Different Roles of the CD2 and LFA-1 T-Cell Co-receptors for Regulating Cytotoxic, Proliferative, and Cytokine Responses of Human Vγ9/Vδ2 T Cells

BackgroundHuman Vγ9/Vδ2 T lymphocytes recognize nonpeptidic antigens in a manner distinct from the classical antigen recognition by aβ T cells. The apparent lack of major histocompatibility (MHC) restriction and antigen processing allows very fast responses against pathogenic insults. To address the potential functional requirement for accessory molecules, we investigated the roles of the CD2 and lymphocyte function-associated antigen (LFA)-1 T-cell co-receptors in antigen-induced activities of human Vγ9/Vδ2 T-cell clones.Materials and MethodsHuman peripheral blood Vγ9/Vδ2 T lymphocytes were cloned and their cytotoxicity against Daudi lymphoma was measured by a standard 51Cr-release assay. The responses of Vγ9/Vδ2 T lymphocytes to nonpeptidic antigens were assessed using DNA synthesis and cytokine ELISA assays. Monoclonal antibodies specific for various molecules with potential T-cell accessory functions were utilized in blocking assays.ResultsAll of our Vγ9/Vδ2 T-cell clones displayed the Th1 phenotype. The anti-LFA-1 antibody strongly inhibited the cytotoxicity of Vγ9/Vδ2 T cells against Daudi B-cell lymphoma; whereas, it had no influence on the antigen-induced cytokine release or proliferation. In contrast, antibodies against CD2 and LFA-3 had no effect on the lytic activity of Vγ9/Vδ2 T cells, but strongly inhibited the cytokine release and proliferation. However, the CD2-LFA-3 interaction was not an absolute requirement for the cytokine release and the DNA synthetic activity of antigen-stimulated Vγ9/Vδ2 T cells, since the inhibitory effect could be reversed by addition of exogenous interleukin 2 (IL-2).ConclusionsThese novel observations indicate that the signals generated by different accessory molecules and IL-2 can contribute in an integrated fashion to the regulation of Vγ9/Vδ2 T cells. These interactions may be important for the effectiveness of Vγ9/Vδ2 T-cell responses.

(2) and alkylamines (3). The recognition of nonpeptidic antigens is mediated through the VͲ9/Vͳ2 TCR (4). The structure of these antigens is very different from that of the major histocompatibility (MHC)-peptide complex (5) and, consequently, the serial trigger mode of activation of T cells may not be applicable to the VͲ9/Vͳ2 subset (6). The recognition of peptide/MHC complex by the TCR is a key element in the activation of Ͱͱ T cells (7). However, the crosslinking of the Ͱͱ TCRs alone appears to be insufficient for the full development of T-cell response (8) and many accessory molecules that contribute to the T-cell activation process have been identified (9).
Lymphocyte function-associated antigen (LFA)-1, a member of the integrin family, is expressed on T cells, B cells, granulocytes, and macrophages (10). Studies with Ͱͱ T cells show that LFA-1 can mediate cell-to-cell adhesion and stimulate various intracellular processes (10). LFA-1 is important for target-cytotoxic Ͱͱ T-cell interactions (11). LFA-1 can also affect other cellular functions of Ͱͱ T-cells, such as apoptosis, proliferation, cytokine production and antigen presentation (10). CD2 belongs to the immunoglobulin superfamily and is expressed on all subsets of T lymphocytes, natural killer (NK) and lymphokine-activated killer (LAK) cells (12). The CD2 ligand, LFA-3, is expressed on human lymphoid cells as a transmembrane-integrated form or phosphoinositol-linked form (13,14). The CD2-LFA-3 interaction delivers an important costimulatory signal and strengthens the adhesion between interacting cells (10,15). It has been demonstrated that antibodies against CD2 can inhibit the antigen-induced proliferation of Ͱͱ T cells (10) and the cytotoxicity of CD8 Ͱͱ T cells (15).
The signaling through the CD2 molecule is dependent on the TCR-associated chain (16). However, there are interesting differences between Ͱͱ T cells and Ͳͳ T cells in this regard. For example, at least two monoclonal antibodies specific for different epitopes of CD2 are necessary to induce the activation of Ͱͱ T cells (16); whereas, a single CD2 antibody is enough to activate Ͳͳ T cells (17). Moreover, the murine Ͳͳ TCR associates with a distinct member of the family, the Ͳ chain of the FcRI (reviewed in ref. 18) and, therefore, signaling through CD2 may operate differently in Ͳͳ and Ͱͱ T cells. However, the situation in primates and other species is unclear.
Human VͲ9/Vͳ2 T cells recognize a broad spectrum of nonpeptidic antigens (5,19). This recognition requires neither antigen processing nor the expression of MHC or MHCrelated molecules (20). Most VͲ9/Vͳ2 T cells can respond to nonpeptidic antigens in the absence of antigen-presenting cells (APCs). However, the presence of APCs can greatly enhance the VͲ9/Vͳ2 T-cell response (20). This suggests that accessory molecules may be involved in these responses. Here we show for the first time that LFA-1 and CD2 have different roles in regulating the VͲ9/Vͳ2 T-cell activities.

Cells and Cell Cultures
Cells were cultured in RPMI 1640 medium supplemented with 2 mM L-glutamine, 100 IU/ml penicillin, 100 Ȑg/ml streptomycin, and 10% fetal bovine serum. The T-cell clones were generated using the method described elsewhere (21). The clonality was confirmed by gene scan analysis of CDR3 polymorphism (data not shown).

TNF-Ͱ Assay
10 4 responder cells were stimulated with the indicated concentrations of isopentenyl pyrophosphate (IPP) purchased from Sigma (Harz, Germany). The supernatants were collected 6 hr later. The concentration of tumor necrosis factor (TNF)-Ͱ was determined using the ELISA assay kit from R&D Systems (Minneapolis, MN) following the manufacturer's instructions. The blocking or control antibodies were added to the culture at the initiation of antigen exposure.

Proliferation Assay
Proliferation assay was performed in 96-well round bottom plates using 10 5 /well responder cells and 10 5 /well irradiated LCL721 cells (11,000 rads, 137 Cs source) as feeders. 48 hr after IPP stimulation, 1 ȐCi/well of [methyl-3 H] thymidine (2.6-3.2 TBq/mM, Amersham Pharmacia Biotech, Piscataway, NJ) was added, the cells were harvested after additional 18 hr and DNA synthesis was measured as described (23). The blocking or control antibodies were added to the culture at the initiation of antigen exposure. To assess the produced IL-2 levels, the supernatants were diluted 1 : 4 with medium, cultured with CTLL-2 cells and their DNA synthetic response was measured as described (23).

Cytotoxicity Assay
Daudi cells were labeled with 100 ȐCi Na 2 [ 51 Cr]O 4 (7.4-18.5 GBq/mg Cr, Amersham Pharmacia Biotech.) for 1 hr at 37ЊC. After three washes with phosphate-buffered saline (PBS) Daudi cells (10 4 /well) were incubated in 96well round bottom plates with VͲ9/Vͳ2 T cells at the indicated effector:target (E:T) cell ratios. The cells were incubated for 6 hr at 37ЊC, and the specific lysis was determined as described (24). The blocking or control antibodies were added to the Daudi cells just before mixing with the effector VͲ9/Vͳ2 T cells.

Analysis of TCR Downregulation
The VͲ9/Vͳ2 T-cell clones were stimulated with the indicated concentrations of IPP. Twenty-four hours later, cells were harvested and washed three times with ice-cold PBS before staining. Cells were stained with Vͳ2-PE on ice for 30 min, washed three times with ice-cold PBS and analyzed by flow cytometry (FASCan® Flow Cytometer, Becton Dickinson, San Jose, CA) using the Quantum Simple Cellular kit following the instructions of the manufacturer (Flow Cytometry Standards Corp., San Juan, Puerto Rico). Dead cells were excluded by propidium iodide (PI) staining. Data analysis was performed using CELLQuest (Becton Dickinson).

Evaluation of Results
The error bars in the figures indicate calculated standard errors (SE) of the measurement. The statistical significance of differences was calculated by Student's t-test. A p value of less than 0.05 was considered significant.

Antigen Presenting Cells Are Required for Optimal VͲ9/Vͳ2 T-Cell Responses to Antigen
Human VͲ9/Vͳ2 T cells recognize nonpeptidic antigens in a MHC-independent manner. Previous studies showed that some VͲ9/Vͳ2 T cells can respond to IPP in the absence of APCs (20). To examine the APC requirements in our VͲ9/Vͳ2 T-cell clones, six clones were stimulated with IPP in the presence or absence of APCs (irradiated LCL721 cells). Four out of six clones responded to IPP in the absence of APCs (Fig. 1A). Nevertheless, the presence of APCs  greatly enhanced their responses. However, two out of six clones absolutely required the presence of APCs for the antigen-induced DNA synthetic response to occur (Fig. 1B). This observation indicates that, at least in some VͲ9/Vͳ2 T cells, surface molecules other than the TCR may be involved in the activation process.
Subsequently, we examined the expression of several of costimulatory molecules, namely CD28, CD2, LFA-1 and LFA-3, by flow cytometry. None of the tested clones express CD28 (Fig.  2), an observation compatible with the results of Gramzinski et al. (25), suggesting that CD28 may not be involved in VͲ9/Vͳ2 T-cell responses. All tested clones expressed CD2, LFA-1 and LFA-3 (Fig. 2). Since CD2 and LFA-1 were expressed on all these clones, we investigated their potential roles in VͲ9/Vͳ2 T-cell responses.

LFA-1 Is Required for VͲ9/Vͳ2 T-Cell Cytotoxicity Against Daudi B-Cell Lymphoma
VͲ9/Vͳ2 T cells are able to kill lymphoma targets such as Daudi cells (21). It has been shown that both LFA-1 and CD2 are involved in the cytotoxic activity of CD8 positive T cells and NK cells (10). To determine the role of LFA-1 and CD2 in the lysis of target cells by VͲ9/Vͳ2 T cells, we performed standard cytotoxicity assays in the presence of specific antibodies. The anti-LFA-1 antibody effectively inhibited the lysis of Daudi lymphoma targets by VͲ9/Vͳ2 T cells (Fig. 3A). However, neither anti-LFA-3 (Fig. 3B) nor anti-CD2 (not shown) had any significant influence on the target lysis.

The CD2-LFA-3 Interaction Is Required for the VͲ9/Vͳ2 T-Cell Response to IPP
Some CD2 antibodies can strongly inhibit the Ͱͱ T-cell antigenic response (10). We examined whether the relevant antibodies can block the response of VͲ9/Vͳ2 T cells to IPP. First, we determined optimal concentrations of anti-CD2 and anti-LFA-3 for blocking (Fig. 4). VͲ9/Vͳ2 T cells were stimulated with IPP in the presence of anti-CD2 or anti-LFA-3. Both antibodies strongly inhibited the antigen-induced DNA synthetic response (Fig. 5A). Since TNF-Ͱ is a predominant cytokine produced by antigen-stimulated VͲ9/Vͳ2 T cells (19), in parallel experiments, we investigated the influence of anti-CD2, anti-LFA-3 and anti-LFA-1 on the production of TNF-Ͱ. The TNFͰ production was strongly inhibited by anti-CD2 and anti-LFA-3 (Fig. 5C). In contrast, anti-LFA-1 had negligible influence on both the proliferative response of VͲ9/Vͳ2 T cells (Fig. 5B) and the TNF-Ͱ production (Fig. 5D).
Similar to Ͱͱ T cells, Ͳͳ T cells have been categorized as Th1 or Th2 cells according to their cytokine profile (26). We analyzed the secretion of interferon (IFN)-Ͳ by VͲ9/Vͳ2 T-cell clones. All tested VͲ9/Vͳ2 T-cell clones released high levels of IFN-Ͳ upon stimulation with IPP (data not shown), which is consistent with the Th1 phenotype (26).

IL-2 Can Restore the VͲ9/Vͳ2 T-Cell Response to IPP
There was a possibility that the CD2/LFA-3 inhibitory effect was due to a decreased production of proliferative cytokines, in particular of IL-2. VͲ9/Vͳ2 T cells were stimulated with IPP in the presence of the anti-CD2 or control antibody. Twenty-four hours later, the supernatants from the cultures were removed and used to culture IL-2-dependent CTLL-2 cells. The supernatants from anti-CD2-treated cells had substantially lower capacity to support the growth of CTLL-2 cells (Fig. 6), suggesting a potential deficit of IL-2. Since it was not easy in this system to exclude the involvement of other cytokines or other mechanisms, we examined whether the addition of exogenous IL-2 could restore the VͲ9/Vͳ2 T-cells response to IPP. Both anti-CD2 (Fig. 7A) and anti-LFA-3treated, IPP-stimulated VͲ9/Vͳ2 T cells (Fig. 7B) released substantial amount of TNF-Ͱ in the presence of exogenous IL-2. The restoration of the TNF-Ͱ release was between approximately 50% (Fig. 7A) and 100% (Fig. 7B) in different experiments. It is noteworthy that the VͲ9/Vͳ2 T-cell clones did not secrete measurable levels of TNF-Ͱ when exposed to IL-2 in the absence of IPP stimulation.

Anti-CD2 and anti-LFA-3 Antibodies Do Not Influence the Antigen-induced TCR Downregulation
The antigenic activation of T cells is accompanied by a significant downregulation of the cell surface TCRs (27). This TCR downregulation process can be modulated by the signal delivered through costimulatory molecules (28). Since anti-CD2 or anti-LFA-3 inhibited the response of VͲ9/Vͳ2 T cells to IPP, we investigated whether or not the antibodies also influenced the IPP-induced downregulation of TCRs. The level of TCR downregulation in the  Continued on next page. presence of anti-CD2 or anti-LFA-3 was comparable to the control (Fig. 8). This indicates that the blocking of CD2-LFA-3 interactions, followed by the functional inhibition of VͲ9/Vͳ2 T cells had no substantial influence on the antigen induced TCR downregulation.

Discussion
Human Ͳͳ T lymphocytes display potent responses to many antigenic entities of bacterial, protozoal, viral and tumor origin (19,(29)(30)(31), but the specifics of their protective function and the mechanisms of their regulation remain unknown. The positive effect of APCs on the Ͳͳ T-cell response to IPP suggests that accessory molecules may influence the response.
The present experiments provide insights into the regulation of VͲ9/Vͳ2 T-cell activities and address the role of specific accessory molecules. It has been shown that both LFA-1 and CD2 are required for CD8 + Ͱͱ T cells to kill target cells (11,15). LFA-1 appears to be the primary molecule that mediates the effectortarget cell contact and CD2 additionally mediates the adhesion among effector and target cells (32). The data in this study strongly sug-gest that, similar to Ͱͱ T cells, the LFA-1 molecule is important in the contact between the target cell and the cytotoxic VͲ9/Vͳ2 T cells. In contrast to Ͱͱ T cells, the potential accessory role of CD2 in VͲ9/Vͳ2 effector-target interactions is dispensable, since anti-CD2 or anti-LFA-3 have practically no detectable effect on VͲ9/Vͳ2 T-cell cytotoxicity against Daudi cells.
Although LFA-1 seems to play an important role in the mechanism of VͲ9/Vͳ2 T-cell cytotoxicity, it is not required for VͲ9/Vͳ2 T cells to respond to IPP. In contrast, in Ͱͱ T cells, LFA-1 can effectively enhance the antigenic response (28). The cell-mediated cytotoxicity is usually dependent on stable effectortarget conjugates that require the LFA-1 function (11). Our results imply that stable cell-cell conjugates may not be required for the IPP response of VͲ9/Vͳ2 T cells.  The mechanism of the inhibition by anti-CD2 antibodies has been studied in Ͱͱ T cells, resulting in a suggestion that anti-CD2 inhibits the expression of the functional IL-2 receptor (33). This is clearly not the case in our experiments with VͲ9/Vͳ2 T-cell clones. Our results indicate that functional IL-2 receptors are expressed on VͲ9/Vͳ2 T cells after stimulation in the presence of anti-CD2. The possible explanation that the expression of IL-2 receptors was caused by in vitro culture of VͲ9/Vͳ2 T cells is unlikely, because the VͲ9/Vͳ2 T cells could only respond to IL-2 after the IPP stimulation. Thus the anti-CD2 inhibitory effect on the proliferation of IPP-stimulated VͲ9/Vͳ2 T cells was more likely to be due to an inadequate secretion of proliferative cytokines. This was further supported by the finding that exogenous IL-2 could reverse the anti-CD2 inhibitory effect on VͲ9/Vͳ2 T-cell proliferation (data not shown).
The anti-CD2 antibody used in this study does not induce any intracellular phosphorylation when binding to the CD2 molecule and its effects are attributable solely to the deficit of CD2 signaling caused by the disrupted interactions between CD2 and LFA-3 (34). The ability of IL-2 to restore the TNF-Ͱ response of V␥9/V␦2 T cells in the presence of anti-CD2 suggests that the CD2 signals induced by the inter-action with LFA-3 can be successfully substituted by signaling through the IL-2 receptor to proceed with the antigen-induced cytokine production. Neither the signaling through CD2 nor the signals through the IL-2 receptor are fully understood. Interestingly, both of them can activate PI3 kinase (35,36), which in turn, activates transcription factors important for entering cell cycle (36). Some accessory molecules mediate costimulatory signals that can modulate the Ͱͱ TCR downregulation induced by antigenic exposure (28). However, our experiments with VͲ9/Vͳ2 T-cell clones indicate that the blocking of the CD2 signal important for the DNA synthetic and cytokine responses does not influence to the process of antigen-induced TCR downregulation in these cells.
The interactions of accessory molecules (such as LFA-3 or ICAMs) with their T-cell co-receptors, such as CD2 or LFA-1, appear to check the plasticity of VͲ9/Vͳ2 T-cell activation, allowing exquisite control of effector functions. Our study shows that the accessory costimulation of T cells stimulated through the TCR is regulated in a somewhat different fashion in Ͳͳ T cells, compared with Ͱͱ T cells. Addressing these differences in detail in future experiments may bring a better understanding of the mechanisms by which a successful co- ordinated immune response is induced. The costimulatory signals are most likely to be involved in the expansion phase of the immune response when proliferative cytokines such as IL-2 are the main limiting factors for generating high numbers of antigen-specific T cells. In addition, similar (but negative) signals may be important for the termination phase of the immune response, when the downmodulation of T-cell immunity is necessary to prevent immunopathology or the development of autoimmunity. Thus, the exact knowledge of unique pathways activated by individual classes of accessory molecules and T-cell co-receptors, their cross talk and the regulation of their expression may be crucial when designing novel strategies for vaccination, immunotherapies and the treatment of autoimmune diseases.