B3GNT7 preserves intestinal barrier integrity: a key defense mechanism against acute pancreatitis progression

Abstract

Mucin-2 (MUC2) plays a crucial role in maintaining intestinal homeostasis during acute pancreatitis (AP). We aimed to investigate the role of B3GNT7 (β-1,3-N-acetylglucosamine transferase) in the O-glycosylation of MUC2 in AP. We used two mouse models of AP induced by caerulein/lipopolysaccharide or L-arginine, and observed colonic goblet cells using electron microscopy, noting that Golgi damage was linked to decreased levels of mature MUC2. Golgi stress–associated proteins (reduced GM130 and elevated GOLPH3) were identified using immunofluorescence. Transcriptome analysis revealed the downregulation of B3gnt7, a glycosyltransferase that is highly enriched in the Golgi apparatus of goblet cells in AP mice. B3GNT7 expression was negatively correlated with pancreatic and colonic pathological scores. In patients with AP, intestinal B3GNT7 levels are markedly reduced and correlated with Ranson scores, C-reactive protein levels, and intestinal permeability markers (serum diamine oxidase and D-lactate). Adeno-associated virus (AAV)-mediated knockdown of B3GNT7 reduced O-glycosylated MUC2 levels, exacerbated pancreatic and systemic inflammation, and worsened intestinal permeability and dysbiosis. In vitro, LPS-treated HT-29 cells exhibited Golgi stress (decreased levels of B3GNT7 and O-glycosylated MUC2) that was reversed by L-glutathione (GSH). These findings demonstrate that B3GNT7 downregulation, mediated by Golgi stress, disrupts MUC2 O-glycosylation, exacerbating AP by impairing intestinal homeostasis.

Introduction

Acute pancreatitis (AP) is a common gastrointestinal disorder characterized by varying degrees of severity and clinical course [1]. Approximately one-fifth of patients with AP develop severe acute pancreatitis (SAP), which has a mortality rate of approximately 20%–30% [2, 3]. Gut dysfunction, a common and serious complication of SAP, contributes to the development of systemic inflammatory response syndrome (SIRS) and multiple organ dysfunction syndrome (MODS) [4, 5].

The severity of intestinal injury has been shown to correlate with adverse outcomes and to serve as a potential predictor of mortality in AP [6]. Therefore, preserving the integrity of the intestinal barrier is essential in clinical practice. The intestinal mucus layer, which is primarily composed of mucin-2 (MUC2), serves as the first line of defense [7]. Our previous study showed that colonic MUC2 levels were significantly decreased in patients with AP. Depletion of MUC2 in a rat model of AP resulted in serious pancreatic and colonic injuries, goblet cell dysfunction, and intestinal dysbiosis, which were partially reversed by supplementation of MUC2 [8]. However, the mechanism underlying the variation of MUC2 levels in AP remains unclear.

MUC2 is a secreted mucin primarily composed of O-glycans, and this heavy O-glycosylation enables MUC2 to resist proteolysis and protect the gastrointestinal tract from pathogens and microbial infections [9, 10]. Specific glycosyltransferases located in the Golgi apparatus play crucial roles in the O-glycosylation of mucin [11]. Under conditions of oxidative stress, DNA damage, and inflammation, insufficient Golgi function augments the expression of glycosylation enzymes via the Golgi stress response [12, 13]. Golgi stress is assessed by evaluating specific markers, including GM130 (a structural protein) and GOLPH3 (a stress-responsive protein), along with ultrastructural changes observed via electron microscopy [13]. In AP, oxidative stress, which affects the structure and physiological function of secondary reaction Golgi stress, plays a mediating role [1, 14]. However, the involvement of Golgi stress in AP remains unclear.

B3GNT7, a gene encoding the specific glycosyltransferase β-1,3-N-acetylglucosamine transferase, is highly expressed in the Golgi apparatus [15]. It influences the integrity of the intestinal barrier by regulating the glycosylation of mucins [16]. The function of colonic B3GNT7 has been clarified in both patients with ulcerative colitis (UC) and a dextran sulfate sodium salt (DSS)-induced mouse model of colitis, showing that its downregulation may contribute to impaired mucin barrier function and exacerbation of colitis [17, 18]. Clinically, intestinal barrier dysfunction is a critical determinant of disease progression and outcomes in patients with AP. Given that B3GNT7 is essential for the O-glycosylation and maturation of MUC2, and that its expression is highly enriched in colonic goblet cells, we hypothesized that B3GNT7 downregulation under Golgi stress conditions may represent a key mechanism linking intestinal barrier impairment to AP severity.

This study aimed to explore the effects of colonic B3GNT7 in regulating the O-glycosylation of MUC2 involved in Golgi stress during AP.

Materials and methods

Human samples

According to the Revised Atlanta Classification in 2012 [19], patients with AP in the Department of Gastroenterology of Shanghai General Hospital who had indications for the placement of a nasojejunal tube to receive enteral nutrition or endoscopic retrograde cholangiopancreatography (ERCP) were enrolled in our study. Healthy controls (HCs) were volunteers who underwent endoscopy. Patients with gastrointestinal infections, inflammatory bowel disease, gastrointestinal tumors, obesity, type 2 diabetes, or fatty liver disease, and pregnant women were excluded. Written informed consent was obtained from all patients. Biopsy specimens from the descending duodenum and blood samples were collected from 22 patients with AP and 13 HCs. We also recorded clinical indicators, such as CRP and white blood cell (WBC) counts. The Ranson score was comprehensively assessed based on multiple clinical and laboratory indicators in patients with AP at admission and 48 h later. Patients with AP were further divided into the moderately severe AP (MSAP) (n = 14, transient organ failure, local complications, or exacerbation of comorbid disease) and severe AP (SAP) (n = 8, persistent organ failure, i.e., organ failure > 48 h) groups [19].

Duodenal biopsy specimens were fixed in 4% formaldehyde or stored at − 80 °C. Blood samples were centrifuged to obtain serum and stored at − 80 °C. This study involving human samples was approved by the Ethics Committee of Shanghai General Hospital (2021035) and registered with the China Clinical Trial Registry (ChiCTR1800017214).

Animals

Male C57BL/6 mice (6- to 8-week old; 20–22 g) were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). They were housed under specific pathogen–free (SPF) conditions with a room temperature of 23 ± 2 °C and a 12-h light/dark cycle.

Experimental design

In this study, we used two mouse models of AP. In caerulein-induced AP (Cae-AP), mice were intraperitoneally injected with caerulein (100 µg/kg, 10 times with 1 h interval between injections), and lipopolysaccharide (LPS) (5 mg/kg) (MedChemExpress, China) was administered by intraperitoneal injection immediately after the final caerulein injection. Cae-AP mice were sacrificed 12 h after the initial injection of caerulein (Fig. S1A). L-arginine-induced AP (L-AP) mice received intraperitoneal injections of 10% L-arginine (4.5 g/kg, pH 7.0) (Sigma-Aldrich, USA) twice at a 1-h interval. L-AP mice were sacrificed 72 h after AP induction (Fig. S1B).

For each AP induction model (caerulein/LPS or L-arginine), mice were randomly assigned to their respective experimental groups (n = 6 per group): a model-specific control group (CON), an AP group, an AAV-B3GNT7 knockdown group, and an AAV-B3GNT7 knockdown + AP group. This design ensured that each AP model was compared against its own appropriate baseline control.

For AAV administration, mice were fasted overnight and pretreated with the intestinal mucosal surface mucolytic agent N-acetyl-L-cysteine (NAC) (20 mM) [20]. NAC (200 µL) was administered into the rectum of the mouse via a microsyringe to clean the colon, which was then emptied for 30 min without sedation [21]. Subsequently, the mice were anesthetized using inhaled isoflurane, and AAV injection (4 × 10¹⁰-1 × 10¹¹ vg) was administered via enema. At 3 weeks after AAV infection, AP was induced in these mice as per the models described earlier.

Samples of the blood, colon, pancreas, liver, and mesenteric lymph nodes were collected and stored at − 80 °C or in 4% paraformaldehyde. Additionally, fresh contents from the ileocecal region were collected for analysis of the intestinal microbiota.

Histopathology

Pancreatic and colon tissues were fixed in 4% paraformaldehyde, and colonic tissues were fixed in Carnoy’s fixative at 4 °C for 48 h and processed by sequential cleaning with double-distilled water (ddH₂O), rehydrating, and dehydrating steps. They were then embedded in paraffin and cut into 4-µm fragments. After periodic acid-Schiff (PAS) or hematoxylin and eosin (H&E) staining, the slides were examined under a light microscope (Leica, Germany) at 100 × and 200 × magnification. Pancreatic histopathological changes were evaluated according to the Schmidt criteria [22]. Colonic pathological damage was scored according to the Araki and Mehmet methods [23, 24].

Immunofluorescence

Paraffin sections were dewaxed, antigen-repaired, and permeabilized with 0.25–0.5% Triton X-100 at room temperature for 20 min. After completion of permeabilization, the slides were blocked with blocking buffer for 1 h at room temperature. Next, they were incubated with the primary antibodies—B3GNT7 (Proteintech, China), Claudin-1 (Abcam, UK), occludin (Abclonal, China), ZO1 (CST, USA), GM130 (Abclonal, China), and GOLPH3 (Abclonal, China)—at 4 °C overnight. Ulex europaeus agglutinin-I (UEA-1) specifically binds to α-1,2-fucosylated glycans and is exclusively used to detect fucosylation levels. For UEA-1 lectin staining, the sections were incubated with it at 37 °C for 1 h. The slides were further incubated with diluted fluorescent secondary antibody for 1 h in the dark at room temperature and stained with dihydrochloride (DAPI) for 5–7 min to visualize the nuclei. Finally, the slides were observed under a fluorescence microscope.

Enzyme-linked immunosorbent assay (ELISA)

Levels of tumor necrosis factor-α (TNFα), interleukin-1β (IL-1β), and IL-6 were measured by Luminex mouse discovery assay kit (R&D Systems, USA) according to the manufacturers’ protocols. Levels of serum diamine oxidase (DAO) and D-lactate were measured using DAO mouse ELISA kit and D-lactate mouse ELISA kit (MultiSciences Biotech, China) according to the provided protocols.

Serum amylase assay

Serum amylase activity was determined with amylase reagents (Proteintech, China) using an Advia 2400 chemistry system (Siemens, Germany) according to the manufacturer’s instructions.

TUNEL and FISH assays

Colonic epithelial cell apoptosis was detected using a One-Step Terminal Deoxynucleotidyl transferase–mediated dUTP-biotin Nick End Labeling (TUNEL) Apoptosis Assay Kit (Red Fluorescence) (Beyotime, China) according to the manufacturer’s instructions. TUNEL-positive cells were counted at 200 × magnification. FISH was used to detect bacterial translocation. Sections of the colon and pancreas were dewaxed, antigen-repaired, and digested with Proteinase K at 37 °C for 20 min. Next, sections were pre-hybridized at 37 °C for 1 h and then incubated with specific probes (EUB338 [59-Cy3-GCTGCCTCCCGTAGGAGT-39]) in a wet box at 37 °C overnight. Sections were washed and stained with DAPI. Images were captured using a fluorescence microscope (Leica, Germany). EUB338, a universal bacterial 16 S rRNA FISH probe, was employed to detect bacterial localization in tissue sections.

FITC-dextran assay

Intestinal permeability was assessed using fluorescein isothiocyanate–labeled dextran (FITC-dextran, average molecular weight 4 kDa, Sigma-Aldrich). Mice were fasted for 4 h with free access to water. FITC-dextran was administered by oral gavage at a dose of 0.5 g/kg body weight. For the caerulein/LPS-AP model, gavage was performed 4 h prior to the first caerulein injection (i.e., at time point − 4 h). Blood samples were collected 4 h after gavage (coinciding with the 12 h AP endpoint), and serum was obtained for fluorescence measurement (excitation 485 nm, emission 535 nm). The concentration of serum FITC-dextran was calculated from a standard curve. This 4-hour circulation time is within the standard and validated window for assessing gut permeability with 4 kDa FITC-dextran [25].

Bacterial cultures and plate counting

Mesenteric lymph nodes and liver tissue were collected in sterile PBS and fully ground using a high-throughput tissue grinder (Onebio Biotech, China). The suspension was plated onto BHIA plates to culture anaerobic bacteria. The plates were incubated for 48 h at 37 °C using Oxoid AnaeroGen 2.5 L and an Oxoid resazurin anaerobic indicator (Thermo Scientific, USA). Plates that had 25–250 colony-forming units (CFU) were counted.

RNA isolation and real-time PCR

Pancreatic and colonic inflammatory cytokines (IL-1β, TNF-α, and IL-6) and the Golgi stress–associated proteins (GM130 and GOLPH3) and B3GNT7 were measured using real-time PCR. Total RNA was isolated from pancreatic and colonic tissues using an EZ-press RNA Purification Kit (EZBioscience, USA). Reverse transcription (RT) was performed using the HyperScript III RT SuperMix for qPCR with a gDNA Remover Kit (EnzyArtisan, China). The 2xS6 Universal SYBR qPCR Mix (EnzyArtisan, China) was used to perform real-time PCR, according to the manufacturer’s instructions. The relative values of gene expression were analyzed using the 2^−ΔΔCt method. The primers used for the detection are listed in Table S2.

Cell culture and processing

HT-29 cells were maintained in DMEM high glucose medium (Gibco, USA), supplemented with 10% heat-inactivated FBS (Gibco, USA) and 1% penicillin/streptomycin at 37 °C in a 5% CO₂ atmosphere [26]. To investigate the role of Golgi stress in LPS-stimulated HT-29 cells, the cells were cultured in 12- and 6-well plates and randomly divided into six groups: control, LPS, GSH, BFA, GSH + LPS, and BFA + LPS. GSH (Sigma-Aldrich, USA) and BFA (MedChemExpress, China) were separately mixed with LPS in DMEM (Gibco, USA). Cells of each group were treated with the respective reagents for 12 h and subsequently harvested for further analysis.

Transmission electron microscopy

Approximately 1 cm of proximal colon tissue was collected, fixed in 2.5% glutaraldehyde for 24 h, dehydrated using graded acetone, embedded in Epon812 and sectioned into 50-nm ultrathin slices using an ultramicrotome [27]. The sections were then stained with uranium dioxide acetate and lead citrate for 10 min at room temperature. The condition of organelles, including the Golgi apparatus and mitochondria within the goblet cells, was examined using a Philips Tecnai-10 transmission electron microscope.

RNA sequencing

Total RNA was extracted from the colon of mice using TRIzol^® Reagent (Plant RNA Purification Reagent for plant tissue) according the manufacturer’s instructions (Invitrogen) and genomic DNA was removed using DNase I (TaKara). The RNA-seq transcriptome library was prepared using a TruSeq RNA sample preparation Kit (Illumina, San Diego, USA). The RNA-seq library was sequenced using an Illumina HiSeq xten/NovaSeq 6000 sequencer (2 × 150 bp read length). Raw paired-end reads were trimmed and quality controlled using SeqPrep (https://github.com/jstjohn/SeqPrep) and Sickle (https://github.com/najoshi/sickle) with default parameters. Clean reads were then separately aligned to the reference genome in orientation mode using TopHat (http://tophat.cbcb.umd.edu/, version 2.0.0) software. To identify DEGs, the expression level of each transcript was calculated according to the fragments per kilobase of exon per million mapped reads (FRKM) method.

16 S ribosomal RNA (rRNA) sequencing

Genomic DNA was extracted from the ileocecal contents using an EZNA Fecal DNA Kit (Omega, USA) following the manufacturer’s instructions. The forward primer 5’-TACGGRAGGCAGCAG-3’ and the reverse primer 5’-AGGGTATCTAATCCT-3’ were utilized for amplification. High-throughput 16 S rRNA sequencing was conducted on the Illumina MiSeq platform (Illumina, USA) in accordance with the standard protocol established by Majorbio Bio-Pharm Technology.

Statistics

Continuous data with a normal distribution are expressed as mean (SD), and variables with a non-normal distribution are expressed as median (IQR). Comparisons between two groups with a normal distribution were performed using t-test. The Mann–Whitney U-test was used for comparisons between two groups with non-normal distribution. Single-factor analysis of variance (ANOVA) was used to compare three or more groups. For categorical variables, data are expressed as numbers and percentages and were evaluated using a Chi-square test. Spearman’s rank correlation coefficient was used to determine the correlations between B3GNT7 and the indicators. All statistical analyses were performed using IBM SPSS Statistics 25. Statistical significance was set at p < 0.05.

Results

Changes of B3GNT7 and Golgi stress in goblet cells of AP mice

We used two mouse models of AP (caerulein + LPS and L-arginine) to elucidate the role of MUC2 in AP. As shown in Fig. 1A, compared with the control (Con) mice, the number of intestinal goblet cells, which are responsible for MUC2 synthesis and secretion [28], was significantly reduced in both models (p < 0.05). We examined the goblet cells using transmission electron microscopy (TEM). In AP mice, mucus particles in goblet cells were reduced and presented with irregular shapes (Fig. 1B). The Golgi apparatus was characterized by marked hypertrophy and capsule expansion, high-intensity fragmented punctate structures, and dispersion into the cytoplasm (Fig. 1B), together with reduced GM130 expression and increased GOLPH3 expression, both of which indicated Golgi stress in goblet cells during AP (p < 0.05) (Fig. 1C–E).

Fig. 1.

Number and morphology of goblet cells, and expression of Golgi stress–related proteins in the colon of acute pancreatitis (AP) mice. (A) Periodic acid-Schiff (PAS)-stained mucus-containing goblet cells (100× magnification). (B) Electron micrograph of goblet cells. (C) Immunofluorescence staining of GM130 (200× magnification). (D) Immunofluorescence staining of GOLPH3 (200×magnification). (D) Quantification of colonic GM130 and GOLPH3

To investigate the downstream molecular mechanisms of Golgi stress in colonic goblet cells during AP, we established an AP mouse model induced by caerulein combined with lipopolysaccharide (LPS). Colonic transcriptomic analysis was conducted and a volcano plot was generated, which revealed 427 differentially expressed genes (DEGs) between the CON and AP groups (Fig. 2A). Cross-referencing the above data with those in a public database of goblet cell–related gene sequences revealed 16 goblet cell–related DEGs, depicted as a Venn diagram (Fig. 2B). As shown in Fig. 2C, among the top three genes in the heatmap, B3GNT7 is localized in the Golgi apparatus and closely linked to intestinal homeostasis via glycosylation of mucin [29]. Using real-time PCR and immunofluorescence (IF), we observed that the expression of colonic B3GNT7 in the caerulein-induced AP (Cae-AP) group was significantly lower than that in the CON group (P < 0.05) (Fig. 2D, E). Changes of colonic B3GNT7 in the L-arginine-induced AP (L-AP) group were similar to those in the Cae-AP group (P < 0.05) (Fig. S1C, D). UEA-1, a highly sugar-specific lectin, exhibits changes in α-1,2-fucosylation within the colonic epithelium [30]. Therefore, UEA-1 expression is an indicator of mature MUC2 (O-glycosylated MUC2). Colocalization of B3GNT7 and mature MUC2, which presented by UEA-1 staining in goblet cells, was further verified by IF in AP (P < 0.05) (Fig. 2E; Fig. S1D). This indicates that B3GNT7–mediated O-glycosylation of MUC2 is impaired during AP.

Fig. 2.

RNA sequencing of colon and validation of B3gnt7 expression in AP mice. (A) Volcano plot of differentially expressed genes (DEGs) between the colons of AP and CON mice identified using transcriptome sequencing. (B) The Venn diagram illustrates the intersection analysis of goblet cell–related DEGs in the AP mice. (C) Heatmap of DEGs in the colonic transcriptome between CON and AP mice. (D) The mRNA expression levels of B3gnt7 in the colon of AP mice. (E) Colocalization of B3GNT7 and mature MUC2 in the colon of AP mice (200× magnification)

We found that MUC2 and B3GNT7 were decreased in colonic goblet cells; however, whether B3GNT7 is related to Golgi stress and the role of B3GNT7 in MUC2 maturation in AP remains unknown.

Inhibition of colonic B3GNT7 aggravates pancreatic injury and intestinal homeostasis in AP

To elucidate the role of B3GNT7 in AP, adeno-associated virus (AAV)-ablated colonic B3GNT7 mice were used (Fig. S2A). Real-time PCR demonstrated that AAV-B3GNT7 significantly and specifically reduced the mRNA and protein levels of B3GNT7 (Fig. S2B-C). There were no significant differences in pancreatic injury or systemic inflammation between the CON group and AAV-B3GNT7 group (P > 0.05) (Fig. 3A-E). In B3GNT7-knockdown AP mice, histological damage to the pancreas was significantly exacerbated, as evidenced by severe pancreatic edema, acinar cell necrosis, and inflammatory cell infiltration (P < 0.05) (Fig. 3A-B). Serum amylase activity and pro-inflammatory factor levels (IL-1β, TNF-α, and IL-6) in the pancreas and serum were significantly elevated in the AAV-B3GNT7 + AP group compared with the AP group (P < 0.05) (Fig. 3C-E).

Fig. 3.

B3GNT7 inhibition aggravated pancreatic injury and systemic inflammation during AP. (A) Histopathologic changes in the pancreas and pancreatic score (upper panel, 100× magnification; lower panel, 200× magnification). (B) Level of pancreatic edema (ww, wet weight; dw, dry weight). (C) Serum amylase activity. (D) The mRNA expression levels of pancreatic inflammatory factors (IL-1β, TNF-α, and IL-6). (E) Levels of serum inflammatory factors (IL-1β, TNF-α, and IL-6)

We also conducted a comprehensive analysis of intestinal homeostasis in AP mice. Compared with the AP group, the AAV-B3GNT7 + AP group exhibited greater intestinal damage characterized by extensive inflammatory cell infiltration, epithelial cell defects, and disordered arrangements (p < 0.05) (Fig. 4A). TUNEL staining revealed a significantly higher colonic epithelial cell apoptosis in the AAV-B3GNT7 + AP group than in the AP group (P < 0.05) (Fig. 4B). Furthermore, the levels of intestinal pro-inflammatory factors (IL-1β, TNF-α, and IL-6) were significantly elevated in the AAV-B3GNT7 + AP group compared with the AP group (P < 0.05) (Fig. 4C). The expression of tight junction proteins (TJPs, including Claudin-1, Occludin, and ZO-1) was significantly reduced in the AAV-B3GNT7 + AP group (P < 0.05) (Figs. 4D-F). Bacterial translocation from the intestinal tract to adjacent and/or remote tissues and organs owing to barrier disruption plays a critical role in the progression of AP [31]. Fluorescence in situ hybridization (FISH) analysis revealed increased bacterial translocation to the colonic mucosa and pancreas in the AAV-B3GNT7 + AP group than in the AP group (P < 0.05) (Fig. 4G-H). Culture of liver and mesenteric lymph node (MLN) suspensions on brain heart infusion agar (BHIA plates revealed a significantly higher number of colony-forming units (CFU) of anaerobic bacteria in the AAV-B3GNT7 + AP group than the AP group (P < 0.05) (Fig. 4I). By measuring the concentrations of FITC-dextran in the serum of mice from the AP group and the AAV-B3GNT7 + AP group (P < 0.05) (Fig. 4J), it was further confirmed that inhibition of intestinal B3GNT7 exacerbates intestinal barrier permeability. These findings imply that B3GNT7 inhibition exacerbates AP-induced pancreatic and intestinal damage.

Fig. 4.

B3GNT7 inhibition exacerbated intestinal injury and permeability and promoted bacterial translocation during AP. (A) Histopathological changes in the colons of AP mice and colonic score (upper panel, 100× magnification; lower panel, 200× magnification). (B) TUNEL staining and quantification of the colon (200× magnification). (C) The mRNA expression levels of IL-1β, TNF-α, and IL-6 in the colon of AP mice. (D) Immunofluorescence staining and quantification of Claudin-1 in the colons of AP mice (200× magnification). (E) Immunofluorescence staining and quantification of occludin in the colons of AP mice (200× magnification). (F) Immunofluorescence staining and quantification of ZO-1 in the colon (200× magnification). (G) FISH analyses of the pancreas and quantification performed using the EUB338 probe (200× magnification). (H) FISH analyses of the colonic epithelium and quantification performed using the EUB338 probe (200× magnification). (I) Colony-forming units (CFUs) were counted on anaerobic culture plates of the liver and mesenteric lymph nodes (MLNs). (J) Content of FITC-dextran in serum

Numerous studies have demonstrated that the gut microbiota undergoes alterations in AP [32–34]. Therefore, the effects of B3GNT7 on the gut microbiota of AP mice were explored using 16 S rRNA sequencing. The CON and AAV-B3GNT7 groups showed no significant difference in α- and β- diversity (Fig. 5A-B). The AAV-B3GNT7 + AP group presented decreased α-diversity in Sobs and Shannon index compared with the AP group (P < 0.05) (Fig. 5A). β- diversity by principal coordinate analysis (PCoA) and non-metric multidimensional scaling analysis (NMDS) revealed differences in microbiota composition among the AP, AAV-B3GNT7 + AP, and CON groups (Fig. 5B). Significant differences in bacterial taxa at the phylum and genus levels among groups were identified using the Kruskal-Walli’s test (Fig. 5C-D). The abundance of Dubosiella, which alleviates inflammation and improves mucosal barrier integrity [35], decreased significantly in both the AP and AAV-B3GNT7 + AP groups (P < 0.05) (Fig. 5D). Despite being elevated in the AAV-B3GNT7 compared to the CON group, Lactobacillus abundance decreased to the lowest level in the AAV-B3GNT7 + AP (P < 0.05) (Fig. 5D).

Fig. 5.

B3GNT7 inhibition aggravated gut dysbiosis during AP. (A) Alpha diversity analysis of the four groups using the Sobs index and the Shannon index. (B) Principal concordance analysis (PCoA) and non-metric multidimensional scaling (NMDS) analysis based on Operational Taxonomic Unit (OTU) abundance. (C) Differential analysis of gut microbiota at the phylum level in each group of mice. (D) Differential analysis of gut microbiota at the genus level in each group of mice

Taken together, these results demonstrate that B3GNT7 inhibition exacerbates intestinal permeability and promotes bacterial translocation, accompanied by a decreased abundance of two beneficial bacteria, suggesting a protective role of B3GNT7 in AP.

The decrease in intestinal B3GNT7 and mature MUC2 is mediated by Golgi stress in vivo and vitro

The aforementioned studies revealed that the expression of B3GNT7 and mature MUC2 decreased and exhibited colocalization under fluorescence microscopy during AP. IF was used to investigate whether B3GNT7 inhibition affects mature MUC2 in AP. Compared with the CON group, the AAV-B3GNT7 group exhibited a significant reduction in the number of mucus-containing goblet cells and MUC2 expression in the colon (P < 0.05) (Fig. 6A-D). Consistent with prior findings, UEA1-positive MUC2 expression was markedly decreased, suggesting impaired glycosylation (P < 0.05) (Fig. 6E-F). Compared with the AP group, the AAV-B3GNT7 + AP group also demonstrated a significant decrease in these indicators (P < 0.05) (Fig. 6A-F). These findings suggest that B3GNT7 participates in MUC2 glycosylation in goblet cells in AP.

Fig. 6.

B3GNT7 inhibition downregulated mature MUC2 expression during AP. (A) PAS-stained colonic sections in AP mice (upper panel, 100× magnification; lower panel, 200× magnification). (B) The number of mucus-containing goblet cells per crypt. (C) Immunofluorescence of MUC2 in the colon. (D) Quantification of MUC-2. (E) UEA1 lectin staining of mature MUC2 in the colon of AP mice. (F) Quantification of UEA1

To elucidate the involvement of B3GNT7 in Golgi stress in lipopolysaccharide (LPS)-stimulated HT-29 cells, we used the Golgi stress agonist BFA and inhibitor L-glutathione (GSH) in our vitro study. First, HT-29 cells were treated with different concentrations of BFA (0, 5, 10, and 15 µg/mL) and GSH (0, 0.5, 1, and 1.5 mmol/L) for 12 h and GOLPH3 expression was detected using real-time PCR. We found that at BFA concentrations of 0, 5, and 10 µg/mL, GOLPH3 expression increased in a concentration-dependent manner (P < 0.05); however, GOLPH3 expression decreased at a BFA concentration of 15 µg/mL although cell viability remained above 90% (P < 0.05) (Fig. S3A). Therefore, BFA was applied at 10 µg/mL in subsequent experiments. Similarly, HT-29 cells were treated with GSH, and 0.5 mmol/L was determined as the ideal concentration for subsequent experiments (Fig. S3B). Compared with the CON group, increased GOLPH3 and decreased GM130 expression indicated Golgi stress in LPS-stimulated HT-29 cells (P < 0.05) (Fig. 7A-E). Next, we explored the role of Golgi stress in B3GNT7 and mature MUC2 expression. Promotion of Golgi stress by BFA resulted in reduced expression of B3GNT7 and O-glycosylated MUC2 in LPS-stimulated HT-29 cells (P < 0.05) (Fig. 7D, F). Inhibition of Golgi stress by GSH upregulated the expression of B3GNT7 and O-glycosylated MUC2 (P < 0.05) (Fig. 7E, G). These results indicate that the downregulation of B3GNT7 mediated by Golgi stress is related to a reduction in O-glycosylated MUC2.

Fig. 7.

The impact of exacerbating or inhibiting Golgi stress on B3GNT7 and mature MUC2 in HT-29 cells. The mRNA expression levels of GOLPH3 (A), GM130 (B), and B3GNT7 (C) in HT-29 cells after BFA or GSH administration. Western blot images and protein quantification of B3GNT7, GOLPH3, and GM130 in HT-29 cells of BFA (D) and GSH (E) administration. Immunofluorescence of B3GNT7 and UEA1 lectin staining of mature MUC2 in HT-29 cells after BFA (F) and GSH (G) administration (200× magnification)

Intestinal B3GNT7 down-regulation is associated with AP severity both in patients and mice

We first used two mouse models of AP to verify the relationship between pancreatic and gut injury and B3GNT7 expression in AP. Spearman analysis indicated a significant negative correlation of B3GNT7 expression in intestinal epithelial cells with pancreatic and intestinal histopathological score (r = -0.8177, P < 0.05; r = -0.8095, P < 0.05) (Fig. 8A; Fig. S4A). We further studied their association in patients with AP. Intestinal specimens were collected via endoscopy from 22 patients with AP and 13 healthy controls (HC). Patients with AP were further divided into moderately severe AP (MSAP) and severe AP (SAP) groups. There were no significant differences in demographic data between the three groups (Table S1). Compared with the HC group, the expression of intestinal B3GNT7 in the MSAP and SAP groups was significantly reduced (P < 0.05) (Fig. 8B-C). B3GNT7 expression in the SAP group was significantly lower than that in the MSAP group (P < 0.05) (Fig. 8B-C). Intestinal B3GNT7 expression in patients of the MSAP and SAP groups was negatively correlated with the Ranson score (r = -0.5640, P < 0.05) and C-reactive protein (CRP) level (P < 0.05, r = -0.6110), whereas it was positively correlated with serum calcium level (P < 0.05, r = 0.7851) (Fig. 8D). Intestinal permeability measured via diamine oxidase (DAO) (P < 0.05, r = -0.8470) and D-lactate (P < 0.05, r = -0.8542) had a negative relationship with B3GNT7 expression (Fig. 8D).

Fig. 8.

Correlation analysis of B3gnt7 expression in the intestine and severity of AP. (A) Correlation analysis of B3gnt7 expression with pancreatic and colonic histological scores was performed using Spearman’s test. (B) The mRNA expression levels of B3gnt7 in duodenum of the patients with AP. (C) Immunofluorescence and quantification of B3GNT7 in duodenum of patients with AP (200× magnification). (C) Correlation analysis of B3gnt7 expression and the Ranson scores (p < 0.05, r = − 0.5640), CRP (p < 0.05, r = − 0.6110), serum calcium (p < 0.05, r = 0.7851), serum DAO (p < 0.05, r = − 0.8470), and D-lactate (p < 0.05, r = − 0.8542)

Based on these findings, we confirmed that decreased intestinal B3GNT7 expression was correlated with the severity of AP.

Discussion

In the present study, we observed a significant reduction in the expression of B3GNT7 and mature MUC2 in the colons of AP mice. B3GNT7 inhibition resulted in more severe pancreatic damage and systemic inflammation, which disrupted intestinal homeostasis and permeability and affected the intestinal mucosal barrier, thereby leading to intestinal dysbiosis and bacterial translocation. Downregulation of B3GNT7 and mature MUC2 was mediated by Golgi stress during AP in vivo and vitro. Additionally, we validated the expression of intestinal B3GNT7 in patients with AP and its correlation with disease severity.

We have previously conducted a series of studies on the potential mechanisms of intestinal barrier dysfunction during AP and demonstrated that MUC2 alleviates acute necrotizing pancreatitis (ANP) by maintaining intestinal homeostasis [8]. Consistent with this, the results of our current study showed that the levels of O-glycosylated MUC2 were significantly decreased in the colons of AP mice. These findings combined with changes in the colonic transcriptome and existing research on DEGs led to the identification of B3GNT7 as the primary target of interest. The mucus layer serves as the first barrier on the colonic surface, with MUC2, synthesized by goblet cells and containing O- and N-linked glycans, being its most important component [28]. B3GNT7, a glycosyltransferase highly expressed in the colon, is involved in the synthesis of glycans and elongating poly-LacNAc chains to provide substrates for fucosylationc [17]. Notably, studies have shown that various signals, including IL-22 and oxytocin, regulate the maturation of protective mucin via B3GNT7-mediated fucosylation [16, 18]. This study demonstrates that the B3GNT7 inhibition downregulates the glycosylation levels of MUC2 during AP, further extending our previous work.

Multiple clinical studies have confirmed the pivotal role of B3GNT7 in the maintenance of intestinal homeostasis. Reduced B3GNT7 expression in the gut has been observed in patients with both colon cancer and UC, underscoring its significance in the intestinal mucosal barrier [36]. Our study specifically examined the function of B3GNT7 in AP and the underlying mechanisms. We found that, compared with the AP group, B3GNT7-ablated AP mice exhibited significantly increased colonic injury and intestinal permeability. Glycosylation of MUC2 plays a key role in maintaining the integrity of the intestinal mucus barrier, regulating the balance of the gut microbiota, and preventing intestinal diseases. Therefore, our study suggests that B3GNT7 disrupts intestinal homeostasis by downregulating glycosylated MUC2. Although the role of B3GNT7 in intestinal diseases such as colitis and colon cancer is well established, our research expands the understanding of its role in AP-related intestinal barrier integrity.

Bacterial translocation refers to the passage of intestinal bacteria and/or bacterial toxins across the intestinal mucosal barrier into the mesenteric lymph nodes and the portal venous system, ultimately entering the systemic circulation and reaching distant organs such as the liver, spleen, lungs, and pancreas. Although the precise routes of bacterial translocation have not been fully elucidated, the blood, ascites, and mesenteric lymph nodes may all serve as important pathways, with intestinal barrier dysfunction representing a “common conduit” for all potential translocation routes [37]. Normal functioning of the intestinal barrier is crucial for preventing bacterial translocation [38]. Dysfunction of the intestinal barrier leads to bacterial translocation, which can trigger enterogenic infections and even sepsis, thereby exacerbating the patient’s condition during AP [5]. In the present study, B3GNT7-knockdown mice exhibited increased intestinal permeability and more pronounced bacterial translocation. Furthermore, significant alterations in the bacterial community structure were observed in B3GNT7-knockdown mice, notably with a marked reduction in the beneficial bacterium Dubosiella, which plays a crucial role in regulating the colonic Treg/Th17 immune balance by producing short-chain fatty acids, particularly propionic acid, and lysine, which in turn inhibits colonic inflammatory responses and enhances the integrity of the mucosal barrier [35]. As bacterial levels in the circulatory system were not directly assessed in this study, the potential involvement of the circulatory route in bacterial translocation remains to be elucidated.

Golgi stress was observed during AP. B3GNT7 is a specific glycosyltransferase localized in the Golgi apparatus. Therefore, we investigated the relationship between changes in B3GNT7 levels and Golgi stress under inflammatory conditions in vitro. The Golgi apparatus is responsible for the post-translational modification and transport of proteins and vesicles [14, 39]. Prolonged Golgi stress can lead to irreparable damage, prompting the body to initiate an autophagy response [12]. Numerous studies have reported the role of Golgi stress in the onset and progression of various diseases [40]. A recent study identified a novel pathway involved in Golgi stress response [41]. When the mucin-type glycosylation capacity of the Golgi apparatus is inadequate, the mucin pathway is activated, enhancing its role in mucin-type glycosylation. This study demonstrated that HT-29 cells exhibit reduced expression of Golgi-specific proteins under conditions of stress. Furthermore, we found that AP indeed activates a Golgi stress response and regulates the downstream intestinal target B3GNT7.

In summary, our findings indicate that the downregulation of intestinal B3GNT7, mediated by Golgi stress, leads to a reduction in mature MUC2, resulting in the disruption of intestinal homeostasis and exacerbation of AP. Colonic B3GNT7 exerts a protective effect in AP mice, positioning it as a promising target for the treatment of AP through the maintenance of intestinal homeostasis. However, the two murine AP models utilized in this study involve distinct pathogenic mechanisms (inflammatory compared to metabolic) and have different time courses. Notably, the endpoint for assessment differed between models (12 h for caerulein/LPS-AP compared to 72 h for L-arginine-AP) based on their established pathophysiology peaks. While this approach allowed us to demonstrate the consistent role of B3GNT7 across different AP etiologies, the temporal discrepancy precludes direct comparison of the absolute severity or the dynamic progression of injury between the two models. Future studies should focus on the validation of this results in clinical samples.

Author contributions

Jia Hu, Yang Fu, and Zehua Huang conceived and designed the research. Jia Hu and Yang Fu performed the research and generated the draft of the manuscript. Nuoming Yin and Binqiang Xu analyzed and interpreted the data. Huizheng Huang, Mengyan Cui, and Junjie Fan provided supports in the animal experiments. Yue Zeng, Qixiang Mei, and Chunlan Huang edited and revised the manuscript.

Funding

This study was funded by the National Natural Science Foundation of China (No. 82270671 and No. 82470675), the National Natural Science Foundation–Youth Foundation (No. 82200714 and No. 82400752), Shanghai General Hospital Characteristic Research Program (CCTR-2022B02), Shanghai Municipal Health Commission Clinical Research Projects (No. 202440170) and Shanghai Jiao Tong University School of Medicine Xinyi Digestive Disease Research (KY-2023-03-01).

Data availability

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

Declarations

Ethics approval and consent to participate

This study involving human samples was approved by the Ethics Committee of Shanghai General Hospital (Approval No. 2021035) and was registered with the China Clinical Trial Registry (ChiCTR1800017214). All procedures performed in studies involving human participants were in accordance with the 1964 Helsinki Declaration and its later amendments. Experimental animal protocols were conducted under specific pathogen–free (SPF) conditions.

Competing interests

The authors declare no competing interests.

New & Noteworthy

This study reveals that downregulation of B3GNT7, mediated by Golgi stress, disrupts MUC2 O-glycosylation, exacerbating intestinal homeostasis and acute pancreatitis (AP). B3GNT7 reduction correlates with increased inflammation, intestinal permeability and gut microbiota dysbiosis, highlighting its critical role in AP pathogenesis.

Patient consent for publication

Written informed consent was obtained from all individual participants included in the study, which covered the publication of de-identified data and associated findings.