (Fig. 5a and b). DCs cultured in the absence of IL-10 served as control. A significantly higher stability with an almost unchanged CD83high/CD14low expression could only be shown for maturation protocols containing ICC.
A well characterised protocol utilising ICC for the generation of mature DCs from CD14+ monocytes has been proposed by Jonuleit et al. (11). These authors have convincingly demonstrated the superiority of mature DCs in vivo by comparing the capacity of mature versus iDCs to induce T-cell responses (25). Moreover, a number of clinical trials have provided evidence for the induction of clinical responses in patients with advanced metastatic disease after the application of DCs generated by this protocol (14,26). Immature DCs, however, have only been applied with minor response rates (27–30). Our results confirm the high degree of maturation/costimulation marker expression by DCs after the application of maturation protocols containing ICC. The addition of other maturation factors to ICC did not induce a further upregulation of these markers. In addition, the strong proliferative response of allogenic CD4+ and CD8+ T-cell fractions supported the superiority of DCs matured by protocols containing ICC and correlated with the magnitude of maturation and costimulation marker expression. Although suggestive in single experiments, the addition of CD40-L to the ICC protocol did not lead to a significant further enhancement of alloproliferation. In our hands, iDCs induced only weak proliferative alloresponses, confirming previous reports (31).
It is clear that the cytokine secretion pattern of distinct DC subsets impacts the resulting T-cell response (32). With the measurements of cytokine profiles on the mRNA level, we intended to answer the question which cytokines could potentially be secreted after application in the setting of a clinical study. Therefore, we chose DCs at day 8 of culture for the preparation of cDNA and RT-PCR. In a clinical trial, cells would be injected at this time. We assessed cytokines that enhance Th1 responses, that exert chemotactic activity on T cells, and cytokines known to mediate immunosuppression, e.g. IL-10 and TGF-β. When compared with iDCs, all maturation protocols containing ICC clearly increased the expression of IL-12p40 and lymphotactin, and decreased the levels for IL-10, TGF-β and IL-16. The influence of maturation protocols on TNF-α secretion was not impressive.
The observed shift in the cytokine signature remained unaltered when further stimuli such as CD40-L were added to the ICC protocol. Thus, our results confirm the impaired capability of monocyte-derived DCs to produce bioactive IL-12p70 (33), whereas DCs maturated by protocols containing ICC produce high amounts of IL-12p40 and very little of the immunosuppressive cytokines IL-10 and TGF-β. In recent studies, it became clear that a number of maturation stimuli, especially interaction with Th cells, as a result of CD40–CD40-L interaction, induce the production of bioactive IL-12p70 temporarily. Although CD40-triggering alone provides a sufficient signal for the induction of IL-12p40, effective induction of IL-12p70 depends on the presence of an additional signal that can be provided by at least two of the Th cell-produced cytokines, interferon gamma (IFNγ) or IL-4 (34). PGE2, however, was shown to selectively induce IL-12p40 only and to inhibit IL-12p70 production (35). From our data we conclude that none of the tested protocols induced a sustained production of bioactive IL-12p70 exceeding 48 h after initiation of maturation procedures. The question, whether DCs would still be able to secrete IL-12p70 in the lymph nodes in vivo after encounter with lymphocytes and CD40–CD40-L interaction as a second stimulus after the above mentioned maturation protocols was addressed by Rouas et al. (36). They found that maturation by Polyriboinosinic Polyribocytidylic Acid [poly(I:C)] triggering TLR3 was the only stimulus preserving the ability of mature DCs to secondarily secrete bioactive IL-12.
Interestingly, DCs after ICC-containing protocols are able to secrete increased quantities of lymphotactin (Lptn, also known as XCL1). Lptn was reported to be efficiently chemotactic towards both CD8+ and CD4+ T cells but only modestly towards natural killer (NK) cells (37). Because of its role as a potent costimulatory activity for CD8+ T-cell proliferation and IL-2 secretion (38), several clinical studies have proposed Lptn as an adjuvant for cellular vaccination (39,40).
The longevity of antigen-bearing DCs in vivo after their application to patients seems crucial for their optimal immunostimulatory functions in regional lymph nodes. Labelled mature DCs home to peripheral lymph nodes within 24 h after intralymphatic, subcutaneous or intradermal injection (41). In order to assess cell death resistance, matured DCs were submitted to a 3-day culture period in media that was deprived of exogenous cytokines. Approximately, 95% of iDCs, DCs maturated by CD40-L alone, and DCs maturated by CpG-ODN did not survive. In contrast, a significantly increased cell survival was observed for DCs matured by protocols containing ICC. The results obtained by counting viable trypan blue exclusing cells on day 8 and 11 could be confirmed by analyzing the extent of cell necrosis by staining day 11 DCs with propidium iodide and submitting cells to flow cytometry. Recently, TNF-α could be identified as a key factor for suppression of cell death after withdrawal of human plasma from DC-cultures (42). It was shown that no other single agent contained in ICC is able to induce a death-protective effect comparable to TNF-α (42). Interestingly, in the present work, the highest resistance to cell death was achieved by a combination of ICC and CD40-L, which points to an additive effect of both TNF-α and CD40-L. It is well recognised that CD40-L alone (which belongs to the TNF family) is able to protect certain DCs by blocking lethal signalling triggered by anti-Fas and by decreasing spontaneous apoptotic cell death during in vitro culture (22).
Overall enhanced survival was paralleled by an increased expression of bcl-XL (Fig. 4b). This is in line with earlier reports showing that besides the improved survival of iDCs, CD40-L, TNF-α or TNF-related activation induced cytokine (TRANCE) may also increase the expression of bcl-XL (22,23). Bcl-XL was described to antagonise the ability of bax and bak to release cytochrome-c from the mitochondrion, a process that would induce caspase-mediated cell death (43).
Conflicting data have been reported for the role of bcl-2. More recently, Kim et al. found that antigen-specific immunostimulation by DNA vaccines was optimally enhanced by engineering constructs to coexpress bcl-XL, rather than other antipoptotic proteins such as bcl-2 (44). These data and the report of Hon et al. (45) showing that bcl-XL-deficient DCs are unable to maintain their survival in the draining lymph node, suggest a functionally superior role for the antiapoptotic bcl-XL gene in maintaining DC survival. Indeed, our own studies of the bcl-XL expression after 2 days of maturation by ICC or ICC plus CD40-L demonstrated a slight but not significant upregulation (results not shown). Twenty-four hours later, bcl-XL was highly significantly upregulated in the matured DCs (Fig. 4b) but not bcl-2. Compared with the data obtained in a mouse model, were bcl-XL RNA expression peaked after 12–18 h, this appears to be late. Species differences may be causative as well as the fact that we measured bcl-XL on the protein level. Clearly, the amount of bcl-XL expressed correlated with the percentage of viable DCs, which were counted 2 days later (Fig. 4a).
As the conventional preparation of monocyte-derived DCs by plastic adherence cannot avoid the contamination by a small percentage of PDCs, it appeared reasonable to assess the specific contribution of these cells. A median fraction of 3.3 ± 1.8% BDCA2+ PDCs was measured on day 6 of culture with GM-CSF and IL-4 in three independent experiments with cells from three different healthy leukapheresis donors. IL-3 was shown to promote the viability and cell proliferation of PDCs. This additional viability and proliferation in comparison to medium controls could be antagonised by the addition of IL-4 in a dose dependent manner (46). However, Grouard et al. reported that besides IL-3 only GM-CSF containing cultures of PDCs showed a trend for a higher proliferation after 3 days of cultivation (47). This effect of GM-CSF on PDC can be explained by a high IL-3 and a low GM-CSF receptor expression on their surface and the well-established overlapping activities of GM-CSF and IL-3 on cells expressing both receptors as both receptors share the same receptor signal transduction unit (betac unit) (48,49). Moreover, in contrast to highly homogenous PDC populations, our starting population of monocytes after plastic adherence with 82 ± 5.2% positive for CD14 and less than 1% expressing CD3, CD19, CD16 and CD56 might propagate cell–cell interactions and also paracrine IL-3 effects crucial for the survival of ‘contaminating’ PDCs. Interleukin-3 is known to be produced by a large panel of immune cells. So far, T cells (50,51), NK cells (52), mast cells (53,54) and cells belonging to the monocyte/macrophage lineage (55) have been reported to synthesise IL-3. While PDCs will obviously not find optimal culture conditions in our system, the above mentioned observations might argue for a survival of a small PDC fraction.
CpG-ODN 2006, which was applied for the stimulation of PDCs in our experiments, has been shown to promote their survival, activation and maturation, as well as a considerable secretion of IFN-α (10). Maturation by CpG-ODN was the only protocol inducing increased levels of TGF-β and IL-10, both immunosuppressive cytokines. Interestingly, when CpG-ODN was added to ICC or ICC plus CD40-L, an increase in IL-10 and TGF-β mRNA levels was not observed, suggesting that as long as ICC is applied the presence or activation of contaminating PDCs will not have adverse effects. On the contrary, CpG-ODN, which is known to induce IFN-α-secretion by PDCs (which itself induces inflammatory cytokines resembling ICC), was neither able to induce a significant maturation on its own nor to replace ICC. These in vitro observations do not contradict the recently described effects observed with CpG-ODN injected into patients’ skin. Here, the draining lymph nodes not only seem to show activation of PDCs but also of CD1a+ monocyte-derived DCs, pointing to a secondary effect of activated PDCs in the in vivo situation (B. Sjuijter, et al., Third International Melanoma Congress, 2006, unpublished).
IL-10 secretion by tumor cells was demonstrated to be an important mechanism by which tumor cells can escape immune-recognition and destruction (56). Several reports have demonstrated that the addition of IL-10 to maturing DC cultures inhibits their differentiation and T-cell stimulatory functions, whereas fully matured DCs induce a stable T-cell alloproliferation after co-culturing with IL-10 (57). Our data support these findings, and beyond that demonstrate that all protocols containing ICC induce a stable CD83high/CD14low expression, that is not affected by high concentrations of IL-10. These observations may be essential for future therapeutic approaches utilising antigen-bearing DCs in cancer therapy.
In conclusion, the clinical application of DCs requires a detailed understanding of their functional potential and of the manipulatory techniques for the optimisation of their efficacy in immunotherapy. Protocols for the generation of monocyte-derived DCs and their subsequent maturation show great variation across clinical studies. The data of our experiments allow a detailed head-to-head comparison of phenotypic and functional properties of immature and mature DCs derived by five maturation protocols. Our results support the use of DC maturation protocols containing ICC plus DC40-L in order to generate highly mature, phenotypically stable, cell-death resistant and T-cell stimulatory DCs with a Th1 type inducing cytokine secretion pattern for clinical application in cancer patients. However, further aspects remain to be clarified, including the optimisation of the DCs migratory capacity and optimal DC loading strategies for tumor antigens.
HH was sponsored by a faculty grant of the Georg-August-University for young researchers and by the German Federal Ministry of Research and Technology (InnoNet programme, grant no 16IN0173).
We thank the Department of Transfusion Medicine of the Medical Faculty of the University of Goettingen for preparing the weekly Leukaphereses. We thank Mrs Manuela Lovasz for expert secretarial assistence.
- 20Verteilungsfreie Methoden in der Biostatistik. Berlin: Springer Verlag, 2000., , .