A total of 24 female hamsters (age, 3 weeks; body weight, 60-70 g) were acquired from Shanghai Yanji Bio-tech Co., Ltd. The animals were housed in a facility maintained at 24˚C under a 12-h light/dark cycle, and were given free access to water and chow. The present study was carried out in accordance with the recommendations of the animal protection guidelines approved by the local animal ethics committee. The study protocol was approved by the Ethics Committee for Medical Experimental Animals, Second Affiliated Hospital of Guangxi Medical University [approval no. 2022(KY-0174); Nanning, China].

The hamsters were randomly divided into the following four groups: Control group (n=6), HTG group (n=6), HTGAP group (n=6), and HTGAP + SC75741 group (n=6). No direct comparison between male and female hamsters was performed; therefore, sex-specific differences could not be assessed and should be addressed in future studies. Hamsters in the HTG, HTGAP and HTGAP + SC75741 groups were fed a high-fat diet (82.8% normal chow + 15% saturated animal fat + 2% cholesterol + 0.2% sodium cholate), whereas hamsters in the control group were fed standard hamster chow.

After the animals were fed standard hamster chow or a high-fat diet for 5 weeks, 0.1-0.2 ml blood was collected from the retro-orbital vein and centrifuged at 1,000 x g for 15 min at 4˚C to isolate serum. Subsequently, serum TG levels were measured to confirm successful establishment of the HTG model using a TG Fluorometric Assay Kit (cat. no. E-BC-F033; Elabscience Bionovation Inc.) according to the manufacturer's instructions. HTG induction was confirmed when serum TG levels in the experimental group were >2-fold higher than those in the control group. The HTGAP model was established by retrograde injection of 6% sodium taurocholate (TCI Tokyo Chemical Industry Co., Ltd.) into the bile and pancreatic duct under isoflurane anesthesia; anesthesia was induced with 5% isoflurane and maintained at 3-4% during the surgical procedure, with reference to established rodent inhalation anesthesia protocols (23). Sodium taurocholate was infused at a volume of 0.1 ml/100 g body weight and a rate of 0.1 ml/min using an infusion pump. Successful cannulation and injection were confirmed by visual observation of pancreatic duct dilatation. The common bile duct was not clamped during the procedure. Sham-operated controls were not included in the present study. The hamsters in the HTGAP + SC75741 group were treated with SC75741 (MedChemExpress), an inhibitor of the NF-κB signaling pathway, through intraperitoneal injection (1 mg/100 g body weight) 30 min before modeling (24). All of the animals were fasted for 6 h before surgical operations. After 24 h of AP induction, euthanasia was performed via intraperitoneal injection of an overdose of pentobarbital sodium (200 mg/kg), in accordance with the approved animal protocol and supporting rodent euthanasia literature (25,26). No hamsters died during anesthesia; however, one hamster in the HTGAP group died unexpectedly during postoperative recovery from anesthesia after model induction. Humane endpoints included severe respiratory distress, inability to eat or drink, persistent recumbency or a moribund appearance; no animals were euthanized early before the planned endpoint. Subsequently, 2-3 ml blood was collected from the abdominal aorta and centrifuged at 1,000 x g for 15 min at 4˚C, and the serum was conserved at -80˚C. Pancreatic tissues used in the present study were fixed in 10% formalin at room temperature for 24 h, embedded in paraffin, and used for histopathological and immunohistochemical analyses. Finally, the animal carcasses were uniformly sent to the Guangxi Zhuang Autonomous Region Hazardous Waste Disposal Center for processing.

The primary study endpoints were pancreatic injury, and changes to the expression levels of tricellulin, claudin-1 and NF-κB pathway-related proteins. Secondary endpoints included F-actin cytoskeletal organization and intercellular morphological changes, and alterations in cell-cell contacts in AR42J cells.

The pancreatic specimens were cut into 3-µm sections and subjected to hematoxylin and eosin staining. Sections were stained with hematoxylin for 2 min, differentiated with 1% hydrochloric acid alcohol for 30 sec and counterstained with eosin for 30 sec, all at room temperature. Pancreatic injury was assessed according to a previously reported scoring system, including edema, inflammatory cell infiltration, hemorrhage and necrosis (27). Morphology was assessed under a light microscope and histopathological evaluation was performed independently by two investigators blinded to group allocation.

Formalin-fixed, paraffin-embedded pancreatic tissue sections were deparaffinized, and antigens were retrieved through high-pressure cooking with citrate buffer (pH 6.0) for 5 min. Subsequently, these tissue sections were subjected to gradient cooling until they reached room temperature and were then washed thoroughly in phosphate-buffered saline (PBS; pH 7.4). The sections were quenched with 3% hydrogen peroxide for 30 min at room temperature, blocked with 3% bovine serum albumin (Beijing Solarbio Science & Technology Co.) at room temperature for 15 min, and incubated overnight at 4˚C with primary antibodies against tricellulin (1:500; cat. no. 13515-1-AP), claudin-1 (1:500; cat. no. 13050-1-AP) and NF-κB p65 (1:300; cat. no. 10745-1-AP) (all from Wuhan Sanying Biotechnology). The next day, the tissue sections were washed and incubated with HRP-based secondary antibody reagents using an SABC-HRP Kit with Anti-Rabbit IgG (IHC&ICC) (cat. no. P0615; Beyotime Biotechnology) for 15 min at room temperature. The slides were then placed in PBS, washed three times on a decolorizing shaker (5 min/wash) and visualized with DAB (50 µl/slide; Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.). Images were acquired using a pathology image analysis system (Leica DMR+Q550; Leica Microsystems GmbH) under identical settings, and quantitative analysis was performed using Image-Pro Plus 6.0 (Media Cybernetics, Inc.). Three randomly selected non-overlapping fields per section were analyzed and quantification was performed in a blinded manner.

The rat pancreatic exocrine cell line AR42J (RRID: CVCL_0143) was obtained from the American Type Culture Collection and cultured to an appropriate density in Dulbecco's Modified Eagle Medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.) in a humidified incubator containing 5% CO₂ at 37˚C. Three tricellulin short hairpin (sh)RNAs were designed and evaluated for lentiviral knockdown experiments: sh1, 5'-GGATAAGCCAGTGTCAGAT-3'; sh2, 5'-GCTTGGATCGCGACGATTA-3'; and sh3, 5'-GCCACATTCCTAAGCCCAT-3'. A non-targeting negative control shRNA sequence (5'-CCTAAGGTTAAGTCGCCCTCG-3') was used as the control. The lentiviral shRNA vector and the corresponding control vector were commercially constructed and packaged by OBiO Technology (Shanghai) Corp., Ltd., and were based on pSLenti-U6-shRNA (Marveld2)-CMV-EGFP-F2A-Puro-WPRE. In addition, a lentiviral overexpression vector carrying tricellulin and the corresponding control vector were also commercially constructed and packaged by OBiO Technology (Shanghai) Corp., Ltd.; these were based on pcSLenti-EF1-EGFP-P2A-Puro-CMV-Marveld2-3xFLAG-WPRE. The efficiency of tricellulin knockdown and overexpression was validated in AR42J cells infected with these lentiviral constructs alone by reverse transcription-quantitative PCR (RT-qPCR) (Fig. S1). For lentiviral infection, AR42J cells were seeded in 6-well plates at 5x10⁵ cells/well and infected at a multiplicity of infection of 100 in the presence of polybrene (10 µg/ml) at 37˚C. After 8 h, the medium was replaced with complete medium and stable cell lines were selected using puromycin (5 µg/ml) until harvest. Cells were harvested 72 h after infection for RT-qPCR validation. Briefly, total RNA was extracted from lentivirus-infected AR42J cells using RNAiso Plus (cat. no. 9109; Takara Bio, Inc.). RT was performed using an RT kit (cat. no. K1622; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. qPCR was performed using SYBR Green Mix (cat. no. 11201ES60; Shanghai Yeasen Biotechnology Co., Ltd.) on a BIOER 9600 Plus real-time PCR system (Bioer Technology Co., Ltd.). Thermocycling was performed as follows: Initial denaturation at 95˚C for 5 min, followed by 40 cycles at 95˚C for 10 sec and 60˚C for 30 sec. Relative mRNA expression was quantified using the 2^-ΔΔCq method (28). The primer sequences used for RT-qPCR, including the internal control GAPDH, are listed in Table I.

To determine the effects of tricellulin on AR42J cells in the HTG environment, lentivirus-infected cells were treated with 66-µM palmitic acid (MilliporeSigma) + 132-µM oleic acid (MilliporeSigma) for 24 h at 37˚C for subsequent experiments. To explore the effects of the NF-κB pathway on AR42J cells, lentivirus-infected cells were treated with 1 µM SC75741 for 24 h at 37˚C for subsequent experiments. Cells cultured without any extra reagents were regarded as the control group. Unless otherwise stated, in vitro experiments were performed in three independent biological replicates.

Cultured cells were lysed in RIPA lysis buffer (Beijing Solarbio Science & Technology Co., Ltd.) and protein concentrations were determined by BCA assay. Total protein samples (20 µg) were then separated by SDS-PAGE on 10% and transferred to a polyvinylidene difluoride membrane. After blocking with 5% non-fat milk for 1 h at room temperature, the membranes were incubated with primary antibodies against β-actin (1:20,000; cat no. 66009-1-Ig), tricellulin (1:1,000; cat. no. 13515-1-AP), claudin-1 (1:1,000; cat. no. 13050-1-AP), NF-κB p65 (1:1,000; cat. no. 10745-1-AP) and NF-κB phosphorylated-p65 (p-p65; 1:2,000; cat. no. 82335-1-RR) (all from Wuhan Sanying Biotechnology) overnight at 4˚C. Subsequently, the membrane was washed three times with Tris-buffered saline-0.1% Tween-20 buffer and incubated for 60 min at room temperature with HRP-conjugated secondary antibodies, including Peroxidase AffiniPure^® Goat Anti-Rabbit IgG (H+L) (1:50,000; cat. no. 111-035-003; Jackson ImmunoResearch) and Peroxidase AffiniPure Goat Anti-Mouse IgG (H+L) (1:50,000; cat. no. 115-035-003; Jackson ImmunoResearch), as appropriate. Protein bands were visualized using Tanon™ Femto-sig ECL chemiluminescent substrate (cat. no. 180-506; Tanon Science and Technology Co., Ltd.) and semi-quantified using Quantity One software (version 4.6.0; Bio-Rad Laboratories, Inc.).

Cells were plated onto coverslips (24x24 mm) in 6-well plates until they reached 55-60% confluency. The cells were then fixed in 4% paraformaldehyde for 15 min at room temperature, permeabilized in 0.1% Triton X-100 for 5 min at room temperature, blocked in PBS containing 2% bovine serum albumin for 20 min at room temperature and stained with tetramethyl rhodamine isothiocyanate-phalloidin (100 nM; 50 µl/well; Beijing Solarbio Science & Technology Co., Ltd.) for 30 min at room temperature in the dark. The samples were washed with PBS three times at 5-min intervals between each step and mounted with an anti-fading medium containing DAPI (Beijing Solarbio Science & Technology Co., Ltd.) for 3 min at room temperature in the dark. F-actin was visualized at 530-550 nm and cell nuclei were observed at 460-495 nm under an upright fluorescence microscope (BX53; Olympus Corporation).

The intercellular TJ ultrastructure of AR42J cells was evaluated by scanning electron microscopy. Cells were washed with PBS and fixed in 3% glutaraldehyde at 4˚C for >2 h, followed by post-fixation in 1% osmium tetroxide. After the cells were dehydrated through a graded ethanol series, samples were dried, mounted on metal stubs, sputter-coated with gold and examined using a scanning electron microscope (H-7650; Hitachi, Ltd.).

Data are presented as the mean ± standard deviation. Normality was assessed using the Shapiro-Wilk test and homogeneity of variance was assessed using Levene's test. For comparisons among multiple groups, one-way analysis of variance was used when parametric assumptions were met, followed by Tukey's post hoc test. Welch's ANOVA followed by Dunnett's T3 test was applied when homogeneity of variance was not met. Because histological scores are ordinal data, they are presented as the median (IQR) and were analyzed using the Kruskal-Wallis test followed by Bonferroni-corrected pairwise Mann-Whitney U tests. Comparisons between two groups were performed using an unpaired Student's t-test. Unless otherwise stated, in vitro experiments were performed in three independent biological replicates. P<0.05 was considered to indicate a statistically significant difference. Statistical analyses were performed using SPSS 20.0 (IBM Corp.).

Serum collected from the hamsters in the high-fat diet groups (HTG, HTGAP and HTGAP + SC75741 groups) was lipemic on visual inspection, whereas that of the control group was clear. Serum TG levels differed significantly among the groups (P=0.001). Serum TG levels in the HTG, HTGAP and HTGAP + SC75741 groups were all significantly higher than those in the control group (P=0.001, P=0.027 and P=0.039, respectively), confirming successful induction of HTG (Table II). No significant difference was observed among the HTG, HTGAP and HTGAP + SC75741 groups (all P>0.05).

On macroscopic examination, pancreatic edema, hemorrhage and necrosis were observed mainly in the HTGAP and HTGAP + SC75741 groups (Fig. 1A). In the HTG group, the pancreas showed no obvious necrosis or hemorrhage on gross examination, and only minimal or no histopathological abnormalities were observed. Notably, the necrotic lesions were mainly concentrated in the head of the pancreas. Occasional biliary fistula and scattered calcification foci were observed in the HTGAP group, but these findings were not predefined endpoints and were not quantitatively analyzed. Histologically, the control group exhibited no obvious edema, inflammatory infiltration or necrosis, whereas the HTGAP and HTGAP + SC75741 groups showed evident edema, inflammatory infiltration, hemorrhage and necrosis (Fig. 1B). Histopathological scores differed significantly among the groups (Kruskal-Wallis test, P<0.001). After Bonferroni-corrected pairwise comparisons, the HTGAP group showed significantly higher pathological scores than the control and HTG groups (both P<0.05) (Table II). The HTGAP + SC75741 group also showed significantly higher pathological scores than the control and HTG groups (both P<0.05), but significantly lower scores than the HTGAP group (P=0.002). No significant difference was observed between the control and HTG groups (P=0.846).

Immunohistochemical analysis showed significant group differences in tricellulin, p65 and claudin-1 expression (Fig. 2; Table II). For tricellulin, the overall difference among groups was significant (P<0.05), and Tukey's post hoc test showed that tricellulin expression was significantly lower in the HTG, HTGAP and HTGAP + SC75741 groups than in the control group (P=0.001, P<0.05 and P<0.05, respectively). In addition, tricellulin expression in the HTGAP group was significantly lower than that in the HTG group (P=0.001), whereas no significant difference was observed between the HTGAP and HTGAP + SC75741 groups (P=0.519).

For p65, the overall difference among groups was also significant (P<0.05). Tukey's post hoc analysis showed that p65 expression was significantly higher in the HTGAP group than in the control, HTG and HTGAP + SC75741 groups (all P<0.05). The HTGAP + SC75741 group also showed significantly higher p65 expression than the control and HTG groups (both P<0.05), whereas no significant difference was found between the control and HTG groups (P=0.290).

For claudin-1, the overall difference among groups was significant (P<0.05). Because the homogeneity of variance assumption was not satisfied, Dunnett's T3 test was used for pairwise comparisons. Claudin-1 expression was significantly lower in the HTGAP group than in the control group (P=0.008) and the HTG group (P<0.05). The HTGAP + SC75741 group also exhibited significantly lower claudin-1 expression than the control group (P=0.016) and the HTG group (P<0.05). No significant difference was observed between the HTGAP and HTGAP + SC75741 groups (P=0.877), nor between the control and HTG groups (P=0.894). These findings indicated that claudin-1 reduction was mainly observed after AP induction rather than under HTG alone.

RT-qPCR validation in AR42J cells transduced with lentiviral constructs alone showed marked upregulation of Marveld2 mRNA in the overexpression group, whereas among the three shRNAs tested, sh1 showed the strongest knockdown efficiency compared with the corresponding negative controls (Fig. S1). The expression levels of tricellulin, claudin-1, and NF-κB p-p65 and total p65 were analyzed by western blotting. The expression levels of tricellulin and claudin-1 were decreased, whereas those of NF-κB p-p65 slightly increased under HTG conditions relative to the baseline values. Total p65 expression showed only modest variation across groups, with lower levels in the tricellulin overexpression and tricellulin overexpression + SC75741 groups, whereas the tricellulin knockdown group remained close to the HTG group (Fig. 3). Overexpression of tricellulin upregulated claudin-1 expression and downregulated NF-κB p-p65 expression in AR42J cells in the HTG environment. When treated with SC75741, an inhibitor of the NF-κB pathway, the expression levels of NF-κB p-p65 further decreased, whereas those of tricellulin and claudin-1 increased. Conversely, knockdown of tricellulin downregulated claudin-1 expression and upregulated NF-κB p-p65 expression in AR42J cells in the HTG environment. SC75741 treatment inhibited NF-κB activation, and was associated with partial recovery of tricellulin and claudin-1 expression compared with the corresponding groups without SC75741 treatment.

Immunofluorescence was performed to analyze how the regulation of tricellulin and treatment with SC75741 impacted F-actin dynamics. Under HTG conditions, the F-actin network showed reduced continuity, weaker cortical distribution and a more disorganized filament pattern compared with the control group (Fig. 4). Tricellulin overexpression was associated with partial restoration of F-actin organization, whereas tricellulin knockdown was associated with more pronounced disruption. These changes remained evident after SC75741 treatment.

Representative scanning electron micrographs showed that control cells were regularly arranged with relatively intact intercellular contacts and smooth surfaces (Fig. 5). Under HTG conditions, cells showed swelling, membrane injury and weakened cell-cell contacts. These changes appeared less pronounced in the tricellulin overexpression group and more severe in the tricellulin knockdown group. Following SC75741 treatment, cellular damage remained evident, particularly in the tricellulin knockdown group.