Berberine (C₂₀H₁₈NO₄·Cl; CAS: 633-65-8; molecular weight: 371.81; high-performance liquid chromatography purity ≥98%) was purchased from Melone Pharmaceutical Co., Ltd (Dalian, China). Irinotecan (CPT11) (C₃₃H₄₅N₄O₉·Cl, CAS: 136572-09-3; molecular weight: 677.18) was purchased from Melone Pharmaceutical Co., Ltd (Dalian, China). β-glucuronidase (GUS, CAS: 9001-45-0, G7396-25kU) from Escherichia coli was obtained from Thermo Scientific Inc. (Waltham, MA, United States). NCM460 human intestinal epithelial cells, CT26 murine colon carcinoma cells and Caco2 human colorectal adenocarcinoma cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA, United States). Roswell Park Memorial Institute (RPMI)-1640, Dulbecco’s modified Eagle’s medium (DMEM), 100 U/ml penicillin/streptomycin, and fetal bovine serum (FBS), were purchased from Gibco BRL (Grand Island, NY, United States). Diethylpyrocarbonate-treated water, dimethyl sulfoxide, paraformaldehyde, and diaminobenzidine were obtained from Sigma-Aldrich (Shanghai, China). We also acquired an enhanced chemiluminescence (ECL) detection kit from Millipore (Billerica, MA, United States). Rabbit antibodies raised against inducible nitric oxide synthase (iNOS) (#18985-1-AP), cyclooxygenase-2 (COX-2) (#12375-1-AP), and Claudin-7 (#10118-1-AP), were obtained from Proteintech Group (Chicago, IL, United States). A β-actin antibody (#4970) was obtained from Cell Signaling Technology (Danvers, MA, United States). Rabbit antibodies against zonula occludens 1 (ZO-1, #A11417), Claudin-1 (#A2196), and Occludin (#A2601), were obtained from ABclonal Technology (Wuhan, China). Trizol and a SuperScript II Reverse Transcriptase kit were purchased from Thermo Scientific Inc. (Waltham, MA, United States). SYBR Premix ExTaq Mix was purchased from Takara Biotechnology (Shiga, Japan).

Both healthy female C57BL/6 mice (8-week-old, 20–22 g) and male BALB/c mice (6-week-old, 20–22 g) were purchased from the Shanghai Laboratory Animal Center in Shanghai University of Traditional Chinese Medicine. All mice were maintained in specific pathogen-free facility and kept under controlled conditions at a humidity of 60–70% and a temperature of 23–25°C in a 12 h light/dark cycle. The mice had access to autoclaved food and drinking water ad libitum. All animal experiments were conducted in accordance with the principles of the declaration and recommendations of the Animal Experimentation Ethics Committee at Shanghai University of Traditional Chinese Medicine (Animal Ethics Number: PZSHUTCM200828001 and PZSHUTCM200724006, respectively).

Experimental intestinal mucositis was induced in mice by the intraperitoneal injection of CPT11 (50 mg/kg). C57BL/6mice were randomly divided into the following groups (n = 6 mice per group): control group, CPT11 group, and CPT11 + Berberine group. Control mice were injected with vehicle (5% DMSO). Except for the control group, all other groups of mice were injected with CPT11 (50 mg/kg) for 14 days to induce intestinal mucositis. Mice in the CPT11 + Berberine group were treated daily with berberine (50 mg/kg/day) from day 1 to day 14. Berberine was dissolved in 0.5% methylcellulose and administered by oral gavage. Body weights were recorded daily, and mice were sacrificed under anesthesia after receiving the last gavage. Blood was collected from the eye orbits of each mouse. Subsequently, the entire colon was removed, and the total length was measured. Then, the feces were collected from the interior of the rectum. Distal colon tissues were collected for hematoxylin–eosin (H&E) staining and Alcian blue (AB)/Periodic Acid-Schiff stain (PAS) staining. The number of goblet cells was determined by AB/PAS staining. Histological injury was evaluated as a combined score of inflammatory cell infiltration (score 0–3) and mucosal damage (score 0–3), as previously described (Ren et al., 2020).

Healthy male BALB/c mice (6-week-old, 20–22 g) were used to establish tumor-bearing animal model. Briefly, CT26 colon cancer cells (1 × 10⁶ in 100 μL of PBS) were harvested and subcutaneously injected into the right flank of each mouse. Three days later, all tumor-bearing animals were divided into the following groups (n = 6 mice per group), control group, CPT11 group, and CPT11 + berberine group. At the end of experiment, all of the mice were euthanized under anesthesia and the tumors were removed and weighed. Tumor volume was then determined by measuring length (L, mm) and width (W, mm) and applying the following formula: V = 0.5 × L × W².

To investigate intestinal permeability in vivo, we used FITC-dextran (MW: 3,000–5,000, Shanghai Aladdin Biochemical Technology Co., Ltd, China) as a tracer. In brief, after fasting for 12 h, mice were administered with FITC-dextran (44 mg/100 g body weight). 4 h later, blood was collected upon euthanasia and allowed to clot for 1 h. After centrifugation (3,000 rpm, 15 min), serum was collected and diluted with the same volume of PBS. Finally, the FITC-dextran concentration was determined by spectrophotofluorometry (Fluoroskan Microplate Fluorometer, Thermo Fisher Scientific) at an excitation wavelength of 480 nm and an emission wavelength of 520 nm. We also determined the serum levels of lipopolysaccharide (LPS) and diamine oxidase (DAO) with ELISA kits (Yingxin laboratory Co., Ltd, Shanghai, China) in accordance with the manufacturer’s instructions (n = 5–6 per group).

NCM460 cells (1.5 × 10⁵ cells/ml) were seeded into transwell cell culture chambers (6.5 mm diameter inserts, 3.0 μm pore size) (Corning Costar, Cambridge, MA) and the growth medium changed every other day. A NCM460 cell monolayer was formed after 10 days of culture. Then, NCM460 cells were administrated with berberine (50 μM), SN38 (250 nM), and SN38 (250 nM) plus berberine (50 μM), for 24 h. The TEER values were then determined with an ohmmeter and chopstick electrodes (Millipore ESR-2; Burlington, MA, United States). Data are presented as the resistance per unit area and was calculated by dividing resistance values by the effective membrane area. Inserts without cells were used as blanks.

The permeability of the NCM460 cell monolayer was determined by analyzing the flux in FITC-dextran. FITC-dextran (1 mg/ml) was added to the apical compartment of the insets. After 6 h of incubation, we collected aliquots of basolateral medium to allow us to measure fluorescence at 480 nm excitation and 520 nm emission wavelengths.

Caco-2 cells were seeded onto coverslips in a 24-well plate at a density of 4 × 10⁵ cells/well and grown to full confluence. After treatment with berberine (50 μM), SN38 (250 nM), and SN38 (250 nM) plus berberine (50 μM), for 24 h, cells were washed three times with PBS and fixed with 4% paraformaldehyde for 10 min at room temperature. Then, cells were permeabilized with 0.1% Triton X-100 for 5 min and then blocked with 3% BSA for 1 h at room temperature. Caco-2 cells were then incubated overnight with rabbit anti-Occludin antibody (1:50, #A2601, ABclonal) at 4°C in the dark. Then, coverslips were washed and incubated with Alexa Fluor 594 goat anti-rabbit IgG secondary antibody (1:200, #AS039, ABclonal) for 60 min at room temperature in the dark. Finally, cells were counterstained with 4′,6-diamidino-2- phenylindole (DAPI) (Beyotime, Jiangsu, China) for 5 min at room temperature and visualized with a fluorescence microscope (Olympus CKX41, Tokyo, Japan).

Proteins were extracted from colon tissues (1–1.5 cm proximal to the anus) and cultured NCM460 cells by homogenization in lysis buffer containing protease and phosphatase inhibitor cocktail tablets (Roche Diagnostics GmbH, Mannheim, GER). The supernatant was collected after centrifugation (4°C, 12,000 g, 15 min). Proteins (30 μg) were separated by 10% SDS-PAGE and transferred onto a PVDF membrane. The membrane was blocked in 5% (w/v) skimmed milk for 2 h at room temperature and then immunoblotted with primary antibody. Then, blots were washed and incubated with a HRP-coupled secondary antibody at room temperature. Finally, blots were visualized by enhanced chemiluminescence (ECL) detection reagents. Protein expression levels were then analyzed by a GS-700 imaging densitometer (Bio-Rad, CA). β-actin was used as an internal control.

Total RNA was extracted using TRIzol Reagent (Life Technologies) according to the manufacturer’s protocol. Complementary DNA (Cdna) was reverse transcribed using the SuperScript Ⅱ Reverse Transcriptase kit (Thermo Fisher Scientific). Real-time PCR was performed using SYBR Green qPCR Master Mix (Applied Biosystems). Primers for the inflammatory mediators and internal reference were as follows: iNOS: forward, 5′-GTC​CTA​CAC​CAC​ACC​AAA​CT-3′, reverse, 5′-ATC​TCT​GCC​TAT​CCG​TCT​C-3′; IL-8: forward, 5′-ATT​CAT​TCC​TCT​CAA​ACT​CAT​T-3′ reverse, 5′-GCC​AAC​AGT​AGC​CTT​CAC-3’; Interleukin 1 beta (IL-1β): 5′-ATT​GTG​GCT​GTG​GAG​AAG​AAG​A-3′, reverse, 5′-TGA​AGG​AAA​AGA​AGG​TG-3′; TNF-α: forward, 5′-CTC​TTC​TCA​TTC​CTG​CTT​GT-3′, reverse, 5′-GTG​GTT​TGT​GAG​TGT​GAG​G-3′; Occludin: forward, 5′-ATG​TCC​GGC​CGA​TGC​TCT​C-3′, reverse, 5′-TTT​GCT​GCT​CTT​GGG​TCT​GTA​T-3’; ZO-1: forward, 5′-GCC​GCT​AAG​AGC​ACA​GCA​A-3′, reverse, 5′-TCC​CCA​CTC​TGA​AAA​TGA​GGA-3’; Claudin-7: forward, 5′-GGC​CAC​TCG​AGC​CTT​AAT​GGT​G-3′ reverse, 5′- CCT​GCC​CAG​CCG​ATA​AAG​ATG​G-3’; β-actin: forward, 5′-GGG​AAA​TCG​TGC​GTG​AC-3′, reverse, 5′-AGG​CTG​GAA​AAG​AGC​CT-3′. qPCR was then performed with the Takara SYBR Green Master Mix Kit and was quantitatively analyzed using the ABI Prism 7900HT Sequence Detection System (Life Technologies, Carlsbad, CA, United States) in accordance with the manufacturer’s instructions. We used β-actin as an internal control.

The in vitro inhibition of bacterial GUS enzymes was assessed by a fluorescent GUS substrate 4-methylumbelliferyl-β-D-glucuronide (4-MUG) (CAS: 6160-80-1, MW: 352.29, ShangHai YuanYe Bio-technology Co., Ltd, China). Reactions consisted of 40 μL of GUS (500 ng/ml), 30 μL of 4-MUG (500 μM), various concentrations of berberine (0, 1.25, 2.5, 5, 10, 20, 40, 60, 80, 100, 200 μM), and assay buffer [75 mM, phosphate buffer, (pH 6.8)]. Reactions were initiated by the addition of 4-MUG and were then incubated for 30 min at 37°C. Then, the end point absorbance was determined by a microplate reader with an excitation wavelength of 350 nm and an emission wavelength of 450 nm.

Next, the inhibitory activity of berberine against E. coli β-glucuronidase was carefully investigated. Inhibition constant (Ki) values were determined by applying various concentrations of 4-MUG in the presence or absence of different concentrations of berberine. To determine the inhibition kinetic type (competitive inhibition, non-competitive inhibition, or uncompetitive), multiple concentrations of 4-MUG and various berberine concentrations were utilized to determine the corresponding reaction rates. The inhibition kinetic type was determined by identifying the intersection point in Dixon and Lineweaver-Burk plots. A second plot, featuring slopes from the Lineweaver-Burk plot versus inhibitor concentrations, was then used to calculate the corresponding inhibition parameter (Ki) value.

The GUS activity in fecal was monitored by the absorbance at 405 nm from the p-nitrophenol release using 4-Nitrophenyl β-D-glucopyranoside (PNPG) at the substrate. Briefly, frozen fecal pellets were homogenized in potassium phosphate (PB) buffer (pH 7.4, 0.01 M). Then the supernatant of all samples were collected after centrifugation (10,000×g for 20 min) and total protein concentration was determined using the pierce BCA Protein Assay Kit (Yeasen Biotech Co., Ltd, Shanghai, China). Subsequently, fecal homogenate from all groups containing 100 μg/ml protein (20 μL) was incubated with 0.5 mM PNPG (20 μL) and PB buffer (80 μL) at 37°C for 30 min. After incubation, 250 μL sodium hydroxide (NaOH) (0.5 M) was added to stop the reaction, and the absorbance was measured at 405 nm by a microplate reader.

GUS-producing bacteria was detected by 4-Methylumbellifery-β-D-Glucuronide (4-MUG) agar culture plate. In brief, the weighed fecal pellets (50-70 mg) were homogenized and centrifuged at 2000 rpm for 5 min. Then, centrifugation of the isolated supernatant was centrifuged at 10,000 rpm for 20 min, and the precipitation was collected then used for the GUS-producing bacteria detection. Finally, the precipitation of all samples were suspended in physiological saline (500 μL), then coated on the culture plate of nutrient agar plus 4-methylumbellifery β-D-glucuronide (4-MUG) and incubated for 12 h at 37°C. GUS-producing bacteria was detected by UV-light in 366 nm.

Berberine (50 mg/kg), or an equivalent volume of vehicle control (200 μL of 0.5% MC-Na) was administered into mice (n = 3/group) orally for three consecutive days. After 30 min of the final administration of berberine, mice were administered (by oral gavage) with fluorescein Di-β-D-Glucuronide (FDGlcU, 7.3 μmol/kg, 0.1 ml per mouse) (CAS: 129,787-66-2, BioRuler, CT, United States). Mice were euthanized immediately prior to imaging with a IVIS Spectrum image system (Bruker) (470 nm excitation/535 nm emission filters). GUS activity was quantified by measuring the mean fluorescence in a given region of interest after subtracting the background; this was held constant throughout the imaging period.

Molecular docking experiments were carried out with AutoDock 4.2.6 program to investigate the interaction between berberine and GUS protein. The three-dimensional structure of GUS protein (PDB code: 3K46 (Wallace et al., 2010). resolution: 2.0 Å) was obtained from the Protein Data Bank. The initial structure was then prepared using AutoDockTools 1.5.6, preserving the original charges of the protein and generating a pdbqt file for docking (Sanner, 1999). The two-dimensional structure of berberine was downloaded from the PubChem database. The MOPAC program was then used to optimize the structure and calculate the PM3 atomic charge (Stewart, 1990; Stewart, 2004). The structure of berberine was also prepared by AutoDock Tools 1.5.6, and the corresponding pdbqt file was generated for docking. The hydrophobic region of GUS was chosen as the binding pocket for docking. The coordinates of the grid box center were then determined (−14.604, −32.129, 47.104) and the number of grid points in the XYZ of grid box was set to 80 × 60 × 60. The grid spacing was 0.375 Å and the number of GA runs was set to 100. The remaining parameters were set to default.

Data were compared between multiple groups by one-way analysis of variance (ANOVA) and GraphPad Prism 7 software (GraphPad Software, La Jolla, CA, United States). Data are shown as mean ± SD. p-values < 0.05 (two-sided) were considered significant (*p < 0.05, **p < 0.01, ***p < 0.001).

Compared with the vehicle treatment control mice, the mice receiving CPT11 treatment exhibited significant body weight loss (Figure 1A). Berberine treatment attenuated the CPT11-induced body weight loss. Furthermore, colon shortening was identified as an indirect symptom, thus indicating the intensity of mucosal inflammation (Ren et al., 2020). As anticipated, the CPT11-induced colon shortening was remarkable, but was significantly mitigated by berberine treatment (Figures 1B,C). Further histopathological analysis showed that CPT11 caused severe intestinal tissue damage, including severe epithelial injury, loss of architecture, muscle thickening, and neutrophil infiltration. Berberine treatment mitigated the CPT11-induced loss of mucosal architecture, ulceration, and cellular infiltration (Figure 1D).

We next investigated the expression levels of multiple inflammatory mediators. The treatment of mice with CPT11 for 14 days significantly increased the levels of COX-2, iNOS, IL-8, IL-1β, and TNF-α in colon tissues (Figures 2A,B). However, the upregulated levels of these inflammatory mediators, as induced by CPT11, were significantly reduced by berberine treatment. Additionally, we observed a significant decrease in the number of goblet cells in the colons of mice receiving CPT11 (Figure 2C). Berberine significantly increased the number of goblet cells in CPT11-treated mice. Furthermore, berberine treatment increased the levels of the tight junction proteins ZO-1 and Claudin-7 (Figure 3B), as well as the mRNA levels of Occludin, ZO-1, and Claudin-7 in CPT11-induced intestinal mucositis mice (Figures 2B,C).

To determine the effects of berberine on the mucosal barrier, we performed immunofluorescence staining to examine the expression of ZO-1 in the colon of CPT11-induced mucositis mouse. CPT11-treated mice showed a disruption of the intestinal intercellular structure and a reduced expression of ZO-1 in the colon (Figure 3A). In addition, intestinal permeability in mouse plasma was assessed by FITC-dextran fluorescence (Table). Mice exposed to CPT11 exhibited higher FITC positive signals, indicating that the intestinal barrier had been disrupted by CPT11. Berberine treatment reduced the FITC signal, thus indicating an improvement in the barrier integrity of CPT11-treated mice. Similar results were observed for the concentrations of lipopolysaccharide (LPS) and diamine oxidase (DAO) (Table 1), both of which are serum indicators of mucosal injury. Collectively, these data indicate that berberine effectively prevents CPT11-induced dysfunction in the intestinal barrier.

We further evaluated the effects of berberine on the intestinal barrier in vitro. NCM460 human intestinal epithelial cells were incubated with berberine, SN38, and SN38 + berberine, respectively. We found that berberine significantly suppressed the SN38-induced trans-endothelial electrical resistance (TEER) decline (Figure 4A), as well as the increase in FITC-Dex permeability (Figure 4B). Furthermore, SN38 effectively inhibited the wound healing of NCM460 cells; however, berberine treatment improved this situation (Figure 4C). Then, we assessed the expression levels of tight junction proteins. We found that treatment of NCM460 cells with berberine significantly improved the expression levels of ZO-1, Occludin, and Claudin-1 (Figure 4D). Then, Caco-2 cells were cultured to form a monolayer to simulate the intestinal barrier in vitro. Immunofluorescence assay showed that the level of Occludin protein was significantly decreased in Caco-2 monolayer treated with SN38. However, pre-incubation with berberine increased the Occludin expression level in Caco-2 cells (Figure 4E). Together, these results indicate that berberine protects the barrier function from SN38-induced injury in vitro.

Intestinal bacterial β-glucuronidase GUS catalyzes SN38G to SN38, and SN38 is considered to be the cause of CPT11-induced intestinal mucositis. To explore whether administration of berberine affacts the activity of GUS in fecal samples, we performed the 4-Nitrophenyl β-D-glucopyranoside (PNPG) assay using fecal lysates. The total GUS activity increased significantly after 14 days of CPT11 exposure (Figure 5A). However, berberine treatment significantly reduced the GUS enzyme activity in mice that had been exposed to CPT11 (p < 0.001) (Figure 5A). Subsequently, we used 4-Methylumbellifery-β-D-Glucuronide (4-MUG) agar to analyze GUS-producing bacteria. The results showed that berberine markedly reduced the levels of GUS-producing bacteria in fecal samples (Figure 5B). Furthermore, a non-fluorescent GUS probe (FDGlcU) was used to verify the attenuated effects of berberine on GUS activity in the intestine. As shown in Figure 5C, mice receiving FDGlcU (7.3 μmol/kg) by oral gavage and exposed to CPT11 showed increased fluorescent signal, which represents increased intestinal GUS activity. However, the fluorescent signal in CPT11 treatment mice was slighter following berberine administration. On the other hand, the LC-MS technology was used to determine the concentration of CPT11 and SN38 in feces. Relatively low levels of SN38 in faecal samples were detected in berberine + CPT11-treated mice compared with CPT11-treated mice (Supplementary Figure S1). Collectively, these data indicate that berberine effectively inhibits bacterial GUS activity, reduces GUS-producing bacteria, and decreases the production of SN38 in a mouse model of CPT11-induced intestinal mucositis.

Next, we performed molecular docking assays to evaluate the potential binding of berberine to E. coli GUS enzyme. Inhibitor 1, as a positive control compound, is an inhibitor of GUS enzyme that has been used in previous studies (LoGuidice et al., 2012; Saitta et al., 2014; Bhatt et al., 2020). We found that both berberine and inhibitor 1 played roles in the catalytic cavity of GUS with relatively low binding energies of −7.8 and −8.6 kcal/mol, respectively. Berberine formed three hydrogen bonds between the hydroxyl groups at C-9 and C-10, and several amino acid residues (Asp163, His162, and Ser 557) (Figure 6A). In addition, berberine associated with Leu561, Leu361, Tyr469, and Tyr472 via Pi-Anion, Alkyl, Pi-Alkyl, and Pi-Pi stacked interactions; these interactions may contribute to the binding affinity between E. coli GUS and berberine. Subsequently, a 4-MUG assay demonstrated the inhibitory effects of berberine on GUS enzyme with an IC₅₀ value of 54.04 ± 5.104 μM (Figure 6B). To further explore the inhibitory mechanisms of berberine on GUS, we prepared Lineweaver-Burk plots (substrate concentration ⁻¹ versus velocity ⁻¹) and found that berberine acted in a non-competitive manner against GUS-mediated 4-MUG hydrolysis (Figure 6C). These results indicate that berberine functions as a potential non-competitive inhibitor against bacterial GUS enzyme.

Finally, we evaluated whether the inhibition of bacterial GUS by berberine affects the anti-tumor efficacy of CPT11 in tumor-bearing mice. A tumor xenograft mouse model was established by injecting mice with CT26 cells, a murine colorectal carcinoma cell line. As shown in Figure 7, compared with controls, the tumor weight (Figure 7B) and volume (Figure 7C) were significantly reduced after CPT11 treatment. Notably, the tumors in mice in the berberine + CPT11 co-treatment group grew more slowly when compared with those in the CPT11 alone treatment group. Then, we investigated the effect of berberine alone on tumor growth and found that berberine inhibited tumor growth in a colorectal cancer xenograft model (Supplementary Figure S2). Collectively, these data indicate that berberine inhibits intestinal GUS activity and attenuates CPT11-induced intestinal mucositis without impairing CPT11-mediated tumor regression.