The human colon cancer cell lines (Caco-2) were obtained from the Cell Resource Center, Peking Union Medical College (Beijing, China). Cells were maintained in Rosewell Park Memorial Institute (RPMI) 1640 medium (Gibco, NY, USA) containing 20% fetal bovine serum (FBS, Gibco), 100 U/ml penicillin and 0.1 mg/ml streptomycin in a humidified (37°C and 5% CO2) incubator. For LPS treatment, Caco-2 cells were incubated in RPMI 1640 and different concentrations of LPS (Sigma-Aldrich, St. Louis, USA) for 24 h.

FA (99% purity, CAS 110773-201614), purchased from the National Institutes for Food and Drug Control (Beijing, China), was dissolved in RPMI 1640 to 10 mM as a stock solution. The experiment concentrations of FA were 0, 25, 50, and 100 μM. Cells were pretreatment with FA for 2 h before LPS treatment.

Transepithelial electrical resistance (TER) of filter-grown Caco-2 monolayers was measured using a Millicell-ERS system (Millipore, Bedford, USA). Cells were seeded on transwell inserts (Corning, Cambridge, USA) with polyethylene terephthalate membrane (0.33 cm², 0.4 μm pore size) at a density of 1.0 ×10⁶ cells/well and monitored daily for about 21 days. When cells reached confluence and completely differentiated, different concentration FA (25, 50, or 100 μM) were added to the apical of the filter for 2 h and then treated with or without LPS. Changes in TER under experimental conditions were expressed as a percentage of the corresponding basal values. Determinations were repeatedly operated on three different sites transwell insert, respectively.

The permeability of the epithelial across Caco-2 cell monolayers was quantified by measuring the flux of the FITC-labeled dextran of molecular mass 4 kDa (FD4; Sigma-Aldrich, St. Louis, USA). In brief, 1 mg/ml of FD4 was added to the apical chamber and the medium was collected from the basolateral chamber after 2 h’s incubation. Fluorescence intensity of FITC was quantified using a fluorescence plate reader (BioTek, Winooski, USA) with excitation at 492 nm and emission at 520 nm.

Cell viability was determined using Cell Counting Kit-8 (CCK-8; Dojindo, Kumamoto, Japan) assay according to the manufacturer’s instructions. In brief, Caco-2 cells (1 × 10⁴ cells) were plated on a 96-well plate and incubated at 37°C with 5% CO₂ for 24 h. Cells were then washed twice with PBS and incubated with LPS (0.001, 0.01, 0.1, 1, 10, or 100 μg/ml) for 24 h or FA (1, 5, 10, 20, 50, 100, 200, or 500 μM) for 48 h. The cells were then treated with 10 µl/well of CCK-8, and incubated at 37°C for 2 h. Absorbance was detected at 450 nm using a microplate reader (PerkinElmer, Inc., Waltham, USA).

Fully confluent cultured Caco-2 cells were pretreated with 100 μM FA for 2 h and then subjected with LPS for 24 h. Cells were scraped off and centrifuged at 1500 rpm for 10 min. The samples fixed in overnight with 4% glutaraldehyde and then post-fixed in cold 1% osmium tetroxide for 1 h, followed by three cacodylate buffer washes. After dehydration in graded ethanol solutions, cells were embedded in Epon-Araldite (EPON 812, Emicron, Shanghai, China). Ultra-thin sections were stained with saturated uranyl acetate in 50% ethanol and lead citrate and examined with transmission electron microscopy (H7650, Hitachi, Ltd., Tokyo, Japan).

Total RNA from Caco-2 cells was extracted using Trizol reagent (Life Technologies, Carlsbad, USA) according to the manufacturer’s protocol. The quality and quantity of the RNA samples were assessed on a Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, USA) using an RNA 6000 Nano kit (Agilent Technologies). Small RNA libraries were generated according to the protocol of Illumina TruSeq™ Small RNA Sample Preparation kit (Agilent Technologies). Affymetrix GeneChip miRNA array analysis was conducted to detect the expression pattern of miRNAs, which were provided by Shanghai Biotechnology Corporation (Shanghai, China). Bioinformatic analyses and visualization of microarray data were performed with MultiExperiment Viewer (MEV) software v.4.6 (TIGR, La Jolla, USA).

Lentivirus-miR-200c-3p (Lv-miR-200c-3p) or lentivirus-miR-200c-3p sponge (Lv-miR-200c-3p spong) were transfected into Caco-2 cells in order to increase or inhibit miR-200c-3p expression, respectively, and lentivirus-negative control (Lv-NC) was used as negative control. The Lv-miR-200c-3p, Lv-miR-200c-3p spong and Lv-NC were designed and synthesized by Hanheng Biotechnology Corporation (Shanghai, China). Concisely, 293T cells were co-transfected with viral packaging vector pMD2.G and psPAX2 (Addgene, Cambridge MA), along with a lentiviral construct expressing a specific miR-200c-3p, miR-200c-3p spong or the empty vector, using LipoFitter™ transfection reagent (Hanheng Biotechnology Corporation) according to the manufacturer’s instructions. The procedure of miR-491-3p synthesis was the same as that of miR-200c-3p. The transfection medium was replaced after 6 h with fresh Dulbecco’s Modified Eagle Medium (Gibco, NY, USA) containing 10% FBS, 100 U/ml penicillin and 0.1 mg/ml streptomycin. At 48 h after incubation, the supernatant of lentivirus was ultracentrifuged, concentrated and used to infect cells, as previously described (Wu et al., 2019). In brief, Caco-2 cells (1 × 10⁶ cells/well) seeded on a 6-well plate were transfected with Lv-miR-200c-3p, Lv-miR-200c-3p spong or Lv-NC at a multiplicity of infection (MOI) of 15, or transfected with Lv- miR-491-3p, Lv- miR-491-3p spong or Lv-NC at a MOI of 20, and incubated at 37°C with 5% CO2 for 48 h. Transfection efficiency was confirmed by quantitative reverse transcription PCR. Then, the cells were treated according to different test requirements.

Total RNA was extracted using TRIzol Reagent (Life Technologies) according to the manufacturer’s protocol. The cDNA was reverse transcribed using the revert aid first-strand cDNA Synthesis kit (Thermo Fisher Scientific). Real-time PCR was performed using SYBR Green qPCR Master Mix (Applied Biosystems). The Occludin, ZO-1 and β-actin primer sequences were synthesized by Sangon Biotech (Shanghai, China). The primers sets were as follows: Occludin, 5′-GCAAAGTGAATGACAAGCGG-3′ (F) and 5′-CACAGGCGAAGTTAATGGAAG-3′ (R); ZO-1, 5′-CGAAGGAGTTGAGCAGGAAA-3′ (F) and 5′-ACAGGCTTCAGGAACTTGAG-3′ (R); β-actin, 5′-TGCAGAAAGAGATCACCGC-3′ (F) and 5′-CCGATCCACACCGAGTATTTG-3′ (R). To generate cDNA of miR-200c-3p, total RNA was reverse transcribed using a miRcute miRNA cDNA Kit (TIANGEN BIOTECH, Beijing, China) according to the manufacturer’s protocol. Real-time PCR was performed using miRcute miRNA (TIANGEN BIOTECH) according to the manufacturer’s protocol. The primers of miR-200c-3p and U6 were purchased from Guangzhou Ruibo Biotechnology Co. LTD. All real-time qPCR analyses were conducted using the 7500 Real-Time PCR system (Applied Biosystems). Relative expression of mRNA and miR-200c-3p were normalized using the endogenous controls β-actin and U6, respectively. The amplification protocols were as follows: 95°C for 10 min, 40 cycles of 95°C for 15 s, 60°C for 34 s, and 72°C for 34 s. Relative expression fold changes were calculated using the 2^−ΔΔCT method.

Total proteins from Caco-2 cells were extracted on ice using RIPA buffer (Beyotime Company, Beijing, China) supplemented with protease inhibitors (Roche, Basel, Switzerland). The concentration of total proteins was measured using the BCA protein assay kit (Beyotime, Jiangsu, China). Then, specified amounts of protein samples were separated by SDS-PAGE and electrotransferred onto nitrocellulose membranes (Pierce, Waltham, MA). After blocking with Odyssey blocking buffer (LI-COR Biosciences, NE, USA) for 2 h, the membranes were incubated with primary antibodies rabbit anti-occludin (1:100, 71–1500, Invitrogen, Carlsbad, USA), rabbit anti-ZO-1 (1:500, 21773-1-AP, Proteintech, Wuhan, China), rabbit anti-PTEN (1:1000, 9559, Cell Signaling Technology, Danvers, MA), rabbit anti-PI3K (1:1000, ab191606, Abcam, Cambridge, United Kingdom), rabbit anti-phospho-PI3K (1:1000, ab182651, Abcam), rabbit anti-Akt (1:1000, 4685, Cell Signaling Technology), rabbit anti-phospho-Akt (1:1000, 4060, Cell Signaling Technology), and rabbit anti-β-actin (1:2000, 4970, Cell Signaling Technology) overnight at 4°C. The membranes were washed three times and incubated with horseradish peroxidase (HRP)-conjugated goat-anti-rabbit IgG secondary antibody (1:5000, A8275, Sigma-Aldrich) for 30 min at room temperature. Then, the protein bands were visualized using enhanced electrogenerated chemiluminescence western blotting detection reagents (Pierce Corporation, Rockford, IL). The gray values of the target band of proteins were quantified using Image J (National Institutes of Health, Bethesda, MD).

Caco-2 cells were cultured on coverslips to fully confluence and subjected to various experimental conditions. After treatment, cells were washed with PBS three times and fixed with 4% paraformaldehyde for 10 minutes at room temperature. Then cells were permeated with 0.1% Triton X-100 for 5 min, followed by blocking with 3% BSA for 1 h at room temperature. Caco-2 cells were incubated with rabbit anti-ZO-1 antibody (1:50, 21773-1-AP, Proteintech) at 4°C overnight. Coverslips were washed and incubated with a secondary antibody Alexa Fluor 488 goat anti-rabbit IgG (1:400, A-11008, Life Technologies, Carlsbad, USA) for 40 min at room temperature in the dark. Then, Cells were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (Beyotime, Jiangsu, China) for 5 min at room temperature and visualized with a fluorescence microscope (Olympus IX71, Tokyo, Japan).

All data were represented as means ± standard deviation (SD). Statistical analysis was performed using the GraphPad Prism 7 program (GraphPad, La Jolla, USA). For data from assay evaluating the effect of LPS on TER values and FD4 flux, the unpaired t test was used. One-way analysis of variance (ANOVA) was performed to compare the statistical differences of data among three or more groups. A P-value of <0.05 was considered statistically significant. All experiments in this study were repeated at least three times.

Firstly, the effects of different concentrations of LPS (0.001 − 100 μg/ml) on epithelial paracellular permeability and cell viability were determined by measuring TER values and FD4 flux in Caco-2 cell monolayers. Figure 1A displayed that LPS at a concentration of above 10 μg/ml significantly decreased the TER value (P < 0.01), whereas LPS at a concentration of below 10 μg/ml had no significant effect on the TER. Figure 1B presented that 10 μg/ml or 100 μg/ml LPS significantly increased the FD4 flux from Caco-2 monolayer apical to the basolateral chamber (P < 0.01). Figure 1C showed that 10 μg/ml LPS had no effect on the cell viability, while 100 μg/ml LPS decreased the cell viability. Consequently, we applied 10 μg/ml LPS concentration to induce intestinal epithelial barrier dysfunction for subsequent experiments.

Effects of FA on LPS-induced intestinal epithelial barrier dysfunction were subsequently explored. To determine whether FA had toxic effects, Caco-2 cells were incubated with 0−500 μM FA for 48 h and cell viability was assessed. In Figure 2A , there were no statistical differences in cell viability, regardless of dose of FA, indicating 0−500 μM FA showed no cytotoxic effects on Caco-2 cells. Then, Caco-2 cells were pretreated with FA (25, 50, and 100 μM) for 2 h, followed by LPS stimulation. Figure 2B displayed that compared with LPS-treated cells, cells pretreated with 50 and 100 μM FA had higher TER values (P < 0.05 and P < 0.01). Similarly, the LPS-induced increase in FD4 flux across the Caco-2 monolayer was significantly attenuated by 50 and 100 μM FA (P < 0.05 and P < 0.01, Figure 2C ).

Compared with control cells, decreased mRNA (P < 0.01, Figure 2D ) and protein (P < 0.01, Figure 2E ) expression levels of occludin and ZO-1 were observed in LPS-treated cells, but not in cells pretreated with 100 μM FA. Compared with LPS-treated cells, 100 μM FA pretreatment increased the mRNA and protein expression of occludin and ZO-1 (P < 0.01). TJs ultrastructure of Caco-2 cells was investigated by transmission electron microcopy ( Figure 2F ). In the control of Caco-2 cells, TJ structure between adjoining cells was intact and densely connected. In LPS-treated Caco-2 cells, the TJ structure was obviously damaged with large gaps between adjoining Caco-2 cells and the electron-dense material was reduced. However, pretreatment with 100 μM FA significantly alleviated LPS-induced TJs disruption.

The miRNA expression profiles of Caco-2 cells treated with LPS and FA were evaluated by microarray hybridization. Using hierarchical clustering, 19 miRNAs were found to be significantly altered in the cells pretreated with FA compared with the cells treated with LPS alone. Of these miRNAs, 10 were up-regulated and nine down-regulated ( Figure 3A ; Table S1 ). The miRNA expression profiles showed that the expression of miR-200c-3p was up-regulated 3.87-fold change in the cells pretreated with FA compared with the cells treated with LPS alone. Furthermore, we validated the expression of miR-200c-3p by qRT-PCR. The results revealed that pretreatment with FA attenuated LPS-induced reduction of miR-200c-3p expression induced by LPS in Caco-2 cells (P < 0.01; Figure 3B ).

The effect of miR-200c-3p or miR-491-3p on the protective effects of FA on LPS-induced intestinal epithelial barrier dysfunction was explored. Overexpression or suppression of miR-491-3p had no effect on the expression of occludin and ZO-1 ( Figure S1 ). Further, Figure 4A showed that compared with Lv-NC transfection, the expression level of miR-200c-3p was significantly enhanced by Lv-miR-200c-3p, but was notably reduced by transfection with Lv-miR-200c-3p spong in Caco-2 cells (P < 0.01). Then, TER values, FD4 flux, tight junction-associated protein occludin, and ZO-1 expression were determined in transfected Caco-2 cells. The protecting effects of FA on intestinal epithelial barrier function were enhanced by transfection with Lv-miR-200c-3p, as shown by elevated TER values (P < 0.01, Figure 4B ), decreased FD4 flux (P < 0.01, Figure 4C ), and increased occludin and ZO-1 proteins expression (P < 0.01, Figure 4D ). However, these results were reversed by suppression of miR-200c-3p (P < 0.05 and P < 0.01, Figures 4B−D ).

Additionally, immunofluorescence staining was employed to detect the change of ZO-1 distribution in Caco-2 cells ( Figure 4E ). In the control group, ZO-1 staining appeared characteristically continuous belt-like pattern around the apical membrane of Caco-2 cells. As expected, there were striking changes in the distributions of ZO-1 in the LPS-treated cells, exhibiting irregularly undulating and staining intensity decreasing. In Caco-2 cells pretreated with FA, the rearrangements of ZO-1 were markedly attenuated, presenting a relatively unambiguous profile. Moreover, miR-200c-3p overexpression further improved the distributions of ZO-1. However, suppression of miR-200c-3p weakened the protective effects of FA on LPS-induced ZO-1 distributions.

To clarify the mechanism underlying FA protecting Caco-2 cells against LPS-induced intestinal epithelial barrier dysfunction, the PTEN/PI3K/AKT pathway was examined by western blotting. As shown in Figure 5A , LPS treatment increased PTEN protein expression (P < 0.01) and decreased expression of phosphorylated PI3K and AKT (P < 0.01). However, compared with LPS-treated cells, cells pretreated with FA had lower expression of PTEN and higher expression of phosphorylated PI3K and AKT (P < 0.01). FA alone increased the expression of phosphorylated PI3K and AKT ( Figure S2 ). Moreover, the overexpression of miR-200c-3p significantly strengthened the effect of FA on the PTEN/PI3K/AKT pathway (P < 0.01). However, suppression of miR-200c-3p notably inhibited the effect of FA on PTEN/PI3K/AKT pathway in LPS-treated cells (P < 0.05 or P < 0.01).

Subsequently, LY294002, the inhibitors of PI3K/Akt signaling pathway was used to suppress PI3K/Akt activity. The expressions of occludin and ZO-1 proteins were then examined. Compared with control cells, increased expression of occludin and ZO-1 proteins were observed in Caco-2 cells pretreated with FA alone, but not in cells treated with LY294002 alone (P < 0.01, Figure 5B ). FA-induced increase in the expression of occludin and ZO-1 proteins was prevented by LY294002 (P < 0.01).