|
|
||||||||

T Cells by Pamidronate1



* Department of Immunology and Cell Biology,
Graduate Schools of Biostudies and Medicine, Kyoto University, Kyoto, Japan; and
PRESTO, JST, Kyoto, Japan
| Abstract |
|---|
|
|
|---|

T cells bearing V
2V
2-TCR recognize various kinds of small nonpeptide Ags, and activation of them by a nitrogen-containing bisphosphonate Ag, pamidronate, requires Ag presentation by cells other than 
T cells, including many human tumor cells. Present results demonstrated that tumor cell lines of nonhuman origins pulsed with pamidronate failed to activate human 
T cells without exception, whereas most if not all human tumor cell lines could do so. 
T cells formed stable conjugates with pamidronate-pulsed human tumor cells and both conjugate formation and 
T cell activation were inhibited significantly by anti-LFA-1 mAb, suggesting the requirement of LFA-1-mediated interaction with APC for efficient 
T cell activation. Consistently, ICAM-1low tumor cell lines pulsed with pamidronate induced no or only weak activation of 
T cells, whereas similarly treated ICAM-1high cell lines could activate them. One of the two ICAM-1low tumor cell lines pulsed with pamidronate induced strong 
T cell activation after ICAM-1 gene transfer. However, another ICAM-1low human cell line as well as murine tumor cell lines pulsed with pamidronate remained totally defective in 
T cell activation even after expression of human ICAM-1. These results suggested that activation of human 
T cells by nonpeptide Ags required species-specific interactions in addition to LFA-1/ICAM-1-mediated cell adhesion with APC. | Introduction |
|---|
|
|
|---|

T cells bearing V
2V
2-
TCR, a predominant 
T cell subset found in adult peripheral blood, specifically recognize various kinds of nonpeptide Ags (reviewed in Ref. 1), such as bacterial pyrophosphomonoesters (2, 3, 4, 5) and alkyl amines (6) as well as synthetic nitrogen-containing bisphosphonates (7, 8, 9) in the V
2V
2- 
TCR-dependent manner (2, 3, 4, 5, 6, 7, 8, 9, 10, 11). However, it remains unknown how the 
TCR recognizes such nonpeptide Ags of small molecular masses. We previously reported that the recognition was dependent on the lysine residues encoded by a germline J
1.2 gene segment proximal to the junctional region (11) located in a putative Ag-binding pocket (12). These lysine residues are not conserved in other human nor murine J
segments, which may in part explain why recognition of these nonpeptide Ags is a unique feature of human V
2J
1.2V
2- 
T cells. Among nonpeptide Ags, nitrogen-containing bisphosphonates such as pamidronate were rather unique, in that activation of the primary 
T cells required the presence of monocytes (9). Furthermore, many if not all human tumor cell lines prepulsed with pamidronate could efficiently activate human 
T cells, whereas no TCR-dependent recognition was observed in mock-treated tumor cells (13). These results strongly suggested that pamidronate had to be "presented" by selected cells other than 
T cells per se for effective 
T cell activation.
In the present study, we show that the optimal activation of 
T cells by pamidronate-pulsed human tumor cells was dependent on the stable cell adhesion mediated by LFA-1 expressed on 
T cells and ICAM-1 on the Ag-pulsed tumor cells. Although human tumor cell lines with high ICAM-1 expression (ICAM-1high) could efficiently present pamidronate to 
T cells, a minor population with little or no expression of ICAMs (ICAM-1low) as well as any of the tumor cell lines of nonhuman origins pulsed with pamidronate failed to activate human 
T cells. Human ICAM-1 expression by cDNA transfer could confer the strong 
T cell-activating capacity in one ICAM-1low human tumor cell line. However, no such induction was observed in another ICAM-1low human tumor as well as mouse tumor cell lines, suggesting the requirement of additional species-specific interaction(s) with Ag-presenting tumor cells for the effective activation of 
T cells by a nonpeptide Ag, pamidronate. These results argue for the presence of a specific mechanism for nonpeptide Ag presentation to human 
T cells, complementing that for 
T cell activation by peptide Ags.
| Materials and Methods |
|---|
|
|
|---|
Monoethyl pyrophosphate was synthesized as described previously (4) and pamidronate was purchased from Novartis Pharma (Nuernberg, Germany).
Human 
T cells
PBMC were separated from healthy volunteers blood using Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) and stimulated with 0.2 mM monoethyl pyrophosphate in Yssels medium supplemented with 2.5% human serum albumin for 2 days. The cells were further cultured in Yssels medium containing 10% human AB sera and IL-2 (100 U/ml) for 14 days. The proportions of V
2V
2-bearing 
T cells in such populations were consistently 95% or more in CD3-positive cells and were used as polyclonal 
T cell lines in the present study.
Tumor cell lines and pulsing with nonpeptide Ag
Tumor cell lines of human and nonhuman origins were supplied from Human Science Research Resource Bank (Osaka, Japan) and cultured according to the original instructions. To pulse the tumor cells with pamidronate, 1.0 x 106 tumor cells were incubated with 0.1 mM pamidronate for 2 h at 37°C and washed extensively with RPMI 1640 containing 10% FCS before use.
Flow cytometry and mAbs
Single- and two-color flow cytometric analyses were conducted using FACScan (BD Biosciences, Mountain View, CA) as described previously (13). mAbs used in the present study were anti-TCR
2 (Immunotech, Marseille, France), anti-IFN-
(BD PharMingen, San Diego, CA), anti-ICAM-1, -2, and -3 (BD PharMingen), anti-LFA-1 (TS1/22), and anti-LFA-1 specific for an activated form (NKI-L16), kindly provided by Dr. T. Kinashi (Kyoto University, Kyoto, Japan).
Proliferation assay
Human 
T cells were cocultured at 1.0 x 105/well with varying numbers (1.2 x 1032.0 x 105) of mitomycin C (100 µg/ml)-treated tumor cells at 37°C for 36 h and pulsed with [3H]thymidine (2 µCi) for 12 h, and radioactivity incorporated in DNA was counted by a MicroBeta scintillation counter (PerkinElmer Life Sciences, Norwalk, CT).
Cytotoxicity assay
Tumor cells (1.0 x 106) pretreated with 0.1 mM pamidronate were labeled with 100 µCi of [51Cr]sodium chromate for 1 h at 37°C. Varying numbers of human 
T cells (0.625 x 1044.0 x 105) were cocultured with 1.0 x 104 tumor cells for 5 h, and 51Cr release was counted by a gamma counter (PerkinElmer Life Sciences).
Intracellular IFN-
staining
Human 
T cells (5.0 x 105/well) were cocultured with pamidronate-pulsed tumor cells (5.0 x 105/well) at 37°C. After 2 h, brefeldin A (10 µg/ml) (Sigma-Aldrich, St. Louis, MO) was added in the culture and the cells were incubated for another 2 h. The cells were then washed and stained with FITC-conjugated anti-TCR-V
2 mAb on ice for 20 min. The cells were washed twice, fixed with 1% paraformaldehyde, permeabilized with 0.5% saponin, and stained with PE-conjugated anti-IFN-
mAb for 20 min. After washing three times, the cells were analyzed using FACScan (BD Biosciences).
Conjugation assay
Tumor cells were stained with PKH26 (Molecular Probes, Eugene, OR) at room temperature for 5 min and human 
T cells were stained with CFSE (Molecular Probes) at 37°C for 15 min. After washing, the labeled 
T cells and tumor cells were incubated for varying periods and were two-color analyzed after gentle vortexing using FACScan.
Human ICAM-1 cDNA transfection
A full-length human (h)3 ICAM-1 cDNA was kindly provided by Dr. O. Takashi (Daiichi Pharmaceutical, Tokyo, Japan). Tumor cell lines were transfected with hICAM-1 cDNA constructed in a pEF-BOS expression vector, kindly provided by Dr. S. Nagata (Osaka University, Suita, Osaka, Japan), by electroporation using ECM830 (BTX). The cells were cultured in the medium containing 1 mg/ml G418 (Nacalaitescque, Osaka, Japan) and screened for the expression using anti-ICAM-1 mAb to establish stable clones highly expressing hICAM-1.
| Results |
|---|
|
|
|---|

T cells
Most human tumor cell lines from various tissues pulsed with pamidronate for 2 h activated human 
T cells in terms of IFN-
production, proliferation, and specific cytotoxicity to varying degrees, while a minor population failed to do so, with representative data being shown in Table I. Without Ag pulsing, none of them could induce IFN-
production or specific proliferation. In contrast, tumor cell lines from nonhuman species similarly pulsed with pamidronate at any doses totally failed to activate human 
T cells (Table I). As summarized in Table II, none of the 33 tumor cell lines pulsed with pamidronate from 12 different species other than human could activate human 
T cells, whereas a significant level of IFN-
was observed in human 
T cells in response to 21of 26 human tumor cell lines pulsed with pamidronate. An effective Ag-pulsing effect was detected by the incubation with pamidronate for as little as 2 h. After the pulsing with 0.1 mM pamidronate for 2 h, no significant toxic effect was observed as judged by propidium iodide staining in both human and nonhuman tumor cell lines (data not shown), excluding a possibility that the different "pulsing effect" reflected differential sensitivity of each cell line to the toxic metabolic effects by pamidronate. The results strongly suggested that effective presentation of pamidronate to human 
T cells by tumor cells required species-specific cellular interactions.
|
|

T cells by pamidronate-pulsed human tumor cells
We previously reported that pamidronate specifically induced the clustering between human 
T cells and accessory monocytes, suggesting the requirement of direct cellular adhesion for effective 
T cell activation (9). As indicated in Fig. 1A, both primary and activated human 
T cells highly expressed LFA-1 in an active form. On the other hand, human tumor cell lines, which could efficiently present pamidronate to 
T cells, expressed ICAM-1, with representative examples being indicated in Fig. 1B. In contrast, a minor population of human tumor cell lines with no or only weak capacity to activate 
T cells exhibited only marginal LFA-1 ligand expression including ICAM-1, -2, and -3 (Fig. 1B). Although not shown, ICAM expression levels on tumor cells were not affected by pamidronate-pulsing. 
T cells and ICAM-1high EJ-1 tumor cells either mock treated or pulsed with pamidronate formed significant cellular conjugates by 30 min after the mixture (Fig. 2A). The conjugate formation with untreated EJ-1 cells was only transient and dissolved nearly to the background level in 2 h. On the other hand, the cellular conjugates between 
T cells and pamidronate-pulsed EJ-1 cells lasted stably at least for 2 h (Fig. 2A). Inclusion of a blocking anti-LFA-1 mAb in the cocultures of 
T cells and pamidronate-pulsed EJ-1 cells significantly inhibited the conjugate formation and subsequent IFN-
production as well (Fig. 2B). These results strongly suggested that LFA-1/ICAM-1-mediated adhesive interaction with pamidronate-pulsed human tumor cells was one of the requirements for effective activation of 
T cells.
|
|

T cell activation by pamidronate-pulsed human tumor cells
These results raised a possibility that failure of 
T cell activation by a minor population of human tumor cell lines pulsed with pamidronate was due to the insufficient expression of ICAM-1 on the cell surface. To examine this, we transfected a human ICAM-1 cDNA into ICAM-1low LK-2 and MKN45 cells, which induced only weak or no detectable activation in 
T cells (see Fig. 1B and Table I), and established ICAM-1high clones (Fig. 3Aa). A pamidronate-pulsed ICAM-1high LK-2 clone induced a much stronger activation of 
T cells than wild-type cells in terms of both IFN-
production and specific proliferation (Fig. 3A, b-d). The results confirmed that inefficient activation of 
T cells by pamidronate-pulsed LK-2 tumor cells was primarily due to the low expression of ICAM-1. In quite a contrast, an ICAM-1high MKN45 clone pulsed with pamidronate remained defective in activating 
T cells (Fig. 3B). Essentially the same results were obtained by using other independent ICAM-1high MKN45 clones. Thus, it was suggested that LFA-1/ICAM-mediated interaction was required but not sufficient for the effective activation of human 
T cells by tumor cells pulsed with pamidronate.
|

T cells to murine tumor cell lines
We finally examined the effect of human ICAM-1 expression on murine tumor cell lines. A mouse B16 melanoma clone transfected with human ICAM-1 cDNA (hICAM-1high B16) expressed a high level of human ICAM-1 comparable to that of EJ-1 human tumor cells (Fig. 4A). However, hICAM-1high B16 cells pulsed with pamidronate failed to activate human 
T cells with respect to either IFN-
production or specific proliferation (Fig. 4, B and C). In addition, human ICAM-1high B16 cells pulsed with pamidronate remained totally resistant to cytotoxicity elicited by human 
T cells (data not shown). Identical results were obtained using hICAM-1high P3U1 mouse myeloma clones (data not shown). These results strongly suggested that the lack of human 
T cell activation by pamidronate-pulsed tumor cells of nonhuman origins was not only due to the defective LFA-1/ICAM-1-mediated interaction with human 
T cells, but also to the absence of additional human-specific element(s) for the effective activation.
|
| Discussion |
|---|
|
|
|---|

T cells bearing V
2J
1.2V
2-TCR recognize several kinds of small nonpeptide Ags with distinct chemical features (1, 10), but the exact mode of Ag recognition remains unknown. Previous reports indicated that homologous aggregation among 
T cells was required for their activation by some of the nonpeptide Ags (14). More recently, it was reported also that human 
T cells formed stable conjugation with the susceptible tumor target cells with synaptic molecular transfer resembling immunological synapse between 
T cells and specific APC (15). Although these results implied that cell-associated presentation of nonpeptide Ags might be required for 
T cell activation, detailed analysis was hampered by the fact that most of the nonpeptide Ags could efficiently activate 
T cells even in the absence of other specific cells (14). In this aspect, we indicated that the primary human 
T cells failed to respond to soluble pamidronate, but recognized the compound in the presence of accessory monocytes (9). In addition, we also showed that many human tumor cells prepulsed with pamidronate could efficiently activate 
T cells (13). Such a requirement of Ag presentation was a common feature of all of the antigenic nitrogen-containing bisphosphonate molecules (S. Kita, Y. Tanaka, and N. Minato, unpublished observation), which might provide an ideal system for detailed analysis of 
T cell activation by nonpeptide Ags.
The present results indicated that the majority of pamidronate-pulsed human tumor cell lines of varying tissue origins (21of 26 lines) activated 
T cells, whereas none of the tumor cell lines of nonhuman origins could activate human 
T cells (0 of 33 lines). After the pulsing procedure, no significant toxicity was detected in all of the tumor cell lines of both human and nonhuman origins as judged by propidium iodide staining. Thus, it was unlikely that 
T cells were activated simply by pamidronate nonspecifically bound to the target tumor cells. Incidentally, the present results also suggested that most human V
2V
2 
T cells could not recognize xenogeneic MHC Ags, while human 
T cells were shown to recognize xenogenic MHC Ags directly and exhibit comparable levels of responses to allogeneic cells (16).
Human 
T cells highly expressed LFA-1 molecules on their cell surface, and ICAM-1 was detected on most of the human tumor cell lines that could present pamidronate. Then, 
T cells formed tight conjugates with the pamidronate-pulsed tumor cells. Inclusion of a blocking anti-LFA-1 mAb in the cocultures significantly inhibited the conjugate formation, resulting in the inhibition of IFN-
production by 
T cells. The results suggested that LFA-1/ICAM-1-mediated adhesive interaction was required for effective 
T cell activation. Consistently, ICAM-1low LK-2 lung tumor cell line gained a strong capacity of activating 
T cells by ICAM-1 cDNA transfection upon pretreatment with pamidronate. Although cell clusters between 
T cells and mock-treated tumor cells were only transient, 
T cells formed stable conjugates with pamidronate-pulsed tumor cells, lasting for at least 2 h. In 
T cells, it is indicated that LFA-1 is activated and reorganized to form stable immunological synapse with APC via "inside-out" signaling mediated by TCR (17, 18, 19, 20). In analogy, we speculate that the 
TCR signaling via cell-associated pamidronate may induce reorganization of LFA-1 to form stable synapse with the Ag-presenting tumor cells.
The present results further indicated that the integrin-mediated interaction was not sufficient for pamidronate-pulsed tumor cells to activate human 
T cells. Thus, a high level of ICAM-1 expression by cDNA transfection failed to confer the capacity of 
T cell activation to the ICAM-1low MKN45 human tumor cell line. Furthermore, expression of human ICAM-1 in murine tumor cell lines could not induce the capacity to activate human 
T cells at all. These results strongly suggested that additional specific interaction other than integrin-mediated adhesion was required for 
T cell activation by pamidronate-pulsed tumor cells.
Although the exact nature of interaction between 
T cells and pamidronate-presenting cells remains to be investigated, at least two possible models are considered, which are not exclusive to each other. First, although the 
TCR may directly interact with the native nonpeptide Ags presented on the target cells, additional costimulatory signals via specific ligands may be needed for the efficient 
TCR signaling like the B7 family molecules in the case of 
T cell activation (21, 22). Although human 
T cells express a significant level of CD28 (our unpublished data), few tumor cell lines exhibit B7-1 or B7-2 expression, and the candidate molecules remain to be identified. Alternatively, 
TCR may recognize nitrogen-containing bisphosphonate Ags only in association with specific molecules on the Ag-presenting tumor cells. Our unpublished results indicated that, in addition to the J
1.2-encoded lysine residues (11), at least two more additional residues in the putative Ag-binding pocket in V
2J
1.2V
2 TCR were required for 
T cell activation by diverse nonpeptide Ags (S. Yamashita, Y. Tanaka, and N. Minato, manuscript in preparation). Thus, it is possible that nonpeptide Ags along with specific Ag-presenting molecules optimally fit to the pocket via anchoring residues in much the same way as peptide Ags in complex with MHC molecules for 
TCR. A number of previous studies provided no evidence for involvement of MHC in 
T cell activation by nonpeptide Ags (14, 23, 24) and again the candidate molecules remain to be seen.
The present results strongly support the hypothesis that effective activation of human 
T cells by nonpeptide Ags depends on the immunological synapse formation in which species-specific interactions take place. Nitrogen-containing bisphosphonate Ags are particularly intriguing, in that the vast majority of human tumor cells can efficiently present them to 
T cells, resulting in the targeted lysis of tumor cells by 
T cells. It is suggested that the putative molecules required for Ag presentation and effective 
T cell activation are rather ubiquitously expressed on many human tumor cells. Molecular understanding of the interaction may provide a new clue for potential immunotherapy of human malignancy by targeting tumor cells with chemically defined nonpeptide Ags for 
T cells.
| Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Nagahiro Minato, Department of Immunology and Cell Biology, Graduate School of Biostudies, Kyoto University, Yoshida-Konoe-Cho, Sakyo-ku, Kyoto 606-8501, Japan. E-mail address: minato{at}imm.med.kyoto-u.ac.jp ![]()
3 Abbreviation used in this paper: h, human. ![]()
Received for publication November 25, 2002. Accepted for publication January 31, 2003.
| References |
|---|
|
|
|---|

cells: a right time and a right place for a conserved third way of protection. Annu. Rev. Immunol. 18:975.[Medline]

T cells. Proc. Natl. Acad. Sci. USA 91:8175.
T cells by nonpeptidic mycobacterial ligands. Science 264:267.
T cells. Nature 375:155.[Medline]

T cells. J. Biol. Chem. 274:32079.
T cells recognize alkylamines derived from microbes, edible plants, and tea: implications for innate immunity. Immunity 11:57.[Medline]

T-cell stimulation by pamidronate. N. Engl. J. Med. 340:737.
T cells by aminobisphosphonates and induction of antiplasma cell activity in multiple myeloma. Blood 96:384.
T cells by aminobisphosphonate antigen. J. Immunol. 166:5508.
T cells. FASEB J. 14:1669.
1.2 segment to the recognition of nonpeptide antigens by human 
T cells. J. Immunol. 167:6773.
T-cell antigen receptor. Nature 411:820.[Medline]

T cells by nonpeptide antigens. J. Immunol. 167:5092.
T cells. Immunity 3:495.[Medline]

T cells stimulated with soluble or cellular antigens. J. Immunol. 168:6336.
T cells. Int. Immunol. 5:1437.
9V
2 T cell broad cytotoxicity and TNF production by nonpeptidic mycobacterial ligands. J. Immunol. 154:5986.[Abstract]This article has been cited by other articles:
![]() |
M. Takahara, M. Miyai, M. Tomiyama, M. Mutou, A. J. Nicol, and M. Nieda Copulsing tumor antigen-pulsed dendritic cells with zoledronate efficiently enhance the expansion of tumor antigen-specific CD8+ T cells via V{gamma}9{gamma}{delta} T cell activation J. Leukoc. Biol., March 1, 2008; 83(3): 742 - 754. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Thompson and M. J. Rogers Bisphosphonates and {gamma}{delta} T-Cells: New Insights into Old Drugs IBMS BoneKEy, August 1, 2006; 3(8): 5 - 13. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. O. Deetz, A. M. Hebbeler, N. A. Propp, C. Cairo, I. Tikhonov, and C. D. Pauza Gamma Interferon Secretion by Human V{gamma}2V{delta}2 T Cells after Stimulation with Antibody against the T-Cell Receptor plus the Toll-Like Receptor 2 Agonist Pam3Cys. Infect. Immun., August 1, 2006; 74(8): 4505 - 4511. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Kato, Y. Tanaka, M. Hayashi, K. Okawa, and N. Minato Involvement of CD166 in the Activation of Human {gamma}{delta}T Cells by Tumor Cells Sensitized with Nonpeptide Antigens J. Immunol., July 15, 2006; 177(2): 877 - 884. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Corvaisier, A. Moreau-Aubry, E. Diez, J. Bennouna, J.-F. Mosnier, E. Scotet, M. Bonneville, and F. Jotereau V{gamma}9V{delta}2 T Cell Response to Colon Carcinoma Cells J. Immunol., October 15, 2005; 175(8): 5481 - 5488. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Rincon-Orozco, V. Kunzmann, P. Wrobel, D. Kabelitz, A. Steinle, and T. Herrmann Activation of V{gamma}9V{delta}2 T Cells by NKG2D J. Immunol., August 15, 2005; 175(4): 2144 - 2151. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Poupot, F. Pont, and J.-J. Fournie Profiling Blood Lymphocyte Interactions with Cancer Cells Uncovers the Innate Reactivity of Human {gamma}{delta} T Cells to Anaplastic Large Cell Lymphoma J. Immunol., February 1, 2005; 174(3): 1717 - 1722. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Conti, R. Casetti, M. Cardone, B. Varano, A. Martino, F. Belardelli, F. Poccia, and S. Gessani Reciprocal Activating Interaction Between Dendritic Cells and Pamidronate-Stimulated {gamma}{delta} T Cells: Role of CD86 and Inflammatory Cytokines J. Immunol., January 1, 2005; 174(1): 252 - 260. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yamashita, Y. Tanaka, M. Harazaki, B. Mikami, and N. Minato Recognition mechanism of non-peptide antigens by human {gamma}{delta} T cells Int. Immunol., November 1, 2003; 15(11): 1301 - 1307. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |