Adoptively-Infused NK Cells Maintain Their Antitumor Effects in Vivo in the Presence of CyclosporineA (CSA)

Abstract

Animal models show infusions of donor NK cells given after allogeneic HCT can prevent GVHD while simultaneously mediating a graft-vs-tumor effect. However, it is unclear whether adoptively infused NK cells can mediate these beneficial effects in the presence of CSA, which is commonly given after HCT to prevent GVHD. In this study, we analyzed the in vitro effects of pharmacological concentrations of CSA on NK cell phenotype, cell proliferation, and tumor cytotoxicity. We also evaluated in vivo whether CSA administration would reduce the anti-tumor effects of adoptively infused NK cells in tumor bearing mice. PBMCs collected from healthy donors were labeled with CFSE then were stimulated in vitro with IL-2 for 7 days in the presence or absence of CSA (1000ng/ml). CFSE proliferation assays on fresh PBMC showed CSA inhibited IL-2 stimulated CD3+ T-cell proliferation more than CD3−/CD56+ NK cell proliferation (mean percentage inhibition of proliferation 49.4% vs. 22.2% for T cells and NK cells respectively; p<0.05). CD3−/CD56+ NK cells were then isolated from PBMCs of healthy donors and expanded in vitro with irradiated EBV-LCL and IL-2 for 10 days. In contrast to T-cells, CSA only minimally inhibited IL-2 induced proliferation of expanded NK cells (mean 9.5% inhibition of proliferation by CFSE staining). T cells and NK cells were next isolated from PBMCs and stimulated with either OKT3, PMA-Ionomycin (PI), or IL-2. A [3H] TdR uptake assay showed T cell proliferation was inhibited at a substantially higher level by CSA (mean stimulation index for OKT3: 0.03, for PI: 0.35, for IL-2: 0.55) compared to that of expanded NK cells (mean stimulation index for IL-2: 0.82, p<0.05). Furthermore, an ELISA assay showed CSA treatment reduced IL-2 induced secretion of INF-g by T cells more than expanded NK cells (mean reduction in INF-g secretion in T cells of 94.7 % vs. 36.5 % in NK cells, p<0.05). Compared to controls, culturing in vitro expanded NK cells in CSA did not alter surface expression of the activating receptors NKp30, NKp42, and NKG2D but did reduce surface expression of NKp44 and TRAIL (mean reduction in surface expression 36% and 36.3% respectively). Cytotoxic granule release assessed by CD107a staining was inhibited by CSA in CD8+ melanoma specific T cells co-cultured with melanoma cells (mean 12.2 % inhibition) in contrast to NK cells co-cultured with K562 cells where CSA increased CD107a expression a mean 29.7% (p<0.05). Furthermore, at a 20:1 E:T ratio, 51Cr cytotoxity assays showed CSA did not reduce the cytotoxicity of in vitro expanded NK cells against renal cell carcinoma (RCC) cells (58% mean lysis) compared to NK cells cultured in control media (55% mean; p=n.s.). In contrast, melanoma specific T-cell killing of tumor targets was significantly lower in CTL cultures containing CSA compared to control media (38.0% vs. 50.2% respectively P<0.05). Next, we assessed the impact of CSA administration on the anti-tumor effects of adoptive NK cell infusions in tumor bearing animals where syngeneic NK cell infusions following bortezomib treatment have been shown to delay tumor progression and prolong survival. BALB/c mice injected with 100,000 luciferase transfected RENCA tumor cells i.v. received 3 weekly treatments with the combination of bortezomib (5ug/mouse i.v.) and 2×106 syngeneic NK cells i.v. with or without daily administration of CSA (15mg/kg sc). Bioluminescence imaging in controls that did not receive CSA showed tumor growth was slower and survival was prolonged in mice receiving adoptive NK cell infusions (median survival 53 days) compared to mice that did not receive NK cells (median survival 30.2 days; p<0.05). CSA administration did not impair the anti-tumor effects of adoptive NK cell infusions; mice receiving CSA and adoptive NK cell infusions had similar tumor growth and survival as recipients of NK cells without CSA (median survival 47 vs. 53 days; p=n.s.). These results show that CSA is significantly more immunosuppressive to T cells compared to NK cells and provide evidence that the anti-tumor effects of adoptively infused in vitro expanded NK cells are maintained even in the presence CSA.