Soy isoflavones and their metabolites modulate cytokine-induced natural killer cell function

Abstract

Soybeans are a rich source of isoflavones that have been linked with anti-inflammatory processes and various health benefits. However, specific mechanisms whereby soy bioactives impact immune cell subsets are unclear. Isoflavones, such as genistein and daidzein, are metabolized by microbes to bioactive metabolites as O-desmethylangolensin (O-DMA) and equol, whose presence has been linked to health benefits. We examined how soy isoflavones and metabolites impact natural killer (NK) cell signaling and function. We observe no impact of isoflavones on viability of healthy donor peripheral blood mononuclear cells (PBMCs) or NK cells, even at high (25 µM) concentrations. However, pre-treatment of PBMCs with physiologically-relevant concentrations of genistein (p = 0.0023) and equol (p = 0.006) decreases interleukin (IL)-12/IL-18-induced interferon-gamma (IFN-γ) production versus controls. Detailed cellular analyses indicate genistein and equol decrease IL-12/IL-18-induced IFN-γ production by human NK cell subsets, but do not consistently alter cytotoxicity. At the level of signal transduction, genistein decreases IL-12/IL-18-induced total phosphorylated tyrosine, and phosphorylation MAPK pathway components. Further, genistein limits IL-12/IL-18-mediated upregulation of IL-18Rα expression on NK cells (p = 0.0109). Finally, in vivo studies revealed that C57BL/6 mice fed a soy-enriched diet produce less plasma IFN-γ following administration of IL-12/IL-18 versus control-fed animals (p < 0.0001). This study provides insight into how dietary soy modulates NK cell functions.

Figure 1

Soy isoflavones and their metabolite compounds. Genistein and daidzein are two of the most abundant isoflavones found in soy. Daidzein can be further metabolized into secondary compounds, O-demthylangolensin (O-DMA) and Equol.

Figure 2

Soy isoflavones and their metabolites do not affect immune cell viability. Healthy human donor PBMCs (a–d) and CD56⁺ NK cells (e) were incubated for 72 hours with genistein, daidzein, O-DMA, and equol. Cells were then assessed for viability.

Figure 3

Genistein and equol abrogate IL-12/IL-18 induced IFN-γ production by human primary PBMCs. Human primary PBMCs were cultured with (a) genistein, (b) daidzein, (c) O-DMA, and (d) equol for 4 hours and then stimulated with 20 ng/ml IL-12 and 50 ng/ml IL-18 for 72 hours. Culture supernatants were tested for IFN-γ by ELISA. Data is representative from 5 healthy donors. Means +/− STD. *p < 0.05, 25 μM concentration compared to DMSO; ^†p < 0.05, 10 µM concentration compared to DMSO; ^‡p < 0.05, 25 µM concentration compared to 10 µM.

Figure 4

Genistein and Equol decrease IL-12/IL-18 induced NK cell IFN-γ production. Human primary PBMCs were cultured with soy compounds for 4 hours and then stimulated with 20 ng/ml IL-12 and 50 ng/ml IL-18 for 72 hours. (a) Cells were initially gated based on forward and side scatter to exclude debris, then cells were gated based on CD3 and CD56 expression as NK cells (CD3⁻CD56^Bright NK cells, CD3-CD56^Dim NK cells, or CD3⁺CD56⁺NKT cells based on evident positive and negative cell populations. Intracellular IFN-γ was then assessed in these populations via intracellular flow cytometry. The positive gate was set based on the level of fluorescence observed in unstimulated cells, which coincided with the level of fluorescence observed in unstained control cells (not pictured). (b) IFN-γ intracellular staining in CD3⁻NK^Dim (open) and CD3⁻CD56^Bright (shaded) NK cells. Cells were pre-treated prior to IL-12/IL-18 stimulus in DMSO (0 μM), (b) genistein, (c) daidzein, (d) O-DMA or (e) equol. Data is representative from 3 healthy donors. Means +/− STD. *p < 0.05, concentration compared to DMSO; ^†p < 0.05, 25 µM concentration compared to 10 µM.

Figure 5

Soy compounds do not affect NK cell cytotoxicity. Human primary NK cells were isolated from healthy donor PBMCs and were cultured with 25 µM of soy compounds (Genistein, Daidzein, O-DMA, Equol) for 4 hours then (a) unstimulated or (b) stimulated with IL-12 (10 ng/ml) overnight. NK cells were co-cultured with ⁵¹Cr labeled target cells (K562) for 4 hours and chromium release was analyzed to determine percent cytotoxicity. Data is representative from 4 healthy donors.

Figure 6

Genistein modulates cellular signaling events. PBMC from healthy donors were cultured with 25 µM of genistein or DMSO (vehicle control) for 4 hours then stimulated with 20 ng/ml IL-12 and 50 ng/ml IL-18 for either 24 hours of additional incubation (long-term), or 15 minutes of additional incubation (short-term) to analyze canonical signal transduction events. (a) Immunoblot analysis of pERK was conducted in short-term lysates to focus on signaling events proximal to initial IL-18 receptor engagement. (b) Densitometry analysis of western blot pERK levels across conditions normalized to total ERK and β-actin levels. (c) Immunoblot analysis of pSTAT1 was conducted in long-term lysates where IFN-γ, an upstream STAT1 activating stimulus was at high levels (d) Densitometry analysis of western blot phosphor-STAT1 levels across conditions normalized to total STAT1 and β-actin levels. (e) Immunoblot analysis of pTyr was conducted in long-term lysates where IFN-γ and pan-pTyr activation was at high levels. Data are presented from three separate donors with similar results. In two of three experiments, total STAT1 or ERK, along with β-actin are included as controls and were run using identical lysates on separate gels, given similar molecular weight of some proteins. In one of three experiments, each phospho-specific Ab blot was stripped and re-probed for detection of loading controls. (Veh = vehicle; 12/18 = IL = 12 + IL-18 stimulation) Immunoblots were cropped for presentation purposes and original blots are provided as Supplemental Fig. 4.

Figure 7

Genistein reduces expression of IL-12/18 induced expression of IL-18Rα. PBMCs from healthy human donors (n = 5) were cultured with 10 or 25 µM of genistein or DMSO (vehicle control) for 4 hours then stimulated with 20 ng/ml IL-12 and 50 ng/ml IL-18 for either 24 hours of additional incubation. IL-12 receptor (IL-12Rβ1) and IL-18 receptor (IL-18Rα) surface expression was analyzed by flow cytometry. Representative dot plots of CD56⁺ NK cells expressing (a) IL-12Rβ1 and (b) IL-18Rα. (c) IL-12Rβ1 or (d) IL-18Rα expression from 5 different donors was analyzed by fold change of unstimulated over IL-12/18 stimulated NK cells. Means +/− STD. *p < 0.01, concentration compared to DMSO control (0 μM).

Figure 8

Soy-enriched diet inhibits IL-12/IL-18 induced IFN-γ responses in vivo. C57BL/6 mice were (a) administered a control or a 0.3% soy-enriched diet for 7 days and then intraperitoneally injected with 0.1ug IL-12 and 1ug IL-18 on days 7 and 8. (b) On day 9 mice were euthanized and plasma was assessed by ELISA for IFN-γ levels. Means +/− STD, *p < 0.0001, soy fed mice compared to control fed mice.