Synthetic Phosphoantigens Enhance Human Vγ9Vδ2 T Lymphocytes Killing of Non-Hodgkin’s B Lymphoma

BackgroundNon-Hodgkin’s B lymphomas (NHL) are often resistant to conventional treatments and, until now, immunotherapeutic approaches against NHL only aimed at inducing αβ anti-tumor effectors. Nevertheless, human blood Vγ9Vδ2 T lymphocytes represent an abundant pool of cytotoxic tumor-reactive cells. Vγ9Vδ2 T cells are strongly activated by natural compounds, from which powerful synthetic ligands have been derived. These synthetic antigens induce efficient Vγ9Vδ2 T cell responses in vitro.Materials and MethodsWe set up a series of Vγ9Vδ2 T cell-activation experiments, including cytotoxic activity and amplification from whole blood cells. Several types of Vγ9Vδ2 effectors were challenged against a panel of 16 B lymphoma cell lines. These tests have been performed in the absence and presence of γδ-specific synthetic ligands to evaluate the effect of such molecules on γδ anti-tumor activity.ResultsWe report here that Vγ9Vδ2 T cells recognize B lymphomas. This recognition is associated with the cytotoxic activity against B-lymphoma cells and/or proliferative responses, and appears to be T-cell antigen receptor (TCR)-dependent. Because few B lymphoma induce a complete set of Vγ9Vδ2 cell responses, a chemical ligand of Vγ9Vδ2 T cells was used to enhance both proliferation and cytotoxic activity of anti-B lymphoma effectors. We show that such synthetic compound improves Vγ9Vδ2 CTL numbers and lysis of B lymphoma lines, especially when the targets are already spontaneously recognized by these effectors.ConclusionsWe report here that human Vγ9Vδ2 T cells anti-B lymphoma response can be improved by use of specific synthetic ligands, which enhance their cytotoxic activity and allows their rapid expansion ex vivo.


Introduction
Non-Hodgkin lymphomas (NHL) are lymphoproliferative disorders developing from B, T, or, rarely, natural killer (NK) cells. B cell NHL arise from the clonal expansion of a B cell developmentally blocked at virtually any stage of maturation (1). Increasing evidence suggests that a significant proportion of NHL B cells remain resistant to conventional chemotherapy (2)(3)(4)(5)(6)(7). Despite their frequent infiltration by CD4 ϩ and CD8 ϩ T cells, B cell NHL rarely induce clinically significant T-cell-mediated responses (8)(9)(10). Clinical data suggest that it is partly due to the low frequency of tumor-infiltrating lymphocytes (TILs) and to their insufficient activated state in vivo [for a recent review see Schultze (11)]. Hence, autologous cytotoxic anti-NHL-specific T lymphocytes can be generated and expanded in vitro solely under very specific conditions, requiring cytokine-enriched media (12,13). However, such effectors do not always acquire significant anti-tumor cytotoxic activity (14). Because use of animal models has demonstrated the essential role of T cells in tumor rejection, most recent immunotherapeutic approaches aim at improving the in vivo activation of cytotoxic CD8 ϩ T lymphocytes (CTL) [for a review see Schultze and Nadler (15)]. Vaccination with B-cell NHL-associated isotype (the tumor's most specific antigenic determinant) is one of the most studied strategies so far (16)(17)(18)(19)(20). Unfortunately, despite significant improvement, this approach still often shows uneven and unconvincing clinical efficacy (21). Therefore, there is an obvious need for characterizing CTL populations with anti-tumor activity against B-cell NHL, as well as for defining simple approaches to amplify such anti-NHL-specific effectors.
In this study, we questioned the ability of synthetic phosphoantigens to improve the anti-B lymphoma activity of V␥9V␦2 T effectors. We provide evidence that ␥␦-specific synthetic ligands could constitute an efficient and convenient tool to enhance the anti-B lymphoma response of human V␥9V␦2 T cells to be tested in future immunotherapeutic approaches.

Purification of Peripheral Blood Mononuclear Cells (PBMC)
Peripheral blood mononuclear cells (PBMC) were prepared from blood from healthy volunteers by centrifugation on Ficoll-Hypaque (Amersham Pharmacia Biotech, Uppsala, Sweden) according to the manufacturer's instructions.

Polyclonal V␥9V␦2 CTL Kill B Cell NHL Lines In Vitro
The in vitro cytotoxicity of two unrelated primary V␥9V␦2 cell lines derived from two different healthy donors was evaluated against 15 non-Hodgkin's B lymphoma cell lines and one Hodgkin's diseasederived B-cell lymphoma line. These B-cell lymphomas were selected such as to represent various stages of B-cell differentiation (see Table 1). The percentages of specific lysis of each polyclonal V␥9V␦2 cells against the panel of NHL are presented in Figure 1. Although overall cytotoxicity of the first polyclonal V␥9V␦2 CTL line is reproducibly lower than that of the second, both polyclonal CTL lines exert comparable levels of cytotoxic response against each individual target. For any tested CTL line, the level of spontaneous cytotoxicity is heterogeneous with regard to the target: seven NHL lines are not spontaneously lysed by V␥9V␦2 CTL in vitro (below 10% specific lysis, see Fig. 1, bottom). Six NHL cell lines spontaneously trigger intermediate levels of killing by V␥9V␦2 cells (10-40% specific lysis: PASC, HLY-1, BL9, RPMI8226, VAL, and DG75) and two NHL lines and the Hodgkin's lymphoma line
spontaneously activate a high level (above 40% of specific lysis; Fig. 1, top) of specific lysis by ␥␦ effectors.
Cytolytic activity of V␥9V␦2 CTL for tumor cells is known to be influenced by a deficit in HLA class I molecule expression at the surface of the targets (32,48,49,52), accounting for an NK-like ␥␦mediated killing of the HLA-class I Ϫ targets K562 and Daudi (32,45,50,51). To rule out such a possibility, we evaluated the level of surface expression of HLA class I molecules on the target lymphoma lines with the W6/32 mAb directed against HLA-A, -B, -C, -E, and -G (Fig. 1). Whereas NHL lines differ in terms of intensity of HLA class I expression, Figure 1 shows that the specific lysis of these targets by V␥9V␦2 CTL does not match to their relative deficit in surface expression of HLA class I molecules. The HLA-class I low lymphoma REC1 and PASC are not killed, whereas several HLA-class I bright lines (OCI-Ly8, DG75, RPMI8226, and VAL) are efficiently lysed (Fig. 1). Thus, in contrast to K562 and Daudi cell lines, the OCI-Ly8, VAL, DG75, BL9, RPMI8226, and HLY-1 NHL are killed by V␥9V␦2 CTL solely through a non-NK-like pathway.
Furthermore, V␥9V␦2 T cells spontaneous cytotoxicity toward these NHL B cell lines is not specifically restricted to Epstein-Barr Virus (EBV)-positive lymphomas (Fig. 1).
Because several B cell lymphoma lines from the above panel activate ␥␦-selective cytotoxicity, we asked whether these irradiated lymphoma could also induce a selective outgrowth of polyclonal V␥9V␦2 T lymphocytes in vitro when co-cultured with freshly purified PBMC from several healthy donors. As expected, Daudi and RPMI8226 cell lines induce expansion of V␥9V␦2 T cells from PBMC (Table 2). Similarly, Hodgkin's lymphoma line L428 triggers both high V␥9V␦2 lysis and V␥9V␦2 CTL expansion from PBMC of two on three donors (Table 2). Interestingly, the NHL line DEAU, which does not induce ␥␦ cytotoxicity, promotes V␥9V␦2 T cell outgrowth from PBMC in two of three donors. None of the other NHL lines tested expand V␥9V␦2 CTL from any PBMC tested, although some of them, like OCI-Ly8, DG75, and VAL, trigger strong V␥9V␦2 T cell cytotoxic activity. Thus, of several NHL B-cell lines activating V␥9V␦2 T cells, few induce both cytotoxicity and amplification responses.
B lymphoma cell lines involved in this study were chosen such as to stretch along the whole B-cell differentiation process (Table 1). This panel comprises B-cell malignancies starting from pregerminal centers (pre-GC) cells, GC/post-GC cells up to immunoglobulin (Ig)-producing cells. Looking for a possible correlate between ␥␦-activating phenotype and B-cell differentiation, we compared the ␥␦-stimulating properties of these 16 B lymphoma and their respective stage of differentiation. This comparison suggested that ␥␦-activating B malignancies span all along B-cell maturation.

Specificity of Spontaneous V␥9V␦2 CTL Cytotoxicity to NHL B-Cell Lines
The unusual pattern of V␥9V␦2 CTL responses when exposed to NHL B-cell lines questioned the nature of the V␥9V␦2 activation pathway by NHL. Using OCI-Ly8 as a model of V␥9V␦2 CTL-activating NHL, we investigated some characteristics of its killing pathway and the involvement of the V␥9V␦2 TCR. As shown in Figure 1, the HLA-class I bright OCI-Ly8 NHL is efficiently and spontaneously killed by V␥9V␦2 T lymphocytes in vitro but fails to induce V␥9V␦2 CTL expansion from primary PBMC cultures.  Table 2. dependent release of perforin (and other soluble mediators) from exocytosis granules. When added to the in vitro mix of OCI-Ly8 and V␥9V␦2 CTL, EGTA totally suppresses killing of OCI-Ly8 by V␥9V␦2 T cells (Fig. 2B). In contrast to earlier reports (59,60), in our experiments V␥9V␦2 T cells kill Jurkat targets in a strictly Ca 2ϩ -dependent way (Fig. 2B). EGTA has, however, no effect on the lysis of Jurkat cells induced by the Fas-agonist (CH11 mAb, Fig. 2B). Taken together, these results demonstrate that OCI-Ly8 lysis by V␥9V␦2 T lymphocytes results exclusively from a release of granzyme-perforine but not from NHL sensibility to Fas-induced apoptosis, thereby suggesting that OCI-Ly8 activates V␥9V␦2 T lymphocytes in a TCR-dependent fashion.
To address this point more directly, OCI-Ly8 and V␥9V␦2 CTL were co-incubated in the presence of increasing quantities of the anti-TCR V␦2 mAb Immu389 (61). Figure 3 shows that this Although Fas-mediated cytotoxicity is assumed to be negligible in 4-hr chromium release assay (53,54), we asked whether the strong lysis of OCI-Ly8 by V␥9V␦2 CTL relies on a marked sensitivity of this NHL line to Fas-L. For this purpose, OCI-Ly8 killing by V␥9V␦2 CTL was tested in presence of the Fas-agonist CH11 mAb (55), which induces the apoptotic death of Fas ϩ -Jurkat cells ( Fig. 2A), or conversely in presence of the Fas-antagonist ZB4 mAb (56), which inhibits CH11-induced apoptosis ( Fig. 2A). In these experiments, killing of Fas-Daudi or of OCI-Ly8 by V␥9V␦2 T lymphocytes was not altered by the Fas antagonist ( Fig. 2A).
Because TCR-mediated activation of CTL usually leads to perforin release [for a review see Shresta et al. (57)], we tested whether OCI-Ly8 line killing involves the release of perforin by V␥9V␦2 T cells. Although it does not alter the Fas-mediated T-cell cytotoxicity (58), EGTA inhibits the calcium-  antibody strongly blocks lysis of OCI-Ly8 as well as the target RPMI8226 by polyclonal V␥9V␦2 CTL. The anti-TCR V␦2 mAb only partially inhibits lysis of HLA-class I Ϫ Daudi lymphoma, which results both from TCR-mediated and NKlike lysis by V␥9V␦2 T cells. As reported (49,50), the anti-TCR V␦2 mAb Immu389 does not inhibit lysis of HLA-class I Ϫ K562 tumor cells (Fig. 3A).
In line with these results, while mAb directed against TCR V␥9 chains reduces the lysis of OCI-Ly8 by V␥9V␦2 T cells, mAbs against CD4 and CD8 seldom expressed on polyclonal ␥␦ cells have no effect on this lysis (Fig. 3B). Taken together, these results support the idea that the cytotoxicity of V␥9V␦2 CTL for NHL target OCI-Ly8 is mediated by the ␥9␦2-TCR. Upon antigen recognition, the TCR mediates initial steps of T-cell activation, ultimately followed by functional T-cell responses. Because ␥9␦2-TCR-mediated recognition of OCI-Ly8 NHL drives further cytotoxic responses, we asked whether NHL stimulation of V␥9V␦2 CTL induces early appearance of activation markers on these effectors. Antigenic activation of T cells induces the surface expression of specific markers, of which CD69 is one of the earliest [for recent reviews, see Tough 2t al. (62) and Marzio et al. (63)]. Thus, following exposure to different tumor cell lines, we tested the induction of CD69 at the surface of freshly derived V␥9V␦2 T lymphocytes. As compared to unstimulated ␥␦ cells alone (negative control, Fig. 4) and PMA-treated ␥␦ cells (positive control, Fig. 4), Daudi, OCI-Ly8 NHL, or RPMI8226 activate CD69 expression (Fig. 4). Conversely, neither of the ␥␦ unstimulatory Jurkat and K562 tumor cells do so (Fig. 4). In these experiments, NHL line OCI-Ly8 induces clear-cut expression of the activation marker CD69 at the surface of polyclonal V␥9V␦2 T lymphocytes as early as 8 hr after co-culture, witnessing early induction by a V␥9V␦2 TCR-mediated activation (Fig. 4).
Taken together, these results demonstrate that although they do not necessarily induce ␥␦ T-cell proliferation, B lymphoma lines may be specifically

Fig. 4. Expression of the early activation surface marker CD69 by polyclonal V␥9V␦2 T-cell line after exposure to different cells.
A freshly derived polyclonal V␥9V␦2 T cells population was exposed to several tumor cells in vitro for 8 hr. CD69-surface expression on V␥9V␦2 T cells was then monitored by FACS analysis. Stimulating lymphoma cells were excluded from analysis by BCECF-staining prior to incubation with the effectors and do not appear here. "Controls" panel shows the staining of V␥9V␦2 T cells with IgG2b-RD1 isotype antibody (gray line) and with PE-conjugated anti-CD69 antibody after incubation with medium alone. preactivated ␥␦ primary cell lines, we questioned the relevance of such a bioactivity for B lymphoma targets within freshly drawn ␥␦ T cells. Thus, the effect of PHD stimulation on anti-NHL V␥9V␦2 fresh CTL was tested by setting cytotoxicity experiments with freshly prepared PBMC effectors. Primary PBMC from two healthy donors with distinct V␥9V␦2 T-cell proportions were tested extemporaneously in a 4-hr chromium release assay, for their recognized by V␥9V␦2 TCR and consequently be killed by these effector cells.

Drug-Induced Amplification of V␥9V␦2 CTL Cell Numbers and Anti-NHL Cytotoxicity
Because ␥␦-stimulating NHL B-cell lines fail to induce a complete set of V␥9V␦2 CTL responses in vitro ( Fig. 1 and Table 2), this limits the clinical potential of spontaneous activated ␥␦ cells. However, novel drugs have been recently designed that trigger the complete set of V␥9V␦2 T cell responses in vitro (41). This prompted us to test whether such synthetic drugs could circumvent the absence of expansion of these ␥␦ effectors while maintaining (or enhancing) their anti-NHL cytotoxicity.
To compensate for the absence of amplification of V␥9V␦2 effectors in response to target B lymphomas (Table 2), we used synthetic drugs to generate high quantities of anti-B lymphoma V␥9V␦2 CTL in vitro. These drugs (39,41) specifically mimic natural V␥9V␦2-T-cell ligands [referred to as phosphoantigens (27,(35)(36)(37)64,65)]. Increasing concentrations of the drug PHD (41) were added to primary PBMC cultures from four healthy donors (Fig. 5). There was a strong increase in V␥9V␦2 T-cell numbers from PBMC of all donors, reaching more than 30-fold (Fig. 5A). This amplification is dose dependent, as shown for donor 1 (Fig. 5B). Finally, Figure  5C shows the maximal percentage of V␥9V␦2 T cells obtained for the three other donors.
Thus, although B-lymphoma lines as stimuli may sometimes fail to amplify specific CTL in vitro, these effectors can nevertheless be conveniently amplified in vitro using synthetic drugs. Because some NHL B-cell lines appear resistant to V␥9V␦2 CTL lysis in vitro, we also investigated the ability of PHD to improve the cytolytic activity of ␥␦ effectors. Here, we analyzed the effect of this drug on the cytotoxic activity of two V␥9V␦2 T-cell lines against the panel of B lymphomas. A final concentration of 20 nM PHD was added in these cytotoxic assays, where the effector-to-target (E/T) ratio decreases from 20/1 to 2.5/1. Figure 6 shows the effect of PHD added to the anti-NHL cytotoxicity of V␥9V␦2 CTL. PHD effect on NHL killing by ␥␦ effectors is heterogeneous, as observed above for their spontaneous anti-NHL cytotoxicity. On the one hand, PHD improves slightly the killing of genuine ␥␦-activating targets such as Daudi, OCI-Ly8, L428, DG754, or VAL (Fig. 6). On the other hand, PHD does not elicit killing of unstimulatory NHL lines (Raji, MIEUL, LIB, or REC1, Fig. 6). Interestingly, PHD increases by about 10-fold the anti-NHL cytolytic potential of V␥9V␦2 CTL against some targets (see lysis of HLY-1, NCI-H 929, and RL in Fig. 6). PHD appears most efficient in enhancing V␥9V␦2 T-cell cytotoxicity toward partially activating NHL B-cell lines.
Because the PHD-induced increase in ␥␦ anti-NHL activity had only been challenged on long-term lysis of the preceding panel of human B lymphomas (Fig. 7). As already observed with long-term polyclonal ␥␦-cell lines (see Fig. 1), freshly prepared PBMC from the two different donors spontaneously exert anti-NHL cytotoxicity, although of different intensity according to the target. Adding PHD (final concentration 200 and 800 nM, without IL-2) to such CTL assays gave roughly a similar pattern of effect as formerly observed using PHD-amplified ␥␦ effectors (Fig. 7). Killing of genuine B lymphoma targets and of stimulating NHL is significantly improved, with a PHD dose effect (see Daudi, OCI-Ly8, L428, DG75, VAL, BL9, Fig. 7), while resistant NHL remain unaffected (see HLY-1, DEAU, MIEUL and Raji, Fig. 7). In these experiments, the PHD effect was not greatly influenced by the ratio of PHD-reactive V␥9V␦2 T cells among PBMC (e.g., the VAL NHL line is killed similarly by PHD-treated PBMC from donor 1 and donor 2, Fig. 7). Furthermore, PHD exposure of freshly prepared PBMC induces a low level of killing of some otherwise resistant cell lines (see NCI-H 929, RL or LIB, Fig. 7).

Discussion
This in vitro study aimed at documenting the cytotoxic potential of human V␥9V␦2 T lymphocytes against non-Hodgkin's B-cell lymphoma lines and the ability of ␥␦ synthetic ligands to improve this anti-B lymphoma activity. We show that this CTL population is spontaneously activated to kill several NHL B-cell lines in vitro. As judged by HLA class I molecule expression at the surface of the targets, this cytotoxicity does not result from an NK-like lysis, but most probably arises from a specific TCR-mediated stimulation. The features of the killing mechanism of the target NHL line OCI-Ly8 by V␥9V␦2 CTL confirms their specific activation. Thus, human V␥9V␦2 T cells represent potential anti-NHL CTL, of high frequency among circulating T cells in blood of healthy donors (approximately 1-10%) (22).
Nevertheless, the activation of V␥9V␦2 CTL by B lymphoma cell lines does not necessarily lead to a  Immunotherapeutic trials against NHL aim at eliciting specific cytotoxic T-cell responses against these cancer cells (21). Generally, the candidate effectors are ␣␤ CD8 ϩ T lymphocytes, because they can be manipulated by two distinct approaches. These effectors are either stimulated in vivo after vaccination with a tumor-specific antigenic determinant (16,17,19,20) or they are expanded ex vivo as CTL against the autologous tumor (12,13,69). The first approach has improved through monitoring the acquired immunity and a better detection of the residual disease. Nevertheless, vaccination still requires further improvement for its generalization in anti-lymphoma protection (21). More specifically, cancer vaccination would benefit from the identification of novel specific tumor antigens (74). Ex vivo generation of specific autologous CTL has proven difficult for the relatively low TILs frequency, their poor intrinsic cytotoxic activity, and their sophisticated culture conditions (12,13,69,75,76). Therefore, a need for the identification of novel effectors of the anti-lymphoma immune response still remains. This study supports new options for the design of antitumor cellular immunotherapies. As yet, however, no report describes autologous V␥9V␦2 T lymphocytes as NHL-TILs in vivo (77,78) or as CTL amplified in vitro from TILs (13); our study indicates that reactive V␥9V␦2 CTL against B lymphomas can be readily generated in vitro. In this context, human V␥9V␦2 T lymphocytes offer several advantages as cellular effectors as compared to ␣␤ cytotoxic T cells. Above all, whereas the usual frequency in blood of almost any ␣␤ CTL is below 0.01%, that of V␥9V␦2 CTL is quite higher, being 1-10% in adults (22). Furthermore, polyclonal V␥9V␦2 CTL cell lines expand within a few days after stimulation of PBMC with specific ligands (35,36,51) (Fig. 5). As effectors of innate immunity, V␥9V␦2 T lymphocytes acquire cytotoxic activity against tumor target without former exposure (79), whereas alloreactive CD8 ϩ CTL have to be primed to become efficient responders to NHL cells (12,66,68,70). Furthermore, once generated from PBMC through phosphoantigenic stimulation, V␥9V␦2 T lymphocytes simultaneously acquire responsiveness to several distinct target cells (34,50,79) (this study). Thus, generating autologous activated ␥␦ T cells is far simpler than expanding ␣␤ CTL against NHL cells in the presence of the patient's tumor cells (12,13). Finally, the recent development of chemical ligands for V␥9V␦2 T cells even reinforces their interest for the design of future antilymphoma immunotherapeutic tests. The synthesis of chemical ligands is easier and less expensive than the purification of natural phosphoantigens from microbial sources. Moreover, for obvious safety reasons, the use of synthetic ligands is easier to control than that of extracts from pathogens such as M. tuberculosis.
Another advantage of the V␥9V␦2 CTL population is that its reactivity and functionality is not MHC restricted (29,41,44,(79)(80)(81). On the other hand, complete set of T-cell activation phenotypes. The NHL DEAU cell line induces their specific amplification from primary PBMC while failing to activate V␥9V␦2 T-cell cytotoxicity. Furthermore, only three of nine B lymphoma activating V␥9V␦2 T-cell cytotoxicities also promote V␥9V␦2 CTL expansion. In this respect, it has often been reported that NHL cells hardly stimulate ␣␤ CTL proliferation in vitro unless expression of several costimulatory molecules is induced at their surface (66)(67)(68)(69)(70)(71)(72).
Absence of V␥9V␦2 CTL amplification following contact with NHL can be conveniently overcome using drugs specific for the ␥␦-TCR. Anti-NHL cytotoxic V␥9V␦2 T lymphocytes expand upon stimulation with natural phosphoantigens such as 3fbPP (35,64) or synthetic analogs such as PHD. This amplification requires IL-2 and very low concentrations of the drug PHD (5-50 nM) within 10 days. Moreover, these molecules can also significantly enhance the anti-NHL cytotoxic activity of both polyclonal V␥9V␦2 T-cell lines and freshly prepared V␥9V␦2 T cells within PBMC. The drug-induced improvement of anti-B lymphoma activity is particularly significant toward targets that otherwise spontaneously induce a low level of V␥9V␦2 T-cell cytotoxicity. Variable levels of basal B lymphoma lysis by PBMC CTL are observed within donors (Fig. 7). It is assumed that V␥9V␦2 T cells account for some of this initial cytotoxicity. However, PHD-dependent increase in B lymphoma lysis is solely mediated by responding V␥9V␦2 T cells; natural phosphoantigens and their synthetic counterparts do not stimulate other cellular effectors (35,41,73) (Fig. 5). Interestingly, PHD not only improves the lysis of ␥␦-sensitive B lymphoma lines, but also confers low cytotoxicity to V␥9V␦2 CTL toward usually resistant cell lines. Unfortunately, one third of the NHL lines tested are not killed, even by effectors stimulated with their specific ligand. We assume that intrinsic NHL-resistance to granzyme-perforinemediated lysis could account for such resistance.
The panel of B-cell lymphoma involved in this study was chosen because it comprises malignant counterparts of B cells at various stages of differentiation (Table 1). Thus, we questioned the existence of a correlation between the ␥␦ stimulation property and the level of B-cell target differentiation. However, because ␥␦-stimulating B cells span all along the different maturation steps, we could not link the V␥9V␦2 T-cell-stimulating property to the B-cell differentiation stage. This finding is in agreement with the assumed ubiquity of the B lymphoma ligands of V␥9V␦2 T cells (32), which remain unknown. Nevertheless, with regard to the nature of these antigens, this study suggests that V␥9V␦2 tumor antigens are most probably not related to (1) maturation-dependent B cell markers, (2) B-cellactivation specific molecules, or (3) B-cell receptor (BCR)-expression and Ig secretion. Thus, one may consider V␥9V␦2 CTL as potential effectors of almost any B malignancy.
weak ␣␤ T-cell response against B lymphoma is partly due to the poor antigen-presenting cell (APC) function of the tumor cells (66,70,82). Hence, V␥9V␦2 CTL are interesting effectors because their cytotoxic response does not depend on a classical presentation of tumor antigens, also circumventing the need for improving ex vivo the APC function of dendritic cells with tumor antigens (83)(84)(85)(86). In conclusion, this study suggests that the V␥9V␦2 population of CTL should be considered as a potential pool of anti-NHL effectors, for which novel powerful stimulating drugs are now available.