Interferon-λ and interleukin 22 act synergistically for the induction of interferon-stimulated genes and control of rotavirus infection

Abstract

The epithelium is the main entry point for many viruses, but the processes that protect barrier surfaces against viral infections are incompletely understood. Here we identified interleukin 22 (IL-22) produced by innate lymphoid cell group 3 (ILC3) as an amplifier of signaling via interferon-λ (IFN-λ), a synergism needed to curtail the replication of rotavirus, the leading cause of childhood gastroenteritis. Cooperation between the receptor for IL-22 and the receptor for IFN-λ, both of which were 'preferentially' expressed by intestinal epithelial cells (IECs), was required for optimal activation of the transcription factor STAT1 and expression of interferon-stimulated genes (ISGs). These data suggested that epithelial cells are protected against viral replication by co-option of two evolutionarily related cytokine networks. These data may inform the design of novel immunotherapy for viral infections that are sensitive to interferons.

Fig. 1. Control of rotavirus infection requires IL-22

Seven day old mice were orally infected with 7×10² ID₅₀ rotavirus and samples were collected at the indicated times following infection. Newborn mice were fostered by wildtype (WT) mothers (a, b) or Ifnlr1^−/− mothers (c, d). Virus replication was analyzed by (a, c) RT-qPCR using RNA obtained from small intestinal tissue, (b, d) ELISA of colon homogenates (see Methods), (e) immunofluorescence staining and (f) by determining the distribution and mean number (± SEM) of infected cells per visual field across the small intestine. The size of the pies represents the mean number of infected cells per visual field and their distribution across the entire small intestine: WT (239 ± 18), Il22^−/− (958.4 ± 10), Ifnlr1^−/− (1187.5 ± 108). (g) H&E staining of small intestinal tissue sections. (h) Clinical score. (i) Body weight at day 19 after infection as the percentage of body weight at day 0. All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a-k) Mean ± SEM. (a, b) n ≥ 6. (c-g) n ≥ 4. (i) n ≥ 3. Data are representative of four (a-b), two (c-d) and two (e-g) individual experiments. i, Data combined from two individual experiments. (a-i) One-way ANOVA, *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant. Scale bars, 50 μm.

Fig. 2. Rotavirus infection induces production of IFN-λ by intestinal epithelial cells and of IL-22 by ILC3s

Groups of suckling wildtype mice were infected with rotavirus. Control groups received PBS. (a) IFN-λ production in whole small intestinal tissue of infected suckling mice at day 1 following infection was determined by ELISA. (b, c) Expression of Ifnl (primer detects both Ifnl2 and Ifnl3) was determined by RT-qPCR using a Taqman assay detecting transcripts of both genes. Small intestinal epithelial cells (IEC) and lamina propria leukocytes (LPL) were isolated from suckling mice at day 1 following infection using the EDTA dissociation method (see Methods) (b) or highly purified by sorting epithelial (EpCAM⁺CD45^–) or hematopoietic cells (EpCAM^–CD45⁺) (c). (d) Il22 expression was determined by RT-qPCR using RNA from IEC and LPL isolated by the EDTA dissociation method from suckling mice at day 1 following infection. (e) Flow cytometry analysis of IL-22 and CD3 expression by gated RORγt⁺ LPL from suckling mice at day 0 (UI, uninfected) and at day 1 following infection. (f) Proportions and (g) absolute numbers of RORγt⁺IL-22⁺CD3⁺ and IL-22⁺CD3^– LPL in uninfected suckling mice at day 1 after infection. (h) Virus load (RT-qPCR of whole small intestinal tissue) from suckling mice of the indicated strains at day 4 after infection. All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a-d, f-h) Mean ± SEM. (a) n ≥ 6. (b, d) n ≥ 5. c, n ≥ 3. (f, g) n = 4. (h) n ≥ 5. Data are representative of two (a, c, h) and three (b, d-g) independent experiments. (a, c) Student’s t test. (b, d, f, g) Two-way ANOVA. (h) One-way ANOVA. *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant. ND, not detectable (or not displayable on the scale indicated).

Fig. 3. IL-1α produced by intestinal epithelial cells controls rotavirus-induced IL-22 production by ILC3s

Suckling mice from the indicated mouse strains were infected with rotavirus. (a-c) Cytokine production in small intestinal tissue of uninfected (UI) and infected (day 1) suckling wildtype mice was determined by ELISA. (d, e) Il1a (d) or Il22 (e) expression by small intestinal LPL and IECs (d) or small intestinal tissue (e) of uninfected and infected (day 1) suckling wildtype mice was determined by RT-qPCR. (f) Virus load in the small intestine of suckling mice was determined by RT-qPCR at day 4 after infection. (g-j) Mice were treated with the indicated neutralizing antibodies at day −1 (g-j) and day 1 (h-j) of infection. (g) IL-22 production by RORγt⁺ LPL was determined by flow cytometry at day 1 following infection. (h) Virus load in colonic tissue at day 4 after infection was determined by ELISA. (i) H&E staining of small intestinal tissue sections. (j) Clinical score. (k, l) Small intestinal lamina propria CD3^–, CD19^–, CD11b^–, KLRG1^–, NK1.1^–, IL-7R⁺, Kit⁺ cells from wildtype animals were highly purified by sorting and cultured in the presence of IL-7, SCF and IL-1α or IL-1β for 3 days. (k) Il22 transcripts in cultured ILC3s were measured by RT-qPCR. (l) IL-22 production was measured by ELISA in culture supernatant. All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a, c) Student’s t test. Data are representative of two (a-c, h-l) or three (d-e, g) independent experiments. (d, e) Two-way ANOVA. (f, j-l) One-way ANOVA. *P < 0.05; **P < 0.01; ***P < 0.001. NS, not significant.

Fig. 4. Control of rotavirus replication by IL-22 depends on IFN-λ receptor signaling

Mice of the indicated strains were infected with rotavirus and samples were collected at the indicated time points after infection. (a, b) Ifnl (a) and Ifnlr1 (b) expression in the small intestine of mock and rotavirus infected suckling mice. (c, d) Virus load was determined by RT-qPCR (c) or ELISA (d) at day 4 after infection of suckling mice. Pooled ELISA data from two independent experiments were normalised to fold ID₅₀ of wildtype mice (d). (e-h) Two week old suckling mice of the indicated strains were subcutaneously injected with PBS or the indicated cytokines 8 hours before infection with rotavirus. Injections were repeated on days 1 and 2 after infection. Virus load was determined at day 3 of rotavirus infection by RT-qPCR of RNA obtained from small intestinal tissue (e, g, h) or by ELISA from colon homogenates (f). All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a-h) Mean ± SEM. (a-c) n ≥ 5. (d) n ≥ 12. (e, f) n ≥ 6. (g, h) n ≥ 4. (a-f) One-way ANOVA. (g, h) Student’s t test. Data are representative of (a, b) three and (c, e-h) two independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant. ND, not detectable (or not displayable on the scale indicated).

Fig. 5. Cooperation between IL-22 and IFN-λ is required for inducing an efficient antiviral state in IECs

(a) Gene expression analysis of IECs isolated from suckling animals sacrificed at day 4 after infection. Heat map visualizing relative expression of selected genes by comparing infected wildtype (WT Inf) with uninfected wildtype (UI) mice and infected Il22^−/− with infected wildtype mice as well as infected Ifnlr1^−/− with infected wildtype mice. (b-e) The intestinal epithelial cell lines IEC6 (b, c) and Caco2 (d, e) were stimulated for 16 hours with the indicated amounts of mouse (IEC6) or human (Caco2) the indicated doses of IFN-λ2 alone or in combination with 100 ng/ml of mouse or human IL-22. The expression of the indicated ISGs was evaluated by RT-qPCR. (f, g) Suckling wildtype mice were injected subcutaneously with indicated amounts of IFN-λ2 and/or 1 μg of IL-22. Three hours after injection, Isg15 (f) and Oasl2 (g) expression in IECs was analyzed by RT-qPCR. (h, i) Adult mice were injected subcutaneously with 1 μg of IL-22 and expression of ISGs was analyzed by RT-qPCR using RNA obtained from IECs isolated from the small intestines 3 hours after cytokine injection. (j, k) Caco2 (j) cells or IEC6 cells (k) were treated with 20 ng/ml of IFN-λ2 and/or 100 ng/ml of IL-22 before infection with poliovirus (j) or VSV (k). All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a-k) Mean ± SEM. (a) n ≥ 5. (b-e) n = 4. (f, g) n ≥ 3. (h, i) n = 3. (j, k) n ≥ 4. (b-k) Student’s t test. Data are representative of (a-g, j, k) three and (h, i) two independent experiments. *P < 0.05; **P < 0.01. NS, not significant.

Fig. 6. IL-22 and IFN-λ synergize for phosphorylation of STAT1

(a-h) Wildtype and Stat3^ΔIEC mice were infected with rotavirus and samples collected at day 4 after infection. Virus load (a) and the expression of STAT1-dependent (b) or STAT3-dependent genes (c-h) was determined by RT-qPCR of RNA from IECs. (i) IEC6 cells were stimulated with the indicated cytokines for 30 min. Phosphorylation of the indicated STAT proteins was determined by flow cytometry (open histograms). Unstimulated IEC6 cells are shown as a control (grey histograms). (j) Wildtype mice were subcutaneously injected with 1 μg of IFN-λ2 and/or 1 μg of IL-22. One hour later, STAT phosphorylation in small intestinal tissues was assessed by immunoblotting. (k) Phospho-STAT (pSTAT) analysis in intestinal epithelial cell lines IEC6 and Caco2 stimulated with 20 ng/ml IFN-λ2 and/or 100 ng/ml IL-22 for 30 min. All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a-h) Mean ± SEM; n ≥ 4. Data are representative of two independent experiments. (a) Student’s t test. (b-h) Two-way ANOVA. *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant.

Fig. 7. Enhancement of IFN-λ-dependent ISG expression by IL-22 requires STAT1 signaling

(a, b) Groups of adult mice of the indicated strains were treated with 1 μg of IL-22. Three hours later, gene expression of the indicated genes was determined by RT-qPCR using RNA isolated from small intestinal tissue. (c) The indicated STAT proteins were knocked down in IEC6 cells using siRNAs. Cells were stimulated with 100 ng/ml IL-22, 100 ng/ml IFN-λ2 or both for 16 hours. Isg15 expression was evaluated by RT-qPCR. (d-f) Suckling mice of the indicated strains were infected with rotavirus. (d) Rotavirus replication in the small intestine was analyzed by RT-qPCR. (e) Analysis of Isg15 expression by small intestinal IECs by RT-qPCR. (f) Virus shedding into colon tissue was analyzed by ELISA. All RT-qPCR data are shown as fold (2^−ΔCt) relative to Hprt. (a-f) Mean ± SEM. (a-c) n ≥ 3. (d-f) n ≥ 4. Data are representative of two (a, b, d-f) and four (c) independent experiments. (a-f) Two-way ANOVA. *P < 0.05; **P < 0.01; ***P < 0.001; NS, not significant.