Minimal requirement for induction of natural cytotoxicity and intersection of activation signals by inhibitory receptors

Abstract

Natural killer (NK) cells provide innate control of infected and neoplastic cells. Multiple receptors have been implicated in natural cytotoxicity, but their individual contribution remains unclear. Here, we studied the activation of primary, resting human NK cells by Drosophila cells expressing ligands for receptors NKG2D, DNAM-1, 2B4, CD2, and LFA-1. Each receptor was capable of inducing inside-out signals for LFA-1, promoting adhesion, but none induced degranulation. Rather, release of cytolytic granules required synergistic activation through coengagement of receptors, shown here for NKG2D and 2B4. Although engagement of NKG2D and 2B4 was not sufficient for strong target cell lysis, collective engagement of LFA-1, NKG2D, and 2B4 defined a minimal requirement for natural cytotoxicity. Remarkably, inside-out signaling induced by each one of these receptors, including LFA-1, was inhibited by receptor CD94/NKG2A binding to HLA-E. Strong inside-out signals induced by the combination of NKG2D and 2B4 or by CD16 could overcome CD94/NKG2A inhibition. In contrast, degranulation induced by these receptors was still subject to inhibition by CD94/NKG2A. These results reveal multiple layers in the activation pathway for natural cytotoxicity and that steps as distinct as inside-out signaling to LFA-1 and signals for granule release are sensitive to inhibition by CD94/NKG2A.

Figure 1

Engagement of individual activation receptors on resting NK cells induces inside-out signals. Resting NK cells were mixed with target cells as indicated. Where indicated, S2 cells were preincubated with a rabbit anti-S2 serum (+ IgG). Cells were incubated for 5 minutes at 37°C, stained with conformation-specific, biotinylated anti–LFA-1 mAbs, washed, and stained with fluorochrome-conjugated anti-CD56 and streptavidin. (A-B) NK cells were gated on forward scatter/side scatter plots, and the profiles show CD56 versus 327C or mAb24 mAb staining, as indicated. Gates indicate the percentage of 327C^high or mAb24^high NK cells. The profiles are representative of 3 or more independent experiments. (C-D) The percentage of 327C^high or mAb24^high CD56^dim NK cells is presented as the mean of 6 donors. Bars indicate SD. *P < .01.

Figure 2

Engagement of individual activation receptors on resting NK cells does not induce degranulation. Resting NK cells were mixed with target cells as indicated. Where indicated, S2 cells were preincubated with a rabbit anti-S2 serum (+ IgG). Cells were incubated for 2 hours at 37°C, then stained with fluorochrome-conjugated anti-CD56 and anti-CD107a mAbs. The percentage increase of CD107a⁺ NK cells after incubation with target cells relative to CD107a⁺ NK cells upon incubation of NK cells alone (ΔCD107a⁺) is presented as the mean of 8 donors. represents target cells as indicated, whereas ■ represents target cells coexpressing ICAM-1 in addition to the other ligands, as indicated. Bars denote SD.

Figure 3

Activation receptor coengagement on resting NK cells enhances CD16-induced degranulation. Resting NK cells were mixed with target cells as indicated. Where indicated, S2 cells were preincubated with a rabbit anti-S2 serum (+ IgG). Cells were incubated for 2 hours at 37°C, thereafter stained with fluorochrome-conjugated anti-CD56 and anti-CD107a mAbs. (A) The percentage of ΔCD107a⁺ CD56^dim NK cells is presented and values are representative of more than 3 independent experiments.

Figure 4

Resting NK-cell degranulation is facilitated by synergy among activation receptors. Resting NK cells were mixed with target cells as indicated. Cells were incubated for 2 hours at 37°C, then stained with fluorochrome-conjugated anti-CD56 and anti-CD107a mAbs. (A) NK cells were gated on forward scatter/side scatter plots, and the profiles show CD56 versus CD107a mAb staining. Gates indicate the percentage of ΔCD107a⁺ NK cells. (B) The percentage of ΔCD107a⁺ NK cells is presented as the mean of 7 donors. Bars indicate SD. **P < .005. (C) The percentage of ΔCD107a⁺ NK cells from 1 representative experiment is shown. Experiments are representative of at least 3 independent experiments.

Figure 5

Coengagement of NKG2D, 2B4, and LFA-1 induces natural cytotoxicity by resting NK cells. Resting NK cells were mixed with (A) S2 (□), S2–ICAM-1 (), S2-CD48-ULBP1 (△), or S2–ICAM-1–CD48–ULBP1 (○) cells, or (B) S2 cells as in panel A preincubated with a rabbit serum raised against S2 cells (■, ♦, ▲, ●). Cells were incubated for 3 hours at 37°C. Specific lysis of S2 cells was calculated from the percentage of propidium iodide–positive S2 cells in duplicate samples, as determined by flow cytometry. The experiment is representative of 3 or more independent experiments.

Figure 6

Inside-out signals on resting NK cells are inhibited by CD94/NKG2A engagement. Resting, NKG2C-depleted NK cells were mixed with target cells as indicated. Where indicated, S2 cells were preincubated with a rabbit anti-S2 serum (+ IgG). Cells were incubated for 5 minutes at 37°C, stained with conformation-specific, biotinylated anti–LFA-1 mAb 327C, washed, and then stained with fluorochrome-conjugated anti-CD56, anti-NKG2A, and streptavidin. (A) NK cells were gated on forward scatter/side scatter plots and CD56^dim expression. Where indicated, S2 cells were preincubated with peptides sp B*27 (blue) or sp C*03 (red). The profiles show NKG2A versus 327C mAb staining on CD56^dim NK cells. Numbers indicate the percentage of 327C^high CD56^dim NK cells within the respective gates. (B) The percentage of 327C^high CD56^dim NK cells is presented as the mean of 4 donors. Bars indicate SD. *P < .05. Experiments are representative of at least 3 independent experiments.

Figure 7

Resting NK-cell degranulation induced by 2B4 and NKG2D synergy or CD16 is inhibited by CD94/NKG2A engagement. Resting NK cells were mixed with S2, S2–HLA-E, or S2–CD48–ULBP1–HLA-E that had been preincubated with peptides as specified. For Fc-receptor stimulation, S2 cells were also preincubated with a rabbit anti-S2 serum (+ IgG). Cells were incubated for 2 hours at 37°C, then stained with fluorochrome-conjugated anti-CD56, anti-CD107a, and anti-NKG2A mAbs. NK cells were gated on forward scatter/side scatter plots and CD56^dim expression. (A) The profiles show NKG2A versus CD107a mAb staining. (B-C) S2 cells were preincubated with serial dilutions of a rabbit anti-S2 serum. The percentage of ΔCD107a⁺ NK cells from 1 representative experiment is shown. Experiments are representative of 3 independent experiments.