Disruption of Tight Junctions in Intestinal Epithelial Cells by Toxic Advanced Glycation End-Products

Abstract

Intestinal epithelial cells (IECs) play a crucial role in forming a protective barrier and regulating the absorption of substances passing through the small intestine. Disrupting the epithelial barrier function can result in intestinal diseases such as inflammatory bowel disease. Glyceraldehyde (GA)-derived advanced glycation end-products (AGEs) (toxic AGEs, TAGE) are AGEs formed by the nonenzymatic Maillard reaction. Although AGEs have been implicated in intestinal barrier breakdown, the associated mechanism remains underexplored. In this study, the effects of accumulated TAGE in IECs were investigated by focusing on tight junctions using Caco-2 cells—a human colorectal epithelial adenocarcinoma cell line. While GA treatment induced the formation of intracellular TAGE in Caco-2 cells, resulting in cell death, the generated intracellular TAGE triggered increased paracellular permeability. In addition, immunofluorescence staining showed that GA treatment decreased the fluorescence intensities of ZO-1 and claudin-7, which are tight junction proteins attached to the plasma membrane. Furthermore, an evaluation of the mechanism behind intestinal barrier breakdown revealed excessive reactive oxygen species (ROS) production and increased expression of reduced nicotinamide adenine dinucleotide phosphate (NAD(P)H) oxidase subunit genes, a mechanism of ROS production, in the GA-treated group compared with the control group. Furthermore, GA treatment induced necrosis and caused cytotoxicity in this condition. Overall, these results suggest that TAGE induction can disrupt tight junctions in IECs via cell injury as a pathway.

INTRODUCTION

The intestines have various functions, including digestion of food, absorption of nutrients, electrolytes, and water, and protection against pathogens and toxic agents. The intestinal epithelium constitutes an essential barrier against the intrusion of harmful substances. It comprises a mucosal layer, a monolayer of epithelial cells, and a lamina propria, comprising immune cells that react rapidly to the presence of foreign substances.¹⁾ Furthermore, the layers of intestinal epithelial cells (IECs) are firmly attached to one another by a tight junction, which is made up of proteins such as zonula occludens (ZO), occludin, and claudins.²⁾ Therefore, intestinal barrier breakdown can result in the invasion of viruses and microbials, contributing to various intestinal disorders, such as inflammatory bowel disease (IBD).

Advanced glycation end-products (AGEs) resulting from the nonenzymatic reactions between sugars and proteins are known for the pathogeneses of several lifestyle-related diseases, such as diabetes mellitus (DM), coronary heart disease, and Parkinson’s disease.³⁾ Some reports suggest a dual function of different AGE types on intestinal barrier function in IECs, with some contributing to disruption^4–11) and others promoting protection.^12,13)

AGEs formed from glyceraldehyde (GA), a fructose/glucose metabolic intermediate, are highly cytotoxic and are called toxic AGEs (TAGE). TAGE is associated with the pathogenesis of DM and its related vascular complications by binding to the receptor for AGEs (RAGE), which triggers intracellular reactive oxygen species (ROS) production. TAGE can damage various cell types, including hepatocytes,¹⁴⁾ vascular endothelial cells,¹⁵⁾ cardiac fibroblasts,¹⁶⁾ and neuroblastoma cells.¹⁷⁾ However, the effect of intracellular TAGE on IECs is unclear. In this study, the effects of TAGE on IECs were investigated by evaluating the intestinal barrier in GA-affected Caco-2 cells. The pathways leading to the disruption of tight junctions by TAGE were also investigated.

MATERIALS AND METHODS

Materials

Caco-2 cells (RCB0988) were provided by the RIKEN BRC through the National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology (MEXT)/Japan Agency for Medical Research and Development (AMED) (Ibaraki, Japan). FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan) supplied the following: aminoguanidine (AG), D-glucose, low-glucose Dulbecco’s modified Eagle’s medium (DMEM) with l-glutamine, Hank’s Balanced Salt Solution with Ca²⁺ and Mg²⁺ (HBSS (+)), MEM Nonessential Amino Acids Solution (NEAA), penicillin-streptomycin solution (PS), thiourea, and urea. Dojindo Laboratories (Kumamoto, Japan) supplied the Cell Counting Kit-8 and CHAPS. While Nacalai Tesque Inc. (Kyoto, Japan) supplied Annexin V-FITC Apoptosis Detection Kit, GA and Chemi-Lumi One L, Nichirei (Tokyo, Japan) supplied fetal bovine serum (FBS). Hydroxyethyl piperazineethanesulfonic acid (HEPES) and the protease inhibitor cocktail complete Mini were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan) and Roche (Basel, Switzerland), respectively. Lucifer yellow CH, CellROX green reagent, 2-(N-morpholino) ethanesulfonic acid (MES) buffer, and Tween 20 were purchased from Thermo Fisher Scientific Inc. (Waltham, MA, U.S.A.). Phosphate-buffered saline (PBS) tablets and ReverTra Ace qPCR RT Master Mix with gDNA Remover were purchased from TaKaRa Bio Inc. (Shiga, Japan). The Trizma^® base and Triton X-100 were purchased from Sigma-Aldrich Co. LLC (St. Louis, MO, U.S.A.). The XL-Bradford assay kit was purchased from APRO Science (Tokushima, Japan). The Agencourt RNAdvance Tissue Kit was purchased from Beckman Coulter Inc. (Brea, CA, U.S.A.). Cell Signaling Technology (Danvers, MA, U.S.A.) supplied the horseradish peroxidase (HRP)-conjugated secondary antibody and the anti-caspase-3 antibody. An anti-TAGE antibody was prepared and purified, as previously described.¹⁸⁾

Cell Culture

Caco-2 cells were maintained in low-glucose DMEM supplemented with 10% FBS, NEAA and PS. Cells were seeded in 6-, 24-, and 96-well plates or 24-well cell culture inserts at a cell density of 1.0 × 10⁶ cells/cm². To differentiate into confluent monolayers, cells seeded in 24- and 96-well plates were cultured for 14 d, and cells seeded in cell culture inserts were cultured for 21 d. The medium was changed every 2 or 3 d. After differentiation, cells were treated with or without AG (15 mM) for 2 h and then with or without GA (0, 2.5, 5, 7.5, or 10 mM) for 24 h. AG was used to block the formation of AGEs in response to GA.¹⁹⁾

Cell Viability Assay

Cell viability was measured using the Cell Counting Kit-8 according to the manufacturer’s instructions. Briefly, the cells treated under each condition were washed twice using a cell culture medium and incubated with WST-8 for 2 h. The absorbance was measured at 450 nm using a spectrophotometer (Synergy H1, Bio Tek, Winooski, VT, U.S.A.). The viability of the control cells was taken as 100%.

Slot Blot Analysis

Slot blotting was performed to detect the total amount of TAGE in the cell extracts treated with the desired drugs in a 6-well plate. This analysis was performed as previously described with some modifications.¹⁹⁾ The treated cells were rinsed with PBS (−) and lysed in buffer (4% CHAPS, 30 mM Tris, 2 M Thiourea, 7 M Urea, and protease inhibitor cocktail). The cell extracts were then stood on ice for 5 min, centrifuged at 10000 × g at 4°C for 15 min, and the supernatant was collected as the cell extract. The protein concentrations of the samples were measured using an XL-Bradford assay kit.

For detection of TAGE, cell extracts with the same amount of protein were placed on polyvinylidene fluoride (PVDF) membranes (0.45 μm; Millipore, Burlington, MA, U.S.A.) located in a slot blot apparatus (Bio-Rad, Burlington, MA, U.S.A.). Then, the samples were blotted under vacuum conditions using an aspirator.

The membranes were blocked at room temperature for 1 h using 5% skimmed milk in PBS (–) containing 0.05% Tween 20 (PBS-T), followed by rinsing thrice with PBS-T. The membranes were shaken slowly with an anti-TAGE antibody at a dilution of 1 : 1000 with 0.5% skimmed milk in PBS (−) at 4°C overnight. After being washed twice with PBS-T, the membranes were incubated with the HRP-linked anti-rabbit immunoglobulin G (IgG) antibody at a dilution of 1 : 2000 with 0.5% skimmed milk in PBS (−) for 1 h. The membranes were washed thrice with PBS-T, and immunoreactive proteins were detected with Chemi-Lumi One L using a luminescent image analyzer (Amersham Imager 600; GE-Healthcare, Chicago, IL, U.S.A.). The immunoreactive bands on the exposed membrane were quantified using digital analyzer software (ImageJ, National Institutes of Health [NIH], Bethesda, MD, U.S.A.).

Transepithelial Electrical Resistance Measurements

Transepithelial electrical resistance (TEER) measurements were performed using EVOM3 (WPI, Sarasota, FL, U.S.A.). We measured TEER values after treatment with or without 15 mM AG for 2 h followed by treatment with or without 7.5 mM GA for 24 h. The raw data were converted to Ω × cm² based on the surface area of the cell culture inserts (0.33 cm²).

Paracellular Marker Lucifer Yellow Permeability Assay

Lucifer yellow flux measurements were performed as previously described.²⁰⁾ To put it simply, we used assay buffer for the apical side, which was HBSS containing 10 mM MES and 0.45% D-glucose (pH 6.5), and assay buffer for the basal side, which was HBSS containing 10 mM HEPES (pH 7.4) and 0.45% D-glucose. The cells were incubated with assay buffer at 37°C for 20 min. The Caco-2 cells were then incubated with assay buffer containing 125 μM lucifer yellow on the apical side and assay buffer on the basal side at 37°C. Samples were taken from the basal side at 30 and 60 min. The concentration of lucifer yellow in the samples was measured using SYNERGY HTX with the wavelengths of excitation and emission at 430 and 535 nm, respectively. The apparent permeability coefficient (P_app) of lucifer yellow was calculated according to the previously described formula:

  

$$P_{\mathrm{a}\mathrm{p}\mathrm{p}}=\frac{dQ}{dt}\times \frac{1}{A\times C_0}$$

where dQ/dt is the amount of lucifer yellow transmitted per unit time, A is the surface area of the cell culture insert membrane (0.33 cm²), and C₀ is the initial concentration of lucifer yellow on the apical side.

Immunofluorescence Staining

Immunofluorescence staining was performed as described.²⁰⁾ In brief, the treated cells in a 96-well plate were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100 solution. After blocking with PBS (−) containing 5% FBS for 30 min, the cells were placed with the first antibodies at a dilution of 1 : 100 with 5% FBS in PBS (−) at 4°C overnight. After being washed thrice with PBS (−), the samples were stained with the secondary IgG antibody at a dilution of 1 : 200 with 5% FBS in PBS (−) at room temperature for 1 h. The cell nuclei were dyed using 0.2 μg/mL DAPI in PBS (−) for 5 min and washed thrice with PBS (−), after which they were examined using a fluorescence microscope LAS-X (Leica, Wetzlar, Germany). The antibodies used in this experiment are listed in Table 1.

Table 1. Antibodies for Immunofluorescence Analysis

Target	Source	Catalog number	Species	Dilution
ZO-1	Fisher Scientific	33–9100	Mouse	1 : 100
Claudin-7	Fisher Scientific	34–9100	Rabbit	1 : 100
Anti-mouse (Alexa Fluor 488)	Fisher Scientific	A11001	Goat	1 : 200
Anti-rabbit (Alexa Fluor 568)	Fisher Scientific	A11011	Goat	1 : 200

Intracellular ROS Assessment

The fluorescent ROS indicator CellROX green reagent was used to detect ROS production in cultured Caco-2 cells in a 96-well plate. The mediums were changed with HBSS (+) containing CellROX green reagent (5 μM) and incubated with the cells at 37°C for 30 min in the dark. Then, the cells were washed with HBSS (+) once gently and observed using a fluorescence microscope LAS-X. Images were quantified for 3 randomly selected fields of view from each well, analyzed using ImageJ, and averaged.

Quantitative RT-qPCR Analysis

RNA extraction, a reverse transcription reaction, and real-time PCR analysis were performed as previously described.²⁰⁾ The housekeeping gene hypoxanthine phosphoribosyl transferase 1 (HPRT1) was used as the normalization standard for mRNA expression levels. The primers used in this experiment were as follows: HPRT1, 5′-CTT TGC TTT CCT TGG TCA GG-3′ and 5′-TCA AGG GCA TAT CCT ACA ACA-3′; nicotinamide adenine dinucleotide phosphate (NAD(P)H) oxidase 1 (NOX1), 5′-AGT CAT CCT CGC AAG TGT GC-3′ and 5′-GGT GCA AGG ATC CAC TTC CAA-3′; and gp-91^phox, 5′-TGC CTT TGA GTG GTT TGC AGA T-3′ and 5′-ATT GGC CTG AGA CTC ATC CCA-3′.

Annexin V-FITC Flow Cytometric Analysis

An Annexin V-FITC Apoptosis Detection Kit was used to identify viable, apoptotic, and necrotic cells. Briefly, both floating and trypsin-treated cells from each condition were washed twice with PBS (−) and stained according to the manufacturer’s instructions. Fluorescence-activated cell analysis was performed using the CytoFLEX system (Beckman Coulter Inc., Brea, CA, U.S.A.), and 10000 events were recorded and represented as dot plots.

Western Blotting

Western blotting was performed as previously described, with some modifications.¹⁴⁾ Similar to the slot blot analysis, treated cell extracts were collected, and protein concentrations were measured using an XL-Bradford assay kit. Equal amounts of protein samples were diluted in sodium dodecyl sulfate sample buffer containing 2-mercaptoethanol and then heated at 95°C for 5 min. The cell lysates were loaded onto AnykD polyacrylamide gradient gels (Bio-Rad, Burlington, MA, U.S.A.) and transferred onto PVDF membranes. The membranes were blocked with 5% skimmed milk in PBS-T and washed thrice with PBS-T. They were then incubated with anti-Caspase-3 antibody (1 : 1000) at 4°C overnight. Washing, secondary antibody incubation, and detection of immunoreactive proteins were performed as described in the slot blot analysis.

Statistical Analysis

Slot blot and Western blotting assays were performed on unrelated triplicate samples. The other experiments were conducted in triplicate and repeated thrice. We used a one-way ANOVA followed by Tukey’s test or Dunnett’s test for the comparison of intergroup differences. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 25.0 (IBM Corporation, Armonk, NY, U.S.A.). Data were presented as the mean ± standard deviation (S.D.), and p-values <0.05 were considered statistically significant.

RESULTS

GA Treatment Induces Cytotoxicity and TAGE Accumulation in Caco-2 Cells

Evaluation of Caco-2 cell viability after 24-h treatment with various GA concentrations revealed a concentration-dependent decrease, identifying the cytotoxic concentrations of GA for Caco-2 cells (Fig. 1a). In addition, examination of intracellular TAGE accumulation upon GA addition revealed an increased amount of intracellular TAGE for the GA concentration exceeding 7.5 mM (Figs. 1b, 1c).

Fig. 1. Cell Death and TAGE Accumulation Were Induced in GA-Treated Caco-2 Cells

(a) Dose-dependent effects of GA on the viability of Caco-2 cells using WST-8 (n = 3). Caco-2 cells were treated with 0, 2.5, 5, 7.5, or 10 mM GA for 24 h. (b, c) The total amount of TAGE in Caco-2 cells was measured using an anti-TAGE antibody (n = 3). Caco-2 cells were treated with 0, 2.5, 5, 7.5, or 10 mM GA for 24 h, and cell extracts were then prepared. The amount of TAGE in Caco-2 cells without 0 mM GA was taken as 1. (a, c) Results are expressed as mean ± S.D. A one-way ANOVA followed by Dunnett’s test was used to compare to 0 mM GA (^*p < 0.05, ^**p < 0.01, ^***p < 0.001). GA: glyceraldehyde; TAGE: toxic advanced glycation end-products.

Notably, 15 mM AG pretreatment inhibited the 7.5 mM GA-triggered decrease in the cell viability of Caco-2 cells (Fig. 2a) and significantly suppressed the GA-induced TAGE accumulation in Caco-2 cells (Figs. 2b, 2c).

Fig. 2. Cell Death and TAGE Accumulation Were Induced in Caco-2 Cells Treated with GA and AG

(a) Effects of GA and AG treatment on the viability of Caco-2 cells using WST-8 (n = 3). Cells were incubated with or without 15 mM AG treatment for 2 h, and cells were incubated with or without 7.5 mM GA for 24 h. (b, c) The total amount of TAGE in Caco-2 cells was measured using an anti-TAGE antibody (n = 3). Caco-2 cells were incubated with or without 15 mM AG treatment for 2 h, and the cells were incubated with or without 7.5 mM GA for 24 h. The amount of TAGE in Caco-2 cells with the control was taken as 1. Results are expressed as mean ± S.D. A one-way ANOVA followed by Tukey’s test was used to compare all samples (^###p < 0.001 vs. Control, ^$$$p < 0.001 vs. AG, ^&&&p < 0.001 vs. AG + GA) (a, c). GA: glyceraldehyde; AG: aminoguanidine; TAGE: toxic advanced glycation end-products.

GA Treatment Induces the Disruption of Permeability in Caco-2 Cells

We examined the permeability of the monolayer of Caco-2 cells by TAGE accumulation. The GA-exposed cells had significantly lower TEER values than the control. However, such GA-triggered decreases in TEER values were suppressed by AG pretreatment (Fig. 3a). Likewise, the GA-induced increases in lucifer yellow permeability were ameliorated by AG pretreatment (Fig. 3b).

Fig. 3. Barrier Function in Caco-2 Cells Treated with GA and AG

(a) TEER values in Caco-2 cells incubated with or without 15 mM AG treatment for 2 h, and the cells were incubated with or without 7.5 mM GA for 24 h (n = 3). (b) Permeability of lucifer yellow in Caco-2 cells incubated with or without 15 mM AG treatment for 2 h, and the cells were incubated with or without 7.5 mM GA for 24 h (n = 3). Results are expressed as mean ± S.D. A one-way ANOVA followed by Tukey’s test was used to compare all samples. (^#p < 0.05 vs. Control, ^###p < 0.001 vs. Control, ^$$$p < 0.001 vs. AG, ^&&p < 0.01 vs. AG + GA, ^&&&p < 0.001 vs. AG + GA.) GA: glyceraldehyde; AG: aminoguanidine; TAGE: toxic advanced glycation end-products.

The immunofluorescence assay conducted to determine the disruption of the tight junction proteins by TAGE accumulation showed GA-induced decreases in fluorescence intensities for ZO-1 and claudin-7, which are tight junction proteins on intestinal cells. However, the GA-exposed cells retained their protein expression levels after AG pretreatment (Fig. 4).

Fig. 4. Expression of Tight Junction Proteins in Caco-2 Cells Treated with GA and AG

Fluorescence images of ZO-1 (green) and claudin-7 (red) expression in Caco-2 cells incubated with or without 15 mM AG treatment for 2 h, and the cells were incubated with or without 7.5 mM GA for 24 h. Experiments were repeated thrice with similar results. DAPI = blue, scale bar = 100 μm. GA: glyceraldehyde, AG: aminoguanidine, TAGE: toxic advanced glycation end-products.

GA Treatment Induces ROS in Caco-2 Cells

To investigate whether ROS affects GA-triggered disruption in tight junctions, we surveyed the level of intracellular ROS in Caco-2 cells and measured it using a fluorescence probe. GA treatment induced excess ROS (Figs. 5a, 5b). GA-treated cells showed significantly higher fluorescence intensities than the control cells, but those that underwent AG pretreatment showed significantly reduced probe fluorescence intensities.

Fig. 5. ROS Production and ROS Production-Related Gene Expression in Caco-2 Cells Treated with GA and AG

(a) Fluorescence images of the cells dyed with CellROX. Caco-2 cells were treated with or without 15 mM AG for 2 h followed by with or without 7.5 mM GA for 24 h, and dyed with 5 μM of CellROX green reagent, a fluorescence probe that responded to intracellular ROS for 30 min. The experiments were repeated thrice with similar results. Scale bar = 100 μm. (b) Quantification of fluorescence intensity in fluorescence photography using Image J (n = 3). Caco-2 cells were treated with or without 15 mM AG for 2 h followed by with or without 7.5 mM GA for 24 h, and dyed with 5 μM of CellROX green reagent, a fluorescence probe that responded to intracellular ROS for 30 min. Images were quantified and averaged for 3 randomly selected fields of view from each well. Each experiment was independently performed thrice in 3 wells. (c, d) RT-qPCR analysis of the expression of genes involved in ROS production, NOX1 (c) and gp91^phox (d), in Caco-2 cells treated with GA and AG (n = 3). Caco-2 cells were treated with or without 15 mM AG for 2 h followed by with or without 7.5 mM GA for 24 h, and total RNA was then extracted. Each experiment was independently performed thrice in 3 wells. (b–d) Results are expressed as mean ± S.D., fold changes from the control value. A one-way ANOVA followed by Tukey’s test was used to compare all samples (^#p < 0.05 vs. Control, ^##p < 0.01 vs. Control, ^$$p < 0.01 vs. AG, ^$$$p < 0.001 vs. AG, ^&&p < 0.01 vs. AG + GA). GA: glyceraldehyde; AG, aminoguanidine; NOX1: NAD(P)H oxidase 1; HPRT1: hypoxanthine phosphoribosyl transferase 1.

We also measured the mRNA expression of ROS-related genes (NOX1 and gp91^phox, which are components of NAD(P)H oxidases that produce ROS). The expression levels of these genes were increased by GA treatment compared with those without GA treatment. By contrast, GA treatment after AG pretreatment significantly reduced expression levels compared to GA treatment alone (Figs. 5c, 5d).

GA Treatment Induces Necrotic Cell Death in Caco-2 Cells

We investigated the effect of GA treatment on apoptosis in Caco-2 cells by assessing Annexin V-FITC/propidium iodide (PI) staining and caspase-3 levels. Caco-2 cells were labeled with Annexin V-FITC and PI to determine the type of cell death induced by AG pretreatment and GA treatment. The percentage of apoptotic cells in GA-treated samples was similar to that in untreated cells (Fig. 6b). Conversely, the percentage of necrotic cells was significantly higher in GA-treated cells than in untreated cells. Notably, GA treatment following AG pretreatment significantly reduced the percentage of necrotic cells (Fig. 6c).

Fig. 6. Type of Cell Death in Caco-2 Cells Treated with AG and GA

(a) The effect of GA treatment on cell death was analyzed using Annexin V-FITC/PI staining and flow cytometry. Caco-2 cells were treated with or without 15 mM AG for 2 h, followed by treatment with or without 7.5 mM GA for 24 h. Collected cells were stained using the Annexin V-FITC Apoptosis Detection Kit. Regions were defined as follows: A = necrotic cells, B = viable cells, and C = apoptotic cells. Each experiment was independently performed 3 times using 3 wells. (b, c) Percentages of apoptotic (C) and necrotic (A) cells from (a) in Caco-2 cells treated with or without GA and AG. Results are expressed as mean ± S.D. Statistical comparisons were performed using one-way ANOVA followed by Tukey’s test (^###p < 0.001 vs. Control, ^$$$p < 0.001 vs. AG, ^&p < 0.05 vs. AG + GA, ^&&&p < 0.001 vs. AG + GA). GA: glyceraldehyde; AG: aminoguanidine; PI: propidium iodide.

Western blot analysis revealed a decrease in caspase-3 protein levels and the presence of slower-migrating caspase-3 bands in GA-treated Caco-2 cells. These changes were attenuated by AG pretreatment (Fig. 7).

Fig. 7. Expression of Caspase-3 in Caco-2 Cells Treated with AG and GA

(a, b) The expression level of caspase-3 in Caco-2 cells was measured using an anti-caspase-3 antibody. Caco-2 cells were treated with or without 15 mM AG for 2 h, followed by treatment with or without 7.5 mM GA for 24 h. Caspase-3 levels in control cells were set to 1. Results are expressed as mean ± S.D. Statistical comparisons were performed using one-way ANOVA followed by Tukey’s test (^##p < 0.01 vs. Control, ^$$p < 0.01 vs. AG, ^&p < 0.05 vs. AG + GA). GA: glyceraldehyde; AG: aminoguanidine.

DISCUSSION

The present study found that TAGE production caused cell death and tight junction disruption in Caco-2 cells that had differentiated into intestinal epithelial-like cells. Intracellular TAGE formation led to cell death in human hepatocyte cancer cell lines and primary cultured hepatocytes.^14,19,21,22) Human cardiac fibroblasts treated with GA also resulted in intracellular TAGE generation, ultimately causing cell death.¹⁶⁾ Furthermore, intercellular TAGE induced cell death more strongly than extracellular TAGE and may suppress autophagy via reduction of LC3-I, LC3-II, and p62 to inhibit their degradation in 1.4E7 cells—a human β-cell line.²³⁾ Similar to their results, we found that intracellular TAGE accumulation in GA-treated Caco-2 cells are cytotoxic and cause cell death.

We then investigated the effect of TAGE production on intestinal barrier function. The intestinal barrier function is one of the most important functions in maintaining a healthy intestinal tract.²⁾ Although the intestinal barrier includes the mucosal layer, tight junctions, adhesive junctions, and desmosomes,¹⁾ we focused on tight junctions in this study. The cells treated with GA showed reduced TEER values, increased permeability of lucifer yellow, and decreased expression of tight junction proteins (Figs. 3, 4). These results suggest that TAGE production results in increased permeability of substances through the paracellular pathway. Intestinal barrier loss can result from tight junction dysfunction or epithelial damage.²⁴⁾ In addition, IECs stimulated with LPS underwent apoptosis-mediated cell death and showed reduced TEER values and expression of tight junction proteins.²⁵⁾ Involvement of different AGE types in tight junction disruption in several cells and organs has been explored. However, some reports suggest that dietary AGEs can harm tight junctions, while others indicate they may protect them.^12,13) For example, the addition of methylglyoxal-derived hydroimidazolone, a dietary AGE, caused intestinal integrity dysfunction in Caco-2 cells and mice.^7,8) Otherwise, exposure to N^ε-(carboxymethyl)lysine, an AGE type, reduced mucin gene expression in the Caco-2/TC7 cell coculture model, although no effect was noted on the intestinal wall permeability.²⁶⁾ These results showed the effects of extracellular AGEs on intestinal barrier function but not of AGEs accumulated in IECs. Our findings suggest that the intracellular production and accumulation of TAGE in IECs are important for tight junction breakdown caused by epithelial damage. In addition to reports indicating that IBD in patients with diabetes is prone to exacerbation and that circulating levels of AGEs are elevated in their blood,^6,27,28) our findings suggest that AGE production in IECs can cause IBD in this patient group. The immunofluorescence staining results suggest that GA addition induced a reduction in the fluorescence intensities of ZO-1 and claudin-7, a tight junction protein, at the plasma membrane (Fig. 4).

The mechanisms behind tight junction breakdown include the decreased expression of the adherence junction and tight junction proteins, dysregulated vesicular trafficking of junctional components, and altered assembly and contractility of the junction-associated actomyosin cytoskeleton.^2,29) ROS has been implicated in intestinal barrier disruption.^1,8,10,30) In this study, significant increases in ROS production and gp91^phox and NOX1 expression levels were observed in GA-treated cells compared with non-GA-treated cells (Fig. 5). TAGE-triggered elevated ROS has been reported in other cell types.^19,31) Extracellular TAGE accumulation enhanced vascular permeability by disrupting adherent junctions and tight junctions via complex signaling, such as ROS and non-ROS pathways in human vascular endothelial cells.¹⁵⁾ Cell death via the RAGE–NAD(P)H oxidase (NOX) pathway has also been reported as a mechanism of AGE-induced ROS production.^10,32) In human arterial endothelial cells, intracellular AGE production was mediated by intracellular NOX activation, resulting in ROS overproduction.³³⁾ Flow cytometry results indicated that GA treatment might induce cytotoxicity primarily through necrosis rather than apoptosis (Fig. 6). Furthermore, Western blot analysis showed that GA treatment caused a reduction in caspase-3 levels and the appearance of slower-migrating caspase-3 bands, observed as a smear. These slower-migrating bands are presumed to be caspase-3 multimers, as glycated proteins are known to form cross-linked structures either between adjacent proteins or within a single protein’s domains.³⁴⁾ A previous study demonstrated that GA-AGE-modified caspase-3 inhibited DNA damage-induced apoptosis and led to necrotic cell death in HepG2 cells and primary hepatocytes.¹⁴⁾ Additionally, activation of NOX has been shown to induce necrotic cell death through a caspase-independent pathway.³⁵⁾ Consistent with these reports, our findings suggest the ROS pathway via TAGE accumulation as a potential cause of cell injury and barrier function disruption in IECs. Flow cytometry results revealed reduced viable cell percentages in the control group. By contrast, GA-induced necrotic cells increased among Caco-2 cells in the log-proliferative phase, a stage without apoptosis or cytotoxicity (Supplementary Fig. 1). Based on the above, we consider that the differentiation of Caco-2 cells into IEC-like cells reduced viable cell percentages compared with the cell viability results due to their susceptibility to apoptosis and flow cytometry manipulation–induced damage. However, unlike the flow cytometry results, Western blotting revealed no expression of cleaved caspase-3—an apoptosis-associated marker—possibly due to the limited detection sensitivity of the antibody used in this study. These considerations require further investigation to elucidate the pathway of GA-induced cell death in Caco-2 cells in more detail. Another possible mechanism for TAGE-induced tight junction disruption is the glycation of tight junction proteins and the cytoskeleton, but many questions remain unanswered. Further studies are needed to identify the proteins targeted for TAGE formation in Caco-2 cells.

This study indicated that intracellular TAGE formation and accumulation in IECs are involved in tight junction failure, leading to intestinal barrier dysfunction by cell injury. Increased oxidative stress by TAGE accumulation also might be implicated as a potential driver of tight junction disruption.

Acknowledgments

This work was supported by JSPS KAKENHI Grant No. JP21H04865 (for M. Takeuchi).

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1)  Dmytriv TR, Storey KB, Lushchak VI. Intestinal barrier permeability: the influence of gut microbiota, nutrition, and exercise. Front. Physiol., 15, 1380713 (2024).
• 2)  Al-Sadi R, Engers J, Abdulqadir R. Talk about micromanaging! Role of microRNAs in intestinal barrier function. Am. J. Physiol. Gastrointest. Liver Physiol., 319, G170–G174 (2020).
• 3)  Twarda-Clapa A, Olczak A, Bialkowska AM, Koziolkiewicz M. Advanced glycation end-products (AGEs): formation, chemistry, classification, receptors, and diseases related to AGEs. Cells, 11, 1312 (2022).
• 4)  Nie C, Li Y, Qian H, Ying H, Wang L. Advanced glycation end products in food and their effects on intestinal tract. Crit. Rev. Food Sci. Nutr., 62, 3103–3115 (2022).
• 5)  Phuong-Nguyen K, McNeill BA, Aston-Mourney K, Rivera LR. Advanced glycation end-products and their effects on gut health. Nutrients, 15, 405 (2023).
• 6)  Snelson M, Lucut E, Coughlan MT. The role of AGE-RAGE signalling as a modulator of gut permeability in diabetes. Int. J. Mol. Sci., 23, 1766 (2022).
• 7)  Qu W, Yuan X, Zhao J, Zhang Y, Hu J, Wang J, Li J. Dietary advanced glycation end products modify gut microbial composition and partially increase colon permeability in rats. Mol. Nutr. Food Res., 61, 1700118 (2017).
• 8)  Lim JM, Yoo HJ, Lee KW. High molecular weight fucoidan restores intestinal integrity by regulating inflammation and tight junction loss induced by methylglyoxal-derived hydroimidazolone-1. Mar. Drugs, 20, 580 (2022).
• 9)  Paparo L, Coppola S, Nocerino R, et al. How dietary advanced glycation end products could facilitate the occurrence of food allergy. J. Allergy Clin. Immunol., 153, 742–758 (2024).
• 10)  Morresi C, Cianfruglia L, Sartini D, Cecati M, Fumarola S, Emanuelli M, Armeni T, Ferretti G, Bacchetti T. Effect of high glucose-induced oxidative stress on paraoxonase 2 expression and activity in Caco-2 cells. Cells, 8, 1616 (2019).
• 11)  Kato S, Itoh K, Ochiai M, Iwai A, Park Y, Hata S, Takeuchi K, Ito M, Imaki J, Miura S, Yakabi K, Kobayashi M. Increased pentosidine, an advanced glycation end-product, in urine and tissue reflects disease activity in inflammatory bowel diseases. J. Gastroenterol. Hepatol., 23 (Suppl. 2), S140–S145 (2008).
• 12)  Jansen FAC, Fogliano V, Rubert J, Hoppenbrouwers T. Dietary advanced glycation end products interacting with the intestinal epithelium: what do we really know? Mol. Metab., 73, 101734 (2023).
• 13)  Cao C, Tang M, Zhao N, Dong S, Wu H. Effects of fish protein with glycation extent on gut microbiota and colonic barrier function in mice fed a high-fat diet. J. Funct. Foods, 85, 3439–3451 (2021).
• 14)  Sakasai-Sakai A, Takata T, Takino J, Takeuchi M. Impact of intracellular glyceraldehyde-derived advanced glycation end-products on human hepatocyte cell death. Sci. Rep., 7, 14282 (2017).
• 15)  Takino J, Sato T, Kanetaka T, Okihara K, Nagamine K, Takeuchi M, Hori T. RasGRP2 inhibits glyceraldehyde-derived toxic advanced glycation end-products from inducing permeability in vascular endothelial cells. Sci. Rep., 11, 2959 (2021).
• 16)  Takata T, Sakasai-Sakai A, Takeuchi M. Intracellular toxic advanced glycation end-products may induce cell death and suppress cardiac fibroblasts. Metabolites, 12, 615 (2022).
• 17)  Ooi H, Koriyama Y. Toxic advanced glycation end-products-dependent Alzheimer’s disease-like alternation in the microtubule system. Curr. Alzheimer Res., 20, 677–681 (2023).
• 18)  Takeuchi M, Makita Z, Bucala R, Suzuki T, Koike T, Kameda Y. Immunological evidence that non-carboxymethyllysine advanced glycation end-products are produced from short chain sugars and dicarbonyl compounds in vivo. Mol. Med., 6, 114–125 (2000).
• 19)  Sakasai-Sakai A, Takata T, Takeuchi M. Intracellular toxic advanced glycation end-products promote the production of reactive oxygen species in HepG2 cells. Int. J. Mol. Sci., 21, 4861 (2020).
• 20)  Shirai K, Qiu S, Minowa H, Hashita T, Iwao T, Matsunaga T. Air-liquid interface culture and modified culture medium promote the differentiation of human induced pluripotent stem cells into intestinal epithelial cells. Drug Metab. Pharmacokinet., 55, 100994 (2024).
• 21)  Takino J, Sato T, Nagamine K, Sakasai-Sakai A, Takeuchi M, Hori T. Suppression of hepatic stellate cell death by toxic advanced glycation end-products. Biol. Pharm. Bull., 44, 112–117 (2021).
• 22)  Kikuchi C, Sakasai-Sakai A, Okimura R, Tanaka H, Takata T, Takeuchi M, Matsunaga T. Accumulation of toxic advanced glycation end-products induces cytotoxicity and inflammation in hepatocyte-like cells differentiated from human induced pluripotent stem cells. Biol. Pharm. Bull., 44, 1399–1402 (2021).
• 23)  Takata T, Sakasai-Sakai A, Takeuchi M. Intracellular toxic advanced glycation end-products in 1.4E7 cell line induce death with reduction of microtubule-associated protein 1 light chain 3 and p62. Nutrients, 14, 332 (2022).
• 24)  Horowitz A, Chanez-Paredes SD, Haest X, Turner JR. Paracellular permeability and tight junction regulation in gut health and disease. Nat. Rev. Gastroenterol. Hepatol., 20, 417–432 (2023).
• 25)  Chen Y, Li Y, Li X, Fang Q, Li F, Chen S, Chen W. Indole-3-propionic acid alleviates intestinal epithelial cell injury via regulation of the TLR4/NF-κB pathway to improve intestinal barrier function. Mol. Med. Rep., 30, 189 (2024).
• 26)  Guibourdenche M, Haug J, Chevalier N, Spatz M, Barbezier N, Gay-Quéheillard J, Anton PM. Food contaminants effects on an in vitro model of human intestinal epithelium. Toxics, 9, 135 (2021).
• 27)  Li Z, Zhao Z, Chen S, Wang X, Wang D, Nie X, Yao Y. Ge-Gen-Qin-Lian decoction alleviates the symptoms of type 2 diabetes mellitus with inflammatory bowel disease via regulating the AGE-RAGE pathway. BMC Complement. Med. Ther., 24, 225 (2024).
• 28)  Yuan JH, Xie QS, Chen GC, Huang CL, Yu T, Chen QK, Li JY. Impaired intestinal barrier function in type 2 diabetic patients measured by serum LPS, Zonulin, and IFABP. J. Diabetes Complications, 35, 107766 (2021).
• 29)  Lechuga S, Braga-Neto MB, Naydenov NG, Rieder F, Ivanov AI. Understanding disruption of the gut barrier during inflammation: should we abandon traditional epithelial cell lines and switch to intestinal organoids? Front. Immunol., 14, 1108289 (2023).
• 30)  Bigagli E, D’Ambrosio M, Cinci L, Fiorindi C, Agostiniani S, Bruscoli E, Nannoni A, Lodovici M, Scaringi S, Giudici F, Luceri C. Impact of preoperative immunonutrition on oxidative stress and gut barrier function in surgical patients with Crohn’s disease. Nutrients, 15, 882 (2023).
• 31)  Takata T, Sakasai-Sakai A, Takino J, Takeuchi M. Evidence for toxic advanced glycation end-products generated in the normal rat liver. Nutrients, 11, 1612 (2019).
• 32)  Yamagishi S, Nakamura K, Matsui T, Inagaki Y, Takenaka K, Jinnouchi Y, Yoshida Y, Matsuura T, Narama I, Motomiya Y, Takeuchi M, Inoue H, Yoshimura A, Bucala R, Imaizumi T. Pigment epithelium-derived factor inhibits advanced glycation end product-induced retinal vascular hyperpermeability by blocking reactive oxygen species-mediated vascular endothelial growth factor expression. J. Biol. Chem., 281, 20213–20220 (2006).
• 33)  Zhang Y, Yang X, Qiu C, Liu F, Liu P, Liu Z. Matrine suppresses AGE-induced HAEC injury by inhibiting ROS-mediated NRLP3 inflammasome activation. Eur. J. Pharmacol., 822, 207–211 (2018).
• 34)  Takeuchi M. Toxic AGEs (TAGE) cause lifestyle-related diseases. Antioxidants, 13, 1372 (2024).
• 35)  Morgan MJ, Kim YS, Liu ZG. TNFα and reactive oxygen species in necrotic cell death. Cell Res., 18, 343–349 (2008).