Lipopolysaccharide induces IFN-γ production in human NK cells

Abstract

Natural killer (NK) cells have been shown to play a regulatory role in sepsis. According to the current view, NK cells become activated via macrophages or dendritic cells primed by lipopolysaccharide (LPS). Recently, TLR4 gene expression was detected in human NK cells suggesting the possibility of a direct action of LPS on NK cells. In this study, effects of LPS on NK cell cytokine production and cytotoxicity were studied using highly purified human NK cells. LPS was shown to induce IFN-γ production in the presence of IL-2 in NK cell populations containing>98% CD56(+) cells. Surprisingly, in the same experiments LPS decreased NK cell degranulation. No significant expression of markers related to blood dendritic cells, monocytes or T or B lymphocytes in the NK cell preparations was observed; the portions of HLA-DR(-bright), CD14(+), CD3(+), and CD20(+) cells amounted to less than 0.1% within the cell populations. No more than 0.2% of NK cells were shown to be slightly positive for surface TLR4 in our experimental system, although intracellular staining revealed moderate amounts of TLR4 inside the NK cell population. These cells were negative for surface CD14, the receptor participating in LPS recognition by TLR4. Incubation of NK cells with IL-2 or/and LPS did not lead to an increase in TLR4 surface expression. TLR4(-)CD56(+) NK cells isolated by cell sorting secreted IFN-γ in response to LPS. Antibody to TLR4 did not block the LPS-induced increase in IFN-γ production. We have also shown that R(e)-form of LPS lacking outer core oligosaccharide and O-antigen induces less cytokine production in NK cells than full-length LPS. We speculate that the polysaccharide fragments of LPS molecule may take part in LPS-induced IFN-γ production by NK cells. Collectively our data suggest the existence of a mechanism of LPS direct action on NK cells distinct from established TLR4-mediated signaling.

Keywords: NK cells; cytokine production; cytotoxicity; flow cytometry; lipopolysaccharide.

FIGURE 1

Co-stimulating influence of LPS on NK cell IFN-γ secretion. Human NK cells were stimulated for 18 h with cytokines and LPS. Supernatants were harvested and analyzed for IFN-γ by ELISA. Data are mean ± SD of triplicate wells in one representative individual experiment. *P < 0.05, ***P < 0.001, comparisons were shown between cells, stimulated by a cytokine with LPS, and cells, stimulated by the cytokine alone. (A) NK cells were incubated in the presence or absence of IL-2 (500 U/ml) or IL-12 (10 ng/ml) and LPS (5 μg/ml). Combination of IL-2 and IL-12 was used as a positive control. (B) NK cells were stimulated with IL-2 (500 U/ml) and different doses of LPS. Results of one experiment are shown from the group of experiments demonstrating an increase of IFN-γ secretion under LPS action (12 different donors). Other experiments showed similar effects but differed in IFN-γ level.

FIGURE 2

Cytometric analysis of NK cell population purity and expression of TLR4. (A) Freshly isolated NK cells were stained with fluorochrome-labeled antibodies CD3-FITC, CD14-PE, CD20-PerCP, and CD56-APC and analyzed by flow cytometry. (B) Here is shown one of typical results from 18 blood samples used in these experiments. Expression of TLR4 was measured using anti-TLR4-FITC antibody (clone HTA125). Forward scatter (FS) versus side scatter (SS) is presented. Three regions of live cells can be distinguished based on morphology. R1 region (95.15% of all events) consists of CD3^-CD56⁺ NK cells (98.6%). CD56^- cells do not express TLR4. A small population of CD56⁺TLR4^dim cells (±0.1%) is detected. On the histogram TLR4 level and isotypic control are shown. R2 region (0.4%) consists of NK cells doublets. R3 region (0.48%) corresponds to granulocytes. (C) Intracellular TLR4 expression in NK cells. Black line – autofluorescence control, green – isotype control, red – fixed cells, labeled with anti-TLR4-FITC.

FIGURE 3

Expression of HLA-DR was analyzed in different cell samples. Freshly isolated NK cells were stained with fluorochrome-labeled antibodies CD56-APC and HLA-DR-FITC and analyzed by flow cytometry. Donor A: one of typical NK cell preparations with very small amount of HLA-DR^bright blood DC cells. Donor B: NK cell preparation containing a fraction of HLA-DR^bright blood DC cells.

FIGURE 4

NK cells were stained with fluorochrome labeled antibodies and sorted using FACSVantage machine. (A) Sort scheme is presented. Gate for lymphocytes was used, then cells were negatively gated for CD3 and CD14, than TLR4-negative CD56⁺ cells were chosen for sorting. (B) Post-sort purity of CD56⁺ cells was no less than 95%, no TLR4-positive cells were detected. (C) Histograms represent three independent experiments with NK cells from three donors. NK were isolated by cell sorting, than stimulated with IL-2 (500 U/ml) and LPS (1 μg/ml) in duplicate wells. Supernatants were harvested and analyzed for IFN-γ by ELISA. Data are mean ± SD, *P < 0.05, ***P < 0.001.

FIGURE 5

NK cells isolated by magnetic separation were preincubated with anti-TLR4 antibody (clone HTA125; 5 μg/ml) for 40 min 37°C, than stimulated by IL-2 (500 U/ml) and different doses of LPS. Supernatants were harvested and analyzed for IFN-γ by ELISA. Data are mean ± SD from one of three experiments with similar results.

FIGURE 6

Comparison of effects of S- and R_e-forms of LPS on IFN-γ production by NK cells. (A) R_e-form of LPS (lipid A) does not alter significantly IFN-γ production in NK cells at the same concentration as the S-form (full-length LPS; 5 μg/ml). Data from one of four experiments with similar results are presented. (B) Significant increase of IFN-γ secretion is induced by low dose of lipid A (10 ng/ml) in NK cell preparation containing a fraction of HLA-DR^bright blood DCs (donor B, Figure 3).