Mast cells disrupt the duodenal mucosal integrity: Implications for the mechanisms of barrier dysfunction in functional dyspepsia

Abstract Background Functional dyspepsia (FD) is a common functional gastrointestinal (GI) disorder, but its pathophysiology is poorly understood. Mast cells (MCs) may play a critical role in the development of FD. Therefore, the aim of this study was to investigate the effect of MCs on barrier function, tight junction (TJ) proteins and related signaling pathways. Methods The expression of the TJ proteins claudin-8, ZO-1 and occludin in biopsy tissues from seven FD patients and five controls was assessed. Based on the in vivo results, we further investigated the effect of (1) MC degranulation in a coculture model of Caco-2/RBL-2H3 cells and tryptase in Caco-2 monolayers, (2) MC degranulation in the presence or absence of a PAR-2 antagonist and (3) MC degranulation in the presence or absence of an ERK1/2 signaling pathway inhibitor. The epithelial integrity of Caco-2 cell monolayers was assessed by measuring the transepithelial electrical resistance (TEER). The expression of TJ proteins was evaluated by western blotting, QT-PCR and immunostaining. Results Epithelial claudin-8, ZO-1 and occludin protein expression were significantly reduced in tissues from FD patients compared with controls. MC degranulation and tryptase decreased the TEER and reduced the expression of TJ proteins in Caco-2 cell monolayers. A PAR-2 antagonist and an ERK1/2 signaling pathway inhibitor significantly reduced the effect of MC degranulation on the TEER and TJ protein expression in Caco-2 cell monolayers. Conclusions MCs disrupt duodenal barrier function by modulating the levels of TJ proteins, and the PAR-2 and ERK1/2 signaling pathways may mediate the pathogenesis of FD.


Introduction
Mast cells (MCs), which are bone marrow-derived granulocytes, are well known to contribute to a variety of immune responses [1,2] and are quite abundant in the gastrointestinal (GI) tract. Tryptase is a natural proteinase that is mainly found within the secretory granules of MCs and is the most abundant protein in MCs. When activated by external stimuli, MCs release inflammatory factors such as tryptase into the surrounding tissues through degranulation and act on adjacent injurious receptors [3]. Tryptase also plays a role in recruiting inflammatory cells, stimulating MC degranulation and amplifying MC effects [4]. Studies have demonstrated that the total number of MCs and the number of activated and degranulated MCs are significantly higher in the duodenal mucosa of FD patients than in that of healthy volunteers (HVs) [5][6][7].
The mucosa of the duodenum is composed of a single layer of columnar epithelial cells, the underlying lamina propria and the muscularis mucosa. The duodenal epithelial barrier is essential for maintaining homeostasis in the gut because it is located between the luminal bacteria and the host's innate immune system. Tight junction (TJ) proteins, including claudins (CLDs), occludin, zonula occludens (ZO), junctional adhesion molecule (JAM) and the adherens junction protein E-cadherin, are major components of the duodenal epithelial barrier and play an important role in controlling cell polarity and adhesion [8,9]. Among these TJ proteins, claudin members [10], which are considered the backbone of the intestinal barrier, are reportedly associated with FD, because their expression is changed in FD patients [11,12], this led us to further consider the potential relationship between MCs and TJ proteins.
FD is a functional GI disorder that has an estimated prevalence of 10-40% in Western countries and 5-30% in Asia [13,14]. Although the pathophysiology of FD remains largely unknown, various mechanisms, such as abnormal GI motility, visceral or central hypersensitivity, genetic factors, increased exposure of the duodenum to acid and altered brain-gut axis interactions, have been implicated in this disease [15][16][17]. Low-grade duodenal inflammation, impaired duodenal barrier function and increased duodenal permeability were recently found to be associated with the development of FD [5,18] and GI symptoms [19]. Previous studies have shown that inflammatory mediators, such as MC products, may contribute to the impairment of intestinal barrier function [11,20]. However, little is known about the mechanism underlying the effect of MCs on barrier function in the duodenum. A previous study showed that the Caco-2/RBL-2H3 cell coculture model is a highly suitable physiological model for investigating the effect of basolateral exposure to substances on permeability [21].
To our knowledge, no study has comprehensively explored the role of duodenal MCs in FD from clinical, cellular, receptor and intracellular signaling pathway perspectives. In this study, we aimed to determine the TJ protein content in vitro and in vivo and further investigate the effect of MCs on barrier function and related signaling pathways using the Caco-2/RBL-2H3 cell coculture model.

Participants
We consecutively recruited seven patients newly diagnosed with FD and five controls who were scheduled to undergo upper GI endoscopy at The Third People's Hospital of Chengdu. The inclusion criteria were the following: (1) aged 18-65 years; (2) the chief complaint was dyspeptic symptoms that met the Rome IV [15] criteria; (3) patients completed upper GI endoscopy and epigastric ultrasounds during the study period; and (4) routine blood examination, liver function test and Helicobacter pylori (H. pylori) test conducted within the last 6 months. The exclusion criteria: (1) history of gastric ulcer, gastric cancer or other types of organic upper GI disease, disease of the pancreas or biliary tract or metabolic disorders; (2) history of abdominal surgery; (3) pregnancy preparation, pregnancy or lactation; (4) abnormal liver function due to a disease, including hepatitis B or hepatitis C or nonalcoholic steatohepatitis; (5) current steroid, antidepressant or nonsteroidal anti-inflammatory drug use; (6) severe liver, heart, kidney or respiratory related dysfunction or severe nervous system diseases or mental illness; and (7) patients who were reluctant to participate in this study.

Duodenal biopsies
During upper gastroduodenoscopy, three biopsy samples were collected with biopsy forceps (Radial Jaw3, outside diameter: 2.2 mm; Boston Scientific, Natick, MA) from the second portion of the duodenum (D2). One biopsy sample was immediately placed in RNAlater solution (Qiagen, Hilden, Germany) and maintained at À20 C until analysis of mRNA levels. One biopsy sample was snap-frozen in liquid nitrogen until analysis of protein levels. One sample was fixed in formalin and embedded in paraffin before staining for TJ proteins. This study was performed in agreement with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of The Third People's Hospital of Chengdu, and written informed consent was obtained from the participants before inclusion.

Cell culture
Human colorectal adenocarcinoma cells and rat basophilic leukemia (Caco-2/RBL-2H3) cells were purchased from the Shanghai Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China), respectively. The cells were cultured in Roswell Park Memorial Institute 1640 medium (RPMI 1640, Gibco, NY) supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) penicillin/streptomycin in an incubator containing 5% CO 2 at 37 C and subcultured in Transwell-Clear chambers (1.12 cm 2 , 0.4 mm pore size, Costar, Cambridge, MA) until the confluency reached approximately 80%. As previously reported [22,23], Caco-2 cells were allowed to differentiate to their small intestinal phenotype after being seeded on Transwell inserts and cultured for 21 d. For the Caco-2/RBL-2H3 cell coculture model, Caco-2 cells (4 Â 10 5 cells/ml) were seeded in the upper chamber of a 12-well Transwell plate. After 19 d of growth and differentiation, RBL-2H3 cells (4 Â 10 5 cells/ml) were inoculated in the lower chamber for 2 d. During coculture, the integrity of the Caco-2 cell monolayer was evaluated by measuring the transepithelial electrical resistance (TEER), and the monolayer was regarded as dense and intact if the TEER was greater than 200 XÁcm 2 .

Construction of the experimental model and various treatments
In our model, each well included an upper and a lower chamber; the upper chamber represented the luminal side of the duodenum and the lower chamber represented the subepithelial side. The cells in the lower chamber were stimulated with C48/80 (10 lg/ml). Inhibition experiments were performed by adding MK-4827 (50 nM) or SCH772984 (60 nM) to the lower chamber for 24 h. The stimulant tryptase was added to the upper and lower chambers for 4 h. Cell culture was terminated after 21 d.

Measurement of the TEER
The resistance across the epithelium was measured with an epithelial voltohmmeter (EVOM2) coupled to a chopstick electrode pair (STX2, both from World Precision Instruments, Sarasota, FL), in accordance with a previously reported method [24,25]. The value of the blank insert was subtracted to obtain the net resistance, which was multiplied by the membrane area to calculate the resistance in area-corrected units (XÁcm 2 ). Prior to measurement, the chopstick electrodes were rinsed with ethanol and equilibrated in medium at room temperature. The mean of three measurements per well was calculated, and the change in the electrical resistance is expressed as the percentage of the baseline resistance, which was the resistance measured at the zero-time point. The TEER was recorded at 0, 12 and 24 h.

Western blotting
Total protein was extracted from duodenal biopsy specimens and Caco-2 cells from the coculture model by incubating the samples in radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime, Wuhan, China) for 30 min on ice and sonicating them for 10 s. The protein concentrations were determined by the bicinchoninic acid (BCA) method using a BCA assay kit (Thermo Scientific, Rockford, IL). Equivalent amounts of protein (20 lg) were eluted in sodium dodecyl sulfate (SDS) sample loading buffer, separated by SDS-PAGE, and transferred to polyvinylidene fluoride membranes. The membranes were blocked for 1 h in 5% (w/v) defatted milk and subsequently incubated with appropriate primary antibodies (1:1000-5000 dilution) overnight at 4 C. The membranes were then washed three times with Tris-buffered saline containing 1% Tween-20 and incubated with horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit secondary antibody (1:5000 dilution). Chemiluminescence was used to visualize the protein bands with a Bio-Image Analysis System (Syngene, Frederick, MD). The density of the western blotting bands was analyzed using ImageJ software (National Institutes of Health, Bethesda, MD). All experiments were performed at least three times and representative results are shown.

Measurement of mRNA expression
Cells grown in Transwell plates for 21 d were washed three times with cold phosphate-buffered saline (PBS). Total RNA was isolated from duodenal biopsy samples or Caco-2 cells using TRIzol reagent (Sigma-Aldrich, St. Louis, MO). Total RNA was reverse transcribed into cDNA using PrimeScript RT Reagent Kit (Perfect Real Time, TaKaRa Biotechnology, Kusatsu, Japan). Quantitative real-time PCR (QT-PCR) was performed with a LightCycler 480 II instrument (Roche, Mannheim, Germany) using SYBR Green (Roche Diagnostics, Basel, Switzerland). The relative expression of these genes was determined by the 2 ÀDDCt method [26]. The levels of the target genes were normalized to the level of the reference gene GAPDH. The primers (Sangon Biotech Shanghai Co. Ltd., Shanghai, China) are listed in Table 1.

Immunofluorescence staining of TJ proteins
Duodenal biopsy specimens were trimmed, fixed, embedded in paraffin and sectioned (at a thickness of 3 lm) prior to staining for TJ-related proteins. Polarized Caco-2 cell monolayers were washed with PBS and fixed in 4% paraformaldehyde for 10 min. The fixed cells were subjected to three 3-min washes with PBS. The membrane was removed from the chamber along the edge and placed on a slide, and nonspecific binding sites were blocked with 10% goat serum in PBS for 30 min at room temperature. The cells were incubated with the following primary antibodies in 1% BSA in PBS overnight at 4 C: anti-ZO-1, anticlaudin-8, and anti-occludin (1:100 dilution). After three washes with PBS, the cells were incubated with Alexa Fluor 594-conjugated and Alexa 488-conjugated goat anti-mouse IgG secondary antibodies (1:100 dilution). The nuclei were stained with 4 0 ,6 0 -diamidino-2-phenylindole (DAPI, Antgene, Wuhan, China). The slides were imaged with a Leica DFC 7000 T camera (Leica, Wetzlar, Germany).

Statistical analysis
All values are presented as the mean ± standard deviations (SDs) of at least three replicate experiments. The data were compared between the control and treated conditions using unpaired Student's t test (two tailed) and Fisher's exact test. The statistical analyses were performed using SPSS version 25.0 software (IBM Corporation, Armonk, NY) or GraphPad Prism version 8.0 software (GraphPad Software Inc., La Jolla, CA). p < 0.05 was considered to indicate significance.

Baseline demographic and clinical characteristics of FD patients and HVs
We compared the demographic and clinical characteristics of FD patients (N ¼ 7) with those of HVs (N ¼ 5) in Table 2. The FD patients did not differ significantly from the HVs with

TJ protein expression is significantly reduced in tissue from FD patients
To determine whether the expression of TJ proteins is reduced in the duodenum, we compared the expression of claudin-8, ZO-1 and occludin in duodenal epithelium samples from FD patients (N ¼ 7) with that of these proteins in samples from HVs (N ¼ 5). The expression of the TJ proteins claudin-8, ZO-1 and occludin was significantly reduced, as revealed by the appropriate molecular size in the western blotting analysis (Figure 1(A-C)). Immunofluorescence ( Figure  1(D)) and QT-PCR also showed that the expression of claudin-8, ZO-1 and occludin was reduced in duodenal tissue (Figure 1(E-G)). Because the PCR results showed no significant difference in the claudin-1, claudin-2, claudin-3 and claudin-4 mRNA levels between the two groups (Supplemental Figure 1(A-E)), claudin-8 was selected as the research object in the following experiments.

Effects of a PAR-2 receptor antagonist on the TEER and TJ protein expression in the Caco-2/RBL-2H3 cell coculture model
Cocultured Caco-2 cells and RBL-2H3 cells were exposed to MK-4827 (PAR-2 antagonist) from the basolateral side. The influence of MC degranulation was determined by measuring the TEER. C48/80 decreased the TEER after 24 h and pretreatment with MK-4827 significantly inhibited the MC-induced decrease in the TEER (Figure 2(A)). Furthermore, treatment with C48/80 or MK-4827 alone had no effect on the TEER (data not shown). Western blotting showed that the decrease in the levels of claudin-8, ZO-1 and occludin in response to C48/80 was blocked by MK-4827 (Figure 2(B-D)). QT-PCR showed that the downregulation of claudin-8, ZO-1 and occludin expression in response to C48/80 was blocked by MK-4827 (Figure 2(E-G)).

Effects of tryptase on the TEER and TJ protein expression in Caco-2 cell monolayers
To better understand the role of tryptase in epithelial permeability, we investigated the effects of tryptase. Caco-2 cell monolayers were exposed to tryptase (100 ng/ml) from the basolateral side. The influence of tryptase was determined by measuring the TEER (Figure 3(A)). Tryptase decreased the TEER after 4 h. Western blotting showed a decrease in the levels of claudin-8, ZO-1 and occludin in response to tryptase (Figure 3(B-D)). QT-PCR also showed that claudin-8, ZO-1 and occludin expression was downregulated (Figure 3(E-G)).
Effects of an ERK pathway inhibitor on the TEER and TJ protein expression in the Caco-2/RBL-2H3 cell coculture model C48/80 decreased the TEER after 24 h and pretreatment with SCH772984 (ERK inhibitor) significantly blocked the MCinduced decrease in the TEER (Figure 4(A)). Furthermore, treatment with C48/80 or SCH772984 alone had no effect on the TEER (data not shown). The decrease in the levels of claudin-8, ZO-1 and occludin in response to C48/80 was blocked by SCH772984, as shown by western blotting (Figure 4(B-D)). QT-PCR showed that the downregulation of claudin-8, ZO-1 and occludin expression in response to C48/ 80 was blocked by SCH772984 (Figure 4(E-G)). These findings were also confirmed by immunofluorescence staining ( Figure  5). Treatment with C48/80 and SCH772984 (24 h) altered the immunofluorescence staining of claudin-8, ZO-1 and occludin. Inhibition of the ERK pathway diminished the effect of MC degranulation on epithelial integrity in the Caco-2/RBL-2H3 cell coculture model.

Discussion
This study, which used a coculture cell model, demonstrated that MC degranulation reduces the expression of TJ proteins claudin-8, ZO-1 and occludin in duodenal epithelial cells and the PAR-2 and ERK1/2 signaling pathways are involved in this process. The MC number appears to be increased in FD [27] and MCs release a wide range of inflammatory mediators and other neurotransmitters. Among these mediators, tryptase is related to duodenal TJ proteins and epithelial permeability [5,28]. Our study also proved that tryptase reduces the TEER and TJ protein expression. The diagnosis and treatment of FD have been a difficult problem and previous studies have focused only on the stomach, our study gives special attention to the duodenum [6]. The results of some previous studies using duodenal biopsies have revealed changes in TJ protein expression in FD patients [5,11,12,29,30]. Vanheel et al. [5] reported that occludin expression is decreased by tryptase release from activated MCs. Komori et al. [29] reported lower expression of ZO-1. In this study, we demonstrated that the expression of the TJ proteins claudin-8, occludin and ZO-1 was downregulated in the duodenal mucosa of patients with FD compared with that of control subjects. This decrease in the expression of ZO-1 and occludin is consistent with previous results. Claudin-1, claudin-3 and other members of claudins but not claudin-8 have been investigated in previous FD studies [11,31], and decreased claudin-8 expression has been previously found in diseases such as lymphocytic colitis and Crohn's disease [32,33]. Interestingly, the expression levels of claudin-8 on duodenal epithelial cells have not been reported. According to the in vivo findings, we provide the    first demonstration that claudin-8 expression was significantly lower in the duodenal mucosa of FD patients than in that of the control subjects. Claudin-8 is expressed in the lung, liver, skeletal muscle, kidney, testis and intestine [33]. As important TJ proteins, members of the claudin family are expressed by epithelial cells and cells of mesodermal origin and are localized on the membrane, within the cytoplasm, or within the nucleus [34]. Occludin and ZO-1 are mainly expressed in the cell membrane and differ from claudin-8, which is also expressed in the cytoplasm and nucleus [35]. The function of these proteins is sealing the paracellular space between adjacent epithelial cells from the outside environment to prevent the uncontrolled paracellular passage of luminal contents and to prevent the loss of solutes and water into the lumen.
Our in vitro experiments revealed that claudin-8, occludin and ZO-1 expression was also significantly decreased in response to MC degranulation in the Caco-2/RBL-2H3 cell model, which is in agreement with the data obtained from biopsy specimens from FD patients. This observation raised the question of which inflammatory mediator and receptor are involved in the MC-induced changes in the TEER and TJ protein expression. The major protease of human MCs is tryptase [36], and recent studies have suggested that tryptase may increase the intestinal permeability [37] and downregulate ZO-1 expression in vitro [38]. MCs have a number of receptors that cause decreases in barrier function and TJ protein expression, such as PAR-2, 5-hydroxytryptamine receptor 3 (5-HT3), transient receptor potential vanilloid 1 (TRPV1) and histamine receptor 1 (H1R) [2]. PAR-2 has been identified as a specific receptor for tryptase [39]. PAR-2 stimulation leads to rapid phosphorylation of mitogen-activated protein kinases (MAPKs), including extracellular regulated protein kinase 1/2 (ERK1/2) and p38 kinase, which have been suggested to play an important role in inflammatory responses [40][41][42].
It has been reported that the ERK1/2 signaling pathway and NF-kappa B pathways are involved in MC-and tryptaseinduced changes in the expression of the TJ proteins occludin and claudin-5 but not that of claudin-8 through the PAR-2 receptor in the brain, liver and intestine [38,[43][44][45]. Among these pathways, the ERK1/2 pathway is the most widely studied signaling pathway. Therefore, we evaluated the levels of ERK1/2 signaling-related molecules that may be involved in the expression of ZO-1, occludin and particularly claudin-8. In this study, we demonstrated that the ERK1/2 signaling pathway is involved in mediating the MC-and tryptaseinduced disruption of the TEER through PAR-2. This conclusion is based on the finding that a PAR-2 agonist and an ERK inhibitor block the detrimental effects of MCs and tryptase on the TEER and TJ protein expression. These data are consistent with those of previous studies showing that ERK activation is involved in ZO-1 and occludin expression in the esophagus [24] and intestine [38] and claudin-8 expression in the brain [43].
This study has several limitations. First, ERK n-terminal kinase belongs to the MAPK family. MAPKs are a group of serine-threonine kinases activated by dual phosphorylation of a threonine residue and a tyrosine residue by an upstream kinase(s) in response to extracellular stimuli. In our study, we did not explore other MAPKs and other pathways involved in the pathogenesis of FD, and further investigations related to the signal transduction mechanisms underlying MC-mediated alterations in TJ protein expression are warranted. Second, the number of clinical samples was small. Third, although the 21-d Caco2 cell model is widely used for small intestinal barrier studies, differences between Caco2 cell monolayers and the duodenum remain. A recent study [46] reported that human duodenal organoid-derived monolayers can serve as a barrier model for duodenal tissue. In future studies on the duodenum, we will consider using organoid-derived monolayers and more suitable cell lines for more exploration. Fourth, large-scale prospective clinical studies will be needed to verify these findings in the future.
In conclusion, MCs decrease the TEER by altering the levels of the TJ proteins ZO-1, occludin and claudin-8. This effect involves the ERK1/2 signaling pathway through PAR-2. These findings may offer insights into the underlying mechanisms of and therapeutic targets for FD.