Loss of Paneth Cell Autophagy Causes Acute Susceptibility to Toxoplasma gondii-Mediated Inflammation

Abstract

The protozoan parasite Toxoplasma gondii triggers severe small intestinal immunopathology characterized by IFN-γ- and intestinal microbiota-mediated inflammation, Paneth cell loss, and bacterial dysbiosis. Paneth cells are a prominent secretory epithelial cell type that resides at the base of intestinal crypts and releases antimicrobial peptides. We demonstrate that the microbiota triggers basal Paneth cell-specific autophagy via induction of IFN-γ, a known trigger of autophagy, to maintain intestinal homeostasis. Deletion of the autophagy protein Atg5 specifically in Paneth cells results in exaggerated intestinal inflammation characterized by complete destruction of the intestinal crypts resembling that seen in pan-epithelial Atg5-deficient mice. Additionally, lack of functional autophagy in Paneth cells within intestinal organoids and T. gondii-infected mice causes increased sensitivity to the proinflammatory cytokine TNF along with increased intestinal permeability, leading to exaggerated microbiota- and IFN-γ-dependent intestinal immunopathology. Thus, Atg5 expression in Paneth cells is essential for tissue protection against cytokine-mediated immunopathology during acute gastrointestinal infection.

Figure 1:. Paneth cell autophagy is microbiota- and IFN-γ-dependent.

(A) Immunofluorescence of intestinal crypts of naïve GFP-LC3 mice for Paneth cells (UEA, red), LC3 (green), and nucleic acid (Syto62, blue). Bar = 20um. Results are representative of three independent experiments. (B) Quantification of GFP-LC3+ cells in the Paneth cells and enterocytes in the intestinal epithelium from naïve (n=3) mice. Results are representative of three independent experiments. ***P<0.0001. (C) Mice were treated with antibiotic water (ampicillin, streptomycin, neomycin, metronidazole, vancomycin) for 10 days and analyzed by immunofluorescence for Paneth cells (UEA, red), LC3 (green), and nucleic acid (Syto62, blue). Quantification of number of GFP-LC3+ cells per crypt (D) and number of GFP-LC3+ puncta per cell (E) in naïve (white) and antibiotic treated (green) mice. The results are representative of three independent experiments involving at least three mice per group. ***P<0.0001. Data shown are mean ± s.d. (F) qRT-PCR analysis of relative IFN-γ expression, measured in the small intestines of control (CONV), antibiotic-treated mice (CONV+Abx), and (G) microbiologically sterile germ-free (GF) mice (n=6) (H) Mice were treated with anti-IFN-γ blocking antibody and analyzed by immunofluorescence for Paneth cells (UEA, red), LC3 (green), and nucleic acid (Syto62, blue). (I) Quantification of number of GFP-LC3+ cells per crypt and (J) number of GFP-LC3+ puncta per cell in control (white) and treated (green) mice. The results are representative of three independent experiments involving at least three mice per group. ***P<0.0001. Data shown are mean ± s.d.

Figure 2:. Loss of intestinal epithelial autophagy increases immunopathology in response to T. gondii.

(A) Survival of WT (open circles) and E-Atg5 KO (black circles) mice during oral infection with 20 cysts ME49 T. gondii from a combination of three experiments, each involving at least three mice per group. P<0.0001. Histological analysis of small intestines of naïve and infected WT and E-Atg5 KO mice with 20 cysts of ME49 T. gondii on day 7 by (B,C) H&E and (D,E) Alcian Blue/PAS staining at 10x (B,D) and 40x (C,E). Image fields are representative of pathology in multiple tissue sections, and chosen sections were selected by blinded observation. (F) Quantification of number of Paneth cells in T. gondii-infected WT and E-Atg5 KO mice on day 7 post infection. Quantification of number of Paneth cells in naïve mice (100%) is marked by the dotted line. Paneth cells were identified based on their morphology of large granule-containing cells and a basolateral nucleus at the base of the intestinal crypt. (G) Analysis of T. gondii parasite loads by qRT-PCR in spleen and (H) liver. (I) qRT-PCR for IFN-γ transcript in the small intestine and (J) ELISA for IFN-γ in sera of naïve and T. gondii-infected mice. (K) qRT-PCR analysis of relative small intestinal TNF. Results are representative of four independent experiments involving at least four mice per group. *P<0.05, **P<0.005. NS = not significant, error bars = mean ± s.d.

Figure 3:. Intestinal organoids lacking Atg5 are more sensitive to cytokine-mediated death.

(A) Intestinal organoids were generated from small intestinal stem cells isolated from crypts of WT and E-Atg5 KO mice (left, scale bar 10 μm). Paneth cells were visualized by light microscopy (center panel, black arrowheads) and immunofluorescence (yellow arrowheads indicating colocalization of Paneth cell markers UEA (red) and lysozyme (green), scale bar 50μm). (B) Organoids were stimulated with TNF (5ng/ml), IFN-γ (100ng/ml), or both cytokines for 12 hours and imaged by light microscopy. Scale bar 50 μm (C) Quantification of death in WT and E-Atg5 KO organoids, characterized by loss of epithelial integrity by light microscopy. (D) qRT-PCR analysis of organoids stimulated with IFN-γ (100ng/ml) for relative Defcr2 and Defa6 expression. Results are representative of four independent experiments involving at least four mice per group. *P<0.05, *P<0.005. Data shown are mean ± s.d.

Figure 4:. Generation of Paneth cell-specific autophagy-deficient mice.

(A) Representative images of TdTomato-Rosa26/Defa4-cre reporter mice and WT (controls). Co-localization of Defa4 Cre-driven TdTomato expression (red) with lysozyme-expressing Paneth cells (green) and DAPI (blue). (B) Representative hematoxylin and eosin (left) and TEM images (right) of WT and PC-Atg5 KO small intestines. Granule-filled Paneth cells outlined in yellow. (C) Quantification of Paneth cells per crypt in naïve WT and PC-Atg5 KO mice and (D) qRT-PCR analysis of relative small intestinal Defcr2 expression. (E) qRT-PCR analysis for relative Defcr2, Defa6 and Lyz1 expression and (F) Cre transcripts in naïve and T. gondii-infected TdTomato-Rosa26/Defa4-cre mice. (G) Representative images of naïve and T. gondii-infected TdTomato-Rosa26/Defa4-cre mice, counter-stained with lysozyme (green) and DAPI (blue). Results are representative of four independent experiments involving at least four mice per group. *P<0.05, NS = not significant, data shown are mean ± s.d.

Figure 5:. Immunopathology of Atg5 deficiency originates from Paneth cells.

(A) Representative 10x H&E images of WT, E-Atg5 KO, and PC-Atg5 KO mice on day 7 post infection. (B) Survival of WT, E-Atg5 KO and PC-Atg5 KO mice during oral infection with 20 cysts ME49 T. gondii, representative of three experiments each involving at least four mice per group. (C) qRT-PCR analysis of relative small intestinal Defcr2 expression in naïve or T. gondii-infected WT, E-Atg5 KO, and PC-Atg5 KO mice. (D) qRT-PCR analysis of Eubacteria loads in the spleen. (E) Analysis of T. gondii parasite loads by qRT-PCR in spleen and (F) and small intestine. (G) Flow cytometric analysis and quantification of CD4+ IFN-γ+ and CD4+ TNF+ cells in the small intestinal lamina propria of T. gondii-infected WT, E-Atg5 KO, and PC-Atg5 KO mice on day 7 post infection. (H) The distribution of microbiota in the lumens of the small intestines was analyzed by in situ hybridization with Eubacteria-specific (red) probes in mice on day 7 post infection and counterstained with DAPI. Scale bar 100μm. (I) Quantification of distance separating villi from microbiota in T. gondii-infected mice (μm). (J) Detection of FITC fluorescence in sera, four hours after oral gavage with FITC-dextran (4kD) in naïve and T. gondii-infected (day 7 post infection) WT, E-Atg5 KO and PC-Atg5 KO mice. (K) Quantification of sera endotoxin levels by LAL assay in naïve and T. gondii-infected (day 7 post infection) WT, E-Atg5 KO and PC-Atg5 KO mice. Results are representative of four independent experiments involving at least four mice per group. *P<0.05, **P<0.005, ***P<0.0005 (unpaired two-tailed Student’s t test). NS = not significant, data shown are mean ± s.d.

Figure 6:. Intestinal inflammation and immunopathology is microbiota-dependent.

(A,B) Flow cytometric analysis and (C,D) quantification of splenic CD4+ IFN-γ+ (A,C) and CD4+ TNF+ (B,D) in T. gondii-infected control or antibiotic-treated WT, E-Atg5 KO, and PC-Atg5 KO mice on day 7 post infection. (E) Histologic visualization and (F) quantification of Paneth cells by Alcian Blue/PAS staining of small intestinal crypts in WT, E-Atg5 KO and PC-Atg5 KO mice infected with T. gondii and additionally treated with antibiotic drinking water (ampicillin, streptomycin, neomycin, metronidazole, vancomycin) for 10 days. (G) qRT-PCR analysis of relative small intestinal Defcr2 and (H) IFN-γ expression in control and antibiotic-treated WT, E-Atg5 KO and PC-Atg5 KO mice. Results are representative of five independent experiments involving at least four mice per group. *P<0.05, ***p<0.0005. NS = not significant, data shown are mean ± s.d.

Figure 7:. Paneth cell loss and intestinal immunopathology is TNF and IFN-γ-dependent.

(A) Histologic visualization of Paneth cells in naïve or T. gondii infected WT, E-Atg5 KO, and PC-Atg5 KO mice by Alcian Blue/PAS staining. WT, E-Atg5 KO, and PC-Atg5 KO mice were treated with anti-IFN-γ or anti-TNF and then infected orally with 20 cysts per mouse of the ME49 strain of T. gondii. Histological analyses of Paneth cells in small intestines were performed on day 7 after infection as described above. (B) Paneth cell quantification and (C) qRT-PCR analysis of relative small intestinal Defcr2 expression in naïve or T. gondii-infected WT, E-Atg5 KO, and PC-Atg5 KO mice. Results are representative of four independent experiments involving at least four mice per group. *P<0.05, ***P<0.0005. NS = not significant, data shown are mean ± s.d.