Exosomes Derived From Natural Killer Cells Exert Therapeutic Effect in Melanoma

Abstract

Objective: Exosomes are nanovesicles that are released from normal and tumor cells and are detectable in cell culture supernatant and human biological fluids. Although previous studies have explored exosomes released from cancer cells, little is understood regarding the functions of exosomes released by normal cells. Natural killer (NK) cells display rapid immunity to metastatic or hematological malignancies, and efforts have been undertaken to clinically exploit the antitumor properties of NK cells. However, the characteristics and functions of exosomes derived from NK cells remain unknown. In this study, we explored NK cell-derived exosome-mediated antitumor effects against aggressive melanoma in vitro and in vivo. Methods: B16F10 cells were transfected with enhanced firefly luciferase (effluc) and thy1.1 genes, and thy1.1-positive cells were immunoselected using microbeads. The resulting B16F10/effluc cells were characterized using reverse transcriptase polymerase chain reaction (RT-PCR), western blotting, and luciferase activity assays. Exosomes derived from NK-92MI cells (NK-92 Exo) were isolated by ultracentrifugation and density gradient ultracentrifugation. NK-92 Exo were characterized by transmission electron microscopy and western blotting. We also performed an enzyme-linked immunosorbent assay to measure cytokines retained in NK-92 Exo cells. The in vitro cytotoxicity of NK-92 Exo against the cancer cells was determined using a bioluminescence imaging system (BLI) and CCK-8 assays. To investigate the possible side effects of NK-92 Exo on healthy cells, we also performed the BLI and CCK-8 assays using the human kidney Phoenix™-Ampho cell line. Flow cytometry and western blotting confirmed that NK-92 Exo induced apoptosis in the B16F10/effluc cells. In vivo, we used a B16F10/effluc cell xenograft model to detect the immunotherapeutic effect of NK-92 Exo. We injected NK-92 Exo into tumors, and tumor growth progression was monitored using the IVIS Lumina imaging system and ultrasound imaging. Tumor mass was monitored after in vivo experiments. Results: RT-PCR and western blotting confirmed effluc gene expression and protein levels in B16F10/effluc cells. B16F10/effluc activity was found to increase with increasing cell numbers, using BLI assay. For NK-92 Exo characterization, western blotting was performed on both ultracentrifuged and density gradient-isolated exosomes. The results confirmed that NK cell-derived exosomes express two typical exosome proteins, namely CD63 and ALIX. We demonstrated by western blot analysis that NK-92 Exo presented two functional NK proteins, namely perforin and FasL. Moreover, we confirmed the membrane expression of FasL. The enzyme-linked immunosorbent assay results indicated that NK-92 Exo can secrete tumor necrosis factor (TNF)-α, which affected the cell proliferation signaling pathway. The antitumor effect of NK-92 Exo against B16F10/effluc cells in vitro was confirmed by BLI (p < 0.001) and CCK-8 assays (p < 0.001). Furthermore, in normal healthy cells, even after 24 h of co-culture, NK-92 Exo did not exhibit significant side effects. In the in vivo experiments, tumors in the vehicle control group were significantly increased, compared with those in the NK-92 Exo-treated group (p < 0.05). Conclusion: The results of the current study suggest that exosomes derived from NK cells exert cytotoxic effects on melanoma cells and thus warrant further development as a potential immunotherapeutic strategy for cancer.

Fig 1

Verification of successful isolation of exosomes. Transmission electron microscopy and western blotting were used to evaluate the phenotype of exosomes recovered from supernatant of NK-92MI cells after culturing them for 3 d. (A) The morphology of exosomes in the recovered pellet was analyzed by transmission electron microscopy (scale bar, 200 nm). (B) The expression of exosome markers (CD63 and ALIX) and negative markers (GM130 and β-actin) was confirmed by western blotting.

Fig 2

Identification of apoptosis-inducing proteins in NK and NK-92 Exo. Fifty micrograms of total proteins from NK cell and NK-92 Exo lysates were loaded into each lane (n = 3). (A) and (B) The apoptosis-inducing proteins FasL and perforin were expressed in both NK cell and NK-92 Exo lysates. FasL was also detected in a membrane proteins enriched extract of NK-92 Exo. (C) and (D) The abundances of FasL and perforin in NK and NK-92 Exo were estimated using the software EvolutionCapt. NK-92 Exo lysates contained higher abundances of FasL and perforin than the NK cell lysates did. Experiments were performed at least in triplicate and values ± SD were plotted. *p < 0.05.

Fig 3

NK cell-derived exosome cytotoxicity against melanoma cells. B16F10/enhanced firefly luciferase (effluc) melanoma cells were co-cultured with two concentrations of NK-92 Exo (5 and 20 µg) and bioluminescence signals were measured at various time points (4, 10, and 24 h). The viability of B16F10/effluc cells decreased at 10 h and 24 h after the co-incubation. Experiments were performed at least in triplicate and values ± SD were plotted. *p < 0.05, ***p < 0.001.

Fig 4

Effects of NK-92 Exo on the proliferation of normal cells. Phoenix-A cells, which are normal human kidney cells, were co-cultured with two concentrations of NK-92 Exo (5 and 20 µg) and bioluminescence signals were measured at 4, 10, and 24 h. NK-92 Exo showed no significant cytotoxicity against the phoenix-A cells even after 24 h (A). To confirm the bioluminescent imaging (BLI) results, a CCK-8 assay was also performed, and the results were consistent with those of BLI (B). Experiments were performed at least in triplicate and values ± SD were plotted.

Fig 5

Detection of cytotoxic activity of NK-92 Exo. (A) The cytotoxic effects of NK-92 Exo on B16F10/effluc cells were inhibited by a FasL inhibitor (AF016). (B) Quantitation of the cytotoxic effect. (C) The results of the CCK-8 assay were consistent with those of the BLI assay. Experiments were performed in triplicate and mean values ± SD were plotted. *p <0.05, **p <0.01, ***p <0.001.

Fig 6

Interaction of NK-92 Exo with melanoma cells. The B16F10/effluc cells were co-incubated with vehicle, NK-92 Exo, and fluorescent dye (DiD)-labeled NK-92 Exo (10 µg). Confocal microscopy images were obtained after 3, 6, and 12 h co-incubation.

Fig 7

NK-92 Exo induced apoptosis in melanoma cancer cells in vitro. (A) The apoptosis of B16F10/effluc cells induced by NK-92 Exo was assessed by flow cytometric analysis using Annexin V. (B) Quantitation of the scatter plots. The apoptotic rate of B16F10/effluc cells was significantly increased by NK-92 Exo. Experiments were performed in triplicate and mean values ± SD were plotted. *p < 0.05, **p < 0.01.

Fig 8

Mechanism of the apoptotic induction in B16F10/effluc cells by NK-92 Exo. (A) Diagram of the apoptosis signaling pathway by NK-92 Exo. (B) Western blot analysis of important proteins in the apoptosis signaling pathway. With NK-92 Exo treatment, the abundances of cleaved caspase-3, cleaved PARP, and cytochrome c increased by 3.78-fold 3.49-fold, and 1.66-fold, respectively.

Fig 9

Anti-tumor effect of NK-92 Exo on melanoma cancer cells in vivo and ex vivo. (A) BLI measurements of the B16F10/effluc activity in mice were performed in the control and treatment groups. In the treatment group, 20 µg of NK-92 Exo was injected into tumors twice. (B) Quantitative BLI of B16F10/effluc activity. (C) BLI of the ex vivo tumor activity. (D) Quantitative BLI of ex vivo B16F10/effluc activity. Values are expressed as the mean ± SD, *p < 0.05 **p < 0.01.

Fig 10

Ultrasound imaging of tumors in vivo. After collecting the B model imaging of (A), the tumor was imaged by ultrasonography. (B) Tumor volumes were measured by 3-D imaging obtained using software. (C) Tumor masses (g) were measured after mice were sacrificed. Values are expressed as the mean ± SD, ***p < 0.001.