Phosphostim-Activated T Cells Kill Autologous Metastatic Renal Cell Carcinoma
http://www.100md.com
免疫学杂志 2005年第2期
Abstract
Metastatic renal cell carcinoma, inherently resistant to conventional treatments, is considered immunogenic. Indeed, partial responses are obtained after treatment with cytokines such as IL-2 or IFN-, suggesting that the immune system may control the tumor growth. In this study, we have investigated the ability of the main subset of peripheral lymphocytes, the V9V2-TCR T lymphocytes, to induce an effective cytotoxic response against autologous primary renal cell carcinoma lines. These T cells were expanded ex vivo using a V9V2 agonist, a synthetic phosphoantigen called Phosphostim. From 11 of 15 patients, the peripheral V9V2 T cells were amplified in vitro by stimulating PBMCs with IL-2 and Phosphostim molecule. These expanded V9V2 T cells express activation markers and exhibit an effector/memory phenotype. They display a selective lytic potential toward autologous primary renal tumor cells and not against renal NC. The lytic activity involves the perforin-granzyme pathway and is mainly TCR and NKG2D receptor dependent. Furthermore, an increased expression of MHC class I-related molecule A or B proteins, known ligands of NKG2D, are detected on primary renal tumor cells. Interestingly, from 2 of the 11 positive cultures in response to Phosphostim, expanded-V9V2 T cells present an expression of killer cell Ig-like receptors, suggesting their prior recruitment in vivo. Unexpectedly, on serial frozen sections from three tumors, we observe a lymphocyte infiltrate that was mainly composed of V9V2 T cells. These results outline that V9V2-TCR effectors may represent a promising approach for the treatment of metastatic renal cell carcinoma.
Introduction
The most common histological type of renal cell carcinoma (RCC)3 is conventional cell carcinoma, which represents up to 70–80% of renal tumors. The metastatic forms of these RCC (MRCC), inherently resistant to conventional treatments, are considered as immunogenic tumors based on the clinical response obtained after treatment with cytokines, IL-2 and IFN-. However, the responses are rare, mostly partial, and it is considered that MRCC patients do not yet benefit from efficient therapies (1, 2). The few spontaneous or cytokine-induced regressions in MRCC patients suggest that the immune system may control the tumor growth. Furthermore, the dense CD3+CD8+ T cell infiltrates in some renal tumors indicate an ongoing local immune response. However, ex vivo, the tumor-infiltrating lymphocytes (TILs) display low lytic capacities toward renal tumor cells (TCs) and present poor proliferation (3, 4). In addition, evidence has been provided indicating that TCs, directly or indirectly, may alter the T cell intracellular signaling, leading to an inhibition of their lytic activities (5, 6). These abnormalities may be partially reversed in vitro and in vivo by immunotherapy treatment with IL-2 or IFN- (7).
Within TILs, MHC-restricted and non-MHC-restricted effector cells have been identified that are considered to be part of the host antitumor immune response (8, 9). Among the innate immune response, although peripheral NK cells function appear normal in MRCC patients (10, 11), these cells present defects in their lytic activity at the tumor site (10). Furthermore, adoptive transfer of lymphokine-activated NK cells in MRCC patients did not clearly improve the clinical response as compared with IL-2 alone (12). As important nonconventional immune effectors, the lytic potential of T lymphocytes against different TC lines in vitro has been described (13, 14). In addition, adoptive transfer of ex vivo expanded human T cells in nude mice transplanted with human nasopharyngal tumors supports the in vivo antitumor effects of these cells (15). Furthermore, a recent pilot study with patients presenting lymphoid malignancies suggest that T lymphocytes, stimulated in vivo, induce an objective clinical response (16). A potential role of T cells to control the development of RCC has been suggested (17, 18), but thus far the lytic capacities of these effectors against primary tumor renal cells are not documented.
T lymphocytes bearing a -TCR represent a minor subset of human peripheral T cells (1 to 10%). Most human peripheral T cells display a disulfide-linked V9V2-TCR, are rare in the adult thymus, but increase with age in the blood, suggesting a positive selection in the periphery due to antigenic stimulation (19). Most of them express the CD45RO+CD95+ effector/memory phenotype (20) and lack CD4 and CD8 expression. They are potent and rapid producers of IFN- and TNF- in response to mycobacterial Ags or uncharacterized nonpeptidic ligands expressed on TC (21), indicating their role to coordinate the interplay between innate and acquired immunity. In nature, ligands that can activate T lymphocytes are produced by Gram-positive or -negative bacteria, inducing in some cases polyclonal expansions of V9V2 T cells (22, 23) with their accumulation in the lesions (24). A series of small phosphorylated nonpeptide metabolites synthesized by Mycobacterium tuberculosis able to stimulate V9V2 T cells in vitro have been identified, from which the most potent is the isopentenyl pyrophosphate compound. All of these molecules share a common structural core, the 3-formyl-1-butyl pyrophosphate, described as a strong inductor of in vitro V2 T lymphocytes amplification. A synthetic analog of isopentenyl pyrophosphate, called bromohalohydrin pyrophosphate (BrHPP) molecule, has been developed that mimics the biological properties of natural phosphoantigens and is easily synthesized and active at nanomolar concentrations (25). In this study, BrHPP molecule, now known as Phosphostim molecule, was used to expand V9V2 T cells from PBMCs obtained from MRCC patients. Recently, another group of compounds, the aminobisphosphonates such as pamidronate or risedronate, has been demonstrated to expand and activate the primary human T cells in vitro and in vivo (26)
The peripheral T cells recognize unprocessed nonpeptidic Ags with conserved molecular patterns such as alkylamines and organophosphates (27), by a non-MHC-restricted mechanism, but requiring cell-cell contact (28, 29). Direct binding of -TCR to phosphoantigen has not been demonstrated, suggesting the existence of an unknown presenting molecule (29, 30). The V9V2 T lymphocytes, amplified and stimulated by bacterial phosphorylated metabolites, recognize and kill TCs in vitro (31, 32, 33), suggesting a cross-reactivity between microbial and tumor-associated Ags (25). Furthermore, it was recently shown that the -TCR recognized, on tumor targets, phosphorylated nonpeptidic Ags synthesized by the mevalonate pathway, resembling those produced by microbial cells. This underlines that the accumulation of mevalonate metabolites in TCs may represent a powerful danger signal that activates the immune response and constitute a novel target of tumor immunotherapy (34).
The aim of the present study was to investigate the ability to expand in vitro peripheral V9V2 T lymphocytes from MRCC patients and to evaluate their capacity to lyse autologous primary tumor renal cells. Here, we show that the V9V2 T lymphocytes are expanded from most MRCC patients PBMCs and display a selective lysis against autologous primary renal TCs and not against renal normal cells (NC). Moreover, V9V2 T cells are detected in situ as an important component of the tumor-infiltrating T cells. These observations outline the role of T population in the natural immune response in RCC.
Materials and Methods
Cell lines and cultures
Tumor and adjacent normal renal parenchyma samples were obtained from 15 patients with conventional RCC who had undergone radical nephrectomy. None of the patients had received preoperative therapy. The autologous primary TC lines were derived from the tumor fragments by enzymatic digestion using collagenase (300 U/ml), DNase I (500 U/ml), and hyaluronidase (3000 U/ml) (Sigma-Aldrich). The same procedure was applied to a normal renal fragment, obtained at distance from the tumor, to derive short term normal renal cells. These cells were cultured in DMEM supplemented with 10% FCS and 1% ultroserG (Life Technologies) and subcultured 1/5 when confluent.
Blood samples from MRCC patients (50 ml) were collected 1–2 months after the nephrectomy. PBMCs were isolated by centrifugation on Ficoll-Hypaque density gradient (Amersham Biosciences). PBMCs from healthy donors (n = 7) were treated in parallel with RCC sample as standard control of amplification. Ten million PBMCs were cultured at 2 x 106/ml in 24-well plates in RPMI 1640 supplemented with 10% v/v fetal clone irradiated (HyClone). Polyclonal V9V2 T cell lines were specifically expanded in the presence of 3 μM Phosphostim (BrHPP molecule; Innate Pharma) and 100 IU/ml IL-2 (Aventis Pharma) during 15 days. Phosphostim was added once at the onset of the culture and then every 3 days; one-half of the culture medium volume was replaced by fresh medium containing only 100 IU/ml IL-2.
Immunochemistry
The 15 renal tumors were staged according to primary tumor-regional lymph nodes-distant metastasis clinical classification and examined for tumor stage and histological grade (I–IV according to Fuhrman classification) as depicted in Table I. Immunostaining was performed for six tumors on 5-μm frozen tissue sections fixed in acetone, using the Ventana ES automated immunohistochemistry system (Ventana Medical Systems). The procedure is based on an indirect biotin-avidin system with an universal biotinylated Ig secondary Ab, diaminobenzidine substrate, and hematoxylin counterstain. Primary mAbs included: anti-CD3 (Ventana Medical Systems); anti-CD8 (Dakocytomation; dilution 1/50); anti-pan--TCR (Immu 510, IgG1, dilution 1/20; Immunotech), anti-V1 (R9.12, IgG1, dilution 1/20; Immunotech), anti-V2 (Immu 389, IgG1, dilution 1/20, Immunotech). A negative control slide was incubated with an irrelevant mAb (mouse IgG; Ventana Medical Systems). Positive controls consisted of small intestinal tissue sections stained with the different mAbs described above.
Table I. Clinical and hematological characteristics of patients with RCC (n = 15)a
Immunophenotypic study
Before in vitro culture, 2 x 105 PBMCs from RCC patients were suspended in PBS containing 2% human AB serum for 10 min to block nonspecific binding site and then incubated at 4°C for 30 min with a combination of FITC-, PE-, and phycoerythrin-cyanin 5 (PC5)-conjugated Abs to determine the percentages of NK cells as well as and T cells: CD3-PC5 (UCHT1, IgG1)/V2-FITC (Immu 389, IgG1)/IgG1-PE; CD8-PE (B9.11, IgG1); CD56-PE (N901, IgG1) (mAbs purchased from Immunotech). At day 14 of culture, the bulk of the expanded V9V2 T cells was analyzed by triple staining using the following Ab combination: from Immunotech, CD3-PC5/V2-FITC/CD8-PE, CD56-PE, CD16-PE (3G8, IgG1), CD69-PE (TP1.55.3, IgG2b), HLA-DR-PE (Immu 357, IgG1), CD45RO-PE (UCHL1, IgG2a), CD158a-PE (EB6, IgG1), CD158b-PE (GL183, IgG1), CD158e-PE (Z27, IgG1), NKG2A-PE (Z199, IgG2b); from R&D Systems, NKG2D-PE (clone 149810, IgG1); and from BD Pharmingen, ILT2-PE (GHI/75, IgG2b). Background levels were measured using isotypic controls purchased from Immunotech. Compensation was set up with single-stained samples, low forward scatter elements (RBC and debris) were excluded from analysis, and 10,000 events were collected and analyzed on a FACS-Sort (BD Biosciences) using Cell Quest software (BD Biosciences).
Primary normal and tumor renal cells were phenotyped by indirect single-color fluorescence. Cells (2 x 105) were suspended in PBS containing 2% FCS for 10 min and incubated at 4°C for 30 min with the following Abs: pan-HLA-I (W6/32, IgG2a); CD54 (IgG1); anti-MHC class I-related molecule (MIC) A and anti-MICB (both provided by Immunex; IgG1); anti-NKG2D-Fc (R&D Systems; human IgG1), anti-human fibroblast (AS02, IgG1); G250-FITC (provided by Dr. F. Hirsch). Cells were washed twice with PBS and then incubated for 20 min at 4°C with FITC-conjugated goat anti-mouse Ig or PE-conjugated goat anti-human Ig.
TaqMan (Applied Biosystems) real time quantitative RT-PCR
One microgram of RNA extracted from primary tumor and normal renal cell lines was reverse transcribed using random hexamers according to the manufacturer’s recommendations (Applied Biosystems). Quantitative real time PCR was performed using 5 μl of diluted cDNA (1 μl in 20 μl of water) in a final volume of 25 μl according to the manufacturer’s recommendations (Applied Biosystems). PCR primers and probes for MHC class I-related (MIC) molecules A and B, UL16-binding protein (ULBP)1, ULBP2 and ULBP3 gene targets were designed by Applied Biosystems and used as the manufacturer’s recommendations. The amount of sample RNA was normalized by amplification of an endogenous RNA control (18S). The relative quantification of the transcripts was derived using the standard curve method (Applied Biosystems User Bulletin 2, ABI PRISM 7700 Sequence Detection system) and were normalized to the 18S mRNA of each cell analyzed.
Cytotoxic assays
Expanded V9V2 T cells were tested for cytotoxicity against allogeneic established renal cell lines (RCC6), autologous primary NC and TC lines, sensitive cell line Daudi, and resistant cell line Raji in a 4-h 51Cr release assay. Target cells were labeled with 100 μCi 51Cr for 60 min. The E:T ratio ranged from 30:1 to 3.75:1. Specific lysis (expressed as percentage) was calculated using the standard formula [(experimental – spontaneous release/total – spontaneous release) x 100] and correspond to the mean of the triplicate. Lysis mediated by V9V2 T cells of normal vs tumor primary renal cells was compared using Student’s test. In some experiments, Immu 510 mAb (anti-pan--TCR IgG1; Immunotech), ON72 mAb (anti-NKG2D, IgG1; Innate Pharma), or control mouse Ig (IgG1) were added at predefined saturating concentrations at the onset of the cytolytic assay before exposure to labeled target cells. In some experiments, effector T cells were preincubated with 10 nM concanamycin A (CMA; Sigma-Aldrich) at 37°C for 20 min to block the perforin-granzyme pathway and then added to labeled target cells. In other experiments, targets wereincubated with saturating concentrations of anti-HLA-BC mAb or mouse control IgG2a for 30 min; then, effector T lymphocytes were added. Data were expressed as the percentages of specific lysis.
Cold target competition assays were performed. A standard 4-h 51Cr release assay was realized in presence of a 40-fold excess of unlabeled targets (DAUDI, RAJI cell lines, RCC6 and NC or TC) that were added to 51Cr labeled TC or NC (2 x 103 per well) before the incubation with autologous activated lymphocytes.
KIR functionality was assessed using a mAb-redirected lysis assay using P815 mastocytoma mouse cells. Briefly, 51Cr labeled P815 target cells (3 x 103) coated with anti-pan--TCR mAb (10 μg/ml) were incubated with expanded V9V2 T effectors (E:T 5:1). TCR-redirected lysis of labeled P815 was modulated by the presence of saturating concentrations of anti-KIR mAbs (5 μg/ml) or mouse control IgG1.
Results
V2 T lymphocytes infiltrate RCC tumors
Fifteen MRCC patients undergoing partial or complete nephrectomy with lymph node resection in some cases are included in this study. They were all diagnosed as conventional RCC. None of them had received any previous treatment. Consecutive sections of frozen tumor fragments were analyzed as to immunohistochemical staining with anti-CD3, anti-CD8, anti-pan--TCR, anti-V1, and V2-TCR mAbs. Before immunostaining, the extend of the inflammatory cells infiltrate was evaluated by H&E- safranin staining. Three of seven tumors analyzed displayed a dense CD3/CD8 T cell infiltrate and were selected for further immunostaining. Fig. 1 shows the immune infiltrate (CD3+ T cells) of tumor sections from patients 1 (Fig. 1A) and 2 (Fig. 1B). As depicted, numerous T cells are detected in situ at the contact with TCs and interestingly are mainly composed of the V 2 subset. The frozen tumor sections from four other patients displayed a moderate immune infiltrate, but in these tumors also the V2 T lymphocytes represent the prevalent tumor infiltrating T cell population (data not shown). Normal renal parenchyma was devoid of any infiltrating T cells.
FIGURE 1. V2 TCR-bearing T cells are the major component of the T cell infiltrate of renal tumors. Immunohistochemical analysis of serial frozen tumor tissue sections from patient 1 (A) and patient 2 (B). Tissue sections were stained using mAb specific for CD3, CD8, pan--TCR, V1 TCR, and V2 TCR. Positive cells were stained in brown. Two of six tumors analyzed are shown.
Phosphostim stimulation induced the expansion of peripheral V2 T cells from PBMCs of MRCC patients
The clinical and hematological characteristics of the patients included in this study are summarized in Table I. Of 15 RCC patients, 2 were lymphopenic (<103 lymphocytes/ml). Before in vitro stimulation, the percentages of NK cells and CD8 T lymphocytes from MRCC patients were in the range of normal donors (8.5–30.1% for NK cells, 13–60.8% for CD8 T cells). In MRCC patients as well as in donors, the percentage of T lymphocytes represented <5% of total PBMCs (range, 0.3–4.1%). PBMCs from these 15 patients and 7 donors, used as standard control of expansion, were cultured with Phosphostim and IL-2 for 2 wk. After 14 days, 11 of 15 cultures derived from patients contained a percentage of V9V2 T lymphocytes ranging between 73.3 and 96.4% and presented a high rate of expansion (>200-fold). The corresponding patients were designated responders to the Phosphostim expansion test (patients 3, 5–9, and 11–15). For the other cultures, poor expansion rate (<30-fold) and low percentages (<20%) of V2 T lymphocytes were obtained; these patients were designated nonresponders to the expansion test and were not included in functional tests due to low cell numbers (patients 1, 2, 4, and 10). In control cultures derived from 7 donors’ PBMCs, 80–96.4% of V2 T cells were obtained after a 2-wk amplification with Phosphostim molecule. Furthermore, the amplification rate of responder patients (range, 248- to 2787-fold expansions; Table I) were similar of those from donors (range, 141- to 1236-fold expansions; data not shown).
Phenotypic analysis of expanded-V2 T cells from MRCC patients’ PBMCs
The Phosphostim-stimulated V9V2 T cells from MRCC patients were phenotyped at day 14 of culture, using the panel of mAbs described in Materials and Methods. The phenotypic characteristics are summarized in Table II, and a representative V9V2 T cell culture is depicted in Fig. 2. Polyclonal V9V2 T cells from MRCC patients display a memory phenotype consistent with a bright expression of CD45RO (>95%; data not shown). They are in an activated state (CD69+HLA-DR+); and compared with cultures from donors, V9V2 T cells from MRCC patients exhibit a significantly increased level of CD8 expression assessed by MFI (532 ± 192 in RCC vs 297 ± 27 in donors, p < 0.09), whereas percentages of CD8 cells were not significantly increased(31.7 ±12.6% vs 24.1 ± 11.3%; p < 0.2). Expanded T cells from patients’ PBMCs, bearing the V9V2 TCR, expressed on their cell surface the activating NKG2D receptors. The percentages varied among MRCC patients (mean, 85.6%; range, 74.2–99.9%) but were similar to those on expanded V9V2 T lymphocytes from a series of donors (mean, 88.9%; range, 67.2–99.9%), except patient 7 that displayed a low NKG2D expression (27.2%).
Table II. Phenotypic analysis of expanded-V2 T lymphocytes from MRCC patients and donorsa
FIGURE 2. Phenotypic analysis of expanded V9V2 T lymphocytes from a RCC patient (patient 12). Expanded-V9V2 T cells were stained with a panel of conjugated mAbs described in Materials and Methods and analyzed by flow cytometry. The histograms show staining on the gated CD3/V2 subset. Open histograms are isotypic IgG control staining.
The control of V9V2 T cell activation involves a balance between stimulatory and inhibitory signaling delivered, at least in part, by the heterodimeric CD94-NKG2-A receptor. There was no clear difference in NKG2-A receptor expression between donors and MRCC patients (Table II). Furthermore, 2 of the 11 patients’ V9V2 T cell cultures exhibited an elevated percentage of KIR2DL/S (CD158b) protein (>25%) (Table II, patients 3 and 7). However, one of seven donors exhibited 70% of CD158b+ expanded-V9V2 T cells. A comparable level of KIR expression on V2 T cells was found on two different cultures from the same patient derived 2 and 10 mo postnephrectomy, indicating a stable expression of these receptors.
Lysis of autologous primary tumor renal cells by expanded-V2 T cells
Tumor and normal renal cells were successfully derived from surgical nephrectomy fragments for 10 responder patients except for patient 9 which presented a fibroblastic contamination of primary culture and thus was not included in functional studies. Primary tumor renal cells are characterized by homogeneous staining with anti-G250 mAb which recognize an Ag expressed exclusively by renal cancer cells. In addition, the absence of contamination of the TC and NC lines by fibroblasts was evaluated by staining with a specific mAb (AS02). Early in vitro passages were used as targets in the functional assays as well as a well-characterized renal TC line, RCC6 (35), that was included in functional studies. The Burkitt’s lymphoma cell lines Daudi and Raji, were used as sensitive and resistant targets, respectively. In a first set of experiments, expanded-V9V2 T cells from MRCC patients were assessed for their lytic potential toward the different targets. These T effectors killed the Daudi cells with a high efficiency, similar to that of donor expanded-V9V2 T cells, indicating that these cells mainly mediate a non-MHC-restricted lysis. Interestingly, the expanded- V9V2 T cells from patients efficiently killed autologous primary renal TCs and RCC6 and displayed a low lytic potential toward their autologous normal renal counterparts similar to that of Raji (Fig. 3A, patient 14). Data depicted in Fig. 3B summarize the cytotoxic activity of V9V2 cultures from 10 RCC patients, evaluated in autologous setting, against tumor and normal renal targets and show that the mean lytic activity of T effectors against primary tumor renal cells was significantly higher than that against normal renal cells (p < 0.05 for all E:T ratios). In addition, cold target competition assays with addition of 40-fold excess of various cold targets were performed to determine whether the same antigenic structure is present on the different targets. Addition of excess of cold Daudi and RCC6 cell lines caused a decrease in lysis of autologous TC and NC lines by V9V2 T cells, whereas Raji cells did not modulate the lysis. Furthermore, primary NC line competed with 51Cr-labeled TC lines (Fig. 3C). These data indicate that T cells recognize a common structure on Daudi and renal TCs and NCs.
FIGURE 3. Expanded-V9V2 T cells from RCC patients recognized and killed primary TCs. A, Cytotoxic activity (4-h 51Cr release assay) of expanded- V9V2 T lymphocytes against autologous primary TC lines (patient 14) was measured for the indicated E:T ratio. The polyclonal V2 T lymphocytes were tested against different targets: Daudi and Raji cell lines as susceptible and resistant targets, respectively; RCC6 renal TC line; and the autologous primary NC and TC lines. Data are presented as the mean percentage of specific target lysis. B, Mean cytolytic activity of expanded- V9V2 T lymphocytes from 10 responder patients’ PBMCs against the pair of primary renal cell lines (TC and NC) as well as Daudi and Raji cell lines. The percentage of lysis for each target cells, at different E:T ratios, represents the mean value of cytotoxicity obtained for 10 patients. p values between primary NCs and TCs, determined by a standard Student test, are indicated. C, Cold target competition assay was performed with expanded-V9V2 T cells from patient 14. The ability of these effector cells to lyse autologous primary TC (C, left) and NC (C, right) were investigated in presence of a 40-fold excess of different cold targets. Data are presented as the mean percentage of specific target lysis. md, Medium.
TCR and NKG2D receptors are involved in RCC target cell lysis
The main cytolytic pathway used by killer cells involves the release of the cytolytic granules in effector-target interface. To determine the involvement of this pathway in V9V2 T cell-mediated lysis, effector cells were preincubated with CMA, which inhibits the perforin-based cytotoxicity by increasing the pH of lytic granules. The lytic potential of expanded-V9V2 T cells was completely abrogated in the presence of CMA during the chromium release assay, indicating that the cytotoxic activity of these effectors involves mainly the perforin-granzyme pathway (Fig. 4A).
FIGURE 4. Mechanisms underlying the cytotoxic activity of expanded- V9V2 T cells. A, The lytic activity of expanded-V9V2 T lymphocytes involves the perforin-granzyme pathway. Expanded-V9V2 T cells (patient 3) were incubated with CMA during the cytotoxicity assay against the pair of primary autologous renal cells and Daudi cell lines. Results are the lysis percentages of targets determined at the indicated E:T ratio. B, The V9V2 T lymphocyte cytolytic activity implicated the TCR and NKG2D receptors. Expanded-V9V2 T cells from patient PBMCs (patient 3) were incubated with anti-pan--TCR mAb (clone IMMU 510) and/or anti-NKGD mAb (clone ON72) before their addition to autologous primary TCs or to RCC6 cells (C). md, Medium; cIg, control Ig.
The majority of amplified V9V2 T cells from MRCC patients express the activating C-type lectin receptor NKG2D (see Table II), that provides a costimulatory signal, amplifying the TCR signal for the lysis of TCs (36). To delineate the respective involvement of NKG2D and TCR signaling in V9V2-mediated lysis, we performed cytolysis assays in presence of saturating concentrations of blocking mAbs. Autologous and allogeneic (RCC6) renal TCs were used as targets. As depicted in Fig. 4B, lysis of autologous TCs by expanded-V9V2 T lymphocytes (patient 3) is clearly decreased in presence of anti-pan--TCR mAb (Immu 510), indicating that this receptor is involved in the lysis of TCs (63% inhibition at E:T 15:1, n = 3). Blocking of the NKG2D receptor (mAb ON72) also decreased the lysis of targets but to a lesser extent (38% of inhibition at E:T 15:1). Furthermore, addition of the two mAbs during the chromium assay had an additional effect (75% inhibition at E:T 15:1). As shown in Fig. 4C, addition of anti-TCR or anti-NKG2D mAbs inhibited the cytotoxic activity of expanded-V9V2 T effectors against the allogeneic renal cell line RCC6. Similar results were obtained from two other V9V2 expanded cell lines from patients, clearly demonstrating the involvement of TCR in the cytotoxicity mediated by these effectors (patients 6 and 7, not shown) and confirming the costimulating role of NKG2D receptor in T cell lysis.
Differential expression of NKG2D ligands on primary normal and tumor renal cells
NKG2D ligands such as MICA, MICB, and ULBP proteins may be induced by cellular stress, and most of them are frequently expressed on TCs (37, 38). This indicates that NKG2D provides a first line surveillance against stressed or transformed cells. A flow cytometry analysis of primary TC and NC lines was performed. As shown in Fig. 5B, primary TC lines exhibit an increase of MICA and MICB protein expression compared with normal counterparts (patient 14). Furthermore, the fusion protein NKG2D Fc, which recognize with a low affinity all ligands of the NKG2D receptor, bind to renal TCs with a high efficiency. In addition, a quantitative analysis by (real time) RT-PCR for the ULBP1,2,3 as well as MICA and MICB genes was performed on the pairs of primary renal cells derived from eight patients (patients 3, 5, 7, 8, and 12–14) indicating that the primary TC lines display increased of ULBP1,2,3 transcripts as compared with primary NC lines. No significant difference between both primary cell lines was observed for the MICA and MICB transcripts (Fig. 5A). Phenotypic analysis of short term primary cells lines indicate that they were devoid of contamination by fibroblasts and that TCs express the G250 tumor-specific Ag. Furthermore, the expression of HLA-I molecules, which can trigger an inhibitory receptors such as CD94-NKG2-A, is similar in both normal and tumor renal cells. In addition, a high and comparable expression of adhesion molecules, ICAM-1, is observed on both normal and tumor renal cells (Fig. 5C).
FIGURE 5. Analysis of NKG2D ligands on primary renal cell lines. A, Quantitative analysis of MICA, MICB, ULBP1, ULBP2, and ULBP3 mRNA transcripts in primary renal TCs and NCs from eight MRCC patients. Results are expressed in quantity of mRNA of the corresponding transcripts normalized to the 18S mRNA of each cell line analyzed. p values between primary NCs and TCs, determined by a standard Student t test, are indicated. B, Phenotype of paired primary TC and NC lines (patient 14). Primary renal cell lines were stained by indirect fluorescence with the panel of mAbs described in Materials and Methods and analyzed by flow cytometry. Open profiles are isotypic IgG control staining. C, Phenotypic analysis of primary tumor and normal renal cells (patient 7). Epithelial cell lines were assessed for tumor-specific G250 Ag, fibroblast contamination (AS02), HLA-I, and ICAM-1 molecules. md, Medium; cIg, control Ig.
KIRs expressed by amplified V2 T cells from MRCC patients are functional
It has been previously shown that peripheral T cells from donors expressed the C-lectin type inhibitory NK receptor CD94-NKG2-A whereas Ig-type receptors (KIRs) are present on discrete subsets (39, 40). In this study, we show that the expression of KIR receptors on expanded V2 T cells was increased in 2 of 10 RCC patients (Table II) and in 1 of 7 donors (Donor 3). To analyze the influence of this inhibitory receptor on lytic activity of V9V2 T cells from patient 3 (CD158b >25%), mAb-redirected lysis experiments with FcR+ murine P815 cells were performed. As shown in Fig. 6A, engagement of TCR on the bulk of V9V2 T cells by specific mAb-induced killing of P815 targets. This lysis is significantly decreased when KIR2DL/S is triggered by anti-CD158b mAb (inhibition, >30%; n = 3), whereas anti-CD158a mAb had no effect. Triggering of KIR also inhibited TCR-induced lysis of KIR+ V9V2 T cells derived from donor 3 (data not shown). Furthermore, the lysis of autologous primary TCs by the bulk of V9V2 T cells was slightly increased in presence of anti-HLA-B/C mAb to block KIR/HLA-C interactions (Fig. 6B), indicating that engagement of KIR2DL/S molecule by specific ligands expressed on target cells may partly protect TCs from lysis by the V9V2 T effectors.
FIGURE 6. KIR receptors exert an inhibitory effect on lytic activity of V9V2 T cells. A, TCR redirected lysis assays of FcR+ P815 target cells using expanded-V9V2 T cells from patient 3, by triggering of -TCR (Immu 510 mAb). Inhibitory effect of KIR2DL/S molecule was tested using an anti-CD158b mAb. The data were obtained at E:T 5:1. B, Modulation of the cytolysis of autologous primary TCs by blocking the HLA-KIR interaction. Targets were preincubated with anti-HLA-B/C mAb or control isotype (cIg) used at 5 μg/ml before addition of the V9V2 T cells. One representative experiment of three is shown.
Discussion
The implication of T lymphocytes in the immune defense against bacterial, microbial, or parasitic infection has been well documented. Once activated, these cells induce immediate effector functions such as cytokine production or cytotoxic activity (30) and are able to develop a memory response against microbial reinfection (41). In addition, recent studies outlined the role for human T cells in recognition and lysis of transformed cells of different origins, mainly of hematopoietic lineage. In solid tumors, such as RCC, an immune T cell response has been described, mainly implicating T lymphocytes, whereas the role of T cells remains poorly documented.
This study gives strong evidence in support of a role of peripheral ex vivo expanded and activated V9V2 T cells in the immune surveillance against renal tumors. The V9V2 T lymphocyte population represents the majority of circulating lymphocytes and recognizes natural and synthetic phosphoantigens. These molecules induce an expansion/activation of this subpopulation cells in primary PBMC cultures of healthy donors. Using a synthetic phosphoantigen, Phosphostim, our results demonstrate that 11 of 15 cultures of patients’ PBMCs respond to in vitro stimulation, resulting in a strong amplification of V9V2 T cells. The patients included in the study had advanced renal tumors (grade III–IV), and most of them had metastasis. Besides lymphopenia (<1000), neither percentages of T cells at day 0 of culture nor percentages of CD3+CD8+ are truly indicative of the responder status. Increasing cell concentration (x2 or 3) at the onset of the culture is not effective to reverse the absence of amplification in nonresponder patients. Given that it is known that immunological responses may be suppressed after surgery and that this effect disappears several months after, the patients who are nonresponders to the expansion test will be followed to determine whether the immunosuppressive status observed in our cultures can be reversed with time after the nephrectomy.
In healthy adults, the circulating V9V2 T lymphocytes are mainly composed of CD45RO+ and CD95+ memory cells. This phenotype is similar to that found for V9V2 T cells from MRCC patients. It was reported that activation of T cells induced an increased expression of CD27 marker which gradually switched off (42). For some patients’ V9V2 cultures, we have detected a low expression of CD27 marker (data not shown) which correlates with the up-regulation of CD56 and CD8 molecules as well as early (CD69) and late (HLA-DR) markers of activation fitting with their effector T cells status. Previous studies reported that in vitro activation in the presence of IL-2 down-regulates CD27 expression, confirming that primed T cells become effector cells. Because our data implicate T cells in the immune response against RCC, it would be interesting to compare the V9V2 T cell populations from healthy and patients before Phosphostim expansion for CD27, CD28, CD45RA/RO, and CCR7 markers to precisely define the distribution of effector/memory T cell subsets (43).
The major aim of our study was to investigate the lytic capacity of the polyclonal expanded V9V2 T cells toward primary renal TCs. From the 10 T cell cultures evaluated in autologous setting, we showed that the amplified V9V2 T lymphocytes present a higher lytic potential toward primary tumor renal cells in comparison with primary normal renal cells. The rare contaminant NK or T cells in the cultures as well as the inhibition of lysis in the presence of TCR- mAb imply that lysis of renal TCs is mediated by T cells. Cold targets competition experiments indicate that a similar structure is recognized by T lymphocytes on normal and tumor renal cells, but additional engagement of other triggering receptors may explain increased lysis toward TCs. Furthermore, our data indicated that besides TCR, NKG2D receptors deliver mainly a costimulating effect involved in the lysis of tumor renal cells. Concerning the expression of the corresponding ligands, MIC molecules were overexpressed on primary TC lines, although no difference was observed for the MICA or MICB transcripts levels. Increased transcripts for ULBP1, 2, and 3 was found in primary renal TCs, and binding of NKG2D-Fc protein was higher on renal TCs. Altogether, these results favor the involvement of the NKG2D receptor in the preferential lysis of renal TCs by V9V2 T cell-mediated killing. In contrast, primary normal and tumor renal cells exhibit similar expression of HLA-I molecules, excluding a differential engagement of inhibitory NK receptors by tumor vs normal renal cells, and indicate that V9V2 T lymphocytes recognize and engage TCR ligands likely preferentially expressed on TC.
The response mediated by peripheral blood V9V2 T cells against the invading pathogen is rapid and potent and requires tight regulation to prevent autoimmunity or immunopathological manifestations. One of the regulatory mechanisms of cytotoxic effectors implicate the inhibitory MHC class I receptors expressed as well by NK cells and some T lymphocyte clones. Circulating T cells from donors express mainly the CD94-NKG2-A complex (80%), which recognizes HLA-E, a nonclassical MHC class I molecule (44). The inhibitory receptor expression on V9V2 T cell clones from donors has been also documented (39), implying that the biological role of the major subset of human T lymphocytes is closely related to that of NK cells. In the present study, 2 of 11 cultures derived from patients’ MRCC PBMCs and 1 of 7 donors present an expression of the KIR2DL/S molecule. This KIR expression was confirmed on PBMCs before culture with Phosphostim, indicating that, as for -TCR T cells, KIRs were induced in vivo (data not shown). This is to our knowledge the first report showing an enhanced KIR expression on peripheralV 9V2 T cells from MRCC patients amplified ex vivo with Phosphostim. Our functional results suggest that this receptor is inhibitory, decreasing the TCR-induced lysis, and may control the lysis of autologous tumor renal cells, although these results should be extended and confirmed on immunoselected KIR+ V9V2 T cells. Altogether, the expression of KIR and the memory/effector phenotype of expanded-V9V2 T cells from MRCC patients reflect Ag-experienced T cells that are likely to respond to a chronic stimulation by a tumor.
At present, it remains unclear whether subsets of human peripheral T cells have the ability to home to damaged tissues such as tumors. The tumor immunosurveillance in vivo has mainly implicated the V1 intraepithelial population (8, 45) but little information about the antitumor potential of peripheral V9V2 T cells has been reported. However, phenotypic analysis of infiltrating T cells in melanoma revealed that they bear the V9V2 TCR and display in vitro a potent non-MHC-restricted lytic potential toward autologous TCs (46, 47). In contrast, a recent study using SCID mice transplanted with human autologous melanoma cells demonstrated that NK cells, V1 and V2 T lymphocytes are able to migrate in tumor site but that only the first two populations are able to prevent and inhibit the tumor growth. Surprisingly, the V2 lymphocyte had no effect even though this population has a higher in vitro cytotoxic potential than do V1 T cells (48). In our study, immunostaining of renal tumor sections showed the presence of the V2 TCR T lymphocytes at close contact with renal TCs. Although the density of the infiltrate may vary from one tumor to another, T cells are well represented in situ, and the major subset of this population bear the V2 TCR, suggesting that this peripheral population has the capacity to migrate to extra lymphoid sites and to home to the tumor.
The high percentages of patient responders to Phosphostim stimulation, the activation state of the expanded-V9V2 effectors and their ability to kill the autologous primary renal TC lines are very encouraging results to develop the use of such compound in vivo. It has been recently reported that treatment of patients presenting multiple myeloma using aminobisphosphonates has induced objective remission or inhibition of disease progression. Interestingly, the in vivo responses correlated with a high expansion rate of cytotoxic T lymphocytes in in vitro cultures of thecorresponding patients (16). These in vivo data, together with our present results, favor the clinical trials using Phosphostim and Il-2 for a new therapeutic approach of MRCC.
Acknowledgments
We thank Stéphane Leborgne for managing of the blood samples from RCC patients and Catherine Gaudin for help in real time quantitative reverse transcription-PCR experiments.
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 Institut National de la Santé et de la Recherche Médicale Grant 4613 (l’Association pour la Recherche sur le Cancer) (to A.C.). E.V. is the recipient of a Conventions Industrielles de Formation par la Recherche Européennes grant from Innate Pharma.
2 Address correspondence and reprint requests to Dr. Anne Caignard, Institut National de la Santé et de la Recherche Médicale Unité 487, Cytokines et Immunologies des Tumeurs Humaines, Institut Gustave Roussy, PR1, 39 rue Camille Desmoulins, F-94805 Villejuif, France. E-mail address: CAIGNARD@IGR.FR
3 Abbreviations used in this paper: RCC, renal cell carcinoma; MRCC, metastatic form of renal cell carcinoma; TIL, tumor-infiltrating lymphocyte; BrHPP, bromohalohydrin pyrophosphate; PC5, phycoerythrin-cyanin 5; TC, tumor cells; NC, normal cells; CMA, concanamycin A; ULBP, UL16-binding protein; MIC, MHC class I-related molecule.
Received for publication July 14, 2004. Accepted for publication October 24, 2004.
References
Escudier, B., C. Chevreau, C. Lasset, J. Douillard, A. Ravaud, M. Fabbro, A. Caty, J. Rossi, P. Viens, J. Bergerat, J. Savary, S. Negrier. 1999. Cytokines in metastatic renal cell carcinoma: is it useful to switch to interleukin-2 or interferon after failure of a first treatment? Groupe Francais d’Immunotherapie. J. Clin. Oncol. 17:2039.
Negrier, S., B. Escudier, F. Gomez, J. Douillard, A. Ravaud, C. Chevreau, M. Buclon, D. Perol, C. Lasset. 2002. Prognostic factors of survival and rapid progression in 782 patients with metastatic renal carcinomas treated by cytokines: a report from the Groupe Francais d’Immunotherapie. Ann. Oncol. 13:1460.
Finke, J., A. Zea, J. Stanley, D. Longo, H. Mizoguchi, R. Tubbs, R. Wiltrout, J. O’Shea, S. Kudoh, E. Klein. 1993. Loss of T-cell receptor chain and p56lck in T-cells infiltrating human renal cell carcinoma. Cancer Res. 53:5613.
Rayman, P., R. Uzzo, V. Kolenko, T. Bloom, M. Cathcart, L. Molto, A. Novick, R. Bukowski, T. Hamilton, J. Finke. 2000. Tumor-induced dysfunction in interleukin-2 production and interleukin-2 receptor signaling: a mechanism of immune escape. Cancer J. Sci. Am. 6:(Suppl. 1):S81.
Uzzo, R., P. Clark, P. Rayman, T. Bloom, L. Rybicki, A. Novick, R. Bukowski, J. Finke. 1999. Alterations in NFB activation in T lymphocytes of patients with renal cell carcinoma. J. Natl. Cancer Inst. 91:748.
Molto, L., P. Rayman, E. Paszkiewicz-Kozik, M. Thornton, L. Reese, J. Thomas, T. Das, D. Kudo, R. Bukowski, J. Finke, C. Tannenbaum. 2003. The Bcl-2 transgene protects T cells from renal cell carcinoma-mediated apoptosis. Clin. Cancer Res. 11:4060.
Bukowski, R., P. Rayman, R. Uzzo, T. Bloom, K. Sandstrom, D. Peereboom, T. Olencki, G. Budd, D. McLain, P. Elson, A. Novick, J. H. Finke. 1998. Signal transduction abnormalities in T lymphocytes from patients with advanced renal carcinoma: clinical relevance and effects of cytokine therapy. Clin. Cancer Res. 4:2337.
Mitropoulos, D., S. Kooi, J. Rodriguez-Villanueva, C. Platsoucas. 1994. Characterization of fresh (uncultured) tumour-infiltrating lymphocytes (TIL) and TIL-derived T cell lines from patients with renal cell carcinoma. Clin. Exp. Immunol. 97:321.
Schendel, D., R. Oberneder, C. Falk, P. Jantzer, S. Kressenstein, B. Maget, A. Hofstetter, G. Riethmuller, E. Nossner. 1997. Cellular and molecular analyses of major histocompatibility complex (MHC) restricted and non-MHC-restricted effector cells recognizing renal cell carcinomas: problems and perspectives for immunotherapy. J. Mol. Med. 75:400.
Schleypen, J., M. V. Geldern, E. Weiss, N. Kotzias, K. Rohrmann, D. Schendel, C. Falk, H. Pohla. 2003. Renal cell carcinoma-infiltrating natural killer cells express differential repertoires of activating and inhibitory receptors and are inhibited by specific HLA class I allotypes. Int. J. Cancer. 106:905.
Gati, A., S. D. Rocha, N. Guerra, B. Escudier, A. Moretta, S. Chouaib, E. Angevin, A. Caignard. 2004. Analysis of the natural killer mediated immune response in metastatic renal cell carcinoma patients. Int. J. Cancer. 109:393.(Emilie Viey, Ga?lle Fromo)
Metastatic renal cell carcinoma, inherently resistant to conventional treatments, is considered immunogenic. Indeed, partial responses are obtained after treatment with cytokines such as IL-2 or IFN-, suggesting that the immune system may control the tumor growth. In this study, we have investigated the ability of the main subset of peripheral lymphocytes, the V9V2-TCR T lymphocytes, to induce an effective cytotoxic response against autologous primary renal cell carcinoma lines. These T cells were expanded ex vivo using a V9V2 agonist, a synthetic phosphoantigen called Phosphostim. From 11 of 15 patients, the peripheral V9V2 T cells were amplified in vitro by stimulating PBMCs with IL-2 and Phosphostim molecule. These expanded V9V2 T cells express activation markers and exhibit an effector/memory phenotype. They display a selective lytic potential toward autologous primary renal tumor cells and not against renal NC. The lytic activity involves the perforin-granzyme pathway and is mainly TCR and NKG2D receptor dependent. Furthermore, an increased expression of MHC class I-related molecule A or B proteins, known ligands of NKG2D, are detected on primary renal tumor cells. Interestingly, from 2 of the 11 positive cultures in response to Phosphostim, expanded-V9V2 T cells present an expression of killer cell Ig-like receptors, suggesting their prior recruitment in vivo. Unexpectedly, on serial frozen sections from three tumors, we observe a lymphocyte infiltrate that was mainly composed of V9V2 T cells. These results outline that V9V2-TCR effectors may represent a promising approach for the treatment of metastatic renal cell carcinoma.
Introduction
The most common histological type of renal cell carcinoma (RCC)3 is conventional cell carcinoma, which represents up to 70–80% of renal tumors. The metastatic forms of these RCC (MRCC), inherently resistant to conventional treatments, are considered as immunogenic tumors based on the clinical response obtained after treatment with cytokines, IL-2 and IFN-. However, the responses are rare, mostly partial, and it is considered that MRCC patients do not yet benefit from efficient therapies (1, 2). The few spontaneous or cytokine-induced regressions in MRCC patients suggest that the immune system may control the tumor growth. Furthermore, the dense CD3+CD8+ T cell infiltrates in some renal tumors indicate an ongoing local immune response. However, ex vivo, the tumor-infiltrating lymphocytes (TILs) display low lytic capacities toward renal tumor cells (TCs) and present poor proliferation (3, 4). In addition, evidence has been provided indicating that TCs, directly or indirectly, may alter the T cell intracellular signaling, leading to an inhibition of their lytic activities (5, 6). These abnormalities may be partially reversed in vitro and in vivo by immunotherapy treatment with IL-2 or IFN- (7).
Within TILs, MHC-restricted and non-MHC-restricted effector cells have been identified that are considered to be part of the host antitumor immune response (8, 9). Among the innate immune response, although peripheral NK cells function appear normal in MRCC patients (10, 11), these cells present defects in their lytic activity at the tumor site (10). Furthermore, adoptive transfer of lymphokine-activated NK cells in MRCC patients did not clearly improve the clinical response as compared with IL-2 alone (12). As important nonconventional immune effectors, the lytic potential of T lymphocytes against different TC lines in vitro has been described (13, 14). In addition, adoptive transfer of ex vivo expanded human T cells in nude mice transplanted with human nasopharyngal tumors supports the in vivo antitumor effects of these cells (15). Furthermore, a recent pilot study with patients presenting lymphoid malignancies suggest that T lymphocytes, stimulated in vivo, induce an objective clinical response (16). A potential role of T cells to control the development of RCC has been suggested (17, 18), but thus far the lytic capacities of these effectors against primary tumor renal cells are not documented.
T lymphocytes bearing a -TCR represent a minor subset of human peripheral T cells (1 to 10%). Most human peripheral T cells display a disulfide-linked V9V2-TCR, are rare in the adult thymus, but increase with age in the blood, suggesting a positive selection in the periphery due to antigenic stimulation (19). Most of them express the CD45RO+CD95+ effector/memory phenotype (20) and lack CD4 and CD8 expression. They are potent and rapid producers of IFN- and TNF- in response to mycobacterial Ags or uncharacterized nonpeptidic ligands expressed on TC (21), indicating their role to coordinate the interplay between innate and acquired immunity. In nature, ligands that can activate T lymphocytes are produced by Gram-positive or -negative bacteria, inducing in some cases polyclonal expansions of V9V2 T cells (22, 23) with their accumulation in the lesions (24). A series of small phosphorylated nonpeptide metabolites synthesized by Mycobacterium tuberculosis able to stimulate V9V2 T cells in vitro have been identified, from which the most potent is the isopentenyl pyrophosphate compound. All of these molecules share a common structural core, the 3-formyl-1-butyl pyrophosphate, described as a strong inductor of in vitro V2 T lymphocytes amplification. A synthetic analog of isopentenyl pyrophosphate, called bromohalohydrin pyrophosphate (BrHPP) molecule, has been developed that mimics the biological properties of natural phosphoantigens and is easily synthesized and active at nanomolar concentrations (25). In this study, BrHPP molecule, now known as Phosphostim molecule, was used to expand V9V2 T cells from PBMCs obtained from MRCC patients. Recently, another group of compounds, the aminobisphosphonates such as pamidronate or risedronate, has been demonstrated to expand and activate the primary human T cells in vitro and in vivo (26)
The peripheral T cells recognize unprocessed nonpeptidic Ags with conserved molecular patterns such as alkylamines and organophosphates (27), by a non-MHC-restricted mechanism, but requiring cell-cell contact (28, 29). Direct binding of -TCR to phosphoantigen has not been demonstrated, suggesting the existence of an unknown presenting molecule (29, 30). The V9V2 T lymphocytes, amplified and stimulated by bacterial phosphorylated metabolites, recognize and kill TCs in vitro (31, 32, 33), suggesting a cross-reactivity between microbial and tumor-associated Ags (25). Furthermore, it was recently shown that the -TCR recognized, on tumor targets, phosphorylated nonpeptidic Ags synthesized by the mevalonate pathway, resembling those produced by microbial cells. This underlines that the accumulation of mevalonate metabolites in TCs may represent a powerful danger signal that activates the immune response and constitute a novel target of tumor immunotherapy (34).
The aim of the present study was to investigate the ability to expand in vitro peripheral V9V2 T lymphocytes from MRCC patients and to evaluate their capacity to lyse autologous primary tumor renal cells. Here, we show that the V9V2 T lymphocytes are expanded from most MRCC patients PBMCs and display a selective lysis against autologous primary renal TCs and not against renal normal cells (NC). Moreover, V9V2 T cells are detected in situ as an important component of the tumor-infiltrating T cells. These observations outline the role of T population in the natural immune response in RCC.
Materials and Methods
Cell lines and cultures
Tumor and adjacent normal renal parenchyma samples were obtained from 15 patients with conventional RCC who had undergone radical nephrectomy. None of the patients had received preoperative therapy. The autologous primary TC lines were derived from the tumor fragments by enzymatic digestion using collagenase (300 U/ml), DNase I (500 U/ml), and hyaluronidase (3000 U/ml) (Sigma-Aldrich). The same procedure was applied to a normal renal fragment, obtained at distance from the tumor, to derive short term normal renal cells. These cells were cultured in DMEM supplemented with 10% FCS and 1% ultroserG (Life Technologies) and subcultured 1/5 when confluent.
Blood samples from MRCC patients (50 ml) were collected 1–2 months after the nephrectomy. PBMCs were isolated by centrifugation on Ficoll-Hypaque density gradient (Amersham Biosciences). PBMCs from healthy donors (n = 7) were treated in parallel with RCC sample as standard control of amplification. Ten million PBMCs were cultured at 2 x 106/ml in 24-well plates in RPMI 1640 supplemented with 10% v/v fetal clone irradiated (HyClone). Polyclonal V9V2 T cell lines were specifically expanded in the presence of 3 μM Phosphostim (BrHPP molecule; Innate Pharma) and 100 IU/ml IL-2 (Aventis Pharma) during 15 days. Phosphostim was added once at the onset of the culture and then every 3 days; one-half of the culture medium volume was replaced by fresh medium containing only 100 IU/ml IL-2.
Immunochemistry
The 15 renal tumors were staged according to primary tumor-regional lymph nodes-distant metastasis clinical classification and examined for tumor stage and histological grade (I–IV according to Fuhrman classification) as depicted in Table I. Immunostaining was performed for six tumors on 5-μm frozen tissue sections fixed in acetone, using the Ventana ES automated immunohistochemistry system (Ventana Medical Systems). The procedure is based on an indirect biotin-avidin system with an universal biotinylated Ig secondary Ab, diaminobenzidine substrate, and hematoxylin counterstain. Primary mAbs included: anti-CD3 (Ventana Medical Systems); anti-CD8 (Dakocytomation; dilution 1/50); anti-pan--TCR (Immu 510, IgG1, dilution 1/20; Immunotech), anti-V1 (R9.12, IgG1, dilution 1/20; Immunotech), anti-V2 (Immu 389, IgG1, dilution 1/20, Immunotech). A negative control slide was incubated with an irrelevant mAb (mouse IgG; Ventana Medical Systems). Positive controls consisted of small intestinal tissue sections stained with the different mAbs described above.
Table I. Clinical and hematological characteristics of patients with RCC (n = 15)a
Immunophenotypic study
Before in vitro culture, 2 x 105 PBMCs from RCC patients were suspended in PBS containing 2% human AB serum for 10 min to block nonspecific binding site and then incubated at 4°C for 30 min with a combination of FITC-, PE-, and phycoerythrin-cyanin 5 (PC5)-conjugated Abs to determine the percentages of NK cells as well as and T cells: CD3-PC5 (UCHT1, IgG1)/V2-FITC (Immu 389, IgG1)/IgG1-PE; CD8-PE (B9.11, IgG1); CD56-PE (N901, IgG1) (mAbs purchased from Immunotech). At day 14 of culture, the bulk of the expanded V9V2 T cells was analyzed by triple staining using the following Ab combination: from Immunotech, CD3-PC5/V2-FITC/CD8-PE, CD56-PE, CD16-PE (3G8, IgG1), CD69-PE (TP1.55.3, IgG2b), HLA-DR-PE (Immu 357, IgG1), CD45RO-PE (UCHL1, IgG2a), CD158a-PE (EB6, IgG1), CD158b-PE (GL183, IgG1), CD158e-PE (Z27, IgG1), NKG2A-PE (Z199, IgG2b); from R&D Systems, NKG2D-PE (clone 149810, IgG1); and from BD Pharmingen, ILT2-PE (GHI/75, IgG2b). Background levels were measured using isotypic controls purchased from Immunotech. Compensation was set up with single-stained samples, low forward scatter elements (RBC and debris) were excluded from analysis, and 10,000 events were collected and analyzed on a FACS-Sort (BD Biosciences) using Cell Quest software (BD Biosciences).
Primary normal and tumor renal cells were phenotyped by indirect single-color fluorescence. Cells (2 x 105) were suspended in PBS containing 2% FCS for 10 min and incubated at 4°C for 30 min with the following Abs: pan-HLA-I (W6/32, IgG2a); CD54 (IgG1); anti-MHC class I-related molecule (MIC) A and anti-MICB (both provided by Immunex; IgG1); anti-NKG2D-Fc (R&D Systems; human IgG1), anti-human fibroblast (AS02, IgG1); G250-FITC (provided by Dr. F. Hirsch). Cells were washed twice with PBS and then incubated for 20 min at 4°C with FITC-conjugated goat anti-mouse Ig or PE-conjugated goat anti-human Ig.
TaqMan (Applied Biosystems) real time quantitative RT-PCR
One microgram of RNA extracted from primary tumor and normal renal cell lines was reverse transcribed using random hexamers according to the manufacturer’s recommendations (Applied Biosystems). Quantitative real time PCR was performed using 5 μl of diluted cDNA (1 μl in 20 μl of water) in a final volume of 25 μl according to the manufacturer’s recommendations (Applied Biosystems). PCR primers and probes for MHC class I-related (MIC) molecules A and B, UL16-binding protein (ULBP)1, ULBP2 and ULBP3 gene targets were designed by Applied Biosystems and used as the manufacturer’s recommendations. The amount of sample RNA was normalized by amplification of an endogenous RNA control (18S). The relative quantification of the transcripts was derived using the standard curve method (Applied Biosystems User Bulletin 2, ABI PRISM 7700 Sequence Detection system) and were normalized to the 18S mRNA of each cell analyzed.
Cytotoxic assays
Expanded V9V2 T cells were tested for cytotoxicity against allogeneic established renal cell lines (RCC6), autologous primary NC and TC lines, sensitive cell line Daudi, and resistant cell line Raji in a 4-h 51Cr release assay. Target cells were labeled with 100 μCi 51Cr for 60 min. The E:T ratio ranged from 30:1 to 3.75:1. Specific lysis (expressed as percentage) was calculated using the standard formula [(experimental – spontaneous release/total – spontaneous release) x 100] and correspond to the mean of the triplicate. Lysis mediated by V9V2 T cells of normal vs tumor primary renal cells was compared using Student’s test. In some experiments, Immu 510 mAb (anti-pan--TCR IgG1; Immunotech), ON72 mAb (anti-NKG2D, IgG1; Innate Pharma), or control mouse Ig (IgG1) were added at predefined saturating concentrations at the onset of the cytolytic assay before exposure to labeled target cells. In some experiments, effector T cells were preincubated with 10 nM concanamycin A (CMA; Sigma-Aldrich) at 37°C for 20 min to block the perforin-granzyme pathway and then added to labeled target cells. In other experiments, targets wereincubated with saturating concentrations of anti-HLA-BC mAb or mouse control IgG2a for 30 min; then, effector T lymphocytes were added. Data were expressed as the percentages of specific lysis.
Cold target competition assays were performed. A standard 4-h 51Cr release assay was realized in presence of a 40-fold excess of unlabeled targets (DAUDI, RAJI cell lines, RCC6 and NC or TC) that were added to 51Cr labeled TC or NC (2 x 103 per well) before the incubation with autologous activated lymphocytes.
KIR functionality was assessed using a mAb-redirected lysis assay using P815 mastocytoma mouse cells. Briefly, 51Cr labeled P815 target cells (3 x 103) coated with anti-pan--TCR mAb (10 μg/ml) were incubated with expanded V9V2 T effectors (E:T 5:1). TCR-redirected lysis of labeled P815 was modulated by the presence of saturating concentrations of anti-KIR mAbs (5 μg/ml) or mouse control IgG1.
Results
V2 T lymphocytes infiltrate RCC tumors
Fifteen MRCC patients undergoing partial or complete nephrectomy with lymph node resection in some cases are included in this study. They were all diagnosed as conventional RCC. None of them had received any previous treatment. Consecutive sections of frozen tumor fragments were analyzed as to immunohistochemical staining with anti-CD3, anti-CD8, anti-pan--TCR, anti-V1, and V2-TCR mAbs. Before immunostaining, the extend of the inflammatory cells infiltrate was evaluated by H&E- safranin staining. Three of seven tumors analyzed displayed a dense CD3/CD8 T cell infiltrate and were selected for further immunostaining. Fig. 1 shows the immune infiltrate (CD3+ T cells) of tumor sections from patients 1 (Fig. 1A) and 2 (Fig. 1B). As depicted, numerous T cells are detected in situ at the contact with TCs and interestingly are mainly composed of the V 2 subset. The frozen tumor sections from four other patients displayed a moderate immune infiltrate, but in these tumors also the V2 T lymphocytes represent the prevalent tumor infiltrating T cell population (data not shown). Normal renal parenchyma was devoid of any infiltrating T cells.
FIGURE 1. V2 TCR-bearing T cells are the major component of the T cell infiltrate of renal tumors. Immunohistochemical analysis of serial frozen tumor tissue sections from patient 1 (A) and patient 2 (B). Tissue sections were stained using mAb specific for CD3, CD8, pan--TCR, V1 TCR, and V2 TCR. Positive cells were stained in brown. Two of six tumors analyzed are shown.
Phosphostim stimulation induced the expansion of peripheral V2 T cells from PBMCs of MRCC patients
The clinical and hematological characteristics of the patients included in this study are summarized in Table I. Of 15 RCC patients, 2 were lymphopenic (<103 lymphocytes/ml). Before in vitro stimulation, the percentages of NK cells and CD8 T lymphocytes from MRCC patients were in the range of normal donors (8.5–30.1% for NK cells, 13–60.8% for CD8 T cells). In MRCC patients as well as in donors, the percentage of T lymphocytes represented <5% of total PBMCs (range, 0.3–4.1%). PBMCs from these 15 patients and 7 donors, used as standard control of expansion, were cultured with Phosphostim and IL-2 for 2 wk. After 14 days, 11 of 15 cultures derived from patients contained a percentage of V9V2 T lymphocytes ranging between 73.3 and 96.4% and presented a high rate of expansion (>200-fold). The corresponding patients were designated responders to the Phosphostim expansion test (patients 3, 5–9, and 11–15). For the other cultures, poor expansion rate (<30-fold) and low percentages (<20%) of V2 T lymphocytes were obtained; these patients were designated nonresponders to the expansion test and were not included in functional tests due to low cell numbers (patients 1, 2, 4, and 10). In control cultures derived from 7 donors’ PBMCs, 80–96.4% of V2 T cells were obtained after a 2-wk amplification with Phosphostim molecule. Furthermore, the amplification rate of responder patients (range, 248- to 2787-fold expansions; Table I) were similar of those from donors (range, 141- to 1236-fold expansions; data not shown).
Phenotypic analysis of expanded-V2 T cells from MRCC patients’ PBMCs
The Phosphostim-stimulated V9V2 T cells from MRCC patients were phenotyped at day 14 of culture, using the panel of mAbs described in Materials and Methods. The phenotypic characteristics are summarized in Table II, and a representative V9V2 T cell culture is depicted in Fig. 2. Polyclonal V9V2 T cells from MRCC patients display a memory phenotype consistent with a bright expression of CD45RO (>95%; data not shown). They are in an activated state (CD69+HLA-DR+); and compared with cultures from donors, V9V2 T cells from MRCC patients exhibit a significantly increased level of CD8 expression assessed by MFI (532 ± 192 in RCC vs 297 ± 27 in donors, p < 0.09), whereas percentages of CD8 cells were not significantly increased(31.7 ±12.6% vs 24.1 ± 11.3%; p < 0.2). Expanded T cells from patients’ PBMCs, bearing the V9V2 TCR, expressed on their cell surface the activating NKG2D receptors. The percentages varied among MRCC patients (mean, 85.6%; range, 74.2–99.9%) but were similar to those on expanded V9V2 T lymphocytes from a series of donors (mean, 88.9%; range, 67.2–99.9%), except patient 7 that displayed a low NKG2D expression (27.2%).
Table II. Phenotypic analysis of expanded-V2 T lymphocytes from MRCC patients and donorsa
FIGURE 2. Phenotypic analysis of expanded V9V2 T lymphocytes from a RCC patient (patient 12). Expanded-V9V2 T cells were stained with a panel of conjugated mAbs described in Materials and Methods and analyzed by flow cytometry. The histograms show staining on the gated CD3/V2 subset. Open histograms are isotypic IgG control staining.
The control of V9V2 T cell activation involves a balance between stimulatory and inhibitory signaling delivered, at least in part, by the heterodimeric CD94-NKG2-A receptor. There was no clear difference in NKG2-A receptor expression between donors and MRCC patients (Table II). Furthermore, 2 of the 11 patients’ V9V2 T cell cultures exhibited an elevated percentage of KIR2DL/S (CD158b) protein (>25%) (Table II, patients 3 and 7). However, one of seven donors exhibited 70% of CD158b+ expanded-V9V2 T cells. A comparable level of KIR expression on V2 T cells was found on two different cultures from the same patient derived 2 and 10 mo postnephrectomy, indicating a stable expression of these receptors.
Lysis of autologous primary tumor renal cells by expanded-V2 T cells
Tumor and normal renal cells were successfully derived from surgical nephrectomy fragments for 10 responder patients except for patient 9 which presented a fibroblastic contamination of primary culture and thus was not included in functional studies. Primary tumor renal cells are characterized by homogeneous staining with anti-G250 mAb which recognize an Ag expressed exclusively by renal cancer cells. In addition, the absence of contamination of the TC and NC lines by fibroblasts was evaluated by staining with a specific mAb (AS02). Early in vitro passages were used as targets in the functional assays as well as a well-characterized renal TC line, RCC6 (35), that was included in functional studies. The Burkitt’s lymphoma cell lines Daudi and Raji, were used as sensitive and resistant targets, respectively. In a first set of experiments, expanded-V9V2 T cells from MRCC patients were assessed for their lytic potential toward the different targets. These T effectors killed the Daudi cells with a high efficiency, similar to that of donor expanded-V9V2 T cells, indicating that these cells mainly mediate a non-MHC-restricted lysis. Interestingly, the expanded- V9V2 T cells from patients efficiently killed autologous primary renal TCs and RCC6 and displayed a low lytic potential toward their autologous normal renal counterparts similar to that of Raji (Fig. 3A, patient 14). Data depicted in Fig. 3B summarize the cytotoxic activity of V9V2 cultures from 10 RCC patients, evaluated in autologous setting, against tumor and normal renal targets and show that the mean lytic activity of T effectors against primary tumor renal cells was significantly higher than that against normal renal cells (p < 0.05 for all E:T ratios). In addition, cold target competition assays with addition of 40-fold excess of various cold targets were performed to determine whether the same antigenic structure is present on the different targets. Addition of excess of cold Daudi and RCC6 cell lines caused a decrease in lysis of autologous TC and NC lines by V9V2 T cells, whereas Raji cells did not modulate the lysis. Furthermore, primary NC line competed with 51Cr-labeled TC lines (Fig. 3C). These data indicate that T cells recognize a common structure on Daudi and renal TCs and NCs.
FIGURE 3. Expanded-V9V2 T cells from RCC patients recognized and killed primary TCs. A, Cytotoxic activity (4-h 51Cr release assay) of expanded- V9V2 T lymphocytes against autologous primary TC lines (patient 14) was measured for the indicated E:T ratio. The polyclonal V2 T lymphocytes were tested against different targets: Daudi and Raji cell lines as susceptible and resistant targets, respectively; RCC6 renal TC line; and the autologous primary NC and TC lines. Data are presented as the mean percentage of specific target lysis. B, Mean cytolytic activity of expanded- V9V2 T lymphocytes from 10 responder patients’ PBMCs against the pair of primary renal cell lines (TC and NC) as well as Daudi and Raji cell lines. The percentage of lysis for each target cells, at different E:T ratios, represents the mean value of cytotoxicity obtained for 10 patients. p values between primary NCs and TCs, determined by a standard Student test, are indicated. C, Cold target competition assay was performed with expanded-V9V2 T cells from patient 14. The ability of these effector cells to lyse autologous primary TC (C, left) and NC (C, right) were investigated in presence of a 40-fold excess of different cold targets. Data are presented as the mean percentage of specific target lysis. md, Medium.
TCR and NKG2D receptors are involved in RCC target cell lysis
The main cytolytic pathway used by killer cells involves the release of the cytolytic granules in effector-target interface. To determine the involvement of this pathway in V9V2 T cell-mediated lysis, effector cells were preincubated with CMA, which inhibits the perforin-based cytotoxicity by increasing the pH of lytic granules. The lytic potential of expanded-V9V2 T cells was completely abrogated in the presence of CMA during the chromium release assay, indicating that the cytotoxic activity of these effectors involves mainly the perforin-granzyme pathway (Fig. 4A).
FIGURE 4. Mechanisms underlying the cytotoxic activity of expanded- V9V2 T cells. A, The lytic activity of expanded-V9V2 T lymphocytes involves the perforin-granzyme pathway. Expanded-V9V2 T cells (patient 3) were incubated with CMA during the cytotoxicity assay against the pair of primary autologous renal cells and Daudi cell lines. Results are the lysis percentages of targets determined at the indicated E:T ratio. B, The V9V2 T lymphocyte cytolytic activity implicated the TCR and NKG2D receptors. Expanded-V9V2 T cells from patient PBMCs (patient 3) were incubated with anti-pan--TCR mAb (clone IMMU 510) and/or anti-NKGD mAb (clone ON72) before their addition to autologous primary TCs or to RCC6 cells (C). md, Medium; cIg, control Ig.
The majority of amplified V9V2 T cells from MRCC patients express the activating C-type lectin receptor NKG2D (see Table II), that provides a costimulatory signal, amplifying the TCR signal for the lysis of TCs (36). To delineate the respective involvement of NKG2D and TCR signaling in V9V2-mediated lysis, we performed cytolysis assays in presence of saturating concentrations of blocking mAbs. Autologous and allogeneic (RCC6) renal TCs were used as targets. As depicted in Fig. 4B, lysis of autologous TCs by expanded-V9V2 T lymphocytes (patient 3) is clearly decreased in presence of anti-pan--TCR mAb (Immu 510), indicating that this receptor is involved in the lysis of TCs (63% inhibition at E:T 15:1, n = 3). Blocking of the NKG2D receptor (mAb ON72) also decreased the lysis of targets but to a lesser extent (38% of inhibition at E:T 15:1). Furthermore, addition of the two mAbs during the chromium assay had an additional effect (75% inhibition at E:T 15:1). As shown in Fig. 4C, addition of anti-TCR or anti-NKG2D mAbs inhibited the cytotoxic activity of expanded-V9V2 T effectors against the allogeneic renal cell line RCC6. Similar results were obtained from two other V9V2 expanded cell lines from patients, clearly demonstrating the involvement of TCR in the cytotoxicity mediated by these effectors (patients 6 and 7, not shown) and confirming the costimulating role of NKG2D receptor in T cell lysis.
Differential expression of NKG2D ligands on primary normal and tumor renal cells
NKG2D ligands such as MICA, MICB, and ULBP proteins may be induced by cellular stress, and most of them are frequently expressed on TCs (37, 38). This indicates that NKG2D provides a first line surveillance against stressed or transformed cells. A flow cytometry analysis of primary TC and NC lines was performed. As shown in Fig. 5B, primary TC lines exhibit an increase of MICA and MICB protein expression compared with normal counterparts (patient 14). Furthermore, the fusion protein NKG2D Fc, which recognize with a low affinity all ligands of the NKG2D receptor, bind to renal TCs with a high efficiency. In addition, a quantitative analysis by (real time) RT-PCR for the ULBP1,2,3 as well as MICA and MICB genes was performed on the pairs of primary renal cells derived from eight patients (patients 3, 5, 7, 8, and 12–14) indicating that the primary TC lines display increased of ULBP1,2,3 transcripts as compared with primary NC lines. No significant difference between both primary cell lines was observed for the MICA and MICB transcripts (Fig. 5A). Phenotypic analysis of short term primary cells lines indicate that they were devoid of contamination by fibroblasts and that TCs express the G250 tumor-specific Ag. Furthermore, the expression of HLA-I molecules, which can trigger an inhibitory receptors such as CD94-NKG2-A, is similar in both normal and tumor renal cells. In addition, a high and comparable expression of adhesion molecules, ICAM-1, is observed on both normal and tumor renal cells (Fig. 5C).
FIGURE 5. Analysis of NKG2D ligands on primary renal cell lines. A, Quantitative analysis of MICA, MICB, ULBP1, ULBP2, and ULBP3 mRNA transcripts in primary renal TCs and NCs from eight MRCC patients. Results are expressed in quantity of mRNA of the corresponding transcripts normalized to the 18S mRNA of each cell line analyzed. p values between primary NCs and TCs, determined by a standard Student t test, are indicated. B, Phenotype of paired primary TC and NC lines (patient 14). Primary renal cell lines were stained by indirect fluorescence with the panel of mAbs described in Materials and Methods and analyzed by flow cytometry. Open profiles are isotypic IgG control staining. C, Phenotypic analysis of primary tumor and normal renal cells (patient 7). Epithelial cell lines were assessed for tumor-specific G250 Ag, fibroblast contamination (AS02), HLA-I, and ICAM-1 molecules. md, Medium; cIg, control Ig.
KIRs expressed by amplified V2 T cells from MRCC patients are functional
It has been previously shown that peripheral T cells from donors expressed the C-lectin type inhibitory NK receptor CD94-NKG2-A whereas Ig-type receptors (KIRs) are present on discrete subsets (39, 40). In this study, we show that the expression of KIR receptors on expanded V2 T cells was increased in 2 of 10 RCC patients (Table II) and in 1 of 7 donors (Donor 3). To analyze the influence of this inhibitory receptor on lytic activity of V9V2 T cells from patient 3 (CD158b >25%), mAb-redirected lysis experiments with FcR+ murine P815 cells were performed. As shown in Fig. 6A, engagement of TCR on the bulk of V9V2 T cells by specific mAb-induced killing of P815 targets. This lysis is significantly decreased when KIR2DL/S is triggered by anti-CD158b mAb (inhibition, >30%; n = 3), whereas anti-CD158a mAb had no effect. Triggering of KIR also inhibited TCR-induced lysis of KIR+ V9V2 T cells derived from donor 3 (data not shown). Furthermore, the lysis of autologous primary TCs by the bulk of V9V2 T cells was slightly increased in presence of anti-HLA-B/C mAb to block KIR/HLA-C interactions (Fig. 6B), indicating that engagement of KIR2DL/S molecule by specific ligands expressed on target cells may partly protect TCs from lysis by the V9V2 T effectors.
FIGURE 6. KIR receptors exert an inhibitory effect on lytic activity of V9V2 T cells. A, TCR redirected lysis assays of FcR+ P815 target cells using expanded-V9V2 T cells from patient 3, by triggering of -TCR (Immu 510 mAb). Inhibitory effect of KIR2DL/S molecule was tested using an anti-CD158b mAb. The data were obtained at E:T 5:1. B, Modulation of the cytolysis of autologous primary TCs by blocking the HLA-KIR interaction. Targets were preincubated with anti-HLA-B/C mAb or control isotype (cIg) used at 5 μg/ml before addition of the V9V2 T cells. One representative experiment of three is shown.
Discussion
The implication of T lymphocytes in the immune defense against bacterial, microbial, or parasitic infection has been well documented. Once activated, these cells induce immediate effector functions such as cytokine production or cytotoxic activity (30) and are able to develop a memory response against microbial reinfection (41). In addition, recent studies outlined the role for human T cells in recognition and lysis of transformed cells of different origins, mainly of hematopoietic lineage. In solid tumors, such as RCC, an immune T cell response has been described, mainly implicating T lymphocytes, whereas the role of T cells remains poorly documented.
This study gives strong evidence in support of a role of peripheral ex vivo expanded and activated V9V2 T cells in the immune surveillance against renal tumors. The V9V2 T lymphocyte population represents the majority of circulating lymphocytes and recognizes natural and synthetic phosphoantigens. These molecules induce an expansion/activation of this subpopulation cells in primary PBMC cultures of healthy donors. Using a synthetic phosphoantigen, Phosphostim, our results demonstrate that 11 of 15 cultures of patients’ PBMCs respond to in vitro stimulation, resulting in a strong amplification of V9V2 T cells. The patients included in the study had advanced renal tumors (grade III–IV), and most of them had metastasis. Besides lymphopenia (<1000), neither percentages of T cells at day 0 of culture nor percentages of CD3+CD8+ are truly indicative of the responder status. Increasing cell concentration (x2 or 3) at the onset of the culture is not effective to reverse the absence of amplification in nonresponder patients. Given that it is known that immunological responses may be suppressed after surgery and that this effect disappears several months after, the patients who are nonresponders to the expansion test will be followed to determine whether the immunosuppressive status observed in our cultures can be reversed with time after the nephrectomy.
In healthy adults, the circulating V9V2 T lymphocytes are mainly composed of CD45RO+ and CD95+ memory cells. This phenotype is similar to that found for V9V2 T cells from MRCC patients. It was reported that activation of T cells induced an increased expression of CD27 marker which gradually switched off (42). For some patients’ V9V2 cultures, we have detected a low expression of CD27 marker (data not shown) which correlates with the up-regulation of CD56 and CD8 molecules as well as early (CD69) and late (HLA-DR) markers of activation fitting with their effector T cells status. Previous studies reported that in vitro activation in the presence of IL-2 down-regulates CD27 expression, confirming that primed T cells become effector cells. Because our data implicate T cells in the immune response against RCC, it would be interesting to compare the V9V2 T cell populations from healthy and patients before Phosphostim expansion for CD27, CD28, CD45RA/RO, and CCR7 markers to precisely define the distribution of effector/memory T cell subsets (43).
The major aim of our study was to investigate the lytic capacity of the polyclonal expanded V9V2 T cells toward primary renal TCs. From the 10 T cell cultures evaluated in autologous setting, we showed that the amplified V9V2 T lymphocytes present a higher lytic potential toward primary tumor renal cells in comparison with primary normal renal cells. The rare contaminant NK or T cells in the cultures as well as the inhibition of lysis in the presence of TCR- mAb imply that lysis of renal TCs is mediated by T cells. Cold targets competition experiments indicate that a similar structure is recognized by T lymphocytes on normal and tumor renal cells, but additional engagement of other triggering receptors may explain increased lysis toward TCs. Furthermore, our data indicated that besides TCR, NKG2D receptors deliver mainly a costimulating effect involved in the lysis of tumor renal cells. Concerning the expression of the corresponding ligands, MIC molecules were overexpressed on primary TC lines, although no difference was observed for the MICA or MICB transcripts levels. Increased transcripts for ULBP1, 2, and 3 was found in primary renal TCs, and binding of NKG2D-Fc protein was higher on renal TCs. Altogether, these results favor the involvement of the NKG2D receptor in the preferential lysis of renal TCs by V9V2 T cell-mediated killing. In contrast, primary normal and tumor renal cells exhibit similar expression of HLA-I molecules, excluding a differential engagement of inhibitory NK receptors by tumor vs normal renal cells, and indicate that V9V2 T lymphocytes recognize and engage TCR ligands likely preferentially expressed on TC.
The response mediated by peripheral blood V9V2 T cells against the invading pathogen is rapid and potent and requires tight regulation to prevent autoimmunity or immunopathological manifestations. One of the regulatory mechanisms of cytotoxic effectors implicate the inhibitory MHC class I receptors expressed as well by NK cells and some T lymphocyte clones. Circulating T cells from donors express mainly the CD94-NKG2-A complex (80%), which recognizes HLA-E, a nonclassical MHC class I molecule (44). The inhibitory receptor expression on V9V2 T cell clones from donors has been also documented (39), implying that the biological role of the major subset of human T lymphocytes is closely related to that of NK cells. In the present study, 2 of 11 cultures derived from patients’ MRCC PBMCs and 1 of 7 donors present an expression of the KIR2DL/S molecule. This KIR expression was confirmed on PBMCs before culture with Phosphostim, indicating that, as for -TCR T cells, KIRs were induced in vivo (data not shown). This is to our knowledge the first report showing an enhanced KIR expression on peripheralV 9V2 T cells from MRCC patients amplified ex vivo with Phosphostim. Our functional results suggest that this receptor is inhibitory, decreasing the TCR-induced lysis, and may control the lysis of autologous tumor renal cells, although these results should be extended and confirmed on immunoselected KIR+ V9V2 T cells. Altogether, the expression of KIR and the memory/effector phenotype of expanded-V9V2 T cells from MRCC patients reflect Ag-experienced T cells that are likely to respond to a chronic stimulation by a tumor.
At present, it remains unclear whether subsets of human peripheral T cells have the ability to home to damaged tissues such as tumors. The tumor immunosurveillance in vivo has mainly implicated the V1 intraepithelial population (8, 45) but little information about the antitumor potential of peripheral V9V2 T cells has been reported. However, phenotypic analysis of infiltrating T cells in melanoma revealed that they bear the V9V2 TCR and display in vitro a potent non-MHC-restricted lytic potential toward autologous TCs (46, 47). In contrast, a recent study using SCID mice transplanted with human autologous melanoma cells demonstrated that NK cells, V1 and V2 T lymphocytes are able to migrate in tumor site but that only the first two populations are able to prevent and inhibit the tumor growth. Surprisingly, the V2 lymphocyte had no effect even though this population has a higher in vitro cytotoxic potential than do V1 T cells (48). In our study, immunostaining of renal tumor sections showed the presence of the V2 TCR T lymphocytes at close contact with renal TCs. Although the density of the infiltrate may vary from one tumor to another, T cells are well represented in situ, and the major subset of this population bear the V2 TCR, suggesting that this peripheral population has the capacity to migrate to extra lymphoid sites and to home to the tumor.
The high percentages of patient responders to Phosphostim stimulation, the activation state of the expanded-V9V2 effectors and their ability to kill the autologous primary renal TC lines are very encouraging results to develop the use of such compound in vivo. It has been recently reported that treatment of patients presenting multiple myeloma using aminobisphosphonates has induced objective remission or inhibition of disease progression. Interestingly, the in vivo responses correlated with a high expansion rate of cytotoxic T lymphocytes in in vitro cultures of thecorresponding patients (16). These in vivo data, together with our present results, favor the clinical trials using Phosphostim and Il-2 for a new therapeutic approach of MRCC.
Acknowledgments
We thank Stéphane Leborgne for managing of the blood samples from RCC patients and Catherine Gaudin for help in real time quantitative reverse transcription-PCR experiments.
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 Institut National de la Santé et de la Recherche Médicale Grant 4613 (l’Association pour la Recherche sur le Cancer) (to A.C.). E.V. is the recipient of a Conventions Industrielles de Formation par la Recherche Européennes grant from Innate Pharma.
2 Address correspondence and reprint requests to Dr. Anne Caignard, Institut National de la Santé et de la Recherche Médicale Unité 487, Cytokines et Immunologies des Tumeurs Humaines, Institut Gustave Roussy, PR1, 39 rue Camille Desmoulins, F-94805 Villejuif, France. E-mail address: CAIGNARD@IGR.FR
3 Abbreviations used in this paper: RCC, renal cell carcinoma; MRCC, metastatic form of renal cell carcinoma; TIL, tumor-infiltrating lymphocyte; BrHPP, bromohalohydrin pyrophosphate; PC5, phycoerythrin-cyanin 5; TC, tumor cells; NC, normal cells; CMA, concanamycin A; ULBP, UL16-binding protein; MIC, MHC class I-related molecule.
Received for publication July 14, 2004. Accepted for publication October 24, 2004.
References
Escudier, B., C. Chevreau, C. Lasset, J. Douillard, A. Ravaud, M. Fabbro, A. Caty, J. Rossi, P. Viens, J. Bergerat, J. Savary, S. Negrier. 1999. Cytokines in metastatic renal cell carcinoma: is it useful to switch to interleukin-2 or interferon after failure of a first treatment? Groupe Francais d’Immunotherapie. J. Clin. Oncol. 17:2039.
Negrier, S., B. Escudier, F. Gomez, J. Douillard, A. Ravaud, C. Chevreau, M. Buclon, D. Perol, C. Lasset. 2002. Prognostic factors of survival and rapid progression in 782 patients with metastatic renal carcinomas treated by cytokines: a report from the Groupe Francais d’Immunotherapie. Ann. Oncol. 13:1460.
Finke, J., A. Zea, J. Stanley, D. Longo, H. Mizoguchi, R. Tubbs, R. Wiltrout, J. O’Shea, S. Kudoh, E. Klein. 1993. Loss of T-cell receptor chain and p56lck in T-cells infiltrating human renal cell carcinoma. Cancer Res. 53:5613.
Rayman, P., R. Uzzo, V. Kolenko, T. Bloom, M. Cathcart, L. Molto, A. Novick, R. Bukowski, T. Hamilton, J. Finke. 2000. Tumor-induced dysfunction in interleukin-2 production and interleukin-2 receptor signaling: a mechanism of immune escape. Cancer J. Sci. Am. 6:(Suppl. 1):S81.
Uzzo, R., P. Clark, P. Rayman, T. Bloom, L. Rybicki, A. Novick, R. Bukowski, J. Finke. 1999. Alterations in NFB activation in T lymphocytes of patients with renal cell carcinoma. J. Natl. Cancer Inst. 91:748.
Molto, L., P. Rayman, E. Paszkiewicz-Kozik, M. Thornton, L. Reese, J. Thomas, T. Das, D. Kudo, R. Bukowski, J. Finke, C. Tannenbaum. 2003. The Bcl-2 transgene protects T cells from renal cell carcinoma-mediated apoptosis. Clin. Cancer Res. 11:4060.
Bukowski, R., P. Rayman, R. Uzzo, T. Bloom, K. Sandstrom, D. Peereboom, T. Olencki, G. Budd, D. McLain, P. Elson, A. Novick, J. H. Finke. 1998. Signal transduction abnormalities in T lymphocytes from patients with advanced renal carcinoma: clinical relevance and effects of cytokine therapy. Clin. Cancer Res. 4:2337.
Mitropoulos, D., S. Kooi, J. Rodriguez-Villanueva, C. Platsoucas. 1994. Characterization of fresh (uncultured) tumour-infiltrating lymphocytes (TIL) and TIL-derived T cell lines from patients with renal cell carcinoma. Clin. Exp. Immunol. 97:321.
Schendel, D., R. Oberneder, C. Falk, P. Jantzer, S. Kressenstein, B. Maget, A. Hofstetter, G. Riethmuller, E. Nossner. 1997. Cellular and molecular analyses of major histocompatibility complex (MHC) restricted and non-MHC-restricted effector cells recognizing renal cell carcinomas: problems and perspectives for immunotherapy. J. Mol. Med. 75:400.
Schleypen, J., M. V. Geldern, E. Weiss, N. Kotzias, K. Rohrmann, D. Schendel, C. Falk, H. Pohla. 2003. Renal cell carcinoma-infiltrating natural killer cells express differential repertoires of activating and inhibitory receptors and are inhibited by specific HLA class I allotypes. Int. J. Cancer. 106:905.
Gati, A., S. D. Rocha, N. Guerra, B. Escudier, A. Moretta, S. Chouaib, E. Angevin, A. Caignard. 2004. Analysis of the natural killer mediated immune response in metastatic renal cell carcinoma patients. Int. J. Cancer. 109:393.(Emilie Viey, Ga?lle Fromo)