Influenza-Activated ILC1s Contribute to Antiviral Immunity Partially Influenced by Differential GITR Expression

Abstract

Innate lymphoid cells (ILCs) represent diversified subsets of effector cells as well as immune regulators of mucosal immunity and are classified into group 1 ILCs, group 2 ILCs, and group 3 ILCs. Group 1 ILCs encompass natural killer (NK) cells and non-NK ILCs (ILC1s) and mediate their functionality via the rapid production of IFN-γ and TNF-α. The current knowledge of ILC1s mainly associates them to inflammatory processes. Much less is known about their regulation during infection and their capacity to interact with cells of the adaptive immune system. The present study dissected the role of ILC1s during early influenza A virus infection, thereby revealing their impact on the antiviral response. Exploiting in vitro and in vivo H1N1 infection systems, a cross-talk of ILC1s with cells of the innate and the adaptive immunity was demonstrated, which contributes to anti-influenza immunity. A novel association of ILC1 functionality and the expression of the glucocorticoid-induced TNFR-related protein (GITR) was observed, which hints toward a so far undescribed role of GITR in regulating ILC1 responsiveness. Overexpression of GITR inhibits IFN-γ production by ILC1s, whereas partial reduction of GITR expression can reverse this effect, thereby regulating ILC1 functionality. These new insights into ILC1 biology define potential intervention targets to modulate the functional properties of ILC1s, thus contributing toward the development of new immune interventions against influenza.

Keywords: cross-talk; glucocorticoid-induced TNFR-related protein; influenza; innate lymphoid cell 1; regulation.

Figure 1

H1N1 infection leads to enhanced ILC1 activation and functionality. Wild-type mice and RAG2^−/−γc^−/− mice were infected i.n. with 2 × 10³ ffu of the H1N1 PR8 strain. (A) Scatter plots represent viral loads of wild-type mice (left panel) and RAG2^−/−γc^−/− mice adoptively transferred with 2 × 10⁵ in vitro-generated ILC1s/animal (right panel; n = 4–8) as determined by foci assay and depicted as ffu per lung at the indicated time points. Lymphocytes derived ex vivo from lungs of infected wild-type mice were incubated at 37°C in medium containing brefeldin and monensin for 3 h prior to flow cytometry staining of markers related to ILC1 activation and functionality. (B) Frequencies and absolute numbers of lung-derived ILC1s. Frequencies and mean fluorescence intensities (MFI) of (C) IFN-γ, (D) TNF-α, (E) TRAIL, and (F) CD49a expressing lung ILC1s. Shown is one out of three independent experiments (n = 4–8). Bars with scatter plots represent MFI and range in frequencies with the horizontal line drawn at the mean. Asterisks denote significant values as calculated by nonparametric Mann–Whitney’s test (viral loads) or One-way ANOVA (frequency and MFI) as compared to uninfected samples; ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05.

Figure 2

GITR expression defines ILC1 functionality. Wild-type mice were infected i.n. with 2 × 10³ ffu of the H1N1 PR8 strain. Lymphocytes derived ex vivo from lungs (for ILC1 profile) or dLNs (mediastinal and cervical, for antigen-presenting cell profile) of infected mice were incubated at 37°C in medium containing brefeldin and monensin for 3 h prior to the flow cytometry staining of markers related to ILC1 activation and functionality as well as DC markers, respectively (n = 5–8). (A) Representative histogram of GITR expression by ILC1s, GITR⁺ ILCs expressed as frequency of ILC1s and MFI of GITR-expressing ILC1s. (B) MFI of IFN-γ, TNF-α, TRAIL, and CD49a of GITR⁺ and GITR⁻ ILC1s. (C) MFI of IFN-γ, TNF-α, TRAIL, and CD49a of GITR^lo and GITR^hi ILC1s. (D) Frequencies of CD11c⁺CD103⁺ DCs and their expression densities of GITR-L. Bars with scatter plots represent MFI and range in frequencies with the horizontal line drawn at the mean. MFI and frequency data are from one out of two independent experiments. Asterisks denote significant values as calculated by One-way ANOVA as compared to uninfected samples; ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05.

Figure 3

GITR expression level defines responsiveness of IL-12- and IL-18-induced ILC1 activation. In vitro-generated ILC1s were stimulated with 100 ng/ml of IL-12 ± IL-18 for 48 h and incubated at 37°C in medium containing brefeldin and monensin for 3 h prior to the flow cytometry of markers related to ILC1 activation and functionality. The bars with scatter plots represent the MFI values (n = 3–4). (A) Representative histogram depicts GITR expression on ILC1s at steady state (gray filled line = fmo control, black line = ILC1s) and (B) MFI of GITR expression by ILC1s post-IL-12 ± IL-18 treatment. Changes in expression densities and frequencies of GITR^lo and GITR^hi ILC1s expressing (C) IFN-γ and (D) TNF-α. MFI data are representative from one out of two independent experiments. Asterisks denote significant values as calculated by nonparametric Kruskal–Wallis test (Dunn’s posttest) as compared to untreated samples; ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05.

Figure 4

ILC1s are engaged in cross-talk with DCs during H1N1 infection in vitro. BMDCs generated from wild-type mice using the FLT-3 ligand stimulation were infected with the H1N1 PR8 strain (at a MOI of 1). In vitro-generated ILC1s cultured overnight with infected or uninfected BMDCs at a 1:1 ratio were stained for flow cytometry analysis after 3 h incubation in media with brefeldin and monensin. MFI of ILC1s expressing (A) IFN-γ, (B) TNF-α, (C) GITR, and (D) CD49a upon coculture with H1N1-infected or uninfected BMDCs. (E) Representative histograms for the expression of CD40, CD80, CD86, and MHC cl. II on infected BMDCs. (F) MFI of CD11c⁺ BMDCs expressing CD40, CD80, CD86, and MHC cl. II after coculture with ILC1s. Bars with scatter plots represent the mean ± SEM (n = 3–4) and MFI data are representative from one out of three independent experiments. Asterisks denote significant values as calculated by nonparametric Kruskal–Wallis test (Dunn’s posttest); ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05.

Figure 5

GITR-expression modulates ILC1 functionality upon coculture with H1N1-infected BMDCs. BMDCs generated from wild-type mice using FLT-3 ligand stimulation were infected with the H1N1 PR8 strain (at a MOI of 1). In vitro-generated ILC1s were cocultured overnight with infected or uninfected BMDCs at a 1:1 ratio. The recombinant mouse GITR-Fc chimera protein was applied to the coculture overnight to manipulate GITR expression. Surface and cytokine staining were performed for flow cytometry analysis after 3 h of incubation in media with brefeldin and monensin. (A) GITR expression by ILC1s and BMDCs represented as MFI and representative histogram. (B) MFI of GITR expression by ILC1s with and without GITR-Fc treatment and representative histogram. MFI of (C) IFN-γ, (D) TNF-α, and (E) CD49a expression after GITR-Fc treated ILC1s cocultured with H1N1-infected BMDCs. (F) MFI of CD40, CD80, CD86, and MHC cl. II expression by infected BMDCs post-GITR-Fc treatment. Bars with scatter plots represent the mean ± SEM (n = 3–4) and MFI data are representative from one out of two independent experiments. Asterisks denote significant values as calculated by nonparametric Mann–Whitney’s test; ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05.

Figure 6

GITR–GITR-L interaction results in increased ILC1 functionality in the course of H1N1 infection. Wild-type mice were treated with either DTA-1 GITR agonist/isotype (500 µg/animal) or GITR-Fc fusion chimera protein (6.25 µg/animal) i.p. After 24 h mice were infected i.n. with 2 × 10³ ffu of the H1N1 PR8 strain. (A) Representative histograms and MFI of GITR expression by ILC1s. Lung-derived lymphocytes of infected wild-type mice 3 dpi were incubated at 37°C in medium containing brefeldin and monensin for 3 h prior to the flow cytometry staining with regard to markers related to ILC1 activation and functionality. (B) MFI and frequencies of lung-derived ILC1s expressing IFN-γ, TNF-α, and TRAIL from DTA-1-treated and control groups. (C) MFI and frequencies of lung-derived ILC1s expressing IFN-γ, TNF-α, and TRAIL from GITR-Fc-treated and control groups. Shown is one out of two independent experiments (n = 4–6). Box plots represent MFI and range in frequencies with the horizontal line drawn at the mean. Asterisks denote significant values as calculated by One-way ANOVA; ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05; n.s., not significant.

Figure 7

Cross-talk of ILC1s with BMDCs and CD8 T cells during H1N1 infection. FLT-3 ligand-differentiated BMDCs were infected with the SIINFEKL expressing H1N1 PR8 influenza strain at a MOI of 1. In vitro-generated ILC1s were cultured overnight with infected or uninfected BMDCs and CD8 T cells derived from OT I mice at a 1:1:1 ratio. Surface markers and cytokines were stained for flow cytometry analysis after 3 h of incubation in media with brefeldin and monensin. MFI of (A,B) IFN-γ and (C,D) TNF-α expression by ILC1s depending on the level of GITR expression. (E) Representative two-dimensional FACS plots of IFN-γ vs. TNF-α from GITR^lo and GITR^hi ILC1s in the absence or presence of CD8 T cells. (F) MFI of IFN-γ and TNF-α-expressing CD8 T cells. CD8 T cells were depleted before and 3 days after H1N1 PR8 infection (2 × 10³ ffu/animal) of wild-type mice by i.p. administration of CD8 T cell-depleting antibodies (200 μg/animal). Lung-derived lymphocytes of infected wild-type mice were analyzed by flow cytometry 6 dpi with regard to markers related to ILC1 activation and functionality. (G) Frequencies of IFN-γ⁺, TNF-α⁺, TRAIL⁺, and GITR⁺ lung-derived ILC1s. Bars with scatter plots represent the mean ± SEM and the in vitro (n = 3–4 technical replicates) and in vivo (n = 5–6) data are representative from one out of two independent experiments. Asterisks denote significant values as calculated by nonparametric Kruskal–Wallis test (Dunn’s posttest) (in vitro data) or One-way ANOVA (in vivo data); ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05; n.s., not significant.

Figure 8

CD4 T cells contribute to enhanced ILC1 functionality during H1N1 infection in vitro and in vivo. FLT-3 ligand-differentiated BMDCs were infected with the OVA peptide (aa323–393)-expressing H1N1 PR8 influenza strain at a MOI of 1. In vitro-generated ILC1s were cocultured overnight with infected or uninfected BMDCs and CD4 T cells derived from OT II mice at a 1:1:1 ratio. Staining of surface makers and cytokines for flow cytometry analysis after 3 h of incubation in media with brefeldin and monensin. MFI of (A,B) IFN-γ and (C,D) TNF-α expression by ILC1s depending on the level of GITR expression. (E) Representative two-dimensional FACS plots of IFN-γ vs. TNF-α from GITR^lo and GITR^hi ILC1s in the absence or presence of CD4 T cells. (F) MFI of IFN-γ and TNF-α-expressing CD4 T cells. CD4 T cells were depleted before and 3 days after H1N1 PR8 infection (2 × 10³ ffu/animal) of wild-type mice by i.p. administration of CD4 T cell-depleting antibodies (200 µg/animal). Lung-derived lymphocytes of infected wild-type mice were stained for flow cytometry analysis 6 dpi with regard to markers related to ILC1 activation and functionality. (G) Frequencies of IFN-γ⁺, TNF-α⁺, TRAIL⁺, and GITR⁺ lung-derived ILC1s. Bars with scatter plots represent the mean ± SEM and the in vitro (n = 3–4 technical replicates) and in vivo (n = 5–6) data are representative from one out of two independent experiments. Asterisks denote significant values as calculated by nonparametric Kruskal–Wallis test (Dunn’s posttest) (in vitro data) or One-way ANOVA (in vivo data); ****p ≤ 0.0001; ***p ≤ 0.001; **p ≤ 0.01; *p ≤ 0.05; n.s., not significant.