Effects of interleukin-18 on natural killer cells: costimulation of activation through Fc receptors for immunoglobulin

Abstract

The antitumor activity of monoclonal antibodies is mediated by effector cells, such as natural killer (NK) cells, that express Fc receptors for immunoglobulin. Efficacy of monoclonal antibodies, including the CD20 antibody rituximab, could be improved by agents that augment the function of NK cells. Interleukin (IL)-18 is an immunostimulatory cytokine that has antitumor activity in preclinical models. The effects of IL-18 on NK cell function mediated through Fcγ receptors were examined. Human NK cells stimulated with immobilized IgG in vitro secreted IFN-γ as expected; such IFN-γ production was partially inhibited by blocking CD16 with monoclonal antibodies. IL-18 augmented IFN-γ production by NK cells stimulated with immobilized IgG or CD16 antibodies. NK cell IFN-γ production in response to immobilized IgG and/or IL-18 was inhibited by chemical inhibitors of Syk and several other kinases involved in CD16 signaling pathways. IL-18 augmented antibody-dependent cellular cytotoxicity (ADCC) of human NK cells against rituximab-coated Raji cells in vitro. IL-18 and rituximab acted synergistically to promote regression of human lymphoma xenografts in SCID mice. Inasmuch as IL-18 costimulates IFN-γ production and ADCC of NK cells activated through Fc receptors in vitro and augments antitumor activity of rituximab in vivo, it is an attractive cytokine to combine with monoclonal antibodies for treatment of human cancer.

Fig. 1

IL-18 and immobilized IgG costimulate IFN-γ production by human NK cells. a Enriched human NK cells were incubated overnight in medium alone or medium containing 1, 10, or 100 ng/mL IL-18 as indicated and then plated overnight in the presence (striped bars) or absence (solid bars) of immobilized IgG. Supernatant IFN-γ levels were measured by ELISA. P value for comparison of medium alone versus medium plus IgG is 0.001 and for comparisons of IgG alone versus IgG plus IL-18 1, 10, or 100 ng/mL is 0.312, 0.033, 0.012, respectively. b Enriched human NK cells were incubated in wells precoated or not with immobilized IgG in medium with or without 100 ng/mL IL-18 as indicated in the presence (striped bars) or absence (solid bars) of F(ab′)2 fragments of CD16 monoclonal antibody 3G8. P value for comparison of presence versus absence of CD16 F(ab′)2 fragments for medium, IgG, IL-18, or IgG plus IL-18 is 0.023, 0.013, 0.019, 0.006, respectively. Results shown represent mean ± SE of data from 3 (a) or 5 (b) separate experiments

Fig. 2

Effect of chemical inhibitors on IFN-γ production by human NK cells. Enriched NK cells were stimulated with 100 ng/mL IL-18 alone (darker solid bars), immobilized IgG alone (lighter solid bars), or both agents (striped bars) in the absence of chemical inhibitors (“Control”) or in the presence of 10 μM Wortmannin, U0126, or SB203580 as indicated. P value for comparison of control versus inhibitor is <0.0001 for each inhibitor when comparing IFN-γ levels produced by NK cell stimulated with IgG alone or IgG plus IL-18. Results shown represent mean ± SE of data from 5 separate experiments

Fig. 3

IL-18 enhances ADCC mediated by human NK cells against rituximab-sensitized lymphoma cells. Enriched human NK cells were incubated overnight in medium with or without 100 ng/mL IL-18 as indicated and then plated at an effector-to-target cell ratio of 10:1 with Raji cells that had been pre-treated or not with rituximab as indicated. P value for comparison of rituximab versus medium is 0.008, and for comparison of rituximab plus IL-18 versus rituximab alone or IL-18 alone is 0.03 and 0.003, respectively. Results shown represent mean ± SE of data from 3 separate experiments

Fig. 4

IL-18 enhances the efficacy of rituximab in SCID mice with human B cell lymphoma xenografts. SCID–ICR mice (n = 6/group) bearing subcutaneous human Ramos lymphoma xenografts were treated with IL-18 or rituximab monotherapy, or with combination of both agents. a Tumor growth in combination therapy groups was significantly suppressed (*P < 0.05, **P < 0.01, ***P < 0.001, two-way ANOVA followed by Bonferroni post-test) as compared to the rituximab monotherapy groups. Moreover, the number of complete tumor regressions was significantly increased in combination therapy groups as compared to rituximab (Table 2) or IL-18 monotherapy (0/6). Mice in the control group in some of the monotherapy groups (#) had to be euthanized due to ethical reasons before the study ended. b Tumor volumes on day 25 show statistically significant tumor growth suppression in all combination therapy groups as compared to their respective rituximab (*P < 0.05, **P < 0.01, t test) monotherapy groups and to the IL-18 monotherapy group (**P < 0.01, one-way ANOVA followed by Bonferroni post-test)