Intestinal Epithelial Axin1 Deficiency Protects Against Colitis via Altered Gut Microbiota

Shari Garrett; Yongguo Zhang; Yinglin Xia; Jun Sun

doi:10.1016/j.eng.2023.06.007

Engineering ›› 2024, Vol. 35 ›› Issue (4) :252 -268. DOI: 10.1016/j.eng.2023.06.007

Research

Article

Intestinal Epithelial Axin1 Deficiency Protects Against Colitis via Altered Gut Microbiota

Shari Garrett ^a^,^b
, Yongguo Zhang ^a
, Yinglin Xia ^a
, Jun Sun ^a^,^b^,^c^,^d^,^*

Author information +

History +

PDF (6810KB)

Abstract

Intestinal homeostasis is maintained by specialized host cells and the gut microbiota. Wnt/β-catenin signaling is essential for gastrointestinal development and homeostasis, and its dysregulation has been implicated in inflammation and colorectal cancer. Axin1 negatively regulates activated Wnt/β-catenin signaling, but little is known regarding its role in regulating host-microbial interactions in health and disease. Here, we aim to demonstrate that intestinal Axin1 determines gut homeostasis and host response to inflammation. Axin1 expression was analyzed in human inflammatory bowel disease datasets. To explore the effects and mechanism of intestinal Axin1 in regulating intestinal homeostasis and colitis, we generated new mouse models with Axin1 conditional knockout in intestinal epithelial cell (IEC; Axin1^ΔIEC) and Paneth cell (PC; Axin1^ΔPC) to compare with control (Axin1LoxP; LoxP: locus of X-over, P1) mice. We found increased Axin1 expression in the colonic epithelium of human inflammatory bowel disease (IBD). Axin1^ΔIEC mice exhibited altered goblet cell spatial distribution, PC morphology, reduced lysozyme expression, and enriched Akkermansia muciniphila (A. muciniphila). The absence of intestinal epithelial and PC Axin1 decreased susceptibility to dextran sulfate sodium (DSS)-induced colitis in vivo. Axin1^ΔIEC and Axin1^ΔPC mice became more susceptible to DSS-colitis after cohousing with control mice. Treatment with A. muciniphila reduced DSS-colitis severity. Antibiotic treatment did not change the IEC proliferation in the Axin1^Loxp mice. However, the intestinal proliferative cells in Axin1^ΔIEC mice with antibiotic treatment were reduced compared with those in Axin1^ΔIEC mice without treatment. These data suggest non-colitogenic effects driven by the gut microbiome. In conclusion, we found that the loss of intestinal Axin1 protects against colitis, likely driven by epithelial Axin1 and Axin1-associated A. muciniphila. Our study demonstrates a novel role of Axin1 in mediating intestinal homeostasis and the microbiota. Further mechanistic studies using specific Axin1 mutations elucidating how Axin1 modulates the microbiome and host inflammatory response will provide new therapeutic strategies for human IBD.

Graphical abstract

Keywords

Axin1 / Bacteria / Microbiome inflammation / Inflammatory bowel disease / Immunity / Microbiome / Paneth cells / Akkermansia muciniphila / Wnt

Cite this article

Download citation ▾

Shari Garrett, Yongguo Zhang, Yinglin Xia, Jun Sun. Intestinal Epithelial Axin1 Deficiency Protects Against Colitis via Altered Gut Microbiota. Engineering, 2024, 35(4): 252-268 DOI:10.1016/j.eng.2023.06.007

登录浏览全文

4963

注册一个新账户忘记密码

1. Introduction

Inflammatory bowel disease (IBD) is a chronic inflammation of the intestines affecting more than six million people globally [1]. The pathogenesis of IBD remains unclear but has been observed to involve complex interactions among environmental factors, gut microbiota, and host epithelial cells, as well as dysregulation of innate and adaptive immunity against a backdrop of genetic pre-disposition [2]. A major pathway implicated in IBD is the dysregulation of intestinal Wnt/β-catenin signaling [3], [4], [5], [6]. Wnt/β-catenin signaling is an essential regulator of gastrointestinal (GI) development and homeostasis [7]. Axin1 was initially identified as an inhibitor of Wnt/β-catenin signaling and a regulator of embryonic axis formation [8]. Axin1 is a scaffold protein that recruits the β-catenin destruction complex, facilitating glycogen synthase kinase 3β (GSK3β)’s phosphorylation of β-catenin, which targets β-catenin for ubiquitination and degradation by the proteasome [9], [10], [11]. However, the role of Axin1 in regulating intestinal inflammation and the development of IBD is unknown.

Axin1 plays a distinct biological role in bacterial infections—specifically host-pathogen interactions [12]. Salmonella decreases Axin1 protein expression in intestinal epithelial cell (IEC) lines at the post-transcriptional level, whereas its overexpression inhibits Salmonella invasion and inflammation in vitro [12]. However, the role of intestinal Axin1 signaling in maintaining mucosal health has not been fully understood in vivo. Because whole-body Axin1 deletion is lethal, few investigators are utilizing a conditional Axin1-knockout mouse model for mechanistic studies [8].

In the current study, we hypothesize that intestinal epithelial Axin1 plays a role in regulating the microbiota and susceptibility to inflammation. In human IBD samples, we found increased expression of Axin1 at the messenger RNA (mRNA) and protein levels. To investigate the molecular mechanism of intestinal epithelial Axin1 regulation, we generated a novel mouse model of IEC conditional knockout of Axin1 (Axin1^ΔIEC) [13]. Altered expression of intestinal Axin1 impairs intestinal epithelial secretory cells and cell differentiation. Furthermore, we generated a mouse model of Axin1^ΔPC to study the tissue-specific role of Paneth cell (PC) Axin1 in response to inflammation. Our studies, for the first time, demonstrate the relationship between intestinal Axin1 and the maintenance of intestinal and microbial homeostasis. Understanding complex interactions among host factors (e.g., Axin1), cellular changes (e.g., PCs), and the microbiota (e.g., Akkermansia muciniphila (A. muciniphila)) in colitis will help to provide novel therapeutic approaches for human IBD.

2. Materials and methods

2.1. Human intestinal biopsies

Slides containing paraffin-embedded colon biopsy samples of patients with ulcerative colitis (UC), Crohn’s disease (CD), and healthy controls were obtained from a tissue microarray from US Biomax, Inc. (CO246; USA).

2.2. Gene expression datasets

We used microarray data registered on the Gene Expression Omnibus (GEO) repository^†. From the GEO repository, we obtained colonic mucosal biopsies from the inflamed mucosa of IBD patients with UC, CD, and healthy controls. Total RNA was isolated via microarray and reported (GEO accession number GSE 16879). Gene expression data for Axin1 from UC control (n = 6), UC (n = 24), CD control (n = 6), and CD (n = 18) patients were extracted and analyzed for our study.

2.3. Experimental animals

Axin1^LoxP (LoxP: locus of X-over, P1) control mice were originally reported by Xie et al. [8]. Axin1^ΔIEC mice were obtained by crossing Axin1^LoxP with villin-cre mice (Jackson Laboratory, USA). Defensin alpha 6 (Defa6)-cre mice were obtained from Dr. Richard Blumberg, Harvard University [14]. Axin1^ΔPC mice were obtained by crossing Axin1^LoxP mice with Defa6-cre mice. Experiments were performed on 6- to 8-week-old male and female littermate mice that were provided with water ad libitum and maintained in a 12 h dark/light cycle. All animals were housed in the Biologic Resources Laboratory at the University of Illinois Chicago (UIC) and utilized in accordance with UIC Animal Care Committee (ACC) and Office of Animal Care and Institutional Biosafety guidelines. Animal work was approved by the UIC Office of Animal Care (ACC15-231, ACC17-218, and ACC18-216).

2.4. Genetic background of mouse strains

Axin1 targeting vector was electroporated into SV129 embryonic stem (ES) cells. These ES cells were selected and assessed for Axin1 disruption at exon 2. The proper ES clones were injected into C57BL/6J blastocysts to produce chimeric animals. Chimeric animals were then further crossed to produce Axin1^LoxP mice that were viable and fertile with no recognizable phenotype. villin-cre (stock No. 004586) transgenic mice had been generated on a C57BL/6 background. The PC-specific Defa6-cre mouse strain [14] was generated by injecting the designed plasmid into the pronucleus of C57BL/6 mice, allowing for cre-recombinase to be expressed under PC-specific promotor, Defa6.

2.5. A. muciniphila strain and growth conditions

A. muciniphila (ATCC BAA-835) was propagated in brain heart infusion broth (BD Diagnostics, USA) supplemented with 3% L-cysteine (Sigma-Aldrich, USA) in an Oxoid AnaeroJar (AG0025A; Thermo Fisher Scientific, USA) with an AnaeroGen Pack (AN0025A; Thermo Fisher Scientific) at 37 °C for 48 h. Cultures were centrifuged at 12 000g for 10 min and resuspended in sterile phosphate-buffered saline (PBS) to 10⁹ colony-forming units (CFU) per 1.5 mL. Culture was placed immediately on ice before oral gavage.

2.6. Colitis induction

Colitis was induced as previously described [15], [16]. Mice were administered 5% dextran sulfate sodium (DSS) (molecular weight (MW): 40-50 kDa; USB Corp., USA) dissolved in filter-purified water ab libitum during the experimental period. Animals were weighed daily. At day 7, mice were sacrificed under anesthesia, and the severity of colitis was quantified by a disease activity index (DAI), determined by percent of weight loss, fecal blood, and diarrhea.

2.7. Co-housing experiment

Male and female Axin1^LoxP and Axin1^ΔIEC or Axin1^ΔPC mice (6-8 weeks old) were co-housed in new cages. Each cage contained three Axin1^LoxP mice and two Axin1^ΔIEC or two Axin1^ΔPC mice, as previously described [17]. After four weeks of co-housing, 5% DSS dissolved in filter-purified drinking water was given ab libitum. Animals were weighed daily. At day 7 after DSS administration, mice were sacrificed under anesthesia, and the severity of colitis was quantified by a DAI, determined by percent of weight loss, fecal blood, and diarrhea.

2.8. DSS-colitis model and A. muciniphila treatment

The animal experiment was performed using 7- to 13-week-old mice. Mice were treated with A. muciniphila (100 μL suspension in PBS) 6 h before the start of 5% DSS treatment. 5% DSS was given ab libitum in water for seven days. The mice were gavaged daily with 1.79 × 10⁹ CFU of A. muciniphila for seven days. After seven days, the mice were sacrificed, and tissue samples were harvested.

2.9. Antibiotic treatment in mice

Axin1^LoxP and Axin1^ΔIEC mice (6-8 weeks old) were randomly divided into two groups: six mice (male: three; female: three) in the non-treatment group, and six mice (male: three; female: three) in the antibiotic-treated group. Antibiotics (1 mg·mL^-1 metronidazole and 0.3 mg·mL^-1 clindamycin) were administered in filtered drinking water. The control group received filtered drinking water without antibiotics. At week 3 after the antibiotic treatment, tissue samples were collected for the indicated studies.

2.10. Histology of mouse colon and small intestine

Intestines were harvested as previously described [15], [16], [18], [19] and were fixed in 10% formalin (pH 7.4), processed, and paraffin embedded. Sections of 4 µm were stained with hematoxylin and eosin (H&E) [20]. Histological damage was scored as described previously [13].

2.11. 5-B romodeoxyuridinc (BrdU) migration

Age-matched 6- to 9-week-old male and female Axin1^LoxP, Axin1^ΔIEC, and Axin1^ΔPC mice were given an intraperitoneal injection of BrdU (160 mg·kg^-1, diluted in PBS; Sigma-Aldrich) and were sacrificed after 2, 12, and 24 h. Migration distance was measured as the distance in micrometers from the base of the crypt to the foremost BrdU-positive enterocyte in jejunal sections.

2.12. Immunoblotting

Mouse ileal and colonic epithelial cells were collected by scraping the tissue and were homogenized as previously described [17], [21]. Equal amounts of normalized protein were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and immunoblotted with primary antibodies, as previously described [12], [22]. Antibodies were visualized by enhanced chemiluminescence. Membranes probed with more than one antibody were stripped before re-probing.

2.13. Immunohistochemistry (IHC)

Intestinal tissues were fixed in 10% buffered formalin and processed using standard techniques, as previously described [15], [16], [22], [23]. Slides were stained with anti-Axin1 (34-5900; Invitrogen, USA), BrdU (1893; Abcam, USA), β-catenin (610154; BD Transduction, USA), and phosphor-β-catenin-S552 (9566; Cell Signaling, USA), and staining intensity was performed as previously described [15], [24].

2.14. Immunofluorescence (IF)

Intestinal tissues were freshly isolated and embedded in paraffin wax after fixation with 10% neutral buffered formalin. IF staining was performed on paraffin-embedded sections (4 μm) of mouse intestine. After the preparation of slides, as described previously [15], sections were incubated with anti-lysozyme (Santa Cruz Biotechnology Inc., USA) antibody overnight at 4 °C. Samples were then incubated with donkey-anti goat Alexa Flour 488 (D1306; Thermo Fisher Scientific) for 1 h at room temperature. Tissues were mounted with SlowFade (S2828; Thermo Fisher Scientific), cover slipped, and sealed. Sections were examined with a Leica SP5 laser scanning confocal microscope (LSM 710; Carl Zeiss Inc., Germany) or an Olympus BX51 fluorescence microscope (Olympus Life-Sciences, USA).

2.15. Fluorescence in situ hybridization (FISH) for intestinal bacteria

FISH was performed using anti-sense single-stranded DNA (ssDNA) probe EUB 388 (5′-GCTGCCTCCCGTAGGAGT-3′) targeting a highly conserved region of the bacterial 16s gene. Tissue sections (4 μm) were baked for 30 min at 60 °C. Sections were deparaffinized in xylene, dehydrated with 100% ethanol, dried, and incubated in 0.2 mol·L^-1 HCl for 20 min, and then heated in 1.0 mol·L^-1 NaSCN for 10 min at 80 °C. Sections were digested with pepsin (4% pepsin in 0.01 mol·L^-1 HCl) for 20 min at 37 °C. Slides were washed in wash buffer (0.30 mol·L^-1 NaCl, 0.03 mol·L^-1 sodium citrate, pH 7.0). Sections were fixed for 15 min in 10% buffered formalin. Probes were hybridized at 5 ng·µL^-1 for 5 min at 90 °C in hybridization buffer (0.9 mol·L^-1 NaCl, 30% formamide, 20.0 mmol·L^-1 Tris-HCl, pH 7.4) and 0.01% SDS and incubated overnight at 37 °C. Slides were washed five times for 5 min at 45 °C in wash buffer. To visualize cell nuclei, sections were stained with 4',6-diamidino-2-phenylindole (DAPI)/antifade solution. Slides were examined with an Olympus BX51 fluorescence microscope.

2.16. Transmission electron microscopy (TEM)

Small intestines were fixed in 4% paraformaldehyde/3% glutaraldehyde in 10 mmol·L^-1 sodium phosphate buffer (pH 7.4) for 48 h. All samples were prepared as previously described [15]. After the resin was polymerized, samples were sliced into 1 mm × 2 mm pieces and examined with a Philips CM 100 electron microscope (the Netherlands) at an accelerating voltage of 80 kV for imaging at the UIC electron microscopy core.

2.17. Multiplex ELISA

A mouse-specific ProcartalPlex mouse cytokine/chemokine convenience 26-plex panel 1 multiplex immunoassay plate (EPXR260-26088-901; Invitrogen) was used to detect serum cytokine levels. The assay was performed according to the manufacturer’s instruction manual using proper standards. Plates were read using a Magpix machine (Luminex, USA).

2.18. Alcian blue staining for goblet cells (GCs)

Alcian blue/periodic acid-Schiff staining was performed on paraffin sections of colon tissue fixed with Carnoy’s fixative [25]. Slides were used for counting GCs. Acidic (blue), neutral (pink), and mixed (purple) GCs were counted in each crypt. GC number was calculated by counting the average number of GCs in three crypts as one point. Three points were randomly selected for each mouse. Mean values for the number of GCs were calculated and analyzed using Welch’s t test.

2.19. Lysozyme IF of PCs

PC staining was performed by means of lysozyme IF. The morphological changes were defined as D0 = normal PCs and D1-D3 = abnormal PCs, as described in previous publications [15], [17], [26].

2.20. Isolation of PCs

Small intestines were harvested. The intestines were washed with PBS and cut lengthwise, then placed in ice-cold PBS for 5 min in a cold room. The PBS was removed and replaced with 2 mmol·L^-1 ethylenediaminetetraacetic acid (EDTA) in PBS, and the samples were rocked in a cold room for 30 min. The 2 mmol·L^-1 EDTA was removed and replaced with 54.9 mmol·L^-1 D-sorbitol and 43.4 mmol·L^-1 sucrose in PBS. The samples were vigorously shaken by hand to dissociate crypts and then filtered through a 100 μm cell strainer. The samples were centrifuged at 150g for 10 min at 4 °C. The supernatant was removed, and the crypts were resuspended in TrypLE Express enzyme at 37 °C for 15 min, with gentle shaking every 5 min. The single-cell suspension was filtered through a 70 μm strainer and centrifuged at 200g for 10 min at 4 °C. The pellet was washed twice with 2 mmol·L^-1 EDTA and 1% fetal bovine serum (FBS) in PBS. After centrifugation, cells were resuspended in 2 mmol·L^-1 EDTA and 1% FBS in PBS and incubated with cluster of differentiation 24 (CD24)-PE antibody in the dark at 4 °C for 30 min on a shaker. Cells were centrifuged and washed two times in flow wash buffer and then resuspended in flow wash buffer for sorting [15].

2.21. Bioinformatic analysis of 16s ribosomal RNA (rRNA) sequencing of fecal data

Fecal samples were harvested and prepped as previously described [17]. Samples were analyzed at the UIC Genome Research Core utilizing the Illumina sequencing platform. The QIIME pipeline [27] was used to process raw sequence data—including read merging, adapter and quality trimming, and chimeric checking—and to generate operational taxonomic units (OTUs) in a closed-reference manner using the UCLUST method, with a threshold of 97% sequence similarity. Taxonomic annotations were assigned to each out using the representative sequence data against the Illumina Curated GreenGenes reference database [28].

2.22. Real-time quantitative polymerase chain reaction (qRT-PCR)

Total RNA was extracted from mouse colonic and small IECs using TRIzol reagent. RNA reverse transcription was done using an iScript complementary DNA (cDNA) synthesis kit, according to the manufacturer’s directions (Bio-Rad Laboratories, Inc., USA). cDNA reaction products were subjected to qRT-PCR using the iQ SYBR green supermix, according to the manufacturer’s instructions. All expression levels were normalized to villin levels in the same sample. Percent expression was calculated as the ratio of the normalized value of each sample to the corresponding control. All qRT-PCR reactions were performed in duplicate.

2.23. qRT-PCR of bacterial DNA

DNA was extracted from mice feces using a stool DNA kit (Omega Bio-Tek, USA), according to the manufacturer’s instructions. 16s ribosomal DNA (rDNA) qRT-PCR reactions were performed on a CFX Connect real time system (Bio-Rad Laboratories, Inc.) and amplified using iTaq Universal SYBR green supermix (1725121; Bio-Rad Laboratories, Inc.), according to the manufacturer’s directions. Primers specific to 16s rRNA were used as an endogenous control to normalize between samples. The relative amount of 16s rDNA in each sample was estimated using the ΔΔC_T method. Primer sequences were designed using Primer-BLAST or obtained from Tables S1 and S2 in Appendix A.

Materials used in this paper are listed in Table S3 in Appendix A.

2.24. Statistical analyses

A single animal was tested individually, except when stated otherwise. Data is expressed as mean ± standard error of mean (SEM), and P values ≤ 0.05 were considered to be statistically significant. A Shapiro-Wilks normality test was performed to detect whether the data significantly departed from normality. Parametric or non-parametric analyses were determined based on whether the variables were normally distributed or not. An F test of equality of variances was performed to test the null hypothesis that two normal populations of groups have the same variance. Differences between two groups with normal distribution were analyzed via an unpaired Student’s t test for equal variances and Welch’s t test for unequal variances. Differences between two groups with non-normal distribution were analyzed via a Wilcoxon rank sum test. Differences among three or more groups were analyzed via one-way analysis of variance (ANOVA) or two-way ANOVA, as appropriate. To adjust for multiple comparisons, P values were adjusted via Tukey’s method. Spearman correlation analysis and scatter plots were performed to detect the correlation between Axin1 and interleukin 6 (IL-6) and tumor necrosis factor α (TNF-α) cytokines. Taxonomic and OTU abundances were analyzed as performed previously [29]. In brief, principal coordinate analysis (PCA) and Shannon diversity analysis were conducted using the R packages phyloseq and vegan. Statistical analyses were performed using GraphPad Prism version 8.0.0 for Windows (GraphPad Software, USA) and R software (R Core Team (2021); R Foundation for Statistical Computing, Austria).

3. Results

3.1. Increased Axin1 in human IBD

The status of Axin1 in the inflamed intestine is unexplored. We analyzed microarray data of human colonic samples using the GEO database. We found the mRNA expression levels of Axin1 were elevated in human UC and CD [28] (Fig. 1(a)). Spearman correlation analysis indicated a positive correlation between Axin1 expression and pro-inflammatory cytokines IL-6 (Fig. 1(b)) and TNF-α (Fig. 1(c)) in CD. To investigate the changes and localization of Axin1 at the protein level, we performed IHC of colonic tissue from healthy control, UC, and CD subjects. UC (Fig. 1(d)) and CD (Fig. 1(e)) colonic tissue showed a significant increase in Axin1 expression in the inflamed mucosa compared normal colon.

3.2. Establishment of an Axin1^Δ^IEC mouse model

We hypothesized that intestinal epithelial Axin1 plays a role in the pathogenesis of colitis. We generated a novel conditional IEC knockout by crossing an Axin1^LoxP mouse strain with villin-cre mice. Expression of Axin1 mRNA in the mouse colon and small intestine showed significant decrease of Axin1 mRNA in the Axin1^ΔIEC mice (Fig. 2(a)). We analyzed the protein level via Western blot using intestinal mucosal scrapings and found a significant reduction of intestinal epithelial Axin1 in both the small intestine and colon of Axin1^ΔIEC mice (Fig. 2(b)).

We examined the localization of Axin1 expression in the intestine. There was no detectable IEC Axin1 staining in either the small intestine or colon (Figs. 2(c) and (d)) of Axin1^ΔIEC mice. In addition, we indirectly determined mucus layer thickness and commensal microbiota location by FISH. We found an increase in commensal bacterial invasion into the ileum (Fig. 2(e)) and colon (Fig. 2(f)) of Axin1^ΔIEC mice. This data confirms the IEC-specific knockdown of Axin1 and that it contributes to decreased mucus layer thickness. This Axin1^ΔIEC model will allow us to determine the mechanisms by which intestinal epithelial Axin1 regulates intestinal homeostasis and colitis.

3.3. Intestinal Axin1 regulation of GC distribution and PC morphology in the small intestine

We next determined whether Axin1 deficiency impaired cell differentiation in the intestine. Axin1^ΔIEC mice show a significant increase in Alcian blue positives (Fig. 3(a)). In addition, there was increased mRNA expression of GC marker mucin 2 (MUC2) in the small intestine of Axin1^ΔIEC mice (Fig. 3(b)), suggesting that Axin1 contributes to intestinal GC differentiation. We analyzed the structure of GCs via TEM (Fig. 3(c)). Axin1^ΔIEC mice had larger mucin granules than Axin1^LoxP mice, despite having similar numbers of mucin granules per GC (Fig. 3(d)).

PCs are specialized in the small intestine that secrete antimicrobial peptides and are crucial for shaping the microbiota and regulating innate immunity [14], [15], [30]. Considering the physiological role, we next assessed PC status in Axin1^ΔIEC mice. We categorized PCs based on their lysozyme morphology via IF staining described previously by Wu et al. [17] (Fig. 3(e)). Normal PCs were indicated as D0, while abnormal PCs were grouped as D1 (disordered), D2 (depleted), and D3 (diffused) lysozyme granule morphologies. We found fewer normal PCs in the Axin1^ΔIEC ileum compared to Axin1^LoxP. Most importantly, we saw a significant increase in the number of abnormal PCs in the Axin1^ΔIEC mice (Fig. 3(f)). These abnormal PCs were associated with decreased lysozyme (Lyz1) at the mRNA level (Fig. 3(g)). We further assessed PC structural morphology by TEM and saw a significant increase in the number of PCs with fewer electron-dense granules in the Axin1^ΔIEC mice (Fig. 3(h)). Collectively this data was associated with decreased lysozyme expression at the protein level (Fig. 3(i)).

**3.4. Altered intestinal microbiome in Axin1^ΔIEC mice**

PCs and GCs are known to modulate the intestinal microbiota profile and function. We examined the bacterial abundance of Axin1^LoxP and Axin1^ΔIEC mice by 16s rRNA sequencing and found that the fecal amplicon profile showed unchanged α- and β-diversity between Axin1^LoxP and Axin1^ΔIEC microbiota (Figs. 3(j) and (k)). However, the top genera abundances showed an enrichment in Akkermansia and a depletion in Rikenellaceae and Clostridiales in Axin1^ΔIEC mice (Fig. 3(l)). Akkermansia is a potential probiotic and a mucolytic bacterium residing in the intestinal mucus layer [31]. These data show that intestinal epithelial Axin1 regulates intestinal secretory cell homeostasis and microbial composition, contributing to increased abundance of Akkermansia.

3.5. Intestinal epithelial Axin1 deficient mice are less susceptible to dextran sodium sulfate (DSS)-induced colitis

To evaluate the physiological impact of loss of IEC Axin1, we utilized a DSS-induced colitis murine model. Axin1^LoxP and Axin1^ΔIEC mice were treated with 5% DSS for 7 days, and parameters of colitis were determined. We found Axin1^LoxP mice lost more weight during DSS administration compared to Axin1^ΔIEC (Fig. 4(a)). Axin1^LoxP DSS mice had apparent blood and diarrhea in their stool according to the DAI [32] (Fig. 4(b)). Accordingly, colon and cecum shortening was observed in DSS-treated Axin1^LoxP mice, compared to DSS-treated Axin1^ΔIEC mice (Figs. 4(c) and (d)). Histological analysis in DSS-treated mice revealed severe inflammation marked by intestinal epithelial destruction and inflammatory cell infiltration (Fig. 4(e)). In contrast, the histological inflammation scores were significantly reduced in DSS-fed Axin1^ΔIEC mice (Fig. 4(f)). We also tested the serum cytokine profile and specifically found a decrease in TNF-α, ΙL-6, and IL-18 in Axin1^ΔIEC mice compared to DSS-treated Axin1^LoxP mice (Figs. 4(g)-(i)). IL-5 and the chemokine eotaxin, which promotes eosinophils into tissues, were elevated in Axin1^ΔIEC mice without DSS challenge (Figs. 4(j) and (k)). In contrast, IL-27 was elevated in Axin1^LoxP mice without DSS challenge (Fig. 4(l)). These data strongly suggest a reduction in inflammation in mice lacking IEC Axin1 during DSS-mediated colitis.

**3.6. Cohousing Axin1^ΔIEC mice with Axin1^LoxP mice increased susceptibility to DSS colitis**

Axin1^ΔIEC mice exhibit alterations in their gut microbiota associated with improved intestinal inflammation outcomes. We hypothesized genetic deletion of Axin1 altered the microbiome contributing to their non-colitogenic phenotype. We co-housed Axin1^ΔIEC and Axin1^LoxP mice for four weeks, followed by 5% DSS challenge for seven days. We found that co-housing decreased the body weight and DAI of co-housed DSS Axin1^ΔIEC mice (Figs. 5(a) and (b)). Co-housing also shortened the colon and cecum in Axin1^ΔIEC mice to levels the Axin1^LoxP mice (Figs. 5(c) and (d)). A. muciniphila in Axin1^ΔIEC mice were decreased after co-housing (Fig. 5(e)). Co-housed Axin1^ΔIEC mice showed increased intestinal epithelial damage Axin1^ΔIEC mice housed alone (Figs. 5(f) and (g)). DSS-treated Axin1^ΔIEC mice had decreased serum TNF-α, IL-6, and IL-18. However, co-housing increased the serum cytokine levels of Axin1^ΔIEC mice to levels similar to those of Axin1^LoxP mice (Figs. 5(h)-(j)). Our results indicate the resistance of Axin1^ΔIEC mice to DSS-induced colitis depends on their gut microbiota.

**3.7. A. muciniphila treatment (AKK) ameliorates DSS-colitis in Axin1^LoxP mice**

We hypothesized that A. muciniphila might be the microorganism driving DSS-colitis protection in Axin1^ΔIEC mice. To test this, we applied AKK to Axin1^LoxP and Axin1^ΔIEC mice during DSS-colitis challenge. We found that AKK improved body weight loss in Axin1^LoxP mice (Fig. 6(a)). In addition, AKK reduced colitis’s severity and lengthened the colon and cecum of Axin1^LoxP mice (Figs. 6(b)-(d)). AKK increased more than 1000 folds in Axin1^LoxP and Axin1^ΔIEC mice with DSS + AKK groups, compared with the DSS groups (Fig. 6(e)). Axin1^ΔIEC mice showed no significant difference in their body weights, DAI, or organ lengths with or without AKK during DSS challenge (Fig. 6). This data indicates that A. muciniphila likely provides the non-colitogenic phenotype in Axin1^ΔIEC mice.

3.8. PC specific deletion of Axin1 results in alteration of intestinal secretory cell lineages

PCs are known to influence the microbiome’s composition [33], [34] and play a critical role in intestinal homeostasis and inflammation [14], [26], [30], [35]. We examined the impact of Axin1 PC on intestinal differentiation by generating an Axin1^ΔPC mouse model. We next isolated PCs by labeling them with an anti-CD24 antibody and sorting them into CD24⁺ and CD24⁻ non-fractions (Fig. 7(a)). We tested Axin1 mRNA expression in the isolated cells by qRT-PCR. PC Axin1 and lysozyme expression in PC-specific knockout of Axin1 (Axin1^ΔPC) mice were reduced compared to Axin1^LoxP PCs (Figs. 7(b) and (c)). IF staining showed PCs from Axin1^ΔPC mice have abnormal lysozyme morphology compared to Axin1^LoxP (Figs. 7(d) and (e)). Alcian blue staining showed that Axin1^ΔPC mice had increased GCs in the small intestine (Fig. 7(f)). Our data indicate that loss of PC Axin1 results in altered secretory cell differentiation, similar to what was observed in the Axin1^ΔIEC model.

3.9. Axin1^ΔPC mice are less susceptible to DSS-induced colitis and are dependent on the gut microbiota

We next determined the impact of loss of PC Axin1 in the disease progression of DSS-colitis. Axin1^ΔPC mice were less susceptible to weight loss and severe disease after DSS administration (Figs. 8(a) and (b)). Axin1^ΔPC mice had longer colons than Axin1^LoxP mice (Fig. 8(c)). Like Axin1^ΔIEC DSS mice, Axin1^ΔPC DSS mice showed less inflammation (Figs. 8(d) and (e)) and decreased IL-6 (Fig. 8(f)). TNF-α was also reduced in the Axin1^ΔPC DSS mice, compared with the Axin1^LoxP mice with DSS treatment (Fig. S1(a) in Appendix A). Axin1^ΔPC mice showed decreased IL-27 even without DSS challenge (Fig. S1(b) in Appendix A). Interestingly, both mouse strains had a decrease and an increase, respectively, in eotaxin and gro-α after DSS-mediated colitis (Figs. S1(c) and (d) in Appendix A). IL-17A was increased in the Axin1^LoxP mice without DSS challenge and significantly decreased in the Axin1^ΔPC DSS mice (Fig. S1(e) in Appendix A). Because the microbiota was shown to be a key player in Axin1^LoxP mice’s vulnerability to DSS-colitis, we examined the transmissibility of the phenotype by performing a co-housing experiment. We challenged Axin1^LoxP and Axin1^ΔPC mice with DSS. After co-housing Axin1^ΔPC with Axin1^LoxPmice, Axin1^ΔPC mice became more susceptible to DSS injury (Fig. 8(g)). Co-housing increased the DAI and shortened the colons of Axin1^ΔPC mice to levels those seen in Axin1^LoxP mice (Figs. 8(h) and (i)). Furthermore, we found an increase in fecal A. muciniphila in the Axin1^ΔPC mice, which was significantly reduced after co-housing (Fig. 8(j)). Like Axin1^ΔIEC mice, Axin1^ΔPC mice had worse histological inflammation and increased levels of IL-6 and TNF-α after co-housing (Figs. 8(k)-(m); Fig. S1(f) in Appendix A). This data indicates that lack of PC Axin1 may be a driving factor in protection against DSS-induced colitis.

3.10. Deletion of Axin1 results in high levels of transcriptional β-catenin and Wnt target genes and proliferation in IECs

In canonical Wnt/β-catenin signaling, Axin1 is known to regulate β-catenin expression by forming a destruction complex that targets it for degradation by the proteasome. We next determined if absence of intestinal epithelial Axin1 results in altered β-catenin and Wnt target gene expression. Axin1^ΔIEC ilea had increased mRNA expression of the Wnt target gene Axin2 (Fig. S2(a) in Appendix A). In addition, Axin1^ΔIEC mice increased cyclin D1 mRNA, compared to Axin1^LoxP mice (Fig. S2(a)). Next, we evaluated transcriptionally activated β-catenin by the expression of phosphorylated β-catenin (S552). We found via Western blots that Axin1^ΔIEC mice had increased expression of β-catenin (S552) (Fig. S2(b) in Appendix A). Lastly, we checked the protein expression of Lgr5 a stem cell marker, and Wnt target gene. Lgr5 protein was increased in Axin1^ΔIEC ilea (Fig. S2(b)). Interestingly, Axin1^ΔPC ilea showed increased mRNA expression of the Wnt target gene Axin2 (Fig. S3(a) in Appendix A). We also found increased level of Lgr5 protein in the Axin1^ΔPC ilea (Fig. S3(b) in Appendix A). We found via IHC staining that Axin1^ΔPC mice showed significantly increased expression of total β-catenin and β-catenin (S552) (Figs. S3(c) and (d) in Appendix A). These data indicate that loss of Axin1 in both IECs and PCs increases β-catenin activation and subsequent Wnt target gene expression.

Wnt/β-catenin signaling is associated with increased proliferation and cell differentiation. BrdU was injected into mice, which were then sacrificed after 2, 12, and 24 hours. BrdU is incorporated into dividing cells during DNA synthesis, remains in place, and is then inherited by daughter cells after division. This experiment was designed to assess proliferation and measure the impact of IEC migration upon the deletion of Axin1 in Axin1^ΔIEC and Axin1^ΔPC mouse models. BrdU IHC staining of jejunum sections showed increased proliferation of IECs in Axin1^ΔIEC and Axin1^ΔPC mice (Figs. S4(a) and (b) in Appendix A). After 2 and 12 hours of BrdU incorporation, there was no difference in cell migration between the control and knockout mice (Figs. S4(a) and (c) in Appendix A). However, a significant increase was found in the cell migration of Axin1^ΔIEC and Axin1^ΔPC after 24 hours (Figs. S4(a) and (c)). We determined proliferation in the small intestine because it has a more rapid rate of proliferation compared with the colon [36]. The colonic changes of β-catenin and proliferation in the Axin^ΔIEC mice showed the same trend as in the small intestine (Fig. S5 in Appendix A).

**3.11. Co-housing or antibiotic treatment changed the proliferation marker proliferating cell nuclear antigen (PCNA) in the Axin1^ΔIEC mice**

Decreased PCNA expression in the ileum was observed by means of IF staining in Axin1^ΔIEC mice that were co-housed with the Axin1^LoxP mice group (Fig. S6 in Appendix A). To further test the role of the microbiome in changing cell proliferation in the intestine, mice were treated with antibiotics (1.0 mg·mL^-1 metronidazole and 0.3 mg·mL^-1 clindamycin) in their drinking water for three weeks (Fig. 9(a)). Through IF staining, we observed decreased PCNA expression in the ileum of Axin1^ΔIEC mice treated with antibiotics (Fig. 9(b)). In contrast, no difference was perceived in the Axin1^LoxP mice with or without antibiotics. In the colon, we also identified decreased PCNA expression in Axin1^ΔIEC mice treated with antibiotics (Fig. 9(c)). Decreased PCNA expression at the protein level was examined using Western blot. In the ileum (Fig. 9(d)) and colon (Fig. 9(e)) of Axin1^ΔIEC mice treated with antibiotics, PCNA was significantly reduced.

4. Discussion

We have demonstrated that the loss of intestinal Axin1 leads to altered intestinal epithelial biology, including an increase in GCs and changes in PC morphology and lysozyme expression. Changes in the intestinal epithelium in Axin1^ΔIEC mice were found to be specifically associated with an altered gut microbiota favoring increased diversity of beneficial microbes, such as A. muciniphila. An increase in A. muciniphila in Axin1^ΔIEC mice was associated with thinning of the mucus barrier. Once induced with DSS, Axin1^ΔIEC mice had less inflammation than Axin1-sufficient animals. Interestingly, co-housing of Axin1^LoxP and Axin1^ΔIEC mice reduced A. muciniphila abundance in Axin1^ΔIEC mice, resulting in vulnerability to DSS colitis. In addition, supplementation with A. muciniphila reduced colitis severity in the control animals, while intestinal epithelial Axin1 knockout mice remained the same. Axin1^ΔPC mice had similar alterations in their small intestinal secretory lineages as those seen in Axin1^ΔIEC mice. Axin1^ΔPC mice had protection against DSS-induced colitis and were more susceptible to colitis after being co-housed with Axin1^LoxP mice (Fig. 10).

In this study, we established two unique experimental models to study the role of Axin1 in intestinal function. Because complete inactivation of Axin1 leads to early embryonic lethality, it is impossible to identify the potential role of Axin1 in later developmental processes using global Axin1 knockout [8]. There is limited literature on Axin1’s protein expression in the intestine. Based on the summary in the Human Protein Atlas, Axin1 is expressed in the small intestine of human samples^†. However, the resolution of the images is not high. There is no publication on the role of Axin1 in the specific cell type of PCs in the intestinal epithelium. We are the first to describe its roles in altering the microbiome and intestinal homeostasis. To investigate the tissue-specific function of Axin1, we developed a system in which Axin1 was conditionally deleted from IECs and PCs utilizing cre-recombinase driven under the villin and Defa6 promoters, respectively. These models allowed us to understand the fundamental role of Axin1 in intestinal and microbial homeostasis and in host responses to inflammatory stimulators.

Our current studies using Axin1 conditional knockout models reveal the physiological functions of Axin1 in IECs and PCs. Our mouse staining analysis demonstrated that Axin1 expression was stronger on the apical side in mice, while its staining in the human epithelium was less clear. This finding might be due to the challenge in identifying high-quality antibodies for Axin1. In the future, we could isolate the villus and crypt in control mice and/or human tissue and determine the difference of Axin1 expression between the two compartments. Doing so would provide further mechanistic insight into Axin1’s contribution to intestinal proliferation and differentiation.

Our data from PC Axin1-deleted mice demonstrate the critical role of Axin1 in innate immunity. We have shown that PC Axin1 confers protection against DSS-induced colitis, and that the gut microbiota contributes to the non-colitogenic phenotype in Axin1^ΔPC mice. We studied both small intestinal and colonic changes related to Axin1 status but focused on PCs in the small intestine for the following reasons:

(1)PCs are mainly present in the small intestine, however, antimicrobial peptides secreted by PCs are released to both the small intestine and colon. Abnormal PCs have been observed in human IBD and colitis models [14], [15], [17], [26], [34].

(2)PCs play a significant role in the intestinal stem cell niche and thereby contribute to cellular proliferation after injury [37], [38], [39].

(3)Previous work has shown that DSS has biochemical and histological effects in the small intestine [40].

(4)DSS-colitis was used to determine whether the same protection in the Axin1^ΔIEC mice was found in the Axin1^ΔPC mice.

PCs play a role in shaping the gut microbiota, including by secreting antimicrobial peptides such as lysozyme. Increased lysozyme mRNA has been observed in the IECs of patients with UC. In addition, lysozyme expression is correlated with the degree of intestinal inflammation [41]. It has also been shown that mice lacking PC lysozyme (Lyz1^-/-) have alterations in their gut bacterial landscape. These alterations were associated with protection from DSS colitis and increased mucolytic bacteria. We have shown that intestinal Axin1 deficiency produced results similar to those seen in Lyz1^-/- mice [34]. Axin1 is classically known as a Wnt/β-catenin signaling regulator. Wnt/β-catenin signaling provokes the differentiation and maturation of PCs, thereby regulating antimicrobial peptide production. Mice lacking the Wnt/β-catenin signaling regulator T-cell factor 1 (TCF1) demonstrated a decrease in lysozyme [35]. Axin1 may regulate lysozyme expression in a Wnt/β-catenin dependent fashion. Our study provides new insights into the molecular mechanism that may contribute to inflammation through intestinal epithelial and PC Axin1.

We have identified increased Axin1 expression at the mRNA and protein level in human UC—findings that are consistent with a previous study reporting Axin1 serum levels being elevated in patients with UC [42]. Activation of Wnt/β-catenin signaling is an aspect of human IBD and colitis animal models [6], [43]. It is likely that Axin1 may regulate intestinal inflammation and IBD through negative regulation of Wnt/β-catenin signaling. However, Axin1 is a multi-domain protein and has been shown to interact with many proteins in several signaling pathways, such as Stress-activated protein kinases (SAPK)/Jun amino-terminal kinases (JNK), p53, transforming growth factor-β (TGF-β), and Wnt signaling [44], [45], [46], [47]. It is unknown how Axin1 may simultaneously regulate multiple signaling pathways in inflammation and infection. We also examined the Axin2- [48] and Wnt/β-catenin-related proliferation in Axin1 knockout mice. It is possible that increased proliferation may contribute to protection from injury in Axin1 knockout mice. This cellular protection may be dependent on the microbiome, because co-housing with Axin1^LoxP mice reduced Akkermansia abundance in Axin1 knockout mice and made them more susceptible to DSS. Moreover, Akkermansia treatment provided protection against DSS-induced colitis in Axin1^LoxP mice. However, this study focused on the altered microbiome and status of Axin1 in the development of colitis. We will further study the Axin1 regulation of IECs at the cellular level in health and disease in future research.

Our data indicate that the protection provided in the Axin1 IEC knockout occurs through modulation of the gut microbiome—specifically, the presence of A. muciniphila. A. muciniphila, a potential probiotic and the only member of the Akkermansia genus, is abundant in the GI tract and plays a critical role in maintaining gut and microbial homeostasis and immunity. Our study shows the direct modulation of Axin1 status on A. muciniphila in the intestine. Reductions in A. muciniphila have been demonstrated in human IBD fecal and mucosal samples [49], [50]. Evidence suggests the protective role of A. muciniphila in diseases such as IBD, type 2 diabetes, and obesity [51], [52], [53]. Furthermore, administration of A. muciniphila ameliorates DSS-induced colitis [32], metabolic disorders [54], and obesity [55] in mice. Mice colonized with A. muciniphila develop immune tolerances toward commensal bacteria [56]. We found that Axin1^ΔIEC mice are protected from DSS-induced inflammation but have a weakened mucus barrier. It has been shown that a depleted mucus barrier results in susceptibility to DSS-induced colitis [57], [58]. However, we found increased MUC2 in Axin1^ΔIEC mice (Fig. 3(b)). Axin1 knockout in IECs did not significantly change the intestinal permeability (data not shown). Moreover, A. muciniphila was enhanced in Axin1^ΔIEC mice. Increased MUC2 and A. muciniphila have been shown to protect against DSS-colitis [32], [58], which collectively may have granted the Axin1^ΔIEC mice increased colitis protection compared with the controls.

Axin1^ΔIEC mice were found to have depletions in Rikenellaceae and Clostridiales. The order Clostridiales is increased in mouse models of colitis and is associated with the clinical course of UC and mucosal inflammation [59], [60]. The role of the family Rikenellaceae is somewhat more unclear, as this family is positively associated with acute DSS treatment in mice [61] and negatively associated with other DSS mouse models [62]. However, microbiota profiling in patients with irritable bowel syndrome (IBS) and IBD showed that Rikenellaceae was unrepresented in the stool of IBD patients compared with controls but was more abundant in IBS patients than in IBD patients [63]. Rikenellaceae resides primarily within the digesta and mucus layer of the colon [64]. Its depletion in the Axin1^ΔIEC mice could be due to the reduced thickness of their mucus layers. In the future, we will investigate the Axin1/A. muciniphila axis and the cellular functions in which host Axin1 may modulate the abundance of A. muciniphila and, subsequently, intestinal mucosal immunity. In conclusion, our study demonstrates a novel and critical role of Axin1 in regulating intestinal epithelial development and microbial homeostasis. PC Axin1 maintains gut and microbial homeostasis, which may be the driving factor in protection against colitis. Loss of intestinal Axin1 may alter innate intestinal immunity through Axin1’s regulation of PC function, which in turn modulates an anti-colitogenic microbiota. Our findings will provide insights into the development of IBD and potential therapeutic strategies for human IBD.

Acknowledgments

We would like to acknowledge the VA Merit Award (1 I01BX004824-01), the National Institute of Diabetes and Digestive and Kidney Diseases/National Institutes of Health Grants (R01 DK105118 and R01DK114126), and the Crohn's & Colitis Foundation Senior Research Award (902766) to Jun Sun. We would like to thank Dr. Mrinalini C. Rao for reading this manuscript and providing insightful suggestions; Figen Seller at the University of Illinois Chicago Electron Microscopy Core for assistance in obtaining TEM images; and the University of Illinois Chicago DNA Services facility for assistance with DNA sequencing. Working model was created using BioRender. R01DK114126 supplement is to promote Diversity in Health-Related Research for PhD student Shari Garrett. The study sponsors play no role in the study design, data collection, analysis, and interpretation of data. The contents do not represent the views of the US Department of Veterans Affairs or the US Government.

Authors' contribution

Shari Garrett performed the cellular and animal studies and the detailed analyses of the results; Yongguo Zhang performed animal studies and analyses of the results; Shari Garrett and Jun Sun prepared the figures and the draft text; Yinglin Xia contributed to the statistical analysis of data and the draft text; and Jun Sun obtained funds, designed the study, and directed the project. All authors contributed to the writing of the manuscript.

Compliance with ethics guidelines

Shari Garrett, Yongguo Zhang, Yinglin Xia, and Jun Sun declare that they have no conflict of interest.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2023.06.007.

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]

S. Alatab, S.G. Sepanlou, K. Ikuta, H. Vahedi, C. Bisignano, S. Safiri, et al. GBD 2017 Inflammatory Bowel Disease Collaborators. The global, regional, and national burden of inflammatory bowel disease in 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet Gastroenterol Hepatol, 5 (1) (2020), pp. 17-30.

[2]	D. Alshehri, O. Saadah, M. Mosli, S. Edris, R. Alhindi, A. Bahieldin. Dysbiosis of gut microbiota in inflammatory bowel disease: current therapies and potential for microbiota-modulating therapeutic approaches. Bosn J Basic Med Sci, 21 (3) (2021), pp. 270-283.

[3]	E.M. Schatoff, B.I. Leach, L.E. Dow. Wnt signaling and colorectal cancer. Curr Colorectal Cancer Rep, 13 (2) (2017), pp. 101-110.

[4]	L. Moparthi, S. Koch. Wnt signaling in intestinal inflammation. Differentiation, 108 (2019), pp. 24-32.

[5]	T.S. Gujral, E.S. Karp, M. Chan, B.H. Chang, G. MacBeath. Family-wide investigation of PDZ domain-mediated protein-protein interactions implicates β-catenin in maintaining the integrity of tight junctions. Chem Biol, 20 (6) (2013), pp. 816-827.

[6]	A. Serafino, N. Moroni, M. Zonfrillo, F. Andreola, L. Mercuri, G. Nicotera, et al. WNT-pathway components as predictive markers useful for diagnosis, prevention and therapy in inflammatory bowel disease and sporadic colorectal cancer. Oncotarget, 5 (4) (2014), pp. 978-992.

[7]	S. Amit, A. Hatzubai, Y. Birman, J.S. Andersen, E. Ben-Shushan, M. Mann, et al. Axin-mediated CKI phosphorylation of β-catenin at Ser 45: a molecular switch for the Wnt pathway. Genes Dev, 16 (9) (2002), pp. 1066-1076.

[8]	R. Xie, R. Jiang, D. Chen. Generation of Axin1 conditional mutant mice. Genesis, 49 (2) (2011), pp. 98-102.

[9]	H.K. Arnold, X. Zhang, C.J. Daniel, D. Tibbitts, J. Escamilla-Powers, A. Farrell, et al. The Axin 1 scaffold protein promotes formation of a degradation complex for c-Myc. EMBO J, 28 (5) (2009), pp. 500-512.

[10]	M. Gavagan, E. Fagnan, E.B. Speltz, J.G. Zalatan. The scaffold protein Axin promotes signaling specificity within the Wnt pathway by suppressing competing kinase reactions. Cell Syst, 10 (6) (2020), pp. 515-525 e5.

[11]	A. Kikuchi. Roles of Axin in the Wnt signalling pathway. Cell Signal, 11 (11) (1999), pp. 777-788.

[12]	Y.G. Zhang, S. Wu, Y. Xia, D. Chen, E.O. Petrof, E.C. Claud, et al. Axin 1 prevents Salmonella invasiveness and inflammatory response in intestinal epithelial cells. PLoS One, 7 (4) (2012), e34932.

[13]	U. Erben, C. Loddenkemper, K. Doerfel, S. Spieckermann, D. Haller, M.M. Heimesaat, et al. A guide to histomorphological evaluation of intestinal inflammation in mouse models. Int J Clin Exp Pathol, 7 (8) (2014), pp. 4557-4576.

[14]	T.E. Adolph, M.F. Tomczak, L. Niederreiter, H.J. Ko, J. Böck, E. Martinez-Naves, et al. Paneth cells as a site of origin for intestinal inflammation. Nature, 503 (7475) (2013), pp. 272-276.

[15]	R. Lu, Y.G. Zhang, Y. Xia, J. Zhang, A. Kaser, R. Blumberg, et al. Paneth cell alertness to pathogens maintained by vitamin D receptors. Gastroenterology, 160 (4) (2021), pp. 1269-1283.

[16]	S. Wu, S. Yoon, Y.G. Zhang, R. Lu, Y. Xia, J. Wan, et al. Vitamin D receptor pathway is required for probiotic protection in colitis. Am J Physiol Gastrointest Liver Physiol, 309 (5) (2015), pp. G341-G349.

[17]	S. Wu, Y.G. Zhang, R. Lu, Y. Xia, D. Zhou, E.O. Petrof, et al. Intestinal epithelial vitamin D receptor deletion leads to defective autophagy in colitis. Gut, 64 (7) (2015), pp. 1082-1094.

[18]	Y.G. Zhang, X. Zhu, R. Lu, J.S. Messer, Y. Xia, E.B. Chang, et al. Intestinal epithelial HMGB 1 inhibits bacterial infection via STAT3 regulation of autophagy. Autophagy, 15 (11) (2019), pp. 1935-1953.

[19]	Y.G. Zhang, R. Lu, S. Wu, I. Chatterjee, D. Zhou, Y. Xia, et al. Vitamin D receptor protects against dysbiosis and tumorigenesis via the JAK/STAT pathway in intestine. Cell Mol Gastroenterol Hepatol, 10 (4) (2020), pp. 729-746.

[20]	A.T. Feldman, D. Wolfe. Tissue processing and hematoxylin and eosin staining. Methods Mol Biol, 1180 (2014), pp. 31-43.

[21]	S. Wu, A.P. Liao, Y. Xia, Y.C. Li, J.D. Li, R.B. Sartor, et al. Vitamin D receptor negatively regulates bacterial-stimulated NF-κB activity in intestine. Am J Pathol, 177 (2) (2010), pp. 686-697.

[22]	D. Jin, Y.G. Zhang, S. Wu, R. Lu, Z. Lin, Y. Zheng, et al. Vitamin D receptor is a novel transcriptional regulator for Axin1. J Steroid Biochem Mol Biol, 165 (Pt B) (2017), pp. 430-437.

[23]	H. Chen, R. Lu, Y.G. Zhang, J. Sun. Vitamin D receptor deletion leads to the destruction of tight and adherens junctions in lungs. Tissue Barriers, 6 (4) (2018), pp. 1-13.

[24]	K.W. Cheon, H.S. Lee, I.S. Parhar, I.S. Kang. Expression of the second isoform of gonadotrophin-releasing hormone (GnRH-II) in human endometrium throughout the menstrual cycle. Mol Hum Reprod, 7 (5) (2001), pp. 447-452.

[25]	Y. Zhang, J. Zhang, Y. Xia, J. Sun. Bacterial translocation and barrier dysfunction enhance colonic tumorigenesis. Neoplasia, 35 (2023), 100847.

[26]	T.C. Liu, B. Gurram, M.T. Baldridge, R. Head, V. Lam, C. Luo, et al. Paneth cell defects in Crohn’s disease patients promote dysbiosis. JCI Insight, 1 (8) (2016), e86907.

[27]	J.G. Caporaso, J. Kuczynski, J. Stombaugh, K. Bittinger, F.D. Bushman, E.K. Costello, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods, 7 (5) (2010), pp. 335-336.

[28]	D. McDonald, M.N. Price, J. Goodrich, E.P. Nawrocki, T.Z. DeSantis, A. Probst, et al. An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. ISME J, 6 (3) (2012), pp. 610-618.

[29]	Y. Xia, J. Sun, D.G. Chen. Statistical analysis of microbiome data with R. Springer, Singapore (2018).

[30]	N.S. Armbruster, E.F. Stange, J. Wehkamp. In the Wnt of Paneth cells: immune-epithelial crosstalk in small intestinal Crohn’s disease. Front Immunol, 8 (2017), 1204.

[31]	P. Paone, P.D. Cani. Mucus barrier, mucins and gut microbiota: the expected slimy partners?. Gut, 69 (12) (2020), pp. 2232-2243.

[32]	X. Bian, W. Wu, L. Yang, L. Lv, Q. Wang, Y. Li, et al. Administration of Akkermansia muciniphila ameliorates dextran sulfate sodium-induced ulcerative colitis in mice. Front Microbiol, 10 (2019), 2259.

[33]	S.R. Lueschow, J. Stumphy, H. Gong, S.L. Kern, T.G. Elgin, M.A. Underwood, et al. Loss of murine Paneth cell function alters the immature intestinal microbiome and mimics changes seen in neonatal necrotizing enterocolitis. PLoS One, 13 (10) (2018), e0204967.

[34]	S. Yu, I. Balasubramanian, D. Laubitz, K. Tong, S. Bandyopadhyay, X. Lin, et al. Paneth cell-derived lysozyme defines the composition of mucolytic microbiota and the inflammatory tone of the intestine. Immunity, 53 (2) (2020), pp. 398-416 e8.

[35]	J. Beisner, Z. Teltschik, M.J. Ostaff, M.M. Tiemessen, F.J.T. Staal, G. Wang, et al. TCF-1-mediated Wnt signaling regulates Paneth cell innate immune defense effectors HD-5 and -6: implications for Crohn’s disease. Am J Physiol Gastrointest Liver Physiol, 307 (5) (2014), pp. G487-G498.

[36]	J. Berlanga-Acosta, R.J. Playford, N. Mandir, R.A. Goodlad. Gastrointestinal cell proliferation and crypt fission are separate but complementary means of increasing tissue mass following infusion of epidermal growth factor in rats. Gut, 48 (6) (2001), pp. 803-807.

[37]	S. Roth, P. Franken, A. Sacchetti, A. Kremer, K. Anderson, O. Sansom, et al. Paneth cells in intestinal homeostasis and tissue injury. PLoS One, 7 (6) (2012), e38965.

[38]	T. Sato, J.H. van Es, H.J. Snippert, D.E. Stange, R.G. Vries, M. van den Born, et al. Paneth cells constitute the niche for Lgr 5 stem cells in intestinal crypts. Nature, 469 (7330) (2011), pp. 415-418.

[39]	X. Mei, M. Gu, M. Li. Plasticity of Paneth cells and their ability to regulate intestinal stem cells. Stem Cell Res Ther, 11 (1) (2020), 349.

[40]	R. Yazbeck, G.S. Howarth, R.N. Butler, M.S. Geier, C.A. Abbott. Biochemical and histological changes in the small intestine of mice with dextran sulfate sodium colitis. J Cell Physiol, 226 (12) (2011), pp. 3219-3224.

[41]	J.M. Kim. Antimicrobial proteins in intestine and inflammatory bowel diseases. Intest Res, 12 (1) (2014), pp. 20-33.

[42]	L. Moraes, M.K. Magnusson, G. Mavroudis, A. Polster, B. Jonefjäll, H. Törnblom, et al. Systemic inflammatory protein profiles distinguish irritable bowel syndrome (IBS) and ulcerative colitis, irrespective of inflammation or IBS-like symptoms. Inflamm Bowel Dis, 26 (6) (2020), pp. 874-884.

[43]	E.M. Bradford, S.H. Ryu, A.P. Singh, G. Lee, T. Goretsky, P. Sinh, et al. Epithelial TNF receptor signaling promotes mucosal repair in inflammatory bowel disease. J Immunol, 199 (5) (2017), pp. 1886-1897.

[44]	W. Luo, S.C. Lin. Axin: a master scaffold for multiple signaling pathways. Neurosignals, 13 (3) (2004), pp. 99-113.

[45]	M. Furuhashi, K. Yagi, H. Yamamoto, Y. Furukawa, S. Shimada, Y. Nakamura, et al. Axin facilitates Smad 3 activation in the transforming growth factor beta signaling pathway. Mol Cell Biol, 21 (15) (2001), pp. 5132-5141.

[46]	S. Salahshor, J.R. Woodgett. The links between Axin and carcinogenesis. J Clin Pathol, 58 (3) (2005), pp. 225-236.

[47]	Y. Zhang, S.Y. Neo, X. Wang, J. Han, S.C. Lin. Axin forms a complex with MEKK1 and activates c-Jun NH(2)-terminal kinase/stress-activated protein kinase through domains distinct from Wnt signaling. J Biol Chem, 274 (49) (1999), pp. 35247-35254.

[48]	C. Harnack, H. Berger, A. Antanaviciute, R. Vidal, S. Sauer, A. Simmons, et al. R-spondin 3 promotes stem cell recovery and epithelial regeneration in the colon. Nat Commun, 10 (1) (2019), 4368.

[49]	C.W. Png, S.K. Lindén, K.S. Gilshenan, E.G. Zoetendal, C.S. McSweeney, L.I. Sly, et al. Mucolytic bacteria with increased prevalence in IBD mucosa augment in vitro utilization of mucin by other bacteria. Am J Gastroenterol, 105 (11) (2010), pp. 2420-2428.

[50]	M. Rajilić-Stojanović, F. Shanahan, F. Guarner, W.M. de. Vos. Phylogenetic analysis of dysbiosis in ulcerative colitis during remission. Inflamm Bowel Dis, 19 (3) (2013), pp. 481-488.

[51]	Y. Xu, N. Wang, H.Y. Tan, S. Li, C. Zhang, Y. Feng. Function of Akkermansia muciniphila in obesity: interactions with lipid metabolism, immune response and gut systems. Front Microbiol, 11 (2020), 219.

[52]	C.T. Shih, Y.T. Yeh, C.C. Lin, L.Y. Yang, C.P. Chiang. Akkermansia muciniphila is negatively correlated with hemoglobin A1c in refractory diabetes. Microorganisms, 8 (9) (2020), 1360.

[53]	T. Zhang, P. Li, X. Wu, G. Lu, C. Marcella, X. Ji, et al. Alterations of Akkermansia muciniphila in the inflammatory bowel disease patients with washed microbiota transplantation. Appl Microbiol Biotechnol, 104 (23) (2020), pp. 10203-10215.

[54]	S. Zhao, W. Liu, J. Wang, J. Shi, Y. Sun, W. Wang, et al. Akkermansia muciniphila improves metabolic profiles by reducing inflammation in chow diet-fed mice. J Mol Endocrinol, 58 (1) (2017), pp. 1-14.

[55]	A. Everard, C. Belzer, L. Geurts, J.P. Ouwerkerk, C. Druart, L.B. Bindels, et al. Cross-talk between Akkermansia muciniphila and intestinal epithelium controls diet-induced obesity. Proc Natl Acad Sci USA, 110 (22) (2013), pp. 9066-9071.

[56]	M. Derrien, P. Van Baarlen, G. Hooiveld, E. Norin, M. Müller, W.M. de. Vos. Modulation of mucosal immune response, tolerance, and proliferation in mice colonized by the mucin-degrader Akkermansia muciniphila. Front Microbiol, 2 (2011), 166.

[57]	J. Fang, H. Wang, Y. Zhou, H. Zhang, H. Zhou, X. Zhang. Slimy partners: the mucus barrier and gut microbiome in ulcerative colitis. Exp Mol Med, 53 (5) (2021), pp. 772-787.

[58]	M. Van der Sluis, B.A.E. de Koning, A.C.J.M. de Bruijn, A. Velcich, J.P.P. Meijerink, J.B. van Goudoever, et al. Muc2-deficient mice spontaneously develop colitis, indicating that MUC 2 is critical for colonic protection. Gastroenterology, 131 (1) (2006), pp. 117-129.

[59]	L.Y. Pei, Y.S. Ke, H.H. Zhao, L. Wang, C. Jia, W.Z. Liu, et al. Role of colonic microbiota in the pathogenesis of ulcerative colitis. BMC Gastroenterol, 19 (1) (2019), 10.

[60]	Y. Nishihara, H. Ogino, M. Tanaka, E. Ihara, K. Fukaura, K. Nishioka, et al. Mucosa-associated gut microbiota reflects clinical course of ulcerative colitis. Sci Rep, 11 (1) (2021), 13743.

[61]	P.M. Munyaka, M.F. Rabbi, E. Khafipour, J.E. Ghia. Acute dextran sulfate sodium (DSS)-induced colitis promotes gut microbial dysbiosis in mice. J Basic Microbiol, 56 (9) (2016), pp. 986-998.

[62]	Z. Gao, K.Y. Chen, O. Mueller, H. Zhang, N. Rakhilin, J. Chen, et al. Microbiota of inflammatory bowel disease models. Annu Int Conf IEEE Eng Med Biol Soc, 2018 (2018), pp. 2374-2377.

[63]	A. Lo Presti, F. Zorzi, F. Del Chierico, A. Altomare, S. Cocca, A. Avola, et al. Fecal and mucosal microbiota profiling in irritable bowel syndrome and inflammatory bowel disease. Front Microbiol, 10 (2019), 1655.

[64]	G.P. Donaldson, S.M. Lee, S.K. Mazmanian. Gut biogeography of the bacterial microbiota. Nat Rev Microbiol, 14 (1) (2016), pp. 20-32.

RIGHTS & PERMISSIONS

THE AUTHOR