TNF induces the production of type 2 cytokines in human group 2 innate lymphoid cells

Capsule summary

TNF receptor II is expressed on ILC2s and TNF is able to induce production of type 2 cytokines in human ILC2s. TNF may play a role in causing or amplifying type 2 immunity contributing to health and disease.

To the Editor:

It is well recognized that group 2 innate lymphoid cells (ILC2s) play a key role in both innate immunity to protect from parasitic pathogens and pathological processes for many type 2 inflammatory diseases including asthma and chronic rhinosinusitis with nasal polyps (CRSwNP).^{1, 2} ILC2-mediated inflammation is mainly triggered via the production of type 2 cytokines including IL-5 and IL-13. Although the epithelial derived cytokines IL-25, IL-33 and thymic stromal lymphopoietin (TSLP) are well appreciated to contribute to ILC2-mediated production of type 2 cytokines, recent studies found that members of the TNF superfamily (TNFSF), including TNFSF11 (receptor activator of NF-κB ligand [RANK-L]), TNFSF15 (TNF-like cytokine 1A [TL1A] and TNFSF18 (glucocorticoid-induced TNFR-related protein ligand [GITR-L]) are also able to induce IL-5 and IL-13 in ILC2s by activation of their respective receptors: TNF receptor superfamily 11A (TNFRSF11A [RANK]), TNFRSF25 (death receptor 3 [DR3]) and TNFRSF18 (GITR).^1–5 Nonetheless, whether TNF itself (also known as TNFα) is able to activate ILC2s and whether TNF receptors, TNFRSF1A (TNF receptor I [TNFRI]) and TNFRSF1B (TNFRII), are expressed on ILC2s has not been investigated. Many TNFRSFs, including RANK, GITR and DR3, share the NF-κB signaling pathway via the activation of TNF receptor associated factor family members, and TNFRII has similar signaling pathways to them. The aim of this study was to examine whether TNFRII is expressed on ILC2s and whether TNF induces production of type 2 cytokines in ILC2s.

We first measured mRNA for TNFRII in human ILC2s. We sorted human ILC2s from peripheral blood from healthy subjects and from nasal polyps (NPs - a type 2 inflammatory manifestation) collected from CRSwNP patients during sinus surgery. We also used human peripheral blood monocytes and the Jurkat cell line as positive and negative controls of TNFRII, respectively. Real-time RT-qPCR was performed using a TaqMan method and detailed methods are given in this article’s Online Repository. We found that TNFRII mRNA was similarly expressed on both blood and NP ILC2s at about 14% of the level on monocytes and was significantly higher than the Jurkat cell line (Fig 1A). We then determined cell surface expression of TNFRII on ILC2s by flow cytometry. Gating strategies identifying human ILC2s are shown in Supplemental Figure E1. We found that TNFRII was expressed on ILC2s from both blood and NPs and levels were significantly higher than Jurkat cells (Fig 1B and C).

Figure 1. TNFRII is expressed on human ILC2s.

Total RNA was extracted from blood monocytes (n=4), blood ILC2s (n=5), NP ILC2s (n=5) and Jurkat cells (n=4). Expression of TNFRII mRNA was analyzed by real-time RT-qPCR using a TaqMan RNA-to-C_T 1-Step Kit (A). Gene expression was normalized to a housekeeping gene, β-actin, and expression levels were shown as ΔCt. Representative histograms of flow cytometric plots for cell surface expression of TNFRII on ILC2s are shown. Levels of cell surface expression of TNFRII on ILC2s from blood from healthy subjects (n=8) and NPs (n=6), Jurkat cells (n=5) and blood monocytes (n=9) are shown by geometric mean fluorescence intensity (gMFI) (B). Comparisons of gMFI ratio of TNFRII to isotype IgG2a are shown in C. Results are shown as mean with SEM. *p< 0.05 and **p< 0.01 by the Mann-Whitney test (A and C) and the Ratio paired t test (B).

We next cultured human ILC2s sorted from blood and NP tissue in RPMI-1640 medium supplemented with 25 IU/mL IL-2 and 10% FBS and investigated the effect of TNF. We found that TNF time- and dose-dependently induced production of IL-5 and IL-13, though IL-5 protein was only detected at 96 hour after stimulation in human blood ILC2s (Fig 2A and E2A). We previously found that ILC2s are activated in NP tissue in vivo and therefore ILC2s sorted from NP tissue release IL-5 and IL-13 without additional stimulation.⁶ We found that TNF further enhanced this spontaneous production of IL-5 and IL-13 in NP ILC2s (Fig 2B). Since TSLP acts in synergy with common ILC2 activators including IL-33, we examined whether TSLP synergistically acts with TNF on blood ILC2s. We found that TSLP significantly and synergistically enhanced TNF-mediated production of type 2 cytokines in blood ILC2s (Fig 2C and E2). To further examine the role of NF-κB in the TNF-mediated production of type 2 cytokines, we stimulated blood ILC2s with TNF and TSLP, and examined the effect of glucocorticoid (dexamethasone, an inhibitor of the NF-κB pathway) and IMD-0354 (an inhibitor of IκB kinase-β that blocks NF-κB nuclear translocation) on IL-5 and IL-13 production. We found that dexamethasone and 10 μM IMD-0354, but not dimethyl sulfoxide (DMSO, vehicle control) completely suppressed TNF-mediated production of IL-5 and IL-13 and partially induced cell death in blood ILC2s (Fig 2D and E3). We also found that 1 μM IMD-0354 partially suppressed cytokine production without enhancement of cell death in ILC2s (Fig E3B and not shown). These results suggest that NF-κB is one of the key transcription factors controlling TNF-mediated induction of type 2 cytokines in human ILC2s. We further found that a STAT6 inhibitor significantly suppressed TNF-mediated production of IL-5 and IL-13 (Fig E4). Since TNF does not directly activate STAT6, the autocrine/paracrine loop of IL-13 may influence TNF-mediated production of type 2 cytokines in ILC2s.

Figure 2. TNF induces production of IL-5 and IL-13 in human ILC2s.

Sorted ILC2s from blood of healthy subjects (A, C-F) and NPs (B) were suspended in RPMI-1640 medium supplemented with 25 IU/ml IL-2, 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin and were stimulated with 5–100 ng/ml TNF (A. n=9) or 20 ng/ml TNF (B. n=7) for 4 days. Blood ILC2s were stimulated with 20 ng/ml TNF, 10 ng/ml TSLP, 10 ng/ml IL-33 or their combination for 4 days (C. n=14, E. n=5). Blood ILC2s were cultured with 20 ng/ml TNF and 10 ng/ml TSLP in the presence or absence of 0.01% DMSO (vehicle control), 100 nM dexamethasone (Dex), 10 μM IMD-0354 (an inhibitor of IKKβ that blocks NF-κB nuclear translocation), 10 ng/ml IL-10, 20 ng/ml TGF-β1, 1000 IU/ml IFN-β, 10 ng/ml IFN-γ or 50 ng/ml IL-27 for 4 days (D. n=6, F. n=8). Each experiment was performed in duplicate with independent donors. The concentrations of IL-5 and IL-13 were measured by Luminex assay. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001 were calculated by the Ratio paired t test (B) and the one-way ANOVA Friedman test (A, C-F).

While TNF induced production of both IL-5 and IL-13 in human blood ILC2s, it induced 15-fold greater quantities of IL-13 than IL-5 (Fig 2C). In contrast, IL-33 induced IL-5 and IL-13 equally in ILC2s (Fig 2E). We also found that the combination of TNF and IL-33 showed additive and/or synergistic effects on the production of type 2 cytokines in human blood ILC2s (Fig 2E). These results suggest that there are different signaling pathways mediating TNF- and IL-33-induced activation of ILC2s. To test for differences between TNF- and IL-33-mediated signaling, we investigated the effects of known inhibitors for IL-33-mediated ILC2 activation (Fig E5)² and found that dexamethasone, TGF-β, IFN-β, IFN-γ and IL-27 but not IL-10 significantly suppressed TNF plus TSLP-mediated production of IL-13 in blood ILC2s (Fig 2D and 2F). These results suggest that signaling pathways that can be suppressed by IL-10 (potentially via STAT3) are present in IL-33-mediated but not in TNF-mediated pathways and that this IL-33-specific pathway may be key to signaling strong induction of IL-5 in ILC2s. Future studies are required to characterize TNF and IL-33 specific signaling pathways in ILC2s.

During helminth infection, epithelial cells release IL-25 and IL-33 and then induce ILC2-mediated production of IL-13 to clear invading worms.¹ Helminth infection also induces production of pro-inflammatory cytokines, including TNF, in immune cells such as macrophages and structural cells such as epithelial cells, via innate pattern recognition receptors such as toll-like receptors.⁷ These findings suggest that TNF, together with other innate cytokines including IL-33, may play a role in the protection from helminth infection by the activation of ILC2s. In the case of human diseases, TNF is known to be elevated in type 2 inflammatory diseases including asthma, however, treatment with TNF blockers did not show clear efficacy in asthmatics.⁸ Asthma is a heterogeneous disease that is characterized by numerous inflammatory endotypes (e.g. eosinophilic or neutrophilic forms).⁹ It is possible that distinct ILC2 activators drive disease in patients with eosinophilic asthma, and further clinical study using endotype analysis based on type 2 markers and ILC2 activators will be required to examine the role of the TNF-ILC2 cascade in asthma. In addition, elevation of TSLP in type 2 inflammatory diseases is expected to produce a synergistic response via TNF and IL-1 family molecules, including TNF and IL-33. Inhibition of TSLP may have greater benefit in type 2 inflammatory diseases than targeting individual ILC2 activators including TNF and IL-33. In conclusion, TNF is able to induce and greatly enhance production of type 2 cytokines in ILC2s and it may play a role in causing or amplifying type 2 immunity contributing to health and disease.