Fig 1.
Dynamics of RELM-β expression during C. rodentium infection.
(A) Enumeration of tissue adherent or luminal C. rodentium following oral gavage. Results show mean values of 3–4 mice/group. (B) qPCR analysis of RELM-β, Muc2, and TFF3 mRNA levels in the distal colons of C57BL/6 mice. Results show mean values of 3–4 mice/group, **P = 0.0014 vs. uninfected; ***P < 0.0001 vs. uninfected. (C) Western blots of RELM-β protein in colonic tissues and stool lysates of uninfected or infected mice. Lanes show data from single mice, representative of >4 mice/group. + indicates rRELM-β positive control. (D) Presence of RELM-β in the serum as detected by ELISA in uninfected or infected (10 DPI) mice. Each data point represents 1 mouse. Error bars = SEM. ** P < 0.001. (E) Western blot of RELM-β within stool of mice 7 days following their natural exposure to C. rodentium infected mice. Each lane follows 1 animal over time. (F) Representative immunofluorescent staining of RELM-β in tissues of uninfected or 7 days post exposure (DPE) mice following natural transmission. n = 4/group. Scale bars = 100 μm. Original magnification = 200X. All results are representative of at least 2 independent experiments.
Fig 2.
Retnlb-/- mice are highly susceptible to C. rodentium-induced colitis.
(A) Body weights following infection of C57BL/6 and Retnlb-/- mice. Error bars = SEM. *P < 0.05; * P < 0.05 *** P < 0.0001 (vs. 0 DPI). Results are representative of at least 3 experiments, n = 3–6/group. (B) Infection survival curve of C57BL/6 mice and Retnlb-/- mice. Starting numbers (i.e at 0 DPI) were WT, n = 9; and Retnlb-/-, n = 11. Results are representative of 3 experiments. (C) Resected colons and ceca of uninfected and 10 DPI C57BL/6 and Retnlb-/- mice. Results are representative of ≥ 4 infections, n = 3–6/group. (D) Representative H&E-stained colon sections from C57BL/6 and Retnlb-/- mice at 10 DPI (n ≥10/group). (E) Ulcer frequency and size in C57BL/6 vs. Retnlb-/- mice at 10 DPI. Bars show average mean values of ≥ 3 independent experiments, each with n = 2–4 mice/group. (F) Histopathology scoring of C57BL/6 vs. Retnlb-/- mice (n ≥ 10 per group). *** P < 0.0001. All error bars = SEM.
Fig 3.
Disease severity of Retnlb-/- mice is associated with greater C. rodentium invasion of colonic crypts.
(A) C. rodentium enumeration in the colon tissue and luminal compartments of Retnlb-/- and C57BL/6 following infection. Results show means of 3–4 (2 and 4 DPI) and 6–11 (6–10 DPI) animals, and are representative of 2 independent experiments **P = 0.0047, *P = 0.0262; ##P = 0.0015 (Retnlb-/- vs. C57BL/6); Mann-Whitney U-test. (B) C. rodentium staining in mouse colonic tissues (6 & 10 DPI). Original magnification = 200X. Results are representative of 4–6 mice/group. (C) Quantitation of C. rodentium LPS-filled crypts in Retnlb-/- and C57BL/6 mice (10 DPI). Results represent mean of total crypts counted in single mice, pooled from 3–6 mice/group. *P < 0.05. (D) Left panel: H&E staining of heavily ulcerated colonic tissue from a Retnlb-/- mouse (10 DPI). M = Mucosa; SM = submucosa. Right Panel: dual FISH staining using a universal EUB338 probe (red) and Gamma-proteobacteria specific probe (green), revealing clusters of C. rodentium (yellow) deep within ulcerated tissue, presumably where crypt bases were located. Cr—C. rodentium. Results are representative of all ulcerated Retnlb-/- mice analyzed (n = 5). (E) Enumeration of systemic C. rodentium burdens in C57BL/6 vs. Retnlb-/- mice (10 DPI). Each data point represents one animal. **P = 0.0037, Mann-Whitney U-test.
Fig 4.
Retnlb-/- mice display defective epithelial proliferative responses predisposing to deep infection of colonic crypts.
(A) H&E staining of infected colons at 8 DPI. Scale bar = 50 μm. Original magnification = 200X. (B) Ki67 staining (green) of infected colons (8 DPI) as well as DAPI staining of host cell nuclei (blue) as shown. Scale bar = 50 μm. Graph on right: Bars represent mean average value of Ki67+ cells/100 IECs, from 3 mice/group. (C) Dual labeling for BrdU (red) and LPS (green) in C57BL/6 and Retnlb-/- mice. Original magnification = 200X. Scale bar = 100 μm. Graph on right: Bars represent the mean average number of BrdU+ cells per 100 cells in 3 mice/group. Error bars = SEM. **P = 0.0068, ***P = 0.0002. Results are representative of at least 3 independent experiments.
Fig 5.
Reduced mucosal CD4+ T-cell numbers within infected Rag1-/- and Retnlb-/- mice, leading to defective epithelial cell proliferation.
(A) CD4 staining in the colons of infected Rag1-/- mice reconstituted with either PBS or CD4+ T cells. (B) Ki67 staining in PBS or CD4+ T cell-reconstituted Rag1-/- mice (10 DPI). Graph on right (B1): Enumeration of Ki67-labeled cells. Bars represent mean average number of Ki67+ cells/100 cells. Error bars = SD. *P < 0.01. (C) LPS staining showing deeper penetration of C. rodentium into the crypts of PBS-reconstituted Rag1-/- mice (arrows) as compared to CD4+ T cell-reconstituted Rag1-/- mice (10 DPI). Scale bar = 100 μm. Original magnification = 200X. (D) CD4 staining in uninfected or 10 DPI C57BL/6 and Retnlb-/- mice. (E) Enumeration of CD4+ cells within colonic sections. Results represent means of cells counted from 5–10 high power fields (400X)/mouse. *P < 0.05. Error bars = SEM. (F) Flow cytometry of lamina propria lymphocytes purified from infected C57BL/6 and Retnlb-/- mice (8 DPI). All results are representative of 2–3 independent experiments, with 3–5 mice/group.
Fig 6.
RELM-β acts as a CD4+ T cell chemoattractant.
(A) qPCR analysis of chemokine genes involved in T-cell recruitment in colonic tissues of uninfected or infected C57BL/6 and Retnlb-/- mice at 6 and 10 DPI. Results are means of n = 4–8 mice/group, pooled from 2 independent experiments. *P < 0.05 Retnlb-/- 6 DPI vs. C57BL/6 UI, Bonferroni’s post-test following 1-way ANOVA. (B) Graph showing percent CD4+ T cells migrating toward RELM-β in a chemotaxis assay. Results are representative of 3 independent experiments, 4 replicates/group. *P < 0.05; **P < 0.001. Error bars = SEM.
Fig 7.
IL-22 induces IEC proliferation in vitro and its production is impaired in Retnlb-/- mice.
(A) qPCR analysis of IL-22 gene transcripts in colonic tissues of infected C57BL/6 and Retnlb-/- mice at 8 DPI. (B) IL-22 protein levels measured via ELISA in colonic (left) and cecal (right) tissues of infected C57BL/6 and Retnlb-/- mice at 8 DPI. Results are means of n = 6 mice/group, *P < 0.05, **P < 0.005, ***P < 0.0005. All error bars = SEM. (C) Ki67 staining and enumeration of Ki-67+ CMT93 cells untreated (PBS) or treated with recombinant mouse IL-22 in culture. Bars represent average number of Ki67+ cells per 60X field. Fig 7A and 7B show results representative of 3 independent experiments with n = 4–10 mice per group and Fig 7C was presented as an average of 3 independent experiments performed with at least 3 replicates per condition. **P < 0.01.
Fig 8.
Reconstitution with recombinant RELM-β protects infected Retnlb-/- mice.
(A) Resected colons + ceca from C57BL/6 mice or Retnlb-/- mice treated with either rRELM-β protein (400 ng) or vehicle (8 DPI). (B) H&E staining of colon tissues at 8 DPI (left panel) and their histopathological scoring (right panel B1) in the above mentioned mice. Scale bar = 100 μm. (C) Flow cytometry of CD4+ cells isolated from infected colons (8 DPI). (D) Colonic Ki67 staining (8DPI) in representative mice of each group, with enumeration (E) on right as in Fig 5E. (F) IL-22 gene transcript (left) and protein (right) levels measured via qPCR and ELISA, respectively, in infected C57BL/6 and Retnlb-/- mice (treated with PBS or rRELM-β protein). *P < 0.05; ***P < 0.0001. Error bars = SEM. Scale bar = 100 μm. Fig 8A, 8B, 8B1 and 8D are representative of 4 independent experiments with n = 3 mice/group. The pathology scoring represents an average of these experiments.