Interactions of CD70, a tumor necrosis factor-related cell surface ligandand its receptor, CD27, are thought to play an important role for T-, B-, and natural killer-cell activation. However, ligation of CD27 can also induce apoptosis. Human glioblastoma is paradigmatic for cancer-associated immunosuppression. We identified CD70 as a radioinducible gene in U87 MG glioma cells. A screening of a panel of human glioma cell lines revealed that 11 of 12 cell lines expressed CD70 mRNA and protein. Two human neuroblastoma cell lines did not express CD70. CD70 mRNA expression was enhanced by irradiation in 8 of 12 glioma cell lines in a p53-independent manner. No alteration in CD70 expression was observed after glioma cell exposure to cytotoxic drugs such as lomustine. CD70 protein was also detected by immunocytochemistry in 5 of 12 glioblastomas and 3 of 4 anaplastic astrocytomas in vivo. CD27 expression was not detected in any glioma cell line, and there was no evidence for autocrine or backward signaling of the CD70 system in human glioma cells. Unexpectedly, CD70 expressed on glioma cells did not increase the immunogenicity of glioma cells in vitro. In contrast, CD70-positive glioma cells induced apoptosis in peripheral blood mononuclear cells (PBMCs) in a CD70-dependent manner. Neutralization of CD70 expressed on glioma cells prevented apoptosis and enhanced the release of tumor necrosis factor-α in cocultures of glioma cells and PBMCs. The effects of CD70-expressing glioma cells on PBMCs were mimicked by agonistic CD27 antibodies. Conversely, the shedding of CD27 by PBMCs was identified as a possible escape mechanism from glioma cell-induced CD70-dependent apoptosis. Thus, induction of B-cell and T-cell apoptosis via interactions of CD70 expressed on glioma cells and CD27 expressed on B and T cells may be a novel way for the immune escape of malignant gliomas.

Human glioblastoma is a highly lethal brain tumor, which kills affected patients by local destructive growth. Although glioblastoma cells display virtually all features of aggressive neoplastic cells, including resistance to multiple apoptotic stimuli, and strong migratory and invasive properties, these tumors hardly ever metastasize outside of the central nervous system. This clinical observation, as well as ample evidence of altered immune function in glioblastoma patients, has raised the hope that immunological approaches may lead to a therapeutic breakthrough in the management of glioblastoma (1). Glioblastoma-mediated immunosuppression has been attributed to various mechanisms, including the release of immunosuppressive TGF-β (2),3 CD95L-dependent maintenance of immune privilege (3, 4, 5), interference with IL-2-mediated T-cell activation (6), and the release of as yet unidentified soluble factors (7, 8).

CD70 (9), a cell surface protein, is the ligand for CD27, a type I glycoprotein and member of the TNF receptor family (10). CD70 is expressed by cells of the lymphoid lineage, notably activated B (11) and T cells (12, 13), but not resting lymphocytes. CD27 is expressed constitutively by T, B (14) and natural killer cells. A reciprocal expression pattern for CD27 and CD70 has been described (15). CD70/CD27 interactions may play an important role in the maturation and activation of B (16, 17), T (18), and natural killer (19) cells. CD8-positive T cells may be activated via CD70, whereas T-helper cells remain largely unaffected (20). Direct inhibitory effects on the generation of plasma cells (21) and on T-cell activation (22) have also been reported. CD27 costimulation may induce CD45 RA+ CD4+ T cells with regulatory function, resulting in a down-regulation of the immune response (23). CD27 can also mediate apoptosis (24). The immune system of CD27 knockout mice shows impaired T-cell memory but no additional immunodeficiency (25). A more severe phenotype was observed in CD70-transgenic mice, which shows chronic T-cell activation associated with depletion of the B-cell compartment, and of naive and CD27-positive T cells (26).

Some lymphoproliferative diseases are associated with reduced CD27 expression, whereas CD70 expression is elevated (27, 28). CD27 was not detected on the affected CD3 cells in 20 of 21 patients suffering from lymphoproliferative disease of granular lymphocytes, whereas the nontransformed cells expressed CD27 constitutively (29). In contrast, CD70 was induced by malignant transformation in these patients. Here we report on the aberrant expression of CD70 by human malignant glioma cells in vitro and in vivo, and delineate a role for CD70/CD27 interactions in mediating immune escape in this type of cancer.

Cell Culture and Reagents.

The human malignant glioma cell lines LN-18, U138MG, U87MG, LN-428, D247MG, T98G, LN-319, LN-229, A172, U251MG, U373MG and LN-308, kindly provided by Dr. N. de Tribolet (University Hospital [CHUV], Lausanne, Switzerland), and SKN-LE and SKN-BE human neuroblastoma cells (American Type Culture Collection, Rockville, MD) were cultured in 75 cm3 Falcon plastic flasks using DMEM supplemented with 1% glutamine (Life Technologies Inc., Paisley, United Kingdom), 10% FCS (Biochrom KG, Berlin, Germany), and penicillin (100 IU/ml)/streptomycin (100 μg/ml). U87MG cells were transfected with the dominant-negative p53V135A mutant by electroporation (30). The CD70-transfected murine pre-B-cell line 300–19 and control-transfected 300–19 B cells carrying a neo control plasmid, kindly supplied by Dr. S. Jacquot (INSERM U519, Centre Hospitalier Universitaire de Rouen, Rouen, France), were maintained in RPMI 1640 (Life Technologies, Inc.) with 10% FCS, G418 (250 μg/ml), and 15 μm β-mercaptoethanol. All of the cell lines were routinely tested for contamination with Mycoplasma by 4′,6-diamidino-2-phenylindole staining (Sigma, Deisenhofen, Germany). Where indicated, the cells were irradiated while adherent using a Gammacell 1000 Elite (Nordion, Kanata, Ontario, Canada). Radiation was applied as a single dose at a central dose rate of ∼9 Gy/min. The pan-specific caspase inhibitor zVAD-fmk was purchased from Bachem (Heidelberg, Germany). Lomustine was kindly provided by Medac (Hamburg, Germany). Cycloheximide and propidium iodide were purchased from Sigma (St. Louis, MO). AnxV-FITC was obtained from PharMingen (Heidelberg, Germany) and human recombinant TNF-α from Roche (Mannheim, Germany). The following antibodies were used: M-T271 mouse antihuman CD27, FITC-labeled, and unlabeled BU69 mouse antihuman CD70 (Ancell, Bayport, MN); HNE.51 mouse antihuman CD70 (Dakopatts, Glostrup, Denmark); C-20 goat antihuman CD70 (Santa Cruz Biotechnology, Santa Cruz, CA); Ki-24 mouse antihuman CD70 and NOK-1 mouse antihuman CD95L (PharMingen); and MAB1835 mouse antihuman TGF-β (R&D Systems, Wiesbaden, Germany). All of the cell lines were tested for the cell surface expression of costimulatory molecules using FITC-labeled antibodies to CD40, CD80, CD86, and CD154 (PharMingen). The cell lines were also analyzed for HLA-A2 expression using the murine BB7.2 antibody. Maxisorp plates used for antibody immobilization were obtained from Nunc (Roskilde, Denmark).

cDNA Array Analysis.

The Atlas Human Cancer 1.2 Arrays (Clontech, Palo Alto, CA) were used to screen for alterations of gene expression by irradiation in U87MG and T98G cells. Details for cDNA probe synthesis, array hybridization, and quantitation were described previously (31). The irradiated to control ratio of ≥1.5 in expression was considered as up-regulation, and a ratio of <0.67 in expression was considered as down-regulation after irradiation.

Northern Blot Analysis.

Total RNA was extracted using the RNeasy RNA purification system (Qiagen, Hilden, Germany). Denatured total RNA (10 μg) was loaded on a 1% agarose gel containing 6.7% formaldehyde. The RNA was separated at 100 V, transferred to a Hybond N+ membrane (Amersham, Freiburg, Germany) using capillary blotting, and cross-linked in a UV stratalinker 1800 (Stratagene, La Jolla, CA) at 1200 J. Methylene blue staining was performed as a loading control. The membrane was preincubated for 2 h in Church buffer [0.25 m sodium dodecyl sulfate, 0.5 m Na2 HPO4/NaH2 PO4 (pH 7.0), 0.5 mm EDTA] at 65°C. The probe was constructed by PCR-amplifying nucleotides 192–535 from a human CD70 expression vector, kindly provided by M. R. Bowman (Genetics Institute, Cambridge, MA), using CD70-specific primers (5′-CTT GGT GAT CTG CCT CGT GG-3′ up and 5′-GCA GCA GGC TGA TGC TAC G-3′ down). Purified PCR product (40 ng) was labeled using 5 μl (∼1.6 Mbq) dCTP and the Rediprime II random labeling system (Amersham). Filters were hybridized overnight at 65°C in a hybridization oven with a rotisserie device using Church buffer. Binding of radioactive probes was visualized and quantified using a phosphorimager (FujiBasReader 1500; Fuji). Signals were normalized to β-actin mRNA expression. RNA from the 12 glioma cell lines under investigation was also probed for CD27 expression using a sCD27 probe amplified from human CD27 cDNA (American Type Culture Collection, Manassas, VA). The primers used were 5′-GGG AAT TCT TGG AGG TGC TAA CT-3 up and 5′-ATG GGC CCC GAA TAA AAT CGG AGC-3 down (32).

Immunoblot Analysis.

CD70 protein levels were analyzed by immunoblot using 20 μg of protein per lane on a 10% acrylamide gel (Bio-Rad, Munich, Germany). After transfer to nitrocellulose (Bio-Rad) the blots were blocked in PBS containing 5% skim milk and 0.05% Tween 20, and incubated overnight at 4°C with 2 μg/ml of CD70 antibody (C20). Visualization of protein bands was accomplished using horseradish peroxidase-coupled antigoat IgG secondary antibody (Sigma) and enhanced chemiluminescence (Amersham).

Flow Cytometry.

Glioma cells were detached nonenzymatically using cell dissociation solution (Sigma). The cells were preincubated in PBS with 2% BSA and incubated with FITC-labeled CD70 antibody (BU69; 20 μg/ml in PBS/BSA) or isotype-matched mouse IgG-FITC as a control. Fluorescence was measured in a Becton Dickinson FACScalibur (Heidelberg, Germany). SFIs were calculated by dividing mean fluorescence obtained with specific antibody by mean fluorescence obtained with control antibody. To validate the data, experiments were repeated with additional monoclonal CD70 antibodies (HNE.51 and KI-24). To examine the cells for CD27 expression, M-T271 antibody was used under the same conditions.

Immunohistochemistry.

Cryosections (0.7 mm) of human gliomas were fixed in acetone and blocked with normal rabbit serum and BSA. All of the samples were stained with three different CD70 antibodies (Dakopatts, PharMingen, and Santa Cruz Biotechnology) at dilutions of 1:25 to 1:100. All of the antibodies gave comparable results at all of the tested dilutions. A biotinylated antimouse or antigoat secondary antibody (Zymed) was used at 1:150. Avidin biotin complex was added, and the staining was developed with diaminobenzidine. Human tonsils were taken as positive control. Normal human brain (temporal lobe) served as a negative control.

T-Cell Reaction against Allogeneic Glioma Cells.

To suppress glioma cell proliferation, the cells were incubated in serum-free medium for 24 h, irradiated at 200 Gy, and maintained in serum-free medium for another 24 h. Although doses in excess of 10 Gy are sufficient to suppress colony formation in all of the glioma cell lines examined here (33), the high dose of 200 Gy was necessary to prevent proliferation within the 1–5 days of the experiments and to suppress [methyl-3H]thymidine incorporation in assays where this incorporation was used to assess immune cell proliferation in the presence of glioma cells. PBMCs were isolated from healthy donors by density gradient centrifugation (Biocoll; Biochrom KG) and characterized for their HLA-A isotype. Monocytes were depleted by adhesion and differential centrifugation. To obtain purified T cells, the samples were depleted of B cells and monocytes using the LymphoKwik T reagent (One Lambda Inc., Canoga Park, CA). The purity of this population was controlled by flow cytometry using antihuman CD3-PE (Becton Dickinson). HLA-A2-negative T cells or PBMCs (105/well) were cocultured with 104 irradiated HLA-A2-positive glioma cells in 96-well plates in triplicates. To minimize interference of CD70 antibody with CD70 expressed by activated immune cells, the glioma cells were preincubated with CD70 antibody for 2 h and washed twice with PBS. After 4 days, the cells were pulsed for 24 h with [methyl-3H]-thymidine (0.5 μCi; Amersham). The cells were harvested with a cell harvester (Inotech, Dottikon, Switzerland), and incorporated radioactivity was bound to a glass fiber filtermat (Wallac, Turku, Finland). The filtermat was wetted with Ultima Gold Scintillation Mixture (Packard, Dreieich, Germany), and radioactivity was determined in a Wallac 1450 Microbeta Plus Liquid Scintillation Counter (Wallac).

Chromium Release Assay.

The glioma cells were seeded into 25-cm2 flasks such that the number of cells reached 106/flask 24 h later. The cells were serum-deprived and irradiated at 200 Gy. After repeated washing and preincubation of the glioma cells with CD70 or control antibodies, 1.5 × 107 PBMCs were added in 5 ml of RPMI 1640 containing 10% FCS. The cells were cocultured for 5 days. PBMCs from the same isolation were maintained in culture without glioma cells in parallel. On day 3, PHA (10 μg/ml; Biochrom) was added to one group of PBMCs. On day 5, PBMCs were collected and counted. The supernatants were collected for cytokine measurements. Glioma cells were detached, suspended in medium (1 ml), and labeled by addition of 50 μCi 51Cr (NEN, Boston, MA). The PHA-stimulated PBMCs were also labeled. The labeled allogeneic glioma cells (104/well) were incubated either alone (spontaneous release) or with effector PBMCs from the cocultures at E:T ratios of 10:1, 20:1, 40:1, or 80:1. The maximum 51Cr release possible was determined by addition of NP40 (100% lysis). To determine autoreactivity, labeled activated PBMCs (105/well) were incubated with numbers of effector PBMCs equivalent to those used for the lysis of the glioma cells. After 4 h, 50 μl of the supernatant were transferred to a Luma-PlateTM-96 (Packard), dried overnight, and measured. Lysis was calculated as [cpm (effector cells) − cpm (spontaneous)]/[cpm (NP40) − cpm (spontaneous)] × 100%. To examine the killing of immune cells by glioma cells, 104 glioma cells were seeded in a 96-well plate. One day later, 105-labeled PBMCs were added. Both PHA-stimulated and nonstimulated PBMCs were tested. Supernatants were collected 18 h later and assayed as above.

TNF-α Bioassay.

l-M cells (5 × 104/well) were seeded in 96-well plates. One day later the serum-containing medium was replaced by 100 μl serum-free DMEM containing cycloheximide (20 μg/ml). The supernatants from PBMCs, glioma cells, or cocultures were diluted to a volume of 100 μl and added. Survival was assessed 16 h later by crystal violet staining. These data were confirmed by ELISA for human TNF-α (Endogen, Woburn, MA).

Determination of sCD27.

Cell culture supernatants were analyzed for sCD27 by sandwich ELISA (CLB, Amsterdam, the Netherlands) according to the manufacturer’s instructions.

Detection of DNA Fragmentation and Apoptosis.

PBMCs were stained with FITC-labeled antibodies to CD4, CD8, CD14, CD19, or CD56, fixed, and permeabilized in ice-cold 70% ethanol. RNA was digested with RNase A (Life Technologies, Inc.). DNA was stained with propidium iodide (50 μg/ml). FACScalibur settings were adjusted so that the G1 peak measured in channel Fl-2A moved to 200 relative fluorescence units. Cells to the left of this peak have less than 2n chromosomes signifying loss of DNA. Aggregated cells were detected in channel Fl-2W and gated out. Given the strong adherence of glioma cells and the excess of PBMCs in these assays, contamination of PBMC populations with glioma cells was negligible as controlled by staining of lymphocyte subpopulations and size. PBMC apoptosis was also analyzed by AnxV-FITC staining. The cells were collected, washed, and resuspended in a buffer containing 10 mm HEPES/NaOH (pH 7.4), 140 mm NaCl, and 2.5 mm CaCl2. AnxV-FITC (1:100) and propidium iodide (50 μg/ml) were added. When analysis of lymphocyte subsets was desired, PBMCs were first incubated with CD4-PE, CD8-PE, CD14-PE, CD19-PE, or CD56-PE monoclonal antibodies.

Statistical Analysis.

Data are representative of experiments performed three times with similar results. Viability studies were performed using triplicate wells. Significance was assessed by t test (∗P < 0.05; ∗∗P < 0.01).

Human Malignant Glioma Cell Lines Express CD70 mRNA and Protein in Vitro and CD70 Protein in Vivo.

U87MG and T98G cells were screened for changes in gene expression in response to irradiation. In U87 MG cells, the Clontech Atlas cDNA array revealed that 42 genes were induced and 36 were down-regulated 1 h after irradiation. At 4 h after irradiation, 15 genes were up-regulated and 10 down-regulated. In T98G cells, the expression of 27 genes was up-regulated, and 66 genes were down-regulated 1 h after irradiation, whereas 6 genes were up- and 22 down-regulated at 4 h after irradiation. Genes known to be induced by irradiation, e.g., p21, were up-regulated in U87 MG cells after 1 h (1.4-fold) and 4 h (2.6-fold) of irradiation. T98G is a p53 mutant cell line, and p21 levels did not change with irradiation in these cells. Surprisingly, CD70 antigen mRNA was found in both control and irradiated U87 MG cells. Whereas no change in CD70 expression was observed at 1 h after irradiation, there was a 5.7-fold increase 4 h after irradiation (data not shown). CD70 mRNA was not detectable in T98G cells with or without irradiation. Northern blot analysis revealed that 11 of 12 malignant glioma cell lines expressed CD70 mRNA (Fig. 1,A). T98G cells were negative. High levels of CD70 mRNA were detected in U138MG, D247MG, LN-319, and U373 MG cells. Immunoblot analysis confirmed that CD70 mRNA was translated into CD70 protein (Fig. 1,B). Flow cytometry showed that the protein was expressed at the cell surface (Fig. 1,C). Quantitative data for CD70 mRNA expression and CD70 protein levels are provided in Table 1. There was a strong correlation between CD70 mRNA expression and CD70 protein expressed at the cell surface (r = 0.9594; P < 0.0001). Two human neuroblastoma cell lines, SKN-BE and SKN-LE, did not show CD70 protein determined by flow cytometry and immunoblot analysis (data not shown). Among 12 human glioblastoma specimens, 5 were positive for CD70. A representative sample is shown in Fig. 2,A. Among 4 anaplastic astrocytomas examined, 3 showed CD70 expression (Fig. 2,B). Morphological features led to the identification of CD70-expressing cells as tumor cells rather than tumor-infiltrating host cells. Normal brain sections revealed no CD70-expressing cells (Fig. 2 C).

p53-independent Modulation of CD70 Expression in Human Malignant Glioma Cell Lines by Irradiation.

Because irradiated glioma cells were used for the ensuing immunological studies, we also determined whether radioinducibility is a general feature of CD70 expression in human glioma cells. All 12 of the cell lines were irradiated at 6 Gy and assessed for CD70 mRNA and protein levels. CD70 mRNA levels increased in 8 cell lines and CD70 protein expression at the cell surface in 7 cell lines (Fig. 1, D–F). The increase in CD70 mRNA in LN-319 cells, which express high levels of CD70 constitutively, did not result in enhanced CD70 protein expression. The highest increase in CD70 protein expression was 89% in U87 MG cells (Fig. 1 F). Kinetic analyses in U87 MG and LN-229 cells showed that the maximum cell surface expression was found at 6 h after irradiation. To assess whether CD70 is also induced at lower doses of irradiation, U87 MG and LN-308 cells were irradiated at 1, 3, or 6 Gy. The SFI were 2.7 at 0 Gy, 3.6 at 1 Gy, 4.9 at 3 Gy, and 5.1 at 6 Gy in U87 MG cells, and 3.0 at 0 Gy, 3.4 at 1 Gy, 3.5 at 3 Gy, and 3.5 at 6 Gy in LN-308 cells (data not shown). Irradiation of U373MG, U138MG, or T98G at 200 Gy resulted in the same changes in CD70 expression detected by flow cytometry as irradiation at 6 Gy (data not shown). U87 MG, the first cell line identified to respond to irradiation with enhanced CD70 expression, is wild type for p53 (34). However, the irradiation-induced increase in CD70 expression was unaffected by ectopic expression of a dominant-negative p53 mutant, p53V135A(31), confirming that p53 does not mediate the induction of CD70 after irradiation. Conversely, the nitrosourea lomustine, a cytotoxic agent that induces p53 expression in U87 MG cells, did not modulate CD70 expression, confirming that enhanced CD70 expression is not a nonspecific response to genotoxic stress (data not shown).

No Autocrine Effects of CD70 Expression in Human Glioma Cells.

To investigate possible autocrine effects of the CD70/CD27 system (35) in glioma cells, we screened the glioma cell lines for the expression of CD27 mRNA using Northern blot, reverse transcription-PCR, and flow cytometry. None of the methods revealed the expression of CD27 in any of the 12 cell lines, using PBMCs as a positive control. Furthermore, ligation of CD70 with specific antibody in LN-18, U138MG, U87MG, T98G, or U373MG cells had no effect on the proliferation or survival of the glioma cells. These data remained negative, even with immobilization of the CD70 antibody on a Nunc Maxisorp surface or cross-linking by a secondary antibody. These overall negative data (not shown) provided no evidence for autocrine or backward signaling of the CD70 system in human glioma cells.

CD70-mediated Inhibition of Alloreactivity by Human Glioma Cells.

None of the 12 glioma cell lines expressed CD80, CD86, or CD154. The cell lines used for the ensuing immunological experiments (T98G, U138MG, and U373MG) were also negative for CD40 and MHC class II expression, but all three expressed the MHC class I molecule HLA-A2 at high levels. We first examined the alloproliferative response of HLA-A2-negative PBMCs or T cells to the glioma cells. The Ki-24, HNE.51, and BU69 antibodies to CD70 have all been described as blocking antibodies.4 When tested in parallel, these antibodies uniformly enhanced the alloproliferative response to the CD70-positive cell lines, U138MG and U373MG, but not the alloproliferative response to the CD70-negative cell line T98G (Fig. 3,A). Because the effects were most prominent with the Ki-24 antibody, an azide-free preparation of this antibody (10 μg/ml) was chosen for all of the ensuing functional experiments. Consistent with this inhibitory effect of CD70, PHA-induced T-cell proliferation (Fig. 3,B) or PBMC proliferation in a MLR (Fig. 3,C) were inhibited much stronger by U138MG and U373MG cells than by T98G cells, and this inhibition was partly relieved by the CD70 antibody. The broad spectrum caspase inhibitor zVAD-fmk mimicked the effects of CD70 antibody in the MLR (Fig. 3 C).

CD70-mediated Induction of PBMC Apoptosis by Glioma Cells.

The number of viable PBMCs was decreased by a factor of 1.8–2.6 after the coculture with U138MG and U373MG cells compared with T98G cells. This cell loss was prevented by CD70 antibody, sCD27, or zVAD-fmk (Fig. 4,A). In PBMCs cocultured with U138MG and U373MG cells, the number of AnxV-positive cells was significantly reduced by CD70 antibody or sCD27. No such effect was seen in PBMCs cocultured with T98G cells (Fig. 4,B). Flow cytometry confirmed cell loss by apoptosis; the percentage of PBMCs in the sub-G1-peak was significantly reduced by CD70 antibody and sCD27 after coculture with U138MG and U373MG cells but not after coculture with T98G cells. zVAD-fmk (100 μm) exerted a protective effect in cocultures with all three of the cell lines (Fig. 4 C). The protection afforded by CD70 antibody or sCD27 was similar to that achieved with zVAD-fmk. Induction of apoptosis was strong in CD4-positive and CD8-positive T cells, which constitute the major subpopulations after 5 day coculture, whereas no significant effect was found for CD56-positive cells (data not shown).

The hypothesis that glioma cells kill PBMCs (reversed lysis) was confirmed in a redirected 51Cr release assay. The coincubation of 51Cr-labeled resting or PHA-stimulated PBMCs with the glioma cells resulted in the release of 51Cr. Lysis was much more prominent with U138MG and U373MG cells than with T98G cells. The addition of CD70 antibody or of sCD27 resulted in decreased lysis of PBMCs cocultured with U138MG or U373MG cells but not T98G cells. zVAD-fmk decreased the lysis of PBMCs in cocultures with all three of the cell lines (Fig. 5,A). Similar results were obtained in PHA-activated PBMCs (Fig. 5,B). After longer stimulation with PHA (48 h), ∼25% of all cells underwent spontaneous apoptosis, which was prevented by zVAD-fmk. The spontaneous 51Cr release by PHA-stimulated PBMCs but not by resting PBMC in the absence of glioma cells was also reduced by CD70 antibodies or sCD27. We speculate that these effects in the 51Cr release assay were mediated by interactions of the reagents (CD70 antibody and sCD27) with CD70 expressed on the activated immune cells. In line with this concept, PHA stimulation greatly enhanced the cell surface expression of CD70 (Fig. 5,D). In contrast, the expression of CD27 remained largely unchanged for 24 h (data not shown) but was considerably decreased among CD4-, CD8-, and CD56-positive cells after 48 h (Fig. 5,C). No cells expressed both CD27 and CD70. Corresponding to the larger number of surviving PBMCs, the number of remaining U138MG and U373MG stimulator cells determined by crystal violet staining was decreased by a factor of 1.7 to 2.5 after 5 days of cocultures when CD70 antibody or sCD27 were present (data not shown). We next sought to determine whether PBMCs surviving the challenge of glioma cell-induced CD70-mediated apoptosis were fully functional or functionally impaired, e.g., less active in lytic assays. However, Fig. 5,E shows that the lytic activity of the surviving PBMCs was unaffected by pre-exposure of the glioma cells to CD70 antibodies (Fig. 5,E). To additionally confirm the specificity of the effects mediated by CD70 expressed on glioma cells, we performed similar experiments in the absence of glioma cells but in the presence of agonistic CD27 antibodies. These experiments revealed induction of apoptosis in resting PBMCs and in the MLR, and apoptosis was blocked by zVAD-fmk (Fig. 5, F and G). The proportion of apoptotic PBMCs detected by flow cytometric analysis of DNA content remained at an almost constant level for 5 days. Because apoptotic cells are likely to be cleared by macrophages in culture (36), the actual number of dying cells may be considerably higher than the level of ∼20% found at a given time point. The amount of 51Cr released by resting and activated PBMCs was significantly increased by the CD27 antibody compared with an isotype control antibody (data not shown). Apoptotic PBMCs detected by AnxV-FITC staining or DNA loss (sub-G0-peak) were additionally characterized by staining with specific antibodies. Subset analysis revealed that several CD27-positive subsets, in particular CD19-positive, CD4-positive, and CD8-positive cells, were affected (data not shown). PBMC activation in coculture with glioma cells was also monitored at the level of TNF-α release into the supernatant (Table 2). TNF-α release was enhanced by CD70 antibody in cocultures containing U138MG and U373MG cells but not in cocultures containing T98G cells. The addition of zVAD-fmk levels elevated TNF-α under all of the conditions. There was no TNF-α release by glioma cells cultured in the absence of PBMCs or in the presence of interferon γ (100 units/ml; data not shown). As in other experiments, we noted that PBMCs from some donors were strongly activated by T98G cells, whereas PBMC from other donors hardly responded to this cell line. This donor-specific variability was less pronounced with U138MG and U373MG cells. Importantly, TNF-α release from PBMC in cocultures with T98G was never affected by the addition of CD70 antibody or sCD27, whereas these reagents produced consistent effects in the other two cell lines. Appropriate control experiments revealed that CD70/CD27-mediated PBMC apoptosis was not mediated by CD95/CD95L interactions or TGF-β, because neutralizing antibodies to CD95L or TGF-β failed to rescue the loss of PBMCs in experiments corresponding to Fig. 4, A and C, and Fig. 5, A and B. The antibodies were used at concentrations shown previously to block human CD95L (37) and TGF-β (2).

CD70-expressing Glioma Cells Induce the Release of sCD27 from PBMC and Decrease CD27 Expression.

The detection of sCD27 in normal human serum (38) indicates that this molecule may have immune regulatory properties. The PBMC preparations used here showed levels of 5 units/ml at day 5 as determined by ELISA. None of the 12 glioma cell lines produced sCD27 when cultured alone, consistent with the failure to detect CD27 mRNA by reverse transcription-PCR and Northern blotting (see above). Coculture with U138MG or U373MG cells significantly enhanced the levels of sCD27 in the supernatant, whereas no such effect was seen with T98G cells. CD70 antibody and zVAD-fmk inhibited this effect (Fig. 6,A). sCD27 is generated by proteolytic cleavage of membrane-bound CD27 (39) after triggering of T cells by antigen (38, 40). Hence we also investigated CD27 expression at the cell surface of the remaining PBMCs after a 5-day coculture. Whereas no change in CD27 expression resulted from coculture with T98G cells, the CD70-positive cell lines U138MG and U373MG induced a reduction of CD27 expression at the surface of the remaining PBMCs. The number of PBMCs with high CD27 expression decreased from 50.7% to 10.9% after coculture with U373 MG cells, suggesting that, after contact with CD70-expressing glioma cells, CD27 is cleaved, released, and not reincorporated into the cell membrane, providing another response of PBMCs to CD70-expressing glioma cells, in addition to apoptosis (Fig. 6 B).

The present study identifies CD70 as a novel candidate molecule mediating immune escape of human malignant gliomas. The induction of CD70 expression by irradiation exemplifies how radiotherapy might suppress antitumor immune responses. Therefore, this observation has implications for the orchestration of multimodality cancer therapy in the clinic. Endogenous expression of CD70 had not been reported previously in nonlymphoid cancer cells. We identified CD70 as a radioinducible gene in U87 MG glioma cells and noted that CD70 expression was rather common among human malignant glioma cell lines (Fig. 1, A–C) and that CD70 was also expressed by neoplastic glial cells in vivo (Fig. 2). CD70 expression was specifically induced by irradiation in glioma cells but not in neuroblastoma cells. Irradiation-induced CD70 expression did not require wild-type p53 activity and was not merely a response to genotoxic stress, because cytotoxic agents such as lomustine did not modulate CD70 expression. In contrast to our expectations, CD70 expressed by glioma cells induced immunosuppression rather than immune stimulation in alloreactivity assays in vitro (Fig. 3). Several lines of experimental evidence disclosed that CD70 expressed by glioma cells promoted apoptosis in the PBMC population (Figs. 4 and 5). The shedding of CD27 by PBMCs may represent an escape mechanism from glioma cell-induced, CD70-dependent apoptosis (Fig. 6). The role for CD70/CD27 interactions in T- and B-cell activation has raised interesting perspectives for the immunotherapy of cancer (41, 42). For instance, virus-mediated CD70 gene transfer into mouse colon carcinoma cells enhanced T-cell responses and induced antitumor immunity (43). Yet, T-cell activation and apoptosis are complex processes that do not only depend on primary signals but also on secondary factors. This complex regulation seems to be necessary, given the need of rapidly controlling immune reactions under different and rapidly changing conditions (44). Thus, many molecules can exert both stimulatory as well as proapoptotic effects on the immune system (45). The costimulatory role of CD70 has been well characterized (46). However, CD70-induced apoptosis mediated by the proapoptotic protein Siva has also been demonstrated (24, 47). The present study provides ample evidence that CD70 expressed by glioma cells evokes immune inhibition rather than immune stimulation (Figs. 3,4,5,6) and, thus, contrasts with several previous studies aiming at exploiting immune stimulatory properties of CD70 (40, 41, 42). It is commonly assumed that cells undergoing malignant transformation can potentially be cleared by the immune system. To escape from the surveillance of the immune system, most tumors down-regulate costimulatory molecules and instead express molecules with immune inhibitory function, including TGF-β or CD95L. Such processes of selection a priori argue against a possible costimulatory role of CD70 expressed on glioma cells. Impaired CD70 signaling may occur when the level of sCD27 is abnormally high. This has been reported for the blood and cerebrospinal fluid of patients suffering from various autoimmune diseases (48, 49, 50, 51). Therefore, one might speculate that high concentrations of sCD27 administered into a postsurgical glioma cavity might promote local antiglioma immunity.

Expression and functional activity of CD70 in a solid nonlymphoid tumor represents one of two novel and unexpected findings of this study. The second major observation was the radioinducibility of CD70 in glioma cells in vitro, which led to its identification by cDNA array analysis. Human glioblastomas are commonly irradiated in 1.8–2 Gy fractions to a total dose of 54–60 Gy. Because single doses of 6 Gy were used here to assess the effects of irradiation in most experiments (Fig. 1), future studies will need to determine whether lower dose fractionated irradiation will also induce CD70 expression in vitro and whether similar effects can be demonstrated in vivo. This study defines a possible pathway by which fractionated radiotherapy may create a locally immunosuppressed environment in human glioblastomas. Thus, CD70 expression maintained by radiotherapy may interfere with immune responses to the tumor during the weeks of radiotherapy. It is tempting to speculate that irradiation-induced CD70 expression might be particularly counterproductive in human cancers known to be immunogenic such as malignant melanoma or renal cell carcinoma.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

Supported by Deutsche Forschungsgemeinschaft (STE 963/1-1). J. W. was supported by the Fortüne program of the University of Tübingen.

3

The abbreviations used are: TGF, transforming growth factor; Ab, antibody; AnxV, Annexin V; sCD27, soluble CD27; CD95L, CD95 ligand; IL, interleukin; MLR, mixed lymphocyte reaction; PBMC, peripheral blood mononuclear cell; PE, phycoerythrin; PHA, phytohemagglutinin; TNF, tumor necrosis factor; zVAD-fmk, N-tert-butoxy-carbonyl-Val-Ala-Asp-fluoromethylketone; SFI, specific fluorescence index.

4

Vth Workshop on Leukocyte Typing.

Fig. 1.

CD70 expression in human glioma cells. A, total RNA was analyzed for the expression of CD70 or β-actin mRNA. B and C, CD70 protein levels were determined by immunoblot (B) and flow cytometry (C). B, CD70 migrating at Mr 55,000. C, representative flow cytometry profiles for CD70-negative T98G cells (left), CD70-positive U373MG cells (middle), and CD70-transfected 300–19 cells (right; bold profile, CD70 antibody; dotted profile, control antibody; see also Table 1, right column). D, the cells were untreated or irradiated at 6 Gy. RNA was isolated 4 h later, and CD70 mRNA levels were quantified by Northern blot. (E and F). CD70 protein in untreated or irradiated cells (6 Gy, 4 h) was quantified by flow cytometry, shown representatively in E for U87 MG cells. The isotype control was not shifted after irradiation at 6 Gy. The numbers indicate the mean fluorescence values used to derive the SFI values. In F, the induction of CD70 expression is expressed as percentages as determined by Northern blot analysis (relative intensities corrected for β-actin) and flow cytometry (SFI6Gy/SFI0Gy − 1); bars, ±SD.

Fig. 1.

CD70 expression in human glioma cells. A, total RNA was analyzed for the expression of CD70 or β-actin mRNA. B and C, CD70 protein levels were determined by immunoblot (B) and flow cytometry (C). B, CD70 migrating at Mr 55,000. C, representative flow cytometry profiles for CD70-negative T98G cells (left), CD70-positive U373MG cells (middle), and CD70-transfected 300–19 cells (right; bold profile, CD70 antibody; dotted profile, control antibody; see also Table 1, right column). D, the cells were untreated or irradiated at 6 Gy. RNA was isolated 4 h later, and CD70 mRNA levels were quantified by Northern blot. (E and F). CD70 protein in untreated or irradiated cells (6 Gy, 4 h) was quantified by flow cytometry, shown representatively in E for U87 MG cells. The isotype control was not shifted after irradiation at 6 Gy. The numbers indicate the mean fluorescence values used to derive the SFI values. In F, the induction of CD70 expression is expressed as percentages as determined by Northern blot analysis (relative intensities corrected for β-actin) and flow cytometry (SFI6Gy/SFI0Gy − 1); bars, ±SD.

Close modal
Fig. 2.

CD70 expression in human gliomas in vivo. CD70 expression was assessed by immunocytochemistry in glioblastoma (A), anaplastic astrocytoma (B), normal brain (C), or tonsil as a positive control (D; magnification ×200 in A and B, and ×100 in C and D).

Fig. 2.

CD70 expression in human gliomas in vivo. CD70 expression was assessed by immunocytochemistry in glioblastoma (A), anaplastic astrocytoma (B), normal brain (C), or tonsil as a positive control (D; magnification ×200 in A and B, and ×100 in C and D).

Close modal
Fig. 3.

Modulation of alloreactivity to glioma cells by CD70. A–C, irradiated glioma cells were preincubated with isotype control antibody or CD70 antibody (Ki-24; 10 μg/ml) for 2 h. In A, the glioma cells were seeded in 96-well plates (104/well), and 105 T cells were added in a total volume of 150 μl RPMI 1640 containing 10% FCS. [Methyl-3H]thymidine (0.5 μCi/well) was added on day 4 for 16 h. [Methyl-3H]thymidine incorporation was measured by liquid scintillation counting. In B, PHA (10 μg/ml) was added to all wells. In C, 105 irradiated HLA-A2-positive stimulator PBMCs were cocultured with 105 HLA-A2-negative responder PBMCs and 104 irradiated glioma cells. Data in A–C are expressed as cpm determined by liquid scintillation counting (∗P < 0.05, ∗∗P < 0.01, CD70 antibody or zVAD-fmk compared with isotype control antibody); bars, ±SD.

Fig. 3.

Modulation of alloreactivity to glioma cells by CD70. A–C, irradiated glioma cells were preincubated with isotype control antibody or CD70 antibody (Ki-24; 10 μg/ml) for 2 h. In A, the glioma cells were seeded in 96-well plates (104/well), and 105 T cells were added in a total volume of 150 μl RPMI 1640 containing 10% FCS. [Methyl-3H]thymidine (0.5 μCi/well) was added on day 4 for 16 h. [Methyl-3H]thymidine incorporation was measured by liquid scintillation counting. In B, PHA (10 μg/ml) was added to all wells. In C, 105 irradiated HLA-A2-positive stimulator PBMCs were cocultured with 105 HLA-A2-negative responder PBMCs and 104 irradiated glioma cells. Data in A–C are expressed as cpm determined by liquid scintillation counting (∗P < 0.05, ∗∗P < 0.01, CD70 antibody or zVAD-fmk compared with isotype control antibody); bars, ±SD.

Close modal
Fig. 4.

CD70-expressing glioma cells kill PBMCs in a CD70-dependent manner. A, irradiated (200 Gy) glioma cells were preincubated with isotype control antibody, CD70 antibody (10 μg/ml) or sCD27 (50 units/ml) for 2 h and washed twice in PBS before PBMCs were added. Alternatively, zVAD-fmk (100 μm) was added at 0 h. The number of viable PBMCs was assessed 5 days later by trypan blue exclusion; bars, ±SD. B, the percentage of AnxV-positive PBMCs was determined after coculture conditions as in A; bars, ±SD. C, after coculture with glioma cells, the PBMCs were fixed, permeabilized, and stained with propidium iodide. Cells to the left of the G1 peak show loss of DNA indicative of apoptosis. In general we noted that the response to T98G cells was more heterogeneous among different PBMCs donors than to U138MG or U373MG cells (compare Figs. A–C). Importantly, the observed effects were never modulated by CD70 antibody or sCD27 in T98G cells, but always in U138MG and U373MG cells.

Fig. 4.

CD70-expressing glioma cells kill PBMCs in a CD70-dependent manner. A, irradiated (200 Gy) glioma cells were preincubated with isotype control antibody, CD70 antibody (10 μg/ml) or sCD27 (50 units/ml) for 2 h and washed twice in PBS before PBMCs were added. Alternatively, zVAD-fmk (100 μm) was added at 0 h. The number of viable PBMCs was assessed 5 days later by trypan blue exclusion; bars, ±SD. B, the percentage of AnxV-positive PBMCs was determined after coculture conditions as in A; bars, ±SD. C, after coculture with glioma cells, the PBMCs were fixed, permeabilized, and stained with propidium iodide. Cells to the left of the G1 peak show loss of DNA indicative of apoptosis. In general we noted that the response to T98G cells was more heterogeneous among different PBMCs donors than to U138MG or U373MG cells (compare Figs. A–C). Importantly, the observed effects were never modulated by CD70 antibody or sCD27 in T98G cells, but always in U138MG and U373MG cells.

Close modal
Fig. 5.

CD70-mediated apoptosis of PBMCs cocultured with glioma cells is mimicked by agonistic CD27 antibodies. A, PBMCs were labeled with 51Cr. After coculture with nonirradiated glioma cells (105 PBMCs; 104 glioma cells/well) for 18 h, 51Cr was determined in the supernatant. Assays were performed in the presence of an isotype-matched control antibody (10 μg/ml), Ki-24 CD70 antibody (10 μg/ml), sCD27 (50 units/ml), or zVAD-fmk (100 μm). B, as in A, but PBMCs (5 × 104) were activated by incubation with PHA (10 μg/ml) for 24 h (∗P < 0.05, ∗∗P < 0.01, coculture in the presence of Ki-24 CD70 antibody, sCD27 or zVAD-fmk compared with coculture plus control IgG, t test). C and D, the expression of CD27 (C) and CD70 (D) by lymphocyte subsets (CD4-, CD8-, CD14-, CD19-, and CD56-positive cells) under resting conditions or 48 h after stimulation with PHA (10 μg/ml) was assessed by double-staining with FITC- and PE-labeled antibodies. E, the lytic activity of PBMCs stimulated with glioma cells was determined by 51Cr release. Before the stimulation period (5 days), glioma cells were incubated with a nonspecific isotype control antibody or Ki-24 CD70 antibody (10 μg/ml) for 2 h. During the effector period (4 h), no antibody was present. The lytic activity of the remaining viable PBMCs is given for various E:T ratios with the number of 51Cr-labeled target cells kept constant. F, resting PBMCs were cultured in the presence of nonspecific IgG or agonistic CD27 antibody (10 μg/ml), without or with zVAD-fmk (100 μm), for 48 h. Dead cells were quantified by flow cytometry as in Fig. 4 C. In G, 105 irradiated HLA-A2-positive stimulator PBMCs were cocultured with 105 HLA-A2-negative responder PBMCs and 104 irradiated glioma cells. On day 4, [methyl-3H]thymidine (0.5 μCi/well) was added. [Methyl-3H]thymidine incorporation was measured 16 h later by liquid scintillation counting; bars, ±SD.

Fig. 5.

CD70-mediated apoptosis of PBMCs cocultured with glioma cells is mimicked by agonistic CD27 antibodies. A, PBMCs were labeled with 51Cr. After coculture with nonirradiated glioma cells (105 PBMCs; 104 glioma cells/well) for 18 h, 51Cr was determined in the supernatant. Assays were performed in the presence of an isotype-matched control antibody (10 μg/ml), Ki-24 CD70 antibody (10 μg/ml), sCD27 (50 units/ml), or zVAD-fmk (100 μm). B, as in A, but PBMCs (5 × 104) were activated by incubation with PHA (10 μg/ml) for 24 h (∗P < 0.05, ∗∗P < 0.01, coculture in the presence of Ki-24 CD70 antibody, sCD27 or zVAD-fmk compared with coculture plus control IgG, t test). C and D, the expression of CD27 (C) and CD70 (D) by lymphocyte subsets (CD4-, CD8-, CD14-, CD19-, and CD56-positive cells) under resting conditions or 48 h after stimulation with PHA (10 μg/ml) was assessed by double-staining with FITC- and PE-labeled antibodies. E, the lytic activity of PBMCs stimulated with glioma cells was determined by 51Cr release. Before the stimulation period (5 days), glioma cells were incubated with a nonspecific isotype control antibody or Ki-24 CD70 antibody (10 μg/ml) for 2 h. During the effector period (4 h), no antibody was present. The lytic activity of the remaining viable PBMCs is given for various E:T ratios with the number of 51Cr-labeled target cells kept constant. F, resting PBMCs were cultured in the presence of nonspecific IgG or agonistic CD27 antibody (10 μg/ml), without or with zVAD-fmk (100 μm), for 48 h. Dead cells were quantified by flow cytometry as in Fig. 4 C. In G, 105 irradiated HLA-A2-positive stimulator PBMCs were cocultured with 105 HLA-A2-negative responder PBMCs and 104 irradiated glioma cells. On day 4, [methyl-3H]thymidine (0.5 μCi/well) was added. [Methyl-3H]thymidine incorporation was measured 16 h later by liquid scintillation counting; bars, ±SD.

Close modal
Fig. 6.

CD70-expressing glioma cells promote sCD27 release by PBMCs. A, PBMCs (1.5 × 107) were cultured alone or cocultured with 1.5 × 106 irradiated glioma cells for 5 days. Glioma cells were preincubated with control IgG or CD70 antibody (10 μg/ml) for 2 h; alternatively, zVAD-fmk (100 μm) was added during coculture. sCD27 levels in the supernatant were determined by ELISA (P < 0.05, ∗∗P < 0.01, compared with control IgG). B, CD27 expression was assessed by flow cytometry using M-T271 CD27 antibody: in PBMCs cultured alone for 5 days (top panel), cocultured with U373 MG cells preexposed to control antibody (middle panel), or to CD70 antibody (bottom panel).

Fig. 6.

CD70-expressing glioma cells promote sCD27 release by PBMCs. A, PBMCs (1.5 × 107) were cultured alone or cocultured with 1.5 × 106 irradiated glioma cells for 5 days. Glioma cells were preincubated with control IgG or CD70 antibody (10 μg/ml) for 2 h; alternatively, zVAD-fmk (100 μm) was added during coculture. sCD27 levels in the supernatant were determined by ELISA (P < 0.05, ∗∗P < 0.01, compared with control IgG). B, CD27 expression was assessed by flow cytometry using M-T271 CD27 antibody: in PBMCs cultured alone for 5 days (top panel), cocultured with U373 MG cells preexposed to control antibody (middle panel), or to CD70 antibody (bottom panel).

Close modal
Table 1

CD70 expression in human glioma cell linesa

CD70/β-actin CD70 protein
mRNA ratioSFI
LN-18 0.0146 1.3 
U138MG 0.1925 10.3 
U87MG 0.0604 2.7 
LN-428 0.0479 2.5 
D247MG 0.0888 3.9 
T98G 0.0048 1.1 
LN-319 0.2747 15.5 
LN-229 0.0317 1.9 
A172 0.0288 1.5 
U251MG 0.0479 6.4 
U373MG 0.4613 31.7 
LN-308 0.0167 
CD70/β-actin CD70 protein
mRNA ratioSFI
LN-18 0.0146 1.3 
U138MG 0.1925 10.3 
U87MG 0.0604 2.7 
LN-428 0.0479 2.5 
D247MG 0.0888 3.9 
T98G 0.0048 1.1 
LN-319 0.2747 15.5 
LN-229 0.0317 1.9 
A172 0.0288 1.5 
U251MG 0.0479 6.4 
U373MG 0.4613 31.7 
LN-308 0.0167 
a

CD70 mRNA expression assessed by Northern blot analysis is expressed as ratios of optical density values normalized to β-actin (Fig. 1,A). CD70 protein quantified by flow cytometry is expressed as SFI values (Fig. 1 E).

Table 2

TNF-α release by cocultures of PBMCs and glioma cellsa

TNF-α [pg/ml]
PBMC alonePBMC + T98GPBMC + U138MGPBMC + U373MG
Control antibody 160 ± 32 992 ± 49b 73 ± 74 152 ± 49 
Ki-24 CD70 antibody 34 ± 86 947 ± 8b 665 ± 62b,c 615 ± 62b,c 
sCD27 148 ± 111 894 ± 74b 714 ± 123b,c 652 ± 111b,c 
zVAD-fmk 174 ± 37 1128 ± 37b,c 689 ± 49b,c 775 ± 62b,c 
TNF-α [pg/ml]
PBMC alonePBMC + T98GPBMC + U138MGPBMC + U373MG
Control antibody 160 ± 32 992 ± 49b 73 ± 74 152 ± 49 
Ki-24 CD70 antibody 34 ± 86 947 ± 8b 665 ± 62b,c 615 ± 62b,c 
sCD27 148 ± 111 894 ± 74b 714 ± 123b,c 652 ± 111b,c 
zVAD-fmk 174 ± 37 1128 ± 37b,c 689 ± 49b,c 775 ± 62b,c 
a

PBMCs (1.5 × 107) were cultured alone or cocultured with 1.5 × 106 irradiated T98G, U138MG, or U373MG cells in 5 ml RPMI 1640 10% FCS for 5 days. TNF-α release was determined by L-M bioassay.

b

P < 0.01, coculture compared to PBMCs alone.

c

P < 0.01, compared to coculture plus control IgG, t test.

1
Weller M., Fontana A. The failure of current immunotherapy for malignant glioma. Tumor-derived TGF-β, T-cell apoptosis, and the immune privilege of the brain.
Brain Res. Rev.
,
21
:
128
-151,  
1995
.
2
Leitlein J., Aulwurm S., Waltereit R., Naumann U., Wagenknecht B., Garten W., Weller M., Platten M. Processing of immunosuppressive pro-TGF-β1,2 by human glioblastoma cells involves cytoplasmic and secreted furin-like proteases.
J. Immunol.
,
166
:
7238
-7243,  
2001
.
3
Weller M., Kleihues P., Dichgans J., Ohgaki H. CD95 ligand: lethal weapon against malignant glioma?.
Brain Pathol.
,
8
:
285
-293,  
1998
.
4
Dietrich P. Y., Walker P. R., Saas P., de Tribolet N. Immunobiology of gliomas: new perspectives for therapy.
Ann. N. Y. Acad. Sci.
,
824
:
124
-140,  
1997
.
5
Saas P., Walker P. R., Hahne M., Quiquerez A. L., Schnuriger V., Perrin G., French L., Van Meir E. G., de Tribolet N., Tschopp J., Dietrich P. Y. Fas ligand expression by astrocytoma in vivo: maintaining immune privilege in the brain?.
J. Clin. Investig.
,
99
:
1173
-1178,  
1997
.
6
Roszman T., Elliott L., Brooks W. Modulation of T-cell function by gliomas.
Immunol. Today
,
12
:
370
-374,  
1991
.
7
Münz C., Naumann U., Grimmel C., Rammensee H. G., Weller M. TGF-β-independent induction of immunogenicity by decorin gene transfer in human malignant glioma cells.
Eur. J. Immunol.
,
29
:
1032
-1040,  
1999
.
8
Zou J. P., Morford L. A., Chougnet C., Dix A. R., Brooks A. G., Torres N., Shuman J. D., Coligan J. E., Brooks W. H., Roszman T. L., Shearer G. M. Human glioma-induced immunosuppression involves soluble factor(s) that alters monocyte cytokine profile and surface markers.
J. Immunol.
,
162
:
4882
-4892,  
1999
.
9
Bowman M. R., Crimmins M. A., Yetz-Aldape J., Kriz R., Kelleher K., Herrmann S. The cloning of CD70 and its identification as the ligand for CD27.
J. Immunol.
,
152
:
1756
-1761,  
1994
.
10
Goodwin R. G., Alderson M. R., Smith C. A., Armitage R. J., VandenBos T., Jerzy R., Tough T. W., Schoenborn M. A., Davis-Smith T., Hennen K., Falk B., Cosman D., Baker E., Sutherland G. R., Grabstein K. H., Farrah T., Giri J. G., Beckman M. P. Molecular and biological characterization of a ligand for CD27 defines a new family of cytokines with homology to tumor necrosis factor.
Cell
,
73
:
447
-456,  
1993
.
11
Akiba H., Oshima H., Takeda K., Atsuta M., Nakano H., Nakajima A., Nohara C., Yagita H., Okumura K. CD28-independent costimulation of T cells by OX40 ligand and CD70 on activated B cells.
J. Immunol.
,
162
:
7058
-7066,  
1999
.
12
Brugnoni D., Airo P., Marino R., Notarangelo L. D., van Lier R. A., Cattaneo R. CD70 expression on T-cell subpopulations: study of normal individuals and patients with chronic immune activation.
Immunol. Lett.
,
55
:
99
-104,  
1997
.
13
Lens S. M., Baars P. A., Hooibrink B., van Oers M. H., van Lier R. A. Antigen-presenting cell-derived signals determine expression levels of CD70 on primed T cells.
Immunology
,
90
:
38
-45,  
1997
.
14
Lens S. M., Tesselaar K., van Oers M. H., van Lier R. A. Control of lymphocyte function through CD27-CD70 interactions.
Semin. Immunol.
,
10
:
491
-499,  
1998
.
15
Orengo A. M., Cantoni C., Neglia F., Biassoni R., Ferrini S. Reciprocal expression of CD70 and of its receptor, CD27, in human long term-activated T and natural killer (NK) cells: inverse regulation by cytokines and role in induction of cytotoxicity.
Clin. Exp. Immunol.
,
107
:
608
-613,  
1997
.
16
Nagumo H., Agematsu K., Shinozaki K., Hokibara S., Ito S., Takamoto M., Nikaido T., Yasui K., Uehara Y., Yachie A., Komiyama A. CD27/CD70 interaction augments IgE secretion by promoting the differentiation of memory B cells into plasma cells.
J. Immunol.
,
161
:
6496
-6502,  
1998
.
17
Agematsu K., Hokibara S., Nagumo H., Shinozaki K., Yamada S., Komiyama A. Plasma cell generation from B-lymphocytes via CD27/CD70 interaction.
Leuk. Lymphoma
,
35
:
219
-225,  
1999
.
18
Stuhler G., Zobywalski A., Grünebach F., Brossart P., Reichardt V. L., Barth H., Stevanovic S., Brugger W., Kanz L., Schlossman S. F. Immune regulatory loops determine productive interactions within human T lymphocyte-dendritic cell clusters.
Proc. Natl. Acad. Sci. USA
,
96
:
1532
-1535,  
1999
.
19
Yang F. C., Agematsu K., Nakazawa T., Mori T., Ito S., Kobata T., Morimoto C., Komiyama A. CD27/CD70 interaction directly induces natural killer cell killing activity.
Immunology
,
88
:
289
-293,  
1996
.
20
Brown G. R., Meek K., Nishioka Y., Thiele D. L. CD27-CD27 ligand/CD70 interactions enhance alloantigen-induced proliferation and cytolytic activity in CD8+ T lymphocytes.
J. Immunol.
,
154
:
3686
-3695,  
1995
.
21
Raman V. S., Bal V., Rath S., George A. Ligation of CD27 on murine B cells responding to T-dependent and T-independent stimuli inhibits the generation of plasma cells.
J. Immunol.
,
165
:
6809
-6815,  
2000
.
22
Sugita K., Tanaka T., Doshen J. M., Schlossman S. F., Morimoto C. Direct demonstration of the CD27 molecule involved in the negative regulatory effect on T cell activation.
Cell. Immunol.
,
152
:
279
-285,  
1993
.
23
Kobata T., Jacquot S., Kozlowski S., Agematsu K., Schlossman S. F., Morimoto C. CD27-CD70 interactions regulate B-cell activation by T cells.
Proc. Natl. Acad. Sci. USA
,
92
:
11249
-11253,  
1995
.
24
Prasad K. V., Ao Z., Yoon Y., Wu M. X., Rizk M., Jacquot S., Schlossman S. F. CD27, a member of the tumor necrosis factor receptor family, induces apoptosis and binds to Siva, a proapoptotic protein.
Proc. Natl. Acad. Sci. USA
,
94
:
6346
-6351,  
1997
.
25
Hendriks J., Gravestein L. A., Tesselaar K., van Lier R. A., Schumacher T. N., Borst J. CD27 is required for generation and long-term maintenance of T cell immunity.
Nat. Immunol.
,
1
:
433
-440,  
2000
.
26
Arens R., Tesselaar K., Baars P. A., van Schijndel G. M., Hendriks J., Pals S. T., Krimpenfort P., Borst J., van Oers M. H., van Lier R. A. Constitutive CD27/CD70 interaction induces expansion of effector-type T cells and results in IFNγ-mediated B cell depletion.
Immunity
,
15
:
801
-812,  
2001
.
27
Ranheim E. A., Cantwell M. J., Kipps T. J. Expression of CD27 and its ligand, CD70, on chronic lymphocytic leukemia B cells.
Blood
,
85
:
3556
-3565,  
1995
.
28
van Oers M. H., Pals S. T., Evers L. M., van der Schoot C. E., Koopman G., Bonfrer J. M., Hintzen R. Q., von dem Borne A. E., van Lier R. A. Expression and release of CD27 in human B-cell malignancies.
Blood
,
82
:
3430
-3436,  
1993
.
29
Zambello R., Trentin L., Facco M., Siviero M., Galvan S., Piazza F., Perin A., Agostini C., Semenzato G. Analysis of TNF-receptor and ligand superfamily molecules in patients with lymphoproliferative disease of granular lymphocytes.
Blood
,
96
:
647
-654,  
2000
.
30
Naumann U., Durka S., Weller M. Dexamethasone-mediated protection from drug cytotoxicity: association with p21WAF1/CIP1 protein accumulation?.
Oncogene
,
17
:
1567
-1575,  
1998
.
31
Huang H., Colella S., Kurrer M., Yonekawa Y., Kleihues P., Ohgaki H. Gene expression profiling of low-grade diffuse astrocytomas by cDNA arrays.
Cancer Res.
,
60
:
6868
-6874,  
2000
.
32
Agematsu K., Kobata T., Sugita K., Freeman G. J., Beckmann M. P., Schlossman S. F., Morimoto C. Role of CD27 in T cell immune response. Analysis by recombinant soluble CD27.
J. Immunol.
,
153
:
1421
-1429,  
1994
.
33
Streffer, J. R., Rimner, A., Rieger, J., Naumann, U., Rodemann, H. P., and Weller, M. Bcl-2 family protein expression modulates radiosensitivity in human glioma cells. J. Neuro-Oncol., in press, 2002.
34
Schmidt F., Rieger J., Wischhusen J., Naumann U., Weller M. Glioma cell sensitivity to topotecan: the role of p53 and topotecan-induced DNA damage.
Eur. J. Pharmacol.
,
412
:
21
-25,  
2001
.
35
Lens S. M., Drillenburg P., den Drijver B. F., van Schijndel G., Pals S. T., van Lier R. A., van Oers M. H. Aberrant expression and reverse signalling of CD70 on malignant B cells.
Br. J. Haematol.
,
106
:
491
-503,  
1999
.
36
Fadok V. A., Bratton D. L., Rose D. M., Pearson A., Ezekewitz R. A., Henson P. M. A receptor for phosphatidylserine-specific clearance of apoptotic cells.
Nature (Lond.)
,
405
:
85
-90,  
2000
.
37
Glaser T., Wagenknecht B., Groscurth P., Krammer P. H., Weller M. Death ligand/receptor-independent caspase activation mediates drug-induced cytotoxic cell death in human malignant glioma cells.
Oncogene
,
18
:
5044
-5053,  
1999
.
38
Hintzen R. Q., de Jong R., Hack C. E., Chamuleau M., de Vries E. F., ten Berge I. J., Borst J., van Lier R. A. A soluble form of the human T cell differentiation antigen CD27 is released after triggering of the TCR/CD3 complex.
J. Immunol.
,
147
:
29
-35,  
1991
.
39
Loenen W. A., De Vries E., Gravestein L. A., Hintzen R. Q., Van Lier R. A., Borst J. The CD27 membrane receptor, a lymphocyte-specific member of the nerve growth factor receptor family, gives rise to a soluble form by protein processing that does not involve receptor endocytosis.
Eur. J. Immunol.
,
22
:
447
-455,  
1992
.
40
Hintzen R. Q., van Lier R. A., Kuijpers K. C., Baars P. A., Schaasberg W., Lucas C. J., Polman C. H. Elevated levels of a soluble form of the T cell activation antigen CD27 in cerebrospinal fluid of multiple sclerosis patients.
J. Neuroimmunol.
,
35
:
211
-217,  
1991
.
41
Nieland J. D., Graus Y. F., Dortmans Y. E., Kremers B. L., Kruisbeek A. M. CD40 and CD70 co-stimulate a potent in vivo antitumor T cell response.
J. Immunother.
,
21
:
225
-236,  
1998
.
42
Couderc B., Zitvogel L., Douin-Echinard V., Djennane L., Tahara H., Favre G., Lotze M. T., Robbins P. D. Enhancement of antitumor immunity by expression of CD70 (CD27 ligand) or CD154 (CD40 ligand) costimulatory molecules in tumor cells.
Cancer Gene Ther.
,
5
:
163
-175,  
1998
.
43
Lorenz M. G., Kantor J. A., Schlom J., Hodge J. W. Anti-tumor immunity elicited by a recombinant vaccinia virus expressing CD70 (CD27L).
Hum. Gene Ther.
,
10
:
1095
-1103,  
1999
.
44
Lenardo M., Chan K. M., Hornung F., McFarland H., Siegel R., Wang J., Zheng L. Mature T lymphocyte apoptosis-immune regulation in a dynamic and unpredictable antigenic environment.
Annu. Rev. Immunol.
,
17
:
221
-253,  
1999
.
45
Siegel R. M., Chan F. K., Chun H. J., Lenardo M. J. The multifaceted role of Fas signaling in immune cell homeostasis and autoimmunity.
Nat. Immunol.
,
1
:
469
-474,  
2000
.
46
Jacquot S., Kobata T., Iwata S., Schlossman S. F., Morimoto C. CD27/CD70 interaction contributes to the activation and the function of human autoreactive CD27+ regulatory T cells.
Cell. Immunol.
,
179
:
48
-54,  
1997
.
47
Yoon Y., Ao Z., Cheng Y., Schlossman S. F., Prasad K. V. Murine Siva-1 and Siva-2, alternate splice forms of the mouse Siva gene, both bind to CD27 but differentially transduce apoptosis.
Oncogene
,
18
:
7174
-7179,  
1999
.
48
Nakajima A., Oshima H., Nohara C., Morimoto S., Yoshino S., Kobata T., Yagita H., Okumura K. Involvement of CD70-CD27 interactions in the induction of experimental autoimmune encephalomyelitis.
J. Neuroimmunol.
,
109
:
188
-196,  
2000
.
49
Font J., Pallares L., Martorell J., Martinez E., Gaya A., Vives J., Ingelmo M. Elevated soluble CD27 levels in serum of patients with systemic lupus erythematosus.
Clin. Immunol. Immunopathol.
,
81
:
239
-243,  
1996
.
50
Hintzen R. Q., Paty D., Oger J. Cerebrospinal fluid concentrations of soluble CD27 in HTLV-I associated myelopathy and multiple sclerosis.
J. Neurol. Neurosurg. Psychiatry
,
66
:
791
-793,  
1999
.
51
Tak P. P., Hintzen R. Q., Teunissen J. J., Smeets T. J., Daha M. R., van Lier R. A., Kluin P. M., Meinders A. E., Swaak A. J., Breedveld F. C. Expression of the activation antigen CD27 in rheumatoid arthritis.
Clin. Immunol. Immunopathol.
,
80
:
129
-138,  
1996
.