EMS chemical ingredients and putative targets
EMS chemical ingredients were collected from The Traditional Chinese Medicine Systems Pharmacology Database 19 (TCMSP, http://lsp.nwu.edu.cn/tcmsp.php). Compounds meeting drug-likeness (DL) ≥ 0.18 and oral bioavailability (OB) ≥ 30% were selected as candidate active ingredients. The target genes of chemical components were collected from the HERB 20 (HERB, http://herb.ac.cn/) and SwissTargetPrediction 21 (SwissTargetPrediction, http://swisstargetprediction.ch/) databases. Target genes were screened based on having a probability > 0 in the SwissTargetPrediction database.
Collection of RA related targets
The GeneCards 22 website (GeneCards, https://www.genecards.org/) and DisGeNET 23 database (DisGeNET, https://www.disgenet.org/home/) were used to search the target genes related to RA. GeneCards is an online database that includes multiomics, clinical, and functional data. Using “rheumatoid arthritis” as the keyword, the disease targets of RA were collected. The DisGeNET database collects variants and genes related to human diseases. Using “rheumatoid arthritis” as the keyword of the “diseases” module, the disease target dataset of RA was obtained (CUI: C0003873). Finally, the repeated targets were removed to construct the target set of the disease.
NF-κB signaling pathway gene acquisition
NF-κB signaling pathway genes were determined from the Reactom 24 (Reactom, https://reactome.org/) and Kyoto Encyclopedia of Genes and Genomes 25 (KEGG, https://www.genome.jp/kegg/) databases. Reactom pathway (ID: R-HSA-1169091, R-HSA-5676590, R-HSA-5668541, R-HSA-2871837, R-HSA-933542, R-HSA-5607761, R-HSA-5676594, R-HSA-933543, R-HSA-193639, R-HSA-209560) genes and KEGG pathway (Pathway: map04064) genes were downloaded as NF-κB signaling pathway genes.
RA and intersection target gene enrichment analysis
DAVID 26, 27 (DAVID, https://david.ncifcrf.gov/home.) is a biological information database and a free online analysis tool. It can provide comprehensive biological function annotation information for large-scale gene or protein lists and help users extract biological information from this database. Gene Ontology (GO) functional enrichment analysis, including biological process (BP), cellular composition (CC) and molecular function (MF), describes important information about the target genes. Pathway enrichment analysis indicates significant signaling pathways by classifying the known genome annotation information. The R package VennDiagram was used to create a Venn diagram for the intersection of EMS-RA-NF-κB signaling pathway target genes. The RA and intersection target genes were separately imported into DAVID (Version 6.8) for GO and pathway enrichment analysis. The identifier was set to “OFFICIAL_GENE_SYMBOL”, the species was set to “Homo sapiens”, and results with P < 0.05 were selected as the restriction. The ggplot2 R package was used to visualize the results.
Protein-protein interaction network and hub genes
The STRING 28 database (STRING, https://string-db.org/) contains a large number of known and predicted protein-protein interaction (PPI) relationships. The STRING database and Cytoscape (Version 3.6.1) were used to analyze the hub genes and build the network. The intersection target genes were input into the STRING database, the species was set to “Homo sapiens”, and a “confidence score” >0.4 was set as the restriction. The obtained PPI network data were imported into Cytoscape (Version 3.6.1), and the hub genes of intersection target genes were obtained using the plug-in cytoHubba 29, 30 with the degree algorithm. Finally, the intersection of the top 3 hub genes of RA and intersection target genes were taken as the key targets, and a PPI network diagram was constructed for visualization by Cytoscape.
Network construction
Cytoscape is software that allows for network visualization and analysis. Its core function is to provide a basic functional layout and query network and to form a visual network based on the combination of basic data. Compounds and intersection target genes were transferred into Cytoscape to create a network diagram. Within these graphical networks, the active ingredients and target genes are presented as nodes, whereas the compound-target interactions are expressed as edges.
Molecular docking
The 3D structures of these target genes, NFKBIA and RELA, were downloaded from the Protein Data Bank 31 (PDB https://www.rcsb.org/) database. Subsequently, berberine wogonin and fumarine were obtained in mol2 file format using the TSMSP database and saved in 3D structure format using ChemOffice. The active ingredient and target protein formats were converted into pdbqt format with PyMOL 2.4.1 to obtain visualization images of molecular docking.
Preparation of the EMS
Er Miao San consists of the following two herbs: Atractylodis Rhizoma (Cangzhu, specimen number 1902120322), the rhizome of Atractylodes lancea (Thunb.) DC. (Compositae), and Phellodendri Cortex (Huangbai, specimen number 1901200062), the bark of Phellodendron chinensis Schneid. (Rutaceae). They were acquired from Anhui Puren Herbal Pieces Co., Ltd (Bozhou, Anhui Province, China) in January 2019, and identified by Dr. Liu SJ (School of Pharmacy, Anhui University of Chinese Medicine). One specimen of each (ID: EMS-19-01) was preserved in the Herbarium of Pharmacy, School of Pharmacy, Anhui University of Traditional Chinese Medicine (Hefei, China). The extraction method can be found in our group's previous research 15.
Induction of the AA model and treatment
All animal experiments were performed at the School of Pharmacy, Anhui University of Traditional Chinese Medicine (Hefei, China). The experimental protocol was approved by the Experimental Animal Ethics Committee and adhered to the guidelines for animal care and use (AHUCM-rats-2020016). Male Sprague-Dawley (SD) rats weighing 160–180 g were acquired from the Animal Department of Anhui Medical University, China, and housed in standard laboratory conditions (under a controlled temperature of 22–26°C with a 12 h light and 12 h dark period). Before the experiment, the rats were adaptively fed for 7 days, where they were allowed free access to food and water. The AA model was established following previously reported methods 32. To induce AA, complete Freund’s adjuvant (CFA, 10 mg/mL) was prepared by suspending heat-killed Mycobacterium butyricum in sterile liquid paraffin. The AA model was induced in rats by subcutaneous injection of 100 µL of CFA into the left hind metatarsal footpad. Control rats were injected with the same amount of physiological saline. After the onset of AA (on approximately the 17th day), the limbs, ears and tail of rats showed varying degrees of redness and swelling, nodules and limitation of movement, which were regarded as the result of the modeling, so the rat model of AA was obtained.
Some of the successful AA model rats were used in pharmacodynamic experiments. On the 15th day after immunization, the rats were randomly divided into normal group, AA model group, EMS (3 g/kg, 1.5 g/kg, 0.75 g/kg) group and MTX (0.5 mg/kg) group. EMS was administered by gavage once a day for 14 days and MTX was administered by gavage once every three days for five times. Meanwhile, rats in the normal and AA model groups were gavaged with the same volume of sodium carboxymethylcellulose solution (10 mL/kg).
Isolation and culture of FLSs
Some of the successful AA model rats were anesthetized after 14 days of normal feeding and soaked in the newly prepared bromogeramine solution, and synovial tissue was extracted on a super clean table. The foreign bodies around the tissue blocks were cut off and washed twice with phosphate buffered saline (PBS) containing 1% penicillin-streptomycin-gentamicin solution (Cat. No. C0223-100 mL, Beyotime, China) in a petri dish and then transferred to Dulbecco’s modified Eagle’s medium (DMEM, Biological Industries, USA) with high glucose supplemented with 20% fetal bovine serum (FBS, Servicebio, China), which was cut into tissue fragments of approximately 1 mm2 using ophthalmic scissors. Pasteur pipettes were used to absorb and inoculate the wall of the culture bottles wetted with 20% FBS, arranged neatly and evenly, supplemented with 3 mL culture medium, covered with the bottle caps and placed in a constant temperature cell incubator of 5% CO2 and 37 ℃ (Thermo Scientific, USA). After 4 hours, the culture bottles were placed, so that the culture medium did not pass through the tissue mass and continued to culture in the carbon dioxide incubator until the primary shuttle-shaped synovial cells spread around the tissue mass. Based on cell growth, the culture medium was changed over time, and the tissue mass was discarded. After the cells grew to 80%-90% confluence in the culture flasks, they were digested with trypsin-EDTA solution (Cat. No. C0201-100 mL, Beyotime, China) and repeatedly blown out with Pasteur pipettes to obtain a single cell suspension, which was then transferred to new culture flasks.
FLSs were cultured in DMEM supplemented with 20% FBS and 1% penicillin/streptomycin. All cells were cultured in a carbon dioxide incubator as described previously. FLSs at passages 3–8 were used for experiments. After three generations of FLSs culture, the shape of FLSs were triangular or fusiform, and immunofluorescence staining showed vimentin+,as shown in Supplementary Fig. S1.
Medicated serum preparation and cell intervention
Normal SD rats were randomly divided into a blank control group and an EMS group (3 g/kg) (calculated using the crude drug), and they were administered intragastrically for 7 days (once a day). One hour after the final administration, blood was taken from the abdominal aorta, centrifuged to prepare medicated serum, inactivated at 56 ℃ for 1 h, filtered through a 0.22 µm microporous filter, and stored at -20 ℃ for later use.
FLSs were treated with 10 ng/mL TNF-α (Cat. No. 400 − 14, PeproTech, USA) to establish an inflammatory model of RA. FLSs were treated with different concentrations of EMS (5%, 10%, 20%). The experiments were divided into the following groups: Control, TNF-α, TNF-α + 5%EMS, TNF-α + 10%EMS and TNF-α + 20%EMS.
Ankle x-ray imaging and histopathology
On the 30th day after immunization, the left posterior ankle joints of rats were collected and placed in a small animal X-ray irradiator. Also, we set the image resolution to 15 LP/min and photography conditions to 22.0 KV, 4 mA, 0.1 mGy.
Then, the excess tissues in the ankle joints were removed, fixed with 4% paraformaldehyde solution, decalcified with EDTA decalcification solution, embedded in paraffin, sliced and stained with hematoxylin-eosin (HE). The pathological changes of ankle joint of rats in each group were observed under microscope. The inflammation, pannus formation, bone erosion, cell infiltration and synovial hyperplasia were classified into 4 grades according to the degree of severity: grade 0, no significant changes; grade 1, mild; grade 2, moderate; and grade 3, severe.
FLS proliferation assays
FLS proliferation was measured using a Cell Counting Kit-8 (CCK-8, Cat. No. BS350B, Biosharp, China) assay. FLSs were seeded into 96-well plates at a density of 5×103 cells per well. Based on the previous cell grouping treatment, after incubating in a cell incubator overnight, CCK-8 (10 µL) was added to each well and the plates were incubated for 2 h before termination of the reactions. Optical density (OD) was detected at 450 nm using a Multiskan Spectrum (Thermo Scientific, USA).
Enzyme-linked immunosorbent assay (ELISA)
The cell supernatant was collected after culturing for 48 h. An ELISA double antibody sandwich method was used, and the specific operation steps were carried out according to IL-1β (Cat. No. EK301B/3–96, Multisciences Co., Ltd.) and IL-6 (Cat. No. EK306/3–96, Multisciences Co., Ltd.). The OD value of each well read by the enzyme labeling instrument was taken as the ordinate, the concentration of the standard was taken as the abscissa, and the curve was drawn. The corresponding concentration range was found according to the OD value of the sample.
Western blot
Cells were harvested in RIPA lysis buffer containing phenylmethanesulfonyl fluoride (PMSF, Cat. No. ST506, Beyotime, China) and phosphatase inhibitors (Cat. No. P1081, Beyotime, China) in a cold environment and then the lysed cells and lysis buffer were collected in a tube, frozen and thawed repeatedly at -20 ℃ three times and centrifuged at 14000 rpm for 15 min to remove cell fragments. A BCA Protein Assay Kit was used to quantify the protein concentration. Then, equal total amounts of protein were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene fluoride (PVDF) membrane. After blocking with 5% nonfat milk, the membranes were incubated with IκBα (Cat. No. 4814, 1:1000, Cell Signaling Technology, USA), p-IκBα (Cat. No. 2859, 1:1000, Cell Signaling Technology, USA), NF-κB p65 (Cat. No. 8242, 1:1000, Cell Signaling Technology, USA) and p-NF-κB p65 (Cat. No. 3033, 1:1000, Cell Signaling Technology, USA) antibodies overnight at 4°C and then washed with Tris buffered saline-Tween 20 (TBST) and secondary antibodies for 2 h at room temperature. The protein bands were developed using the enhanced chemiluminescence reagent and scanned by the Amersham Imager (AI) 600 images imaging system (GE Healthcare Bio-Sciences AB, USA). The intensity of each band was determined using ImageJ software.
Immunofluorescence
Cells cultured on 24-well plates were washed with PBS, fixed in 4% paraformaldehyde (Cat. No. BL539A, Biosharp, China) for 30 min. Then, the cells were washed with PBS three times and blocked with Immunol staining blocking buffer (Cat. No. P0102, Beyotime, China) for 1 h at room temperature, followed by incubation with primary antibodies overnight at 4°C. After washing three times for 5 min, secondary antibodies (Cat. No. SA00013-4, Proteintech, USA) was added and incubated at room temperature for 1 h in the dark. After washing twice for 5 min, the cells were counterstained with DAPI staining solution (Cat. No. C1005, Beyotime, China) for 5 min in the dark and examined under an immunofluorescence microscope (Leica, Germany).
Statistical analysis
Data are presented as the mean ± standard deviation (SD). The differences between groups were analyzed by one-way analysis of variance (ANOVA) using Prism 8 software (GraphPad Software, USA). A p value < 0.05 indicated statistical significance.