Exploring gastroprotective effect and mechanisms of epigallocatechin gallate‐impregnated black garlic in acidic ethanol‐induced gastric ulcer model mice

Abstract

Background

Peptic ulcers, particularly gastric ulcers (GUs), are multifactorial disorders with unclear mechanisms. Black garlic (BG) has shown gastroprotective effects. This study evaluated the effects of BG and epigallocatechin gallate‐impregnated BG (EG) in acidic ethanol (AE)‐induced GU model mice. Mice were pretreated with BG or EG (50–200 mg/kg) for 7 days before ulcer induction. Oral acute toxicity evaluations for BG and EG were also conducted.

Results

Oral acute toxicity evaluations showed elevated serum renal function indices at 5 g/kg of BG and 10 g/kg of EG but without histopathological liver or kidney damage. Both BG and EG pretreatments reduced ulcer indices, with EG offering stronger and significant protection. EG pretreatments also showed higher gastric mucosal accumulations. They significantly decreased AE‐induced malondialdehyde levels and restored glutathione, superoxide dismutase, and catalase. AE‐induced serum levels of tumor necrosis factor‐alpha (TNF‐α), interleukin‐6 (IL‐6), interleukin‐1 beta (IL‐1β), as well as gastric transcript levels of TNF‐α, IL‐6, IL‐1β, cyclooxygenase‐2, inducible nitric oxide synthase, and myeloperoxidase, were suppressed. EG further reduced AE‐induced phosphorylations of extracellular signal‐regulated kinase, c‐jun‐N‐terminal kinase, p38, nuclear factor kappa‐light‐chain‐enhancer of activated B cells (NF‐κB), and inhibitor of κB.

Conclusion

EG2 showed stronger gastroprotective activity than BG by modulating oxidative stress, inflammatory response, and mitogen‐activated protein kinase (MAPK)/NF‐κB signaling pathways. This study supports the application of EG as a value‐added functional food ingredient for GU prevention. © 2025 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

INTRODUCTION

Gastric ulcers (GUs), which can cause complications, stomach and duodenal cancers, and even death, are one of the most important gastrointestinal problems. The pathogenesis of GU, including Helicobacter pylori, administration of non‐steroidal anti‐inflammatory drugs (NSAID), the accumulation of hydrochloric acid (HCl), hypoxia, ischemia, alcohol consumption, and smoking, is multifactorial but not completely elucidated. ¹ , ² Animal models of GUs were used to select extracts or compounds with gastroprotective effects. ³ Carbenoxolone (CBX), a derivative of glycyrrhetinic acid, is well known for its anti‐inflammatory properties and therapeutic effects on peptic ulcers. It promotes gastric mucosal regeneration, reduces epithelial cell turnover, and suppresses gastric acid secretion – actions that contribute to its protective and healing effects at ulcer sites. ⁴ These multifaceted mechanisms make CBX a robust benchmark for evaluating the therapeutic potential of other candidate compounds within the same pathological model.

Black garlic (BG), a functional food product derived from Allium sativum L., is produced by subjecting fresh garlic to a controlled thermal aging process under high temperature (60–90 °C) and relative humidity (RH, 70–90%) for several days to weeks without any additives. ⁵ , ⁶ During this process, Maillard reactions, polysaccharide degradation, and enzyme conversions lead to significant physicochemical changes that enhance the bioactivity of the product while reducing its pungent odor and irritancy. ⁷ , ⁸ , ⁹ Compared to raw garlic, BG is enriched with health‐promoting compounds such as S‐allyl cysteine (SAC), S‐allyl‐mercapto‐cysteine, flavonoids, and polyphenols, which are responsible for its antioxidant, anti‐inflammatory, hepatoprotective, anti‐cancer, anti‐obesity, antiallergenic, and cardio‐protective activities. ⁶ , ⁷ , ¹⁰ , ¹¹ , ¹²

Recent innovations in BG processing have been aimed at improving its functional quality and reducing the fermentation time. In particular, non‐thermal and phytochemical‐assisted pretreatments, including heat treatment, enzymatic treatment, freezing, fermentation, ultrasonic treatment, pulsed electric field, and high‐pressure processing, have been shown to modulate the formation of key bioactives during the shorten aging period. ¹¹ , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ However, epigallocatechin gallate (EGCG) impregnation has garnered interest for its ability to suppress the formation of the potentially harmful by‐product, 5‐hydroxymethylfurfural (5‐HMF), by scavenging reactive intermediates like 3‐deoxyglucosone, while simultaneously promoting the accumulation of antioxidants such as SAC and total phenolics. ¹⁸ , ¹⁹ Such optimized processing methods offer new opportunities to develop BG with tailored health benefits and improved safety profiles.

BG extracts and a synthetic SAC, PMK‐S005, have demonstrated antioxidant and anti‐inflammatory properties, they have been shown to mitigate mucosal damage or promote mucosal healing in animal models of GUs. ²⁰ , ²¹ , ²² , ²³ However, the possible signaling pathway(s) or molecular mechanisms of their gastroprotective effects remain unclear. Furthermore, catechins such as EGCG not only improve BG bioactivity through processing but can also provide synergistic gastroprotective effects due to their own radical scavenging and mucosal‐modulating properties. ¹⁸ Despite the accumulation of evidence supporting the health‐promoting effects of BG, little is known about its comparative biological efficacy of BG from different processes. Therefore, the present study aimed to evaluate and compare the gastroprotective effects and underlying mechanisms of aqueous BG extracts prepared using a conventional aging method and those subjected to soaking with EGCG prior to aging in acidic ethanol (AE)‐induced GU model mice. Antioxidative and anti‐inflammatory mechanisms, and possible signaling pathways were also analyzed. The findings will contribute to a better understanding of how processing techniques influence the functional properties of BG and their implications for gastro‐intestinal health.

MATERIALS AND METHODS

Preparation of black garlic extracts

Fresh garlic (A. sativum L.) was cultivated in Cihtong Township, Yunlin County, Taiwan. For conventional BG, whole bulbs were aged in a controlled chamber at 70 °C and 80% RH for 30 days. For EGCG‐impregnated BG (EG), garlic bulbs were impregnated with 50 g/L EGCG solution at 40 °C for 12 h and then followed by the same aging condition as BG. The aged samples were freeze‐dried, ground into powder, and extracted with distilled water (1:10 w/v) for 20 min with sonication as previously described. ¹⁸ , ¹⁹ The extracts were filtered, concentrated under vacuum, lyophilized and stored at −20 °C until use.

Animals and experimental design

Five‐week‐old male BALB/cByJNarl mice (20–22 g) were obtained from the National Laboratory Animal Center (Taipei, Taiwan). Mice were housed under specific pathogen‐free conditions (12 h of light/dark cycle, 22 ± 2 °C, 55 ± 5% RH) with free access to Laboratory Rodent Diet 5001 and RO water. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC protocol no. 2658) of Chung Shan Medical University. The mice were acclimatized in the Animal Center for more than 1 week. Before the start of the experiments, they were fasted for 12 h and weighed ≥ 25 g.

Oral acute toxicity evaluation

Mice were randomly assigned to seven groups (n = 5) according to their body weight. Except for the phosphate‐buffered saline (PBS) treatment group, mice received a single dose of BG or EG at 1, 5, or 10 g/kg body weight by intragastric administration (IG, Fig. 1(A)). Body weight, food and water intakes of each mouse were recorded during the experiment period, and then body weight changes after 14 days, food and water intakes per day per mouse were calculated. After fasting for 12 h, all mice were anaesthetized and sacrificed by overdose of isoflurane (Baxter, CA, USA) on the 15th day. Blood was immediately collected by cardiac puncture. Serum samples were collected to assess liver and renal toxicities via alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN) and creatinine (CREA) levels using an automated analyzer (Hitachi 7080; Hitachi, Tokyo, Japan), as previously described. ²⁴ Liver and kidney tissues were fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned and stained with hematoxylin and eosin (H&E) for histopathological evaluation as in a previous publication. ¹

Figure 1.

Experimental design and results of oral acute toxicity assay for control mice (PBS) and mice treated with 1, 5, and 10 g/kg of BG and EG for 14 days, respectively. (A) Experimental design and time chart of acute toxicity assay. Effects of 1, 5 and 10 g/kg of BG and EG treatments for 14 days on serum indicators of liver (B) and kidney (C). (D) No significant histopathological changes of liver and kidney tissues were demonstrated by H&E staining. Scale bars: 100 μm. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the PBS group.

Induction of GU and pretreatment regimen

Mice were randomly assigned to eight groups (n = 5) according to their body weight. Except the control group (CON) and the positive control group received CBX, mice were pretreated with BG (100 or 200 mg/kg) or EG (50, 100, or 200 mg/kg) daily for 7 days with IG. On day 8, GUs were induced by oral gavage with AE (0.3 mol/L hydrochloric acid (HCl) in 50% ethanol, 10 mL/kg) as previously established. ¹ The CBX group mice received 100 mg/kg CBX 1 h before AE administration. Furthermore, 2 h after the induction of AE, mice were sacrificed and then bloods were collected by cardiac puncture and stomachs were harvested for determination of ulcer index (UI), histological staining, and biochemical assays.

UI assessment and histopathology

Excised stomachs were opened along the higher curvature, rinsed with PBS, and images of the stomach were captured with a digital camera. UI was calculated as the percentage of the total mucosal ulcerated area relative to the total gastric area using ImageJ software. Gastric tissues were fixed, embedded, and stained with H&E to evaluate mucosal damage. Periodic acid–Schiff (PAS) staining was performed to assess mucosal glycoprotein distribution. ¹

Oxidative stress and antioxidant markers

Malondialdehyde (MDA) and reduced glutathione (GSH) levels and activities of superoxide dismutase (SOD) and catalase (CAT) activities were measured in gastric tissues for every mouse using commercial assay kits (Elabscience Biotechnology Inc., Bethesda, MD, USA), following manufacturer protocols. ¹ , ²⁴

Inflammatory cytokines and enzyme assays

Serum levels of tumor necrosis factor‐alpha (TNF‐α), interleukin‐1 beta (IL‐1β), and interleukin‐6 (IL‐6) for every mouse were quantified using ELISA kits (Invitrogen Inc., Carlsbad, CA, USA). Nitrite concentrations in the sera were determined using the Griess reagent kit (Sigma‐Aldrich, St Louis, MO, USA). Gastric myeloperoxidase (MPO) activities were measured using an MPO activity assay kit purchased from Abcam Corp. (Cambridge, UK).

Measurements of gastric mRNA levels of inflammatory cytokines by real‐time PCR

Total messenger RNAs (mRNAs) for every mouse were extracted from each stomach using Tri Reagent (Sigma‐Aldrich) and then reverse transcribed using the Superscript III system (Life Technologies Corp., Carlsbad, CA, USA) and the oligo‐dT primer. Levels of TNF‐α, IL‐1β, IL‐6, cyclooxygenase‐2 (Cox‐2), inducible nitric oxide synthase (iNos), and glyceraldehyde 3‐phosphate dehydrogenase (GAPDH) mRNAs were detected by real‐time polymerase chain reaction (PCR) performed on the Bio‐Rad CFX Real‐Time PCR detection system with Fast SYBR Green Master Mix (Thermo Fisher Scientific Inc., Waltham, MA, USA). The primers used are listed in Table 1. GAPDH mRNA levels were used as internal controls for assays. Relative mRNA levels were quantified using the comparative 2^(−ΔΔCt) method and normalized against GAPDH mRNA levels. Relative mRNA levels of TNF‐α, IL‐6, and IL‐1β in the CON group were defined as 1.0.

Table 1.

Primers used for real‐time polymerase chain reaction (PCR)

Target transcript	Forward or backward	Sequences
TNF‐α	Forward	5′‐CTGTAGCCCACGTCGTAGC‐3′
	Backward	5′‐TTGAGATCCATGCCGTTG‐3′
IL‐1β	Forward	5′‐TGCCACCTTTTGACAGTGATG‐3′
	Backward	5′‐TGTGCTGCTGCGAGATTTGA‐3′
IL‐6	Forward	5′‐TCTGGGAAATCGTGGAAATGAG‐3′
	Backward	5′‐TCTCTGAAGGACTCTGGCTTTGTC‐3′
Cox‐2	Forward	5′‐GAGTGGGGTGATGAGCAACTATTCC‐3′
	Backward	5′‐GGCAATGCGGTTCTGATACTGG‐3′
iNos	Forward	5′‐TGGAGCGAGTTGTGGATTGTCC‐3′
	Backward	5′‐AGCCTCTTGTCTTTGACCCAGTAGC‐3′
GAPDH	Forward	5′‐GTGTTCCTACCCCCAATGTGT‐3′
	Backward	5′‐ATTGTCATACCAGGAAATGAGCTT‐3′

Note: TNF‐α, tumor necrosis factor‐alpha; IL‐6, interleukin‐6; IL‐1β, interleukin‐1 beta; Cox‐2, cyclooxygenase‐2; iNos, inducible nitric oxide synthase; GAPDH, glyceraldehyde 3‐phosphate dehydrogenase.

Western blot analysis

Total gastric proteins for every mouse were extracted using RIPA buffer (Sigma‐Aldrich) and quantified using a Bio‐Rad Protein Assay (Bio‐Rad Laboratories, Inc., Hercules, CA, USA). Next, 60 μg of proteins were subjected to 12% SDS‐PAGE (sodium dodecyl sulphate‐polyacrylamide gel electrophoresis), transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA) and probed with antibodies against Cox‐2, total and phosphorylated forms of extracellular signal‐regulated kinase (ERK), c‐jun‐N‐terminal kinase (JNK), p38, and inhibitor of κB alpha (IκBα), phosphorylated nuclear factor kappa‐light‐chain‐enhancer of activated B cells (NF‐κB) p65, and β‐actin. The bands were visualized using enhanced chemiluminescence (Biokit Biotechnology Inc., Miaoli, Taiwan) and quantified using AlphaEase FC 6.0 software (BIOZ, Los Altos, CA, USA). The levels of β‐actin were served as loading controls for normalization of band densities.

Primary antibodies to NF‐κB p65 (GTX102090), p38 (GTX110520), and phospho‐p38 (GTX486146) were obtained from GeneTex Inc. (Irvine, CA, USA). Primary antibodies to ERK (#9102), phospho‐ERK (#9101), JNK (#9252), phospho‐NF‐κB p65 (#3033), Cox‐2 (#4842), IκBα (#4812), and phospho‐IκBα (#2859) were obtained from Cell Signaling Technology Inc. (Beverly, MA, USA). Anti‐phospho‐JNK antibody (sc‐6254) was obtained from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA), and anti‐β‐actin monoclonal antibody (AC‐15) was purchased from Novus Biologicals (Centennial, CO, USA).

Immuno‐histochemical examination

Gastric paraffin sections for every mouse were incubated in hydrogen peroxide block (TA‐060‐HP; Epredia Technology Inc., Taoyuan, Taiwan), then blocked with protein block (TA‐060‐PBQ; Epredia Technology Inc.). The sections were probed with NF‐κB primary NF‐B antibody (1:200 dilution, #8242; Cell Signaling Technology Inc.), followed by horseradish peroxidase (HRP) conjugated rabbit antibody (D39; Origene Technology Inc., Rockville, MD, USA). Sections were developed with DAB (TA‐060‐QHSX and TA‐002‐QHCX; Epredia Technology Inc.) and counterstained with hematoxylin. Photographs were obtained using the TissueFax system (TissueGnostics GmbH, Vienna, Austria) and quantification was performed according to the scoring system. ²⁵ Multiply the proportion of high positive signals by 3, the proportion of medium positive signals by 2, and the proportion of low positive signals by 1, and adding them together to get the total value of each mouse in each group.

Statistical analysis

All data are presented as mean ± standard deviation (SD) using Microsoft Office Excel 2016 software (Microsoft Crop., Redmond, WA, USA). Statistical analyses were determined using SPSS for Windows, version 25 (SPSS, Inc., Chicago, IL, USA) and one‐way analysis of variance (ANOVA) followed by Tukey's post hoc comparison. The value of P < 0.05 indicated statistical significance.

RESULTS

Evaluating potential toxicity for oral administration of EG or BG. Initially, we conducted oral acute toxicity analyses of EG and BG by IG of EG and BG at doses of 1, 5, and 10 g/kg in mice, respectively (Fig. 1(A)). After 14 days, body weights, water intakes and food consumptions were recorded, and serum indicators for liver and kidney injuries, including ALT, AST, BUN, and CREA, were analyzed. Changes in body weight after 14 days and water intakes and food consumption per day per mouse for different experimental groups were calculated (Table 2). Compared to the PBS group, mice treated with 5 g/kg BG exhibited significant decreases in body weight changes due to a significant reduction of water intake. However, mice treated with 10 g/kg of BG exhibited significant increases in body weight changes but significant reductions in food and water intakes. However, all EG‐treated mice showed significant reductions in body weight changes, while their water and food intakes were higher than those of the PBS‐treated mice. When comparing BG and EG treated mice at the same dose, the body weight changes in the 1 and 10 g/kg EG‐treated mice were significantly lower than those of BG treated mice, whereas the food intakes in the 5 and 10 g/kg EG‐treated mice were significantly higher than those in the BG‐treated mice, and water intakes in the 5 g/kg EG‐treated mice were significantly higher than those in the BG‐treated mice. In summary, EG treatments promote water and food intakes, while effectively reducing the increases of body weight.

Table 2.

The calculated body weight changes after 14 days, and food and water intakes per day per mouse of different groups of mice

Group	Body weight change (g)	Food intake (g)	Water intake (mL)
PBS	2.18 ± 0.46	4.36 ± 0.36	4.61 ± 0.57
BG (1 g/kg)	1.60 ± 0.35	6.08 ± 1.22	4.03 ± 0.57
BG (5 g/kg)	1.25 ± 0.45 a	4.48 ± 0.02	3.84 ± 0.04 a
BG (10 g/kg)	3.85 ± 1.20 a	3.80 ± 0.21 a	3.32 ± 0.42 a
EG (1 g/kg)	0.81 ± 0.16 a , b	5.60 ± 1.14	5.19 ± 0.94
EG (5 g/kg)	0.90 ± 0.62 a	6.02 ± 0.78 a , b	5.15 ± 1.25 b
EG (10 g/kg)	0.90 ± 0.62 a , b	5.54 ± 1.25 b	4.61 ± 1.87

Note: EG, epigallocatechin gallate‐impregnated black garlic.

^a (P < 0.05) denotes significant differences compared with the phosphate‐buffered saline (PBS) group.

^b (P < 0.05) denotes significant differences compared with the same dose of the black garlic (BG) group.

Results of serum liver indexes did not show significant differences in both ALT and AST activities in all experimental groups (Fig. 1(B)). Otherwise, BUN levels of 5 g/kg BG and CREA levels of 5 and 10 g/kg BG and 10 g/kg EG were significantly elevated compared to the PBS group (Fig. 1(C)). However, no abnormalities were observed in these kidney tissues by histopathological analyses (Fig. 1(D)). In summary, even at an extremely high dose of 10 g/kg, neither BG nor EG showed obvious toxicity to the liver and kidneys of mice. While doses of BG at 5 g/kg and above significantly elevated serum renal function indices, which may indicate potential concerns regarding kidney health of the extremely high BG dose.

Gastroprotective effects of BG and EG on AE‐induced GU

After 1 week of IG of various doses of BG or EG, GUs were induced in mice using AE to evaluate and compare the gastroprotective effects of EG and BG. AE caused extensive hemorrhagic necrosis in the gastric mucosa, leading to severe hemorrhagic ulcers in mice in the AE group (Fig. 2(A), AE). Compared to the AE group, the severity of hemorrhagic necrosis was slightly reduced in the BG pretreatment groups (Fig. 2(A), BG100 + AE, BG200 + AE), though the differences were not significant and there was no dose‐dependent effect (Fig. 2(B)). On the contrary, the EG pretreatment groups showed obvious reductions in hemorrhagic necrosis (Fig. 2(A), EG50 + AE, EG100 + AE, EG200 + AE) and showed dose‐dependent gastroprotective effects (Fig. 2(B)). Both the EG100 + AE and EG200 + AE groups showed significant decreases in UI. Furthermore, the UI in the CBX pretreatment group was also significantly lower than in the AE group.

Figure 2.

(A) Macroscopic gross lesions of gastric mucosa. (B) The ulcer index (UI, percentage of ulcer area) was calculated for each group of mice. PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 100 mg/kg (BG100) or 200 mg/kg (BG200) BG, and mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100) or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

Gastroprotective effects of BG and EG on AE‐induced gastric mucosal histopathology

Histopathological examinations were performed to demonstrate the protective effects of EG or BG on the gastric mucosa. After H&E staining, normal mucosa is lined with regular straight gastric cavities containing pale eosinophilic acid‐secreting parietal cells secreting eosinophilic acid in the middle sections, and the darker staining zymogen secreting chief cells line the basal region in glandular stomach in the CON and CBX + AE (Fig. 3). Otherwise, multifocal degeneration/necrosis with hemorrhage in the mucosal layer of the stomach was observed in the AE group (Fig. 3, AE). Pretreatments with EG or BG alleviated AE‐induced mucosal hemorrhage, edema, and inflammation compared to the AE group.

Figure 3.

Histopathological findings of stomachs in the AE‐induced GU for 2 h in mice. PBS‐treated mice (CON), mice treated with AE only (AE), mice pretreated with 100 mg/kg (BG100 + AE) or 200 mg/kg (BG200 + AE) BG before AE treatment, and mice pretreated with 50 mg/kg (EG50 + AE), 100 mg/kg (EG100 + AE) or 200 mg/kg (EG200 + AE) EG before AE treatment, and mice pretreated with 100 mg/kg CBX before AE treatment (CBX + AE). Arrows indicate multifocal degeneration/necrosis of the gastric mucosal layer with associated hemorrhage. The dashed line indicates the area of the affected region. Scale bars: 100 μm.

Gastroprotective effects of EG and BG on AE‐induced gastric mucosa

In the highly acidic environment of the stomach, the gastric mucosa protects the gastric tissues from acidic injury. We analyzed glycoproteins in the mucosa using PAS staining, which could also indicate mucosal damage. PBS treated gastric tissue exhibited strong deep magenta signals in the mucosal layer (Fig. 4, CON), while the AE group showed much weaker signals than the CON group, indicating mucosal damage (Fig. 4, AE). Additionally, although the BG100 and BG200 pretreated groups showed signals, they were lower than those of the CON group (Fig. 4, BG100 + AE, BG200 + AE). The EG and CBX pretreated groups exhibited significantly higher signals than the BG pretreated groups, indicating that the EG groups show more protective effects on the gastric mucosa compared to the BG groups (Fig. 4, EG50 + AE, EG100 + AE, EG200 + AE, and CBX + AE). These results are consistent with the finding that the UI of the EG treatment groups is lower than those of the BG treatment groups (Fig. 2(B)). In summary, since the EG treatment groups show lower UI and more gastric mucosa than the BG treatment groups at the same dose, we conclude that EG is more effective in gastric protection than BG for AE‐induced GU. The potential mechanism(s) and signaling pathway(s) of EG in gastric protection were analyzed in the following sections.

Figure 4.

Histological examination of stomach tissues with PAS staining. The red area represents distribution of glycoproteins in the mucosa. PBS‐treated mice (CON), mice treated with AE only (AE), mice treated with 100 mg/kg (BG100 + AE) or 200 mg/kg (BG200 + AE) BG before treatment with AE, and mice treated with 50 mg/kg (EG50 + AE), 100 mg/kg (EG100 + AE) or 200 mg/kg (EG200 + AE) EG before treatment with AE, and mice treated with 100 mg/kg CBX before treatment with AE (CBX + AE). Scale bars: 100 μm.

Gastroprotective effects of EG on AE‐induced lipid peroxidation and antioxidative enzyme activities

wIn our subsequent mechanistic analyses, we focused exclusively on EG rather than BG. This decision was based on the lack of significant gastroprotective effects observed for BG in our AE‐induced GU model, as well as the superior efficacy demonstrated by EG in the same model. Additionally, prioritizing EG aligns with ethical guidelines to minimize the use of laboratory animals by avoiding redundant or uninformative experimental groups.

Reactive oxygen species (ROS), generated by ethanol metabolism, react with fatty acids in the cell membrane to produce MDA. Therefore, high levels of MDA indicate more severe mucosal tissue damage in AE‐induced GU. Our results showed that the AE group induced significant increases in MDA levels compared to the CON group, while all the EG and CBX treatment groups significantly reduced MDA levels, and EG showed a dose‐dependent reducing effect (Fig. 5(A)). Ethanol‐induced ROS can be neutralized by the tissue antioxidant, GSH, and antioxidative enzyme activities may also be affected. Compared to the CON group, the AE group significantly decreased GSH levels (Fig. 5(B)) and significantly reduced the activities of SOD (Fig. 5(C)) and CAT (Fig. 5(D)). Compared to the AE group, the EG treatment groups showed higher levels of GSH, only EG100 significant, and higher SOD and CAT, both EG100 and EG200 significant. The positive therapeutic drug, CBX, also significantly increased GSH level and both SOD and CAT activities.

Figure 5.

Analyses of gastric MDA levels (A), GSH levels (B), SOD activities (C), and CAT activities (D). PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100), or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg of CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

Gastroprotective effects of EG on AE‐induced nitric oxide (NO) and MPO

Nitric oxide (NO) is crucial for maintaining the integrity of the gastric mucosa, as it regulates blood flow, mucus secretion, and acid production, which are essential for protecting the stomach lining. ²⁶ AE treatment significantly reduced gastric nitrate levels (Fig. 6(A)). Compared to the AE group, the EG100 treatment group significantly increased nitrate levels, indicating that EG100 enhanced gastric mucosa as shown in Fig. 4. CBX treatment also increased nitrate levels but not significantly (Fig. 6(A)). MPO is a marker of ethanol‐induced neutrophil infiltration in the gastric mucosa and is associated with the appearance of AE‐induced gastric lesions. ²⁷ Our results showed that AE treatment significantly induced MPO expression (Fig. 6(B)). Compared to the AE group, all pretreated groups with EG and CBX showed significant reductions in MPO activities, in which the EG100 treatment group showed the lowest MPO activity. Since the EG100 pretreated group shows significant increases in gastric nitrate levels and the lowest MPO activity, EG of 100 mg/kg may be the optimal gastroprotective dose among the three tested doses.

Figure 6.

Analyses of (A) nitrate concentration and (B) MPO activity in gastric tissues. PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100), or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg of CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

Gastroprotective effect of EG on AE‐induced pro‐inflammatory cytokines

One of the pathological mechanisms of AE‐induced GU is the imbalance in the expression of pro‐inflammatory and anti‐inflammatory cytokines. ¹ , ²⁸ To demonstrate whether EG treatment can reduce expressions of pro‐inflammatory cytokines in AE‐induced GU, we analyzed the levels of pro‐inflammatory factors – including TNF‐α, IL‐1β, and IL‐6, in mouse sera. Compared to the CON group, AE treatment significantly increased serum levels of all TNF‐α, IL‐1β, and IL‐6 by 1.56, 6.25, and 9.7‐folds, respectively (Fig. 7(A)). Compared to the AE group, only the EG100 pretreated group showed a significant reduction in serum TNF‐α levels, both the EG50 and EG100 pretreated groups showed significant reductions of serum IL‐1β levels, and all the EG pretreated groups showed significant reductions of serum IL‐6 levels (Fig. 7(A)). The CBX pretreatments also showed significant reductions in serum TNF‐α, IL‐1β, and IL‐6 levels. Therefore, both EG and CBX pretreatments could significantly reduce serum levels of pro‐inflammatory cytokines induced by AE treatment.

Figure 7.

Analyses of serum levels (A) and gastric transcript levels (B) of TNF‐α, IL‐1β, and IL‐6. PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100), or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg of CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

Furthermore, real‐time quantitative PCR was used to detect gastric mRNA levels of TNF‐α, IL‐1β, and IL‐6. Compared to the CON group, AE treatment significantly increased gastric mRNA levels of TNF‐α, IL‐1β, and IL‐6 by 7.6 times, 33.4 times, and 64.3 times, respectively (Fig. 7(B)). Similarly, compared to the AE group, all pretreated EG groups showed significant reductions in gastric TNF‐α and IL‐1β levels, and both the pretreated EG100 and EG200 pretreated groups showed significant reductions in gastric IL‐6 levels (Fig. 7(B)). The CBX pretreatments also showed significant reductions of gastric TNF‐α, IL‐1β, and IL‐6. Interestingly, reductions in IL‐6 levels by EG are dose‐dependent. In summary, both EG and CBX pretreatments could reduce the expressions of pro‐inflammatory cytokines induced by AE treatment in both sera and gastric tissues.

Gastroprotective effect of EG on the AE‐induced Cox‐2 and iNos induced by AE

The inflammatory response induced by AE is also associated with Cox‐2 and iNos. ²⁹ AE increases Cox‐2 expressions in gastric tissues. Our results showed that AE induced Cox‐2 RNA and protein expressions (Fig. 7(A),(C)). Although all pretreatment groups with EG reduced Cox‐2 expressions, only the group pretreated with EG100 pretreated group significantly reduced both transcript and protein levels. CBX pretreatment also significantly reduced both transcript and protein levels of Cox‐2.

Furthermore, studies have shown increased expressions of iNos in patients with peptic ulcers. ³⁰ Our results in Fig. 7(B) indicated that AE treatment significantly induced iNos expressions, while pretreatments with EG and CBX significantly reduced these elevations. Furthermore, EG pretreatments inhibit the AE‐induced increases of iNos levels in a dose‐dependent manner with higher doses of EG showing better effects. These results are consistent with previous results (Figs. 6(B) and 7), AE treatment induces an inflammatory response, leading to the expressions of MPO, iNos and Cox‐2, while EG and CBX pretreatments reduce this inflammatory response, thereby decreasing the expressions of MPO, iNos and Cox‐2.

Gastroprotective effects of EG on AE‐induced MAPK signaling

The AE‐induced inflammatory response is known to be related to the mitogen‐activated protein kinase (MAPK) signaling pathway. ²⁸ In this study, the relationship between the gastroprotective effects of the EG and MAPK signaling pathway has been studied. Our results showed that AE treatment significantly increased the ratios of phosphorylated MAPK to total MAPK family proteins, including ERK, JNK, and p38, compared to the CON group (Fig. 9). Compared to the AE group, the ratios of phosphorylated MAPKs to total MAPKs were reduced in all EG pretreatment groups. However, only EG100 pretreatment showed significant reductions in all MAPKs. CBX pretreatment also significantly reduced all ratios of phosphorylated MAPKs to total MAPKs. Therefore, pretreatments with EG and CBX can reduce AE‐induced MAPK signaling, thereby reducing the inflammatory response.

Figure 9.

Analyses of gastric expression levels of total and phosphorylated MAPKs. Expression levels of β‐actin were used as internal controls. The ratios of p‐ERK to ERK (upper right) p‐JNK to JNK (middle right), and p‐p38 to p38 (lower right) were calculated. PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100), or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg of CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

Gastroprotective effects of EG on AE‐induced NF‐κB signaling

NF‐κB is a transcription factor associated with the inflammatory response. To analyze the molecular mechanism by which EG pretreatment reduces the AE‐induced inflammatory response, we first examined NF‐κB expression using immuno‐histochemical staining (Fig. 10(A)). Statistical analysis of signal intensity and area showed that AE treatment significantly induced NF‐κB expressions compared to the CON group (Fig. 10(A), lower right panel). All pretreatments with EG and CBX significantly reduced the AE‐induced NF‐κB expressions. Furthermore, we analyzed the expression levels of proteins related to the NF‐κB signaling pathway using Western blots. AE treatment obviously promoted the expressions of p‐IκB and p‐NF‐κB (Fig. 10(B),(C)). The ratios of p‐IκB to IκB and p‐NF‐κB to β‐actin were significantly increased. Only the EG100 pretreatment group showed significant reductions in AE‐induced ratios of p‐IκB to IκB, but all EG pretreatment groups significantly reduced the AE‐induced ratios of p‐NF‐κB/β‐actin. CBX pretreatment also significantly reduced the ratios of p‐IκB to IκB and p‐NF‐κB/β‐actin. These results indicate that EG and CBX pretreatments effectively reduce AE‐induced NF‐κB activations. Among the groups pretreated with EG, EG100 shows the highest ratios in reducing p‐IκB and p‐NF‐κB expressions, suggesting that EG100 can most effectively reduce inflammatory responses by inhibiting the AE‐induced MAPK and NF‐κB signaling pathways induced by AE.

Figure 10.

Analyses of gastric expression levels of total NF‐κB (A), total and phosphorylated IκBα (B), and phosphorylated NF‐κB (C). (A) Immuno‐histochemical analyses of total NF‐kB expressions. Scale bars: 100 μm. (B) Western blot analyses of total and phosphorylated IκBα (B), and phosphorylated NF‐κB (C) expressions in gastric tissues. Expression levels of β‐actin were used as internal controls. The ratios of p‐IκBα to IκBα (B) and p‐NF‐κB to β‐actin (C) were calculated. PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100), or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg of CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

DISCUSSION

Ethanol rapidly permeates the gastric mucosa, causing membrane damage, cell loss, and erosion. Since the symptoms of ethanol‐induced GU in mice closely resemble those of humans, the ethanol‐induced GU model is one of the commonly used models for GU research. ³¹ The presence of HCl increases membrane permeability, exacerbating ethanol‐induced oxidative stress, leading to corrosive mucosal injury and cell necrosis. Consequently, gastric mucosal changes induced by AE are more pronounced than those caused by ethanol alone. ³² Our findings reveal that AE treatment induces significant mucosal hemorrhagic necrosis (Fig. 2), which is consistent with previous studies. ¹ , ³ The results of histopathology and PAS staining indicate that AE treatment causes cell death and decreases mucus secretion (Figs. 3 and 4). Pretreatments with BG and EG reduce AE‐induced necrosis (Fig. 3) and increase mucus secretion (Fig. 4). It is hypothesized that increased mucus secretion is one of the protective mechanisms of BG or EG against AE‐induced GU. Additionally, the EG pretreatment groups demonstrate stronger gastroprotective effects than the BG pretreatment groups with the same doses against AE‐induced GU. Therefore, the results in this study demonstrate and compare the gastroprotective effects of EG and BG, indicating that EG is more superior than BG in AE‐induced GU model mice.

Ethanol metabolism generates ROS, which react with membrane lipids to produce MDA. ³³ The increase in gastric MDA content is associated with exacerbated gastric mucosal damage, making MDA a marker for oxidative stress in gastric tissues induced by AE. Although exogenous AE induces ROS in gastric tissues, the antioxidant system within gastric tissues can neutralize ROS. However, AE‐induced ROS leads to gastric tissue damage and necrosis, resulting in a reduction in SOD activity. This reduction facilitates the accumulation of superoxide anions (O²⁻), causing irreversible damage. Elevated oxidative stress also depletes the GSH level and decreases CAT activity. Our study demonstrated that AE treatment induces MDA production and reduces GSH level and both SOD and CAT activities (Fig. 5), causing irreversible damage to the antioxidant system. However, pretreatment with EG significantly reduces MDA level and restores the level of GSH level and both of SOD and CAT activities. These findings indicate that the antioxidant capacity constitutes one of the molecular mechanisms of EG, thus protecting against AE‐induced gastric injury.

The fermentation process of BG not only reduces the pungency of raw garlic, but also enhances its antioxidant capacity, immune modulating effects, and anticancer activity, making it a milder and nutritionally enriched alternative. ¹⁰ Previous studies have shown that BG mitigates a NSAID, indomethacin, induced GU primarily by decreasing MDA levels and increasing antioxidative enzyme activities, particularly SOD and CAT. ²⁰ , ²³ However, the high dose of BG (200 mg/kg) does not show a significant gastroprotective effect in AE‐induced GU model mice (Figs. 2 and 3) as in previous GU model rats. It may be due to differences in the induced GU model animals, the dosages used, and the administration time periods. Previous investigations were conducted by indomethacin‐induced GU models in rats over 30 days of BG at doses of 100 and 200 mg/kg ²⁰ or at a single high dose of BG (275 mg per rat). ²³ In contrast, the current study employed an AE‐induced ulcer model in mice by a more aggressive mucosal damage mechanism, and BG was administered at 100–200 mg/kg for only 7 days. Since the dose conversion between animals and human based on body surface area is 12.3 for mice and 6.2 for rats, ³⁴ the relative dose used in mice should be 1.98‐fold of dose used in rats. The dosages employed in the two previous studies were at least 1.98 times greater than those administered for BG in this investigation. Moreover, indomethacin disrupts the Cox pathway and induces oxidative stress, which BG can counteract; however, AE induces multifaceted tissue injury involving direct chemical damage, inflammation, and oxidative stress, potentially requiring stronger bioactive support. Additionally, EGCG‐impregnation enriches beneficial compounds like SAC while reducing the content of potentially harmful 5‐HMF, conferring a clear advantage in this specific model. Collectively, these factors likely explain divergent outcomes and support the focus on EG in the current mechanistic analyses.

The aging process of BG also generates a potentially harmful compound such as 5‐HMF. ³⁵ Our research team has shown that the product of the fermentation process, EG, has reduced 5‐HMF levels while enhancing antioxidant capacity. ¹⁸ Therefore, EG retains the antioxidant potential of BG in protecting against GUs while simultaneously reducing 5‐HMF levels, thus enhancing its safety for consumption. ¹⁸ Furthermore, a synthetic SAC, PMK‐S005, prevents both ethanol‐induced and NSAIDs‐induced acute gastric mucosal damages in rats through anti‐inflammatory effects and enhancement of the gastric defense system. ²¹ , ²² Thus, EG shows superior gastroprotective effects against AE‐induced GU than BG in this study, which may account for the decrease of 5‐HMF and the increase of SAC in EG. Interestingly, the results of body weight changes and food and water intake in the oral acute toxicity indicate that the consumption of an extremely large amount of EG causes increases in food and water intakes but decreases in body weight gains in mice (Table 2). The promoting effect of EG on metabolism needs further studies.

In addition to the antioxidant system, the intrinsic defense mechanism of gastric tissue includes NO, which plays a protective role in the gastric mucosa by promoting blood flow, inhibiting leukocyte adhesion and activation, and reducing inflammatory responses. ³⁰ Studies have shown that physiologically functional NO is synthesized by neuronal NO synthase or endothelial NO synthase, with low levels of NO sufficient to maintain normal physiological functions. It has been reported that the development of gastric diseases is associated with a reduction in constitutive NO synthases in the gastric mucosa. ²⁶ Our study demonstrated that AE treatment dramatically reduced gastric NO levels (Fig. 6(A)), while pretreatment with EG100 significantly increased NO production, indicating improved gastric protection. These findings are consistent with previously results of Camellia japonica against GU published by Akanda and Park. ³⁶ However, NO produced by iNos can combine with free radicals to trigger intestinal inflammatory responses and contribute to ulcer formation through ROS generation. ³⁷ Increased iNos expression has been observed in patients with chronic ulcerative colitis and peptic ulcers. ³⁸ Similarly, in our AE‐induced GU model, an up‐regulation of gastric iNos expression is observed, which is significantly reduced by EG pretreatment (Fig. 8(B)). These results also align with findings from previous studies of C. japonica and α‐mangostin. ³⁶ , ³⁹

Figure 8.

Analyses of gastric transcript levels of Cox‐2 (A) and iNos (B), and protein levels of Cox‐2 (C). PBS‐treated mice (CON) and mice treated with AE alone (AE). Mice pretreated with 50 mg/kg (EG50), 100 mg/kg (EG100), or 200 mg/kg (EG200) EG, and mice pretreated with 100 mg/kg of CBX (CBX), then treated with AE to induce GU were underlined. * (P < 0.05), ** (P < 0.01), and *** (P < 0.001) denote significant differences compared to the AE group.

Ethanol‐induced ROS result in mucosal damage, which not only triggers the protective effects of the endogenous antioxidant system, but also initiates an inflammatory response characterized by the release of pro‐inflammatory cytokines such as TNF‐α, IL‐6, and IL‐1β. ⁴⁰ , ⁴¹ Macrophages secrete TNF‐α, leading to neutrophil infiltration into the gastric mucosa. Consequently, inflammatory diseases, including mucosal inflammation, are associated with elevated levels of TNF‐α in sera. ⁴² In addition, elevated IL‐6 levels at inflammatory sites activate neutrophils, which release ROS. ROS contribute to local tissue damage in GU through oxidative pathways. ⁴³ Therefore, reducing the productions of TNF‐α and IL‐6 can decrease neutrophil infiltration and subsequently reduce ROS production. ⁴² Studies have also indicated that neutrophil infiltration is associated with the development of acute gastritis. ⁴⁴ MPO, an enzyme found in neutrophils, serves as a marker for neutrophil infiltration. ⁴⁵ Neutrophil infiltration and the production of free radicals through their metabolism cause inflammation‐associated tissue damage. ⁴⁶ Therefore, inhibiting neutrophil infiltration in gastric tissue can protect against GU. ²⁷ Our results show that AE treatment significantly increases gastric MPO levels (Fig. 6(B)), indicating neutrophil infiltration. Our results also demonstrate that AE treatment induces the expressions of TNF‐α, IL‐6, and IL‐1β in sera and gastric tissues (Fig. 7). However, EG pretreatments significantly reduced AE‐induced gastric MPO levels and expression levels of TNF‐α, IL‐6, and IL‐1β in both sera and gastric tissues, demonstrating that EG pretreatment reduces AE‐induced inflammatory responses, thus protecting the integrity of the stomach. Therefore, our findings suggest that EG, like BG, exhibits anti‐inflammatory properties in gastric protection. ¹ , ¹²

Cox is a key enzyme in prostaglandin synthesis, and Cox‐1 being highly expressed in the gastrointestinal tract to maintain mucosal integrity. Cox‐2, however, is primarily induced during inflammation; when the gastric mucosa is exposed to harmful substances, expressions of Cox‐2 mRNA and protein are up‐regulated. ⁴⁷ Previous studies have shown that both Cox‐1 and Cox‐2 possess immunomodulatory functions. ⁴⁸ Selective Cox‐2 inhibitors have been used as anti‐inflammatory agents to protect gastrointestinal function. ⁴⁷ Reducing Cox‐2 expression can mitigate the parameters of chronic GUs. ⁴⁹ Our results revealed that AE treatment induces an inflammatory response, accompanied by increased expression of Cox‐2 (Fig. 8(A),(C)). Pretreatment with EG decreases AE‐induced inflammation and reduced Cox‐2 expression, which is consistent with studies of oxyresveratrol and naringenin in ethanol‐induced GU. ²⁹ , ⁵⁰ These findings support the notion that EG possesses anti‐inflammatory properties that mitigate the AE‐induced inflammatory response.

Ethanol‐induced ROS regulate the expressions of pro‐inflammatory factors through the MAPK signaling pathway, thus exacerbating tissue inflammation and damage. ¹ Studies have shown that inhibiting the MAPK pathway can reduce the ethanol‐induced inflammatory response and gastric injury. ²⁸ , ⁵¹ Our study revealed that AE treatment significantly increases phosphorylations of all MAPK families, ERK, JNK, and p38 – while EG pretreatment suppresses these MAPK phosphorylations (Fig. 9). The suppressions are consistent with the role of EG in reducing the expressions of pro‐inflammatory factors (Fig. 7) and mitigating the gastric injury observed in our study (Figs. 2 and 3). These findings also align with our previously published results of the Machilus zuihoensis Hayata bark. ¹

Furthermore, ROS can enhance the transcription of pro‐phosphorylation inflammatory factors by activating the activation of the NF‐κB pathway. Ethanol‐induced TNF‐α alone is sufficient to activate NF‐κB; therefore, ethanol‐induced TNF‐α, in conjunction with ROS, cooperatively facilitates the NF‐κB activation. TNF‐α facilitates NF‐κB activation by promoting IκB phosphorylation, allowing NF‐κB to translocate into the nucleus. ⁵² Our results demonstrate that AE treatment induces IκB phosphorylation, and then significantly increases the expressions of NF‐κB, ratios of p‐IκBα/IκBα and p‐NF‐κB/β‐actin (Fig. 10). However, EG100 pretreatment decreases the AE‐induced phosphorylation of IκBα and p‐IκBα/IκBα ratios and then decreases NF‐κB expressions and p‐NF‐κB /β‐actin ratios. These results indicate that EG100 pretreatment protects against AE‐induced gastric injury at least in part by inhibiting NF‐κB activation. Other studies of C. japonica and Canolol have also shown that blocking NF‐κB activation supports ulcer repair in ethanol‐induced GU. ³⁶ , ⁵³ Thus, we hypothesize that EG pretreatment inhibits the AE‐induced MAPK and NF‐κB pathways, thus suppressing the expressions of inflammatory factors, ultimately protecting against AE‐induced GUs.

Although both EG and CBX demonstrated significant gastroprotective effects in the AE‐induced GU model, their underlying mechanisms and pharmacological targets are similar only in inflammatory pathways. CBX is a synthetic derivative of glycyrrhizin known for its mucoprotective properties. Its efficacy is attributed to enhancement of gastric mucus secretion and the inhibition of inflammatory mediators including Cox‐2 and iNos along with suppression of neutrophil infiltration and MPO activity. ⁴ In this study, EG primarily acts through potent antioxidant activity by neutralizing ROS and restoring endogenous antioxidant systems, including GSH, SOD, and CAT. It also inhibits the phosphorylation of MAPK signaling components (ERK, JNK, and p38) and suppresses NF‐κB activation via down‐regulation of phosphorylated IκB and p65, thereby attenuating the AE‐induced expressions of pro‐inflammatory cytokines (TNF‐α, IL‐1β, and IL‐6) and inflammatory mediators and MPO activity. While both agents modulate overlapping inflammatory pathways, EG further contributes to gastric protection through its antioxidant capacity, highlighting its potential as a multifunctional food‐based therapeutic alternative to conventional drugs like CBX.

In conclusion, this study demonstrates that EG and BG exhibit gastroprotective effects in AE‐induced GU model mice. EG shows a greater ability to reduce UI compared to BG at the same dose. We obtain a non‐linear nature for the EG dose–response relationship with a potential optimal dose of 100 mg/kg in mice. The gastroprotective mechanisms of EG involve primarily improving mucus secretion, improving antioxidant and anti‐inflammatory capacities. Furthermore, the anti‐inflammatory mechanism of EG is through the NF‐κB and MAPK signaling pathways. This study highlights the promising potential of EG as a value‐added functional food ingredient for GU prevention, emphasizing its unique gastroprotective mechanisms. Studies are underway to explore the protective effects of BG and EG on animal models of liver and kidney diseases and non‐alcoholic fatty liver disease.

AUTHOR CONTRIBUTIONS

Ming‐Shiun Tsai: conceptualization, methodology, formal analysis, investigation, writing – original draft. Gunn‐Guang Liou: methodology, resources, data curation. Chang‐Wei Hsieh: supervision, resources, writing – review and editing. Jiunn‐Wang Liao: histopathological analysis, validation, visualization. Pin‐Yen Lai: software, investigation, data curation. Chao‐Kai Chang: resources, supervision, writing – review and editing. Wan‐Jou Lee: software, investigation, data curation. Sue‐Hong Wang: project administration, funding acquisition, supervision, writing – review and editing.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.