Virus-plus-susceptibility gene interaction determines Crohn's disease gene Atg16L1 phenotypes in intestine

Abstract

It is unclear why disease occurs in only a small proportion of persons carrying common risk alleles of disease susceptibility genes. Here we demonstrate that an interaction between a specific virus infection and a mutation in the Crohn's disease susceptibility gene Atg16L1 induces intestinal pathologies in mice. This virus-plus-susceptibility gene interaction generated abnormalities in granule packaging and unique patterns of gene expression in Paneth cells. Further, the response to injury induced by the toxic substance dextran sodium sulfate was fundamentally altered to include pathologies resembling aspects of Crohn's disease. These pathologies triggered by virus-plus-susceptibility gene interaction were dependent on TNFalpha and IFNgamma and were prevented by treatment with broad spectrum antibiotics. Thus, we provide a specific example of how a virus-plus-susceptibility gene interaction can, in combination with additional environmental factors and commensal bacteria, determine the phenotype of hosts carrying common risk alleles for inflammatory disease.

Figure 1. Murine norovirus (MNV) infection triggers Paneth cell abnormalities in Atg16L1 mutant mice

(A-B) Atg16L1 hypomorph (Atg16L1^HM) and WT mice in the MNV-free barrier facility were orally inoculated with 3×10⁷ particle forming units (pfu) of MNV CR6 for 7 days or left untreated (n > 6 mice for each group). (A) Light microscopy images of ileal sections stained with PAS-alcian blue. Dotted line denotes crypt unit containing several Paneth cells each, and arrowheads indicate typical granules. For the two representative images of Atg16L1^HM mice infected with MNV CR6, red dotted circle denotes aggregated granules and red arrows indicate granules with abnormal staining and size. Scale bar represents 10 μm. (B) Indirect immunofluorescence of ileal sections stained for lysozyme (red) and nuclei (blue). Dotted line denotes crypt unit. Scale bar represents 10 μm. (C) Four types of lysozyme distribution patterns observed in Paneth cells: normal (D₀), disordered (D₁), depleted (D₂), diffuse (D₃). (D) Percentage of Paneth cells displaying each of the four types of lysozyme distribution patterns from WT and Atg16L1^HM mice that were uninfected or inoculated with MNV CR6 (n > 5,700 cells from 3 mice/condition, *p < 0.05, **p < 0.01, ***p < 0.001, mean +/−SEM). See also Figure S1.

Figure 2. Properties of MNV associated with Paneth cell abnormalities in Atg16L1 mutant mice

(A) Light microscopy of ileal sections stained with PAS-alcian blue of WT and Atg16L1^HM mice 7 days post infection with 3×10⁷ pfu of MNV CW3 (n = 6 mice/genotype). Dotted line denotes crypt unit and arrowheads indicate Paneth cell granules. Scale bar represents 10 μm. (B) Number of Paneth cells displaying abnormal lysozyme distribution from WT and Atg16L1^HM mice in (A) (n > 6100 cells from 3 mice/condition, mean +/−SEM). See also Figure S2.

Figure 3. MNV infection induces a distinct transcriptional response in Paneth cells from Atg16L1 mutant mice

Factorial design analysis was performed to identify gene sets showing differential gene expression in Paneth cells from WT and Atg16L1^HM mice with or without MNV CR6 infection. Venn diagram displays the subsets of genes found to be differentially expressed (p < 0.05 as defined using a linear model and exhibiting a fold-change > 1.5) in the respective comparisons as indicated (circles 1 to 3). The ‘differential-of-differential’ analysis in the factorial design identified genes that respond differently to infection in Atg16L1^HM mice compared to WT mice (Zones A, B, C, and D bounded by the red circle). These were found to be enriched for genes associated with intracellular protein traffic, protein targeting and localization, and amino acid metabolism. Also noteworthy is the sizable representation of genes associated with carbohydrate metabolism. The expression profiles for genes in these biological process categories are displayed as Log₂-transformed fold-changes in the heatmaps. The ‘differential-of-differential’ analysis also identified genes that were not significantly altered in response to infection when compared within each genotype but exhibited significant differences when comparing responses to infection between genotypes as highlighted in Zone A. Genes with transporter functions or those with ligase activity were significantly enriched in this subset. See also Figure S3 and S4.

Figure 4. Atg16L1 mutant mice display a virus-dependent aberrant response to DSS in the colon

(A) WT and Atg16L1^HM mice were orally inoculated with MNV CR6 for 7 days or uninfected, and subsequently given 2.5% DSS for an additional 7 days at which point intestines where harvested. Light microscopy images of H+E stained sections of ulcerated regions located immediately adjacent to the ano-rectal junction are shown. Yellow double-headed arrows indicate the muscularis propria thickness. In the MNV CR6-infected Atg16L1^HM sample, lymphoid aggregates are indicated by yellow dashed circles, and the black dashed region contains submusocal fibrosis and inflammation. Scale bar represents 500 μm. (B) Table summarizing the outcome of DSS treatment. All intestines were harvested at the end of DSS treatment for analysis. + and − refer to the presence or absence of inflammatory hallmarks of Crohn’s disease respectively (n ≥ 6 mice/condition). (C) Quantification of the muscularis propria thickness in the region adjacent to the anorectal junction from DSS treated mice. 7 days and 0 days refer to the amount of time mice were inoculated with the virus prior to DSS treatment. The Increase in muscle thickness was normalized to the average of uninfected WT mice treated with DSS (*p < 0.05, ***p < 0.001, mean +/−SEM). (D) Number of lymphoid aggregates in the region adjacent to the ano-rectal junction from DSS treated mice (***p < 0.001, mean +/−SEM). N.D. refers to not detected. See also Figure S5.

Figure 5. Virus-infected Atg16L1 mutant mice display mucosal atrophy in the ileum in response to DSS

(A) Light microscopy of H+E stained ileal sections from the same mice in Figure 4. Scale bar represents 200 μm. (B) Quantification of average villus height normalized to uninfected WT mice treated with DSS. 7 days and 0 days refer to the amount of time mice were inoculated with the virus prior to DSS treatment (an average of 286 total villi from at least 5 mice/condition was measured, **p < 0.01, ***p < 0.001, mean +/−SEM). See also Figure S5.

Figure 6. TNFα and IFNγ inhibition or antibiotics treatment ameliorates DSS-induced disease in virus-infected Atg16L1 mutant mice

(A) Isotype control or blocking antibodies against TNFα and IFNγ were administered to WT and Atg16L1^HM mice infected with MNV CR6 for 7 days followed by additional 7 days of DSS treatment at which point intestines where harvested. The muscularis propria thickness in the region adjacent to the ano-rectal junction was quantified and normalized to the average of WT mice receiving isotype control antibodies (n ≥ 6 mice/condition, *p < 0.05, mean +/−SEM). (B) Number of lymphoid aggregates in the region adjacent to the ano-rectal junction from mice in (A) (*p < 0.001, mean +/−SEM). (C) Quantification of average villus height from mice in (A) normalized to the average of WT mice receiving isotype control antibodies (an average of 263 total villi from at least 5 mice/condition was measured, *p < 0.05, ***p < 0.001, mean +/−SEM). (D) WT and Atg16L1^HM mice were orally inoculated with MNV CR6 for 7 days and then given 2.5% DSS and broad spectrum antibiotics concurrently for another 7 days at which point intestines where harvested. The muscularis propria thickness was normalized to the average of uninfected WT mice treated with DSS from Figure 4D (n = 6 mice/condition, *p < 0.05, mean +/−SEM). (E) Number of lymphoid aggregates from mice in (D) (***p < 0.001, mean +/−SEM). (F) Quantification of average villus height from mice in (C) normalized to the average of uninfected WT mice treated with DSS from Figure 5B (an average of 320 total villi from at least 5 mice/condition was measured, *p < 0.05, mean +/−SEM). See also Figure S5.