Fig 2: Inhibition of colitis by human TNF blockade in humanized TNF model of colitis.A Humanized TNF colitis model. Rag1-/- or hTNF-KI x Rag1-/- mice were reconstituted with naive T cells from either WT or hTNF-KI mice. Mice were treated once they had lost > 5 % of their initial weight twice per week for two or three weeks with various anti-TNF agents or respective controls (10 mg/kg; i.p.). B Weight of hTNF-KIxRag1-/- mice reconstituted with naive T cells from hTNF-KI mice and treated with infliximab (n = 6) or Fc control (n = 5) (10 mg/kg) twice per week. Data are representative of two independent experiments. C Colitis inflammation scores of hTNF-KIxRag1-/- mice reconstituted with naive T cells from hTNF-KI mice upon 2 weeks treatment with infliximab (n = 4) or isotype control (n = 8). Data are representative of two independent experiments. D Colitis inflammation scores of hTNF-KIxRag1-/- mice reconstituted with naive T cells from hTNF-KI mice upon 3 weeks treatment with infliximab (n = 6) or isotype control (n = 5). Frequencies of CD11b+ (CD45+CD11b+) E, Th1 (CD45+CD4+TCRß+IFN?+) F, Th17 (CD45+CD4+TCRß+IL-17A+) G and Th1/17 (CD45+CD4+TCRß+IL-17A+IFN?+) H cells after 3 weeks of infliximab (n = 6) or Fc control treatment (n = 5). All data in D–H are representative of two independent experiments. Data represent mean values ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as calculated by Student’s t-test, ns not significant.

Fig 3: TNF controls IL-22BP expression in DCs.A TNF blockade increases levels of bioactive colonic IL-22 that can be inhibited by recombinant IL-22BP. Supernatants derived from colonic explants of Rag1-/- mice which received naive hTNF-KI T cells and treated for 2 weeks with anti-TNF, were added to human colonic Colo205 cell line with or without rmIL-22BP (1,25 mg/ml). Production of human IL-10 as an indicator of bioactive IL-22 has been determined 48 h later. B Expression of IL-22BP mRNA in sorted CD45+CD11c+MHCII+CD103+ and CD45+TCRß+ cells from the colon of mice treated with Fc control or infliximab for 2 weeks. Naive hTNF-KI T cells were transferred to Rag1-/- recipients, anti-TNF (infliximab; 10 mg/kg; i.p. twice per week) was administered once mice had lost 5% of their initial weight. C Expression of IL-22BP in colon and spleen from naive WT and TNF-/- mice. D Expression of IL-22BP in spleen from naive WT, T-TNF-/-, B-TNF-/-, and M-TNF-/- mice. E Expression of IL-22BP in sorted CD4+CD11c+MHCII+ or CD8+CD11c+MHCII+ cells isolated from spleen of naïve WT and TNF-/- mice. CD4+CD11c+MHCII+ cells from TNF-/- mice were incubated with rmTNF (100 ng/ml; 4 h). F Levels of human IL-22BP in human monocyte-derived DCs (moDCs) stimulated with TNF (100 ng/ml), LPS (100 ng/ml), PAM3Cys (5 mcg/ml) for indicated times. G Correlation between hTNF and hIL-22BP levels in sera of IBD patients. All mouse data are representative of two independent experiments. Data represent mean values ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as calculated by Student’s t-test; ns not significant. The Pearson correlation was used for correlative analyses. The significance level was set to p = 0.05.

Fig 4: T cell-derived TNF modulates levels of IL-22 in vivo via control of IL-22BP expression during established colitis.A Levels of IL-22 in ex vivo colonic explants from mice treated with Fc control or infliximab for 2 weeks. B IL-22 mRNA levels in the colons of mice treated with Fc control or infliximab for 2 weeks. C IL-22BP mRNA levels in the colons of mice treated with Fc control or infliximab for 2 weeks. D Serum IL-22BP concentration in sera, E Ratio of IL-22 and IL-22BP mRNA levels in the colons of mice treated with Fc control or infliximab for 2 weeks. Naive hTNF-KI T cells were transferred to Rag1-/- recipients, anti-TNF (infliximab; 10 mg/kg; i.p. twice per week) was administered once mice lost 5% of their initial weight. F Representative images of colon sections stained with Hematoxylin/Eosin and immunohistochemically stained for Ki67 (brown) and CD3 (red), G weight changes, H corresponding inflammation scores, I percentage of proliferating colonic epithelial cells after treatment of colitic mice with either Fc control, infliximab, anti-IL-22 or infliximab/anti-IL-22 (Fc control and infliximab - 10 mg/kg; anti-IL-22 - 10 mcg/mouse; i.p, twice per week) for 2 weeks. Naive hTNF-KI T cells were transferred to Rag1-/- recipients, antibodies were administered once mice had lost 5% of their initial weight. Scale bar is equal to 100 µm in x100 magnification and 25 µm in x400 magnification. J Colonic inflammation score at various time points after start of anti-mTNF treatment. Percentage of proliferating colonic epithelial cells K, IL-22BP mRNA levels L in the colons of mice treated with Fc control or anti-mTNF for 2 weeks. All data are representative of two independent experiments. Data represent mean values ± SEM. *p < 0.05, **p < 0.01,***p < 0.001, as calculated by Student’s t-test; ns not significant.

Fig 5: T cell-derived TNF is pathogenic during established colitis in hTNFRp75-KIxRag1-/- mice.A Weight of hTNFRp75-KIxRag1-/- mice reconstituted with naive T cells from hTNF-KI mice and treated with Infliximab or Fc control (10 mg/kg) twice per week. IL-22 levels in colonic explants B and expression of IL-22BP mRNA in colon C of hTNFRp75-KIxRag1-/- mice reconstituted with naive T cells from hTNF-KI mice and treated with-infliximab for 2 weeks. Frequencies of CD11b+ (CD45+CD11b+) D of granulocytes (CD45+CD11b+Ly6G+Ly6C+) and inflammatory monocytes (CD45+CD11b+Ly6G-Ly6C+) E, Th1 (CD45+CD4+TCRß+IFN?+) F, Th17 (CD45+CD4+TCRß+IL-17A+) G and Th1/17 (CD45+CD4+TCRß+IL-17A+IFN?+) H cells 2 weeks after infliximab or Fc control therapy. All data are representative of two independent experiments. Data represent mean values ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, as calculated by Student’s t-test; ns not significant.