Collecting WPL active ingredients and targets

The active ingredients were obtained from the Traditional Chinese Medicine Integrated Database (TCMID, http://www.megabionet.org/tcmid/) [33] and Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, http://sm.nwsuaf.edu.cn/lsp/tcmsp.php) [34]. However, the active ingredients of Hericium erinaceus (Bull.) Pers and Gecko were not found in the TCMID or TCMSP. WPL-related targets were retrieved from TCMID, TCMSP, and the Search Tool for Interacting Chemicals (STITCH, http://stitch.embl.de) [35], then screened with a medium compound-target association score greater than 400 in the STITCH database. The target information was standardised using NCBI (https://www.ncbi.nlm.nih.gov/). Next, a drug-like analysis of WPL active components was carried out according to the quantitative estimate of drug-likeness (QED) proposed by Bickerton [36] with a QED ≥ 0.3. Subsequently, WPL active compounds and targets were screened using a binomial statistical model [37].

We searched for Hericium erinaceus (Bull.) Pers active components and targets in Herb Ingredients’ Targets (HIT, http://lifecenter.sgst.cn/hit/) and deleted some active constituents with no targets [38]. Additionally, calycosin 7-O-glucoside and notoginsenoside R1 targets, which have high drug activities, were retrieved from the PharmMapper (http://www.lilab-ecust.cn/pharmmapper) platform with a median Norm Fit ≥ 0.6008 and 0.4921, respectively, followed by further searches with a median Norm Fit ≥ 0.72915 and 0.586, respectively.

Collecting GPL targets

GPL targets were collected from the GeneCards (https://www.genecards.org) database with “gastric precancerous lesions” as the search words, then screened with relevance scores ≥ 5.

Protein–protein interaction (PPI) network construction

The PPI network was constructed using the STRING database (https://cn.string-db.org/) using common targets obtained from GPL cure targets and WPL-related targets using the Venny 2.1 tool. The PPI network was visualised using Cytoscape software 3.8.0.

Function enrichment analysis

Gene Ontology (GO) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed using the clusterProfiler package 3.15.4. A significant enrichment analysis was set at P ≤ 0.01.

Molecular docking analysis

Zinc (http://zinc15.docking.org) [39] and the Protein Data Bank (PDB, http://www.rcsb.org/pdb/) [40] were used to download the chemical structure of compounds and the three-dimensional structure of targets, respectively. Molecular docking analysis was performed using AutoDock Vina 1.1.2 [41] and plotted using PyMol 2.3.0 [42].

Reagents and antibodies

Methanol, acetonitrile, and HPLC-grade acetic acid were purchased from Merck (Darmstadt, Germany). The reference substances notoginsenoside R1, calycosin-7-O-beta-d-glucoside, astragaloside A, and adenine-9-β-d-ribofuranoside, with HPLC-verified purities > 98%, and vitamin B12 (VitB12) were obtained from Meilun Biotechnology Co., Ltd. (Dalian, China). MNNG was provided by the Tokyo Chemical Industry (TCI) (Tokyo, Japan). Primary antibodies against inducible nitric oxide synthase (iNOS, Abcam, Cambridge, MA, USA, ab178945), cyclooxygenase-2 (COX-2, Abcam, ab179800), NOX-2 (Abcam, ab129068), NOX-4 (Abcam, ab154244), NF-κB p65 (8242S, Cell Signalling Technology (CST), Beverly, MA, USA), phospho-NF-κB p65 (CST, 3033S), IκB-α (CST, 4814S), phospho-IκB-α (CST, 2859S), CD4+ (CST, 93518S), interferon-gamma (IFN-γ, CST, 8455S), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, CST, 5174S) () were used for western blotting or immunohistochemistry (IHC). The secondary horseradish peroxidase (HRP)-conjugated antibody (8114S) was from CST. All other reagents and chemicals used in this study were of analytical grade.

Preparing and phytochemically analysing WPL

WPL consists of nine herbs: Huang Qi (Guangdong Tiancheng Prepared Herbal Medicine Co., Ltd., China, 200501), Bai Zhu (Guangdong Zhixin traditional Chinese medicine Tabl Co., Ltd., China, 201101), Fu Ling (Guangdong Zhixin traditional Chinese medicine Tabl Co., Ltd., 201002), E Zhu (Guangdong medicinal materials co., China, E0520011), Tai Zi Shen (Guangdong medicinal materials co., T3520011), San Qi (KANGMEI, China, 201201201), Bai Hua She Cao (Guangdong medicinal materials co., B2020912), and Hou Gu Jun (Guangzhou Ziyunxuan Pharmaceutical Co., Ltd., China, 201001) Voucher specimens were deposited at the School of Pharmaceutical Sciences, Guangzhou University of Chinese Medicine (Guangzhou, China). All these herbs were extracted in water ten times the amount of the herbs and boiled twice for 2 h each. Next, San Qi and Hou Gu Jun powders were added and mixed. It was uniformly concentrated at 60 °C under reduced pressure, then vacuum-dried, and crushed to obtain 1 g of extract equivalent to 3.33 g of the crude drug. For in vivo management, WPL was dissolved in water (450 mg/mL).

WPL was dissolved in methanol (10 mg/mL), and a 5-μm filter or WPL solution was filtered by HPLC–MS/MS analysis to phytochemically examine the WPL components.

WPL was characterised using the negative and positive ionisation modes of TSQ Quantum Ultra-EMR HPLC–MS/MS (Thermo Scientific, USA). The operating parameters were as follows: an ion source applying ESI and an ion spray voltage of 3000 V in both positive and negative modes. For two modes, the low rate of sheath gas and the low rate of auxiliary gas were 30 mL/min and 10 mL/min, respectively, and the probe heater and capillary temperatures were set at 350 °C and 300 °C, respectively. The skimmer level was set to 1. Parallel reaction monitoring (PRM) scanning was performed to obtain MS data at a resolution of 70,000. Additionally, sample analyses were performed at 9, 17, and 18 normalised collision energies (NCE), and the MS detection resolution was 17,500. The TSQ Tune software Trace Finder (Thermo Scientific) was used to set up the instrument and process the quantitative data.

HPLC analysis was conducted on a Phenomenex C18 column (50 mm × 2.1 mm, 5 μm) with mobile phases A (acetonitrile) and B (0.1% formic acid) (v/v). The flow velocity was set to 1 mL/min, and the injection volume was 5 μL. A linear gradient (initial 0% A, 5% A 0–0.5 min; 5–95% A 0.5–1 min; 95% A 1.0–2.5 min; 95–5% A 2.5–3 min) was used for elution. The retention duration and HPLC–MS/MS data of WPL extracts and the reference compounds were compared to identify compounds, including notoginsenoside R1, calycosin-7-O-beta-d-glucoside, astragaloside A, and adenine-9-β-d-ribofuranoside.

Quantitative determination was performed by using six-point regression curves. Four reference compounds were used to prepare a stock solution diluted with methanol to a suitable concentration. HPLC–MS/MS was conducted to analyse the standard and WPL solutions. The peak area of each reference compound was calculated from the calibration curves. Compound contents in WPL were expressed as mg/g was and was calculated after the correlation between the peak area and the calibration curves of each analyte (Table 2).

Table 2 Quantitatively determined parameters of reference compounds in WPLAnimal experiments

Male BALB/c mice (2-month-old, 18–22 g) were purchased from the Laboratory Animal Centre of the Guangzhou University of Chinese Medicine. All mice were maintained at 23 ± 1 °C with sufficient water and food. The mice were randomly divided into five groups (n = 10 per group): vehicle control (H2O), MNNG (150 μg/mL), MNNG + WPL (3.75 g/kg/day and 15 g/kg/day), and MNNG + VitB12 (1 mg/kg). VitB12 was applied as the positive control for GPLs. The mice were administered an MNNG solution (150 μg/mL) with free drinking water for 6 weeks. After constructing the GPL mouse model by administering MNNG for 6 weeks, the respective groups orally received equal volumes of WPL, VitB12, or the vehicle for 8 weeks. The mice were sacrificed after the last administration, and stomach tissues were collected for further experiments. MNNG was dissolved in H2O to a concentration of 2 mg/mL and stored at 4 °C in the dark. This study was approved by the Institutional Animal Care and Use Committee of Guangzhou University of Traditional Chinese Medicine (NO.20201028003).

Histological examination

The stomach tissues were fixed with 4% PFA in 0.1 mol/L phosphate buffer overnight, embedded in paraffin, and cut into 4-μm slices. Standard techniques were used to stain the sections with haematoxylin and eosin (H&E) to observe histological changes.

Immunohistochemistry

Formalin-fixed, paraffin-embedded gastric mucosal tissues were selected to perform IHC. Sections were de-paraffinized and rehydrated with xylene and graded ethanol to water, followed by antigen retrieval and tissue blocking using 3% normal non-immune serum. Next, they were incubated with the primary antibodies IFN-γ (Abcam, ab9657, rabbit, polyclonal, 1/400) and CD4 (Abcam, ab203034, rabbit, polyclonal, 1/200) per the kit protocol (KGOS60, KeyGEN, Nanjing, China) at 4 °C overnight. After being washed with phosphate-buffered saline (PBS), the sections were incubated with HRP-conjugated secondary antibodies for 30 min at an ambient temperature. Tissue sections were stained with 3,3ʹdiaminobenzidine (DAB) and haematoxylin. The IHC results were observed under a microscope and analysed using Image-Pro Plus software (version 6.0; Media Cybernetics, Silver Spring, MD, USA) with the average combined optical density (IOD) to positive part (IOD/pixel) ratios.

Western blotting

After preparation with radioimmunoprecipitation assay (RIPA) lysis buffer, protein samples were quantitated using a bicinchoninic acid (BCA) protein assay. Equal amounts of proteins were separated by sodium dodecyl-sulphate polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene fluoride (PVDF) membranes which were blocked with 5% bovine serum albumin (BSA) solution, and incubated with primary antibodies at 4 °C overnight. The next day, the membrane was rinsed with Tris-buffered saline and Tween 20 (TBST), followed by incubation with HRP-conjugated secondary antibodies for 2 h at room temperature. Protein levels were detected using a chemiluminescence assay kit (Millipore, Massachusetts, USA). The density analysis of immunoblots was conducted using ImageJ software (NIH, USA).

Quantitative reverse transcription PCR (RT-qPCR)

Total RNA was isolated from gastric mucosal tissues using a TRIzol reagent. Per the manufacturer’s protocol, we used the One-Step PrimeScript RT-PCR Kit to synthesise complementary DNA (cDNA) from total RNA. Next, qPCR was performed using ChamQTM Universal SYBR qPCR Master Mix (the reagents above were obtained from Vazyme Biotech Co., Ltd., Nanjing, China) and analysed on an ABI 7500 sequence detection system. The experiments were performed in triplicate, and the primer sequences were as follows: tumour necrosis factor alpha (TNF-α), forward 5′-AGGGTCTGGGC CATAGAACT-3′ and reverse 5′-CCACCACGCTCTTCTGTCTAC-3′; interleukin (IL)-6 (IL-6), forward 5′-ATGATGAGAAACGAGCCAATTG-3′ and reverse 5′-GCTTTGGCTTCTTTCTTACGAG-3′.

Statistical analysis

All quantitative measures are described as the average ± standard deviation (SD). Two independent groups were compared using an unpaired t-test. Differences between more than two groups were compared using a one-way analysis of variance (ANOVA), followed by Dunnett’s post-hoc test. P < 0.05 was regarded as statistically significant. All data analyses were performed using GraphPad Prism 7.0 (GraphPad Software, San Diego, CA, USA). All experiments were conducted in triplicates.