- Email: [email protected]
Dendritic cell-based immunogens for B-cell chronic lymphocytic leukemia
Cancer Letters 245 (2007) 275–283 www.elsevier.com/locate/canlet
Dendritic cell-based immunogens for B-cell chronic lymphocytic leukemia Thomas Allgeier a, Silke Garhammer a, Elfriede No¨ßner a, Ulrich Wahl a, Konrad Kronenberger a, Martin Dreyling b, Michael Hallek b,c, Ralph Mocikat a,* a
GSF-Institut fu¨r Molekulare Immunologie, Marchioninistr. 25, D-81377 Mu¨nchen, Germany b Ludwig-Maximilians-Universita¨t, III. Medizinische Klinik, Mu¨nchen, Germany c Universita¨t zu Ko¨ln, Klinik I fu¨r Innere Medizin, Ko¨ln, Germany
Received 29 July 2005; received in revised form 22 November 2005; accepted 16 January 2006
Abstract Hybrids generated from tumor cells and dendritic cells (DC) have been proposed as tools for treating malignant disease. Here, we study the underlying principles and the feasibility for the adjuvant therapy of human B cell chronic-lymphocytic leukemia (B-CLL). CLL cells and allogeneic DC were only mixed or additionally fused. Using a combination of FACS and fluorescence microscopic analyses, we show that DC–CLL hybrids can be successfully generated. However, fusion frequencies have to be critically evaluated because the number of fused cells is overestimated when based on FACS analyses alone. The capability of activating patients’ PBMC was examined by measuring cytokine secretion in co-culture assays. We made a systematic comparison of the immunostimulatory capacities of different stimulator cell populations, including DC–CLL fusion samples, unfused mixtures of DC and CLL cells as well as DC or tumor cells alone. Surprisingly, even unfused mixtures had a pronounced tumor-directed immunostimulatory effect. This could be explained by the capture of antigens from surrounding leukemia cells by DC during cocultivation. Although fusion frequencies were low, PBMC stimulation was significantly more effective when the mixtures were subjected to cell fusion. The most potent stimulus was provided by DC–CLL fusion samples derived from mature DC, probably due to their enhanced costimulatory capacity. In summary, DC–tumor cell hybrids might be feasible in the treatment of B-CLL. It should be considered that FACS analysis is not sufficient to assess fusion frequencies and that interactions between unfused DC and CLL cells also result in PBMC activation. q 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Antigen presentation; Cell fusion; Tumor vaccination; Low-grade lymphoma; T-cell activation
1. Introduction Experimental strategies for cancer treatment using dendritic cells (DC) as vaccines have attracted much interest in the past years. DC are potent antigenpresenting cells (APC) that are highly effective at * Corresponding author. Tel.: C49 89 7099 313; fax: C49 89 7099 300. E-mail address: [email protected] (R. Mocikat).
stimulating naı¨ve T lymphocytes [1–3]. DC capture exogenous proteins and present antigenic peptides to CD4CT cells in association with major histocompatibility complex (MHC) class II molecules [3]. In contrast, intracellular antigens (Ag) are presented by MHC class I molecules to CD8 C cytotoxic T lymphocytes [4]. It is, however, widely accepted that exogenous proteins can also be directed to the endogenous presentation pathway, a process referred to as cross-presentation [5,6]. DC are further able to
0304-3835/$ - see front matter q 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2006.01.019
276
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
deliver second signals necessary for efficient T-cell stimulation, which are provided by interleukins or costimulatory surface molecules [7]. To elicit specific immune responses against tumorassociated antigens (TAA) DC vaccines were developed that were modified to present tumor-derived peptides in the context of MHC molecules. Being able to provide both signals required for T cell activation, DC-based vaccines overcome the insufficient presentation and co-stimulation exerted by the cancer cells. DC, which can easily be generated from peripheral blood-derived precursors in vitro [8], were pulsed with antigen protein or synthetic peptides [9,10] or transduced with viral vectors encoding TAAs [11,12]. Murine tumor models showed that DC loaded with TAAs induce effective antitumor responses [10,13,14]. However, induction of immunity against a single defined epitope may be ineffective in heterogeneous tumor cell populations and may select for antigen loss mutants. Furthermore, it is often limited by the paucity of known TAAs and it is MHC type-restricted. Vaccination protocols using whole engineered tumor cells circumvent these difficulties because immune responses against multiple antigens including those yet unidentified can be induced. We showed in several in vivo models that polyclonal antitumor responses are superior to immunity against a single antigen [15–19]. DC-based protocols developed to target multiple TAAs include pulsing with tumor lysates [14,20–23], loading with apoptotic [24] or necrotic tumor cells or transfection with RNA from tumor cells [25]. Another approach, which combines the antigen-presenting capacity of DC with the antigenic polyvalency of whole tumor cells is the fusion of the malignant cells with DC. Hybrid cells generated from carcinoma cells or hematological neoplasias were effective in human and in mouse models [26–34]. B cell chronic-lymphocytic leukemia (B-CLL) is a disease which still has a poor prognosis despite recent advances in chemotherapy. Therefore, particularly in younger patients, new therapeutic strategies are warranted. In this paper we show that vaccination with DC fused to B-CLL cells might be a feasible approach for immunotherapy. Using this model, we demonstrate that the fusion frequencies reported earlier for other tumor models might have been overestimated. Nevertheless, peripheral blood mononuclear cells (PBMC) are effectively stimulated in this system against autologous tumor cells, but part of the tumor antigen-directed immunity is due to unfused DC that have taken up tumor-derived antigens.
2. Materials and methods 2.1. Generation of DC PBMC were isolated from healthy donors by Ficoll– Hypaque density-gradient centrifugation. CD14C monocytes were enriched by magnetic cell separation (Miltenyi Biotec, Bergisch-Gladbach, Germany) according to the manufacturer’s protocol to O95% purity, as assessed by flow cytometry. The CD14C monocytes were cultured in 6-well plates with X-VIVO medium (Cambrex, Verviers, Belgium) containing 2 mM L-glutamine, 100 U/ml penicillin and 100 U/ml streptomycin, 1000 U/ml granulocyte–macrophage colony-stimulating factor (Berlex, Richmond, VA, USA), 800 U/ml interleukin- (IL-) 4 (Promocell, Heidelberg, Germany) for 8 days at a density of 1.5!106 cells per ml. On day 4, medium was replaced. For DC maturation, 250 U/ml tumor necrosis factor-a (TNF-a; Promocell), 1 mM/ml prostaglandin E2 (Sigma, Mu¨nchen, Germany), 1000 U/ml IL-6 (R&D Systems, Wiesbaden, Germany) and 20 ng/ml IL-1b (Promocell) were added and the cells were cultured for 36 h. Maturation was assessed by flow cytometry. 2.2. Preparation and fusion of B-CLL cells B-CLL cells were obtained from the peripheral blood of patients after approval by the local ethics committee and informed consent. They were isolated by Ficoll–Hypaque density-gradient centrifugation and frozen for later use. For fusion, CLL cells were thawed and cultured overnight at 37 8C in RPMI 1640 medium (Gibco/BRL, Karlsruhe, Germany) containing 10% human serum, 2 mM L-glutamine and antibiotics. Tumor cells and DC were then mixed in a ratio of 1:1, 3:1 or 5:1 and washed with PBS (Gibco/BRL). After removing the PBS, 1 ml of warm polyethylene glycol (PEG 1500, Boehringer, Mannheim, Germany) was added to the cell pellet and incubated for 1 min while gently moving the tube. Then the suspension was slowly diluted with 2,5 ml PBS over a period of 4 min. The cells were washed and resuspended in RPMI 1640 supplemented with 15% heatinactivated human serum, 2 mM L-glutamine and antibiotics and incubated at 37 8C. In some experiments, an electrofusion was performed using the Multiporator apparatus (Eppendorf, Hamburg, Germany). Up to 106 DC and CLL cells, respectively, were suspended in hypoosmolar buffer and exposed to an electric field at 20–60 V according to the manufacturer’s operating manual. 2.3. FACS analyses For phenotyping, DC were labelled with the following phycoerythrine- (PE-) coupled mouse antibodies (Ab): anti-CD1a (Immunotech, Krefeld, Germany), anti-CD11c (Leinco Technologies, St Louis, USA), anti-CD14 (Immunotech), anti-CD40 (Pharmingen, San Diego, CA, USA), anti-CD54 (Pharmingen), anti-CD80 (Pharmingen), anti-CD83 (Pharmingen), anti-CD86
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
(Pharmingen), anti-HLA-DR (Pharmingen), anti-MNR (Pharmingen). Ab staining was done in PBS containing 2% FCS and 12 mM EDTA for 30 min at 4 8C. In each experiment, isotype controls were used and dead cells were excluded by propidium iodide staining (ICN, Eschwege, Germany). Cells were then analyzed using a FACSCalibur flow cytometer (Becton Dickinson, Heidelberg, Germany). The fusion efficacy was examined by staining of CLL and DC markers. The DC moiety was detected by a biotinylated anti-CD11c Ab followed by streptavidin-FITC, the CLL moiety was detected by a combination of PE-conjugated anti-CD5 and anti-CD19 Ab. 2.4. Immunocytology Alternatively, the fusion frequency was quantitated by fluorescence microscopy. The cells were stained with PEconjugated anti-CD5 and anti-CD19 mouse Ab according to the FACS protocol. Then the cells were fixed on poly-Llysine-coated slides and washed once with Tris-buffered saline pH 7.4. After fixation in acetone for 10 min and a washing step, the cells were incubated with a rat anti-mouse Ab conjugated with Cy3. The slides were washed and incubated with a biotinylated mouse anti-human CD11c Ab, washed again and stained with streptavidin-FITC. To amplify the FITC signal, the slides were then incubated with a biotinylated anti-avidin Ab followed by streptavidin-FITC. Finally, the cells were fixed again in acetone for 10 min. Each step was performed for 40 min in a humidified chamber. In some experiments, DC and CLL cells were labelled with Cell Tracker (Molecular Probes, Eugene, OR, USA) prior to fusion. DC were stained with Cell Tracker Green BODIFY and CLL cells with Cell Tracker Orange CMTMR at a density of 2!106 cells per ml in 2 ml PBS for 40 min at 37 8C. The cells were washed twice with PBS before they were subjected to the fusion protocol. Then they were seeded on poly-Llysine-coated slides and fixed with acetone. About 2000 cells were evaluated in a fluorescence microscope. For confocal microscopy, cells attached to poly-L-lysin-coated slides and stained with Ab as described above were processed on a Leica TCS SP2 confocal microscope. 2.5. Stimulation assay Stimulation of patient PBMC was done as described previously [35]. In brief, 1!105 PBMC were cultured with 5!104, 105 or 2!105 irradiated (100 Gy) DC, CLL cells, a DC/CLL mixture or fusion cells in 96-well plates in 200 ml of RPMI 1640 medium supplemented with 15% heat-inactivated human serum, 2 mM glutamine, antibiotics, 50 U/ml IL-4 and 20 U/ml IL-2. After 8 days, the PBMC were restimulated with 105 irradiated tumor cells in 10 ml fresh medium or with medium alone. On day 13, supernatants were collected and frozen. In some stimulation experiments, DC were fixed with 1% paraformaldehyde (PFA) for 10 min at room temperature prior to or following mixing and fusing with CLL cells. All stimulations were performed at least in triplicate. Statistical
277
analyses were done using Student’s t-test. A different outcome of two experimental groups means that the results are significant with P!0.05. 2.6. TNF-a assay and IFN-g enzyme-linked immunosorbent assay (ELISA) PBMC stimulation was quantitated by measuring the concentrations of TNF-a and interferon-g (IFN-g). Intracellular FACS staining demonstrated that the cytokine was not derived from CLL cells in the stimulation assays but from CD3 C T cells [35]. The TNF-a concentration in the supernatants was measured by analyzing the cytoxicity of the supernatant against WEHI 164. WEHI cells were suspended at a density of 2.4!104 cells per well in 40 ml RPMI 1640 supplemented with L-glutamine and 10% FCS and 2 mg/ml actinomycin D in flat-bottom 96-well plates. Forty microlitres of supernatant from each stimulation assay were transferred to the WEHI suspension. For quantification, a standard titration of TNF-a was included. After incubation for 20 h at 37 8C (6,5% CO2) in a humified chamber, 20 ml of MTS/PMS solution (10 mg/ml in PBS) were added to each well. After another 4 h of incubation, 25 ml of a 10% aqueous SDS solution were added. Then, cell viability as indicated by colour development was measured at 492 nm. For the detection of IFN-g in the supernatant, an IFN-g ELISA kit was used (Becton-Dickinson, Heidelberg, Germany). The ELISA plates were coated overnight with mouse anti-human IFN-g Ab. After incubation with the supernatant, the proteins were detected with a biotinylated anti-human IFN-g Ab and avidine-horseradish peroxidase. The colour reaction was started with o-phenylene-diamine (Sigma, Mu¨nchen, Germany) and H2O2. In each IFN-g ELISA, a standard was used to quantitate the level of IFN-g.
3. Results and discussion 3.1. Generation of DC–CLL hybrids To evaluate the feasibility of DC–tumor cell hybrids for immune modulation in B-CLL patients, leukemic cells from 8 patients were fused to DC using PEG or electrofusion. Patient characteristics are summarized in Table 1. Because DC precursors occur in low frequency in the peripheral blood of leukemic patients (not shown) we used DC from allogeneic healthy donors. Since the hybrid cells will also express those MHC class I and class II molecules which are derived from the CLL cells, tumor peptides will be presented to the patient’s T-lymphocytes even when patient and DC donor are not compatible with regard to their MHC type. Donor DC were extensively characterized by FACS analyses before and after maturation. Representative results are
278
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
Table 1 Characteristics of B-CLL samples used in this study Patient
Age
Sex
Binet stage
WBC
% lymphocytes
% CD5/CD19
% CD3
1 2 3 4 5 6 7 8
54 61 55 84 75 83 45 60
M M M F F M M M
B B B C C B A C
9200 17,100 160,000 49,900 250,000 75,600 29,200 250,000
94 92 96 83 98 90 81 98
90 87 97 93 93 97 82 99
6 7 2 6 6 2 10 1
compiled in Table 2. No alterations of the DC surface markers were observed after the fusion step. Fusion frequencies were examined by FACS analyses using anti-CD11c Ab detecting the DC moiety and a combination of anti-CD5 and anti-CD19 Ab recognizing the CLL moiety. A typical result is shown in Fig. 1a. Since staining with anti-CD19 or anti-CD5 alone was too weak, cells labelled with a single PEconjugated Ab and anti-CD11c-FITC coincided with the lower right quadrant. Therefore, the population in the upper right quadrant represented events involving CD5/CD19 double-positive, i.e. malignant B cells. Double fluorescence was observed at a frequency of 5–6% with slightly increased efficiencies when mature DC (mDC) were used (Fig. 1d). Varying the CLL/DC ratio (1:1, 3:1 or 5:1) did not result in different fusion frequencies. Fusion efficiencies obtained with the electrofusion protocol were in the same range (not shown). However, electrofusion was not further investigated because the absolute number of cells that could be placed into the fusion chamber was not sufficient for the subsequent PBMC stimulation experiments. When DC and CLL cells were mixed in the absence of either PEG or electric pulses, double-positive events were observed in the FACS at an average frequency of about 3% (Fig. 1b and d). This indicated that cell aggregates may have been formed and that FACS analyses may overestimate the fusion efficiency. We therefore examined the fusion success by fluorescence microscopy. DC and tumor cells were separately labelled with different lipophilic dyes before they were combined. Alternatively, the mixed cell populations either treated with PEG or not were stained with the Ab that were also used for FACS analysis. By this procedure we could unambiguously identify true fusion events. The arrow head in Fig. 1c shows an event where a CLL membrane has been integrated into a DC membrane. Such pictures were never seen in samples that were not treated with PEG. Fusion was also
confirmed by confocal microscopy (not shown). As expected, the fusion efficiencies determined by microscopy differed from those obtained by FACS about twofold (Fig. 1d). This shows that only 50% of the double-positive events identified in the FACS are cell fusions, whereas the other half may represent spontaneous cell aggregations. 3.2. Stimulation of patient PBMC by DC/CLL immunogens with and without cell fusion We next tested the potential of the DC / CLL immunogens to induce antitumor responses in stimulation assays in vitro. The stimulation protocol was exactly as described previously [35]. Patients’ PBMC were primed with irradiated CLL cells, unmodified DC, DC/CLL cell mixtures or DC–CLL cell fusion samples. One week later, activation of PBMC against tumor cells was determined by restimulation with unmodified irradiated CLL cells and quantitative determination of TNF-a and IFN-g release. Stimulation indices were calculated as the ratios of cytokine secretion in response to CLL cells after priming with the different stimulators in comparison to priming with CLL cells alone. The stimulation index obtained after priming with CLL cells and restimulation with CLL cells was defined as 1. Results of IFN-g quantitation are summarized in Fig. 2. The pattern of TNF-a release exactly paralleled that of IFN-g secretion. Studies using intracellular FACS have shown that upon Ag presentation by APC, the cytokines Table 2 Expression of selected DC surface markers before (iDC) and after (mDC) maturation
iDC mDC
CD11c
CD80
CD83
CD86
HLA-DR
25.8 (2.1) 57.4 (6.1)
2.3 (0.7) 7.0 (0.9)
0.8 (0.1) 1.6 (0.1)
13.5 (2.3) 26.3 (4.2)
7.6 (1.7) 14.3 (1.2)
The numbers indicate the ratios of the mean fluorescence intensities of the respective markers and the isotype controls. Standard deviations are given in parentheses.
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
279
Fig. 1. Characterization of DC–CLL cell hybrids. The cells were mixed at a ratio of 1:1. (a) FACS analysis of cell populations after PEG fusion of immature DC and CLL cells. Cells were stained with PE-conjugated anti-CD5 and anti-CD19 Ab and with biotinylated anti-CD11c Ab and streptavidin-FITC. The population in the upper right quadrant represents events involving CD5/CD19 double-positive, i.e. malignant B cells, because staining of either CD5 or CD19 alone was too weak so that cells labelled with a single PE-labelled Ab and FITC coincided with the lower right quadrant. (b) FACS analysis of a mixture of immature DC and CLL cells without fusion step. Staining was as described for Fig. 1a. (c) Typical fluorescence microscopy image of a fusion experiment. Staining of cells was as described in Fig. 1a. Brightfield image is shown on the left, the FITC signals are visualized on the intermediate and the PE signals on the right panel. The arrow-head denotes a fusion event where a CLL membrane has been integrated into a DC membrane. Unlabelled cells are peripheral blood cells other than malignant B-lymphocytes. (d) Number of doublepositive events after mixing or fusing DC and CLL cells as determined by FACS and by fluorescence microscopy. Compilation of 21 independent experiments (all patients were examined by FACS and by microscopic analyses).
measured are derived from CD3C cells but not from CD14C or CD19C cells. Omitting T cells in the coculture assays abrogated cytokine release. To assess the allogeneic effect of DC, patient PBMC were primed with immature DC (iDC) from healthy
donors without tumor-specific restimulation (Fig. 2a, second column). There was a clear response compared to experiments where priming and restimulation were done with autologous cells (first column). This allogeneic response induced by DC did not significantly
280
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
and includes no tumor-specific component. However, when priming was done with a mixture of DC and CLL cells, the stimulation indices markedly increased (fourth column). This effect was further enhanced when the DC and CLL cells had been subjected to the fusion procedure (fifth column). The stimulating effect was even more pronounced when mDC were used (Fig. 2b). The immunogenicity of the fusion samples was not a result of the fusion procedure per se because CLL cells exposed to PEG in the absence of DC had no increased effect (not shown). The difference between iDC and mDC might be due to the higher frequency of hybrid cells caused by the increased fusion efficiency of mDC in comparison to iDC (Fig. 1d). However, when fusion frequencies and stimulation indices were analyzed in all individual experiments, no correlation between these parameters was found. Furthermore, the activating potential of fusion samples generated from iDC could not be improved by enriching the hybrid cells through preparative FACS sorting (not shown). Thus, the enhanced activating effect of mDC–CLL hybrids is likely to be attributable to their more potent co-stimulatory capacity rather than to their slightly increased fusion frequency. Accordingly, the allogeneic effect of mDC was also enhanced as compared to iDC (Fig. 2b, first column and Fig. 2a, second column). Fig. 2. Activation of patient PBMC after consecutive incubation with the indicated stimulators. IFN-g secretion was determined after the second stimulation which was performed using tumor cells or medium. Stimulation indices were calculated as described in the text because absolute cytokine levels were consistent within single experiments but varied between different patients. Thus, the results are only comparable between different patients when normalized to restimulation with unmodified tumor cells in the same experiment. The average absolute IFN-g concentration in the read-out with CLL cells to which all stimulations were normalized was 140 pg/ml. Maximum cytokine release as observed after priming with fusion samples was in the range of 1–2 ng/ml. Not all patients could be included in the different experiments because sufficient cell numbers were not always available. There was, however, no correlation between any clinical parameter and the outcome of fusion or stimulation experiments. The error bars indicate standard deviations. Different results were significant at P! 0.05 (Student’s t-test). (a) Stimulatory potential of allogeneic immature DC and immature DC/CLL cell mixtures with or without fusion step. The first column represents the stimulation index obtained after priming with CLL cells and restimulation with CLL cells, which was defined as 1. Average values from 9 experiments are shown. (b) Stimulation indices obtained by co-cultivation experiments using mature DC (nZ5).
differ from the activation achieved by priming with DC and restimulation with tumor cells (Fig. 2a, second vs. third column). This shows that the response observed in the latter setting might be due to the allogeneic effect
3.3. Immunostimulatory mechanism of DC/CLL cell mixtures Priming with DC/CLL cell mixtures yielded higher stimulation indices than priming with DC alone (Fig. 2a and b) suggesting the presence of a tumor-related component in addition to the allogeneic effect. This antitumor response might be explained by uptake of tumor-associated antigens by DC during co-cultivation with tumor cells followed by presentation of antigenic peptides to T lymphocytes. Alternately, it was possible that T cells were stimulated due to a concerted action of CLL cells presenting tumor-derived peptides to T lymphocytes, and DC that simultaneously delivered the requisite co-stimulatory signals. To distinguish between these possibilities, DC were fixed with PFA before mixing or fusing to CLL cells. PFA-treated DC can provide co-stimulatory signals, but they are not capable of internalizing cells, cellular fragments or proteins and directing antigenic peptides to the cell surface [36]. As shown in Fig. 3, the stimulation index was reduced to background levels when PFA-fixed DC were mixed with CLL cells and used for stimulation
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
Fig. 3. Inhibition of PBMC activation after PFA fixation of DC. Before mixing to or fusing with CLL cells, immature DC were treated with PFA. PBMC activation was done as described in Fig. 2, stimulators used for priming and for restimulation are indicated. Standard deviations are shown as error bars.
of patients’ PBMC. Thus, DC/CLL cell mixtures without fusion are able to induce antitumor immune responses because DC capture antigens from CLL cells during co-cultivation. A marked decrease of the immunostimulatory potential was also found for fusion samples that were derived from PFA-treated DC (Fig. 3). This result was predictable because not only unfused DC which are still present at great numbers in the fusion assay but also the hybrid cells need an intact processing machinery for presenting immunogenic peptides. However, cell fusion might lead to integration of tumor cell membranes containing MHC molecules that are already loaded with CLL-derived peptides into the DC membrane. This could explain why the stimulation observed after fusion of CLL cells with PFA-treated DC was not completely abrogated. 3.4. Concluding remarks A crucial point for protocols using DC–tumor cell hybrids is the fusion efficiency. The comparison of FACS data and fluorescence microscopy in our study suggests that the frequency of DC–tumor cell fusion has to be critically assessed. Thus, unfused DC/CLL cell mixtures yielded double-positive events in the FACS which were probably due to cell aggregations. The consequences of this finding have not been considered so far. Therefore, in previous investigations that mostly rely on FACS analyses, fusion efficiencies might have been overestimated. Using cell populations that contained fusion products as well as unfused DC and CLL cells, autologous patient PBMC could readily be stimulated. Because the
281
stimulation indices did not correlate with the frequency of hybrid cells we assume that the maximal stimulation indices we obtained cannot be further increased. This limitation might be explained by the paucity of responder cells in the patients’ peripheral blood and by additional functional defects of T cells. In a recent study, fusions of B-CLL cells with autologous DC sometimes stimulated autologous T cells more efficiently than naı¨ve DC [37]. However, the stimulating effects of unfused cell mixtures were not investigated in that paper. Using a mouse melanoma model another group reported that physical interactions between unfused DC and tumor cells resulted in an immunogen that was equally effective as hybrid cells [38]. We show that DC that are mixed to CLL cells are capable of priming PBMC, but to a lower extent than those samples that had been subjected to cell fusion. Although in our study the fusion frequencies were as low as about 3% and the most part of the immunostimulatory effect seemed to be attributable to unfused DC and tumor cells, a specific effect of cell fusion could be seen. Mature DC proved superior to immature DC. Therefore, despite low fusion frequencies, immunization with mDC–tumor cell hybrids may be a feasible approach for the adjuvant treatment of B-CLL patients.
Acknowledgements We thank Dr D.J. Schendel for her ongoing support and Dr S. King for critically reading the manuscript. Expert technical assistance by A. Brandl is gratefully acknowledged. The work was supported by Else-Kro¨ner-Fresenius-Stiftung, HGF-Strategiefonds ‘Infektionsabwehr und Krebspra¨vention’ and by Nationales Genomforschungsnetz. This work is part of T. Allgeier’s doctoral thesis at the Medical Faculty of the Ludwig-Maximilians-Universita¨t Mu¨nchen. References [1] J. Banchereau, R.M. Steinman, Dendritic cells and the control of immunity, Nature 392 (1998) 245–252. [2] J. Banchereau, F. Brie`re, C. Caux, J. Davoust, S. Lebecque, Y.J. Liu, et al., Immunobiology of dendritic cells, Annu. Rev. Immunol. 18 (2000) 767–811. [3] R.M. Steinman, The dendritic cell system and its role in immunogenicity, Annu. Rev. Immunol. 9 (1991) 271–296.
282
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
[4] J.W. Yewdell, J.R. Bennink, Cell biology of antigen processing and presentation to MHC class I molecule restricted T lymphocytes, Adv. Immunol. 52 (1992) 1–123. [5] F.R. Carbone, M.J. Bevan, Class I-restricted processing and presentation of exogenous cell-associated antigen in vivo, J. Exp. Med. 171 (1990) 377–387. [6] W.R. Heath, F.R. Carbone, Cross-presentation, dendritic cells, tolerance and immunity, Annu. Rev. Immunol. 19 (2001) 47–64. [7] R.M. Steinman, M. Witmer-Pack, K. Inaba, Dendritic cells: antigen presentation, accessory function and clinical relevance, Adv. Exp. Med. Biol. 329 (1993) 1–9. [8] N. Romani, S. Gruner, D. Brang, E. Kampgen, A. Lenz, B. Trockenbacher, et al., Proliferating dendritic cell progenitors in human blood, J. Exp. Med. 180 (1994) 83–93. [9] M.A. Ossevoort, M.C. Feltkamp, K.J. van Veen, C.J. Melief, W.M. Kast, Dendritic cells as carriers for a cytotoxic T-lymphocyte epitope-based peptide vaccine in protection against a human papillomavirus type 16-induced tumor, J. Immunother. 18 (1995) 86–94. [10] J.I. Mayordomo, T. Zorina, W.J. Storkus, L. Zitvogel, C. Celluzzi, L.D. Falo, et al., Bone marrow-derived dendritic cells pulsed with synthetic tumour peptides elicit protective and therapeutic antitumour immunity, Nat. Med. 1 (1995) 1297– 1302. [11] J. Gong, L. Chen, D. Chen, M. Kashiwaba, Y. Manome, T. Tanaka, et al., Induction of antigen-specific antitumor immunity with adenovirus-transduced dendritic cells, Gene Ther. 4 (1997) 1023–1028. [12] A.L. Ribas, H. Butterfield, W.H. McBride, V.B. Dissette, A. Koh, C.M. Vollmer, et al., Characterization of antitumor immunization to a defined melanoma antigen using genetically engineered murine dendritic cells, Cancer Gene Ther. 6 (1999) 523–536. [13] L. Zitvogel, J.I. Mayordomo, T. Tjandrawan, A.B. DeLeo, M.R. Clarke, M.T. Lotze, et al., Therapy of murine tumors with tumor peptide-pulsed dendritic cells: dependence on T cells, B7 costimulation, and T helper cell 1-associated cytokines, J. Exp. Med. 183 (1996) 87–97. [14] R.C. Fields, K. Shimizu, J.J. Mule, Murine dendritic cells pulsed with whole tumor lysates mediate potent antitumor immune responses in vitro and in vivo, Proc. Natl Acad. Sci. USA 95 (1998) 9482–9487. [15] R. Mocikat, M. Selmayr, S. Thierfelder, H. Lindhofer, Triomabased vaccination against B cell lymphoma confers long-lasting tumor immunity, Cancer Res. 57 (1997) 2346–2349. [16] M. Selmayr, J. Strehl, J.-P. Kremer, E. Kremmer, A. Doenecke, M. Hallek, et al., Induction of tumor immunity by autologous B lymphoma cells expressing a genetically engineered idiotype, Gene Ther. 6 (1999) 778–784. [17] J. Strehl, M. Selmayr, J.-P. Kremer, L. Hu¨ltner, H. Lindhofer, R. Mocikat, Gene therapy of B-cell lymphoma with cytokine gene-modified trioma cells, Int. J. Cancer 83 (1999) 113–120. [18] M. Selmayr, H. Menzel, J.-P. Kremer, S. Thierfelder, R. Mocikat, B cell lymphoma idiotypes chimerized by gene targeting can induce tumor immunity, Cancer Gene Ther. 7 (2000) 501–506. [19] K. Kronenberger, A. Dieckmann, M. Selmayr, J. Strehl, U. Wahl, H. Lindhofer, et al., Impact of the lymphoma idiotype on in vivo tumor protection in a vaccination model based on targeting antigens to antigen-presenting cells, Blood 99 (2002) 1327–1331.
[20] L.A. Lambert, G.R. Gibson, M. Maloney, B. Durell, R.J. Noelle, R.J. Barth, Intranodal immunization with tumor lysate-pulsed dendritic cells enhances protective antitumor immunity, Cancer Res. 61 (2001) 641–646. [21] Z. Abdel-Wahab, P. DeMatos, D. Hester, X.D. Dong, H.F. Seigler, Human dendritic cells, pulsed with either melanoma tumor cell lysates or the gp100 peptide (280–288), induce pairs of T-cell cultures with similar phenotype and lytic activity, Cell. Immunol. 186 (1998) 63–74. [22] S.K. Nair, D. Snyder, B.T. Rouse, E. Gilboa, Regression of tumors in mice vaccinated with professional antigen-presenting cells pulsed with tumor extracts, Int. J. Cancer 70 (1997) 706–715. [23] M. Schnurr, P. Galambos, C. Scholz, F. Then, M. Dauer, S. Endres, et al., Tumor cell lysate-pulsed human dendritic cells induce a T-cell response against pancreatic carcinoma cells: an in vitro model for the assessment of tumor vaccines, Cancer Res. 61 (2001) 6445–6450. [24] M.L. Albert, B. Sauter, N. Bhardwaj, Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs, Nature 392 (1998) 86–89. [25] L.M. Liau, K.L. Black, R.M. Prins, S.N. Sykes, P.L. DiPatre, T.F. Cloughesy, et al., Treatment of intracranial gliomas with bone marrow-derived dendritic cells pulsed with tumor antigens, J. Neurosurg. 90 (1999) 1115–1124. [26] J. Gong, N. Nikrui, D. Chen, S. Koido, Z. Wu, Y. Tanaka, et al., Fusions of human ovarian carcinoma cells with autologous or allogeneic dendritic cells induce antitumor immunity, J. Immunol. 165 (2000) 1705–1711. [27] J. Gong, D. Avigan, D. Chen, Z. Wu, S. Koido, M. Kashiwaba, et al., Activation of antitumor cytotoxic T lymphocytes by fusions of human dendritic cells and breast carcinoma cells, Proc. Natl Acad. Sci. USA 97 (2000) 2715–2718. [28] J. Gong, D. Chen, D. Kufe, Induction of antitumor activity by immunization with fusions of dendritic and carcinoma cells, Nat. Med. 3 (1997) 558–561. [29] L. Lespagnard, P. Mettens, A.-M. Verheyden, N. Tasiaux, K. Thielemans, S. van Meirvenne, et al., Dendritic cells fused with mastocytoma cells elicit therapeutic antitumor immunity, Int. J. Cancer 76 (1998) 250–258. [30] J. Wang, S. Saffold, X. Cao, J. Krauss, W. Chen, Eliciting T cell immunity against poorly immunogenic tumors by immunization with dendritic cell-tumor fusion vaccines, J. Immunol. 161 (1998) 5516–5524. [31] J. Gong, S. Koido, D. Chen, Y. Tanaka, L. Huang, D. Avigan, et al., Immunization against murine multiple myeloma with fusions of dendritic cells and plasmacytoma cells is potentiated by interleukin-12, Blood 99 (2002) 2512–2517. [32] J. Wang, S. Saffold, X. Cao, J. Krauss, W. Chen, Eliciting T cell immunity against poorly immunogenic tumors by immunization with dendritic cell–tumor fusion vaccines, J. Immunol. 161 (1998) 5516–5524. [33] J. Gong, D. Chen, M. Kashiwaba, Y. Li, L. Chen, H. Takeuchi, et al., Reversal of tolerance to human MUC1 antigen in MUC1 transgenic mice immunized with fusions of dendritic and carcinoma cells, Proc. Natl Acad. Sci. USA 95 (1998) 6279–6283. [34] M. Lindner, V. Schirrmacher, Tumour cell–dendritic cell fusion for cancer immunotherapy: comparison of therapeutic efficiency of polyethylen–glycol versus electro-fusion protocols, Eur. J. Clin. Invest. 32 (2002) 207–217.
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283 [35] U. Wahl, E. No¨ßner, K. Kronenberger, R. Gangnus, H. Pohla, M.S. Staege, et al., Vaccination against B-cell chronic lymphocytic leukemia with trioma cells: preclinical evaluation, Clin. Cancer Res. 9 (2003) 4240–4246. [36] E. No¨ßner, R. Gastpar, V. Milani, A. Brandl, P.J.S. Hutzler, M.C. Kuppner, et al., Tumor-derived heat shock protein 70 peptide complexes are cross-presented by human dendritic cells, J. Immunol. 169 (2002) 5424–5432.
283
[37] P. Kokhaei, M.R. Rezvany, L. Virving, A. Choudhury, H. Rabbani, ¨ sterborg, et al., Dendritic cells loaded with apoptotic tumour A. O cells induce a stronger T-cell response than dendritic cell–tumour hybrids in B-CLL, Leukemia 17 (2003) 894–899. [38] C.M. Celluzzi, L.D. Falo, Physical interaction between dendritic cells and tumor cells results in an immunogen that induces protective and therapeutic tumor rejection, J. Immunol. 160 (1998) 3081–3085.
Dendritic cell-based immunogens for B-cell chronic lymphocytic leukemia Thomas Allgeier a, Silke Garhammer a, Elfriede No¨ßner a, Ulrich Wahl a, Konrad Kronenberger a, Martin Dreyling b, Michael Hallek b,c, Ralph Mocikat a,* a
GSF-Institut fu¨r Molekulare Immunologie, Marchioninistr. 25, D-81377 Mu¨nchen, Germany b Ludwig-Maximilians-Universita¨t, III. Medizinische Klinik, Mu¨nchen, Germany c Universita¨t zu Ko¨ln, Klinik I fu¨r Innere Medizin, Ko¨ln, Germany
Received 29 July 2005; received in revised form 22 November 2005; accepted 16 January 2006
Abstract Hybrids generated from tumor cells and dendritic cells (DC) have been proposed as tools for treating malignant disease. Here, we study the underlying principles and the feasibility for the adjuvant therapy of human B cell chronic-lymphocytic leukemia (B-CLL). CLL cells and allogeneic DC were only mixed or additionally fused. Using a combination of FACS and fluorescence microscopic analyses, we show that DC–CLL hybrids can be successfully generated. However, fusion frequencies have to be critically evaluated because the number of fused cells is overestimated when based on FACS analyses alone. The capability of activating patients’ PBMC was examined by measuring cytokine secretion in co-culture assays. We made a systematic comparison of the immunostimulatory capacities of different stimulator cell populations, including DC–CLL fusion samples, unfused mixtures of DC and CLL cells as well as DC or tumor cells alone. Surprisingly, even unfused mixtures had a pronounced tumor-directed immunostimulatory effect. This could be explained by the capture of antigens from surrounding leukemia cells by DC during cocultivation. Although fusion frequencies were low, PBMC stimulation was significantly more effective when the mixtures were subjected to cell fusion. The most potent stimulus was provided by DC–CLL fusion samples derived from mature DC, probably due to their enhanced costimulatory capacity. In summary, DC–tumor cell hybrids might be feasible in the treatment of B-CLL. It should be considered that FACS analysis is not sufficient to assess fusion frequencies and that interactions between unfused DC and CLL cells also result in PBMC activation. q 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Antigen presentation; Cell fusion; Tumor vaccination; Low-grade lymphoma; T-cell activation
1. Introduction Experimental strategies for cancer treatment using dendritic cells (DC) as vaccines have attracted much interest in the past years. DC are potent antigenpresenting cells (APC) that are highly effective at * Corresponding author. Tel.: C49 89 7099 313; fax: C49 89 7099 300. E-mail address: [email protected] (R. Mocikat).
stimulating naı¨ve T lymphocytes [1–3]. DC capture exogenous proteins and present antigenic peptides to CD4CT cells in association with major histocompatibility complex (MHC) class II molecules [3]. In contrast, intracellular antigens (Ag) are presented by MHC class I molecules to CD8 C cytotoxic T lymphocytes [4]. It is, however, widely accepted that exogenous proteins can also be directed to the endogenous presentation pathway, a process referred to as cross-presentation [5,6]. DC are further able to
0304-3835/$ - see front matter q 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2006.01.019
276
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
deliver second signals necessary for efficient T-cell stimulation, which are provided by interleukins or costimulatory surface molecules [7]. To elicit specific immune responses against tumorassociated antigens (TAA) DC vaccines were developed that were modified to present tumor-derived peptides in the context of MHC molecules. Being able to provide both signals required for T cell activation, DC-based vaccines overcome the insufficient presentation and co-stimulation exerted by the cancer cells. DC, which can easily be generated from peripheral blood-derived precursors in vitro [8], were pulsed with antigen protein or synthetic peptides [9,10] or transduced with viral vectors encoding TAAs [11,12]. Murine tumor models showed that DC loaded with TAAs induce effective antitumor responses [10,13,14]. However, induction of immunity against a single defined epitope may be ineffective in heterogeneous tumor cell populations and may select for antigen loss mutants. Furthermore, it is often limited by the paucity of known TAAs and it is MHC type-restricted. Vaccination protocols using whole engineered tumor cells circumvent these difficulties because immune responses against multiple antigens including those yet unidentified can be induced. We showed in several in vivo models that polyclonal antitumor responses are superior to immunity against a single antigen [15–19]. DC-based protocols developed to target multiple TAAs include pulsing with tumor lysates [14,20–23], loading with apoptotic [24] or necrotic tumor cells or transfection with RNA from tumor cells [25]. Another approach, which combines the antigen-presenting capacity of DC with the antigenic polyvalency of whole tumor cells is the fusion of the malignant cells with DC. Hybrid cells generated from carcinoma cells or hematological neoplasias were effective in human and in mouse models [26–34]. B cell chronic-lymphocytic leukemia (B-CLL) is a disease which still has a poor prognosis despite recent advances in chemotherapy. Therefore, particularly in younger patients, new therapeutic strategies are warranted. In this paper we show that vaccination with DC fused to B-CLL cells might be a feasible approach for immunotherapy. Using this model, we demonstrate that the fusion frequencies reported earlier for other tumor models might have been overestimated. Nevertheless, peripheral blood mononuclear cells (PBMC) are effectively stimulated in this system against autologous tumor cells, but part of the tumor antigen-directed immunity is due to unfused DC that have taken up tumor-derived antigens.
2. Materials and methods 2.1. Generation of DC PBMC were isolated from healthy donors by Ficoll– Hypaque density-gradient centrifugation. CD14C monocytes were enriched by magnetic cell separation (Miltenyi Biotec, Bergisch-Gladbach, Germany) according to the manufacturer’s protocol to O95% purity, as assessed by flow cytometry. The CD14C monocytes were cultured in 6-well plates with X-VIVO medium (Cambrex, Verviers, Belgium) containing 2 mM L-glutamine, 100 U/ml penicillin and 100 U/ml streptomycin, 1000 U/ml granulocyte–macrophage colony-stimulating factor (Berlex, Richmond, VA, USA), 800 U/ml interleukin- (IL-) 4 (Promocell, Heidelberg, Germany) for 8 days at a density of 1.5!106 cells per ml. On day 4, medium was replaced. For DC maturation, 250 U/ml tumor necrosis factor-a (TNF-a; Promocell), 1 mM/ml prostaglandin E2 (Sigma, Mu¨nchen, Germany), 1000 U/ml IL-6 (R&D Systems, Wiesbaden, Germany) and 20 ng/ml IL-1b (Promocell) were added and the cells were cultured for 36 h. Maturation was assessed by flow cytometry. 2.2. Preparation and fusion of B-CLL cells B-CLL cells were obtained from the peripheral blood of patients after approval by the local ethics committee and informed consent. They were isolated by Ficoll–Hypaque density-gradient centrifugation and frozen for later use. For fusion, CLL cells were thawed and cultured overnight at 37 8C in RPMI 1640 medium (Gibco/BRL, Karlsruhe, Germany) containing 10% human serum, 2 mM L-glutamine and antibiotics. Tumor cells and DC were then mixed in a ratio of 1:1, 3:1 or 5:1 and washed with PBS (Gibco/BRL). After removing the PBS, 1 ml of warm polyethylene glycol (PEG 1500, Boehringer, Mannheim, Germany) was added to the cell pellet and incubated for 1 min while gently moving the tube. Then the suspension was slowly diluted with 2,5 ml PBS over a period of 4 min. The cells were washed and resuspended in RPMI 1640 supplemented with 15% heatinactivated human serum, 2 mM L-glutamine and antibiotics and incubated at 37 8C. In some experiments, an electrofusion was performed using the Multiporator apparatus (Eppendorf, Hamburg, Germany). Up to 106 DC and CLL cells, respectively, were suspended in hypoosmolar buffer and exposed to an electric field at 20–60 V according to the manufacturer’s operating manual. 2.3. FACS analyses For phenotyping, DC were labelled with the following phycoerythrine- (PE-) coupled mouse antibodies (Ab): anti-CD1a (Immunotech, Krefeld, Germany), anti-CD11c (Leinco Technologies, St Louis, USA), anti-CD14 (Immunotech), anti-CD40 (Pharmingen, San Diego, CA, USA), anti-CD54 (Pharmingen), anti-CD80 (Pharmingen), anti-CD83 (Pharmingen), anti-CD86
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
(Pharmingen), anti-HLA-DR (Pharmingen), anti-MNR (Pharmingen). Ab staining was done in PBS containing 2% FCS and 12 mM EDTA for 30 min at 4 8C. In each experiment, isotype controls were used and dead cells were excluded by propidium iodide staining (ICN, Eschwege, Germany). Cells were then analyzed using a FACSCalibur flow cytometer (Becton Dickinson, Heidelberg, Germany). The fusion efficacy was examined by staining of CLL and DC markers. The DC moiety was detected by a biotinylated anti-CD11c Ab followed by streptavidin-FITC, the CLL moiety was detected by a combination of PE-conjugated anti-CD5 and anti-CD19 Ab. 2.4. Immunocytology Alternatively, the fusion frequency was quantitated by fluorescence microscopy. The cells were stained with PEconjugated anti-CD5 and anti-CD19 mouse Ab according to the FACS protocol. Then the cells were fixed on poly-Llysine-coated slides and washed once with Tris-buffered saline pH 7.4. After fixation in acetone for 10 min and a washing step, the cells were incubated with a rat anti-mouse Ab conjugated with Cy3. The slides were washed and incubated with a biotinylated mouse anti-human CD11c Ab, washed again and stained with streptavidin-FITC. To amplify the FITC signal, the slides were then incubated with a biotinylated anti-avidin Ab followed by streptavidin-FITC. Finally, the cells were fixed again in acetone for 10 min. Each step was performed for 40 min in a humidified chamber. In some experiments, DC and CLL cells were labelled with Cell Tracker (Molecular Probes, Eugene, OR, USA) prior to fusion. DC were stained with Cell Tracker Green BODIFY and CLL cells with Cell Tracker Orange CMTMR at a density of 2!106 cells per ml in 2 ml PBS for 40 min at 37 8C. The cells were washed twice with PBS before they were subjected to the fusion protocol. Then they were seeded on poly-Llysine-coated slides and fixed with acetone. About 2000 cells were evaluated in a fluorescence microscope. For confocal microscopy, cells attached to poly-L-lysin-coated slides and stained with Ab as described above were processed on a Leica TCS SP2 confocal microscope. 2.5. Stimulation assay Stimulation of patient PBMC was done as described previously [35]. In brief, 1!105 PBMC were cultured with 5!104, 105 or 2!105 irradiated (100 Gy) DC, CLL cells, a DC/CLL mixture or fusion cells in 96-well plates in 200 ml of RPMI 1640 medium supplemented with 15% heat-inactivated human serum, 2 mM glutamine, antibiotics, 50 U/ml IL-4 and 20 U/ml IL-2. After 8 days, the PBMC were restimulated with 105 irradiated tumor cells in 10 ml fresh medium or with medium alone. On day 13, supernatants were collected and frozen. In some stimulation experiments, DC were fixed with 1% paraformaldehyde (PFA) for 10 min at room temperature prior to or following mixing and fusing with CLL cells. All stimulations were performed at least in triplicate. Statistical
277
analyses were done using Student’s t-test. A different outcome of two experimental groups means that the results are significant with P!0.05. 2.6. TNF-a assay and IFN-g enzyme-linked immunosorbent assay (ELISA) PBMC stimulation was quantitated by measuring the concentrations of TNF-a and interferon-g (IFN-g). Intracellular FACS staining demonstrated that the cytokine was not derived from CLL cells in the stimulation assays but from CD3 C T cells [35]. The TNF-a concentration in the supernatants was measured by analyzing the cytoxicity of the supernatant against WEHI 164. WEHI cells were suspended at a density of 2.4!104 cells per well in 40 ml RPMI 1640 supplemented with L-glutamine and 10% FCS and 2 mg/ml actinomycin D in flat-bottom 96-well plates. Forty microlitres of supernatant from each stimulation assay were transferred to the WEHI suspension. For quantification, a standard titration of TNF-a was included. After incubation for 20 h at 37 8C (6,5% CO2) in a humified chamber, 20 ml of MTS/PMS solution (10 mg/ml in PBS) were added to each well. After another 4 h of incubation, 25 ml of a 10% aqueous SDS solution were added. Then, cell viability as indicated by colour development was measured at 492 nm. For the detection of IFN-g in the supernatant, an IFN-g ELISA kit was used (Becton-Dickinson, Heidelberg, Germany). The ELISA plates were coated overnight with mouse anti-human IFN-g Ab. After incubation with the supernatant, the proteins were detected with a biotinylated anti-human IFN-g Ab and avidine-horseradish peroxidase. The colour reaction was started with o-phenylene-diamine (Sigma, Mu¨nchen, Germany) and H2O2. In each IFN-g ELISA, a standard was used to quantitate the level of IFN-g.
3. Results and discussion 3.1. Generation of DC–CLL hybrids To evaluate the feasibility of DC–tumor cell hybrids for immune modulation in B-CLL patients, leukemic cells from 8 patients were fused to DC using PEG or electrofusion. Patient characteristics are summarized in Table 1. Because DC precursors occur in low frequency in the peripheral blood of leukemic patients (not shown) we used DC from allogeneic healthy donors. Since the hybrid cells will also express those MHC class I and class II molecules which are derived from the CLL cells, tumor peptides will be presented to the patient’s T-lymphocytes even when patient and DC donor are not compatible with regard to their MHC type. Donor DC were extensively characterized by FACS analyses before and after maturation. Representative results are
278
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
Table 1 Characteristics of B-CLL samples used in this study Patient
Age
Sex
Binet stage
WBC
% lymphocytes
% CD5/CD19
% CD3
1 2 3 4 5 6 7 8
54 61 55 84 75 83 45 60
M M M F F M M M
B B B C C B A C
9200 17,100 160,000 49,900 250,000 75,600 29,200 250,000
94 92 96 83 98 90 81 98
90 87 97 93 93 97 82 99
6 7 2 6 6 2 10 1
compiled in Table 2. No alterations of the DC surface markers were observed after the fusion step. Fusion frequencies were examined by FACS analyses using anti-CD11c Ab detecting the DC moiety and a combination of anti-CD5 and anti-CD19 Ab recognizing the CLL moiety. A typical result is shown in Fig. 1a. Since staining with anti-CD19 or anti-CD5 alone was too weak, cells labelled with a single PEconjugated Ab and anti-CD11c-FITC coincided with the lower right quadrant. Therefore, the population in the upper right quadrant represented events involving CD5/CD19 double-positive, i.e. malignant B cells. Double fluorescence was observed at a frequency of 5–6% with slightly increased efficiencies when mature DC (mDC) were used (Fig. 1d). Varying the CLL/DC ratio (1:1, 3:1 or 5:1) did not result in different fusion frequencies. Fusion efficiencies obtained with the electrofusion protocol were in the same range (not shown). However, electrofusion was not further investigated because the absolute number of cells that could be placed into the fusion chamber was not sufficient for the subsequent PBMC stimulation experiments. When DC and CLL cells were mixed in the absence of either PEG or electric pulses, double-positive events were observed in the FACS at an average frequency of about 3% (Fig. 1b and d). This indicated that cell aggregates may have been formed and that FACS analyses may overestimate the fusion efficiency. We therefore examined the fusion success by fluorescence microscopy. DC and tumor cells were separately labelled with different lipophilic dyes before they were combined. Alternatively, the mixed cell populations either treated with PEG or not were stained with the Ab that were also used for FACS analysis. By this procedure we could unambiguously identify true fusion events. The arrow head in Fig. 1c shows an event where a CLL membrane has been integrated into a DC membrane. Such pictures were never seen in samples that were not treated with PEG. Fusion was also
confirmed by confocal microscopy (not shown). As expected, the fusion efficiencies determined by microscopy differed from those obtained by FACS about twofold (Fig. 1d). This shows that only 50% of the double-positive events identified in the FACS are cell fusions, whereas the other half may represent spontaneous cell aggregations. 3.2. Stimulation of patient PBMC by DC/CLL immunogens with and without cell fusion We next tested the potential of the DC / CLL immunogens to induce antitumor responses in stimulation assays in vitro. The stimulation protocol was exactly as described previously [35]. Patients’ PBMC were primed with irradiated CLL cells, unmodified DC, DC/CLL cell mixtures or DC–CLL cell fusion samples. One week later, activation of PBMC against tumor cells was determined by restimulation with unmodified irradiated CLL cells and quantitative determination of TNF-a and IFN-g release. Stimulation indices were calculated as the ratios of cytokine secretion in response to CLL cells after priming with the different stimulators in comparison to priming with CLL cells alone. The stimulation index obtained after priming with CLL cells and restimulation with CLL cells was defined as 1. Results of IFN-g quantitation are summarized in Fig. 2. The pattern of TNF-a release exactly paralleled that of IFN-g secretion. Studies using intracellular FACS have shown that upon Ag presentation by APC, the cytokines Table 2 Expression of selected DC surface markers before (iDC) and after (mDC) maturation
iDC mDC
CD11c
CD80
CD83
CD86
HLA-DR
25.8 (2.1) 57.4 (6.1)
2.3 (0.7) 7.0 (0.9)
0.8 (0.1) 1.6 (0.1)
13.5 (2.3) 26.3 (4.2)
7.6 (1.7) 14.3 (1.2)
The numbers indicate the ratios of the mean fluorescence intensities of the respective markers and the isotype controls. Standard deviations are given in parentheses.
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
279
Fig. 1. Characterization of DC–CLL cell hybrids. The cells were mixed at a ratio of 1:1. (a) FACS analysis of cell populations after PEG fusion of immature DC and CLL cells. Cells were stained with PE-conjugated anti-CD5 and anti-CD19 Ab and with biotinylated anti-CD11c Ab and streptavidin-FITC. The population in the upper right quadrant represents events involving CD5/CD19 double-positive, i.e. malignant B cells, because staining of either CD5 or CD19 alone was too weak so that cells labelled with a single PE-labelled Ab and FITC coincided with the lower right quadrant. (b) FACS analysis of a mixture of immature DC and CLL cells without fusion step. Staining was as described for Fig. 1a. (c) Typical fluorescence microscopy image of a fusion experiment. Staining of cells was as described in Fig. 1a. Brightfield image is shown on the left, the FITC signals are visualized on the intermediate and the PE signals on the right panel. The arrow-head denotes a fusion event where a CLL membrane has been integrated into a DC membrane. Unlabelled cells are peripheral blood cells other than malignant B-lymphocytes. (d) Number of doublepositive events after mixing or fusing DC and CLL cells as determined by FACS and by fluorescence microscopy. Compilation of 21 independent experiments (all patients were examined by FACS and by microscopic analyses).
measured are derived from CD3C cells but not from CD14C or CD19C cells. Omitting T cells in the coculture assays abrogated cytokine release. To assess the allogeneic effect of DC, patient PBMC were primed with immature DC (iDC) from healthy
donors without tumor-specific restimulation (Fig. 2a, second column). There was a clear response compared to experiments where priming and restimulation were done with autologous cells (first column). This allogeneic response induced by DC did not significantly
280
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
and includes no tumor-specific component. However, when priming was done with a mixture of DC and CLL cells, the stimulation indices markedly increased (fourth column). This effect was further enhanced when the DC and CLL cells had been subjected to the fusion procedure (fifth column). The stimulating effect was even more pronounced when mDC were used (Fig. 2b). The immunogenicity of the fusion samples was not a result of the fusion procedure per se because CLL cells exposed to PEG in the absence of DC had no increased effect (not shown). The difference between iDC and mDC might be due to the higher frequency of hybrid cells caused by the increased fusion efficiency of mDC in comparison to iDC (Fig. 1d). However, when fusion frequencies and stimulation indices were analyzed in all individual experiments, no correlation between these parameters was found. Furthermore, the activating potential of fusion samples generated from iDC could not be improved by enriching the hybrid cells through preparative FACS sorting (not shown). Thus, the enhanced activating effect of mDC–CLL hybrids is likely to be attributable to their more potent co-stimulatory capacity rather than to their slightly increased fusion frequency. Accordingly, the allogeneic effect of mDC was also enhanced as compared to iDC (Fig. 2b, first column and Fig. 2a, second column). Fig. 2. Activation of patient PBMC after consecutive incubation with the indicated stimulators. IFN-g secretion was determined after the second stimulation which was performed using tumor cells or medium. Stimulation indices were calculated as described in the text because absolute cytokine levels were consistent within single experiments but varied between different patients. Thus, the results are only comparable between different patients when normalized to restimulation with unmodified tumor cells in the same experiment. The average absolute IFN-g concentration in the read-out with CLL cells to which all stimulations were normalized was 140 pg/ml. Maximum cytokine release as observed after priming with fusion samples was in the range of 1–2 ng/ml. Not all patients could be included in the different experiments because sufficient cell numbers were not always available. There was, however, no correlation between any clinical parameter and the outcome of fusion or stimulation experiments. The error bars indicate standard deviations. Different results were significant at P! 0.05 (Student’s t-test). (a) Stimulatory potential of allogeneic immature DC and immature DC/CLL cell mixtures with or without fusion step. The first column represents the stimulation index obtained after priming with CLL cells and restimulation with CLL cells, which was defined as 1. Average values from 9 experiments are shown. (b) Stimulation indices obtained by co-cultivation experiments using mature DC (nZ5).
differ from the activation achieved by priming with DC and restimulation with tumor cells (Fig. 2a, second vs. third column). This shows that the response observed in the latter setting might be due to the allogeneic effect
3.3. Immunostimulatory mechanism of DC/CLL cell mixtures Priming with DC/CLL cell mixtures yielded higher stimulation indices than priming with DC alone (Fig. 2a and b) suggesting the presence of a tumor-related component in addition to the allogeneic effect. This antitumor response might be explained by uptake of tumor-associated antigens by DC during co-cultivation with tumor cells followed by presentation of antigenic peptides to T lymphocytes. Alternately, it was possible that T cells were stimulated due to a concerted action of CLL cells presenting tumor-derived peptides to T lymphocytes, and DC that simultaneously delivered the requisite co-stimulatory signals. To distinguish between these possibilities, DC were fixed with PFA before mixing or fusing to CLL cells. PFA-treated DC can provide co-stimulatory signals, but they are not capable of internalizing cells, cellular fragments or proteins and directing antigenic peptides to the cell surface [36]. As shown in Fig. 3, the stimulation index was reduced to background levels when PFA-fixed DC were mixed with CLL cells and used for stimulation
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
Fig. 3. Inhibition of PBMC activation after PFA fixation of DC. Before mixing to or fusing with CLL cells, immature DC were treated with PFA. PBMC activation was done as described in Fig. 2, stimulators used for priming and for restimulation are indicated. Standard deviations are shown as error bars.
of patients’ PBMC. Thus, DC/CLL cell mixtures without fusion are able to induce antitumor immune responses because DC capture antigens from CLL cells during co-cultivation. A marked decrease of the immunostimulatory potential was also found for fusion samples that were derived from PFA-treated DC (Fig. 3). This result was predictable because not only unfused DC which are still present at great numbers in the fusion assay but also the hybrid cells need an intact processing machinery for presenting immunogenic peptides. However, cell fusion might lead to integration of tumor cell membranes containing MHC molecules that are already loaded with CLL-derived peptides into the DC membrane. This could explain why the stimulation observed after fusion of CLL cells with PFA-treated DC was not completely abrogated. 3.4. Concluding remarks A crucial point for protocols using DC–tumor cell hybrids is the fusion efficiency. The comparison of FACS data and fluorescence microscopy in our study suggests that the frequency of DC–tumor cell fusion has to be critically assessed. Thus, unfused DC/CLL cell mixtures yielded double-positive events in the FACS which were probably due to cell aggregations. The consequences of this finding have not been considered so far. Therefore, in previous investigations that mostly rely on FACS analyses, fusion efficiencies might have been overestimated. Using cell populations that contained fusion products as well as unfused DC and CLL cells, autologous patient PBMC could readily be stimulated. Because the
281
stimulation indices did not correlate with the frequency of hybrid cells we assume that the maximal stimulation indices we obtained cannot be further increased. This limitation might be explained by the paucity of responder cells in the patients’ peripheral blood and by additional functional defects of T cells. In a recent study, fusions of B-CLL cells with autologous DC sometimes stimulated autologous T cells more efficiently than naı¨ve DC [37]. However, the stimulating effects of unfused cell mixtures were not investigated in that paper. Using a mouse melanoma model another group reported that physical interactions between unfused DC and tumor cells resulted in an immunogen that was equally effective as hybrid cells [38]. We show that DC that are mixed to CLL cells are capable of priming PBMC, but to a lower extent than those samples that had been subjected to cell fusion. Although in our study the fusion frequencies were as low as about 3% and the most part of the immunostimulatory effect seemed to be attributable to unfused DC and tumor cells, a specific effect of cell fusion could be seen. Mature DC proved superior to immature DC. Therefore, despite low fusion frequencies, immunization with mDC–tumor cell hybrids may be a feasible approach for the adjuvant treatment of B-CLL patients.
Acknowledgements We thank Dr D.J. Schendel for her ongoing support and Dr S. King for critically reading the manuscript. Expert technical assistance by A. Brandl is gratefully acknowledged. The work was supported by Else-Kro¨ner-Fresenius-Stiftung, HGF-Strategiefonds ‘Infektionsabwehr und Krebspra¨vention’ and by Nationales Genomforschungsnetz. This work is part of T. Allgeier’s doctoral thesis at the Medical Faculty of the Ludwig-Maximilians-Universita¨t Mu¨nchen. References [1] J. Banchereau, R.M. Steinman, Dendritic cells and the control of immunity, Nature 392 (1998) 245–252. [2] J. Banchereau, F. Brie`re, C. Caux, J. Davoust, S. Lebecque, Y.J. Liu, et al., Immunobiology of dendritic cells, Annu. Rev. Immunol. 18 (2000) 767–811. [3] R.M. Steinman, The dendritic cell system and its role in immunogenicity, Annu. Rev. Immunol. 9 (1991) 271–296.
282
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283
[4] J.W. Yewdell, J.R. Bennink, Cell biology of antigen processing and presentation to MHC class I molecule restricted T lymphocytes, Adv. Immunol. 52 (1992) 1–123. [5] F.R. Carbone, M.J. Bevan, Class I-restricted processing and presentation of exogenous cell-associated antigen in vivo, J. Exp. Med. 171 (1990) 377–387. [6] W.R. Heath, F.R. Carbone, Cross-presentation, dendritic cells, tolerance and immunity, Annu. Rev. Immunol. 19 (2001) 47–64. [7] R.M. Steinman, M. Witmer-Pack, K. Inaba, Dendritic cells: antigen presentation, accessory function and clinical relevance, Adv. Exp. Med. Biol. 329 (1993) 1–9. [8] N. Romani, S. Gruner, D. Brang, E. Kampgen, A. Lenz, B. Trockenbacher, et al., Proliferating dendritic cell progenitors in human blood, J. Exp. Med. 180 (1994) 83–93. [9] M.A. Ossevoort, M.C. Feltkamp, K.J. van Veen, C.J. Melief, W.M. Kast, Dendritic cells as carriers for a cytotoxic T-lymphocyte epitope-based peptide vaccine in protection against a human papillomavirus type 16-induced tumor, J. Immunother. 18 (1995) 86–94. [10] J.I. Mayordomo, T. Zorina, W.J. Storkus, L. Zitvogel, C. Celluzzi, L.D. Falo, et al., Bone marrow-derived dendritic cells pulsed with synthetic tumour peptides elicit protective and therapeutic antitumour immunity, Nat. Med. 1 (1995) 1297– 1302. [11] J. Gong, L. Chen, D. Chen, M. Kashiwaba, Y. Manome, T. Tanaka, et al., Induction of antigen-specific antitumor immunity with adenovirus-transduced dendritic cells, Gene Ther. 4 (1997) 1023–1028. [12] A.L. Ribas, H. Butterfield, W.H. McBride, V.B. Dissette, A. Koh, C.M. Vollmer, et al., Characterization of antitumor immunization to a defined melanoma antigen using genetically engineered murine dendritic cells, Cancer Gene Ther. 6 (1999) 523–536. [13] L. Zitvogel, J.I. Mayordomo, T. Tjandrawan, A.B. DeLeo, M.R. Clarke, M.T. Lotze, et al., Therapy of murine tumors with tumor peptide-pulsed dendritic cells: dependence on T cells, B7 costimulation, and T helper cell 1-associated cytokines, J. Exp. Med. 183 (1996) 87–97. [14] R.C. Fields, K. Shimizu, J.J. Mule, Murine dendritic cells pulsed with whole tumor lysates mediate potent antitumor immune responses in vitro and in vivo, Proc. Natl Acad. Sci. USA 95 (1998) 9482–9487. [15] R. Mocikat, M. Selmayr, S. Thierfelder, H. Lindhofer, Triomabased vaccination against B cell lymphoma confers long-lasting tumor immunity, Cancer Res. 57 (1997) 2346–2349. [16] M. Selmayr, J. Strehl, J.-P. Kremer, E. Kremmer, A. Doenecke, M. Hallek, et al., Induction of tumor immunity by autologous B lymphoma cells expressing a genetically engineered idiotype, Gene Ther. 6 (1999) 778–784. [17] J. Strehl, M. Selmayr, J.-P. Kremer, L. Hu¨ltner, H. Lindhofer, R. Mocikat, Gene therapy of B-cell lymphoma with cytokine gene-modified trioma cells, Int. J. Cancer 83 (1999) 113–120. [18] M. Selmayr, H. Menzel, J.-P. Kremer, S. Thierfelder, R. Mocikat, B cell lymphoma idiotypes chimerized by gene targeting can induce tumor immunity, Cancer Gene Ther. 7 (2000) 501–506. [19] K. Kronenberger, A. Dieckmann, M. Selmayr, J. Strehl, U. Wahl, H. Lindhofer, et al., Impact of the lymphoma idiotype on in vivo tumor protection in a vaccination model based on targeting antigens to antigen-presenting cells, Blood 99 (2002) 1327–1331.
[20] L.A. Lambert, G.R. Gibson, M. Maloney, B. Durell, R.J. Noelle, R.J. Barth, Intranodal immunization with tumor lysate-pulsed dendritic cells enhances protective antitumor immunity, Cancer Res. 61 (2001) 641–646. [21] Z. Abdel-Wahab, P. DeMatos, D. Hester, X.D. Dong, H.F. Seigler, Human dendritic cells, pulsed with either melanoma tumor cell lysates or the gp100 peptide (280–288), induce pairs of T-cell cultures with similar phenotype and lytic activity, Cell. Immunol. 186 (1998) 63–74. [22] S.K. Nair, D. Snyder, B.T. Rouse, E. Gilboa, Regression of tumors in mice vaccinated with professional antigen-presenting cells pulsed with tumor extracts, Int. J. Cancer 70 (1997) 706–715. [23] M. Schnurr, P. Galambos, C. Scholz, F. Then, M. Dauer, S. Endres, et al., Tumor cell lysate-pulsed human dendritic cells induce a T-cell response against pancreatic carcinoma cells: an in vitro model for the assessment of tumor vaccines, Cancer Res. 61 (2001) 6445–6450. [24] M.L. Albert, B. Sauter, N. Bhardwaj, Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs, Nature 392 (1998) 86–89. [25] L.M. Liau, K.L. Black, R.M. Prins, S.N. Sykes, P.L. DiPatre, T.F. Cloughesy, et al., Treatment of intracranial gliomas with bone marrow-derived dendritic cells pulsed with tumor antigens, J. Neurosurg. 90 (1999) 1115–1124. [26] J. Gong, N. Nikrui, D. Chen, S. Koido, Z. Wu, Y. Tanaka, et al., Fusions of human ovarian carcinoma cells with autologous or allogeneic dendritic cells induce antitumor immunity, J. Immunol. 165 (2000) 1705–1711. [27] J. Gong, D. Avigan, D. Chen, Z. Wu, S. Koido, M. Kashiwaba, et al., Activation of antitumor cytotoxic T lymphocytes by fusions of human dendritic cells and breast carcinoma cells, Proc. Natl Acad. Sci. USA 97 (2000) 2715–2718. [28] J. Gong, D. Chen, D. Kufe, Induction of antitumor activity by immunization with fusions of dendritic and carcinoma cells, Nat. Med. 3 (1997) 558–561. [29] L. Lespagnard, P. Mettens, A.-M. Verheyden, N. Tasiaux, K. Thielemans, S. van Meirvenne, et al., Dendritic cells fused with mastocytoma cells elicit therapeutic antitumor immunity, Int. J. Cancer 76 (1998) 250–258. [30] J. Wang, S. Saffold, X. Cao, J. Krauss, W. Chen, Eliciting T cell immunity against poorly immunogenic tumors by immunization with dendritic cell-tumor fusion vaccines, J. Immunol. 161 (1998) 5516–5524. [31] J. Gong, S. Koido, D. Chen, Y. Tanaka, L. Huang, D. Avigan, et al., Immunization against murine multiple myeloma with fusions of dendritic cells and plasmacytoma cells is potentiated by interleukin-12, Blood 99 (2002) 2512–2517. [32] J. Wang, S. Saffold, X. Cao, J. Krauss, W. Chen, Eliciting T cell immunity against poorly immunogenic tumors by immunization with dendritic cell–tumor fusion vaccines, J. Immunol. 161 (1998) 5516–5524. [33] J. Gong, D. Chen, M. Kashiwaba, Y. Li, L. Chen, H. Takeuchi, et al., Reversal of tolerance to human MUC1 antigen in MUC1 transgenic mice immunized with fusions of dendritic and carcinoma cells, Proc. Natl Acad. Sci. USA 95 (1998) 6279–6283. [34] M. Lindner, V. Schirrmacher, Tumour cell–dendritic cell fusion for cancer immunotherapy: comparison of therapeutic efficiency of polyethylen–glycol versus electro-fusion protocols, Eur. J. Clin. Invest. 32 (2002) 207–217.
T. Allgeier et al. / Cancer Letters 245 (2007) 275–283 [35] U. Wahl, E. No¨ßner, K. Kronenberger, R. Gangnus, H. Pohla, M.S. Staege, et al., Vaccination against B-cell chronic lymphocytic leukemia with trioma cells: preclinical evaluation, Clin. Cancer Res. 9 (2003) 4240–4246. [36] E. No¨ßner, R. Gastpar, V. Milani, A. Brandl, P.J.S. Hutzler, M.C. Kuppner, et al., Tumor-derived heat shock protein 70 peptide complexes are cross-presented by human dendritic cells, J. Immunol. 169 (2002) 5424–5432.
283
[37] P. Kokhaei, M.R. Rezvany, L. Virving, A. Choudhury, H. Rabbani, ¨ sterborg, et al., Dendritic cells loaded with apoptotic tumour A. O cells induce a stronger T-cell response than dendritic cell–tumour hybrids in B-CLL, Leukemia 17 (2003) 894–899. [38] C.M. Celluzzi, L.D. Falo, Physical interaction between dendritic cells and tumor cells results in an immunogen that induces protective and therapeutic tumor rejection, J. Immunol. 160 (1998) 3081–3085.