Mitigation of indomethacin-induced gastrointestinal damages in fat-1 transgenic mice via gate-keeper action of ω-3-polyunsaturated fatty acids

Abstract

Non-steroidal anti-inflammatory drugs (NSAIDs) damage the gastrointestinal (GI) epithelial cell membranes by inducing several signals through lipid raft organization after membrane incorporation, whereas ω-3 polyunsaturated fatty acids (PUFAs) relieve inflammation, reduce oxidative stress, and provide cytoprotection, consequent to lipid raft disorganization. Therefore, we hypothesized that ω-3 PUFAs can protect the GI from NSAID-induced damages by initiating the gatekeeper action of cell membranes, subsequent to anti-inflammatory and anti-oxidative actions. Administration of indomethacin (IND) leads to the formation of lipid rafts and activation of caveolin-1; however, no such observations were made upon co-administration of eicosapentaenoic acid (EPA) and IND. In addition, the EPA-induced lipid raft disorganization, caveolin-1 inactivation, and cellular cytotoxicity were inhibited when target cells were knocked-out using G-protein coupled receptor 120 (GPR 120). EPA significantly attenuated IND-induced oxidative damage and apoptosis. IND administration induced significant ulceration, bleeding, and oedema in the stomach or small intestine of wild-type (WT) mice; however, such severe damages to the GI significantly decreased in fat-1 transgenic (TG) mice (P < 0.001), which exhibited decreased cyclooxygenase-2 expression and apoptosis, decreased interleukin-1β and FAS concentrations, and increased heme oxygenase-1 concentration. Our study indicates that the gatekeeper function of ω-3 PUFAs improves GI safety when administered with NSAID.

Figure 1

Inhibitory effect of EPA on lipid rafts organization in gastric epithelial cells after 500 μM IND administration (a) SEM (upper and middle) and confocal analysis of GM1 (lower) were done in RGM-1 cells, non-transformed gastric mucosal cells pretreated with EPA (10 μM) for 1 hr before 16 h of 500 μM IND administration. Confocal microscopy analysis was done with GM1 staining in RGM-1 cells and their localization of GM1 is seen as green fluorescence via FITC-conjugated GM1 (Magnification at x 800). (b) Dot-blot analyses and Western blot for lipid raft fractionated by sucrose density gradient centrifugation RGM-1 cells were treated with vehicle alone (Control) or with 500 μM IND and with 10 μM EPA or not as indicated. The distribution of GM1 in the gradient fractions was determined using 2 μg of protein of each fraction and HRP-CTB. (c) Immunoprecipitation was done with caveolin-1 antibody from each homogenate and Western blotting was done with phosphorylated-tyrosin and caveolin-1, respectively. Also western blot with NOX-1 was done. Full length blots are presented in Supplementary Figure 1a. (d) Cell survival by MTT assay between ns siRNA-transfected and GPR120 siRNA-transfected RGM-1 cells. In RGM-1 cells transfected with either ns siRNA or GPR120 siRNA, MTT assay was done under the challenge with 10 μM concentration of EPA for 1 hr before IND-treatment (500 μM, 16 h). (e) Western blotting for caveolin-1 in fractions obtained by sucrose density gradient centrifugation RGM-1 cells were transfected with GPR120 siRNA and RGM-1 cells were treated with vehicle alone (Control) or with 500 μM IND and with 10 μM EPA or not as indicated. Each fraction was resolved on SDS-PAGE gels and Western blotting using caveolin-1 antibody.

Figure 2

Changes of NOX, apoptosis, and oxidative stress according to IND and combination of IND and EPA (a) Cell survival by MTT assay and Cell counting. They were done in RGM-1 cells under the challenge with 10 μM EPA for 1 hr before IND treatment (500 μM, 16 h). (b) Western blot for Bcl-2, Bax, and PARP RGM-1 cells were pretreated with EPA (10 μM) for 1 hr and stimulated with IND for 16 hr. Expression of apoptosis mediator was analyzed by immunoblotting. All experiments were done in triplicate. Mean ± SE was calculated from independent experiments. Full length blots are presented in Supplementary Figure 1(b–d). (c) RT-PCR for NOXs family RGM-1 cells were pretreated with EPA (10 μM) for 1 hr and stimulated with IND for 16 hr. Expression of NOX family was analyzed by RT-PCR. (d) FACS analysis of DCF-DA Cells were treated 10 μM of EPA for 1 hr before IND treatment (500 μM, 16 h) then incubated with fluorescent probe, H₂DCFDA for 30 min. (e) ESR measurement ESR spectra of the DMPO‐OH adducts arising in the Fenton reaction. After administration of EPA, the signal intensity of ESR spectra was decreased in a dose-dependent manner.

Figure 3

Mitigating effects of ω-3 PUFAs in IND-induced gastric damage (a) Protocol for IND-induced gastric damages. The schematic overview: the experimental protocol for IND-induced gastric ulcer using WT and fat-1 TG mice. The experimental animal (n = 10) were divided 4 groups, WT with CMC, WT with IND, fat-1 TG with CMC, and fat-1 TG with IND. (b,c) Lipid profiles of WT and fat-1 TG mice stomach before and after IND administration. The mice were sacrificed 16 h after treatment with IND (50 mg/kg). Stomach tissue of fat-1 TG mice and WT mice was analyzed by LC/MS/MS as described in Materials and Methods. (d) Gross gastric lesion according to WT and fat-1 TG mice before and after IND (50 mg/kg, po). Gross lesion pictures of representational photo in each group were noted. (e) Pathological scores according to group gastric changes were quantified from H&E stained sections. In stomach was stained by H&E staining and the pathologic scores were represented as mean ± SD of 10 animals. Statistical significance with the controls was analyzed by one-way ANOVA. (f) TUNEL assay for detection of apoptosis. Stomach tissue slides stained with TUNEL staining to detect the apoptotic positivity cells. Graph shows the number of TUNEL-positive cells per high power field (magnification × 400) from at least 10 fields. Each value represents the mean ± SD for 10 mice. (g) ELISA assay for IL-1β (upper) and IL-6 (lower). The mucosal cytokine levels of IL-1β and IL-6 were assayed by ELISA method, respectively. All experiments were done in triplicate. Each value represents the mean ± SD for 10 mice.

Figure 4

Molecular mechanisms to explain mitigating effects of ω-3 PUFAs in IND-induced gastric damage (a) Changes of COX-2 according to group Equal amounts of total protein extracted from stomach tissues were subjected to Western blotting using COX-2 antibody. (b) Cytokine array for inflammatory mediators. Equal amounts of total protein extracted from stomach tissues were subjected to cytokine array for inflammatory mediators. The relative intensity of each cytokine was showed (*denoted P < 0.01; versus WT-CMC). (c) Changes of apoptosis mediators according to group. Equal amounts of total protein extracted from stomach tissues were subjected to Western blotting using Bcl-2 and FAS antibodies. (d) Changes of HO-1 and HSPs according to group. Equal amounts of total protein extracted from stomach tissues were subjected to Western blot analysis using HO-1, HSP60, and HSP70 antibodies.

Figure 5

Mitigating effects of ω-3 PUFAs against IND-induced small intestine damage (a) Protocol for IND-induced small intestinal damages. The schematic overview: the experimental protocol for IND-induced small intestine injury using WT and fat-1 TG mice. The experimental animal (n = 10) were divided 4 groups, WT with CMC, WT with IND, fat-1 TG with CMC, and fat-1 TG with IND. (b) Comparison of concentration of endogenous ω-3 PUFAs between WT and fat-1 TG mice. The mice were sacrificed 48 hr after treatment with IND (30 mg/kg, po). Small intestine tissue of fat-1 TG mice and WT mice was analyzed by LC/MS/MS as described in Materials and Methods. (c) Gross lesion and total pathological score of IND-induced small intestine injury. Gross lesion pictures of representational photo in each group were noted (upper). In small intestine was stained by H&E staining and the pathologic scores were represented as mean ± SD of 10 animals (lower). Statistical significance with the controls was analyzed by one-way ANOVA. (d) Changes of COX-2 according to group Equal amounts of total protein extracted from small intestine tissues were subjected to Western blotting using COX-2 antibody. (e) TUNEL assay for detection of apoptosis. Small intestine tissue slides stained with TUNEL staining to detect the apoptotic positivity cells. Graph shows the number of TUNEL-positive cells per high power field (Magnification at × 400) from at least 10 fields. Each value represents the mean ± SD for 10 mice. (f) Changes of apoptosis mediators according to group Equal amounts of total protein extracted from small intestine tissues were subjected to Western blotting using Bcl-2 and FAS antibodies. Full length blots are presented in Supplementary Figure 2(a–c). (g) Changes of Claudin-1 and ZO-1 in small intestine according to group. The expression of Claudin-1 and ZO-1 were investigated by immunohistochemistry. Each value represents the mean ± SD for 10 mice.

Figure 6. Summary explaining the mitigating action of ω-3 PUFAs against NSAIDs-induced GI damage.

As a proposed pathway for ω-3 PUFAs-mediated inhibition of IND-induced GI damages, our study significantly opens the possibility of ω-3 PUFAs conjugated NSAID or the combination of ω-3 PUFAs and NSAID as potential GI-safer NSAIDs.