A distinct subpopulation of human NK cells restricts B cell transformation by EBV

Abstract

NK cells constitute the first line of defense against pathogens and transformed cells. They mature in secondary lymphoid organs, including tonsils, where common pathogens, such as EBV, enter the host and potentially imprint differentiating cells, which then patrol the body via the blood stream. Therefore, we set out to characterize a distinct human NK cell population in tonsils that produces high amounts of the immunomodulatory and antiviral cytokine IFN-γ. We found that the tonsilar IFN-γ(high) NK cell subset is CD56(bright)NKG2A(+)CD94(+)CD54(+)CD62L(-), is present in tonsils ex vivo and is more mature than other CD56(bright) NK cells in tonsils and less mature than other NK cells in blood, shows very low plasticity even after prolonged cytokine stimulation, accumulates in tonsils of EBV carriers, and is able to potently restrict EBV-induced transformation of B cells. Thus, we characterized a distinct and stable IFN-γ(high) NK cell subpopulation that can specifically restrict malignant transformation of EBV-infected B cells. This subset should be exploited for future development of cell-based therapeutic approaches in EBV-associated malignancies.

Figure 1. The human tonsilar IFN-γ^high NK cell immunoresponsive subset (IR) is characterized by being CD56^brightNKG2A⁺CD94⁺CD54⁺CD62L⁻

Purified NK cells were stimulated with 10 ng/ml IL-12, after 12 h monensin was added and cells where then analyzed after 18 h by flow cytometry. A: Gating strategy for analysis of distinct IFN-γ high Natural Killer cell subset after IL-12 stimulation. B: The IFN-γ high subset in tonsils is CD56^brightNKG2A⁺CD94⁺ and C: CD54⁺CD62L⁻. D: Gated on this subset, these Natural Killer cells are almost exclusively IFN-γ positive after IL-12 stimulation (10 ng/ml as above). Gating strategy: Lymphocyte population in FSC/SSC, Live cells (Live/Dead cell stain Aqua negative), Natural Killer cells (CD3⁻CD56⁺), subset gating as depicted in A. At least three independent experiments were performed and data were summarized from more than 10 healthy donors. Statistical significance was calculated with the two-tailed nonparametric Mann Whitney test. All error bars represent the SEM. Asterisks indicate statistical significance (*p ≤0.05, **p ≤0.005, ***p ≤0.0005).

Figure 2. The NK cell IR subset is present in tonsils ex vivo and accumulates upon stimulation with IL-12

A: % of subset of total NK cells in the tonsil and of CD56^bright cells. Subset accumulates after 18 h stimulation with IL-12 (10 ng/ml). B: Significantly less tonsilar IR cells express c-kit (CD117), whereas C: significantly more blood IR cells express this molecule, which is expressed by more immature NK cells. D: More of the tonsilar IR subset expresses KIR molecules compared to total or CD56^bright NK cells. E: IR cells expressed similar levels of CXCR3 and F: The NKG2A MFI (subset versus CD56^bright total) was not different between IR cells and the rest of CD56^bright cells, pregated on NKG2A positive cells. G: There was a higher frequency of CD16⁺ IR cells than CD16 expression in the rest of CD56^bright cells in TMCs and also a higher level of expression (as measured by MFI ratio (subset/NK)). H: On the contrary in PBMCs a higher frequency of the rest of CD56^bright cells expressed CD16 than IR cells and to a higher level as measured by MFI ratio (subset/NK). Gating strategy: Lymphocyte population in FSC/SSC, Live cells (Live/Dead cell stain Aqua negative), Natural Killer cells (CD3⁻CD56⁺), subset gating: CD56^brightNKG2A⁺CD94⁺CD54⁺CD62L⁻ as depicted in Supplemental Figure 3 A. Data represent at least three independent experiments and at least three donors. Statistical significance was calculated with the two-tailed nonparametric Mann Whitney test. All error bars represent the SEM. Asterisks indicate statistical significance (*p ≤0.05, **p ≤0.005, ***p ≤0.0005).

Figure 3. The NK cell IR subset shows low plasticity even after prolonged cytokine stimulation

NK cells were sorted into IR cells, the rest of CD56^bright and CD56^dim NK cells, stimulated with IL-15 100pg/ml, IL-12 10ng/ml or IL-2 100U/ml for 6 days and NKG2A positivity and CD62L negativity was analysed. CD56 expression changed after prolonged stimulation as shown in Supplemental Figure 4 and discussed. A: Sorting strategy, circled subsets are compared. D, E: IR subset regardless of CD56 expression shows very low plasticity in expression of NKG2A positivity (B) and the CD62L negativity (C) even after prolonged stimulation with cytokines; D: IR cells retain their superior ability for IFN-γ production compared to the rest of CD56^bright cells as shown by the MFI ratio (subset/NK) after prolonged stimulation with IL-12 or IL-2. General Gating strategy: Lymphocyte population in FSC/SSC, live cells (Live/Dead cell stain Aqua negative), Natural Killer cells (CD3⁻CD56⁺), IR: CD56^brightNKG2A⁺CD94⁺CD54⁺CD62L⁻ as depicted in Supplemental Figure 3 A, further subset gating for sorting as in 1A, for analysis after cytokine stimulation NKG2A and CD62L were assessed. Data represents at least three independent experiments with a total of 3–6 donors. All error bars represent the SEM. Asterisks indicate statistical significance (*p ≤0.05, **p ≤0.005, ***p ≤0.0005).

Figure 4. IR NK cells accumulate in EBV-infected tonsils

A: Tonsils of children carrying the Epstein-Barr Virus (EBV), a chronic and cancerogenic virus infection entering the human body via the tonsils, harbor significantly more of this subset of NK cells than those of non-infected children; B: IR subset exists in peripheral blood ex vivo even though this subset is enriched in tonsils and there is significantly less of this subset in the blood of EBV carrying children than in non-infected, adults have less of the IR subset than children; C: Increased numbers of IR in EBV-positive tonsils after IL-12 stimulation, i.e. mean frequency 1.6 fold in EBV negative donors and 1.25 fold in EBV positive donors. Gating strategy: Lymphocyte population in FSC/SSC, Live cells (Live/Dead cell stain Aqua negative), Natural Killer cells (CD3⁻CD56⁺), subset gating CD56^brightNKG2A⁺CD94⁺CD54⁺CD62L⁻ as depicted in Supplemental Figure 3 A. Data represent at least three independent experiments with a total of 4–12 donors. Statistical significance was calculated with the two-tailed nonparametric Mann Whitney test. All error bars represent the SEM. Asterisks indicate statistical significance (*p ≤0.05, **p ≤0.005, ***p ≤0.0005).

Figure 5. Immunoresponsive tonsilar NK cells (IR) potently restrict EBV-induced B cell transformation

Gating strategy for B cell analysis as shown in A. 100’000 autologous B cells were infected with EBV and cocultured with the indicated numbers of sorted NK cells and IL-12 (10 ng/ml) for 10 days. The distinct IR subset has the ability to significantly restrict EBV-induced B cell transformation in physiological numbers (500 NK versus 100’000 B cells, i.e. 0.5% or a five fold increase 2’500 NK versus 100’000 B cells i.e. 2.5%) (B). Gating strategy for sorted IR subset: Lymphocyte population in FSC/SSC, live cells (Live/Dead cell stain Aqua negative), Natural Killer cells (CD3⁻CD56⁺), subset gating: CD56^brightNKG2A⁺CD94⁺CD54⁺CD62L⁻ as exemplary depicted in Supplemental Figure 3 A. Data from three independent experiments with a total of 4–5 donors were summarized. Statistical significance was calculated with the two-tailed nonparametric Mann Whitney test. All error bars represent the SEM. Asterisks indicate statistical significance (*p ≤0.05, **p ≤0.005, ***p ≤0.0005).