Atg16L1^T300A/T300A (Atg16L1^f/f or T300A KI) mice on a C57BL/6J background have been described previously (10). Atg16L1^f/f ERT2Cre⁺ mice were generated by crossed Atg16L1^f/f mice with ERT2Cre mice, which were treated with tamoxifen every other day for 4 times to activate ERT2Cre, and Atg16L1 was deleted to generate Atg16L1 KO mice. All mice within one experiment were cohoused for 4–5 weeks after weaning in the third week of birth to minimize differences between each line (12). All mice were bred and maintained in the SPF barrier at the Institute of Microbiology, Chinese Academy of Sciences (IMCAS). All animal studies were performed according to approved protocols from the IMCAS of Medicine Institutional Animal Care and Use Committee. The genotypes of all mice were genotyped with the following genotyping primers:

Atg16L1^T300A/T300A:

5′-CCTGGAGCTGGGCAGTCAGGTTGGGCTCCATG-3′

5′-GCTGCTTCCCTGTCAGTCAACTGTG-3′

ERT2^Cre:

5′-AAAGTCGCTCTGAGTTGTTAT-3′

5′-GGAGCGGGAGAAATGGATATG-3′

5′-CCTGATCCTGGCAATTTCG-3′

Cohoused, male and female, 7–8-week-old Atg16L1^T300A/T300A mice and their littermate control mice were used. For mild experimental colitis induction, mice were administered with 1.5% DSS (M.W. = 36,000–50,000 Da; MP Biomedicals) in their drinking water for 7 days, followed by regular drinking water for 3 days. Body weight was monitored, with changes in body weight calculated relative to initial body weight. Postmortem whole colons were harvested by blunt dissection, and colon length was measured.

Distal colon segments in the same defined region were collected carefully and were transferred immediately to 4% phosphate-buffered formaldehyde (4% PFA) solution for 24 h and embedded in paraffin. Sections of 5 μm were stained with hematoxylin and eosin (H&E) and then scanned with a light microscope (Zeiss Axio Imager A2) for further histopathological analysis. At least three slides were randomly selected and observed by a blinded pathologist. The tissue damage of each colon was scored based on the degree of epithelial damage and inflammatory infiltrate in the mucosa, submucosa, and muscularis/serosa, as previously described (13).

The isolated small intestine or colon was opened longitudinally after the adipose tissues were stripped off. Using a scissor, the intestine was cut into 2–3-cm equal sections. Each section was placed onto a wax plate, and each end was pinned down so the section was slightly stretched with the mesenteric line uppermost; any excess mesentery was trimmed. Tissues were fixed in 4% PFA for 1 h before being rinsed three times for 5 min in PBS. The tissues were flooded with fluorescein isothiocyanate–labeled H. pomatia lectin (Invitrogen, L11271) for 2 h at room temperature. Tissues were mounted on glass slides and examined under a Zeiss Axio Imager A2 fluorescence microscope with a ×10 objective.

Colon tissues were prepared for fluorescence in situ hybridization (FISH), analysis as previously described (14). Briefly, after mouse intestines were fixed in 4% PFA, and after deparaffinization and rehydration in hybridization buffer (0.9 M NaCl, 0.1% SDS, and 20 mM Tris–HCl, pH 7.4), the tissues were incubated 2 hours at 50°C in the dark with Alexa 532-conjugated Eubacteria EUB338 (5′-GCTGCCTCCCGTAGGAGT-3′) or non-EUB338 probe for bacterial 16S rRNA genes (15). The sections were then washed three times with a hybridization solution for 15 min, counterstained with DAPI, and mounted. The sections were imaged with an inverted microscope (Leica SP8) using a Leica ×10 objective. Distances between intestinal epithelial cells and luminal bacteria were quantified by using the ImageJ tool.

Mouse distal ileum or colon samples were fixed in 4% PFA in PBS for 24 h, dehydrated, and embedded in paraffin. Tissue slices (5 μm in thickness) were mounted on positively charged glass and dewaxed.

For periodic acid–Schiff (PAS) staining, sections (5 μm) were cut and stained with PAS. Briefly, the slides were incubated for 5 min in 0.5% periodic acid solution, washed with ddH₂O, and then incubated for 20 min in Schiff reagent, washed 5 min in tap water. The slides were then counterstained with hematoxylin for 1 min, after which the slides were dehydrated, mounted, and observed with a light microscope (Zeiss Axio Imager A2).

Sections of mouse intestines from euthanized mice were washed with ice-cold PBS and cut into pieces about 3 mm in length. Selected tissues were fixed in 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M sodium phosphate buffer (PB buffer, pH 7.4) at 4°C overnight. Then, samples were rinsed 5–7 times in PB buffer, postfixed in 1% osmic acid for 2 h, and washed 5–7 times in PB buffer. Dehydration was done with an acetone gradient followed by infiltration with Epon 812 resin and baked overnight at 60°C. Hardened blocks were cut using a Leica Ultracut Ultramicrotome (Leica EM UC6) and stained using 2% uranyl acetate and lead citrate. Images were obtained using a transmission electron microscope (JEM-1400) at 80 kV. Goblet cells were measured from five animals for each genotype.

Epithelial cells were obtained from freshly isolated colons and flushed from both ends with sterile PBS. Colons were then opened longitudinally and washed with PBS. Mucus was removed by gently scraping the villus with a cover glass slide. Intestines were washed and incubated in 10 ml PBS containing 30 mM EDTA and 1.5 mM DTT on ice for 20 min (16). Intestines were then removed and briefly washed in PBS and incubated in 10 ml PBS containing 30 mM EDTA at 37°C at 220 RPM for 10 min. The cells were subjected to 30-s vigorous shaking, then centrifuged at 1,000g for 5 min at 4°C, washed in PBS, and then resuspended in PBS buffer. Cell pellets constituting isolated intestinal epithelial cell fractions were lysed for Western blot to detect Atg16L1 and LC3.

Colonic epithelium cells were isolated as described above (16). Isolated cells were lysed in RIPA buffer with protease inhibitors and subjected to SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes. Then, PVDF membranes were blocked by exposure to PBST containing 5% non-fat dry milk for 1 h at room temperature. The blocked membranes were probed by incubation with the indicated antibody at 4°C overnight. Proteins were detected with HRP-conjugated secondary antibody and visualized by ECL (GE). The following antibodies were used: anti-Atg16L1 (MBL, M150-3), anti-LC3 (Novus Biologicals, NB100-2331), and anti-actin (Santa Cruz Biotechnology, sc-8432).

RNA was extracted from isolated colonic epitheliums with TRIzol Reagent (Invitrogen) following the manufacturer’s instructions. cDNA was generated using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Real-time PCR was performed using the SYBR Green Master Mix (Applied Biosystems) on an ABI 7500 thermal cycler in duplicate. Gene-expression values were calculated by using 2^-△△Ct normalized to HPRT. Primers for quantitative PCR were used as follows:

Mouse–Muc2-F: ACGATGCCTACACCAAGGTC

Mouse–Muc2-R: TGATCTTCTGCATGTTCCCA

Mouse HPRT-F: GTCCCAGCGTCGTGATTAGC

Mouse HPRT-R: TGGCCTCCCATCTCCTTCA

Stool samples for 16S rRNA amplicon surveys were collected into a 2-ml sterile tube and stored at -80°C prior to DNA extraction. Mucus scraping samples were collected as previously described (17). Briefly, following sacrifice of mice, the distal ileum and whole colon were excised and filleted open. Large fecal matter was gently removed if present, and a brushing of the colonic mucosal surface was performed to test for the mucosal-associated microbiome. Then, the collected fecal pellets and mucus scrapings were processed for DNA isolation according to the recommended manual (Qiagen). PCR amplification, library preparation, library-quality inspection, and quantification were performed, and the set TAG sequence was used to distinguish the samples. 16S rRNA sequencing of the V4 region was performed on an Illumina HiSeq 2500 high-throughput sequencing platform to sequence the qualified libraries.

Data are expressed as mean ± SEM. Two groups were compared by using a non-parametric test analysis of variance followed by a Mann–Whitney test analysis. Multiple-comparison analyses were performed using one/two-way ANOVA followed by Tukey or Bonferroni post hoc tests using GraphPad Prism 7 (GraphPad Software Inc.). p < 0.05 was considered as indicating a statistically significant difference.

Previous studies have shown that IBD development is associated with dysbiosis (2). Atg16L1 is an autophagy gene involved in the handling of intracellular bacteria. Cells or mice bearing Atg16L1 T300A polymorphism have a reduced capacity of bacterial clearance (10, 18). Although it has been reported that the Atg16L1 T300A mutation is associated with alterations in gut microbiota, there is a lack of further detailed investigation, and the factors that contributed to altered gut microbiota have never been studied. Thus, to determine the effect of Atg16L1 T300A on commensal bacteria at a steady state, we performed experiments by using cohoused WT and Atg16L1^T300A/T300A mice to reduce the influence of microbiome heritability as a confounding factor. Because fecal samples are widely used in mouse studies to survey the microbiome and its relationship to IBD, we first performed 16S rRNA gene sequencing on feces of WT and Atg16L1^T300A/T300A mice to determine the composition and diversity of the fecal associated microbiota. We found that the taxonomy-based analysis did not have broad alterations in the phylum and family composition of commensal bacteria between Atg16L1^T300A/T300A and WT mice ( Figures 1A, B ), and alpha-diversity analyses based on the Shannon index and Bray–Curtis distance also showed no significant differences between Atg16L1^T300A/T300A mice and WT controls ( Figure 1C ). Next, we performed unweighted UniFrac PCoA analysis to investigate the differences in species complexity and structural alterations in gut microbial communities (β-diversity), and we found a minor difference between the WT and Atg16L1^T300A/T300A microbiotas ( Figure 1D ). Moreover, we applied the linear discriminant analysis (LDA) effect size (LEfSe) method for high-dimensional biomarker discovery to interpret the microbiota between Atg16L1^T300A/T300A and WT commensal composition. We found nine differentially abundant taxonomic clades with an LDA score higher than 2.0, indicating that these bacteria significantly differ in WT and Atg16L1^T300A/T300A mouse feces. Among the nine differentially abundant taxonomic clades, the opportunistic pathogens Ruminococcaceae and Cyanobacteria were enriched in Atg16L1^T300A/T300A mice ( Figure 1E ) and have been reported to be increased in fecal samples from the mouse model of ulcerative colitis (19, 20) and in human patients of Crohn’s disease (21). Thus, this finding indicated that the Atg16L1 T300A variant slightly influenced the microbial composition in fecal samples at a steady state.

While fecal samples are the predominant material used for microbial community analysis, it may not be the ideal sample for analysis of the gut microbiome. Because IBD is characterized by a compromised mucosal barrier and inappropriate immune activation by commensals mislocalized to the mucosa (22), and bacteria reside in the colonic mucus gel layer, and the colonic mucus microbiome was more closely correlated with colitis severity than that of the alterations in the fecal microbiomes (23). Thus, investigating the microbial communities of the mucus layer may provide more important information that is key to understanding the mechanisms by which the Atg16L1 T300A variant affected CD development. In the next study, we investigated the interrelationship between the Atg16L1 T300A variant and the composition of microbiota in the ileum and colon mucosa of Atg16L1^T300A/T300A and WT mice. For this end, colons and terminal ileum were isolated from WT and Atg16L1^T300A/T300A mice, and the content was removed. Then, the mucosa scraping sample was collected and used to perform 16S rRNA gene amplicon surveys to detect associated mucosal bacteria. The UniFrac PCoA analysis of microbial communities in the mucosa indicated that the overall microbiome diversity is comparable between WT and Atg16L1^T300A/T300A mice ( Figure S1A ). Moreover, the relative taxa abundance is different in colon mucosa between the WT and Atg16L1^T300A/T300A mice. The most abundant taxa at the phylum and family levels in the colon are shown in ( Figures 2A, B ). At the phylum level, the abundance of Firmicutes, Campilobacterota, and Deferribaterota was amplified, while Bacteroidetes was declined in Atg16L1^T300A/T300A mice compared to that of the WT group. This observation was consistent with the reports that decreased abundance of Bacteroidetes can contribute to disease pathogenesis in IBD (24, 25). In addition, we also found that the levels of Verrucomicrobiota were reduced in Atg16L1^T300A/T300A mouse colonic mucosa, and a similar observation has been reported in IBD patients (26, 27).

At the family level, the relative abundance of Helicobacteraceae, Ruminococcaceae, and Deferribacteraceae in the Atg16L1^T300A/T300A mice was increased compared with that of the WT mice ( Figure 2B ). Helicobacteraceae played a pathogenic role in the development of CD in a considerable proportion of children (28, 29), and Ruminococcaceae and Deferribacteraceae were increased in DSS-treated colitis mice (30). Interestingly, the relative abundances of Akkermansiaceae in Atg16L1^T300A/T300A mice were decreased dramatically compared with those of WT mice. Akkermansia muciniphila (A. muciniphila), a mucin-degrading bacterium, has been proved associated with healthy mucosa and plays a protective role in regulating intestinal epithelium and mucosal function (31). A. muciniphila reduction has been shown in ulcerative colitis (UC) and Crohn’s disease (CD), both in clinically active diseases and during the remission phase. Moreover, it has been reported that A. muciniphila can act as a biomarker to predict CD (31, 32). Moreover, a study in 2011 demonstrated an inverse correlation between A. muciniphila levels and the severity of acute appendicitis (33).

Venn diagrams are extensively used for the visualization of relationships between different group datasets to illustrate their similarities and differences. At the genus level, Venn diagram analysis indicated that the T300A variant led to reduced microbiota diversity compared with WT control mice ( Figure 2C ) (86 and 16), suggesting that Atg16L1 T300A polymorphism reduced the mucosa-associated bacterial composition. In addition, Wilcoxon rank-sum test analysis based on genus level found that several taxonomic clades significantly differed between Atg16L1^T300A/T300A mice and those of cohoused WT controls. For instance, the proportions and relative abundance of Akkermansia, Blautia, and norank_o_Clostridia_vadinBB60_group diminished significantly in the colon of Atg16L1^T300A/T300A mice ( Figure 2D and Figure S1B ), while the opportunistic pathogens such as norank_f_Lachnospiraceae, norank_f_Ruminococcaceae, Tyzzerella, Mucispirillum, and Lachnospiraceae_FCS020_group significantly increased in Atg16L1^T300A/T300A mice ( Figure 2E and Figure S1B ), and the changes of the above bacteria were associated with IBD development (34–37).

To gain more insight into the changes of microbiota resident in colon mucus in Atg16L1^T300A/T300A mice, we next analyzed the microbiota abundance from the genus to species levels by using the LEfSe method as a visualization tool. We found more than twenty differentially abundant taxonomic clades with an LDA score higher than 2.0 in Atg16L1^T300A/T300A mice. Many of them were consistent with Wilcoxon rank-sum test analysis ( Figure 2F ) and had been reported associated with IBD patients (26, 31). These results indicated altered IBD-associated microbiota resident in Atg16L1^T300A/T300A mice even under a steady state.

We next analyzed the microbiota composition in ileum mucosa. We found that the T300A risk allele altered the composition of mucosa-associated bacteria in Atg16L1^T300A/T300A mice and was similar to what was observed in the colon mucus. A Venn diagram analysis indicated that Atg16L1T300A led to a reduction in the abundance of genus compared with WT mice (95 and 36) ( Figure 2G ), suggesting that Atg16L1 T300A polymorphism reduced the mucosa-associated bacterial diversity in the ileum as well. Wilcoxon rank-sum test analysis indicated that the proportions and relative abundance of probiotic bacteria Akkermansia decreased dramatically in the ileum, and the relative abundance of opportunistic pathogens Tyzzerella and norank_f_Ruminococcaceae, profoundly overrepresented bacteria in IBD, were also increased in Atg16L1^T300A/T300A mouse ileum mucosa ( Figure 2H and Figure S1C ). LEfSe analysis from genus to species further confirmed the finding that the T300A risk allele significantly altered the abundance of bacteria residing in the mucus gel layer of the ileum, and approximately twenty differentially abundant taxonomic clades with LDA score higher than 2.0 in Atg16L1^T300A/T300A mouse ileum mucosa were found ( Figure 2I ).

Collectively, these findings indicated that the mucus-associated gut microbiome composition in the ileum and colon of Atg16L1^T300A/T300A mice displayed an IBD-associated microbe species even under the steady state in our facility.

The data above showed that Akkermansia was decreased dramatically in the mucus of Atg16L1^T300A/T300A mice in both colon and ileum mucosa. Akkermansia, as a gram-negative and strictly anaerobic bacterium, tended to widely colonize the nutrient-rich intestinal mucus gel layer (MGL) and was particularly effective in increasing mucus thickness and gut barrier function to maintain the homeostasis of the intestinal microbiota (38). Colonization with Akkermansia has been reported to promote mucosal wound healing (39, 40). Its metabolic and mucolytic activity makes Akkermansia a key species in the mucus layer, stimulating beneficial mucosal microbial networks. Because the MGL is composed mainly of mucin (Muc2) secreted by goblet cells, we hypothesized that the decreased Akkermansia composition in the Atg16L1^T300A/T300A ileum and colon mucosal layer may be correlated with the defect of the mucin secretion (41–43). To test this possibility, we first investigate the morphology of the goblet cell in the colon. Colon sections from Atg16L1^T300A/T300A and cohoused WT mice were performed with PAS staining. PAS staining labels highly glycosylated proteins, most notably in goblet cell mucins. We found that Atg16L1^T300A/T300A mice had larger areas of mucin ( Figures 3A, B ) and enlarged goblet cells within the surface epithelial cuffs than WT controls ( Figure 3C ). However, the total number of goblet cells per crypt did not change compared to WT controls ( Figure 3D ), suggesting that the Atg16L1 T300A variant did not regulate the differentiation of goblet cells in the gut. These results are similar to what was observed in mice with a complete absence of autophagy protein Atg5 (44).

Similarly, although the total number of goblet cells per crypt was comparable in the ileum between WT and Atg16L1^T300A/T300A mice ( Figures 3E, H ), the ileum goblet cells of Atg16L1^T300A/T300A mice also contained larger areas of cytoplasmic mucin and an increased number of enlarged goblet cells ( Figures 3F, G ), which is in contrast to the previous report that the enlarged goblet cell number in the small intestinal was unchanged (45). The discrepancy between the two studies was probably due to the difference in the microbiota of the two-mouse facility.

Goblet cells are a secretory lineage that specializes in mucus production. Goblet cells secrete mucus, which forms a layer that overlays the epithelium and forms an important barrier to intestinal microbes (46). Although it has been reported that Atg16L1^T300A/T300A results in enlarged goblet cells within the surface epithelial cuffs (45), the consequence of enlarged goblet cells in Atg16L1^T300A/T300A mice has never been studied. Intestinal mucus is composed of multiple highly glycosylated proteins called mucins. Mucins are stored in secretory granules; constitutive secretion occurs by fusion of individual mucin granules with the apical plasma membrane (47). Mucin granules can accumulate within the goblet cell cytoplasm leading to increased size, which can be secondary to either increased mucin production or diminished secretion. To examine whether the enlarged goblet cell in Atg16L1^T300A/T300A mice was due to impaired secretion, we first used transmission electron microscopy to visualize the theca of goblet cells, which is normally packed with mucin granules. We found that in WT mice, once the theca-containing mucin granules reach the apical surface of the intestinal epithelium, they fuse with the epithelium, releasing the stored mucins and associated proteins into the intestinal lumen. In contrast, the goblet cell from Atg16L1^T300A/T300A mice was featured with an increased accumulation of intracellular mucin granules and an apparent inability of these granules to fuse with the apical surface of the intestinal epithelium ( Figure 4A ), which resulted in a deficiency of mucin release. To further confirm the defect of mucin secretion in Atg16L1^T300A/T300A mice, we detected mucin in the lumen of the intestine by whole-mount staining. To this end, colons obtained from cohoused WT and Atg16L1^T300A/T300A mice were collected and opened longitudinally after the adipose tissues were stripped off. 2–3-cm equal intestine sections were fixed and stained with fluorescein isothiocyanate–labeled H. pomatia lectin at room temperature, after which the images were taken immediately above the colon mucosal surface. We found that mucus secreted into the colon lumen was reduced in Atg16L1^T300A/T300A mice compared with that of cohoused WT controls ( Figure 4B ). These results indicate that Atg16L1^T300A/T300A goblet cells had a defect in granule exocytosis and a lack of intact mucus layer.

The diminished secretion of mucin in the lumen of the Atg16L1^T300A/T300A mice could be due to impaired mucin gene expression. Muc2 is the predominant mucin secreted by goblet cells, so we analyzed Muc2 expression in the colon epithelium from cohoused WT and Atg16L1^T300A/T300A mice by real-time PCR. We found no difference in the expression of Muc2 in cohoused WT and Atg16L1^T300A/T300A mice ( Figure 4C ), supporting the conclusion that the reduced mucin observed in the colon was not due to the gene expression but the deficiency of secretion.

The MGL lining the colon is integral to preventing bacterial invasion and maintaining intestinal homeostasis in health and disease. MGL defects allow bacteria to contact the colonic surface directly and are commonly observed in IBD (48). Having observed defected mucin secretion of goblet cells in Atg16L1^T300A/T300A mice, we next determined whether the reduced secretion of mucin could impair the homeostasis of mucus lining. We examined the spatial distance between mucosal surfaces and the luminal contents by fluorescence in situ hybridization (FISH) assay with a universal 16S ribosomal RNA (rRNA) gene probe. We found that the spatial distance of mucosal surfaces and the luminal contents were approximately ~45 μm (from the villus tip) in WT mice ( Figures 4D, I ). In contrast, a marked decrease in spatial distance between mucosal surfaces and the luminal contents was observed in Atg16L1^T300A/T300A mice (~25 μm). We also examined mucin secretion in the ileum as above. The results showed decreased mucus secretion in the ileum lumen of Atg16L1^T300A/T300A mice compared with WT controls ( Figure 4E ).

Collectively, these findings indicate that Atg16L1T300A polymorphism regulated the secretion capacity of intestinal goblet cells, which may contribute to microbiota dysbiosis in Atg16L1^T300A/T300A mice.

To further investigate the mechanism by which Atg16L1T300A polymorphism regulates goblet cell secretion, we next investigated whether the impaired secretion function of Atg16L1 T300A goblet cells is due to the defective autophagic process which is required for proper secretion of mucus granules. We crossed Atg16L1^f/f mice with ERT2Cre mice to generate Atg16L1^f/f&ERT2Cre⁺ mice. Tamoxifen was orally administrated to Atg16L1^f/f&ERT2Cre⁺ mice to generate Atg16L1-KO mice. Colon epithelial cells collected from WT, Atg16L1^T300A/T300A, and Atg16L1-KO mice were lysed and immunoblotted with the anti-Atg16L1 antibody; we found that Atg16L1 expression was comparable between WT and Atg16L1^T300A/T300A mice, but almost completely lost in Atg16L1-KO mice ( Figure 4F ). These results indicated that Atg16L1 T300A mutation did not influence the Atg16L1 expression and tamoxifen administration activates ERT2Cre and deleted Atg16L1 expression. We next studied autophagy induction in the intestine epithelial cell (mainly containing goblet cell). The colon epithelial cell was collected from WT, Atg16L1^T300A/T300A, and Atg16L1 KO mice as above, and autophagy induction was detected by LC3 conversion. We observed that LC3II conversion was impaired significantly in intestinal epithelial cells of Atg16L1^T300A/T300A mice compared with those of cohoused WT mice, indicating that autophagy was impaired in Atg16L1^T300A/T300A intestinal epithelial cells; LC3II was almost completely lost in epithelial cells of Atg16L1-KO mice ( Figure 4F ).

To detect the effect of autophagy on mucus secretion directly, we performed whole-mount staining in both Atg16L1^T300A/T300A and Atg16L1-KO mice as above, and the results showed that Atg16L1 KO mice had more drastic effects on mucin secretion; mucus secreted was almost completely lost in Atg16L1-KO mice compared with that of WT controls ( Figure 4G ). These findings suggested that impaired mucin secretion in Atg16L1^T300A/T300A mice is autophagy-dependent, which is consistent with the previous report that autophagy was required for mucin granule exocytosis of goblet cells (44). Finally, we performed FISH staining with the colon of WT, Atg16L1^T300A/T300A, and Atg16L1-KO mice to examine the spatial distance between mucosal surfaces and the luminal contents. The result showed that the distance between the apical surface of the epithelium and the luminal contents in Atg16L1-KO mice reduced dramatically compared with that of Atg16L1^T300A/T300A and WT mice, and commensal bacteria were directly attached to the epithelia of Atg16L1-KO mice ( Figures 4H, I ). These findings indicated that reduced mucin secretion of goblet cells in Atg16L1 T300A mice is autophagy-dependent.

Dextran sodium sulfate (DSS)-induced colitis is a model commonly used to study host–microbial interactions characterized by a defect in mucus gel layer and microbiota alterations (17). In addition, the administration of DSS disrupts the intestinal barrier and effectively mimics the clinical and histological features of IBD patients (49). Therefore, we next investigated whether Atg16L1^T300A/T300A mice will be more sensitive to DSS-induced colitis. Accordingly, WT and Atg16L1^T300A/T300A mice were administrated with 1.5% DSS in drinking water for 7 days followed by 3 days of regular drinking water, and then the extent of colitis was evaluated. We observed that the weight loss in Atg16L1^T300A/T300A mice was comparable with WT mice ( Figure 5A ). However, the colon length of Atg16L1^T300A/T300A mice was shorter than that of WT mice ( Figures 5B, C ). In addition, the colon section was obtained and H&E staining was performed. We found that Atg16L1^T300A/T300A mice exhibited more intense inflammation in the colon section compared than did WT mice, as indicated by blinded assessment of histopathologic scores ( Figures 5D, E ). Overall, our data indicated that the presence of Atg16L1 T300A resulted in defects in mucin granule exocytosis and impaired mucus layer formation that altered the microbial community in Atg16L1^T300A/T300A mice. Because there are many other defects in Atg16L1^T300A/T300A mice, the dysbiosis observed in Atg16L1^T300A/T300A mice is probably one of the important factors that contributed to the Atg16L1^T300A/T300A variant’s susceptibility to Crohn’s disease.