Animal models and experimental design

Healthy female C57/BL mice, aged 7–8 weeks and weighing between 18 and 20 g, were sourced from Chengdu Dashuo Technology Biology Co., Ltd. To create a model of chronic radiation proctitis, a single high-dose irradiation was administered to the pelvic area. The mice were anesthetized via an intraperitoneal injection of 1% pentobarbital sodium and positioned supine within a 4-mm thick lead box to target the pelvic region (Figure S1A). The irradiation was conducted using a Rad Source RS2000Pro medical X-ray irradiator from the State Key Laboratory of Biotherapy at West China Hospital, Sichuan University, delivering a dose of 10 Gy at a rate of 1.32 Gy/min [18].

The study involved categorizing the mice into four distinct groups: the blank control group, the irradiation-only group, the Liriodendron gavage group, and the Liriodendron enema group, with each group consisting of 16 mice. The Liriodendron compound (C34H46O18) was sourced from Tianjin Furui Xiang Technology Co., Ltd., and its chemical structure is illustrated in Figure S2. Following a 24-h period post-irradiation, the groups receiving Liriodendron were treated with a dosage of 100 mg/kg, either through gavage or enema, while the blank control and irradiation-only groups were given a weekly administration of a 1% dimethyl sulfoxide solution (DMSO) for a duration of 8 weeks. All mice were kept in a pathogen-free environment with ample access to food and water. Tissue samples were collected at the end of the 8-week period following radiation therapy. The Animal Care and Use Committee of the Animal Research Institute at Sichuan Provincial People's Hospital approved all experimental procedures, under the ethical reference number Ethics (Research) No. 65 of 2017.

Fecal occult blood test (FOBT)

At the 1 st, 4th, and 8th weeks following irradiation, fresh fecal samples were gathered from each group of mice into sterile EP tubes and combined with a small volume of saline. Subsequently, fecal occult blood test strips were employed to assess the presence of occult blood in the feces of each group, using the following scoring system: negative: no blue-green color appeared within 3 min, resulting in a score of 0; weakly positive: a blue hue was detected within 30–60 s, earning a score of 1; positive: a blue-green color was seen immediately, scoring 2; strongly positive: blood was visible to the naked eye, with an immediate dark blue color, scoring 3; strong positive: blood was also visible to the naked eye, with a deep blue color appearing instantly, also scoring 3.

Histological and immunohistochemical analysis

Samples were collected from the top one centimeter of the mouse's anal region. The preserved tissues underwent dehydration through an automated machine, followed by embedding, with the sections rehydrated in water.

Hematoxylin and eosin (H&E) staining involves several steps: first, immerse the sample in hematoxylin for 10–20 min, followed by a rinse in tap water for 1–3 min. Next, treat with acid alcohol for 5–10 s, rinse again, and then use warm water until the sample turns blue. After another rinse, place it in 85% alcohol for 3–5 min. Eosin is then applied for 3–5 min, followed by rinsing, gradient alcohol dehydration, xylene clearing, and sealing with neutral gum. For MASSON staining, the procedure begins with similar sectioning and dewaxing. The sample is incubated overnight in potassium dichromate, then differentiated back to blue with hematoxylin. Titrate with Lichun red for 10 min, rinse, and treat with phosphomolybdic acid for 1 min. Stain with aniline blue for 2 min, dehydrate using alcohol, clear with xylene, and seal with neutral glue. Images were captured using a 3DHISTECH Panoramic 250 scanner from Budapest, Hungary, initially at 40 × for an overview, followed by 100 × and 400 × for detailed examination. The positive expression area was evaluated using Image-Pro Plus 6.0 software, calculating the percentage as the ratio of positive area to total area (in pixels). Statistical analysis was performed using SPSS 23.0 for t-tests, with results presented as mean ± SD.

Immunofluorescence procedure involved several steps: first, the paraffin sections were rehydrated; then, antigen retrieval was performed. A blocking solution containing goat serum was applied dropwise and allowed to sit at room temperature for 20 min. Following this, a CD4 antibody (ab183685 from Abcam, California, USA) was added dropwise and incubated overnight at 4 °C. The sections were washed three times for 5 min each with PBS. Next, a secondary antibody (FITC-conjugated goat anti-rabbit) (GB22303 from Servicebio, Wuhan, China) was added dropwise and incubated for 30 min at 37 °C, followed by another three washes in PBS. DAPI was then added dropwise, and the sections were incubated at room temperature for 10 min before being washed three more times in PBS. Finally, the slides were sealed with an anti-fade mounting medium. Observations were made at 100 × magnification, and images were captured at both 100 × and 400 × magnifications across three different fields. The Image-J analysis software was used to measure the integrated density (IntDen) and area of the images, from which the mean gray value (mean) was derived. The mean fluorescence intensity for each sample was calculated based on the average of two images. Additionally, a CD8 antibody (ab217344 from Abcam) and a secondary antibody (CY3-labeled goat anti-mouse) (GB21301 from Servicebio) were utilized. In the resulting images, DAPI-stained nuclei appeared blue, CD8 expression was indicated in red, and CD4 expression was shown in green.

Cytokine and protein quantification

The enzyme-linked immunosorbent assay (ELISA) was conducted using TNF-α, IFN-γ, IL-10, and TGF-β1 from Ruixinbio in Quanzhou, China, employing a double antibody sandwich method. Following incubation and thorough washing, any unbound substances were eliminated, resulting in the formation of a sandwich complex of solid-phase antibodies, antigens, and enzymatic antibodies on the microtiter plate's surface. Substrates A (0.01% hydrogen peroxide) and B (0.1% TMB) were introduced, leading to a blue product catalyzed by HRP, which turned yellow upon the addition of a stopping solution (2 M sulfuric acid). The optical density (OD) was recorded at a wavelength of 450 nm using an enzyme marker (Rayto, RT-6100). The OD exhibited a positive correlation with the concentrations of the analytes in the samples. The standard concentrations were plotted on the horizontal axis (comprising six standard wells and one additional well, totaling seven concentration points), while the corresponding OD values were plotted on the vertical axis. A four-parameter logistic curve fit (4-pl) was generated using software to establish a standard curve, enabling the calculation of sample concentrations based on their OD. The correlation coefficient (r-value) for the calibration product dose–response curve was found to be 0.9900.

Immunoblotting: Tissue samples were collected from the uppermost centimeter of the mouse anus. The rectal tissues were lysed using 200 μL of P0013B RIPA strong lysis buffer (Beyotime, Shanghai) supplemented with 1 mM PMSF, and then homogenized with a glass homogenizer (YJQ0928Q). Following centrifugation at 10,000–14,000 g for 3–4 min, the supernatant was utilized for the BCA protein concentration assay (Therm Scientific, USA). A quarter volume of 5* SDS-PAGE loading buffer (P1040 Solarbio) was mixed with the lysate and heated at 100 °C for 10 min. The protein samples were then separated via SDS-PAGE and transferred onto PVDF membranes, which were blocked using a protein-free rapid closure solution (Yarase, Shanghai) for 15–20 min. The membranes were incubated overnight at 4 °C with either glucose transporter α-SMA antibody (ER1003 HUABIO Hangzhou) or GAPDH mouse mAb (HRP conjugate ZENBIO), followed by an additional hour at room temperature. After five washes with TBST, the membranes were treated with an HRP-conjugated secondary antibody for 2 h. After three more washes with TBST (BL608A, Biosharp, Guangzhou, China), the membranes were exposed to ECL solution (Millipore) for 1–3 min, and the protein bands were visualized using the Invitrogen imaging system (GoldBand protein marker, YEASEN, Shanghai).

Gene and transcript expression analysis

mRNA Sequencing: Tissue samples were collected from the top one centimeter of the mouse anus. The sequencing was carried out by Shanghai Bioscience Co. Rectal samples from each group of mice were preserved in liquid nitrogen. We randomly chose 4 samples from the irradiation-only group and 6 from the Liriodendron gavage group for RNA extraction and transcriptomic analysis. PolyA mRNA was isolated using Oligo(dT) magnetic beads, and the RNA was fragmented to around 300 bp. The first strand of cDNA was created with a 6-base random primer, followed by the synthesis of the second strand. The library was constructed through PCR amplification for fragment enrichment and size selection at 450 bp. The quality of the library was evaluated using an Agilent 2100 Bioanalyzer. Libraries with distinct index sequences were mixed in specific ratios, diluted to 2 nM, and denatured into single strands. Paired-end (PE) sequencing was conducted on the Illumina platform utilizing Next-Generation Sequencing (NGS). Gene expression was assessed based on clean read counts, with differential expression analyzed using DESeq2. GO enrichment analysis was performed with Goatools, and KEGG pathway analysis was executed using KOBA, applying Fisher's exact test for both analyses.

Real-time RT-PCR: Following established protocols, total RNA was isolated from the liver tissues of mice using TRIzol reagent (Invitrogen, California, USA). For the reverse transcription process, 1 μg of total RNA was utilized, and the SYBR Green Pro Taq HS Premix qPCR Kit (AG11701 Accurate Biology, Changsha, China) along with the Evo M-MIV reverse transcription premix kit (AG11728 Accurate Biology, China) were employed for the real-time qPCR analyses. The CFX Manager system (BioRad, California, USA) was used to measure mRNA expression levels.

GAPDH-F ATGATTCCACCCATGGCAAATTC

GAPDH-R GACTCCACGACGTACTCAGC

JUND-F TCTTGGGCTGCTCAAACTCG

JUND-R CCTTCGGGTAGAGGAACTGC

AP-1-F TGGGCACATCACCACTACAC

AP-1-R TCTGGCTATGCAGTTCAGCC

CXCL9-F CCGAGGCACGATCCACTAC

CXCL9-R AGGCAGGTTTGATCTCCGTT

Smad3-F TGAAGAAGCTCAAGAAGACGGG.

Smad3-R GAGGGAGCCCCTTCCGAT.

Cellular assays

Reactive oxygen species (ROS) present in intestinal tissues were analyzed using flow cytometry. Samples were taken from the first centimeter of the mouse anus, rinsed with chilled PBS, and then minced into smaller fragments before being homogenized. The resulting cell suspension was filtered through a 200-mesh sieve and centrifuged at 1200 rpm for 5 min. The supernatant was removed, and the pellet was washed twice with PBS, followed by centrifugation at 300 g for another 5 min to obtain the cell precipitate. A 10 μmol/L solution of DCFH-DA was prepared, with 1 mL added to each tube and incubated at 37 °C for 20 min. After a subsequent centrifugation at 1500 rpm for 5 min, the supernatant was discarded, and the cells were washed three times to eliminate any remaining DCFH-DA. Positive controls were established using a serum-free medium diluted at a 1:1000 ratio. Probes were added, with Rosup also diluted to 1:1000 for the positive control. The cells were incubated at 37 °C for 30 min, washed three times, and centrifuged at 1500 rpm for 5 min. The supernatant was discarded, and the cells were resuspended in PBS. Flow data were analyzed using CytExpert software, and the ROS levels in rectal tissues were evaluated with an independent sample T-test in GraphPad Prism after exporting the data.

Apoptosis in intestinal cells was assessed using the TUNEL staining technique. Initially, the tissue sections underwent 2–3 washes with PBST solution, followed by treatment with Triton X-100 and BSA solutions for 30 min each. After additional washes with PBST to remove excess sealing solution, a mixture of TdT enzyme reaction solution and TMR labeling solution was applied to the sections, which were then incubated for one hour under controlled temperature and light protection. Following this, the sections were washed again 2–3 times with PBST and treated with DAPI solution for 5 min while protected from light. After two more washes with PBST, an autofluorescence quencher was used for sealing. The sections underwent another series of washes with PBST and DAPI, followed by sealing with an anti-fluorescence quencher. The apoptosis in mouse intestinal tissues across different groups was examined under a light microscope, images were captured and archived, and the count of TUNEL-positive cells in each visual field was analyzed using Image J software.

Statistical analysis

The statistical analysis was conducted using SPSS version 23.0, where independent-sample T tests were applied to the data, which was presented as mean ± standard deviation (mean ± SD). To assess the normality of the data distribution, both two independent samples t-tests and the Kolmogorov–Smirnov test were employed. The survival curve data were evaluated through the Log-rank (Mantel-Cox) test, while the Prism 9 Program (Graph Pad, San Diego, CA, USA) facilitated the Two-way ANOVA for statistical evaluation. A P-value of less than 0.05 was deemed indicative of a statistically significant difference.