FoxO1 regulates TLR4/MyD88/MD2‐NF‐κB inflammatory signalling in mucosal barrier injury of inflammatory bowel disease

Abstract In this study, FoxO1 transgenic mice (transgenic, FoxO1‐Tg) and C57BL/6 wild‐type (wild‐type, FoxO1‐WT) mice were used to establish chronic colitis by drinking water containing dextran sulphate sodium (DSS). Afterwards, we observed the life changes in mice and assessed the pathological changes by H&E tissue staining. In addition, the TLR4/MyD88/MD2‐NF‐κB inflammatory signals were detected. As a result, under DSS treatment, the activation level of TLR4/MyD88/MD2‐NF‐κB inflammatory signal was higher in FoxO1‐Tg mice than that in FoxO1‐WT mice. Meanwhile, the intestinal mucosal tissue damage was more severe, the down‐regulation of tight junction protein level was more significant and the life quality was decreased to a higher degree in FoxO1‐Tg mice compared with those in FoxO1‐WT mice. Caco‐2 cells were used to mimic the intestinal mucosal barrier model for in vitro assays. In addition, lentiviral packaging FoxO1 overexpressing plasmid was transfected into Caco‐2 cells for FoxO1 overexpression. TNF‐α intervention was performed for intestinal mucosal inflammatory response model. Consequently, the down‐regulation of FoxO1 inhibited the activation of TLR4/MyD88/MD2‐NF‐κB inflammatory signal, decreased the mucosal barrier permeability and up‐regulated the expression of tight junction protein. By contrast, the overexpression of FoxO1 increased the mucosal barrier permeability and down‐regulated the level of tight junction protein.

The occurrence of IBD mainly damages the local mucosal structure, which is characterized by the loss of tight junction proteins, the damage of intestinal epithelial cells and the destruction of villus structure, among which, local inflammatory response plays an important role. 3,4 Toll-like receptors (TLRs) are the main receptors that mediate inflammatory responses. After stimulation by HMGB1, LPS,etc, TLRs can further transmit signals to activate downstream NF-κB, thereby promoting the expression of inflammatory factors such as interleukin (IL) and tumour necrosis factor (TNF) and interferon (IFN). Among them, TLR4 is a relatively critical receptor, and TLR4, MD2 and MyD88 combine to form a complex, 5 playing an important role in the early stage of inflammatory signals. However, the regulation mechanism of this signal in IBD has not been clarified at present.
Transcription factors (TFs) are a group of proteins that regulate the transcription process and regulate protein expression through transcriptional and translational modulation of mRNA, further regulating protein functions. 6,7 A variety of TFs have been found to be involved in UC. Our team has found that the expression of FoxO1, a transcription factor, is increased in the intestinal tissue of patients with IBD. FoxO1, a member of the Fox family, is involved in apoptosis, stress, DNA damage, metabolism and etc Moreover, FoxO1 has been found to promote the carcinogenesis and tumour progression in a variety of tumours.
However, to date, the role of FoxO1 has not been reported in IBD. The expression of FoxO1 and TLR4 is high in IBD. Previous studies have also reported the association between TLR4 and FoxO1 in anti-inflammatory response. 8 However, whether there is a regulatory relationship between them has not been reported. As IBD is a risk factor of colorectal cancer, we cautiously speculate that FoxO1 might play a role in the transformation from colitis to colorectal cancer. Herein, in this study, we mainly investigated the role of FoxO1 in IBD. All patients underwent colonoscopy, followed by acquisition of colon tissue for biopsy. All patients signed the written informed consent. Intestinal protein from patients was extracted to detect the protein expression of FoxO1 by ELISA kit. Moreover, paraffin-embedded tissue was subjected to immunohistochemical staining to detect the expression of FoxO1.

| Establishment and grouping of animal models
Transgenic mice (FMS-FoxO1-GFP-transgenic, Tg mice, Saiye Biotechnology Co., Ltd.) with high expression of FoxO1 and wild-type C57BL/6 mice (wild-type, WT mice) were fed under the same environment. After adaptive feeding for one week, WT and Tg mice were randomly divided into FoxO1-WT, FoxO1-Tg, FoxO1-WT-DSS and FoxO1-Tg-DSS groups (N = 5 in each group). Chronic colitis model was constructed by daily feeding of 2.5% DSS in drinking water at 1-5, 8-12, 15-19 and 22-26 days, whereas distilled water was supplied for the rest of the time, which lasted for a total of four cycles. DAI scores and mouse weight loss rate were measured on days 1, 3 and 5 of each cycle. Mice were killed on day 29 by carbon dioxide asphyxiation.
Afterwards, mice were placed in the supine position, disinfected with iodine, and the abdominal skin and muscle layer were quickly cut off to separate the terminal ileum until to the rectum. The resected tissue was washed with PBS for three times, fixed with 4% paraformaldehyde and embedded with paraffin for further analysis.

| Evaluation of bodyweight and DAI score in mice with chronic colitis
The DAI score was used to evaluate the pathological state of mice. In addition, the mental state, activity, hair gloss, appetite, faecal traits and bodyweight of mice were observed daily. According to the criteria of DAI score (shown in Table 1), the bodyweight, faecal traits and occult blood of mice were assessed and recorded on days 1, 3 and 5 of each cycle. In terms of bodyweight loss rate, bodyweight of mice was measured on days 1, 3 and 5 of each cycle, followed by calculation of ratio between the bodyweight and initial bodyweight. After the mice were killed, the colon length was determined (shown in cm).
Moreover, the gross morphological score of colonic injury in mice was performed according to the relevant criteria (shown in Table 2).

| Histopathological examination of mouse intestinal tissue (H&E staining)
Paraffin-embedded colon tissue was serially cut into 4-μm-thick sections, dewaxed with xylene, dehydrated with gradient concentrations of 100%, 95%, 80% ethanol, rinsed with tap water for 2 minutes, stained with haematoxylin for 3 minutes, rinsed with tap water for 2 minutes, treated with 1% hydrochloric acid ethanol for 2 seconds, rinsed with tap water for 2 minutes, reacted with 1% ammonia for 20 seconds, stained with 0.5% eosin ethanol for 10 seconds, dehydrated with gradient ethanol, treated with xylene for transplantation and sealed with neutral gum. Finally, the pathological changes in intestinal tissue were observed under light microscope.

| Detection of tight junction protein expression by immunohistochemistry (IHC)
The paraffin-embedded intestinal tissue samples were cut into 4-μmthick sections, baked at 60°C for 2 hours, dewaxed in xylene 3 times (5 minutes each), immersed in absolute ethanol for 5 minutes, immersed with 95% ethanol 2 times (2 minutes each), immersed in 85% ethanol once for 2 minutes; rinsed with tap water for 5 minutes and rinsed with distilled water for 3 minutes. Afterwards, antigen retrieval was performed in 0.01 mol/L citrate buffer (pH = 6.0) by microwave at 98°C for 20 minutes, followed by cooling down at room temperature for 30 minutes and then rinsing with distilled water. The sections were incubated with 3% hydrogen peroxide to eliminate endogenous peroxidase at room temperature for 10 minutes. Afterwards, sections were blocked with 2% bovine serum albumin (BSA) at 37°C for 30 minutes to prevent the non-specific binding between antigen and antibody. After discarding BSA, the sections were incubated with proper primary antibody at 37°C for 2 hours. The primary antibodies

| Detection of colonic mucosal permeability by FITC-D
After relevant intervention, mice were fasted and free of water 4 hours before killing. Fluorescein isothiocyanate-dextran (SM Biochemicals, FITC-D, MW 4000) was administered orally (60 mg FITC-D /100 g), followed by collection of serum. The fluorescence density of each sample was measured by fluorescence spectrophotometer, and the serum concentration of FITC-D was examined as well.

| Relative expression levels of proteins by Western blot
In total, 100 mg of colon tissue was cut into pieces using sterile surgical scissors. For cell assay, cells were collected, washed with

| Construction of cell lines with FoxO1 overexpression/down-regulation
Small interfering RNA (siRNA) was used to silence

| Determination of transepithelial electrical resistance (TEER)
The cells were seeded in Transwell chamber at a density of 4 × 10 5 cells/per well (200 μL per well). After inoculation for 24 hours, the medium was changed, which was thereafter changed every other day to monitor TEER. TEER was measured by Millicell resistance meter. Briefly, the two electrodes of the resistance meter were inserted into the top side and basal side of the chamber and immersed within the liquid to measure TERR. The assay was conducted at 37°C, and three points from different directions of each Transwell chamber were taken, and the assay was conducted in triplicate.
The resistance value was expressed as ohm/cm 2 (Ω/cm 2 ). In consideration of the intrinsic electrical resistance of the Transwell membrane, the standard TEER value = (measured value − blank control value)/0.33 cm 2 .

| Statistical analysis
Measurement data were shown as mean ± standard deviation ( ± s).
One-way ANOVA was used for comparison among multiple groups, and LSD as post hoc, SNK test was utilized for comparison between groups. SPSS 18.0 software was used for statistical analysis. A P < .05 was considered as statistical significance.  Figure 1).

| Effects of FoxO1 by evaluating bodyweight, DAI score and pathological condition of mice with chronic colitis
During the experiment, FoxO1-WT-Con and FoxO1-Tg-Con mice were good in general condition, with normal stool and gradually increased bodyweight. After DSS intervention, the hair was dull, the amount of drinking water was decreased, the frequency of defecation was increased, with the appearance of mucopurulent bloody stool, and the bodyweight was significantly decreased.
Moreover, the bodyweight loss was significantly more obvious in the FoxO1-Tg-DSS group than that in the FoxO1-WT-DSS group.
After DSS intervention, the expression of FoxO1 was significantly elevated in FoxO1-Tg-DSS and FoxO1-WT-DSS groups (shown in Figure 2).

| Effects of FoxO1 on the expression of tight junction protein in mice with chronic colitis
The expression of tight junction proteins, including ZO-1 and occludin in the intestinal mucosa of mice was relatively higher in FoxO1-WT-Con and FoxO1-Tg-Con groups, without significant difference between the two groups (P < .05). After DSS intervention and FoxO1 overexpression, the expression of tight junction proteins was down-regulated, which was more obvious following FoxO1 overexpression. As shown in Figure 3, the expression levels

| Effects of FoxO1 on intestinal mucosal villus structure, permeability and expression of inflammatory signals in mice with chronic colitis
Under TEM, the intestinal mucosal villus structure was neatly arranged in FoxO1-WT-Con and FoxO1-Tg-Con mice, with clear structure and no change in the villus. After DSS intervention, the intestinal mucosa villus structure was significantly changed in F I G U R E 2 Effects of FoxO1 by evaluating bodyweight, DAI score and pathological condition of mice with chronic colitis. A, The intestinal histopathological scores in mice showed that the pathological scores were 0 in FoxO1-WT-Con and FoxO1-Tg-Con groups. However, after DSS intervention, the pathological scores were significantly higher in FoxO1-Tg-DSS group than those of FoxO1-WT-DSS group. Comparison between groups, * P<.05. B, The change in intestinal tissue length in mice: The length of intestinal tissue was significantly lower in FoxO1-Tg-DSS group than that in FoxO1-WT-DSS group. Comparison between groups, * P<.05. C, DAI scores in mice showed no significant changes in FoxO1-WT-Con and FoxO1-Tg-Con mice during the experiment, whereas DAI scores were significantly changed in FoxO1-Tg-DSS and FoxO1-WT-DSS groups. Comparison with FoxO1-WT-DSS group at the same time-point, * P<.05. D, Bodyweight changes of mice: The bodyweight was gradually increased in FoxO1-WT-Con and FoxO1-Tg-Con mice under normal feeding, without significant difference. However, the bodyweight was decreased in FoxO1-Tg-DSS and FoxO1-WT-DSS groups. Comparison with the FoxO1-WT-DSS group at the same time-point, * P<.05. E, HE staining showed that the intestinal mucosa epithelium was intact, and the intestinal gland composed of lamina propria and mucosa muscle layer was arranged regularly in FoxO1-WT-Con group under light microscope. However, the colonic mucosa was defective, with decreased goblet cells and destructive or even disappeared gland, and large number of lymphocytes were infiltrated in the submucosa and even the muscle layer in FoxO1-WT-DSS group, which was more severe in FoxO1-Tg-DSS group than FoxO1-WT-DSS group.

| Cell permeability changes in inflammatory models after down-regulation/overexpression of FoxO1 in Caco-2 cells
After silencing FoxO1 by siRNA (the protein level of FoxO1 was down-regulated), cells were divided into Control, siRNA-Negative However, the cell permeability, resistance down-regulation and apoptotic rate of the siRNA-FoxO1 group were significantly different from those of the other two groups. Down-regulation of FoxO1 could significantly inhibit cell permeability and resistance (shown in Figure 5).
The lentiviral packaging pcDNA-FoxO1 plasmid was transfected into Caco-2 cells to significantly elevate FoxO1 expression, and cells were further divided into Control, pcDNA-Negative and pcDNA-FoxO1 groups. Inflammatory model was constructed by TNF-α, which caused significant apoptosis. The apoptotic rate was significantly up-regulated, cell permeability was increased, whereas resistance was down-regulated in pcDNA-FoxO1 cells,

F I G U R E 4
Effects of FoxO1 on intestinal mucosal villus structure, permeability and expression of inflammatory signals in mice with chronic colitis. A, Changes in the intestinal mucosal microvillus structure of mice. The intestinal mucosal microvillus was neatly arranged in FoxO1-WT-Con and FoxO1-Tg-Con mice, with clear structure. The intestinal mucosa microvillus structure was significantly changed in FoxO1-WT-DSS and FoxO1-Tg-DSS groups, with obvious loss of microvillus structure. In addition, the lesion degree was significantly more severe in FoxO1-Tg-DSS group than that in FoxO1-WT-DSS group. B, FITC-D permeability assay revealed that the permeability was significantly higher in FoxO1-WT-DSS and FoxO1-Tg-DSS groups than that in Con group. Comparison between groups * P<.05. C,D, Western blot assay for TIL4/MyD88/MD2-NF-kB signalling. The activation levels of TIL4/MyD88/MD2-NF-kB signalling were lower in FoxO1-WT-Con and FoxO1-Tg-Con groups. After DSS intervention, the levels of key proteins, TLR4, MyD88 and MD2 were up-regulated. Comparison between groups, * P<.05 compared with those in Control and pcDNA-Negative groups (shown in Figure 6).

| Down-regulation/overexpression of FoxO1 on inflammatory signalling and tight junction protein expression in inflammatory models in Caco-2 cells
After TNF-α intervention, the TIL4/MyD88/MD2-NF-kB signal was significantly activated, with significantly up-regulated level of p-p65, and other key proteins, including TLR4, MyD88 and MD2. After  Figure 7).

| D ISCUSS I ON
Inflammatory bowel disease is an intestinal inflammatory disease including CD and UC. CD, a chronic granulomatous lesion, can affect the entire colon and terminal ileum, 9 whereas UC is a chronic, non-specific inflammation that mainly affects the colonic mucosa and submucosa. 10 Comparison between groups, * P<.05. D,E, After TNF-α intervention, the apoptotic level was not significantly different between the Control group and the siRNA-Negative group, whereas the apoptotic rate was significantly increased in the siRNA-FoxO1 group. Comparison between groups, * P<.05. F, Cell resistance assay showed no significant difference between the Control group and the siRNA-Negative group, whereas cell resistance was significantly down-regulated in the siRNA-FoxO1 group. Comparison between groups, * P<.05. G, The permeability change of FITC-D was not significantly different between the Control group and the siRNA-Negative group, whereas the permeability was significantly up-regulated in the siRNA-FoxO1 group. Comparison between groups, * P<.05 Similarly, in this study, we also found that overexpression of FoxO1 could lead to further down-regulation of tight junction proteins, resulting in mucosal damage. At the cellular level, Caco-2 cells are commonly used to construct intestinal mucosal barrier model. After overexpression of FoxO1, TLR4 signalling was activated and permeability was increased in the inflammatory response model. By contrast, after down-regulation of FoxO1, the activation level of TLR4 signal was down-regulated, and the permeability was decreased.
These results indicate that FoxO1 can regulate the inflammatory response of IBD by modulating TIL4/MyD88/MD2-NF-kB signalling.
However, whether FoxO1 predominantly regulates tight junction barrier or inflammatory signalling is uncertain. It is also not clear which signalling comes first during IBD. We speculate that FoxO1, as a transcription factor, may promote the transcription and expression of one protein in TIL4/MyD88/MD2-NF-kB signal, thus stimulating the activation of the whole signal. However, the specific protein has not been found, which is also the direction of future research.
To sum up, in this study, we report that the transcription factor The down-regulation of FoxO1 inhibited the down-regulation of ZO-1 and occludin proteins, whereas after overexpression of FoxO1, the down-regulation of ZO-1 and occludin proteins was more obvious. Comparison between groups, * P<.05 mechanism of immunology is not involved in this study, which is also the limitation of this study. The immunoregulation and mechanism of FoxO1 in inflammatory bowel disease is the future research content of our group, which is also necessary to further explain the role of FoxO1.

CO N FLI C T O F I NTE R E S T
No Competing interests.

E TH I C A L A PPROVA L
The study was approved by Ethics Committee.

CO N S E NT FO R PU B LI C ATI O N
All authors approved the publication of the article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data are available.