Lack of Vitamin D Receptor Leads to Hyperfunction of Claudin-2 in Intestinal Inflammatory Responses

Abstract

Background

Vitamin D3 and vitamin D receptor (VDR) are involved in the pathogenesis of inflammatory bowel disease (IBD) and bacterial infection. Claudin-2 is a junction protein that mediates paracellular water transport in epithelia. Elevation of Claudin-2 is associated with active IBD. However, VDR involved in epithelial junctions under inflammation and infection remains largely unknown. We investigated the mechanisms on how VDR and Claudin-2 are related in inflamed states.

Methods

Using cultured VDR^-/- enteroids, human intestinal epithelial cells, VDR^-/- mice with Salmonella- or DSS-colitis, and human IBD samples, we investigated the mechanisms how VDR and Claudin-2 are related in inflamed states.

Results

After Salmonella infection had taken place, we observed significantly enhanced Claudin-2 and an increased bacterial invasion and translocation. A lack of VDR regulation led to a robust increase of Claudin-2 at the mRNA and protein levels post-infection. In DSS-treated VDR^-/- mice, Claudin-2 was significantly increased. Location and quantification of Claudin-2 protein in the mouse colons treated with DSS also confirmed these results. Inflammatory cytokines were significantly higher in the serum and mRNA levels in intestine, which are known to increase Claudin-2. Furthermore, in inflamed intestine of ulcerative colitis patients, VDR expression was low and Claudin-2 was enhanced. Mechanistically, we identified the enhanced Claudin-2 promoter activity through the binding sites of NF-κB and STAT in inflamed VDR^-/- cells.

Conclusions

Our studies have identified a new role for intestinal epithelial VDR in regulating barrier functions in the context of infection and inflammation.

INTRODUCTION

Tissue barrier functions are determined by tight junction (TJ) structural components.^1–6 Defective intestinal barrier function has been implicated in inflammatory bowel disease (IBD) and can predict relapse during clinical remission.⁷ Claudin-2 forms a water channel to permit the paracellular passage of water through its pore in epithelia.⁸ Colonic Claudin-2 expression is uniquely restricted to the crypt base, which is the proliferative zone.^{9, 10} Increased Claudin-2 tends to make intestinal epithelial cells leakier.^{7, 10–17} Elevation of Claudin-2 is associated with active human IBD.^{18, 19}

Impaired epithelial barriers are not only associated with chronic inflammation but also infection. Salmonella enterica serotypes are invasive enteric pathogens spread through fecal contamination of food and water sources. Salmonella induce the disruption of intestinal barriers during infection.^20–24 Our previous study has demonstrated that Salmonella manipulate the distribution of Claudin-2 and elevated its protein expression in intestine.²⁴ Intestinal epithelial cells with Claudin-2 knockdown had significantly less internalized Salmonella than control cells with normal Claudin-2 expression.²⁴

The nuclear vitamin D receptor (VDR) mediates most known functions of 1,25-dihydroxyvitamin D. Vitamin D and VDR are key players in calcium homeostasis.²⁵ It also plays a protective role in infection and inflammation.^{26, 27} Vitamin D deficiency has been implicated in patients with IBD.^28–32 Vitamin D receptor is identified as an IBD risky gene.^33–37 Tight junction proteins (eg, ZO-1 and Occludin) are known to maintain a barrier.³⁸ ZO-1 and Occludin are upregulated in intestinal epithelial cells by 1,25 (OH)₂D₃ treatment in vitro.^{39, 40} Claudin-2 has been identified as a target gene of the VDR signaling in cells without any inflammatory stimulation.^{41, 42} Therefore, it is interesting to study how VDR is involved in regulating Claudin-2 and barrier functions under inflammation and infection. Insights into the mechanisms responsible for VDR and intestinal barrier dysfunction are needed, especially in disease models in vivo.

In the current study, we hypothesize that a lack of intestinal VDR leads to hyperfunction of Claudin-2 and impaired barrier functions in infection and inflammatory responses. Using VDR^-/- mice, organoids, human samples, and cultured human intestinal epithelial cells, we performed a series of molecular and biochemical experiments in vivo and in vitro to investigate VDR regulation of Claudin-2 expression in infection and inflammation. We investigated the mechanisms on how VDR and Claudin-2 are related in inflamed states. Our findings reveal a novel role of VDR in regulating barrier functions in inflammation and infectious disease.

MATERIALS AND METHODS

Human Tissue Samples

This study was performed in accordance with approval from the University of Rochester Ethics Committee (RSRB00037178). Colorectal tissue samples were obtained with informed consent from the sigmoid colon of patients exhibiting no apparent intestinal pathology and from the mucosa of patients undergoing anterior resection. All pathologic diagnoses were evaluated by pathologists. The ulcerative colitis (UC) cohort consisted of 16 patients with previous diagnosis of UC (10 female, 6 male, mean age = 66.6 years, range of 51 to 83), while the age-matched control cohort consists of 10 patients (6 female, 4 male, mean age = 65.7 years, range of 44 to 85) without previous diseases in gastrointestinal tract. All patient samples were obtained at the University of Rochester Medical Center from 2008 to 2012. In this study, no Crohn’s disease (CD) patients were included.

Animals

VDR^+/- and VDR^−/− mice on a C57BL6 background were obtained by breeding heterozygous VDR^+/− mice.⁴³ VDR^+/- and VDR^-/- mice were used to compare the difference between VDR-null and 1 allele of the VDR gene. We also confirmed similar results were obtained when VDR^+/+ mice were used (data not shown). Mice were fed a normal chow diet and under 12:12 light/dark cycles. Experiments were performed on 2 to 3 months old mice. The animal work was approved by the UIC Office of Animal Care and the Rush University Committee on Animal Resources. Euthanasia method was sodium pentobarbital (100 mg per kg body weight) intraperitoneal (IP) injection, followed by cervical dislocation. All experiments were carried out in accordance with the approved guidelines.

Culture of Mouse Small Intestinal Organoids and Treatment With Salmonella

Organoids from small intestine of VDR^+/-and VDR^-/- mice were prepared and maintained as previously described.^{44, 45} Mini gut medium [advanced Dulbecco’s Modified Eagle’s Medium (DMEM)/F12 supplemented with HEPES, L-glutamine, N2, and B27] was added to the culture, along with R-Spondin, Noggin, and epidermal growth factor (EGF). At day 7 after passage, organoids were colonized by Salmonella for 30 minutes, then washed, and incubated for 1 hour in Mini gut medium with Gentamicin (500 µg/mL).

Bacterial Strains and Growth Condition

Wild-type Salmonella Typhimurium ATCC14028 was used in this study. Nonagitated microaerophilic bacterial cultures were prepared.⁴⁶

Streptomycin Pretreated Mouse Model

Animal experiments were performed using specific-pathogen-free female VDR^+/- and VDR^−/− mice that were 6 to 8 weeks old, as previously described.⁴⁷ Water and food were withdrawn 4 hours before oral gavage with 7.5 mg/mouse of streptomycin (100 µL of sterile solution or 100 µL of sterile water in control). Afterward, animals were supplied with water and food ad libitum. Twenty hours after streptomycin treatment, water and food were withdrawn again for 4 hours before the mice were infected with 1x10⁷ CFU of S. typhimurium [100 µL suspension in Hank’s Balanced Salt Solution (HBSS)] or treated with sterile HBSS (control). Three days after infection, mice were sacrificed, and tissue samples from the intestinal tracts were removed for analysis.

Salmonella Burden in Intestine, Spleen, and Liver

Intestine, liver, and spleen were dissected from each mouse, put into 14-mL tubes with 5 mL of sterile phosphate-buffered saline (PBS), cut into pieces with scissors, then homogenized adequately using a Polytron PT2100 (Kinematica, Switzerland). Each homogenate was diluted 1000× to 10,000× with Luria Broth (LB), plated (100 µL) on MacConkey agar plates, and incubated at 37°C overnight. Colony-forming units were quantified.

Induction of Mouse Colitis and Assessment of Clinical Disease

Mice were fed 3% (wt/vol) dextran sulfate sodium (DSS, molecular mass = 36 to 50 kD; USB Corp. Cleveland, OH, USA) dissolved in filter-purified and sterilized water.⁴⁸ At day 3 after DSS administration, mice were scarified. Body weight and animal symptoms, including the extent of diarrhea and rectal bleeding, were closely monitored during and after DSS treatment. Symptom scores were determined by assessing the degree of body weight loss, stool consistency (diarrhea), and hemoccult positivity or gross blooding in each animal according to previously published methods.⁴⁰

Histology

For each animal, histological examination was performed of the cecum. Cecum length and inflammation were key features in this Salmonella-colitis mouse model.⁴⁹ Cecum length was measured using a ruler. Histological parameters were quantified in a blinded fashion using modified validated scoring systems.^48–50 Six independent parameters were measured: severity of inflammation, depth of injury, crypt damage, submucosal edema, polymorphonuclear cells (PMNs) in the lamina propria, and mean number of goblet cells. Histological scores were multiplied by a factor reflecting the percentage of tissue involvement (×1, 0%–25%; ×2, 26%–50%; ×3, 51%–75%; ×4, 76%–100%), and values were added to obtain a total score.

Cell Culture

Human epithelial CaCO2 and SKCO15 cells were maintained on transwell inserts (0.33 or 4.67 cm², 0.4 mm pore; Costar, Cambridge, MA, USA) in DMEM supplemented with 10% fetal bovine serum, penicillin-streptomycin (Penicillin, 100 I.U./mL/Streptomycin, 100µg/mL), and L-glutamine (4.5g/L). Human epithelial HCT116 cells, VDR^+/− mouse embryonic fibroblasts (MEF), and VDR^−/− MEF cells were cultured in DMEM medium supplemented with 10% (vol/vol) fetal bovine serum, as previously described.^{52, 53}

Immunoblotting

Cultured cells were rinsed twice with ice-cold HBSS, lysed in protein-loading buffer (50 mM Tris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol), and then sonicated. Mouse colonic epithelial cells were collected by scraping the tissue from the colon of the mouse, including the proximal and distal regions.^{51, 52} The cells were sonicated in lysis buffer [10 mM Tris, pH 7.4, 150 mM NaCl, 1 mM ethylene diamine tetraacetic acid (EDTA), 1 mM ethylene glycol tetraacetic acid (EGTA), pH 8.0, 1% Triton X-100] with 0.2 mM sodium ortho-vanadate, and protease inhibitor cocktail. The protein concentration was measured using the BioRad Reagent (BioRad, Hercules, CA, USA). Equal amounts of protein were separated by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and immunoblotted with primary antibodies. The following antibodies were used: anti-claudin-2, anti-claudin-3, anti-claudin-7 (Invitrogen, Carlsbad, CA, USA), anti-p-STAT3, anti-STAT3, anti-p-IκBα, anti-IκBα, anti-p-SAPK/JNK, anti-SAPK/JNK (Cell Signal, Beverly, MA, USA), anti-Villin, and anti-VDR (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), or anti-β-actin (Sigma-Aldrich, Milwaukee, WI, USA) antibodies and were visualized by enhanced chemiluminescence (ECL). Membranes that were probed with more than one antibody were stripped before reprobing.

Immunofluorescence

Colonic tissues were freshly isolated and embedded in paraffin wax after fixation with 10% neutral buffered formalin. Immunofluorescence was performed on paraffin-embedded sections (4 μm), after preparation of the slides as described previously,⁵³ followed by incubation for 1 hour in blocking solution (2% bovine serum albumin, 1% goat serum in HBSS) to reduce nonspecific background. The tissue samples were incubated overnight with primary antibodies at 4°C. The following antibodies were used: anti-Claudin-2, anti-Claudin-7 (Invitrogen, Grand Island, NY, USA), anti-Salmonella typhimurium (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA). Samples were then incubated with secondary antibodies (goat anti-mouse Alexa Fluor 488 or goat anti-rabbit Alexa Fluor 488, Molecular Probes, CA; 1:200) for 1 hour at room temperature. Tissues were mounted with SlowFade Antifade Kit (Life technologies, s2828, Grand Island, NY, USA), followed by a coverslip, and the edges were sealed to prevent drying. Specimens were examined with a Zeiss laser scanning microscope (LSM 710 (Carl Zeiss Inc., Oberkochen, Germany).

Vitamin D Receptor and Claudin-2 Expression in Human IBD

Immunohistochemistry (IHC) of VDR and immunofluorescence (IF) of Claudin-2 in human normal and ulcerative colitis samples staining were scored by a well-trained pathologist. Immunohistochemistry of VDR and IF of Claudin-2 staining was initially assessed as a product of staining intensity (0, no staining; 1, minimal; 2, slight; 3, moderate; 4, marked intensity). We randomly took 3 scores derived from each samples. As a result, the final analytical sample consisted of 30 normal scores and 48 UC scores.⁵⁴ Statistical analyses were performed using Student t test.

In Vivo Permeability Measurements Across Colonic Mucosa

Mice were treated with Salmonella for 3 days. Fluorescein Dextran (Sigma-Aldrich, St. Louis, MO, USA; Molecular weight 3000 Da, diluted in HBSS) was gavaged (50 mg/g mouse). Four hours later, mouse blood samples were collected for fluorescence intensity measurement.²³

Mouse Cytokine

Mouse blood samples were collected by cardiac puncture and placed in tubes containing EDTA (10 mg/mL). Mouse cytokines were measured using a mouse cytokine 10-Plex Panel kit (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Briefly, beads of defined spectral properties were conjugated to protein-specific capture antibodies and added along with samples (including standards of known protein concentration, control samples, and test samples) into the wells of a filter-bottom microplate, where proteins bound to the capture antibodies over the course of a 2-hour incubation. After washing the beads, protein-specific biotinylated detector antibodies were added and incubated with the beads for 1 hour. After removal of excess biotinylated detector antibodies, the streptavidin-conjugated fluorescent protein R-phycoerythrin (streptavidin-RPE) was added and allowed to incubate for 30 minutes. After washing to remove unbound streptavidin-RPE, the beads were analyzed with the Luminex detection system PerkinElmer CS1000 Autoplex Analyzer (PerkinElmer Inc., Waltham, MA, USA).

Transient Transfections

Transient transfections were performed with Lipofectamine 2000 (Invitrogen, San Diego, CA, USA) in accordance with the manufacturer’s instructions. Cells were seeded on 60-mm dishes overnight before transfection with plasmids. Plasmid DNA solution was mixed with the transfection reagent at a ratio of 1:1 before being added to cells. After a 24-hour transfection period, the proteins were extracted for analysis.

Vitamin D Receptor siRNA

SKCO15 cells were grown in 12-well plates. The cells were transfected with on-Target plus smart pool human VDR siRNA (Dharmacon Inc., Lafayette, MO, USA) or scrambled siRNA control (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) using Surefect reagent (SABiosciences, Frederick, MD, USA). After 72-hour transfection, cells were colonized by Salmonella for 30 minutes, then washed and incubated in DMEM with Gentamicin (500 µg/mL) within the indicated time.

Transcriptional Activation

Deletions of different domains of the Claudin-2 promoter plasmids were constructed as previously described.⁴¹ After a 24-hour transfection period, cells were colonized by Salmonella for 30 minutes, washed, and incubated for 30 minutes in DMEM with Gentamicin (500 µg/mL). The cells were lysed, and luciferase activity was determined using the Dual Luciferase Reporter Assay System (Promega, Madison, WI, USA) with a TD-20/20 luminometer (Turner Designs, Sunnyvale, CA, USA). Firefly luciferase activity was normalized to Renilla luminescence activity, and the activity was expressed as relative units.

Real-Time Quantitative Polymerase Chain Reaction Analysis

Total RNA was extracted from mouse epithelial cells or cultured cells using TRIzol reagent (Invitrogen, Grand Island, NY, USA). The RNA integrity was verified by electrophoresis. RNA reverse transcription was performed using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s protocol. The RT cDNA reaction products were subjected to quantitative real-time polymerase chain reaction (RT-PCR) using CTFX 96 Real-time system (Bio-Rad, Hercules, CA, USA) and SYBR green supermix (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s protocol. All expression levels were normalized to β-actin levels of the same sample. Percent expression was calculated as the ratio of the normalized value of each sample to that of the corresponding untreated control cells. All real-time PCR reactions were performed in triplicate. Optimal primer sequences were designed using Primer-BLAST or were obtained from Primer Bank primer pairs listed in Table 1.

TABLE 1.

Real-time PCR Primers

Primers name	Sequence
mβ-actinF	5΄-TGTTACCAACTGGGACGACA-3΄
mβ-actinR	5΄-CTGGGTCATCTTTTCACGGT-3΄
mVDRF	5΄-GAATGTGCCTCGGATCTGTGG-3΄
mVDRR	5΄-ATGCGGCAATCTCCATTGAAG-3΄
mClaudin-2F	5΄-GCAAACAGGCTCCGAAGATACT-3΄
mClaudin-2R	5΄-GAGATGATGCCCAAGTACAGAG-3΄
mTNF-αF	5΄-CCCTCACACTCAGATCATCTTCT-3΄
mTNF-αR	5΄-GCTACGACGTGGGCTACAG-3΄
mIL-6-F	CTGCAAGAGACTTCCATCCAG
mIL-6-R	AGTGGTATAGACAGGTCTGTTGG

Statistical Analysis

All data are expressed as the mean ± SD. All statistical tests were 2-sided. All P values <0.05 were considered statistically significant. The distribution of each outcome variable was tested before statistical analysis. The test results showed that the outcome data are normally distributed. Thus, the differences between samples were analyzed using Student t test for 2 groups and using 1-way ANOVA for more than 2 groups, respectively. Multiple comparisons of mean body weight were performed using 2-way ANOVA. Because we were interested in comparing the mean differences among treatment groups, the group effects at each post-treatment days of 1, 2, and 3 were tested using generalized linear mixed models. The P values in ANOVA analysis and generalized linear mixed models were adjusted using the Tukey method to ensure accurate results. Statistical analyses were performed using SAS version 9.4 (SAS Institute, Inc., Cary, NC, USA).

RESULTS

Claudin-2 Protein Was Significantly Increased in Salmonella-Colonized VDR^-/- Cells

We have reported a Salmonella-infected organoid culture system suitable for studying host-bacterial interactions.⁵⁵ We used the VDR^+/- and VDR^-/- organoids infected with Salmonella (Fig. 1A). Vitamin D receptor protein level was lower in the VDR^-/- organoids at the basal level. Salmonella infection induced significantly higher Claudin-2 upregulation in the VDR^-/- organoids compared with the VDR^+/- organoids with normal level of VDR (Fig. 1B). In the loss function study, we knocked down VDR in human SKCO15 cells using VDR-siRNA. We found that cells with lower VDR expression had higher Claudin-2 upregulation after Salmonella infection compared with the cells with normal level of VDR (Fig. 1C). Overall, these data indicate that a lack of VDR leads to a robust increase of Claudin-2 post Salmonella infection.

FIGURE 1.

Claudin-2 protein expression was significantly increased in Salmonella-colonized VDR^-/- cells. A, The micrographs showed representative VDR^+/- and VDR^-/- organoids with or without Salmonella infection. Each condition was examined in triplicates with multiple (>10) organoids in each sample. Each experiment was repeated three times. B, Claudin-2 protein expression increased 9.3-fold in VDR^-/- organoids and 2.3-fold in VDR^+/- organoids infected with Salmonella compared with control. (Data are expressed as mean ± SD; n = 3, Student t test, *P < 0.05). C, Claudin-2 protein expression was upregulated in VDR-siRNA SKCO15 cells compared with C-siRNA SKCO15 cells infected with Salmonella. SKCO15 cells were transfected with human VDR siRNA or scrambled siRNA control according to the manufacturer’s instructions. At 72 hours post-transfection, cells were colonized by Salmonella for 30 minutes, washed, and incubated for 30 minutes in DMEM with Gentamicin (500 µg/mL). Cell lysis samples were collected for western blots. (Data are expressed as mean ± SD; n = 3, 1-way ANOVA test; *P < 0.05).

More Severe Salmonella-Infection in the VDR^-/- Mice

Based on the in vitro data in Fig. 1, we further hypothesize that infection and inflammation may take over and drive hyperreglation of Claudin-2 in intestines lacking VDR regulation. We thus investigated VDR and Claudin-2 in experimental Salmonella-colitis mice using VDR^-/- and VDR^+/- mice (Fig. 2). In the present study, we used VDR^+/- mice because we wanted to compare the difference between VDR-null and 1 allele of the VDR gene. Similar results were seen when VDR^+/+ mice were used (data not shown). VDR^-/- mice had more severe infection and lost the integrity of intestinal epithelium post-infection compared with the Salmonella-infected VDR^+/- mice (Fig. 2A). The inflammation score was significantly higher in the infected VDR^-/- mice. Cecum length was smaller in the infected VDR^-/- mice compared with the infected VDR^+/- mice (Fig. 2B), suggesting more inflammation induced by Salmonella in mice lacking VDR. Furthermore, we found that Salmonella infected VDR^-/- mice had much more body weight loss than VDR^+/- mice during day 1 and day 3 post-infection (adjusted P value <0.0001) . Compared with VDR^+/- and VDR^-/- mice, the Salmonella-infected mice also significantly dropped more weight than those without the Salmonella-infected mice at day 3 (adjusted P value <0.0001) (Fig. 2C). Taken together, our data suggest more severe Salmonella infection in the VDR^-/- mice.

FIGURE 2.

VDR^-/- mice were more susceptible to Salmonella infection than VDR^+/- mice. A, Representative H&E staining and inflammation scores of intestine tissues from VDR^+/- and VDR^-/- mice with or without Salmonella infection. (Data are expressed as mean ± SD; n = 7 mice/group, 1-way ANOVA test; *P < 0.05). B, Cecum shortening was observed in the VDR^-/- mice infected with Salmonella. The cecum length was significantly shorter in the VDR^-/- mice infected with Salmonella compared with the VDR^+/- mice infected with Salmonella. (Data are expressed as mean ± SD; n = 7 mice/group, 1-way ANOVA test; *P < 0.05.) C, Relative body weight changes in VDR^+/- and VDR^-/- mice infected with Salmonella or not. (Data are expressed as mean ± SD; n = 8 mice/group, 2-way ANOVA.) D, Claudin-2 protein expression increased 8.8-fold in VDR^-/- mice intestinal epithelial cells and 2.6 fold in VDR^+/- mice intestinal epithelial cells infected with Salmonella compared with control. Increased folds of normalized Claudin-2 expressin after Salmonella treatment to control. (Data are expressed as mean ± SD; n = 3, Student t test, *P < 0.05). E, Claudin-2 mRNA was upregulated in VDR^-/- mice intestinal epithelial cells infected with Salmonella. (Data are expressed as mean ± SD; n = 3, 1-way ANOVA test; *P < 0.05). F, Location and quantification of Claudin-2 protein in colons 3 days post-infection with Salmonella. Data are from a single experiment and are representative of 5 mice per group. G, Intestine permeability increased in VDR^-/- mice infected with Salmonella. Mice were treated with Salmonella for 3 days. Fluorescein Dextran (Molecular weight 3000 Da, diluted in HBSS) was gavaged (50 mg/g mouse). Four hours later, mouse blood samples were collected for fluorescence intensity measurement. (Data are expressed as mean ± SD; n = 7 mice/group, 1-way ANOVA test; *P < 0.05).

Uncontrolled Claudin-2 Elevation in VDR^-/- Mice with Salmonella-Colitis

When performed in vivo, VDR deletion led to reduced Claudin-2 in the basal level of cells without any treatment. After Salmonella infection in VDR^-/- mice, we observed significantly enhanced Claudin-2 protein by pathogenic Salmonella 8 hours post-infection, which is the early stage of infection. The basal levels of claudin-2 protein in VDR^-/- organoids and VDR^-/- mouse expression are lower compared with VDR^+/- organoids and VDR^+/- mouse, respectively. After Salmonella infection, the similar levels of VDR^+/- and VDR^-/- Claudin-2 were observed. After normalization to the basal levels accordingly, the fold changes were obtained: Salmonella infection induced significant increase of Claudin-2 (Fig. 2D). There was robust increase of Claudin-2 mRNA in inflammatory intestinal epithelial cells that lacked VDR (Fig. 2E).

Vitamin D receptor deletion led to a small amount of Claudin-2 restricted to the crypt base without any treatment. Post-infection, the distribution of Claudin-2 moved from the crypt base to the middle of the crypt in the infected intestine (Fig. 2F). Further, the fluorescence data for intestinal permeability showed that Salmonella infection induced higher permeability in both VDR^+/- and VDR^-/- mice, whereas the VDR^-/- mice had significantly high permeability post-infection (Fig. 2G).

Salmonella Translocation in Liver and Spleens of the VDR^-/- Mice

We investigated the Salmonella load in the intestine and found significantly increased bacterial burden in VDR^-/- cecum compared with the VDR^+/- mice (Fig. 3A). Bacterial translocation was examined in liver and spleen 4 days post-infection (Fig. 3B and 3C). More Salmonella were detected in the liver and spleen of VDR^+/- mice, suggesting leaky tight junctions that allow for translocation of gut bacteria to other organs. Culture of Salmonella clones further showed that significantly higher numbers of bacterial translocation to liver and spleen of the VDR^-/- mice post-Salmonella infection compared with the VDR^+/- mice (Fig. 3D).

FIGURE 3.

Deletion of VDR leads to increased levels of inflammatory cytokines and enhanced Salmonella invasion. A, Salmonella location and quantification (green) in the ceca of VDR^+/- and VDR^-/- mice. VDR^+/- and VDR^-/- mice were infected with Salmonella for 3 days. (Data are expressed as mean ± SD; n = 5, 1-way ANOVA test; *P < 0.05). B, Salmonella location and quantification (green) in the liver of VDR^+/- and VDR^-/- mice and C, Salmonella location and quantification (green) in the spleen of VDR^+/- and VDR^-/- mice. VDR^+/- and VDR^-/- mice were infected with Salmonella for 3 days. D, Salmonella concentrations in the liver and spleen of VDR^-/- mice. Most of Salmonella were translocated to the liver and spleen 3 days VDR^-/- after being infected with Salmonella. (Data are expressed as mean ± SD; n = 5, 1-way ANOVA test; *P < 0.05). E, Serum levels of IL-6 and TNF-α were significantly higher in Salmonella-infected VDR^-/- mice compared with VDR^+/- mice. (Data are expressed as mean ± SD; n = 5, 1-way ANOVA test; *P < 0.05).

Elevated Level of Intestinal Claudin-2 is Associated With Active Inflammation in VDR^-/- Mice

An elevated level of intestinal Claudin-2 is associated with active inflammation.^{15, 18} We then measured the serum inflammatory cytokines by ELISA (Fig. 3E). We found that Interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α) were significantly increased in the infected VDR^-/- mice compared with the VDR^+/- mice post-infection. We thus hypothesize that inflammation may further drive hyperfunction of Claudin-2 in absence of VDR. Then, we investigated the changes of Claudin-2 in an experimental DSS-induced colitis model. The body weight of the DSS-induced VDR^-/- mice dropped more than that of the VDR^+/- mice at day 3 (adjusted P value <0.05) (Fig. 4A): 15% body weight loss in VDR^-/- mice and 5% body weight loss in VDR^+/- mice after DSS treatment for 3 days. We also observed increased intestinal permeability in VDR^-/- mice compared with VDR^+/- mice 3 days post-DSS (Fig. 4B). In DSS-treated VDR^-/- mice, Claudin-2 protein was significantly increased compared with VDR^+/- mice treated with DSS. Claudin-2 protein was 5-fold increased in DSS-treated VDR^-/- mice compared with the 2-fold increase in VDR^+/- mice treated with DSS (Fig. 4C). Location and quantification of Claudin-2 protein in the mouse colons 3 days post-DSS-treatment also confirmed these results (Fig. 4D). Inflammatory cytokines IL-6 and TNF-α are known to increase Claudin-2.^56–58 In the DSS-treated VDR^-/- mice, we found significantly higher IL-6 and TNF-α in the serum, as shown in Fig. 4E. In the colonic mucosal samples, we also found higher IL-6 and TNF-α at the mRNA levels (Fig. 4F).

FIGURE 4.

Lack of VDR regulation led to increased expression of Claudin-2 protein and inflammatory cytokines IL-6 and TNF-α in the DSS-colitis VDR^-/- mice. A, Relative body weight changes in VDR^+/- and VDR^-/- mice with or without DSS treatment. (Data are expressed as mean ± SD; n = 6 mice/group, 2-way ANOVA). B, DSS increased VDR^-/- mice intestine permeability. Mice were treated with DSS for 3 days, and Dextran-3000 was gavaged (50 mg/g mouse). Four hours later, mouse blood samples were collected for fluorescence intensity measurement. (Data are expressed as mean ± SD; n = 6, 1-way ANOVA test; *P < 0.05). C, Claudin-2 protein robustly increased in VDR^-/- mice intestinal epithelial cells infected with Salmonella. Increased folds of internalized Claudin-2 expression after Salmonella treatment compared with control. (Data are expressed as mean ± SD; n = 3, Student t test; *P < 0.05). D, Location and quantification of Claudin-2 protein in colons 3 days post-treatment with DSS. Data are from a single experiment and are representative of 5 mice per group. E, IL-6 and TNF-α in the serum were significantly higher in DSS-treated VDR^-/- mice compared with VDR^+/- mice. (Data are expressed as mean ± SD; n = 5, 1-way ANOVA test; *P < 0.05). F, IL-6 and TNF-α mRNA levels in the intestine were significantly higher in DSS-treated VDR^-/- mice compared with VDR^+/- mice. (Data are expressed as mean ± SD; n = 5, 1-way ANOVA test; *P < 0.05).

Enhanced Claudin-2 Promoter Activity Through the Binding Sites of NF-κB and STAT

Because VDR is a transcriptional factor, we reason that VDR is the upstream regulator in linking barrier functions in the experimental colitis model and infection. Absence of VDR leads to activation of NF-κB and higher risk of chronic inflammation.^{59, 60} We further hypothesize that a robust increase of Claudin-2 in inflamed intestinal epithelial cells may be due to lacking VDR regulation at the transcriptional level, thus allowing the inflammatory cytokines to take over. To study the molecular mechanism of Claudin-2 gene regulation, we mutated the promoter of Claudin-2 at the Cdx, NF-kB, or STAT binding sites (Fig. 5A). In human HCT116 cells, we transfected the plasmid with wild type (WT), ∆NF-kB, ∆STAT, or ∆Cdx, then colonized the cells with Salmonella. Our data showed that ∆NF-kB and ∆STAT led to significantly lower activity of Claudin-2 promoter, whereas ∆Cdx did not change the activity of Claudin-2 promoter (Fig. 5B). In CaCO2 cells, we had similar observations (Fig. 5C). Therefore, we identified the enhanced Claudin-2 promoter activity through the binding sites of NF-κB and STAT.

FIGURE 5.

Salmonella enhanced Claudin-2 promoter activity through the binding sites of NF-κB and STAT. A, A schematic representation of transcriptional binding sites in the WT Claudin-2 promoter and deletion mutants. B, WT or mutant Claudin-2 reporter plasmids were transfected into HCT116 and C, CaCO2 cells and then colonized with Salmonella. Luciferase activity was measured in the cell monolayers colonized with Salmonella 30 minutes and then incubated with DMEM for 30 minutes. Dual luciferase assays were performed and firefly luciferase activity was normalized to renilla luciferase activity. (Data are expressed as mean ± SD; n = 3, 1-way ANOVA test; *P < 0.05).

Claudin-2 Protein Expression Was Upregulated in VDR^-/- Cells After Salmonella Infection Through the NF-κB and STAT3 Pathways

To study the functional regulation for enhanced Claudin-2 through the NF-κB and STAT pathways in the Salmonella infection, we treated cells with various inhibitors (Fig. 6). When VDR^-/- cells were treated with both NF-κB and STAT3 inhibitors, there was a significant suppression in Salmonella-induced Claudin-2 expression in VDR^-/- MEF cells, which was also observed in VDR^+/- MEF cells colonized with Salmonella (Fig. 6A). The NF-κB or STAT3 inhibitors could abolish the enhanced Claudin-2 in Salmonella-infectedVDR^+/- MEF cells but not in Salmonella-infected VDR^-/- MEF cells (Fig. S1). Salmonella-induced Claudin-2 was suppressed by blocking the stress-activated protein kinase/jun-amino-terminal kinase (SAPK/JNK) pathways.²⁴ After the treatment of VDR^-/- and VDR^+/- cells with SAPK/JNK inhibitor, we observed that Salmonella-induced upregulated Claudin-2 expression has been abolished in both VDR^+/- and VDR^-/- MEF cells (Fig. 6B).

FIGURE 6.

Claudin-2 protein expression was up-regulated in Salmonella-infected VDR^-/- cells in NF-κB and STAT3 pathways. A, Claudin-2 protein expression after Salmonella infection can be inhibited by combined NF-κB/STAT3 inhibitors in VDR^+/- and VDR^-/- MEF cells. MEF cells were first pretreated with 20 µM of stattic and 10 µM of BAY11-7085 for 2 hours, then colonized with Salmonella for 30 minutes, then followed by 2-hour incubation of 20 µM of stattic and 10 µM of BAY11-7085. (Data are expressed as mean ± SD; n = 3, 1-way ANOVA test; *P < 0.05). B, Claudin-2 protein expression after Salmonella infection can be inhibited by SAPK/JNK inhibitor SP600125 in VDR^+/- and VDR^-/- MEF cells. MEF cells were first pretreated with 50 µM of SP600125 for 1 hour, then colonized with Salmonella for 30 minutes, and then incubated for 2 hours of SP600125 at the concentration of 50 µM. (Data are expressed as mean ± SD; n = 3, 1-way ANOVA test; *P < 0.05).

We did coknockdown of VDR and Claudin-2 in intestinal epithelial SKCO15 cells (Fig. S2). We found that the forceful decreased expression of Claudin-2 partly restored the epithelial defects in VDR-depleted epithelial cells infected with Salmonella. We also checked p-Iκβα expression, an indicator of NF-κB activation. Western blot data showed that cells with VDR-deficient and reduced expression of Claudin-2 had less p-Iκβα post-infection compared with cells without coknockdown (Fig. S2). These data indicate less inflammatory response to Salmonella infection when VDR and Claudin-2 were both reduced.

Reduced VDR Expression Is Associated With Enhanced Claudin-2 in Ulcerative Colitis Patients

In inflamed intestine of ulcerative colitis patients, VDR expression is low (Fig. 7A). We showed that the increased Claudin-2 was associated with the reduction of intestinal VDR in human IBD intestine (Fig. 7B). Our study included 10 normal cases and 16 cases with ulcerative colitis (UC). Compared with normal intestines, UC patients’ intestines had a statistically significantly lower VDR and higher Claudin-2 proteins expression (Fig. 7C). We further examined whether cytokines recapitulate the effects of Salmonella. The SKCO15 cells were transfected with VDR-siRNA followed by TNF-alpha or IFN-gamma treatment (Fig S3). We found that both TNF-alpha and IFN-gamma could trigger higher Claudin-2 expression in the VDR-deficient epithelial cells.

FIGURE 7.

Lack of intestinal VDR led to hyperfunction of Claudin-2 in inflammatory responses. A, VDR IHC staining in normal and UC human colon. Images were representative of experiments that we carried out in triplicates (normal, n = 10; UC, n = 16). B, Claudin-2 staining in normal and UC human colon. Images were representative of experiments that we carried out in triplicates (normal, n = 10; UC, n = 16). C, Compared with normal intestines, UC patients’ intestines had a statistically significantly lower VDR and higher Claudin-2 proteins expression. (Data are expressed as mean ± SD; normal, n = 10; UC, n = 16; Student t tes; *P < 0.05). D, A working model of VDR in intestinal inflammation and tight junctions. In inflamed intestine of ulcerative colitis patients, VDR expression was low and Claudin-2 was high. Inflammatory cytokines (eg, IL-6 and TNF-α), induced by bacterial infection and chronic inflammation, were significantly higher in the intestine. These cytokines take over to further enhance Claudin-2 in absence of VDR.

DISCUSSION

We demonstrate that lack of intestinal VDR leads to hyperfunction of Claudin-2 in inflammatory responses. There was a significant increase of Claudin-2 in inflammatory cells lacking VDR. In a Salmonella-colitis model and a DSS-induced colitis model, VDR deletion consequences led to more severe leakage of intestine and inflammation in intestine of VDR^-/- mice. Importantly, observation in experimental colitis models was consistent with dysregulation of VDR in inflamed human intestine with robustly increased Claudin-2. Functionally, we find an increase of Claudin-2 in inflammatory intestinal epithelial cells (IECs) lacking VDR. Increased Claudin-2 confers more severe leakage, enhanced permeability, and inflammation in VDR^-/- intestine. Inflammatory cytokines in VDR^-/- cells further increase Claudin-2 post-infection. Mechanistically, we identified the enhanced Claudin-2 promoter activity through the binding sites of NF-κB and STAT in VDR^-/- cells. This study highlights an important mechanism for VDR regulation of Claudin-2 critical to intestinal homeostasis.

Vitamin D receptor target genes include antimicrobial peptide (AMP) cathelicidin precursor (LL-37)⁶¹ and β defensin.⁶² We have identified that Claudin-2⁴¹ and autophagy regulator ATG16L1²⁶,⁶³ are VDR target genes. We further study the physiological role and mechanism of intestinal epithelial VDR as an upstream regulator of TJs. Here, we showed without any stimulation that Claudin-2 is weak by immunistaining and significantly decreased in colon of VDR^-/- mice. However, Salmonella infection enhanced Claudin-2 in VDR^-/- mice. Claudin-2 staining is significantly enhanced which may explain the observed leaky and bacterial translocation in the Salmonella-infected mice. These data suggest that TJs without stimulation may be very different from those with infection. Interestingly, there is no change of Claudin-3 and Claudin-7 in the WT mice and VDR^-/- mice with or without Salmonella infection, indicating the regulatory role of VDR is selectively on certain junction proteins in intestinal inflamamtion.

The vitamin D receptor has multiple critical functions in regulating innate and adaptive immunity, intestinal homeostasis, host response to invasive pathogens and commensal bacteria, and tight junction structure.^{40, 61, 64–70} Severely disrupted TJs and increased permeability were reported in the DSS-treated VDR^-/- colonic epithelial cells in vivo.⁴⁰ There is a specific regulatory role of VDR on Claudin-2 under both healthy and infectious states. The pathobiological importance of the VDR regulation of Claudin-2 is very interesting. Claudin-2 is enhanced in the inflamed gut in patients with IBD.^{15, 18} In the UC-activated inflamed intestine, VDR is low and Claudin-2 is enhanced, suggesting the other regulators that take over and enhance the level of Claudin-2.

We dissected the transcriptional regulation of Claudin-2 under inflamed states in the current study. Based on others’ studies and our studies on the promoter of Claudin-2,^{4, 16, 56} we know that multiple factors contribute to the upregulation of Claudin-2 at the transcriptional level. TNF-α and IL-1β also contribute to elevated Claudin-2 in vitro.^{57, 71} Interleukin-6 enhances Claudin-2 promoter activity in a Cdx-binding, site-dependent manner.⁵⁶ Interleukin-22 induces Claudin-2 upregulation to drive diarrhea and pathogen clearance.⁷² Recent evidence also suggests that Claudin-2 is relevant to neoplastic transformation and growth via epidermal growth factor receptor (EGFR) transactivation.⁷³ Absence of VDR in the intestine leads to activation of NF-κB and higher risk of chronic inflammation.^{59, 60} Our study has demonstrated that we need to inhibit both NF-kB and STAT3 promoters to abolish the Salmonella-induced inflammation in the VDR^-/- cells. Thus, chronic inflammation enhances the expression of Claudin-2.

In the current study, the regulatory role of intestinal VDR has not been studied in the tissue-specific knockout mouse models. We only investigated the samples from UC patients, not Crohn’s disease patients. We plan to study both UC and CD samples in the future. Intestinal permeability and inflammation are regulated by various proteins (eg, Junctional Adhesion Molecule [JAM-A] and myosin IIA).^{48, 74} It will be very interesting to identify other regulators that modulate the level of Claudin-2 and tissue barriers in inflamed VDR-deficient tissues in human IBD.

In summary, our study provides new insights into VDR on TJs in infected and inflamed intestine. Lacking VDR may lead to the hyperfunction of Claudin-2 and increase permeability in the inflammatory state, as we see in the murine colitis models and human IBD samples (Fig. 7D). Our studies highlight the complex role of VDR in bacterial infection and chronic inflammation.^{71, 75} It will provide ideas for developing new therapeutic targets by enhancing function of intestinal epithelial VDR for alleviating infection and chronic inflammation.

Author contributions: YZ, RL, and YX contributed to the acquisition, analysis and interpretation of data, drafting the manuscript, and statistical analysis. DZ, EP, and ECC contributed to technical or material support and drafting the manuscript. JS contributed to the study concept and design, obtaining funding, supervising the study, analysis and interpretation of data, and writing the manuscript.

Supported by the NIDDK/National Institutes of Health grants R01 DK105118, R01DK114126, and DOD BC160450P1 given to JS.