Interleukin-1beta selectively expands and sustains interleukin-22+ immature human natural killer cells in secondary lymphoid tissue

Abstract

Among human natural killer (NK) cell intermediates in secondary lymphoid tissue (SLT), stage 3 CD34(-)CD117(+)CD161(+)CD94(-) immature NK (iNK) cells uniquely express aryl hydrocarbon receptor (AHR) and interleukin-22 (IL-22), supporting a role in mucosal immunity. The mechanisms controlling proliferation and differentiation of these cells are unknown. Here we demonstrate that the IL-1 receptor IL-1R1 was selectively expressed by a subpopulation of iNK cells that localized proximal to IL-1beta-producing conventional dendritic cells (cDCs) within SLT. IL-1R1(hi) iNK cells required continuous exposure to IL-1beta to retain AHR and IL-22 expression, and they proliferate in direct response to cDC-derived IL-15 and IL-1beta. In the absence of IL-1beta, a substantially greater fraction of IL-1R1(hi) iNK cells differentiated to stage 4 NK cells and acquired the ability to kill and secrete IFN-gamma. Thus, cDC-derived IL-1beta preserves and expands IL-1R1(hi)IL-22(+)AHR(+) iNK cells, potentially influencing human mucosal innate immunity during infection.

Figure 1. IL-1R1 expression during NK cell development

(A) IL-1R1 mRNA expression ex vivo in flow cytometry purified human stage 1-4 SLT NK developmental intermediates assessed via Real-Time PCR. Fold difference in IL-1R1 mRNA is relative to that in stage 1, arbitrarily normalized to 1 (n = 4). (B-D) Total CD3⁻CD19⁻ tonsillar mononuclear cells were stained for surface expression of CD34, CD117, CD94, and IL-1R1, and gated for each stage as described (Freud et al., 2006). (B) Histograms show IL-1R1 (filled) and isotype (empty) surface staining in stage 2 (top) or stage 3 (bottom) cells from a representative donor (n = 4). (C,D) Shown within each gated SLT NK developmental intermediate populations are average: (C) % of cells with IL-1R1 surface expression (n = 4), and (D) density of IL-1R1 surface expression, as represented by the ratio of IL-1R1 mean fluorescence intensity (MFI) to isotype MFI (n = 6). Data in A, C, D presented as mean ± SEM (*, P ≤ 0.05; **, P ≤ 0.005; ***, P ≤ 0.0005).

Figure 2. IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells ex vivo

(A-C) Tonsillar mononuclear cells were stained ex vivo for surface expression of CD117, IL-1R1, and (B,C) intracellular IL-22. (A) Gated on Lin⁻ (CD3⁻CD14⁻CD16⁻CD19⁻CD34⁻CD94⁻) lymphocytes, dot plot indicates gates for IL-1R1^hi (upper) or IL-1R1^lo (lower) subpopulations of CD117⁺ stage 3 iNK cells as sorted in a representative donor (n = 18). (B,C) Gating as indicated in (A), histograms depict staining with isotype (empty) or α-IL-22 (filled) antibody in a representative donor (n = 5). (D) IL-22 and AHR mRNA measured via Real-Time PCR ex vivo in flow cytometry sorted IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells. “Fold difference” was quantified relative to expression in the IL-1R1^lo subpopulation, arbitrarily normalized to 1 (P ≤ 0.05). Data in D presented as mean ± SEM (n = 6). (E) Using gating as indicated in (A), histograms depict a representative donor stained ex vivo with isotype (empty) or antibody specific for the indicated antigen (filled), within total stage 3 iNK cells or IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells (n ≥ 4). The IL-1R1⁻CD117⁻ and IL-1R1⁺CD117⁻ subsets noted in (A) were both non-reactive for intracellular IL-22 protein staining.

Figure 3. Treatment with IL-15 and IL-1β promotes the selective expansion of stage 3 iNK cells

(A,B) “Fold change” was calculated and averaged from 11 donors as: absolute number of cells enumerated (by Trypan blue exclusion) after 14 d of culture in IL-15 and IL-1β divided by the absolute number of cells enumerated after 14 d of culture in IL-15 alone. For example, starting with 5,000 stage 3 iNK cells/well, four representative donors cultured in IL-15 and IL-1β expanded to 90,000, 285,000, 162,000, and 40,000, while the same four donors’ stage 3 iNK cells cultured in IL-15 alone expanded to 20,000, 55,0000, 45,000 and 10,000 cells, respectively. Results revealed a selective growth advantage among stage 3 iNK cells in response to culture with IL-15 and IL-1β (**, P < 0.005; n = 11), (B) which was completely abrogated in the presence of IL-1RA (P < 0.05; n = 4). (C,D) Proliferation assessed via EdU incorporation in (C) total stage 3 iNK cells (P < 0.05; n = 8) or (D) IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells (*, P = 0.01; n = 3) after culture for 14 d in the presence of IL-15 or IL-15 and IL-1β. Data in A-D depicted as mean ± SEM.

Figure 4. IL-1β sustains IL-22 and AHR expression in IL-1R1^hi stage 3 iNK cells

(A-D) IL-22 and AHR transcript expression following 14 d in the presence of IL-15 or IL-15 and IL-1β, shown in (A,B) total (“bulk”) stage 3 iNK cells, or (C,D) the IL-1R1^hi subpopulation of stage 3 iNK cells. Results depicted as the mean ± SEM, and P ≤ 0.05 in each (n ≥ 4). (E) Dot plots depict surface IL-1R1 and intracellular staining with IL-22 or isotype antibody in IL-1R1^hi stage 3 iNK cells from a representative donor cultured for 14 d in the presence of IL-15 or IL-15 and IL-1β (n = 6). All IL-22⁺ cells co-expressed IL-1R1. (F) Histograms from a representative donor (n = 4) depict staining with IL-22 (filled) or isotype (empty) antibody in the IL-1R1^hi subpopulation of stage 3 iNK cells d 0, and within IL-1R1⁺ cells remaining after culture for 14 d with IL-15 and IL-1β (G) IL-1R1 MFI assessed via flow cytometry after a 6 h culture in the presence of IL-15 or media alone (n = 3). (H) Representative staining for surface CD94 and intracellular IL-22 protein at d 14 in cells cloned from either a single IL-1R1^lo stage 3 iNK cell or an IL-1R1^hi stage 3 iNK cell. Isotype staining on far right.

Figure 5. IL-1β inhibits differentiation of IL-1R1^hi stage 3 iNK cells

Surface expression of (A,D,E) CD94 and (B,C) CD56 assessed after 14 d of in vitro culture of sorted (A,B) total, or (C,D,E) IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells with IL-15 or IL-15 and IL-1β. (F,G) CD94 and CD56 were also assessed after bulk stage 3 iNK cells were cultured for 14 d in the presence of IL-15, IL-15 and IL-1β, or IL-15 with IL-1β and IL-1RA. (H) Histograms depict staining with isotype (empty) or antibody specific for the indicated antigen (filled) after culture of IL-1R1^hi stage 3 iNK cells for 14 d in the presence of IL-15 or IL-15 and IL-1β. Arrows indicate high density expression for IL-1R1 and IL-22 (also see Figure 4F). Data in A, B, E, F, and G presented as the mean ± SEM (*, P ≤ 0.05; n ≥ 4). Histograms in C, D, and H depict staining in a representative donor (n ≥ 4).

Figure 6. IL-1β impedes the IL-1R1^hi subpopulation of stage 3 iNK cells from acquiring IFN-γ production and degranulation

(A,C) Total or (B,D) IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells sorted from SLT were cultured for 14 d in vitro with IL-15 or IL-15 and IL-1β. Cells were replated in equal numbers and either: (A,B) stimulated for 12 h with IL-15 + IL-12 + IL-18 to assess IFN-γ secretion via (A) ELISA (P < 0.005; n = 7) or (B) intracellular flow (n = 8); or (C,D) incubated overnight with K562 targets (4:1 E/T ratio) and assessed for surface expression of the degranulation marker CD107a by flow cytometry (n = 3; for C, P = 0.14). Data in B depicts IFN-γ expression after gating on the population which had acquired CD94 surface expression. Data presented as mean ± SEM (*, P ≤ 0.01).

Figure 7. Stage 3 iNK cells reside in the vicinity of IL-1β⁺ CD11c^hi cDCs within SLT

(A,B) Intracellular staining of unstimulated SLT mononuclear cells from a representative donor (n = 3) with isotype (empty) or α-IL-1β (filled) antibody shown after gating on (A) CD45⁺CD3⁺ T cells and CD45⁺CD14⁺ monocytes and macrophages, or (B) CD123^hiCD11c^lo pDCs and CD123^loCD11c^hi cDCs subsets. (C) Immunohistochemical staining depicted in serial sections of human tonsil from a representative donor (n = 4). Yellow areas indicate co-expression, as shown for CD11c^hi and IL-1β (top). CD117⁺ cells did not co-express CD11c, but were located in proximity of CD11c^hi cDCs (bottom). Magnification, 400x. Bar, 30 μM. (D) EdU incorporation in sorted IL-1R1^hi and IL-1R1^lo subpopulations of stage 3 iNK cells after 6 h with α-MEM medium alone, or in the presence of autologous BDCA-1⁺ cDCs (4:1 iNK:cDC ratio) with either isotype or α-IL-15 and α-IL-1β blocking antibodies. Depicted is % of CD117⁺ lymphocytes which were also EdU⁺ (*, P = 0.05; n = 3). Data in D presented as mean ± SEM.