The alarmin IL33 orchestrates type 2 immune-mediated control of thymus regeneration

Abstract

As the primary site of T-cell development, the thymus dictates immune competency of the host. The rates of thymus function are not constant, and thymus regeneration is essential to restore new T-cell production following tissue damage from environmental factors and therapeutic interventions. Here, we show the alarmin interleukin (IL) 33 is a product of Sca1⁺ thymic mesenchyme both necessary and sufficient for thymus regeneration via a type 2 innate immune network. IL33 stimulates expansion of IL5-producing type 2 innate lymphoid cells (ILC2), which triggers a cellular switch in the intrathymic availability of IL4. This enables eosinophil production of IL4 to re-establish thymic mesenchyme prior to recovery of thymopoiesis-inducing epithelial compartments. Collectively, we identify a positive feedback mechanism of type 2 innate immunity that regulates the recovery of thymus function following tissue injury.

Fig. 1. Selective requirement for the alarmin IL33 in thymus regeneration.

Analysis of thymus regeneration (indicated TEC and thymocyte subsets) at 35 days post-SLI in WT (black circle) and either Il25^−/− (a) or Il33^−/− (b) (white circle) mice. For (a), TEC WT n = 6 and Il25^−/− n = 7 and thymocytes WT and Il25^−/− n = 6 animals over 2 independent experiments. For (b), TEC WT n = 6 and Il25^−/− n = 7 and thymocytes WT and Il25^−/− n = 7 animals over 2 independent experiments. All error bars show the mean ± SEM. P-values were obtained using two-tailed, unpaired Student’s t tests.

Fig. 2. IL33 boosts thymus regeneration.

a Injection regime for the administration of recombinant IL33 (rIL33), 2.5μg/mouse/injection or PBS control. Three injections were given prior to sub-lethal irradiation (SLI) (1 × 425 rad) and one injection post-SLI exposure. Mice were harvested either 7 or 35 days after SLI. b Effects of IL33 administration on TEC recovery at 7 days post-SLI, and TEC and thymocyte recovery at 35 days post-SLI (c). For (b), PBS and rIL33 n = 8 animals obtained over 2 independent experiments. For (c), TEC PBS and rIL33 n = 6 and thymocytes PBS and rIL33 n = 8 animals obtained over 2 independent experiments. All bars show mean ± SEM. P-values were obtained using two-tailed, unpaired Student’s t tests.

Fig. 3. IL25 does not enhance thymus regeneration.

a Injection regime for the administration of recombinant IL25 (rIL25), 2.5μg/mouse/injection or PBS control. Three injections were given prior to sub-lethal irradiation (SLI) (1 × 425 rad) and one injection post-SLI exposure. Mice were harvested either 7 or 35 days after SLI. b Effects of IL25 administration on TEC recovery at analysis 7 days post-SLI, and TEC and thymocyte recovery at 35 days post-SLI (c). For (b), PBS and rIL25 n = 8 animals obtained over 2 independent experiments. For (c), TEC PBS and rIL25 n = 8 and thymocytes PBS and rIL25 n = 8 animals obtained over 2 independent experiments. All bars show mean ± SEM. P-values were obtained using two-tailed, unpaired Student’s t tests.

Fig. 4. IL33 Expands ILC2 In The Thymus.

a Injection regime for the administration of rIL33 or rIL25 in Red5xRagGFP mice, 2.5μg/mouse/injection or PBS control, four injections given. b Representative FACs plots, where n = 6, to illustrate ILC gating and effect of PBS/rIL33/rIL25 administration, total ILC are gated as CD45⁺RagGFP⁻CD4⁻CD8^-TCRβ^-IL7Rα⁺Lin⁻ cells, with ILC2 further identified as KLRG1⁺NK1.1⁻ cells. c Quantitation of effects of rIL33 and rIL25 administration on total thymus cellularity, ILC2 and IL5⁺ILC2 in Red5xRagGFP mice. Significance is of rIL33 or rIL25 treatment compared to PBS controls, (n = 6 animals for each condition). All analysis was carried out across two independent experiments. All statistical analysis was obtained using one-way ANOVA with Dunnett’s multiple comparisons. All bars show mean ± SEM.

Fig. 5. IL33 expands ILC2 To enable intrathymic expansion of eosinophils.

a Injection regime of Il7ra^Cre (control) and Il7ra^CrexRora^f/fl mice with PBS or 2.5μg rIL33, with three injections prior to SLI (1 × 500 rad) and then one injection post-SLI. b Quantitation of TEC recovery (n = 5–7) in control and ILC2 KO mice at 35 days post-SLI following PBS or rIL33 treatment. c Thymic eosinophil numbers in control and ILC2 KO mice at day 7 post-SLI and following PBS or rIL33 treatment (n = 6 animals from 2 independent experiments). Statistical significance is shown comparing Il7ra^Cre (Cre) PBS (n = 8 animals from 4 independent experiments) with Il7ra^Cre (Cre) rIL33 (n = 5 animals from 4 independent experiments) or Il7ra^CrexRora^fl/fl(ILC2KO) PBS (n = 6 animals from 4 independent experiments) with Il7ra^CrexRora^fl/fl(ILC2KO) rIL33 (n = 7 animals from 3 independent experiments). Eosinophils were identified as CD45⁺CD4⁻CD8^-TCRβ^-TER119⁻CD11b⁺Siglec-F⁺. All statistical analysis was obtained using one-way ANOVA with Šídák’s multiple comparisons. All bars show mean ± SEM.

Fig. 6. IL33 expression selectively maps to Sca1^± thymic mesenchyme.

a Representative FACs plots illustrating the proportion of IL33^cit expressing CD45⁺ thymic cells and total CD45^- thymic stromal cells (blue histogram). Red histograms represent florescence levels in control WT mice. b Quantitation of IL33^cit expression in defined thymic stromal populations; cTEC (CD45^-EpCAM-1⁻UEA1⁻Ly51⁺), mTEC (CD45^-EpCAM-1⁻UEA1⁺Ly51⁻), mTEC^hi (CD45^-EpCAM-1⁻UEA1⁺Ly51⁻MHCII⁺CD80⁺), mTEC^lo(CD45^-EpCAM-1^-UEA1⁺Ly51⁻MHCII^-CD80⁻), Endothelium (TER119⁻CD45⁻EpCAM-1⁻CD31⁺), Mesenchyme (TER119⁻CD45⁻EpCAM-1⁻CD31⁻Integrinα7⁻), Pericytes (TER119⁻CD45⁻EpCAM-1⁻CD31⁻Integrinα7⁺), (where n = 7 animals for TEC subsets and n = 5 animals for all other subsets, obtained from 2 independent experiments). c Representative FACs plots showing Sca-1 expression in thymic mesenchyme, and IL33^cit expression within Sca-1⁺ and Sca-1⁻ subsets. Blue histograms are from IL33^cit mice, red histograms from control WT mice. d Total thymus cellularity, Sca-1⁺ mesenchyme numbers/proportions in WT mice following SLI exposure at indicated time points (n = 6 animals obtained from 2 independent experiments) with statistical analysis from a one-way ANOVA comparing means of each group to D0 with Dunnett’s multiple comparisons. e Comparative analysis of TEC (blue line) and Sca-1⁺ mesenchyme (red line) recovery at indicated time points in WT mice (n = 6 animals obtained from 2 independent experiments). f FACs plots showing IL4Rα expression in Sca-1⁺ mesenchyme from WT (blue histogram) mice, where against Il4ra^−/− mice (red histogram) were used for control staining levels. g Quantitation of Sca1⁺ mesenchyme in WT and Il4ra^−/− mice at steady state (D0) (WT n = 5 and Il4ra^−/− n = 6 animals obtained from 2 independent experiments), from a two-tailed unpaired Student’s t test. h Injection and irradiation regime of 5μg recombinant IL4-complexes/PBS into WT mice. Analysis of Sca-1⁺ mesenchyme was performed at day 7 and day 14 after injection regime, n = 6 minimum. i Numbers of Sca-1⁺ thymic mesenchyme cells at day 7 and day 14 post-SLI following PBS (D7 n = 6 and D14 n = 7 animals) or IL4c (D7 n = 7 and D14 n = 7 animals) injection, obtained from 2 independent experiments. Statistics from a two-tailed unpaired Student’s t test. All analysis was carried out across at least two independent experiments. All error bars show mean ± SEM.

Fig. 7. Eosinophil production of IL4 controls thymus regeneration.

a Representative FACs of total thymus cells from IL4^GFP mice, with analysis of CD1d-PBS57 tetramer and Siglec-F expression shown in GFP cells at D0 (a) and D1 post-SLI (b). c, d Show quantitation of total IL4^GFP, eosinophils and NKT cells within thymus at day 0 and day 1 post-SLI, n = 6 animals obtained from 2 independent experiments. e Representative FACs plots identifying ILC2 in thymus of IL4-GFP mice at D0 and D1 post-SLI. f Quantitation of ILC2 and IL4-GFP⁺ ILC2 in thymus of IL4-GFP mice D0 (black circle) and D1 post SLI (white circle), n = 8 animals obtained from 2 independent experiments. g Quantitation of Sca-1⁺ mesenchyme in WT (black circle) and ΔdblGATA (white circle) mice at day 0 (WT n = 13 from 4 independent experiments, ΔdblGATA n = 8 from 2 independent experiments) and day 14 (WT n = 14 from 4 independent experiments, ΔdblGATA n = 8 from 2 independent experiments) post-SLI. h Injection regime of PBS or 5μg rIL4-complex into WT/ΔdblGATA mice, three injections prior to SLI and one after, mice harvested 14 days or 35 days post-SLI. i Analysis day 14 post-SLI of Sca-1⁺ mesenchyme in ΔdblGATA injected as (h) (n = 6 animals obtained from 2 independent experiments). j Images show gross thymus size at day 35 post-SLI in WT and ΔdblGATA mice injected with either PBS or rIL4c. k Analysis at day 35 post-SLI of total thymus cellularity, and TEC and thymocyte subsets in ΔdblGATA mice following either PBS (black circle) or IL4c injection (white circle), where for TEC analysis PBS n = 5 and IL4c n = 6 and for thymocyte analysis PBS n = 6 and IL4c n = 7 obtained from 2 independent experiments. Statistical analysis was performed using a two-tailed unpaired Student’s t test. All significance shows mean ± SEM.