(A) Mean (± SEM) body weights and (B) survival in mice with inducible inactivation of FAS in the gut. Littermates in a series of experiments received injections of Tamoxifen or Vehicle on Days 1–5. FAS-villin (FAS floxed alleles with Villin-Cre) + Tamoxifen, n=46; FAS floxed (without Cre) + Tamoxifen, n=28; FAS-villin + Vehicle, n=7; FAS floxed + Vehicle, n=4, Villin-Cre (without FAS floxed) + Tamoxifen, n=9. For the survival curve, the control symbol represents a composite of all the control groups since there were no deaths in any controls. FAS-villin symbols represent FAS-villin mice treated with tamoxifen on days 1–5. P<0.01 for survival by logrank testing. (C) In vivo detection of orally administered tracer (10 kDa dextran). Representative imaging is shown on the left and quantitation of fluorescent imaging (presented as pixels detected over a standard area of the abdomen) on the right. *P<0.01 (with data presented as mean ± SEM) by two-tailed unpaired t-test for 12 control and 14 FAS-villin mice. (D) TNFα (25 control and 15 FAS-villin mice), (E) endotoxin (31 control and 23 FAS-villin mice), and (F) LBP (27 mice for each condition) serum levels assayed on day 9 as denoted in A. *P<0.01 (with data presented as mean ± SEM) by two-tailed unpaired t-test. (G) Levels of secreted cytokines and (H) MPO activity from colons (5 control and 4 FAS-villin) 9 days after initiation of tamoxifen injections. Mean values ± SEM are plotted with P values shown or *P<0.05 by two-tailed unpaired t-test. (I) Histology scores (mean ± SEM) for the ceca of 6 control and 8 FAS-villin mice. (Also see Fig. S2.)

(A) Staining for chromogranin A (which detects most enteroendocrine cells) and (B) Alkaline phosphatase (biomarker of the enterocytic lineage) in colons of control (FAS floxed animals without Cre treated with tamoxifen) or FAS-villin (FAS floxed animals with villin-Cre treated with tamoxifen) mice. (C) Alcian blue staining (to detect goblet cells) in mouse colon. (D) Muc2 immunostaining in mouse colon. (E) mRNA levels for Muc2 and (F) Tff3 in mouse colon at days 7 and 9 as denoted in Fig. 2A (n=5 per condition; mean values ± SEM are plotted; *P<0.05 by t-test). (G) Disruption of the colonic mucus layer in FAS-villin mice detected by Muc2 immunocytochemistry. The white dashed lines denote the epithelial cell surface while the yellow dashed lines indicate the luminal border of the inner mucus layer. (H) Average thickness of the inner mucus layer measured at the proximal colon. Measurements were made perpendicular to the inner mucus layer at 5–7 different sites for each colon. Data represent thickness determinations for 5 mice per condition. Mean values ± SEM are plotted. *P<0.05 (t-test). (I) Merged images of Muc2 (red) and DBA (green) staining in the colons of control and FAS-villin mice. Scale bar: 50 μm. (Also see Fig. S3)

(A) Western blotting for FAS and actin in colon extracts from diabetic mice. For panels on the top, streptozotocin (STZ) or vehicle (control) was injected into C57BL/6 male mice. For panels on the bottom, non-diabetic mice with no expression of a mutant Kir6.2 and hyperglycemic mice with overexpression of a mutant Kir6.2 induced by RIP-Cre were studied. (B) Serum fluorescence (expressed as relative fluorescence units or RFUs, mean ± SEM) following gavage of mice with FITC-dextran (4 kDa). (C) Detection of disruption of the inner mucus layer by anti-Muc2 staining in STZ-treated mice. The white dashed lines mark the epithelial cell surface while the yellow dashed lines identify the luminal border of the inner mucus layer. (D) Colons were fixed in methanol-Carnoy fixative without disruption of fecal contents following by determination of the thickness of the inner mucus layer. Measurements were made perpendicular to the inner mucus layer at 5–7 different sites for each colon. Data represent thickness determinations for 6–7 mice per condition. Mean values ± SEM are plotted. *P<0.05 (t-test). (E) Alcian blue staining for mucus-producing goblet cells in the colons of control or STZ mice. The arrow indicates the inner adherent mucus layer. (F) Fluorescent in situ hybridization for bacteria (red) in colon from wild type and FAS-villin mice. Nuclei are stained with DAPI. As expected, bacteria are detected by red color in the lumen, but arrows denote bacterial penetration of the epithelium in FAS-villin colon. (G) Fluorescent hybridization of bacteria in colon from control or STZ mice. The double-headed arrow identifies a clear space separating bacteria and epithelium in control mice. Single arrows denote bacterial penetration into crypts in STZ mice. Scale bars: 50μm.(A lso see Fig. S5.)

(A) Left panel: Western blotting of FAS and actin in intestinal extracts from tamoxifen-treated villin-Cre mice (without floxed FAS alleles), tamoxifen-treated FAS floxed mice (without villin-Cre), tamoxifen-treated FAS villin-Cre mice (with both floxed FAS alleles and villin-Cre), and from vehicle-treated FAS villin-Cre mice. Right panel: Western blotting of FAS and actin in brain, lung, liver, spleen, and fat recovered from FAS floxed (FASf/f) and FAS-villin (with both floxed FAS alleles and villin-Cre) mice following tamoxifen treatment. (B) Upper panel: Western blotting of FAS and actin in colon from FAS-villin animals at various time points after tamoxifen treatment. Bottom panel: q-RT-PCR for intestinal FAS expression in control (C) or FAS-villin mice on days 3–9 after initiation of tamoxifen treatment (n=3 per condition, data presented as mean ± SEM). White indicates control while red indicates tamoxifen treatment. (C) Confirmation of intestinal deletion of FAS in FAS-villin mice by immunocytochemistry. The first and third rows show FAS immunostaining (green) in jejunum and colon of control and FAS-villin mice after tamoxifen treatment. The second and fourth rows show DAPI staining (blue). Scale bars: 50 μm.(A lso see Fig. S1.)

(A) Weight of donor mice following initiation of tamoxifen treatment on day 1. (B) Weight of recipient germ-free mice. (C) Body composition of recipient germ-free mice. (D) Body weight, (E) feces wet/dry ratio, (F) serum endotoxin, and (G) serum TNFα in control and FAS-villin mice with and without daily antibiotic treatment. P values are shown or *P<0.05. Data are presented as mean ± SEM. (Also see Table S1.)

(A) Muc2 protein content assayed by ELISA (mean ± SEM) in colon extracts from mice at day 7 as denoted in Fig. 2A. (B) Muc2 protein content (mean ± SEM) secreted from LS174T cells following lentivirus-mediated shRNA knockdown of FAS. Verification of the knockdown is shown by Western blotting in the bottom panels. (C) The HA-tagged Muc2 N-terminus (N-Muc2) was expressed in COS7 cells, which were treated with a control (scrambled) or FAS-specific siRNA. The top panel shows Western blotting with an antibody to FAS, and the bottom panels blotting with an anti-HA antibody. (D) HA-tagged FAS was overexpressed in COS cells using a retrovirus, then N-Muc2 was expressed in the same cells. The top panel verifies expression of the tagged FAS. The bottom panels show Western blotting of N-Muc2, distinguished from FAS by differences in protein size. (E) Biotin switch assay demonstrates S-palmitoylation of N-Muc2 by detecting a biotin moiety that is hydroxylamine (NH₂OH)-dependent. (F) Sequence alignment of human, mouse and rat N-Muc2 with the conserved Cys12 residue (mouse) indicated in green and the conserved Cys163 and Cys166 residues (mouse) indicated in yellow. (G) Western blotting of lysates or culture supernatants for N-Muc2 mutants. (H) Biotin switch assay using wild type N-Muc2 and N-Muc2 mutants. (I) Biotin switch assay for N-Muc2 S-palmitoylation in control (scrambled) and FAS knockdown cells. (Also see Fig. S4.)

(A) Glucose levels in mice injected with STZ on day 0 (solid arrow), then implanted with pumps delivering saline or insulin (7 mice per condition) on day 3 (open arrow). (B) Western blotting of FAS protein in colons of mice at day 12. (C) Body weight of treated mice. (D) Inner mucus layer thickness for control (n=4), saline-treated diabetic mice (STZ+saline, n=7), and insulin-treated diabetic mice (STZ+insulin, n=7). (E) Alcian blue images from the animals shown in panel D. (F) Intestinal permeability assay of animals after gavage with FD4 (n=5 for non-diabetic controls and n=7 for each diabetic group). P values are shown or *P<0.05 (data presented as mean ± SEM).