Human Vγ9Vδ2 T cells are characterized by a unique specificity for certain tumors (e.g., Daudi), cells presenting so-called phosphoantigens such as isopentenyl pyrophosphate (IPP), or cells treated with aminobisphosphonates. We now report conversion of hematopoietic and nonhematopoietic tumor cell lines into Vγ9Vδ2 T cell activators by means of short hairpin RNA-mediated knockdown of expression of the IPP-consuming enzyme, farnesyl pyrophosphate synthase (FPPS). FPPS knockdown cells activated Vγ9Vδ2 T cells, as measured by increased levels of CD69 and CD107a, killing of FPPS knockdown cells, and induction of IFN-γ secretion. The IPP-synthesis-inhibiting drug mevastatin reduced Vγ9Vδ2 T cell activation by FPPS knockdown cells but not activation by the phosphoantigen bromohydrin pyrophosphate. In conclusion, our data support the concept of Vγ9Vδ2 T cells as sensors of a dysregulated isoprenoid metabolism and suggest therapeutic down-modulation of FPPS expression as an additional tool to target tumor cells to Vγ9Vδ2 T cell-mediated immunosurveillance.

The majority of circulating human γδ T cells expresses a TCR comprising the variable segments Vγ9 and Vδ2 (Vγ9Vδ2 TCR). These cells massively expand in various infectious diseases, exhibit MHC-unrestricted cytolysis, and, in the presence of IL-2, produce Th1-type cytokines such as IFN-γ and TNF-α (1, 2). In addition, there is increasing evidence for an important role in host defense (3, 4) and tumor surveillance (5, 6). A hallmark of Vγ9Vδ2 T cells is their unique Ag reactivity. Several types of Vγ9Vδ2 T cell activators are known. At first the three types of phosphoantigens which comprise (E)-4-hydroxy-3-methyl-but-enyl pyrophosphate (HMB-PP),6 isopentenyl pyrophosphate (IPP), and bromohydrin pyrophosphate or Phosphostim (BrHPP). HMB-PP is the immediate precursor of IPP in the 1-deoxy-d-xylulose-5-phosphate pathway of isoprenoid synthesis, which is common to plants, many bacteria, and apicomplexan protozoa but absent in mammals (1, 7). IPP is found in all organisms. Its Vγ9Vδ2 T cell-activating potency is several orders of magnitude lower than that of HMB-PP but it is the most potent Vγ9Vδ2 T cell activating metabolite of mammals, which exclusively use the mevalonate pathway of isoprenoid synthesis (1). BrHPP and related substances are highly potent synthetic IPP analogs (8). The exact mode of Vγ9Vδ2 T cell activation by these “phosphoantigens” is not yet clear, but is known to be TCR mediated (9) and to require cell-cell contact (10).

The second type of Vγ9Vδ2 T cell activators are certain tumor cells, such as the human B cell lymphoma Daudi. It has been proposed that increased levels of intracellular IPP (11) are somehow recognized by the Vγ9Vδ2 T cells. Increased IPP levels have also been proposed to induce Vγ9Vδ2 T cell activation early during bacterial infections (12).

In line with the concept that Vγ9Vδ2 cells act as sensors of increased cellular IPP levels, mevastatin, which blocks 3-hydroxy-3-methyl-glutaryl (HMG)-CoA-reductase, and consequently the synthesis of downstream metabolites such as IPP, inhibits Daudi cell mediated-Vγ9Vδ2 T cell activation (11). Alternatively, several cell surface molecules have been suggested to act as direct Vγ9Vδ2 TCR ligands or modifiers of Vγ9Vδ2 TCR ligand interactions (13, 14). Ectopically expressed F1-ATPase complex, in conjunction with apolipoprotein A, binds recombinant Vγ9Vδ2 TCR (14). How far the modulation of IPP levels may affect the Vγ9Vδ2 TCR binding or expression of these postulated ligands has not been investigated.

Aminobisphosphonates (e.g., zoledronate or pamidronate) also induce Vγ9Vδ2 T cell activation and proliferation in vitro and in vivo (15, 16), trigger Vγ9Vδ2 TCR mediated lysis of tumor lines, and induce Vγ9Vδ2 T cell-mediated kill of autologous myeloma cells in bone marrow biopsies (15, 16, 17). These findings imply that Vγ9Vδ2 T cells could mediate immunotherapy of tumors, which is corroborated by clinical studies where an objective tumor response could be observed (5, 17, 18). Similar to phosphoantigens, Vγ9Vδ2 T cell activation by aminobisphosphonates requires cognate recognition by the TCR, but Vγ9Vδ2 T cell activation by both classes of substances varies insofar that Vγ9Vδ2 TCR recognizes cells incubated with aminobisphosphonates and washed subsequently but not cells pulsed with phosphoantigens. Because aminobisphosphonates bind and inhibit IPP-consuming enzymes such as farnesyl pyrophosphate synthase (FPPS) and geranylgeranyl pyrophosphate synthase (GGPPS) (19), it has been proposed that aminobisphosphonate-mediated Vγ9Vδ2 T cell activation reflects accumulation and subsequent presentation of the Vγ9Vδ2 TCR ligand IPP (11). FPPS-inhibition has also been suggested as a mechanism for the Vγ9Vδ2 T cell activation by alkylamines (20). However, a direct interaction between Vγ9Vδ2 TCR and aminobisphosphonate or aminobisphosphonate-modified cellular compounds cannot be excluded.

In this study, we use a novel approach to convert tumor cells to Vγ9Vδ2 T cell activators. We demonstrate that short hairpin RNA (shRNA)-mediated reduction of expression of the IPP-consuming enzyme FPPS faithfully mimics the Vγ9Vδ2 T cell-activating action of aminobisphosphonates and may provide an alternate or supplementary strategy for promoting Vγ9Vδ2 T cell-mediated immunosurveillance.

Raji, Daudi, and 293T cells were obtained from American Type Culture Collection. HepG2 cells were a gift of Ann-Kristin Müller (Zentrum für Infektionsforschung, Würzburg, Germany), and AsPC-1 was obtained from DSMZ. The 293T cells were propagated in DMEM (Invitrogen) and the other lines in RPMI 1640 medium supplemented with 10% FCS, 200 U/ml penicillin, 200 U/ml streptomycin, 1 mM sodium pyruvate, 4.8 mM l-glutamine, 1× nonessential amino acids (Invitrogen), and 0.025 mM 2-ME (Sigma-Aldrich). Zoledronate was obtained from Novartis Pharma. Generation of short-term human primary Vγ9Vδ2 T cell lines was done as described previously (21). Briefly, PBMC were cultured at 1 × 106/ml for up to 15 days in the presence of 1 μM BrHPP (Innate Pharma) and 100 IU/ml rhuIL-2 (Miltenyi Biotec). Half of the medium was replaced at days 3, 7, and 10 by fresh medium containing IL-2. At the time of the experiments, cell lines contained 30–60% Vγ9Vδ2 T cells, as shown by FACS analysis. Highly purified Vγ9Vδ2 T cell lines were obtained by enriching 7-days-cultured short-term lines by magnetic cell separation using the TCRγ/δ microbead kit (Miltenyi Biotec) according to the manufacturer’s instructions. The purity of positively selected γδ T cells (>98%) was confirmed by FACS analysis. Cells were cultured in medium without IL-2 for 1–3 days before the assay.

RT-PCR was performed to clone human FPPS and GGPPS from total RNA which was isolated from 5 × 106 Raji cells following the protocol of the RNeasy MiniKit (Qiagen). The cDNA was synthesized according to the manufacturer’s RT-PCR protocol supplied with a First Strand cDNA synthesis kit (Fermentas). PCR was performed with Phusion Hot Start high-fidelity DNA polymerase (Finnzymes), using the following FPPS- or GGPPS-specific primers (MWG Biotec): Forward FPPS: 5′-TAG CCC AAT TGG CCG CCA CCA TGC CCC TGT CCC GCT-3′, Reverse FPPS: 5′-AGT CCG GAT CCC GCT TTC TCC GCT TGT AGA T-3′; Forward GGPPS: 5′-GCT ACC AAT TGA CCA TGG AGA AGA CTC AAG AAA CAG-3′, Reverse GGPPS: 5′-AGC ATG GAT CCC CTT CAT TTT CTT CTT TGA ACA TC-3′. To express FPPS-enhanced yellow fluorescent protein (EYFP) or GGPPS-EYFP in mammalian cells, PCR products were cloned into the EcoRI and BamHI sites of the Moloney murine leukemia virus-derived retroviral vector pIZ-MCS-EYFP (22). A bicistronic message generated from this vector codes for both the EYFP fusion protein and downstream of an internal ribosomal entry site for the zeocin-resistant protein.

To generate FPPS knockdown cells, Expression ArrestLMP microRNA-adapted retroviral shRNAmir Vector (MSCV-LMP) was obtained from Open Biosystems. For construction of LMP/AsRed, the red fragment of pred/MCS which contains the AsRed fluorescent protein (red) was ligated into the NcoI and PacI sites of MSCV-LMP, to replace the GFP coding region. Two specific shRNA for FPPS (Open Biosystems) (V2HS_228248: AAGGTATATTGCTGTTGACAGTGAGCGACCAGCAGTGTTCTTGCAATATTAGTGAAGCCACAGATGTAATATTGCAAGAACACTGCTGGCTGCCTACTGCCTCGGAATATACCTT and V2HS_113816: AAGGTATATTGCTGTTGACAGTGAGCGCCCTCCTGGAAGCATGTATCTATAGTGAAGCCACAGATGTATAGATACATGCTTCCAGGAGGTTGCCTACTGCCTCGGAATATACCTT) and one unrelated control shRNA AsGRII (Open Biosystems) (V2MM_18443: AAGGTATATTGCTGTTGACAGTGAGCGCGGCGATACCAGGATTCAGAAATAGTGAAGCCACAGATGTATTTCTGAATCCTGGTATCGCCTTGCCTACTGCCTCGGAATATACCTT) were cloned into XhoI and EcoRI sites of LMP/AsRed, respectively. Abbreviations used: AS, empty vector LMP/AsRed; ASGr2, recombinant LMP/AsRed encoding unrelated control shRNA GRII; AS22 (V2HS_228248) and AS11 (V2HS_113816), recombinant LMP/AsRed encoding FPPS-specific shRNA.

Recombinant retroviral particles were generated by using the three vector-packaging-system as described by Soneoka et al. (23). Transduced cells were selected by supplementing the culture medium with 3 μg/ml puromycin (Sigma-Aldrich), or 250 μg/ml zeocin (Invitrogen), respectively, or cell sorting using a FACSVantage (BD Biosciences).

A commercially available IFN-γ ELISPOT kit (Mabtech) containing a precoated 96-well plate was used (Mabtech, code 3420–2APW-2). After 2 h of blocking of unspecific binding sites with cell culture medium containing 5% FCS (200 μl/well), medium was removed. Subsequently, 5 × 103 Vγ9Vδ2 T cells (numbers were calculated from staining of short term lines with FITC-anti-human Vδ2 TCR Ab (Beckman Coulter; clone IMMU 389) were cocultured with 2.5 × 104 target cells, with or without treatment of 1 μM mevastatin (Sigma-Aldrich), respectively (final volume of 150 μl/well). Plates were incubated for 16–18 h in 37°C, 5% CO2. Thereafter, cells were removed and ELISPOTs were developed and counted by an ELISPOT Reader (ELR03, AID).

Whole-cell lysates were prepared by lysing 1.5 × 106 cells in 50 μl lysis buffer (50 mM (pH 7.4) Tris-HCL, 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 μg/ml aprotinin, leupeptin, and pepstatin, 1 mM Na3VO4, 1 mM NaF) combined with a protease inhibitor mixture, Complete Mini (Roche). Proteins were separated by 12% SDS-PAGE, and transferred to a Roti-polyvinylidene difluoride membrane (Carl Roth). FPPS expression levels were determined using rabbit anti-FPPS polyclonal Ab (Abgent; animal ID RB4786). As loading control, β-tubulin expression was detected by anti-β-tubulin mAb (BD Pharmingen; clone 5H1).

Cytometric analysis was performed on a FACSCalibur instrument using Cell Quest or FlowJo software. For evaluation of CD69 expression, 1 × 105 effector cells (fresh PBL) were cocultured with 1 × 105 target cells in 96-well tissue culture plates. After 16–18 h of incubation, cells were harvested and stained for 30 min with FITC-anti-human Vδ2 TCR Ab plus PE-conjugated mouse anti-human CD69 (Beckman Coulter; clone TP1.55.3). For evaluation of the CD107a expression, effector cells (1 × 105 fresh PBL or 3 × 104 primary γδ T cell line) were cocultured with 1 × 105 target cells plus 10 μl conjugated mouse anti-human CD107a/LAMP-1 (BD Pharmingen; clone H4A3) in 96-well tissue culture plates. After 16–18 h (Fig. 3B) or 4 h incubation (Table I), cells were harvested and stained for 30 min with FITC-anti-human γδ TCR Ab (BD Pharmingen; clone B1) or FITC-anti-human Vδ2 TCR Ab.

Table I.

Tumors of hematopoietic and nonhematopoietic origin transduced with FPPS-specific shRNA AS22 activate Vγ9Vδ2 T cellsa

Stimulator cellsExpressed shRNA% CD107a Positive in Vδ2 T cells
Medium0.1 μM BrHPP
AsPC-1 AS 3.7 17.6 
AS22 9.1 28.2 
HepG2 AS 7.3 76.2 
AS22 53.1 79.5 
293T AS 4.9 63.2 
AS22 18.0 63.2 
Raji AS 5.8 85.3 
AS22 42.6 85.7 
Daudi None 26.6 85.3 
None None 3.6 6.8 
Stimulator cellsExpressed shRNA% CD107a Positive in Vδ2 T cells
Medium0.1 μM BrHPP
AsPC-1 AS 3.7 17.6 
AS22 9.1 28.2 
HepG2 AS 7.3 76.2 
AS22 53.1 79.5 
293T AS 4.9 63.2 
AS22 18.0 63.2 
Raji AS 5.8 85.3 
AS22 42.6 85.7 
Daudi None 26.6 85.3 
None None 3.6 6.8 
a

Cells (3 × 104) of a Vγ9Vδ2 T cell line were incubated for 4 h with 105 cells of the indicated cell lines in the presence of PE-anti-CD107a Ab. Thereafter cells were harvested and costained with FITC-anti-Vδ2 TCR Ab.

Effector cells (1 × 105) were cocultured with the same number of target cells in 96-well tissue culture plates. After 16–18 h incubation, cells were harvested and analyzed for IFN-γ secretion using the IFN-γ secretion assay detection kit (Miltenyi Biotec) according to manufacturer’s instructions. The cells were counterstained with a FITC-conjugated mAb against human Vδ2 TCR.

FACS-based cytotoxicity assays were performed as described (24). After 4 h, cells were harvested, brought up to 400 μl with 1× PBS, stained with the nucleic acid-specific ToPro-3 iodide stain (0.5 μM, 15 min on ice) and further analyzed by flow cytometry.

To test for the efficacy of different shRNAs to reduce FPPS expression, Raji cells stably expressing both an FPPS-EYFP fusion protein (Raji-FPPSYFP) and FPPS-specific shRNA constructs were established. Fig. 1,A shows similar expression of shRNA constructs in the different cell lines generated as indicated by equally high AsRed fluorescence in FL3 (AsRed), and widely varying levels of FL1-fluorescence (EYFP) (Fig. 1 B). Cells transduced with FPPS-targeting vectors AS11 and AS22 showed a 5–10-fold reduction of FL1, compared with cells transduced with empty vector (AS) or a vector comprising an unrelated shRNA control vector (ASGr2). Additionally, we tested for specificity of the shRNA construct by generating cells expressing the prenyltransferase GGPPS as EYFP fusion protein. In contrast to the FPPS-EYFP, the AS22 vector had no effect on the GGPPS-EYFP expression (data not shown). This is of especial interest because aminobisphosphonates bind and inhibit not only FPPS but also GGPPS (19). It also suggests that shRNA knockdown of the different enzymes of isoprenoid synthesis may provide feasible tools in further dissecting their role in cellular functions.

FIGURE 1.

Suppression of FPPS protein expression by shRNA (AS22 or AS11). The Raji cell line expressing FPPS-EYFP fusion protein (Raji-FPPSYFP) was transfected, with either empty vector control (AS), control shRNA (ASGr2), or FPPS-specific shRNA (AS22 or AS11). After 2 wk of puromycin selection, expression of shRNA constructs (A) and FPPS-YFP were tested by FACS analysis (B). C, Immunoblot analysis of FPPS-expression levels in wild-type Raji cells stably expressing empty vector AS or FPPS-specific shRNA AS22 and AS11. Blots were probed with β-tubulin to verify equal loading. Results are representative of three independent experiments. LTR, long terminal repeat; IRES, internal ribosome entry site.

FIGURE 1.

Suppression of FPPS protein expression by shRNA (AS22 or AS11). The Raji cell line expressing FPPS-EYFP fusion protein (Raji-FPPSYFP) was transfected, with either empty vector control (AS), control shRNA (ASGr2), or FPPS-specific shRNA (AS22 or AS11). After 2 wk of puromycin selection, expression of shRNA constructs (A) and FPPS-YFP were tested by FACS analysis (B). C, Immunoblot analysis of FPPS-expression levels in wild-type Raji cells stably expressing empty vector AS or FPPS-specific shRNA AS22 and AS11. Blots were probed with β-tubulin to verify equal loading. Results are representative of three independent experiments. LTR, long terminal repeat; IRES, internal ribosome entry site.

Close modal

After interference with FPPS had been demonstrated by flow cytometry, AS22 and AS11 were used to target wild-type Raji cells. As shown by immunoblot, AS22 and AS11 significantly reduce FPPS expression compared with empty vector control AS (Fig. 1 C). Densitometric analysis revealed a relative reduction of FPPS protein by 90% for Raji/AS22, by 62% for Raji/AS11 cells, respectively.

To investigate the capacity of FPPS knockdown cells to activate primary γδ T cells, total PBMC from healthy donors were stimulated in cocultures with different target cells as described in Materials and Methods. Subsequently, cells were stained for CD69 and CD107a and tested in an IFN-γ secretion assay. The proportion of activated Vγ9Vδ2 T cells was tested electronically by gating on Vδ2 TCR positive cells, which varied between 1 and 3% of total PBMC (data not shown). The frequency was determined by expression of the early activation marker CD69, CD107a as marker for the active cytolytic machinery, and of IFN-γ secreting cells as described in Materials and Methods. Fig. 2, A and B, show the induction of CD69 and CD107a by Raji-FPPS knockdown cells as well as by aminobisphosphonates and phosphoantigen, but not by cells expressing only control vector constructs (Raji/AS and Raji/ASGr2). In three of three experiments, Raji/AS22 shRNA knockdown cells induced up-regulation of CD69 and CD107a expression, which was always stronger than by Raji-AS11 cells. AS11 cells always induced CD69 expression by the Vγ9Vδ2 T cells but varied with respect to induction of CD107a expression between hardly detectable induction (data not shown) or induction reaching about half of that found with Raji/AS22 cells (Fig. 2,B). IFN-γ secretion was tested in three experiments and was, again, more efficient with Raji/AS22 than with Raji/AS11 cells (Fig. 2,C). The differential degree of activation by both types of FPPS knockdown cells correlates with the differential reduction of FPPS (Fig. 1 C). Cells transduced with vector controls (Raji/AS and Raji/ASGr2) did not activate Vγ9Vδ2 T cells. Finally, Raji-FPPS knockdown cells induced no significant up-regulation of CD69 and CD107a on NK cells and αβ T cells, indicating specific activation of the γδ T cells (data not shown).

FIGURE 2.

FPPS knockdown in Raji cells increases CD69 and CD107a expression and IFN-γ production by primary Vδ2 T cells. The contour plots depict the gating strategy used to generate the data depicted in the bar histograms. A, Representative FACS analysis of CD107a induction after cocultures of PBMC with indicated Raji cells, BrHPP, Zoledronate, or medium alone. B, Representative FACS analysis of CD69 induction after cocultures of PBMC with indicated Raji cells, BrHPP, Zoledronate, or medium alone. C, Proportion of IFN-γ secreting cells among Vδ2 TCR-positive PBL. FSC, Forward scatter.

FIGURE 2.

FPPS knockdown in Raji cells increases CD69 and CD107a expression and IFN-γ production by primary Vδ2 T cells. The contour plots depict the gating strategy used to generate the data depicted in the bar histograms. A, Representative FACS analysis of CD107a induction after cocultures of PBMC with indicated Raji cells, BrHPP, Zoledronate, or medium alone. B, Representative FACS analysis of CD69 induction after cocultures of PBMC with indicated Raji cells, BrHPP, Zoledronate, or medium alone. C, Proportion of IFN-γ secreting cells among Vδ2 TCR-positive PBL. FSC, Forward scatter.

Close modal

Raji-FPPS knockdown cells were tested for their capacity to induce CD107a up-regulation as well as IFN-γ production in polyclonal Vγ9Vδ2 T cell enriched cell lines. CD107a surface expression of the γδ TCR positive cells was assessed by electronic gating on the γδ TCR positive cells and the proportion of CD107a positive cells was determined. Fig. 3,A shows that CD107a surface expression was strongly up-regulated after over-night coculture with AS22 cells. Furthermore, we observed an increased size of CD107a-positive Vγ9Vδ2 cells, compared with cocultures with control Raji cells. We tested the capacity of FPPS knockdown cells to induce IFN-γ secretion. Fig. 3,B depicts a drastic increase in the frequency of IFN-γ secreting cells in an ELISPOT assay of cocultures of a short term γδ T cell line with FPPS knockdown cells, but not in the negative control cultures. The activation by AS22-transduced cells was more efficient than that by AS11-transduced FPPS knockdown cells, which, again, inversely, correlates with the different levels of residual FPPS reached by the knockdown constructs. AS- and AS22-transduced Jurkat cells were also tested for their Vγ9Vδ2 T cell activating properties using the same experimental setting as that shown in Fig. 3 B. AS22-transduced Jurkat clearly induced IFN-γ release (245 spots) while AS-transduced cells did not (one spot).

FIGURE 3.

FPPS knockdown in Raji cells induces effector-functions in Vγ9Vδ2 T cells. A, FACS analysis of CD107a surface expression in Vγ9Vδ2 T cells. Cells (3 × 104) of a Vγ9V2δ T cell line were incubated overnight with 105 cells of the indicated Raji lines in the presence of PE-anti-CD107a Ab. Thereafter cells were harvested and costained with FITC-anti-γδ TCR Ab. Events shown are gated on the γδ TCR-positive population. The numbers represent the percent of CD107a-positive cells in the γδ T cell gate. B, ELISPOT IFN-γ assays of overnight cocultures of 3 × 103 Vγ9Vδ2 T cells with 2.5 × 104 of the indicated Raji cells. C, A Vγ9Vδ2 T cell-enriched cell line was tested for cytotoxic activity against Raji cells transduced with control vector (Raji/AS; squares), transduced with FPPS knockdown vector (Raji/AS22; triangles), or Daudi cells (circles) as described in Materials and Methods. E:T ratios are indicated. Duration of the assay was 4 h. FSC, Forward scatter.

FIGURE 3.

FPPS knockdown in Raji cells induces effector-functions in Vγ9Vδ2 T cells. A, FACS analysis of CD107a surface expression in Vγ9Vδ2 T cells. Cells (3 × 104) of a Vγ9V2δ T cell line were incubated overnight with 105 cells of the indicated Raji lines in the presence of PE-anti-CD107a Ab. Thereafter cells were harvested and costained with FITC-anti-γδ TCR Ab. Events shown are gated on the γδ TCR-positive population. The numbers represent the percent of CD107a-positive cells in the γδ T cell gate. B, ELISPOT IFN-γ assays of overnight cocultures of 3 × 103 Vγ9Vδ2 T cells with 2.5 × 104 of the indicated Raji cells. C, A Vγ9Vδ2 T cell-enriched cell line was tested for cytotoxic activity against Raji cells transduced with control vector (Raji/AS; squares), transduced with FPPS knockdown vector (Raji/AS22; triangles), or Daudi cells (circles) as described in Materials and Methods. E:T ratios are indicated. Duration of the assay was 4 h. FSC, Forward scatter.

Close modal

Next, we tested Vγ9Vδ2 T cell enriched-lines for their capacity to mediate a dose-dependent killing of FPPS knockdown cells in comparison to control vector-transduced cells and Daudi cells. Fig. 3,C shows killing of AS22-transduced Raji cells, AS control Raji cells, and Daudi by a Vγ9Vδ2 T cell-enriched cell line. The killing of AS22 cells, as measured on a per cell basis, was ∼10-fold more efficient than killing of untransduced cells, but still less efficient than that of Daudi cells. Fig. 3 is representative for three independent experiments.

To exclude that observed effects were indirect and do not result from activation of other cell populations such as αβ T cells, NK cells, or monocytes, which then may provide help to the γδ T cells, CD107a up-regulation and IFN-γ secretion was also tested using highly purified Vγ9Vδ2 T cells. Fig. 4 shows both assays for specific activation of the highly purified γδ T cell lines (>98% purity) by FPPS knockdown cells.

FIGURE 4.

FPPS knockdown in Raji cells induces effector-functions in highly purified Vγ9Vδ2 T cells. Highly purified short-term γδ T cells lines were enriched by MACS for γδ T cells (>98% purity), rested for 1 (A) or 3 days (B), and tested for CD107a surface expression or frequency of IFN-γ secreting cells, respectively. A, Cells (3 × 104) of purified Vγ9V2δ line T cells were incubated overnight with 105 cells of the indicated cell lines in the presence of PE-anti-CD107a Ab. Thereafter cells were harvested and costained with FITC-anti-γδ TCR Ab. The numbers represent the percent of CD107a positive γδ T cells. B, Results of an ELISPOT IFN-γ assay of overnight cocultures of 5 × 103 purified Vγ9V2δ line T cells with 2.5 × 104 of the indicated Raji cells.

FIGURE 4.

FPPS knockdown in Raji cells induces effector-functions in highly purified Vγ9Vδ2 T cells. Highly purified short-term γδ T cells lines were enriched by MACS for γδ T cells (>98% purity), rested for 1 (A) or 3 days (B), and tested for CD107a surface expression or frequency of IFN-γ secreting cells, respectively. A, Cells (3 × 104) of purified Vγ9V2δ line T cells were incubated overnight with 105 cells of the indicated cell lines in the presence of PE-anti-CD107a Ab. Thereafter cells were harvested and costained with FITC-anti-γδ TCR Ab. The numbers represent the percent of CD107a positive γδ T cells. B, Results of an ELISPOT IFN-γ assay of overnight cocultures of 5 × 103 purified Vγ9V2δ line T cells with 2.5 × 104 of the indicated Raji cells.

Close modal

To test whether FPPS-specific shRNA induces a Vγ9Vδ2 T cell-activating phenotype in cells of nonhematopoietic origin, AsPC-1 (pancreatic adenosarcoma), HepG2 (hepatocellular carcinoma), and 293T cells (human embryonic kidney cell line expressing large T Ag) were transduced with AS or AS22 shRNA. Analysis of the cells was done as described in Fig. 3,A, but duration of coculture was restricted to 4 h. As shown in Table I, only AS22-transduced cells, Daudi cells, or addition of BrHPP led to a clear activation of the Vγ9Vδ2 T cell line.

One explanation for Vγ9Vδ2 T cell activation by aminobisphosphonates is their capacity to block FPPS in the target cell, leading to the accumulation and subsequent presentation of IPP to the Vγ9Vδ2 T cells. To test this hypothesis, we treated FPPS knockdown cells with mevastatin, which lowers IPP levels as a consequence of blocking HMG-CoA-reductase activity. IFN-γ secretion induced by both types of FPPS knockdown cells was significantly inhibited by mevastatin (Fig. 5). Similarly, we observed reduced induction of IFN-γ secretion induced by aminobisphosphonate zoledronate-treated cells. Addition of the exogenous ligand BrHPP to mevastatin-treated cells reversed the block and led to efficient IFN-γ production.

FIGURE 5.

Effects of mevastatin treatment on induction of IFN-γ secretion by FPPS knockdown Raji and zoledronate-treated Raji cells. ELISPOT IFN-γ assays of overnight cocultures of 3 × 103 γδ cells with 2.5 × 104 of the indicated RAJI cells were performed in the presence of the given concentrations of mevastatin. Percent activation indicates the activation in comparison to activation without mevastatin which is set as 100% activation. The numbers in brackets indicate the spots found for 100% activation with the respective stimulus.

FIGURE 5.

Effects of mevastatin treatment on induction of IFN-γ secretion by FPPS knockdown Raji and zoledronate-treated Raji cells. ELISPOT IFN-γ assays of overnight cocultures of 3 × 103 γδ cells with 2.5 × 104 of the indicated RAJI cells were performed in the presence of the given concentrations of mevastatin. Percent activation indicates the activation in comparison to activation without mevastatin which is set as 100% activation. The numbers in brackets indicate the spots found for 100% activation with the respective stimulus.

Close modal

In conclusion, this paper shows that shRNA-mediated inhibition of FPPS expression is sufficient to induce a Vγ9δ2 T cell stimulatory activity in otherwise nonstimulatory tumor cells. It also provides independent support of the concept that increased intracellular IPP levels are instrumental in Vγ9Vδ2 T cell activation by tumor lines, which so far has been based on a correlation between IPP levels and Vγ9Vδ2-T cell activation as well as by observations with enzyme inhibitors, such as HMG-CoA-reductase inhibitors (e.g., mevastatin) and FPPS inhibitors (e.g., aminobisphosphonates) (11). A special benefit of the use of shRNA in comparison to enzyme inhibitors is that only the transduced tumor cells and not the Vγ9Vδ2 T lymphocytes have been targeted, which is important because the mevalonate pathway is fundamental for cell survival and growth of all eukaryotic cells (25). Furthermore, FPPS-specific shRNA may help to elucidate the inhibition by aminobisphosphonates of tumor growth and survival, which has been proposed to be a consequence of increased IPP levels and the formation of toxic ATP-analogs (ApppI) (26). It will be interesting to see whether similar observations on IPP and ApppI levels will be made with the FPPS knockdown cells, especially because we could not observe such differences between Raji-FPPS and control knockdowns, at least not under the culture conditions used (high glucose and glutamine, saturating oxygen) (J. Li, V. Kunzmann, and T. Herrmann, unpublished data).

In any case, the Vγ9Vδ2 T cell activation by the FPPS knockdown cells strongly supports the current hypothesis that intracellular accumulation of phosphorylated mevalonate metabolites (such as IPP) allows identification of stressed (e.g., infected) (12) or transformed cells (11) and shows that reduced FPPS activity is a key mechanism in stimulation of Vγ9Vδ2 T cells by tumor cells. Indeed, it is tempting to speculate that Vγ9Vδ2 T cell-mediated immunosurveillance may continuously act on tumors with aberrant isoprenoid metabolism, and thereby contribute to the rather small number of human tumors exerting Vγ9Vδ2 T cell activating properties found so far.

The exact mode of recognition of tumor cells and IPP by Vγ9Vδ2 T cells still remains unclear. Beside overproduction of phosphorylated metabolites generated from the mevalonate pathway, recognition of tumor cells by Vγ9Vδ2 T cells requires not only the expression of the Vγ9Vδ2 TCR, as indicated by Ab blocking and gene transfer approaches (9), but probably also species-specific interactions of adhesion (i.e., LFA-1/ICAM-1) and costimulatory signals (27), or even a species-specific Vγ9Vδ2 T cell Ag-presenting molecule (28, 29, 30). How far other molecules, such as the ectopically expressed F1-ATPase, which has been claimed to serve as the Vγ9Vδ2 TCR ligand expressed by Daudi tumor cells, are involved in IPP recognition remains unclear, but we observed neither after treatment with aminobisphosphonates, nor an increased cell-surface binding of β-chain F1-ATPase specific mAbs in Raji-FPPS knockdown cells (J. Li, V. Kunzmann, and T. Herrmann, unpublished data).

Altogether, our data provide a proof of concept that mevalonate pathway dysregulation in tumor cells is sufficient to activate immunosurveillance by Vγ9Vδ2 T cells and may provide a strategic platform to develop novel compounds for immunotherapy of human malignancies.

We thank Niklas Beyersdorf for critical reading of the manuscript and Ingolf Berberich for providing the pIZ-MCS-EYFP vector.

The authors have no financial conflict of interest.

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 Interdiziplinäres Zentrum für Klinische Forschung Grant No. 01KS9603.

6

Abbreviations used in this paper: HMB-PP, (E)-4-hydroxy-3-methyl-but-enyl pyrophosphate; BrHPP, bromohydrin pyrophosphate; EYFP, enhanced yellow fluorescent protein; FPPS, farnesyl pyrophosphate synthase; GGPS, geranylgeranyl pyrophosphate synthase; HMG, 3-hydroxy-3-methyl-glutaryl; IPP, isopentenyl pyrophosphate; shRNA, short hairpin RNA.

1
Morita, C. T., C. Jin, G. Sarikonda, H. Wang.
2007
. Nonpeptide antigens, presentation mechanisms, and immunological memory of human Vγ2Vδ2 T cells: discriminating friend from foe through the recognition of prenyl pyrophosphate antigens.
Immunol. Rev.
215
:
59
-76.
2
Thedrez, A., C. Sabourin, J. Gertner, M. C. Devilder, S. Allain-Maillet, J. J. Fournie, E. Scotet, M. Bonneville.
2007
. Self/non-self discrimination by human γδ T cells: simple solutions for a complex issue?.
Immunol. Rev.
215
:
123
-135.
3
Wang, L., A. Kamath, H. Das, L. Li, J. F. Bukowski.
2001
. Antibacterial effect of human Vγ2Vδ2 T cells in vivo.
J. Clin. Invest.
108
:
1349
-1357.
4
Chen, Z. W., N. L. Letvin.
2003
. Vγ2Vδ2+ T cells and anti-microbial immune responses.
Microb. Infect.
5
:
491
-498.
5
Kunzmann, V., M. Wilhelm.
2005
. Anti-lymphoma effect of γδ T cells.
Leuk. Lymphoma
46
:
671
-680.
6
Kabelitz, D., D. Wesch, W. He.
2007
. Perspectives of γδ T cells in tumor immunology.
Cancer Res.
67
:
5
-8.
7
Eberl, M., M. Hintz, A. Reichenberg, A. K. Kollas, J. Wiesner, H. Jomaa.
2003
. Microbial isoprenoid biosynthesis and human γδT cell activation.
FEBS Lett.
544
:
4
-10.
8
Espinosa, E., C. Belmant, F. Pont, B. Luciani, R. Poupot, F. Romagne, H. Brailly, M. Bonneville, J. J. Fournie.
2001
. Chemical synthesis and biological activity of bromohydrin pyrophosphate, a potent stimulator of human γδ T cells.
J. Biol. Chem.
276
:
18337
-18344.
9
Bukowski, J. F., C. T. Morita, Y. Tanaka, B. R. Bloom, M. B. Brenner, H. Band.
1995
. Vγ2Vδ2 TCR-dependent recognition of non-peptide antigens and Daudi cells analyzed by TCR gene transfer.
J. Immunol.
154
:
998
-1006.
10
Morita, C. T., E. M. Beckman, J. F. Bukowski, Y. Tanaka, H. Band, B. R. Bloom, D. E. Golan, M. B. Brenner.
1995
. Direct presentation of nonpeptide prenyl pyrophosphate antigens to human γδ T cells.
Immunity
3
:
495
-507.
11
Gober, H. J., M. Kistowska, L. Angman, P. Jeno, L. Mori, G. De Libero.
2003
. Human T cell receptor γδ cells recognize endogenous mevalonate metabolites in tumor cells.
J. Exp. Med.
197
:
163
-168.
12
Kistowska, M., E. Rossy, S. Sansano, H. J. Gober, R. Landmann, L. Mori, G. De Libero.
2008
. Dysregulation of the host mevalonate pathway during early bacterial infection activates human TCR γδ cells.
Eur. J. Immunol.
38
:
2200
-2209.
13
Chen, H., X. He, Z. Wang, D. Wu, H. Zhang, C. Xu, H. He, L. Cui, D. Ba, W. He.
2008
. Identification of human T cell receptor γδ-recognized epitopes/proteins via CDR3δ peptide-based immunobiochemical strategy.
J. Biol. Chem.
283
:
12528
-12537.
14
Scotet, E., L. O. Martinez, E. Grant, R. Barbaras, P. Jeno, M. Guiraud, B. Monsarrat, X. Saulquin, S. Maillet, J. P. Esteve, et al
2005
. Tumor recognition following Vγ9Vδ2 T cell receptor interactions with a surface F1-ATPase-related structure and apolipoprotein A-I.
Immunity
22
:
71
-80.
15
Kunzmann, V., E. Bauer, J. Feurle, F. Weissinger, H. P. Tony, M. Wilhelm.
2000
. Stimulation of γδ T cells by aminobisphosphonates and induction of antiplasma cell activity in multiple myeloma.
Blood
96
:
384
-392.
16
Kunzmann, V., E. Bauer, M. Wilhelm.
1999
. γ/δ T-cell stimulation by pamidronate.
N. Engl. J. Med.
340
:
737
-738.
17
Wilhelm, M., V. Kunzmann, S. Eckstein, P. Reimer, F. Weissinger, T. Ruediger, H. P. Tony.
2003
. γδ T cells for immune therapy of patients with lymphoid malignancies.
Blood
102
:
200
-206.
18
Dieli, F., D. Vermijlen, F. Fulfaro, N. Caccamo, S. Meraviglia, G. Cicero, A. Roberts, S. Buccheri, M. D'Asaro, N. Gebbia, et al
2007
. Targeting human γδ T cells with zoledronate and interleukin-2 for immunotherapy of hormone-refractory prostate cancer.
Cancer Res.
67
:
7450
-7457.
19
Guo, R. T., R. Cao, P. H. Liang, T. P. Ko, T. H. Chang, M. P. Hudock, W. Y. Jeng, C. K. Chen, Y. Zhang, Y. Song, et al
2007
. Bisphosphonates target multiple sites in both cis- and trans-prenyltransferases.
Proc. Natl. Acad. Sci. USA
104
:
10022
-10027.
20
Thompson, K., J. Rojas-Navea, M. J. Rogers.
2006
. Alkylamines cause Vγ9Vδ2 T-cell activation and proliferation by inhibiting the mevalonate pathway.
Blood
107
:
651
-654.
21
Rincon-Orozco, B., V. Kunzmann, P. Wrobel, D. Kabelitz, A. Steinle, T. Herrmann.
2005
. Activation of Vγ9Vδ2 T cells by NKG2D.
J. Immunol.
175
:
2144
-2151.
22
Schmidt, D., A. Nayak, J. E. Schumann, A. Schimpl, I. Berberich, F. Berberich-Siebelt.
2008
. Blimp-1Deltaexon7: a naturally occurring Blimp-1 deletion mutant with auto-regulatory potential.
Exp. Cell Res.
314
:
3614
-3627.
23
Soneoka, Y., P. M. Cannon, E. E. Ramsdale, J. C. Griffiths, G. Romano, S. M. Kingsman, A. J. Kingsman.
1995
. A transient three-plasmid expression system for the production of high titer retroviral vectors.
Nucleic Acids Res.
23
:
628
-633.
24
Olin, M. R., K. Hwa Choi, J. Lee, T. W. Molitor.
2005
. γδ T-lymphocyte cytotoxic activity against Mycobacterium bovis analyzed by flow cytometry.
J. Immunol. Methods
297
:
1
-11.
25
Kurakata, S., M. Kada, Y. Shimada, T. Komai, K. Nomoto.
1996
. Effects of different inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, pravastatin sodium and simvastatin, on sterol synthesis and immunological functions in human lymphocytes in vitro.
Immunopharmacology
34
:
51
-61.
26
Monkkonen, H., J. Kuokkanen, I. Holen, A. Evans, D. V. Lefley, M. Jauhiainen, S. Auriola, J. Monkkonen.
2008
. Bisphosphonate-induced ATP analog formation and its effect on inhibition of cancer cell growth.
Anticancer Drugs
19
:
391
-399.
27
Kato, Y., Y. Tanaka, H. Tanaka, S. Yamashita, N. Minato.
2003
. Requirement of species-specific interactions for the activation of human γδ T cells by pamidronate.
J. Immunol.
170
:
3608
-3613.
28
Green, A. E., A. Lissina, S. L. Hutchinson, R. E. Hewitt, B. Temple, D. James, J. M. Boulter, D. A. Price, A. K. Sewell.
2004
. Recognition of nonpeptide antigens by human Vγ9Vδ2 T cells requires contact with cells of human origin.
Clin. Exp. Immunol.
136
:
472
-482.
29
Wei, H., D. Huang, X. Lai, M. Chen, W. Zhong, R. Wang, Z. W. Chen.
2008
. Definition of APC presentation of phosphoantigen (E)-4-hydroxy-3-methyl-but-2-enyl pyrophosphate to Vγ2Vδ2 TCR.
J. Immunol.
181
:
4798
-4806.
30
Sarikonda, G., H. Wang, K. J. Puan, X. H. Liu, H. K. Lee, Y. Song, M. D. Distefano, E. Oldfield, G. D. Prestwich, C. T. Morita.
2008
. Photoaffinity antigens for human γδ T cells.
J. Immunol.
181
:
7738
-7750.