Repression of TXNIP–NLRP3 axis restores intestinal barrier function via inhibition of myeloperoxidase activity and oxidative stress in nonalcoholic steatohepatitis

2.1 Ethics statement

The study was conducted with the approval of the Animal Ethics Committee of China-Japan Friendship Hospital. All animal experiments were performed in line with the approval of the Guide for the Care and Use of Laboratory Animal by International Committees.

2.2 Model establishment and high-fat and high-fructose diet inducement

Healthy 8–10-week-old male C57BL/6J mice (weight: 21 ± 3 g) in clean grade were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Specific TXNIP knockout mice were obtained through the following procedures. First, TXNIP^flox/flox and Tie2-Cre/⁺ mice (purchased from Jackson Laboratory, Bar Harbor, Maine) were hybridized to obtain TXNIP^flox/floxTie2-Cre/⁺ mice. TXNIP^flox/floxTie2-Cre/⁺ mice were hybridized with TXNIP^flox/flox mice and their generation was specific TXNIP knockout mice (TxNIP^−/−) model. C57BL/6 wild-type (WT) mice and TXNIP^−/− mice were fed with common forage with free access to food and water. After 3-day of adaptive feeding, the mice in the normal group (n = 6) were still fed with common forage, whereas mice in the model group were performed with intragastric administration of high-fat and high-fructose emulsion (10% cholesterol, 40% corn oil, 15% sucrose, 2% protein, 1% bile salt, 3.1% propylene glycol, and 3.6% Twain-80) with a dosage based on 1 ml/100 g every morning for 20 weeks (stopped for two days after every five consecutive days). Next, to clarify the roles of specific TXNIP knockout in the MPO activity and ROS level in the intestinal mucosa of mice, six WT mice and six TXNIP^−/− mice were randomly selected and conducted with intraperitoneal injection of 2,2,6,6-tetramethyl piperidine-1-oxyl (TEMPO, 1 mg•kg•day) at the same time every day (Ding, Liu, Bi, & Zhang, 2017), ad libitum water for 4 weeks, as the TEMPO group. To study the effect of specific TXNIP knockout on the intestinal microflora in mice, six WT mice and six TXNIP^−/− mice were randomly selected, treated with intraperitoneal injection of equal dose of phosphate buffer solution (PBS) and fed with water dissolved with compound antibiotics (ampicillin 1 g/ml, vancomycin 0.5 g/ml, neomycin 1 g/ml, and metronidazole 1 g/ml) for 4 weeks successively (Ochoa-Reparaz et al., 2009), as the antibiotic group. Subsequently, six WT mice and six TXNIP^−/− mice were randomly selected and treated with intraperitoneal injection of equal dose of PBS and ad libitum to water, as the model mice. Following 24 weeks, mice were administrated with fluorescein isothiocyanate (FITC)-dextran (40 mg/100 g) through stomach tube when fasting, and injected with 200 μl of fluorescent staining ROStar intraperitoneally. Four hours later, mice were killed using CO₂ and then the abdomen was exposed for liver tissues and intestinal mucosa collection, 1/3 of the tissues was fixed in 4% formaldehyde and the remaining was stored in a freezer at − 70°C for further usage.

2.3 Intestinal epithelial cell (IEC) treatment

Intestinal mucosa tissues from WT and TXNIP^−/− mice were cut into pieces (3 mm²), washed with Hanks solution and incubated with Dispase I (2.4 U/ml) at 37°C for 30 min. After cell suspension was filtered via a 74-μm nylon mesh, cells were collected, washed, and centrifuged. Then cells were incubated with the addition of phycoerythrin (PE)-labeled-monoclonal antibody anti-mice CD326 (Ep-CAM, a specific antigen for epithelial cells) at 4°C for 30 min, rinsed with PBS and suspended. Then intestinal epithelial cells (IECs) were screened out by a flow cytometer (MOFLO S2500). The living cells were separated from particles and fragments by FSC, and then divided into CD326 negative and CD326 positive cells by the activity of PE. The positive group was small intestinal epithelial cells (Alenghat et al., 2013). Next, cells were adjusted to the density of 5 × 10⁵, cultured with Dulbecco's modified Eagle's Medium (DMEM) containing 20% fetal bovine serum (FBS), added with 1% combination of penicillin and streptomycin, reacted with 0.25% trypsin containing ethylene diaminetetraacetic acid (EDTA), subcultured at the ratio of 1: 2 and cultured in an incubator containing 5% CO₂ at 37°C. After 24-hr of subculture, the medium was replaced. In the following days, the medium was replaced once every day. The cells were subcultured once every 5–6 days. IECs in the logarithmic growth were given different stimulation and starved for 12 hr before experiment with serum-free medium. IECs from WT and TXNIP^−/− mice were assigned into the control group (without any treatment) and the lipopolysaccharide (LPS) group (48-hr culture with the addition of LPS [10 μg/ml, Sigma, Deisenhofen, German] at the 5th day of culture; Evans, Flint, Somers, Eyden, & Potten, 1992). After that, the supernatant in each group was collected, concentrated to about 1/20 using a concentrated filter, and then the MPO activity was determined with a colorimetry in accordance with the instructions of the MPO kit. IECs in each group were added with dichlorofluorescin diacetate (DCFH-DA) at the ratio of 1:1,000, with the final concentration of 10 μmol/l, then incubated in an incubator at 37°C for 20 min, and washed three times with serum-free culture solution. Following, the intensity of fluorescence was detected at the extension wavelength of 488 nm and the emission wavelength of 525 nm using a fluorometer.

2.4 Cell transfection and grouping

IECs from TXNIP^−/− mice in the logarithmic growth were seeded into 6-well plates and adjusted to a density of 2 × 10⁵ cells/well. Two hours before the transfection, serum-free DMEM was prepared and then each well was incubated with the replacement of 2 ml of serum-free Eagle's minimum essential medium (MEM; Gibco, Carlsbad, California) without penicillin/streptomycin. After 2 hr, cells were preheated in a warm oven with 1.5 ml of opti-MEM. Empty vector and pcDNA-TxNIP (Shanghai Genechem Co., Ltd., Shanghai, China), Opti-MEM and Lipofectamine® 2000 transfection reagent (Invitrogen Inc., Carlsbad, California) were mixed well and allowed to stand for 5 min. Afterward, the mixture was added into the medium for IECs from TXNIP^−/− mice and mixed evenly. After 24-hr culture, the medium was replaced by a fresh one. After the transfection of 48 hr, cells were treated with trypsin and then collected for the following experiments (Li et al., 2017). Next, cells were classified into the TXNIP^−/− group (transfection with empty plasmids, as negative control) and the pcDNA-TxNIP group (transfection with TXNIP overexpression plasmids). IECs from WT mice in the logarithmic growth and transfected cells were fasted for 12 hr before experiment and the cells were divided into the WT group, the TXNIP^−/− group and the pcDNA-TxNIP group. At the 5th day of culture, cells were added with 10 μg/ml LPS (Sigma, Deisenhofen, German) and incubated for 48 hr (Evans et al., 1992). Successively, the MPO activity and ROS level in IECs in each group were determined.

2.5 RNA isolation and quantitation

Total RNA was extracted using the Trizol extraction kit (16096020, Thermo Fisher Scientific Inc., Waltham, Massachusetts) and then 5 µg of RNA was prepared and reversely transcribed into complementary DNA (cDNA) in accordance with the instructions of the reverse-transcription quantitative polymerase chain reaction (RT-qPCR) kit (ABI Company, Oyster Bay, New York). The reaction system was 25 µl, including 300 ng of cDNA, 1 × PCR buffer, deoxyribonucleotide triphosphates (dNTPs; 200 μmol/l), forward primers (80 pmol/l) and reverse primers (80 pmol/l), and 0.5 U Taq enzyme (S10118, Shanghai Yuanye Biotechnology Co., Ltd., Shanghai, China). The reaction conditions consisted of predenaturation at 94°C for 5 min, followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 54.5°C for 30 s and extension at 72°C for 30 s, and a final elongation at 72°C for 10 min. The product was preserved at 4°C. The primer sequences of TXNIP, NLRP3, ZO-1, occludin, TNF-α, IL-1β, and β-actin are shown in Table 1. β-actin was used as the internal reference of TXNIP, NLRP3, ZO-1, occludin, TNF-α, IL-1β. The 2^−ΔΔCt represented the ratio of the expression of the target gene in the experiment group and the control group. The formula was as follows: ΔΔCT = ΔCt _{experimental group} − ΔCt _{control group}, where ΔCt = Ct _miRNA − Ct _U6. Ct was the amplification cycles when the real-time fluorescence intensity of the reaction reached the set threshold, and the amplification was in the logarithmic growth. The experiment was independently repeated three times.

2.6 Western blot analysis

Obtained protein samples were separated with sodium dodecyl sulfate-polyacrylamide gel (100 g/l), and then transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with 5% skim milk, and then incubated with the following rabbit antimouse antibodies at 4°C overnight and washed three times with PBS at room temperature, 5 min per time. The rabbit antimouse antibodies were NLRP3 (ab214185; 1:1,000), TXNIP (ab114981, 1:1,000) and β-actin (ab8227, 1:1,000). All antibodies were purchased from Abcam (Cambridge, Massachusetts). On the following day, the membranes were washed three times with Tris-buffered saline with Tween-20 (TBST; Wuhan Boster Company, Wuhan, Hubei, China). Horseradish peroxidase (HRP)-labeled goat antirabbit IgG (ab150077, 1:5,000, Abcam) was then incubated with the membranes at room temperature for 1 hr as the secondary antibody. The membranes were washed with TBST, then immersed in a mixture of A solution and B solution with equal dose obtained from a chemiluminescence (ECL) reaction solution (BB-3501, Amersham Pharmacia Biotech, Chicago, Illinois), and exposed using a gel imaging instrument. Finally, the membranes were imaged using a Bio-Rad imaging analyzer (Bio-Rad Laboratories, Hercules, California) and analyzed with the Quantity One v4.6.2. The ratio of the gray value of the target band to β-actin band was utilized as the relative expression of the protein. The experiment was independently repeated three times.

2.7 Coimmunoprecipitation

IECs were treated with lysate buffer (p0013, Beyotime Biotechnology Co., Ltd., Shanghai, China) on ice, centrifuged for 15 min at 40,256 × g and incubated at 4°C overnight with the mixture of 500–1,000 μg of proteins and 1 μg of TXNIP antibodies. Then cells were incubated at 4°C overnight with the addition of A/G gel beads (p2006, Beyotime Biotechnology Co., Ltd., Shanghai, China), centrifuged for 20 s at 25,764 × g, followed by the removal of the supernatant. The precipitation was washed three times with radioimmunoprecipitation assay (RIPA) buffer, and the sample buffer was added and boiled. Finally, NLRP3 protein content was determined via western bolt analysis.

2.8 Immunofluorescence assay

IECs were seeded into 24-well plates which were precoated with polylysine. Upon cell adhering to walls, the culture medium was absorbed. Then cells were washed gently three times with warmed 0.01 mol/l PBS, fixed in 4% polyoxymethylene at room temperature for 30 min, and washed three times with 0.01 mol/l PBS (10 min each time). Cells were added with blocking solution (Beyotime Biotechnology Co., Ltd., Shanghai, China) at 37°C for 60 min. After the blocking solution being discarded, cells were incubated at 4°C overnight with the addition of primary antibodies, including rabbit antimouse ZO-1 (ab96587, 1:400), rabbit antimouse occludin (ab31721, 1:400), both of which were purchased from Abcam. After three times washing with 0.01 mol/l PBS (10 min each time), cells were cultured for 1 hr under the environment void of light with the addition of secondary antibodies, including Alexa Fluor 594 donkey antirabbit (A21202, 1:400, Life Technologies, Gaithersburg, Maryland) and Alexa Fluor 488 donkey antimouse (A21207, 1:400, Life Technologies). Subsequently, cells were stained with 2,4-diamadino-6-phenylindole (Beyotime Biotechnology Co., Ltd.) for 5 min at the room temperature, added with appropriated fluorescence quenching (Beyotime Biotechnology Co., Ltd.), and imaged with a Molecular Devices. Molecular Devices MetaXpress Image Acquisition and Analysis Software was used for data analysis. The experiment was independently conducted three times.

2.9 Transepithelial electrical resistance (TEER)

Next, we established an IEC monolayer model. IECs in the logarithmic growth were extracted and incubated with trypsin, prepared into cell suspension and then inoculated into Transwell chambers at a density of 5 × 10⁵ cell/well, followed by independent culture. The apical side of the chamber was 0.4 μm and the area was 1.12 cm². Apical chamber was added with 0.5 ml of cell suspension and the basolateral chamber was added with 1.5 ml of culture solution, making the liquid level of internal and external ones basically flat. The culture solution was changed every other day at the first week, afterwards, the solution was replaced every day. Following 5-day inoculation, transepithelial electrical resistance (TEER) was assessed using an EVOM2 epithelial volt-ohm meter to identify the compactness and integrity of monolayer cells.

Electrical resistance was measured from the three angels of Transwell chambers and obtained the mean values: TEER (Ω·cm²) = (R1 − R0) × A, in which R1 and R0 represented the value of electrical resistance (Ω) with cell culture well and cell-free negative control well, respectively, and A represented polycarbonate membrane area (cm²). TEER reaching 500–1,500 Ω·cm² was suggested that monolayer cells were close and complete. Three parallel controls were set in the experiment.

2.10 Reactive oxygen species (ROS) level detection

Intestinal mucosa tissues of mice were fixed in 4% polyoxymethylene for 3 days, dehydrated with 20% sucrose solution for 2 days, covered with a octenidine-gel (OCT) and then sectioned via a frozen slicer (CM1900, Leica, Germany) at the thickness of 30 μm. The sections were adherent onto the glass slide, washed three times with PBS (5 min each time), dried, sealed with glycerol and imaged under a fluorescence microscope at the extension wavelength of 490 nm and the emission wavelength of 530 nm. The Image ProPlus software was used to analyze the average optical density (OD), which referred to the relative ROS level in the intestinal mucosa.

2.11 MPO activity measurement

A total of 100 mg of intestinal mucosa tissues from mice were prepared into 5% homogenate with homogenate medium provided by the MPO measurement kit. The above procedure was processed in strict accordance with the instructions of the MPO measurement kit. MPO activity was expressed as unit of enzyme activities that per gram of intestinal mucosa contained, and recorded as U/g. The unit of enzyme activities was defined as 1 μmol H₂O₂ resolved at 25°C in one minute.

2.12 FITC-dextran permeability assay

Mice were treated with CO₂, and the abdomen was open with celiac artery blood extracted, followed by anticoagulation with heparin. Intestinal mucosa tissues were centrifuged at 1,610 × g for 10 min and the plasma was obtained for later usage. Next, 25 μl of plasma was added with 100 μl of PBS. A fluorescence spectrophotometer was adopted to determine the intensity of fluorescence at the extension wavelength of 480 nm and the emission wavelength of 520 nm. A relative curve was drawn and the sample concentration was calculated. The experiment was independently repeated three times.

2.13 Detection of intestinal flora population

After feeding with compound antibiotic, the fresh feces of mice was obtained using an aseptic method, then placed into an aseptic tube and immediately stored in a freezer at − 80°C. Next, 0.2 g of feces was added with 200 μl of Tris-EDTA (TE) buffer containing lysozyme (20 mg/ml) and mixed with a pipette. Then the feces was incubated at room temperature for 40 min, added with 400 μl of digestion buffer provided by the kit and mixed evenly using a pipette. After 3 μl of protease K treatment and even mixture, samples were preserved at 55°C for 2–3 hr, followed by a centrifugation. Then the supernatant was added with 260 μl of absolute ethanol, mixed well and transcribed into the extraction cartridge of bacterial genomic DNA. The 3-times centrifugation was followed and the liquid collected into the centrifugation tube was the bacterial genomic DNA. The reaction volume of RT-qPCR contained 10 μl of 10 × buffer, 1 μl of 2 μmol/l forward and reverse sequences, respectively, 2 μl of DNA plate, 2.5 μl of MgCl₂, 1 μl of dNTP, 0.4 μl of Taq enzyme, and brought up to 25 μl with double-distilled water. The reaction condition consisted of predenaturation at 95°C for 5 min, denaturation at 95°C for 20 s, and annealing at 50°C for prevotella and bifidobacterium, 55°C for Escherichia coli, Clostridium, and Lactobacillus for 45 s, extension at 72°C for 45 s, with a total of 40 cycles, and a final extension at 72°C for 5 min, preserving at 4°C. PCR products were processed with agarose electrophoresis, and then the target fragments were retrieved according to the instructions of the DNA product purification kit. Nucleic acid quantitation was used to detect the concentration of purified products. All retrieved products were treated with a quantitative one-step SYBR Green-based RT-qPCR system. The system software was used to automatically analyze Ct and generate standard curves based on the recoded fluorescence value. The absolute content of the bacteria was calculated by substituting the Ct of the sample into the standard curve. Full sequences of 16S rRNA of five bacteria were searched in the Genebank, which were compared with the selected primer sequences in the experiment using BLAST database (www.ncbi.nlm.nih.gov/BLAST). The standard curve is shown in Table 2.

2.14 Hematoxylin-eosin (HE) staining

Mice were treated with abdomen exposure and then the left lobe of liver was fixed with 4% paraformaldehyde, paraffin embedded routinely, sectioned (5 μm) and stained with hematoxylin-eosin (HE) according to the instructions of the HE kit (Beyotime Biotechnology Co., Ltd., Shanghai, China; Wang et al., 2017). The experiment was independently conducted three times. The remaining liver tissues were preserved in a freezer at − 70°C for the further experiments.

2.15 Biochemical parameter detection

We collected 100 μl of plasma from mice and utilized automatic biochemical analyzer (BS200, Mindray, Shenzhen, Guangdong, China) and the corollary reagent to detect the levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) in serum. The experiment was independently repeated three times.

2.16 Enzyme-linked immunosorbent assay (ELISA)

The procedure was conducted in accord with the instructions of an enzyme-linked immunosorbent assay (ELISA) detection kit for LPS (Uscn Life Science, Inc., Wuhan, Hubei, China). The samples were diluted 10 times successively. Liver tissues (100 mg) were mixed with 2 ml of 0.5% human thyroid antibody (HTAB), and then the subcellular component was treated three times with ultrasonication (30 s per time). After centrifugation at 17,891 × g for 15 min at 0–4°C, the supernatant was diluted at the ratio of 1:5, and the corresponding reagents were diluted based on the instructions. Afterwards, the standardized well (addition with standard samples), the blank well (without any reagent), and the sample well to be detected (addition with samples to be detected) were set with parallel wells. Standard samples (100 μl) at different concentrations and diluted samples were added and incubated at 37°C for 2 hr. After the liquid being discarded, 100 μl of A detection solution was dropped and incubated together at 37°C for 1 hr. The liquid in the plate was discarded. The plate was rinsed with washing solution and dried using absorbent paper. After 3-times washing, 100 μl of B detection solution was dropped and incubated together for 30 min. Following 5-times washing, each well was added with 90 μl of substrate solution, developed at 37°C for 15–25 min, treated with 50 μl of termination solution and shaken gently to mix well until the color turned from blue to yellow. The OD value of each well was measured using a microplate reader (Thermo, Scientific, Milford, Massachusetts) at the wavelength of 450 nm. Cell growth curve was drawn with the concentration of standard sample as ordinate and the OD value of standard sample as abscissa. Standard curve method was performed to calculate the concentration of unknown samples, with three replications set for each sample.

2.17 Statistical analysis

All data were processed by SPSS 21.0 statistical software (IBM Corp. Armonk, New York). The measurement data were expressed as mean ± standard deviation. The t test was used to compare data between two groups and one-way analysis of variance (ANOVA) for comparisons among multiple groups. Enumeration data were expressed as the percentage (%) and the comparisons between groups were analyzed using Chi square test. A p value <0.05 was indicative of significant statistical difference.

3.1 LPS disrupts IEC barrier function by enhancing MPO activity and ROS level

Initially, to investigate the effect of LPS on the barrier function of IECs, CD326 positive cell subsets were separated by flow cytometry (Figure 1a), which was IECs of mice and treated with LPS. RT-qPCR (Figure 1b), western blot analysis (Figure 1c), immunofluorescence assay (Figure 1f) were used to measure the expression and location of ZO-1 and occludin in the control and LPS groups. MPO kit, ELISA, and resistance measuring instrument were used to measure the MPO activity (Figure 1d), ROS level (Figure 1e), TEER (Figure 1g) in the two groups. In the control group, ZO-1 and occludin mainly distributed in the cell membrane, showing strong fluorescence signals and close connections between cells without obvious fracture. The messenger RNA (mRNA) levels of ZO-1 and occludin in the control and LPS groups were 0.512 ± 0.051 and 0.385 ± 0.038, and the protein levels were 0.215 ± 0.021 and 0.235 ± 0.023, respectively. The MPO activity was 27.49 ± 2.71 U/L, ROS level was 398.35 ± 39.83 mol/L, and TEER was 156 ± 8 Ohm/cm². In the LPS group, ZO-1 and occludin showed much lower fluorescence signal with fraction, unconnected cells, and exhibited a little fluorescence signal in the cytoplasm, the mRNA and protein levels as well as TEER were significantly decreased, whereas the MPO activity and ROS level were significantly increased (all p < 0.05). These findings demonstrated that the LPS treatment might promote the inhibition of ZO-1 and occludin expression, increase of the MPO activity and ROS level, as well as permeability of IECs, thus weakening the barrier function of IECs.

3.2 LPS disrupts IEC barrier function by upregulating expression of TXNIP and NLRP3 and the interaction of TXNIP–NLRP3

Subsequently, to further investigate the mechanism of LPS in IEC barrier function via the regulation to TXNIP–NLRP3 pathway, we used RT-qPCR to detect the expression of TXNIP and NLRP3, and coimmunoprecipitation to detect the interaction of TXNIP and NLRP3 protein in the control and LPS groups. As shown in Figure 2a,b, in IECs, the expression of TXNIP and NLRP3 were significantly increased and the content of TXNIP–NLRP3 protein was also significantly increased (p < 0.05), suggesting that the interaction of TXNIP–NLRP3 was significantly strengthened (Figure 2c,d). It can be concluded that LPS might upregulate the expression of TXNIP and NLRP3 and increase the interaction of TXNIP–NLRP3, disrupting IEC barrier function.

3.3 TXNIP knockout elevates ZO-1 and occludin expression yet lowers MPO activity and ROS level in IECs

To explore the effects of TXNIP on IEC barrier function, we conducted IEC separation from WT mice and TXNIP^−/− mice, TXNIP overexpression in TXNIP^−/− IECs, IECs in each group treated with LPS. Then immunofluorescence assay, RT-qPCR and western blot analysis were utilized to detect how ZO-1 and occludin expressed in the WT, TXNIP^−/− and pcDNA-TXNIP groups. MPO kit, ELISA, and resistance measuring instrument were used to measure the MPO activity, ROS level, and TEER, respectively. As displayed in Figure 3e, in the WT group, ZO-1, and occludin mainly distributed in a mass in the cytoplasm, indicating weakened fluorescence signal and little connection between cells with obvious fracture. The mRNA levels of ZO-1, occludin and NLRP3 were 0.365 ± 0.044, 0.322 ± 0.053, and 0.861 ± 0.052 and the protein levels were 0.384 ± 0.051, 0.355 ± 0.031, and 0.921 ± 0.042, respectively (Figure 3a,b). The MPO activity was 29.82 ± 2.98 U/L, ROS level was 395.26 ± 39.52 mol/l, and TEER was 101 ± 7 Ohm/cm² (Figure 3c,d,f). Although in the TXNIP^−/− group, ZO-1, and occludin showed much stronger fluorescence signal without fraction and closely connected cells as compared with the WT group (Figure 3e), the expression of ZO-1, occludin, and NLRP3 was significantly increased (Figure 3a,b), whereas the MPO activity and ROS level were significantly decreased (Figure 3c,d) and TEER was significantly increased (Figure 3f; all p < 0.05). However, the above factors between the WT and pcDNA-TXNIP groups showed no significant differences (p > 0.05). Compared with the TXNIP^−/− group, in the pcDNA-TXNIP group, ZO-1, and occludin showed much lower fluorescence signal and unconnected cells with remarkably increased fraction, the expression of ZO-1, occludin, and NLRP3 was decreased, whereas the MPO activity and ROS level were markedly increased and TEER was significantly decreased (all p < 0.05). These findings demonstrated that TXNIP knockout might elevate ZO-1 and occludin expression yet lower the MPO activity and ROS level and lessen permeability of IECs, thus restoring the barrier function of IECs.

3.4 TXNIP knockout represses the MPO activity and oxidative stress in intestinal mucosa

It has been reported that TEMPO has mitochondria-targeted antioxidant activity and protects the kidney from aldosterone-induced damage (Ding et al., 2017), so we want to explore the protective effect of TEMPO on intestinal mucosa. Furthermore, to manifest the effects of TXNIP on the MPO activity and ROS level in the intestinal mucosa induced by high-fat and high-fructose diet in mice with steatohepatitis, WT, and TXNIP^−/− mice models were established and then assigned into the normal, model, and TEMPO groups. As presented in Figure 4a,b, the MPO activity and ROS level in intestinal mucosa in the model and TEMPO groups were significantly elevated as compared with those in the normal group (p < 0.05). Compared with the model group, MPO activity, and ROS level in the intestinal mucosa in the TEMPO group were significantly descended (p < 0.05). In comparison to WT mice, MPO activity, and ROS level in the intestinal mucosa in the model and TEMPO groups from TXNIP^−/− mice were significantly declined (all p < 0.05). After the TEMPO treatment, in WT mice, the clearance rate of ROS was 35.3%, and that of TXNIP^−/− mice was 45.8%, indicating that the clearance rate of ROS was significantly increased in TXNIP^−/− mice. The present findings suggested that high-fat and high-fructose diet can trigger elevated the MPO activity and oxidative stress, which can be inhibited by the knockout of TXNIP, and the TEMPO treatment can further repress oxidative stress.

3.5 TXNIP knockout protects intestinal mucosal barrier function by reducing Escherichia coli

To demonstrate the effects of TXNIP on the intestinal mucosal barrier function of mice with steatohepatitis induced by high-fat and high-fructose diet, we used the immunofluorescence assay to detect how ZO-1 and occludin in intestinal epithelial tissue was located, RT-qPCR to detect the mRNA levels of TXNIP, ZO-1, and occludin, and FITC-Dextran to detect the permeability in the intestinal epithelial tissue from the normal and model groups. The results showed that compared with the normal group, ZO-1 and occludin in intestinal epithelial tissue exhibited mass distribution in cytoplasm and discontinuous gaps in fluorescent stained cells. Moreover, the connection between cells and cells was not close, whereas the expression of TXNIP, ZO-1, and occludin were significantly reduced and the FITC content in the intestinal epithelial tissue was significantly enhanced (all p < 0.05). In comparison to WT mice, TXNIP^−/− mice showed no TXNIP expression whereas ZO-1 and occludin expressions were significantly increased (p < 0.05). ZO-1 and occludin displayed honeycomb linear fluorescence distributed around cell membrane and relatively complete outline; in addition, positive fluorescence staining intensity in cytoplasm decreased and the FITC content in intestinal epithelial tissue was noticeably ascended (all p < 0.05; Figure 5a–f).

To demonstrate the effects of TXNIP on the change of intestinal flora induced by high-fat and high-fructose diet, we classified mice into the normal, model, and antibiotic groups and utilized RT-qPCR to evaluate the content of intestinal flora in each group. As shown in Table 3, compared with the normal group, the number of Clostridium, Prevotella, Bifidobacterium, and Lactobacillus was obviously decreased whereas E. coli was obviously increased in the model and antibiotic groups (all p < 0.05). Compared with the model group, the number of Clostridium, Prevotella, Bifidobacterium, and Lactobacillus was obviously increased whereas E. coli was obviously decreased in the normal and antibiotic groups (all p < 0.05). Compared with WT mice, the number of Clostridium, Prevotella, Bifidobacterium, and Lactobacillus was obviously increased, whereas E. coli was obviously decreased in TXNIP^−/− mice in the model and antibiotic groups (all p < 0.05). After the antibiotic treatment, in WT mice, the clearance rate of E. coli was 18.9%, and that in TXNIP^−/− mice was 28.8%, suggesting that the clearance rate of E. coli in TXNIP^−/− mice was much higher than that in WT mice. We demonstrated that TXNIP knockout can prevent the disorders in intestinal flora that resulted from high-fat and high-fructose diet and increase the clearance rate of E. coli. The above findings indicated that high-fat and high-fructose diet can accelerate the expression of TXNIP in intestinal mucosa, decrease the expression of ZO-1 and occludin and continuous distribution, increase the permeability of intestinal mucosa and intestinal flora disorder protects intestinal mucosal barrier function, thus making IEC barrier function disrupted. However, TXNIP knockout can restore intestinal mucosal barrier function triggered by high-fat and high-fructose diet.

3.6 TXNIP knockout inhibits liver tissue injury

To explore whether TXNIP was involved in the process of steatohepatitis resulted from high-fat and high-fructose diet, we performed the RT-qPCR, automatic biochemical analyzer, ELISA, and HE staining. The results are shown in Figure 6a–e. Compared with the normal group, TNF-α and IL-1β expression, serum levels of ALT and AST, and LPS concentration were significantly elevated in the model group (all p < 0.05). Compared with WT mice, TNF-α and IL-1β expression, serum levels of ALT and AST, and LPS concentration were significantly descended in the model group from TXNIP^−/− mice (all p < 0.05). In the normal group, liver tissue of mice showed no fatty degeneration and inflammatory cell infiltration. While in the model group from WT mice, liver tissue exhibited fat vacuoles with different size, more inflammatory cell infiltration and necrosis; in the model group from TXNIP^−/− mice, liver tissue exhibited much less inflammatory cell infiltration and no noticeable decrease in fat vacuoles (Figure 6f). The results above manifested that TXNIP knockout can repress the expressions of inflammatory factors, lowers LPS concentration, and pathological injury of liver tissues, hence protecting against the development of steatohepatitis.