Characterization of Tumor Reactivity of Human Vγ9Vδ2 γδ T Cells In Vitro and in SCID Mice In Vivo

Dieter Kabelitz, Daniela Wesch, Elke Pitters and Margot Zöller
J Immunol December 1, 2004, 173 (11) 6767-6776; DOI: https://doi.org/10.4049/jimmunol.173.11.6767

Abstract

Human Vγ9Vδ2 γδ T cells are selectively activated by bacterial phosphoantigens and aminobisphosphonates and exert potent cytotoxicity toward various tumor cells. In this study we have characterized the cytotoxic reactivity of γδ T cell lines established from healthy donors by stimulation with aminobisphosphonate alendronate toward melanoma MeWo and pancreatic adenocarcinomas Colo357 and PancTu1 lines in vitro and in vivo upon adoptive transfer into SCID mice. Lysis of all tumor cells was enhanced when γδ effector cells were preactivated with phosphoantigens. Recognition of MeWo was TCR dependent, as shown by anti-TCR Ab blockade, whereas only the phosphoantigen-mediated increased, but not the basal, lysis of Colo357 and PancTu1 was inhibited by anti-TCR Ab. Furthermore, lysis of Colo357, but not that of MeWo or PancTu1, was completely inhibited by the pan-caspase inhibitor zVAD, indicating different recognition and effector mechanisms involved in the γδ T cell/tumor cell interactions. Upon transfer into SCID mice, alendronate-activated γδ T cells given together with IL-2 and alendronate significantly prolonged the survival of SCID mice inoculated with human tumor cells. The best results were thus obtained when γδ T cells were repetitively given five times over a period of 30 days. With this protocol, human γδ T cells prolonged the mean survival of mice inoculated with MeWo melanoma from 28.5 to 87.3 days (p < 0.0001) and in the case of PancTu1 adenocarcinoma from 23.0 to 48.4 days (p < 0.0001). We conclude that an effective γδ T cell-based immunotherapy might require activation of endogenous γδ T cells with aminobisphosphonate (or phosphoantigen) and IL-2, followed by adoptive transfer of in vitro expanded γδ T cells.

Two general concepts have been put forward for cellular immunotherapy of malignant diseases, i.e., immunization strategies with the aim of eliciting and boosting specific antitumor immune responses, and adoptive transfer of effector cells with potent antitumor activity, such as CTL or NK cells. Tumor-specific Ags have been identified (1), and various strategies of tumor vaccination with peptide-, DNA-, or dendritic cell-based vaccines have entered clinical evaluation, with only limited efficacy to date (2). Alternatively, adoptive immunotherapy based on the transfer of large numbers of in vitro expanded antitumor effector cells, including tumor-specific, MHC class I-restricted CD8+ CTL and NK cells is under investigation (3, 4). In view of the limited clinical success of cancer immunotherapy to date, new strategies need to be explored. Apart from conventional MHC class I-restricted CD8+ CTL and classical CD3-negative NK cells, unconventional T cells, including NKT cells and γδ T cells, might contribute to the immune defense against tumors. In fact, recent evidence indicates that γδ T cells play a significant role in the local immune surveillance and, moreover, can be activated to kill a variety of tumor cells.

Approximately 5–10% of peripheral blood CD3+ T cells express a heterodimeric γδ TCR instead of the conventional αβ TCR. Most γδ T cells lack CD4 and CD8 Ags, in line with their MHC-nonrestricted mode of ligand recognition (5). Fifty to 95% of all γδ T cells in the peripheral blood of healthy adults express Vδ2 paired with Vγ9, whereas Vδ1 cells coexpressing various Vγ genes are primarily localized within the epithelia of the small intestine (6, 7, 8). Human Vγ9Vδ2 T cells recognize phosphorylated metabolites of the bacterial nonmevalonate isoprenoid biosynthesis pathway, so-called phosphoantigens, as well as various tumor cells, including Daudi Burkitt’s lymphoma (9, 10, 11, 12, 13, 14). Vδ1 cells, the second most frequent subset of human γδ T cells, recognize stress-induced MHC class I-related genes, MHC class I chain-related protein A (MICA) and B (MICB) (15). MICA/MICB are also ligands for the NK cell receptor NKG2D, which is expressed on some Vδ1 and Vδ2 γδ T cells and delivers an activating signal (16, 17). Therefore, the cellular response of Vδ1 T cells toward MICA/MICB-expressing cells is integrated from signals generated through TCR- as well as NKG2D-dependent ligand recognition (18). Additional cellular Ags recognized by human γδ T cells include members of the CD1 family, notably CD1c (19). The effector functions of γδ T cells are not fundamentally different from those of αβ T cells. γδ T cells produce a variety of cytokines and exert potent cytotoxic activity toward various target cells, including many tumor cells (5). Some γδ T cells produce additional cytokines that are not commonly made by αβ T cells, such as keratinocyte growth factor-1 and fibroblast growth factor-9, which presumably contribute to the role of local γδ T cells in tissue repair (20, 21, 22).

Although their precise function in the immune system is not yet completely understood, substantial evidence indicates a pivotal role for γδ T cells in monitoring the integrity of epithelial cells and thus in tumor defense. Experiments with gene-targeted TCRδ−/− mice clearly revealed increased skin tumor rates upon exposure to the tumor promoter PMA or inoculation of squamous cell carcinoma (23). The role of γδ T cells in tumor defense was also revealed when TCRδ−/− mice were inoculated with B16F10 melanoma cells or treated with methylcholanthrene (24). Furthermore, it was recently shown that γδ T cells (together with NK cells), but not αβ T cells, rejected spontaneous disseminated B cell lymphomas inoculated into a variety of gene-targeted recipient mice (25). γδ T cells, which kill autologous tumor cells, have also been isolated from tumor-infiltrating lymphocytes in various solid tumors, including renal carcinoma, colorectal cancer, and lung carcinoma (26, 27, 28, 29), and from leukemia patients (30). Therefore, it is tempting to explore the potential of γδ T cells for the immunotherapy of human tumors. Several studies have demonstrated the capacity of human γδ T cells upon adoptive transfer to improve the survival of SCID mice inoculated with human nasopharyngeal carcinoma, melanoma, or Daudi lymphoma cells (31, 32, 33). Such strategies are aided by the availability of synthetic phosphoantigens that are potent and selective activators of human Vγ9Vδ2 T cells (34). Interestingly, aminobisphosphonates, which are in clinical use for the treatment of osteoporosis and bone metastasis, are also potent activators of the very same Vγ9Vδ2 T cell population (35, 36, 37, 38). This raises the possibility that some already licensed drugs might be used to boost tumor-reactive γδ T cells in vivo or, alternatively, to expand tumor-reactive γδ T cells in vitro for subsequent adoptive immunotherapy (39, 40).

In the present study we have investigated the tumor reactivity of human Vγ9Vδ2 T cells activated and expanded in vitro with the aminobisphosphonate alendronate against three solid tumor cell lines. Upon adoptive transfer into SCID mice, optimal therapeutic efficacy against the melanoma MeWo and the pancreatic adenocarcinoma PancTu1 required the repetitive transfer of alendronate-activated γδ T cells. Based on our results, we propose that an effective γδ T cell-based tumor immunotherapy might require both the endogenous activation of γδ T cells with aminobisphosphonates or synthetic phosphoantigens and an adoptive transfer of in vitro expanded γδ T cells.

Materials and Methods

Activation of peripheral blood γδ T cells and establishment of γδ T cell lines

PBMC from healthy adult blood donors were isolated by Ficoll-Hypaque density gradient centrifugation. PBMC (1 × 106/ml) were cultured in medium RPMI 1640 supplemented with 2 mM l-glutamine, antibiotics, 10 mM HEPES, and 10% heat-inactivated human serum and incubated at 37°C in a humidified atmosphere of 5% CO2 in air. PBMC were stimulated with bromohydrin pyrophosphate (BrHPP)3 (200 nM) or aminobisphosphonate alendronate (5 μM) in the presence of 100 U/ml rIL-2 as indicated in Results. To establish Vγ9Vδ2 T cell lines, phosphoantigen- or alendronate-activated PBMC were restimulated twice at 14-day intervals with the respective Ag in the presence of irradiated PBMC feeder cells and rIL-2. Both protocols induced selective and dramatic expansion of Vγ9Vδ2 T cells. After the second restimulation, γδ T cells were separated from remaining non-γδ T cells using the MACS system (Miltenyi Biotec, Bergisch Gladbach, Germany) to deplete CD4+, CD8+, TCRαβ+, and CD56+ cells. Additional large-scale expansion of Vγ9 cells (that continued to coexpress Vδ2) was performed in rIL-2-supplemented medium with repetitive restimulation using PHA and irradiated PBMC and EBV-transformed B cell lines as feeder cells as previously described (41). Dead feeder cells were removed 3–5 days after restimulation by Ficoll-Hypaque density gradients.

Determination of γδ T cell expansion in vitro

A previously described flow cytometry method, termed standard cell dilution assay (SCDA), was used to selectively measure the expansion of Vγ9Vδ2 T cells among the phosphoantigen- or alendronate-stimulated PBMC (42). Briefly, samples were removed from cell cultures 5–8 days after initiation, washed, and stained with biotinylated anti-Vγ9 mAb 7A5 (43), followed by PE-labeled streptavidin. Shortly before analysis in a FACScan (BD Bioscience, Heidelberg, Germany), propidium iodide (0.2 μg/ml) and a known number of FITC-labeled and fixed standard cells were added. Based on the ratio to FITC-positive standard cells, the absolute number of viable Vγ9 T cells per microculture well was calculated (42). The proliferation of peritoneal exudate cells (PEC) and spleen cells (SC) isolated ex vivo from SCID mice inoculated with human tumor and γδ T cells was measured by uptake of tritiated thymidine during 8-h culture.

Flow cytometry

The following mAb were used for characterization of γδ T cell lines or tumor cells: Vδ2, CD94, and NKG2A were obtained from Immunotech (Marseille, France); Vδ1 was obtained from Endogen (Woburn, MA); NKG2D was obtained from R&D Systems (Wiesbaden, Germany), MICA was obtained from Immatics (Tubingen, Germany); and CD95/Fas and pan anti-TCRγδ mAb TCRγδ-1 were obtained from BD Bioscience (Heidelberg, Germany). HLA class I mAb w6/32 (American Type Culture Collection, Manassas, VA), anti-H-2Kd (K9-18; provided by Dr. G. Hammerling, German Cancer Research Institute) (44), and anti-Vγ9 mAb 7A5 (43) were purified from hybridoma supernatants. Cells incubated with unconjugated mAb were subsequently incubated with the appropriate anti-mouse PE-labeled secondary Ab. Negative controls were incubated with a nonbinding primary Ab and the same secondary reagents. For double-fluorescence analysis, cells were first labeled with a biotinylated Ab and PE-labeled streptavidin, followed by exposure to the second, FITC-labeled Ab. Flow cytometry followed routine procedures. The production of human cytokines (IFN-γ and TNF-α) by PEC and SC isolated from SCID mice was determined by intracellular flow cytometry essentially as previously described (45). Briefly, adherent cell-depleted (two periods of 1-h plastic adherence) PEC and SC were cultured for 24 h in the presence of 20 U/ml rIL-2 and 5 μM alendronate. Thereafter, cells were fixed, permeabilized, and stained with biotinylated anti-human TNF-α or IFN-γ mAb (BD Biosciences), followed by PE-conjugated streptavidin (45). All samples were measured on FACScan or FACSCalibur flow cytometers (BD Biosciences, Heidelberg, Germany) using CellQuest software.

Cytotoxicity assay

MeWo (melanoma; ECACC), Colo357 (pancreatic adenocarcinoma; ECACC), and PancTu1 (pancreatic adenocarcinoma; provided by Dr. H. Kalthoff, Section of Molecular Oncology, Universitätsklinikum Schleswig-Holstein Campus Kiel, Kiel, Germany) were labeled with 100 μCi of sodium 51Cr and used as targets in standard 4-h 51Cr release assay with titrated numbers of γδ T cells as effector cells. Specific lysis was calculated as follows: % specific lysis = cpmtest − cpmspontaneous/cpmmax −cpmspontaneous, where spontaneous release was determined in medium only, and maximal release was determined in Triton-lysed target cells. To determine the Fas sensitivity of tumor cells, 51Cr-labeled target cells were incubated with the agonistic anti-Fas/CD95 mAb CH11 (1 μg/ml; Kamiya Biomedical, Seattle, WA). To investigate the modulation of cytotoxic activity, effector cells were preincubated for 30–60 min before the assay with one of the following: anti-Vγ9 mAb 7A5 or control mouse IgG (Sigma-Aldrich, St. Louis, MO; 10 μg/ml); pan-caspase inhibitor zVAD-fmk (100 μM; Bachem, Bubendorf, Switzerland); phosphoantigen BrHPP (200 nM; Innate Pharma, Marseille, France); or alendronate (5 μM; Merck, Rahway, NJ).

SCID mouse/human tumor model

For in vivo experiments, the melanoma line MeWo and the pancreatic adenocarcinoma line PancTu1 were used. SCID mice were conditioned by irradiation (300 rad) and anti-asialo-GM1 (WAKO, Osaka, Japan) treatment according to the manufacturer’s suggestions (46). Mice received a single i.p. injection of 5 × 106 MeWo or PancTu1 cells. According to the i.p. application of titrated numbers of these two human tumor lines in SCID mice, 5 × 106 tumor cells were in the range of 10-fold (MeWo) and 20-fold (PancTu1) the tumor dose 100% take. The high numbers of tumor cells were inoculated to allow for full-fledged tumor growth within the period before graft-vs-host (GvH) reactions developed in the SCID mouse reconstituted with human PBL. Mice received concomitantly with the tumor cells an i.p. injection of human rIL-2 (300 ng) or bisphosphonate alendronate (10 μg) plus rIL-2 (300 ng). The application of rIL-2 and bisphosphonate was repeated at 4-day intervals. Where indicated, mice received, in addition, γδ T cells (described above) on day 0 (2 × 107), day 4 (5 × 106), day 10 (6 × 106), day 20 (1 × 107), and day 30 (1 × 107). In our experience, the danger of lethal GvH reactions rapidly increases with additional applications of human PBL. In fact, three of the mice receiving 5-fold γδ T cells died within 5 days after the last γδ T cell application without any sign of tumor growth, but evidence for GvH disease (destruction of gut epithelium, bloody stool, and icteric liver). These mice were not included in the presented experiments.

Results

In vitro activation of γδ T cells by alendronate

PBMC from healthy adult blood donors containing 3–6% γδ T cells were cultured with titrated concentrations of the phosphoantigen BrHPP (34) or alendronate in the presence of 100 U/ml rIL-2. The cellular expansion of Vγ9 γδ T cells was monitored by SCDA, as described in Materials and Methods. As shown for three representative donors in Fig. 1⇓, 1–10 μM alendronate induced a dramatic expansion of Vγ9 T cells. At an optimal concentration (5 μM), alendronate stimulated selective growth of Vγ9 T cells to almost similar levels as did optimal concentrations (200 nM) of the synthetic phosphoantigen BrHPP. The outgrowth of γδ T cells in response to alendronate required the addition of exogenous IL-2, because γδ T cells did not produce T cell growth factors after alendronate or phosphoantigen stimulation (not shown) (47). γδ T cell lines initiated by alendronate stimulation were propagated in rIL-2-supplemented medium and restimulated with alendronate and irradiated feeder cells twice at 14-day intervals. At that time, cell cultures contained 50–80% Vγ9 γδ T cells. To obtain sufficient cell numbers for adoptive transfer experiments, Vγ9 cells were purified from these cultures by negative MACS and further expanded with occasional restimulation as required (41). With this protocol, Vγ9 T cells could be readily expanded to 2 × 109 cells, starting from 50 ml of blood.

FIGURE 1.
FIGURE 1.

In vitro expansion of Vγ9 γδ T cells in response to BrHPP and alendronate. PBMC (105) from three healthy donors (a, b, and c) were cultured in 96-well microtiter plates with BrHPP (200 nM) or the indicated concentration of alendronate in the presence of 100 U/ml rIL-2. Absolute numbers of viable Vγ9 T cells per microculture well were measured by SCDA on day 6 or 7. Results represent the mean of duplicate determinations.

In vitro cytotoxicity of Vγ9Vδ2 γδ T cell lines

We used the melanoma line MeWo and the pancreatic adenocarcinomas PancTu1 and Colo357 as targets for Vγ9 γδ T cells in vitro. As shown in Fig. 2⇓, all three cell lines strongly expressed HLA class I and variable amounts of MICA. PancTu1 and Colo357 were strongly Fas/CD95-positive, whereas MeWo lacked CD95 expression. Fas sensitivity was tested in a 4-h 51Cr release assay using the agonistic anti-CD95 mAb CH11. Although Colo357 was Fas-sensitive, both MeWo and PancTu1 proved Fas-resistant (not shown). The Fas resistance of PancTu1 despite strong surface expression is well in line with the known defect in Fas signaling in these cells (48). Vγ9Vδ2 T cell lines were established from several healthy donors by initial activation and restimulation with alendronate, purified by negative selection, and further expanded as described in Materials and Methods. The cells uniformly expressed Vδ2, Vγ9, NKG2D, and variable levels of CD94, NKG2A, and CD56, and lacked CD16 and Vδ1 expression (Fig. 3⇓a). These γδ T cell lines were tested for cytotoxicity in a standard 4-h 51Cr release assay against the various tumor targets. As shown in Fig. 3⇓b, MeWo and Colo357 were efficiently lysed by the various γδ T cell lines, whereas PancTu1 was lysed by some, but not all, lines tested. To characterize the cytotoxic activity in more detail, we analyzed the effects of anti-TCR mAb blockade, preactivation of γδ T cells with phosphoantigen, and caspase inhibition with the pan-caspase inhibitor zVAD. The results of representative experiments are presented in Fig. 4⇓. In contrast to PancTu1 and Colo357, lysis of MeWo was consistently inhibited by anti-TCR mAb 7A5 (but not by control Ig), suggesting TCR-dependent target cell recognition (Fig. 4⇓a). The cytolytic activity of Vγ9Vδ2 T cells can be enhanced by phosphoantigens (49, 50). As shown in Fig. 4⇓b, preincubation of γδ effector cells with BrHPP or alendronate enhanced the killing of all three tumor cells. In this regard, BrHPP was generally more effective than alendronate. Notably, the preactivation of γδ T cells with BrHPP or alendronate induced significant killing of PancTu1 cells even in cases where only minimal lysis was observed in the absence of phosphoantigens (see Fig. 4⇓b). Finally, we observed striking differences with regard to inhibition of cytotoxicity by the pan-caspase inhibitor zVAD. As shown in Fig. 4⇓c, killing of Colo357 by γδ T cells was completely inhibited by 100 μM zVAD, whereas lysis of MeWo and PancTu1 was not significantly affected. Next, we asked whether the increased lysis after phosphoantigen preincubation (Fig. 4⇓b) involved TCR-dependent target recognition in cases where lysis in the absence of phosphoantigen was not inhibited by anti-TCR mAb. To this end, EP3 or DU γδ T cell lines were incubated for 30 min in the absence or the presence of BrHPP. Thereafter, control Ig or anti-Vγ9 mAb 7A5 was added for another 30 min before tumor target cells were added. As shown in Fig. 5⇓a (left), lysis of Colo357 by EP3 cells in the absence of BrHPP was not inhibited by 7A5 mAb (in line with the results in Fig. 4⇓a). In contrast, BrHPP-increased lysis was substantially inhibited by 7A5 mAb (but not by control Ig), but strikingly only to the level of spontaneous lysis mediated in the absence of BrHPP (Fig. 5⇓a, right; compare with Fig. 5⇓a, left). This suggests that killing of Colo357 by EP3 cells involves TCR-dependent (i.e., induced by BrHPP and inhibited by anti-TCR mAb) and TCR-independent (not inhibited by anti-TCR mAb) target recognition. In contrast, spontaneous (TCR-dependent or -independent) cytotoxicity was minimal when DU γδ T cells were tested against PancTu1 target cells (Fig. 5⇓b, left). Again, lysis was drastically increased when effector cells were preincubated with BrHPP, and the phosphoantigen-triggered cytotoxicity was inhibited by 7A5 mAb, but not by control Ig (Fig. 5⇓b, right).

FIGURE 2.
FIGURE 2.

Phenotypic characterization of the tumor cells used in this study. Colo357, MeWo, and PancTu1 were stained with mAb against HLA class I (w6/32), MICA, and CD95/Fas as indicated (solid lines). Appropriate isotype controls are shown as dotted lines.

FIGURE 3.
FIGURE 3.

Phenotype of established γδ T cell lines and cytolytic effector function. γδ T cell lines were generated as described in Materials and Methods. Cell lines were stained with the indicated mAb (solid lines) or appropriate isotype controls (dotted lines) and analyzed on a FACScan (a). The Vγ9Vδ2 T cell lines EP3 (▴), JQ (○), DU (•), HO (□), and DK (▵) were tested at the indicated E:T cell ratios against 51Cr-labeled target cells MeWo, PancTu1, and Colo357 as indicated (b). The lines were repeatedly tested with comparable results.

FIGURE 4.
FIGURE 4.

Modulation of tumor cell lysis. γδ T cell lines were preincubated for 30 min with a) 10 μg/ml anti-Vγ9 mAb 7A5 (○) or control IgG (▵); b) 200 nM BrHPP (○) or 5 μM alendronate (▵); or c) 100 μM zVAD (○) before the addition of 51Cr-labeled tumor targets MeWo, PancTu1, and Colo357 as indicated. Lysis in the absence of these reagents is indicated (•). The results are representative of three to five experiments performed with the Vγ9Vδ2T cell lines JQ (Mewo, Colo357), DK (MeWo,PancTu1, Colo357), EP3 (PancTu1), and the Vγ9 clone A36DN8 (PancTu1).

FIGURE 5.
FIGURE 5.

Inhibition of phosphoantigen-mediated increase in cytotoxicity by anti-TCR mAb. EP3 (a) and DU (b) γδ T cells were incubated for 30 min with 200 nM BrHPP or medium. Thereafter, 10 μg/ml anti-Vγ9 mAb 7A5 (○) or control Ig (▵) was added for another 30 min before 51Cr-labeled target cells Colo357 (a) or PancTu1 (b) were added. Control lysis in the absence of these reagents is indicated (•). One of two similar experiments is shown.

In vivo antitumor activity of Vγ9Vδ2 γδ T cell lines

The SCID mouse/human tumor model system was used to analyze the antitumor activity of γδ T cells in vivo against MeWo melanoma and PancTu1 pancreatic adenocarcinoma. In the first experiment, groups of seven or eight mice received 5 × 106 MeWo cells i.p. together with 2 × 107 EP3 γδ T cells (which displayed intermediate cytotoxic activity toward MeWo in vitro; see Fig. 3⇑b) i.p. on day 0, 8 × 106 EP3 γδ T cells on day 4, and human rIL-2 (300 ng) and/or 10 μg alendronate on days 0, 4, 11, 25, 32, and 39. As presented in Fig. 6⇓, γδ T cells given alone on days 0 and 4 did not improve survival (mean survival time, 29.4 ± 5.4 vs 30.4 ± 6.2 days). In contrast, survival was significantly prolonged when alendronate with or without additional rIL-2 was given together with γδ T cells (mean survival time, 52.3 ± 19.3 days with alendronate, 48.9 ± 22.3 days with alendronate plus rIL-2). Although mice receiving MeWo cells only developed miliary tumor nodules in the peritoneal cavity, bloody ascites, and multiple liver metastases, it is of interest that five of seven mice receiving γδ T cells showed few, but large, solid tumor nodules in the peritoneal cavity, did not develop ascites and had only two to five small metastatic nodules in the liver. Multiple liver metastases in γδ T cell receiving mice were seen only in the two mice that survived for >70 days. The latter observations could have indicated that tumor cells possibly started to expand only after exhaustion of γδ T cells. To support the hypothesis and also to determine whether alendronate alone exerted some antitumor activity in this setting (51), we performed two additional in vivo experiments with another Vγ9 γδ T cell line that was given repeatedly and included appropriate controls for possible effects of alendronate and rIL-2. As shown in Fig. 7⇓a, alendronate and rIL-2 alone or in combination did not have any effect on the survival of SCID mice inoculated i.p. with 5 × 106 MeWo cells. As in the previous experiment, the application of DK γδ T cells on days 0 and 4 (together with alendronate and rIL-2) significantly prolonged the mean survival time. Much longer survival was achieved, however, when γδ T cells were given repeatedly at 10-day intervals. In this study the mean survival time was prolonged from 28.5 to 87.3 days, whereas mice receiving γδ T cells on days 0 and 4 only had a mean survival time of 44.9 days. The difference in the mean survival time of mice receiving γδ T cells two vs five times was highly significant (p = 0.0021). Similar to the previous experiment, mice receiving γδ T cells developed only few, rather than many, tumor nodules. Repeated applications of γδ T cells were also accompanied by a corresponding decrease in liver metastases. One mouse receiving five doses of γδ T cells was killed 180 days after tumor cell application and was tumor free, and an additional four mice of this group had no visible metastases at autopsy. Very similar results were obtained with PancTu1 tumor cells (Fig. 7⇓b). Again, alendronate and rIL-2, alone or in combination, did not significantly increase the survival of SCID mice. Repetitive inoculation of DK γδ T cells (displaying strong in vitro cytotoxicity against PancTu1; see Fig. 3⇑b) significantly increased the mean survival time, and there was a clear correlation between the frequency of γδ T cell application and survival, as shown in Fig. 7⇓b. The best results were thus obtained when γδ T cells were given five times, on days 0, 4, 10, 20, and 30, which increased the mean survival time to 48.4 ± 6.4 days compared with the survival time of 25.4 ± 2.8 days in the alendronate plus IL-2 only group (p < 0.0001) and the survival time of 32.0 ± 2.4 days of mice receiving two treatments with γδ T cells (p = 0.0007). PancTu1 cells settled frequently in the pancreatic gland, but rarely formed liver metastasis. However, the mice suffered from cachexia and became icteric. Together, these results demonstrate that human Vγ9Vδ2 γδ T cells exert significant antitumor activity against two different human tumor types in the adoptive SCID mouse transfer model, but require repetitive transfer for optimal results.

FIGURE 6.
FIGURE 6.

In vivo activity of γδ T cells against MeWo tumor. SCID mice were irradiated with 300 rad and received anti-asialo-GM1 2 days before i.p. inoculation of 5 × 106 MeWo cells. Mice received an i.p. injection of 2 × 107 Vγ9-expressing EP3 γδ T cells on day 0 and 8 × 106 Vγ9-expressing EP3 γδ T cells on day 4. As indicated, mice received 300 ng of human rIL-2 and/or 10 μg alendronate i.p. on days 0, 4, 11, 18, 25, 32, and 39. Mean survival time was 30.4 ± 6.2 days for the control group (MeWo only); 29.8 ± 5.4 days for the MeWo plus γδ T cells group; 52.3 ± 19.3 days for the MeWo, γδ T cells, plus alendronate group; and 48.9 ± 22.3 days for the MeWo, γδ T cells, alendronate, plus rIL-2 group.

FIGURE 7.
FIGURE 7.

Repetitive application of human γδ T cells prolongs survival of SCID mice transplanted with human tumors. Groups of eight (a) or five (b) SCID mice were inoculated with 5 × 106 MeWo (a) or PancTu1 (b) tumor cells. At the indicated time points, 107 DK (a and b) γδ T cells were given i.p. together with 300 ng of rIL-2 and 10 μg of alendronate.

Characterization of human cells after adoptive transfer into SCID mice

To follow the fate of γδ T cells as well as their state of activity in the tumor-bearing mice, two animals per group were killed at various time points after the last transfer of γδ T cells. The application of rIL-2 and bisphosphonate had no impact on the number of cells recovered from the PEC or SC, and leukocytes could not be recovered from remnant lymph node cells (LNC). The number of recovered PEC increased steadily by the repeated transfer of human γδ T cells (Fig. 8⇓a). In the spleen, a major increase in recovery was only observed after repeated application of γδ T cells (Fig. 8⇓b). Also, LNC, albeit few, were only recovered in SCID mice receiving four or five doses of γδ T cells (data not shown). The recovery of leukocytes was independent of whether the mice had been inoculated with PancTu1 or MeWo.

FIGURE 8.
FIGURE 8.

Recovery and phenotype of lymphoid cells from SCID mice inoculated with human tumor cells and γδ T cells. The numbers of PEC and SC recovered from SCID mice on day 20 (mice receiving rIL-2, alendronate, or alendronate plus rIL-2) or day 34 (mice receiving two, three, or four injections of γδ T cells, alendronate, plus rIL-2) after inoculation with PancTu1 or MeWo cells are shown in a and b. The origin of the recovered PEC and SC was determined by flow cytometry with mAb directed against murine H-2Kd and human MHC class I. Mice that had received γδ T cells, alendronate, plus rIL-2 two times (days 0 and 4), three times (days 0, 4, and 10), or four times (days 0, 4, 10, and 20) were killed on day 34. The total number of recovered cells as well as the number of HLA class I- and H-2Kd-positive cells are shown in c.

We have shown previously that human αβ T cells survive for a limited period of time in the SCID mouse (52). This also holds true for γδ T cells, as apparent by the evaluation of the percentage of human and mouse MHC class I-positive cells (Fig. 8⇑c). Fourteen days after the last of multiple injections (days 0, 4, 10, and 20) of γδ T cells, the majority of PEC and SC (Fig. 8⇑c) and all LNC (not shown) expressed human MHC class I Ags. However, the percentage of human leukocytes decreased steadily with time in the xenogeneic host. Thus, relatively few human cells were recovered on day 34 from mice that had received human γδ T cells twice (days 0 and 4) or three times (days 0, 4, and 10) compared with mice that had been given a fourth injection on day 20 (Fig. 8⇑c). Apparently, the decrease in human leukocytes was accompanied by an increase in murine macrophages, as demonstrated in Table I⇓ for mice receiving four injections of γδ T cells. More than 60% of PEC and >80% of SC were of human origin, and the vast majority of HLA class I+ cells expressed CD3 and γδ TCR, as also revealed by double-fluorescence staining with anti-CD3 and a pan anti-γδ mAb (data not shown), indicating that there was no obvious change in the phenotype of the injected human lymphocytes. It should also be noted that even on days 34 after tumor cell inoculation, murine NK cells had not recovered and thus could be excluded from contributing to tumor growth retardation. However, there was a slight increase in the number of peritoneal murine macrophages in mice receiving γδ T cells compared with mice receiving rIL-2 only. To explore the functional activity of γδ T cells in the tumor-bearing mouse, we evaluated their ex vivo proliferative capacity and cytokine expression pattern (Fig. 9⇓) PEC and SC preparations from mice that had received γδ T cells were depleted of adherent cells. The remaining cell population was stained to >90% by an HLA class I-specific mAb and could thus be considered as predominantly human leukocytes. All mice were killed on day 34 after tumor cell inoculation, i.e., γδ T cells from mice receiving two or three injections had survived for 30 and 24 days in the SCID mice. These T cells still proliferated in the presence of rIL-2 and alendronate, but at a much decreased rate compared with T cells passaged for 14 days in the SCID mouse (Fig. 9⇓a). Notably, the latter mice, which had received four injections of γδ T cells, did not show any visible tumor growth at autopsy. Furthermore, γδ T cells recovered from tumor-bearing SCID mice that had been cultured ex vivo in the presence of irradiated tumor cells displayed strongly reduced proliferative activity (data not shown). Persisting functional activity of adherent cell-depleted PEC and SC populations was also confirmed at the level of cytokine production. As revealed by flow cytometry of cells that had been maintained in culture for 20 h in the presence of rIL-2 and alendronate, a considerable fraction of PEC (Fig. 9⇓b) and SC (Fig. 9⇓c) expressed human IFN-γ or TNF-α. In line with the decrease in proliferative activity, the fraction of cytokine-expressing human leukocytes decreased strongly during prolonged persistence in the xenogeneic host (Fig. 9⇓, b and c).

FIGURE 9.
FIGURE 9.

Ex vivo proliferative capacity and cytokine production of PEC and SC. PEC and SC (after two periods of 1-h plastic adherence), collected 34 days after tumor cell application, were cultured for 3 days in the presence of 20 U/ml rIL-2 and 5 μM alendronate, with [3H]thymidine (10 μCi/ml) added during the last 16 h of culture. Mean cpm ± SD of triplicate cultures (2 × 104 leukocytes) are shown in a. Nonadherent PEC (b) and SC (c) derived from MeWo- or PancTu1-inoculated mice were cultured for 24 h in the presence of 20 U/ml rIL-2 and 5 μM alendronate. Cells were fixed and permeabilized, and the percentages of human IFN-γ- and TNF-α-expressing cells were evaluated by flow cytometry.

View this table:
Table I.

Phenotype of lymphoid cells from SCID mice inoculated with human tumor cells and γδ T cells

Taken together, human γδ T cells survived and expanded in the SCID mouse, particularly when applied repeatedly. However, their functional activity vanished with time, which might be due to tumor-induced immunosuppression.

Discussion

Both experimental and clinical studies suggest that γδ T cells play a significant role in the control of epithelial homoeostasis and tumor cell development (23, 27, 53, 54). Activated human γδ T cells exert broad cytotoxic activity toward many different tumor cell types (see Ref.54 for a review). Therefore, it is tempting to explore the potential usage of γδ T cells for the immunotherapy of human cancers. Such strategies are aided by the fact that there are ligands available that selectively activate subsets of human γδ T cells. Thus, all Vγ9Vδ2 T cells (which represent the dominant γδ T cell subset and can account for up to 5% of all T cells in the peripheral blood) recognize natural microbial phosphoantigens and synthetic analogues such as BrHPP (9, 10, 11, 12, 34). Moreover, the very same γδ T cells are also activated by aminobisphosphonates, licensed drugs that are in clinical use for the treatment of osteoporosis and bone metastasis (35, 36, 37, 38). These drugs (or synthetic phosphoantigens) in combination with IL-2 might thus be used to boost human γδ T cell activity in vivo, a strategy that is currently under investigation (39, 40). Unfortunately, murine γδ T cells do not recognize phosphoantigens. Therefore, conventional mouse models cannot be used to explore the possible antitumor activity of phosphoantigen-activated γδ T cells in vivo.

In the present study we have characterized the tumor reactivity of alendronate-activated human γδ T cells in vitro and in an adoptive transfer model in vivo. Most Vγ9Vδ2 γδ T cell lines established from healthy blood donors efficiently killed the melanoma line MeWo and the pancreatic adenocarcinoma Colo357, whereas the susceptibility of PancTu1 was more donor dependent. The γδ T cell lines used in our study varied in their expression of CD56, but were CD16 (FcγRIII)-negative (Fig. 3⇑a). Interestingly, a recent report suggested that Vγ9Vδ2 T cells can be subdivided into functionally distinct subgroups on the basis of the CD16 expression. Cytotoxic activity and perforin expression were confined to the CD16-positive subpopulation of CD27−CD62L− Vδ2 γδ T cells (55). In contrast, Lafont et al. (56) have shown that freshly isolated Vγ9Vδ2 γδ T cells are CD16-negative, but acquire CD16 expression upon TCR-mediated activation. In their study, CD16 expression reached a plateau after ∼3 wk of stimulation, which was subsequently down-regulated to an undetectable level 4–6 wk after stimulation (56). Although our results clearly indicate that at least some CD16-negative Vγ9Vδ2 T cells can also exert potent cytotoxic activity, it is worth considering establishing alendronate (or phosphoantigen)-derived γδ T cell lines for adoptive immunotherapy from γδ T cells that have acquired CD16 during the early phases of in vitro culture in the future. In the present experiments, lysis of tumor cells by γδ T cell lines was certainly not dependent on CD16 (57), because our effector cells were CD16-negative. Although the underlying effector/target cell interaction is not clear in the case of Colo357 and PancTu1 tumor cells (see below), TCR-dependent target cell recognition was involved in the recognition of MeWo cells, because lysis was drastically inhibited by anti-TCR mAb. Presently, we have not yet characterized the MeWo Ag(s) recognized by the human Vγ9Vδ2 T cells. Gober et al. (58) have recently shown that some tumor cells produce phosphorylated mevalonate metabolites that stimulate Vγ9Vδ2 T cells. Interestingly, aminobisphosphonates are well-known inhibitors of farnesyl pyrophosphate synthase (59), thereby leading to an intracellular accumulation of γδ-stimulating phosphorylated mevalonate intermediates. Therefore, it is possible that human γδ T cells are stimulated by aminobisphosphonates via two mechanisms: 1) the direct presentation through APC (37), and 2) the intracellular accumulation of mevalonate metabolites due to inhibition of farnesyl pyrophosphate synthase (58, 59). In contrast to MeWo, spontaneous killing of Colo357 and PancTu1 by several γδ T cell lines was not inhibited by anti-TCR mAb, suggesting TCR-independent triggering of the cytotoxic effector function in these instances. Both tumor cells expressed MICA, and the γδ T cell lines expressed NKG2D and NKG2A, activating and inhibitory receptors for MICA and HLA-E, respectively (16, 60). Therefore, it is possible that the spontaneous cytotoxicity of γδ T cells toward Colo357 and PancTu1 tumor targets resulted from MICA-NKG2D receptor interactions. In line with published data (49, 50), the cytotoxic activity of γδ T cell lines toward all three analyzed tumor targets was strongly enhanced by preactivation with BrHPP and to some extent by alendronate. Interestingly, the BrHPP-augmented lysis of Panc Tu1 and Colo 357 was inhibited by anti-TCR mAb, suggesting that lysis of these two tumors by γδ T cell lines involved TCR-independent spontaneous as well as phosphoantigen-triggered TCR-dependent tumor cell recognition.

We also used the pan-caspase inhibitor zVAD-fmk to analyze the possible contribution of caspases to the execution of γδ T cell-mediated tumor cell lysis. In this study we reproducibly observed close to complete inhibition of lysis of Colo357, but not of either MeWo nor PancTu1. Among the three tumor cells analyzed in this study, Colo357 was the only one susceptible to anti-Fas/CD95-mediated apoptosis, whereas MeWo was Fas/CD95-negative and PancTu1 proved Fas/CD95-resistant despite strong cell surface expression (48). The inhibitory effect of zVAD in the case of Colo357 target cells is thus in line with a role of Fas/Fas ligand-dependent target cell lysis triggered by γδ T cells (61). It is most likely that the lysis of Fas/CD95-negative MeWo and of Fas/CD95-resistant PancTu1 cells involved other cytolytic effector machinery, such as the perforin/granzyme or the granulysin pathways (62, 63). Together, the results of our in vitro studies with three different tumor target cell and five different γδ effector T cells revealed substantial heterogeneity with respect to the involvement of TCR-dependent target recognition and the impact of caspase inhibition in target cell destruction.

Previous investigators have demonstrated the antitumor efficacy of human γδ T cells in vivo after adoptive transfer into SCID mice inoculated with human tumor cells. In this model system, antitumor activity of human γδ T cells has been demonstrated against Daudi lymphoma, nasopharyngeal carcinoma, and melanoma (31, 32, 33). The two major subsets of human γδ T cells, i.e., Vδ2 and Vδ1, differ in various respects, such as tissue localization and spectrum of ligands recognized by the respective TCR (13, 64). Therefore, these subsets should differ in their in vivo antitumor efficacy, depending on the type of tumor and perhaps additional parameters, such as the migratory properties of Vδ1 vs Vδ2 T cells (64). However, the TCR V gene expression of adoptively transferred γδ T cells has not been characterized in all previous studies. In the case of adoptive transfer of Daudi-reactive γδ T cells (33), it can be assumed that the human γδ T cells expressed Vδ2 and Vγ9, because anti-Daudi reactivity is exclusively associated with this γδ T cell population (65). No information on the TCR V gene usage of nasopharyngeal carcinoma-reactive human γδ T cells transferred into SCID mice was provided in the study by Zheng and colleagues (31). Lozupone et al. (32) compared the efficacies of Vδ1 and Vδ2 T cells (as well as NK cells) on the growth of autologous melanoma xenografts in SCID mice in relationship to the route of tumor and γδ T cell application. In their studies, the efficacies of Vδ1 and Vδ2 T cells were dependent on the route of γδ T cell transfer. Although both Vδ2 and Vδ1 γδ T cells (as well as NK cells) were effective when inoculated s.c. together with the melanoma, only Vδ1 γδ T cells exerted antitumor activity when given i.v (32). In our studies we observed potent antitumor activity of alendronate-activated Vγ9Vδ2 T cells transferred i.p. into SCID mice together with pancreatic adenocarcinoma PancTu1 or melanoma MeWo. Although we have not yet explored alternative routes of γδ T cell application (to avoid SCID mouse-related handicaps in γδ T cell targeting toward the tumor), our present results substantially extend the previous studies. Thus, we show for the first time that a Fas/CD95-resistant pancreatic adenocarcinoma is susceptible to adoptive human γδ T cell therapy in the SCID mouse model. In this as well as in the MeWo melanoma model, we observed a clear correlation between the survival time of SCID mice after inoculation of 5 × 106 tumor cells and the frequency of adoptive γδ T cell transfers. Optimal results were thus observed when γδ T cells together with alendronate and IL-2 were given repetitively five times. Following this protocol, survival was highly significantly (p < 0.005) prolonged compared with that using only the two-time application. The five-time repetitive application of γδ T cells also led to highest recovery of PEC and lymphoid cells in the spleen, of which the vast majority were human HLA class I-positive CD3+ γδ T cells. A significant fraction of these cells produced human TNF-α and/or IFN-γ, which might have contributed to the efficacy of adoptively transferred γδ T cells. Murine NK cells did not reappear in significant numbers, thereby excluding a contribution of these cells to the observed antitumor activity. From the autopsy of mice, which were killed for the ex vivo analysis of γδ T cell survival and functional activity, it became apparent that with very few exceptions tumor growth was completely prevented for the period of γδ T cell application. This finding suggests that continuation of the application of γδ T cells, which was prohibited by the development of GvH disease, might have been curative. Whether the rather rapid decline in the recovery of γδ T cells and in their functional activity, particularly in PancTu1-bearing mice, is an inherent feature of γδ T cells, is due to the xenogeneic environment, or is brought about by tumor-mediated immunosuppression remains to be explored. Particularly the cachexia and the icterus developed by PancTu1-bearing mice as well as the strong inhibition of γδ T cell proliferation in the presence of irradiated tumor cells argue at least for an unfavorable contribution of the tumor.

The level of γδ T cell reactivity toward a given tumor cell is likely to result from the integration of positive and negative signals mediated through TCR-dependent ligand recognition and additional receptor-ligand interactions via inhibitory and activating NK receptors (66). In this regard, it is of interest that all γδ T cell lines used in this study strongly expressed the activating NKG2D receptor and variable levels of the inhibitory receptor NKG2A. Taken together, our present results together with those of previous studies suggest that adoptive transfer of activated γδ T cells might be a useful approach in the treatment of human malignancies. In fact, the first tumor responses after aminobisphosphonate plus IL-2 therapy have been reported in patients with lymphoid malignancies (39). We suggest that optimal therapeutic efficacy in tumor patients might require combined therapy with aminobisphosphonates (or phosphoantigens) plus IL-2, and the adoptive transfer of in vitro expanded autologous γδ T cells. Tumor patients potentially responsive to γδ T cell-based immunotherapy might be identified on the basis of strong in vitro reactivity of their γδ T cells toward aminobisphosphonates and/or phosphoantigens.

Acknowledgments

We gratefully acknowledge the technical assistance of Kyoung-Ae Yoo-Ott, Hoa Ly, Monika Kunz, and Susanne Hummel. Alendronate was kindly provided by Merck, and BrHPP was supplied by Innate Pharma (Marseille, France).

Footnotes

  • 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 This work was supported by The Werner and Klara Kreitz Foundation and a grant from the Medical Faculty (to D.W.). This study is part of the M.D. thesis of E.P.

  • ↵2 Address correspondence and reprint requests to Dr. Dieter Kabelitz, Institute of Immunology, UK S-H Campus Kiel, Michaelisstrasse 5, D-24105 Kiel, Germany. E-mail address: kabelitz{at}immunologie.uni-kiel.de

  • ↵3 Abbreviations used in this paper: BrHPP, bromohydrin pyrophosphate; GvH, graft-vs-host; LNC, lymph node cell; PEC, peritoneal exudate cell; SC, spleen cell; SCDA, standard cell dilution assay; MICA/B, MHC class I chain-related protein A/B.

  • Received June 17, 2004.
  • Accepted September 24, 2004.

References

  1. ↵
    Van der Bruggen, P., Y. Zhang, P. Chaux, V. Stroobant, C. Panichelli, E. S. Schultz, J. Chapiro, B. J. van den Eynde, F. Brasseur, T. Boon. 2002. Tumor-specific shared antigenic peptides recognized by human T cells. Immunol. Rev. 188:51.
    OpenUrlCrossRefPubMed
  2. ↵
    Finn, O. J.. 2003. Cancer vaccines: between the idea and the reality. Nat. Rev. Immunol. 3:630.
    OpenUrlCrossRefPubMed
  3. ↵
    Dudley, M. E., J. R. Wunderlich, T. E. Shelton, J. Even, S. A. Rosenberg. 2003. Generation of tumor-infiltrating lymphocyte cultures for use in adoptive transfer therapy for melanoma patients. J. Immunother. 26:332.
    OpenUrlCrossRef
  4. ↵
    Dudley, M. E., S. A. Rosenberg. 2003. Adoptive-cell-transfer therapy for the treatment of patients with cancer. Nat. Rev. Cancer 3:666.
    OpenUrlCrossRefPubMed
  5. ↵
    Hayday, A.. 2000. γδ T cells: a right time and a right place for a conserved third way of protection. Annu. Rev. Immunol. 18:975.
    OpenUrlCrossRefPubMed
  6. ↵
    Porcelli, S., M. B. Brenner, H. Band. 1991. Biology of the human γδ T-cell receptor. Immunol. Rev. 120:137.
    OpenUrlCrossRefPubMed
  7. ↵
    Hinz, T., D. Wesch, F. Halary, S. Marx, A. Choudhary, B. Arden, O. Janssen, M. Bonneville, D. Kabelitz. 1997. Identification of the complete expressed human T-cell receptor Vγ repertoire by flow cytometry. Int. Immunol. 9:1065.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    Deusch, K., F. Luling, K. Reich, M. Classen, H. Wagner, K. Pfeffer. 1991. A major fraction of human intraepithelial lymphocytes simultaneously expresses the γδ receptor, the CD8 accessory molecule and preferentially uses the Vδ1 gene segment. Eur. J. Immunol. 21:1053.
    OpenUrlCrossRefPubMed
  9. ↵
    Constant, P., F. Davodeau, M. A. Peyrat, Y. Poquet, G. Puzo, M. Bonneville, J. J. Fournié. 1994. Stimulation of human γδ T-cells by nonpeptidic mycobacterial ligands. Science 264:267.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    Jomaa, H., J. Feurle, K. Luhs, V. Kunzmann, H. P. Tony, M. Herderich, W. Wilhelm. 1999. Vγ9/Vδ2 T cell activation induced by bacterial low molecular mass compounds depends on the 1-deoxy-d-xylulose-5-phosphate pathway of isoprenoid biosynthesis. FEMS Immunol. Med. Microbiol. 25:371.
    OpenUrlPubMed
  11. ↵
    Altincicek, B., J. Moll, N. Campos, G. Foerster, E. Beck, J. F. Hoeffler, C. Grosdemange-Billiard, M. Rodriguez-Concepcion, M. Rohmer, A. Boron, et al 2001. Cutting edge: human γδ T cells are activated by intermediates of the 2-C-methyl-d-erythritol 4-phosphate pathway of isoprenoid biosynthesis. J. Immunol. 166:3655.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Belmant, C., E. Espinosa, R. Poupot, M. Guiraud, Y. Poquet, M. Bonneville, J. J. Fournié. 1999. 3-Formyl-1-butyl pyrophosphate: a novel mycobacterial metabolite activating human γδ T cells. J. Biol. Chem. 274:32079.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Kabelitz, D., A. Glatzel, D. Wesch. 2000. Antigen recognition by human γδ T lymphocytes. Int. Arch. Allergy Immunol. 122:1.
    OpenUrlPubMed
  14. ↵
    Fisch, P., M. Malkovsky, S. Kovats, E. Sturm, E. Braakman, B. S. Klein, S. D. Voss, L. W. Morissey, R. DeMars, W. J. Welch, et al 1990. Recognition by human Vγ9/Vδ2 T cells of a GroEL homolog on Daudi Burkitt’s lymphoma cells. Science 250:1269.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Groh, V., A. Steinle, S. Bauer, T. Spies. 1998. Recognition of stress-induced MHC molecules by intestinal epithelial γδ T cells. Science 279:1737.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    Bauer, S., V. Groh, J. Wu, A. Steinle, J. H. Philipps, L. L. Lanier, T. Spies. 1999. Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science 285:727.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    Das, H., V. Groh, C. Kuijl, M. Sugita, C. T. Morita, T. Spies, J. F. Bukowski. 2001. MICA engagement by human Vγ2Vδ2 T cells enhances their antigen-dependent effector function. Immunity 15:83.
    OpenUrlCrossRefPubMed
  18. ↵
    Wu, J., V. Groh, T. Spies. 2002. T cell antigen receptor engagement and specificity in the recognition of stress-inducible MHC class I-related chains by human epithelial γδ T cells. J. Immunol. 169:1236.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Spada, F. M., E. P. Grant, P. J. Peters, M. Sugita, A. Melián, D. S. Leslie, H. K. Lee, E. van Donselar, D. A. Hanson, A. M. Krensky, et al 2000. Self recognition of CD1 by γδ T cells: implications for innate immunity. J. Exp. Med. 191:937.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    Boismenu, R., W. L. Havran. 1994. Modulation of epithelial cell growth by γδ T cells. Science 266:1253.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Workalemahu, G., M. Foerster, C. Kroegel. 2004. Expression and synthesis of fibroblast growth factor-9 in human γδ T-lymphocytes: response to isopentenyl pyrophosphate and TGF-β/IL-15. J. Leukocyte Biol. 75:657.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    Hayday, A., R. Tigelaar. 2003. Immunoregulation in the tissues by γδ T cells. Nat. Rev. Immunol. 3:233.
    OpenUrlCrossRefPubMed
  23. ↵
    Girardi, M., D. E. Oppenheim, C. R. Steele, J. M. Lewis, E. Glusac, R. Filler, P. Hobby, B. Sutton, R. E. Tigelaar, A. C. Hayday. 2001. Regulation of cutaneous malignancy by γδ T cells. Science 294:605.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    Gao, Y., W. Yang, M. Pan, E. Scully, M. Girardi, L. H. Augenlicht, J. Craft, Z. Yin. 2003. γδ T cells provide an early source of interferon γ in tumor immunity. J. Exp. Med. 198:433.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Street, S. E. A., Y. Hayakawa, Y. Zhan, A. M. Lew, D. MacGregor, A. M. Jamieson, A. Diefenbach, H. Yagita, D. I. Godfrey, M. J. Smyth. 2004. Innate immune surveillance of spontaneous B cell lymphomas by natural killer cells and γδ T cells. J. Exp. Med. 199:879.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    Choudhary, A., F. Davodeau, A. Moreau, M. A. Peyrat, M. Bonneville, F. Jotereau. 1995. Selective lysis of autologous tumor cells by recurrent γδ tumor-infiltrating lymphocytes from renal carcinoma. J. Immunol. 154:3932.
    OpenUrlAbstract
  27. ↵
    Maeurer, M. J., D. Martin, W. Walter, K. Liu, L. Zitvogel, K. Halusczak, H. Rabinowich, R. Duquesnoy, W. Storkus, M. T. Lotze. 1996. Human intestinal Vδ1+ lymphocytes recognize tumor cells of epithelial origin. J. Exp. Med. 183:1681.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Zocchi, M. R., C. Rugarli, M. Ferrarini. 1990. Selective lysis of autologous tumor by TCSδ1+ γδ tumor infiltrating lymphocytes from human lung carcinomas. Eur. J. Immunol. 20:2685.
    OpenUrlCrossRefPubMed
  29. ↵
    Ferrarini, M., S. Heltai, S. M. Pupa, S. Menard, M. R. Zocchi. 1996. Killing of laminin receptor-positive human lung cancers by tumor infiltrating lymphocytes bearing γδ+ T-cell receptors. J. Natl. Cancer Inst. 88:436.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    Dolstra, H., H. Fredrix, A. van der Meer, T. de Witte, C. Figdor, E. van de Weil-van Kemenade. 2001. TCRγδ cytotoxic T lymphocytes expressing the killer cell-inhibitory receptor p58.2 (CD158b) selectively lyse acute myeloid leukaemia cells. Bone Marrow Transplant. 27:1087.
    OpenUrlCrossRefPubMed
  31. ↵
    Zheng, B. J., K. W. Chan, S. Im, D. Chua, J. S. T. Sham, P. C. Tin, Z. M. He, M. H. Ng. 2001. Anti-tumor effects of human peripheral γδ T cells in a mouse tumor model. Int. J. Cancer 92:421.
    OpenUrlCrossRefPubMed
  32. ↵
    Lozupone, F., D. Pende, V. L. Burgio, C. Castelli, M. Spada, M. Venditti, F. Luciani, L. Lugini, C. Federici, C. Ramoni, et al 2004. Effect of human natural killer and γδ T cells on the growth of human autologous melanoma xenograft in SCID mice. Cancer Res. 64:378.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    Malkovska, V., F. K. Cigel, N. Armstrong, B. Storere, R. Hong. 1992. Antilymphoma activity of human γδ T-cells in mice with severe combined immune deficiency. Cancer Res. 52:5610.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    Espinosa, E., C. Belmant, F. Pont, B. Lucani, R. Poupot, F. Romagné, H. Brailly, M. Bonneville, J. J. Fournié. 2001. Chemical synthesis and biological activity of bromohydrin pyrophosphate, a potent stimulator of human γδ T cells. J. Biol. Chem. 276:18337.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Kunzmann, V., E. Bauer, M. Wilhelm. 1999. γ/δ T-cell stimulation by pamidronate. N. Engl. J. Med. 340:737. (Lett.).
    OpenUrlCrossRefPubMed
  36. ↵
    Kunzmann, V., E. Bauer, J. Feurle, F. Weiβinger, H. P. Tony, M. Wilhelm. 2000. Stimulation of γδ T cells by aminobisphosphonates and induction of antiplasma cell activity in multiple myeloma. Blood 96:384.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Das, H., L. Wang, A. Kamath, J. F. Bukowski. 2001. Vγ2Vδ2 T-cell receptor-mediated recognition of aminobisphosphonates. Blood 98:1616.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    Miyagawa, F., Y. Tanaka, S. Yamashita, N. Minato. 2001. Essential requirement of antigen presentation by monocyte lineage cells for the activation of primary human γδ T cells by aminobisphosphonate antigen. J. Immunol. 166:5508.
    OpenUrlAbstract/FREE Full Text
  39. ↵
    Wilhelm, M., V. Kunzmann, S. Eckstein, P. Reimer, F. Weissinger, T. Ruediger, H. P. Tony. 2003. γδ T cells for immune therapy of patients with lymphoid malignancy. Blood 102:200.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Dieli, F., N. Gebbia, F. Poccia, N. Caccamo, C. Montesano, F. Fulfaro, C. Arcara, M. R. Valerio, S. Meraviglia, C. Di Sano, et al 2003. Induction of γδ T-lymphocyte effector functions by bisphosphonate zoledronic acid in cancer patients in vivo. Blood 102:2310. (Lett.).
    OpenUrlFREE Full Text
  41. ↵
    Kabelitz, D., S. Wesselborg. 1992. Life and death of a superantigen-reactive human CD4+ T cell clone: staphylococcal enterotoxins induce death by apoptosis but simultaneously trigger a proliferative response in the presence of HLA-DR+ antigen-presenting cells. Int. Immunol. 4:1381.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    Pechhold, K., N. Craighead, D. Wesch, D. Kabelitz. 2002. Measurement of cellular proliferation. S. H. E. Kaufmann, and D. Kabelitz, eds. Methods in Microbiology, Vol. 32: Immunology of Infection 2nd Ed.77. Academic Press, London.
  43. ↵
    Janssen, O., S. Wesselborg, B. Heckl-Östreicher, A. Bender, S. Schondelmaier, K. Pechhold, G. Moldenhauer, D. Kabelitz. 1991. T-cell receptor/CD3-signalling induces death by apoptosis in human T-cell receptor γδ-positive T-cells. J. Immunol. 146:35.
    OpenUrlAbstract
  44. ↵
    Koch, N., G. Hämmerling, N. Tada, S. Kimura, U. Hämmerling. 1982. Cross-blocking studies with monoclonal antibodies against I-A molecules of haplotypes b, days and k. Eur. J. Immunol. 12:909.
    OpenUrlCrossRefPubMed
  45. ↵
    Wesch, D., A. Glatzel, D. Kabelitz. 2001. Differentiation of resting human peripheral blood γδ T cells towards Th1- or Th2-phenotype. Cell. Immunol. 212:110.
    OpenUrlCrossRefPubMed
  46. ↵
    Shpitz, B., B. J. Fernandes, J. B. Mullen, J. C. Roder, S. Gallinger. 1994. Improved engraftment of human tumours in SCID mice pretreated with radiation and anti-asialo GM1. Anticancer Res. 14:1927.
    OpenUrlPubMed
  47. ↵
    Pechhold, K., D. Wesch, S. Schondelmaier, D. Kabelitz. 1994. Primary activation of Vγ9 expressing γδ T cells by mycobacterium tuberculosis: requirement for Th1-type CD4 T cell help and inhibition by interleukin-10. J. Immunol. 152:4984.
    OpenUrlAbstract
  48. ↵
    Ungefroren, H., M. Voss, M. Jansen, C. Röder, D. Henne-Bruns, B. Kremer, H. Kalthoff. 1998. Human pancreatic adenocarcinomas express Fas and Fas ligand yet are resistant to Fas-mediated apoptosis. Cancer Res. 58:1741.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    Sicard, H., T. Al Saati, G. Delsol, J. J. Fournié. 2001. Synthetic phosphoantigens enhance human Vγ9Vδ2 T lymphocytes killing of non-Hodgkin’s lymphoma. Mol. Med. 7:711.
    OpenUrlPubMed
  50. ↵
    Kato, Y., Y. Tanaka, F. Miyagawa, S. Yamashita, N. Minato. 2001. Targeting of tumor cells for human γδ T cells by nonpeptide antigens. J. Immunol. 167:5092.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    Shipman, C. M., M. J. Rogers, J. F. Apperley, R. G. G. Russell, P. I. Croucher. 1997. Bisphosphonates induce apoptosis in human myeloma cell lines: a novel anti-tumour activity. Br. J. Haematol. 98:665.
    OpenUrlCrossRefPubMed
  52. ↵
    Riedle, S., M. Rösel, S. Seiter, M. Zöller. 1998. Effector cell targeting by bispecific antibodies in the therapy of human malignant melanoma: evaluation in the SCID mouse model. Int. J. Cancer. 75:908.
    OpenUrlCrossRefPubMed
  53. ↵
    Kobayashi, H., Y. Tanaka, J. Yagi, H. Toma, T. Uchiyama. 2001. γ/δ T cells provide innate immunity against renal cell carcinoma. Cancer Immunol. Immunother. 50:115.
    OpenUrlCrossRefPubMed
  54. ↵
    Ferrarini, M., E. Ferrero, L. Dagna, A. Poggi, M. R. Zocchi. 2002. Human γδ T cells: a nonredundant system in the immune-surveillance against cancer. Trends Immunol. 23:14.
    OpenUrlCrossRefPubMed
  55. ↵
    Angelini, D. F., G. Borsellino, M. Poupot, A. Diamantini, R. Poupot, G. Bernardi, F. Poccia, J. J. Fournié, L. Battistini. 2004. FcγRIII discriminates between two subsets of Vγ9Vδ2 effector cells with different responses and activation pathways. Blood 104:1801.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    Lafont, V., J. Liautard, J. P. Liautard, J. Favero. 2001. Production of TNF-α by human Vγ9Vδ2 T cells via engagement of FcγRIIIA, the low affinity type 3 receptor for the Fc portion of IgG, expressed upon TCR activation by nonpeptidic antigen. J. Immunol. 166:7190.
    OpenUrlAbstract/FREE Full Text
  57. ↵
    Cooper, M. A., T. A. Fehniger, M. A. Caliguri. 2001. The biology of human natural killer-cell subsets. Trends Immunol. 22:633.
    OpenUrlCrossRefPubMed
  58. ↵
    Gober, H. J., M. Kistowska, L. Angman, P. Jenö, L. Mori, G. de Libero. 2003. Human T cell receptor γδ T cells recognize endogenous mevalonate metabolites in tumor cells. J. Exp. Med. 197:163.
    OpenUrlAbstract/FREE Full Text
  59. ↵
    Bergstrom, J. D., R. G. Bostedor, P. J. Masarachia, A. A. Reszka, G. Rodan. 2000. Alendronate is a specific, nanomolar inhibitor of farnesyl diphosphate synthase. Arch. Biochem. Biophys. 373:231.
    OpenUrlCrossRefPubMed
  60. ↵
    Borrego, F., J. Kabat, T. B. Sanni, J. E. Coligan. 2002. NK cell CD94/NKG2A inhibitory receptors are internalized and recycle independently of inhibitory signaling processes. J. Immunol. 169:6102.
    OpenUrlAbstract/FREE Full Text
  61. ↵
    Zhu, H., H. O. Fearnhead, G. M. Cohen. 1995. An ICE-like protease is a common mediator of apoptosis induced by diverse stimuli in human monocytic THP. 1 cells. FEBS Lett. 374:303.
    OpenUrlCrossRefPubMed
  62. ↵
    Koizumi, H., C. C. Liu, L. M. Zheng, S. V. Joag, N. K. Bayne, J. Holoshitz, J. D. Young. 1991. Expression of perforin and serine esterases by human γδ T-cells. J. Exp. Med. 173:499.
    OpenUrlAbstract/FREE Full Text
  63. ↵
    Pardo, J., P. Perez-Galan, S. Gamen, I. Marzo, I. Monleon, A. A. Kaspar, S. A. Susin, G. Kroemer, A. M. Krensky, J. Naval, et al 2001. A role of the mitochondrial apoptosis-inducing factor in granulysin-induced apoptosis. J. Immunol. 167:1222.
    OpenUrlAbstract/FREE Full Text
  64. ↵
    Kabelitz, D., D. Wesch. 2003. Features and functions of γδ T lymphocytes: focus on chemokines and their receptors. Crit. Rev. Immunol. 25:339.
    OpenUrl
  65. ↵
    Bukowski, J. F., C. T. Morita, Y. Tanaka, B. R. Bloom, M. B. Brenner, H. Band. 1995. Vγ2Vδ2 TCR-dependent recognition of non-peptide antigens and Daudi cells analyzed by TCR gene transfer. J. Immunol. 154:998.
    OpenUrlAbstract
  66. ↵
    Fisch, P., A. Moris, H. G. Rammensee, R. Handgretinger. 2000. Inhibitory MHC class I receptors on γδ T cells in tumour immunity and autoimmunity. Trends Immunol. 21:187.
    OpenUrlCrossRef
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section