Human gammadelta T lymphocytes induce robust NK cell-mediated antitumor cytotoxicity through CD137 engagement

Abstract

Natural killer (NK) cells are innate effector lymphocytes that control the growth of major histocompatibility complex class I negative tumors. We show here that γδ T lymphocytes, expanded in vitro in the presence isopentenylpyrophosphate (IPP), induce NK cell-mediated killing of tumors that are usually resistant to NK cytolysis. The induction of cytotoxicity toward these resistant tumors requires priming of NK cells by immobilized human immunoglobulin G1 and costimulation through CD137L expressed on activated γδ T lymphocytes. This costimulation increases NKG2D expression on the NK-cell surface, which is directly responsible for tumor cell lysis. Moreover, culturing peripheral blood mononuclear cells with zoledronic acid, a γδ T lymphocyte activating agent, enhances NK-cell direct cytotoxicity and antibody-dependent cellular cytotoxicity against hematopoietic and nonhematopoietic tumors. Our data reveal a novel function of human γδ T lymphocytes in the regulation of NK cell-mediated cytotoxicity and provide rationale for the use of strategies to manipulate the CD137 pathway to augment innate antitumor immunity.

Figures

Figure 1

γδ T lymphocytes increase hIgG1-induced activation of NK cells. (A) Purified natural killer (NK) cells and isopentenylpyrophosphate plus interleukin-2 (IPP+IL-2) expanded γδ T lymphocytes were cocultured at a 4:1 ratio in the presence or absence of plate-immobilized human immunoglobulin G1 (hIgG1; 2.5 μg/mL) for 48 hours. The expression of CD69 was analyzed by flow cytometry. Histograms represent gated CD3−CD56+ NK cells. An example of gating strategy is shown in supplemental Figure 1. (B) Purified NK cells (2 × 105 cells/well) were cultured with indicated numbers of IPP+IL-2 expanded γδ T lymphocytes in the presence of immobilized hIgG1 (2.5 μg/mL) for 48 hours. Expression of activation markers CD69 and CD54 were assessed by flow cytometry and plotted as a percentage of CD69 and CD54 positive gated NK cells. Representative data from 1 of 10 independent experiments is shown.

Figure 2

γδ T lymphocytes induce NK cell–mediated cytotoxicity against various tumor cell lines. (A) Purified NK cells were cultured with bulk IPP-expanded peripheral blood mononuclear cells (PBMCs) at a 2:1 ratio for 48 hours (ie, 2 NK cells/1 IPP-expanded PBMC). IPP-expanded PBMCs in these experiments contained 70% to 80% γδ T lymphocytes, so the actual ratio of NK cells to IPP-expanded γδ T lymphocytes was approximately 2:0.8. NK cells were repurified from the cultures by immunomagnetic depletion of non-NK cells. Representative dot plots of NK + γδ T cells (right) and NK cells purified after 48 hours of culture (left) are shown. (B) Cytolytic activity of NK cells purified after coculture with expanded γδ T cells and immobilized hIgG1 (2.5 μg/mL) for 48 hours was analyzed in a standard 4-hour 51Cr-release assay against indicated tumor targets. Data are presented as mean ± SD of triplicate samples and are representative of 7 independent experiments. *P < .05 compared with NK cells cultured with hIgG1 alone.

Figure 3

Cell-to-cell contact is required for the activation of NK cells by γδ T lymphocytes. Trans-well experiments were performed by culturing purified NK cells in lower wells coated with hIgG1 (2.5 μg/mL). Expanded γδ T lymphocytes were added either to the lower wells (cell-to-cell contact) or to the upper wells (soluble factors). The ratio of NK to γδ T cells was 4:1. After 48 hours of culture, the expression of CD69 and CD54 was analyzed by flow cytometry. The bar diagrams depict the percentage of CD69 and CD54 expressing cells in gated NK populations. Representative data from 1 of 3 independent experiments is shown.

Figure 4

Expression of costimulatory ligands and receptors on γδ T lymphocytes and NK cells. (A) Fresh γδT lymphocytes from healthy donors (top histograms) or γδT lymphocytes expanded in the presence of IPP+IL-2 (bottom histograms) were stained with mAbs specific for CD80, CD86, CD252 (OX40L) and CD137L (41BBL). The expression of indicated costimulatory ligands on gated CD3+γδTCR+ cells is shown. (B) NK cells cultured with media alone or immobilized hIgG1 with or without in vitro expanded γδ T lymphocytes for 48 hours were stained with mAbs specific for CD28, CD152 (CTLA-4), CD134 (OX40), and CD137 (4-1BB). Overlays of histograms representing gated CD3−CD56+ NK cells are shown. Depicted data represent 1 of 5 independent experiments.

Figure 5

Blocking of CD137L partially inhibits γδ T lymphocyte–induced cytolytic activity of NK cells. (A) Purified NK cells were cocultured with IPP+IL-2 expanded γδT lymphocytes at a 4:1 ratio in the presence of immobilized hIgG1 (2.5 μg/mL) for 48 hours. In some groups, soluble CD137Ig fusion protein at 10 μg/mL was added to block CD137 receptor and ligand interactions. The bar diagram represents the percentage of cells expressing CD54 in gated CD3−CD56+ NK-cell population. Representative data from 1 of 4 independent experiments is shown. (B) Purified NK cells were cocultured with either mock (left histograms) or CD137L-transfected P815 (middle histograms) at a 4:1 ratio in the presence of immobilized hIgG1 (2.5 μg/mL). In some wells containing NK cells and CD137L-transfected P815 tumors, soluble CD137Ig fusion protein (10 μg/mL) was added (right histograms). After 48 hours of culture, cells were stained for CD54 and CD25. Histograms represent cells gated on CD56+CD3− NK population. (C) Soluble CD137Ig fusion protein (10 μg/mL) was included during the culture of purified NK cells and γδ T lymphocytes (4:1 ratio) in hIgG1-precoated plates for 48 hours. Cytotoxicity of NK cells repurified after culture was analyzed in a standard 4-hour 51Cr-release assay against the TU167 SCCHN cell line. Data are presented as mean ± SD of triplicate samples and are representative of 4 independent experiments. *P < .05 compared with NK cells cultured in the presence of CD137Ig blocking. (D) NK cells were purified from PBMCs of healthy donors and cocultured with irradiated mock or CD137L-transfected P815 cells at a 4:1 ratio. Expanded γδ T lymphocytes were used as a positive control for NK-cell activation. After 48 hours of culture, NK cells were repurified and used as effectors against TU167 SSCHN target cells. Data are presented as mean ± SD of triplicate samples and representative of 3 independent experiments. *P < .05 compared with NK cells cultured with mock transfected P815.

Figure 6

CD137 ligation on NK cells results in enhanced NKG2D expression that is involved in tumor cell killing. (A) NK cells purified from PBMCs of 11 individual donors were cocultured in the presence of expanded γδ T lymphocytes (4:1 ratio) on plates precoated with hIgG1. After 48 hours of culture, the expression of NKG2D was analyzed on NK cells. Dots represent individual values of NKG2D expression on gated NK cells. Horizontal lines represent average values of NKG2D expression in indicated groups. (B) Cytotoxic activity of NK cells purified after 48 hours of culture with in vitro expanded γδ T lymphocytes was measured in a standard 4-hour 51Cr-release assay against TU167 squamous cell carcinoma of the head and neck (SCCHN). Blocking anti-NKG2D antibodies or isotype control IgG were added into the wells containing purified NK cells and TU167 targets for the duration of the cytotoxicity test. Data are presented as mean ± SD of triplicate samples and are representative of 2 independent experiments. *P < .05 compared with isotype control. (C) CD137Ig (10 μg/mL) was added to wells containing NK cells and IPP+IL-2 expanded γδ T lymphocytes. After 48 hours of culture, cells were stained with anti-NKG2D mAb. The histograms depict NKG2D expression on gated CD56+CD3− NK cells. Numbers in brackets indicate mean fluorescent intensity of NKG2D expression. (D) Purified NK cells were cultured with irradiated mock or CD137L-transfected P815 cells (4:1) on plates precoated with hIgG1. After 48 hours the expression of NKG2D was analyzed by fluorescence-activated cell sorting (FACS). Expanded γδ T lymphocytes were used as a positive control for NK-cell activation. Numbers in brackets indicate mean fluorescent intensity of NKG2D expression.

Figure 7

Zoledronate, a γδ T lymphocyte–activating agent, enhances NK-cell activation and cytotoxicity. (A) Purified NK cells were cultured for 96 hours in the presence of media, TU167 cells alone, TU167 + 10 μg/mL hIgG1 (isotype control), or TU167 + 10 μg/mL cetuximab. CD137 expression on CD56+ NK cells was analyzed by FACS. Two representative experiments are shown. (B) Whole PBMCs were incubated in the presence of media, 10 μg/mL rituximab, 15μM zoledronate, or a combination of rituximab with zoledronate. The expression of CD69 on gated CD3−CD56+ NK cells was analyzed by FACS 96 hours after initiation of the cultures. A representative of 3 independent experiments is depicted. (C) Whole PBMCs were cultured with media (circles) or zoledronate (squares). Alternatively, γδ T lymphocyte–depleted PBMCs were cultured with zoledronate (triangles) for 96 hours. NK cells were purified from the groups described. NK-cell direct cytotoxicity (left plots) or antibody-dependent cellular cytotoxicity (right plots) was measured in a standard 4-hour 51Cr-release assay against TU167 SCCHN or Ramos B-cell lymphoma targets. Data are presented as mean ± SD of triplicate samples and are representative of 2 independent experiments. *P < .05 compared with NK cells purified from γδ T lymphocyte–depleted cultures.