Abstract Kaempferol is an active compound found in traditional Chinese medicine epimedium soup, which exhibits potent anti-inflammatory and antioxidant properties. Nevertheless, the mechanism of action in rheumatoid arthritis remains unclear. This study constructed targets protein interaction networks by utilizing the String platform. The analysis of GO function and kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment was performed on key target genes. Target gene validation was conducted through microarray analysis. Molecular docking was employed to evaluate the relationship between kaempferol and various key targets. In vitro experiments were conducted to elucidate kaempferol’s mechanism of action on rheumatoid arthritis. Topological analysis of the protein protein interaction (PPI) network identified 10 core targets. Mitogen activated protein kinase 8 (MAPK8), peroxisome proliferator-activated receptor gamma (PPARG), and nuclear factor kappa-B (NF-kB) were all differentially expressed in the microarray dataset and all belonged to the target genes of kaempferol. Furthermore, kaempferol exhibited the highest binding affinity for MAPK8. In vitro cellular experiments demonstrated that kaempferol suppressed autophagy, and ameliorated abnormal proliferation and inflammation in rheumatoid arthritis fibroblast-like synoviocytes (RA-FLS) cells by activating the MAPK8/NOD-like receptor protein 3(NLRP3) signaling pathway. Keywords: Rheumatoid arthritis, Fuzi Decoction, Kaempferol, MAPK8 Subject terms: Molecular biology, Endocrinology, Pathogenesis, Rheumatology Introduction Rheumatoid arthritis (RA) is an autoimmune disease characterized by chronic, symmetrical inflammation of the joints, commonly affecting those in the hands, wrists, knees, and feet^[34]1,[35]2. Additionally, RA can manifest systemic symptoms including fatigue, fever, and weight loss. Untreated, RA may lead to joint destruction and disability^[36]3. Fuzi Decoction, an ancient Chinese herbal remedy, is extensively employed in the management of RA and other arthritic conditions^[37]4. Its primary constituents comprise Fuzi, Baizhu, Renshen, Baishao, Fuling, and various other herbs. This soup is known for its multifaceted effects, including dispelling wind, dispersing cold, promoting blood circulation, removing blood stasis, enhancing immunity, mitigating inflammation, and improving digestion^[38]5. In our investigation, we discovered that kaempferol, a key component of Fuzi Decoction, plays a pivotal role in the context of RA. Kaempferol, a natural flavonoid, is among the active constituents of traditional Chinese medicine Fuzi Decoction. It is extensively found throughout the plant kingdom, particularly in various vegetables, fruits, and herbs^[39]6. Kaempferol is recognized for its anti-inflammatory properties, which aid in mitigating inflammatory responses and alleviating inflammatory conditions. Its mechanism involves modulation of inflammation-associated signaling pathways, including inhibition of inflammatory mediator release and cytokine expression^[40]7–[41]9. Additionally, kaempferol functions as an antioxidant can counteracte free radicals generated within the body and mitigate oxidative stress. This action aids in minimizing oxidative damage to cells and tissues^[42]10,[43]11. MAPK8, also known as Mitogen-Activated Protein Kinase 8 or c-Jun N-terminal kinase 1(JNK1), is a protein kinase classified within the mitogen-activated protein kinase (MAPK) family. MAPK8 is involved in cell signaling pathways, with a particularly significant role in cellular stress response, apoptosis, inflammation, and various other biological processes^[44]12,[45]13. MAPK8 is frequently activated in response to diverse exogenous and endogenous stimuli, including cytokines, growth factors, and environmental stresses. Upon activation, MAPK8 phosphorylates downstream target proteins, thereby regulating cellular physiological and pathological processes^[46]14. Given its pivotal role in cell signaling pathways, MAPK8 has been extensively investigated and is regarded as a potential therapeutic target for conditions such as cancer, inflammation, and other diseases^[47]15,[48]16. This study integrated network pharmacology, molecular docking, and in vitro cellular experiments to explore the potential key targets and signaling pathways of kaempferol, a bioactive compound in the traditional Chinese medicine Compound Fuzi Decoction, in the context of RA. Materials and methods Prediction of common targets in Chinese herbal compound fuzi decoction and RA Targets related to Fuzi Decoction Chinese herbal medicine were retrieved from the TCMSP database ([49]http://tcmspw.com/tcmsp.php), while those linked to RA were sourced from GeneCards ([50]https://www.genecards.org/), OMIM ([51]https://omim.org/), PharmGkb ([52]https://www.pharmgkb.org/), TTD ([53]http://db.idrblab.net/ttd/), and DrugBank ([54]https://www.drugbank.ca/). The two sets of target intersection yielded potential common targets associated with kaempferol and RA. Protein–protein interaction network construction and key target screening The predicted potential common targets were imported into the STRING database ([55]https://cn.string-db.org/) for protein–protein interaction (PPI) analysis. Then, the analysis results were imported into Cytoscape 3.9.0 software. Network diagrams of interactions between common protein targets were obtained. According to the node-based topological centrality measure, the median of Degree, Betweenness Centrality, and Closeness Centrality were used to screen and get the key targets of the active ingredients of epiphyllum soup, such as kaempferol, against rheumatoid arthritis. Degree denotes the number of direct neighbors of a node; the more nodes are connected to a node, the more nodes the greater the influence of the node. Intermediary Centrality is the number of shortest paths through a node; the smaller the value, the less important the node is. Proximity Centrality is the reciprocal of the average shortest paths between a node and other nodes; the smaller the value, the slower the signal transmission between nodes. Key target GO function and KEGG pathway enrichment analysis The key targets were subjected to GO function and KEGG pathway enrichment analyses using R. A p-value threshold of p < 0.05 and a minimum enrichment value of 1.5 were applied for both GO function and KEGG pathway analyses. Subsequently, the enrichment results for GO functions and KEGG pathways were visualized by histograms and bubble plots. “Drug-target-disease” network construction Utilizing Cytoscape 3.9.0 software, drug-target-disease network diagrams were assembled. These diagrams were generated by importing both drug-active ingredients and disease-common targets into the Cytoscape 3.9.0 platform, facilitating the visualization of the network interactions. Microarray data analysis Data sets [56]GSE55235, [57]GSE55457, and [58]GSE77298 were retrieved from the GEO database ([59]https://www.ncbi.nlm.nih.gov/geo/) and utilized to detect Differentially Expressed Genes (DEGs) employing the R package “Limma”. DEGs were defined as logFC > 1 or logFC < -1. Specifically, DEGs with logFC < −1 and p-value < 0.05 were considered statistically significant. Molecular docking of kaempferol to key targets Leigh obtained the 3D structure of kaempferol (PubChem CID: 5280863) from the PubChem database ([60]https://pubchem.ncbi.nlm.nih.gov/) to be used as a ligand file for subsequent molecular docking. The drug kaempferol was purchased from MCE (catalog number: HY-14590). Target protein 3D structures were sourced from the RCSB database ([61]https://www.pdbus.org/) and utilized as receptor files. AutoDocktools (version 1.5.6) facilitated the extraction of ligand small molecules from receptors, side chain repair, and hydrogen atom addition. Subsequently, AutoDock Vina (version 1.1.2) was employed for molecular docking. A protein binding affinity of less than -5 kcal/mol indicated some binding activity with the compound. Docking results of compounds and proteins exhibiting optimal conformations were analyzed and visualized by PyMol (version 2.5.7) software. Cell culture Normal human fibroblast-like synoviocytes (HFLS) and RA-FLS were procured from the American Type Culture Collection (ATCC, Rockville, MD, USA). The cells were cultured in dulbecco’s modified Eagle’s medium (DMEM, Life Technologies/Gibco, Grand Island, NY, USA) supplemented with 15% fetal bovine serum (Gibco) and 1% antibiotics (comprising 100 U/mL penicillin and 100ug/mL streptomycin, Sigma-Gibco, Grand Island, NY, USA). Subsequently, the cells were incubated in a 37 °C incubator with 5% CO2. Cell transfection and processing Overexpression MAPK8 plasmid (AD-MAPK8) was constructed using pcna3.1 plasmid as a vector. RA-FLS cells were cultured using six-well plates, and when the cells reached 70%-90% confluence, AD-MAPK8 was transfected into RA-FLS cells using Lipo2000 (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Cells were replaced with fresh double-antibody serum-free medium about one hour before transfection, using 3ug plasmid and 6ul Lipo2000 per well, diluted and mixed separately using DMEM medium, and left to stand at room temperature for five minutes, the Lipo2000 dilution solution was added to the plasmid dilution solution and left to stand for 30 min, the mixture was evenly dripped into the cell medium, and then the cells were replaced with normal serum-containing freshly formulated medium after 6 h. After 6 h, the cells were replaced with normal serum-containing freshly prepared medium and continued to be cultured for 24 h, and then the cells were used for experiments. Cell scratch experiment Initially, use a marker to draw an even line across the back of the 6-well plate. Subsequently, add approximately 5 × 10^5 cells to the wells. On the following day, utilize the marker’s tip perpendicularly to the cell layer’s plane, to create a scratch along the previously drawn line on back of the plate. Following the completion of scratching, incubate the cells in a 37°℃, 5% CO2 incubator. At 0 and 24 h, remove the cells to respectively observe, measure, and photograph the scratch width under a microscope. Using Image J, randomly scratch six to eight horizontal lines and calculate the mean of intercellular distance. Cell counting kit-8 (CCK-8) cell viability assay RA-FLS cells were seeded into 96-well plates at a density of 3 × 10^3 cells per well. After 6–12 h, the cells were treated with kaempferol for 24 h. Subsequently, 10 μL of CCK-8 solution (Biosharp, Beijing, China) was added to each well, and the cells were further incubated for 1–4 h. The absorbance of the cells at 450 nm was then measured by an enzyme marker to assess the impact of kaempferol on cell viability. Detection of cellular reactive oxygen species (ROS) levels The dichlorodihydrofluorescein diacetate (DCFH-DA) green fluorescent probe (Beyotime Biotechnology, Shanghai, China) was employed for measuring ROS levels. DCFH-DA was diluted 1000 times with DMEM to create a working solution. Subsequently, 1 ~ 2 ml of this solution was applied to completely cover the cells on the culture plate, then which was incubated at 37℃ in darkness for 20 min. Afterward, the cells were washed three times with phosphate buffer saline (PBS) buffer. Subsequently, the cells were imaged by an inverted fluorescence microscope, and the alterations in ROS levels of RA-FLS cells were quantified by ImageJ software. Detection of mitochondrial membrane potential levels The mitochondrial membrane potential (JC-1) kit from Beyotime Biotechnology in Shanghai, China, was employed for measuring mitochondrial membrane potential. Inoculate the cultured cells and incubate them overnight at 37 ℃. According to the JC-1 kit instructions, JC-1 was diluted 1000 times to prepare a working solution. 1 ~ 2 ml working solution was applied to completely cover the cells on the culture plate in a 15-min incubation at 37 ℃. Subsequently, the cells were washed three times with PBS buffer. Images were captured by an inverted fluorescence microscope, and the membrane potential of RA-FLS cells was analyzed by ImageJ software. Western blot Total proteins were extracted from cells using a prepared sodium dodecyl sulfate (SDS) lysis buffer for protein blotting analysis. Proteins were separated using SDS–polyacrylamide gels, and according to the concentration of the cell samples, 10 ug of protein per well was added to each lane for electrophoresis at a uniform rate, initially using a voltage of 80 V. Waiting for the protein samples to enter the separator gel and using a voltage of 120 V for rapid electrophoresis, the proteins were rapidly electrophoresed to the bottom layer. After electrophoresis, the SDS-PAGE gel was removed, and a polyvinylidene difluoride membrane (PVDF, Merck Millipore, Germany) was activated and placed on the top layer of the gel, and the blot was transferred to the PVDF membrane by electrotransferring the blot at 200 mA for 2 h. Subsequently, the membranes were blocked with 5% skimmed milk for 2 h. Antibodies including glyceraldehyde-3-phosphate dehydrogenase(GAPDH) (Proteintech, #60,004–1-IG), phosphor-mitogen-activated protein kinase (P-MAPK) (Wanleibio, #WL01813), mitogen-activated protein kinase (MAPK) (Wanleibio, #WL05246), peroxisome proliferator-activated receptor gamma (PPARG) (Wanleibio, #WL01800), phosphor-nuclear factor kappa-light-chain-enhancer of activated B cells (P-NFKB) (Wanleibio, #WL02169), NFKB (Bioss, # bsm-33117 M), NLRP3 (Bioss, #bs-41293R), IL-1β (Wanleibio, #WLH3903), uperoxide dismutase (SOD2) (Boster, #BA4566), microtubule-associated protein 1 light chain 3 beta (LC3B) (Cell signaling technology, #12,741), P62 (Cell signaling technology, #16177S), and proliferating cell nuclear antigen (PCNA) (Proteintech, #10,205–2-AP) used. The Polyvinylidene fluoride (PVDF) membranes were incubated within the mentioned antibodies overnight at 4 °C. Subsequently, a horseradish peroxidase (HRP)-labeled secondary antibody (ZSGB-BIO, Beijing, China) was applied for 1 h and detected using the electro-chemi-luminescence (ECL) luminescence system. Immunofluorescence RA-FLS cells were seeded in 48-well plates at a density of 1 × 10^4 cells per well. After cell treatment, fixation was carried out using 4% paraformaldehyde for 30 min at room temperature, followed by PBS washing. Next, cells were exposed to 0.1% Triton-X for 20 min, followed by another round of washing. The cells were blocked with 5% bovine serum albumin (BSA, Solarbio, Beijing, China) for 60 min at room temperature. Subsequently, they were incubated overnight at 4 °C with antibodies (diluted 1:200) targeting LC3B, P62, etc. Subsequently, cells underwent another washing step and were incubated in the dark for 1 h with secondary antibodies: either goat anti-mouse antibody (Proteintech) or goat anti-rabbit antibody (Proteintech). Following that, cells were incubated in the dark with 4′,6-DiAmidino-2-PhenylIndole (DAPI) (Beyotime Biotechnology, Shanghai, China) for 10 min. Finally, the samples were sealed with an anti-fluorescence quencher (Beyotime Biotechnology, Shanghai, China) and observed with an orthogonal fluorescence microscope. qRT-PCR Total RNA was extracted from cells utilizing TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA). TRIzol was added to cell samples and left to stand for 5 min at room temperature. Subsequently, 200 μL of chloroform was added to each sample, vortexed, and after a 5-min stand, centrifuged at 12,000 rpm and 4°℃ for 15 min. The samples were separated into 3 layers, and 400 μl isopropanol was added. After standing for 10 min, they were centrifuged at 4000 rpm for 10 min at 12°℃, and the supernatant was collected. The pellet was washed with 70% ethanol, centrifuged, dried, and finally reconstituted in diethylpyrocarbonate (DEPC) water. Ultimately, the concentration and purity of the extracted RNA were assessed by ultraviolet (UV) spectrophotometry. Total RNA was reverse transcribed according to the cDNA Reverse Transcription Kit instructions. Quantitative PCR was conducted by synergetic binding reagent (SYBR) Green (Beyotime Biotechnology, Shanghai, China) on a Roche LightCycler 480 real-time fluorescence quantitative polymerase chain reaction (PCR) system. Upon completion of the reaction, the cycle threshold (Ct) was determined, and the relative ribonucleic acid (RNA) level for each sample was calculated by the 2^−ΔΔCt method, normalized to GAPDH levels. Primer sequences are shown in Table [62]1. Table 1. Gene names and primer sequences. Primer sequence IL-1β Forward CAGAAGTACCTGAGCTCGCC Reverse GAAGCCCTTGCTGTAGTGGT MAPK8 Forward CAGCCCTCTCCTTTAGGTGC Reverse TAACCGACTCCCCATCCCTC PPARG Forward TGCATTCTGCTTAATTCCCTTTCC Reverse CTGTGTCAACCATGGTCATTTCGTT NFKB Forward ACTCGCCACCCGGCTTC Reverse TCACTAGAGGCACCAGGTAGT MMP2 Forward TGCAGTGGGGGCTTAAGAAG Reverse ATCACTAGGCCAGCTGGTTG MMP9 Forward GGACAAGCTCTTCGGCTTCT Reverse TCGCTGGTACAGGTCGAGTA N-Cadherin Forward ATGGGAAATGGAAACTTGATGGC Reverse CAGTTGCTAAACTTCACTGAAAGG Vimentin Forward GGACCAGCTAACCAACGACA Reverse AAGGTCAAGACGTGCCAGAG GAPDH Forward CAAGGTCATCCATGACAACTTTG Reverse GTCCACCACCCTGTTGCTGTAG [63]Open in a new tab Statistical analysis Continuous variables are presented as mean ± standard deviation (SD). The overall significance of the results was assessed with one-way ANOVA with GraphPad Prism 5 (GraphPad Software, La Jolla, CA, USA). For comparing differences between groups, a P-value less than 0.05 was considered statistically significant. Results Network pharmacology target gene prediction RA-associated genes were retrieved from GeneCards, On-line Mendelian Inheritance in Man (OMIM), PharmGkb, Therapeutic Target Database (TTD), and DrugBank databases, and a total of 4142 disease-related genes were obtained (Fig. [64]1A). Targets associated with Fuzi Decoction were extracted from the Traditional Chinese Medicine SystemsPharmacology (TCMSP) database, which yielded a total of 57 key genes. By intersecting the two sets of targets, 35 potential common targets of Fuzi Decoction and RA were identified (Fig. [65]1B). The drug-target-disease network was built by Cytoscape, resulting in the identification of 29 active ingredients and 35 key targets (Fig. [66]1C). Fig. 1. [67]Fig. 1 [68]Open in a new tab Web pharmacology target gene prediction. (A) GeneCards, OMIM, PharmGkb, TTD, and DrugBank databases screened for RA disease genes. (B) TCMSP database screened the target genes related to the Chinese herbal medicine of Fuzi Decoction, and the disease target genes and drug target genes were taken to be intersected. (C) Cytoscape constructed a “drug-target-disease” network, Red represents small molecule compounds and yellow represents relevant target genes. Protein–protein interaction network analysis Protein interactions among the 35 target genes were analyzed through Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) and visualized by Cytoscape 3.9.0 software. The data comprised degree centrality, proximity centrality, and inter-degree centrality, derived from the network topology centrality^[69]17. Based on previous studies we used the screening conditions of: degree value ≥ 10, inter-degree centrality ≥ 0.03, and proximity centrality ≥ 0.4. Ten targets comformed to the screening criteria and signified their importance in the network. Table [70]2 displays the gene names, degree values, betweenness centrality (BC), and closeness centrality (CC) of the key targets. Figure [71]2 illustrates the interaction network diagram of the key targets. Table 2. Gene names, degree value, BC and CC of key targets. Degree value BC CC IL6 27.0 126.00841 0.8292683 NFKBIA 19.0 26.461227 0.68 ICAM1 18.0 81.29322 0.6666667 BCL2 23.0 46.381218 0.75555557 NR3C1 15.0 11.656864 0.6296296 CASP3 23.0 45.529327 0.75555557 ESR1 24.0 102.657555 0.77272725 PPARG 25.0 90.979935 0.7906977 MAPK8 16.0 74.8215 0.6415094 NFKB 21.0 47.482487 0.7234042 [72]Open in a new tab Fig. 2. [73]Fig. 2 [74]Open in a new tab Protein–protein interaction network analysis. STRING performed protein interaction analysis on 35 target genes, visualized by Cytoscape 3.9.0 software, to screen core target genes. GO function and KEGG pathway enrichment analysis Functional annotation and enrichment analysis were performed on the 10 core genes, resulting in the enrichment of 78 biological processes, 6 cellular components, and 26 molecular functions. These processes were primarily related to (Fig. [75]3A,B,C). KEGG pathway enrichment analysis revealed enrichment in 96 signaling pathways, predominantly associated with inflammation, oxidation, cell autophagy, cell cycle, and other pathways. The TOP15 enriched KEGG pathway annotations and their potential targets are displayed in Table [76]3, and a bubble diagram illustrating this is presented in Fig. [77]3D^[78]18–[79]20. Fig. 3. [80]Fig. 3 [81]Open in a new tab GO function and KEGG pathway enrichment analysis. (A) Functional enrichment analysis of biological processes of 10 core genes. (B) Functional enrichment analysis of cellular components of 10 core genes. (C) Functional enrichment analysis of 10 core gene molecules. (D) KEGG pathway enrichment analysis. Table 3. Annotation of KEGG pathways with TOP15 enrichment degree and the involved potential targets. KEGG pathways p value Target count Targets TNF signaling pathway 9.28E–10 6 CASP3/IL6/NFKB/NFKBIA/ICAM1/MAPK8 IL-17 signaling pathway 3.30E–08 5 CASP3/IL6/NFKB/NFKBIA/MAPK8 Apoptosis 2.12E–07 5 BCL2/CASP3/NFKB/NFKBIA/MAPK8 NOD-like receptor signaling pathway 1.01E-06 5 BCL2/IL6/NFKB/NFKBIA/MAPK8 Cytosolic DNA-sensing pathway 1.06E–06 4 CASP3/IL6/NFKB/NFKBIA NF-kappa B signaling pathway 3.93E–06 4 BCL2/NFKB/NFKBIA/ICAM1 Toll-like receptor signaling pathway 4.57E–06 4 IL6/NFKB/NFKBIA/MAPK8 HIF-1 signaling pathway 0.000219466 3 BCL2/IL6/NFKB cAMP signaling pathway 0.001823869 3 NFKB/NFKBIA/MAPK8 p53 signaling pathway 0.003112277 2 BCL2/CASP3 MAPK signaling pathway 0.004181108 3 CASP3/NFKB/MAPK8 Rheumatoid arthritis 0.004871884 2 IL6/ICAM1 Mitophagy-animal 0.00594539 2 NFKB/MAPK8 PI3K-Akt signaling pathway 0.006855121 3 BCL2/IL6/NFKB FoxO signaling pathway 0.009472045 2 IL6/MAPK8 Autophagy-animal 0.01473775 2 BCL2/MAPK8 JAK-STAT signaling pathway 0.014908261 2 BCL2/IL6 Focal adhesion 0.021814716 2 BCL2/MAPK8 Ras signaling pathway 0.028908245 2 NFKB/MAPK8 MicroRNAs in cancer 0.047696931 2 BCL2/CASP3 [82]Open in a new tab Microarray data analysis Differentially expressed genes (DEG) were identified by the R package “Limma”. Among the 10 core genes, only MAPK8, PPARG, and NFKB showed differential expression across all three datasets. In the [83]GSE55235 dataset, 713 genes were up-regulated, and 237 genes were down-regulated (Fig. [84]4A). In the [85]GSE55457 dataset, 784 genes were up-regulated, and 33 genes were down-regulated (Fig. [86]4B). In the [87]GSE77298 dataset, 626 genes were up-regulated, and 141 genes were down-regulated (Fig. [88]4C). MAPK8, PPARG, and NFKB exhibited up-regulation consistently across all three datasets (Table [89]4). These three genes were further processed to detect protein and mRNA expression in HFLS versus RA-FLS, respectively, and MAPK8, PPARG, and NFKB appeared to be significantly increased in RA-FLS compared to HFLS (Fig. [90]4E,F). The molecular structure of kaempferol is shown below (Fig. [91]4D). Fig. 4. [92]Fig. 4 [93]Open in a new tab Microarray data analysis. (A) [94]GSE55235 variance analysis. (B) [95]GSE55457 differential analysis. (C) [96]GSE77298 differential analysis (D) Chemical structure of kaempferol (E) Western blot detection of protein expression of MAPK8, PPARG and NFKB in HFLS, RA-FLS cells. (F) qRT-PCR to detect RNA expression of MAPK8, PPARG, and NFKB in HFLS, and RA-FLS cells. *P < 0.05; **P < 0.01; P < 0.001. Table 4. Differential genes. P. Value logFC Target [97]GSE55235 PPARG 1.91E–05 1.37093938 Kaempferol Frutinone A MAPK8 7.62E–05 2.02493917 Kaempferol NFKB 5.18E–03 1.0478269 Kaempferol [98]GSE55457 PPARG 6.20E–02 0.6940568 Kaempferol Frutinone A MAPK8 4.93E–04 1.3109564 Kaempferol NFKB 7.25E–03 0.513227 Kaempferol [99]GSE77298 PPARG 1.33E–01 1.1229224 Kaempferol Frutinone A MAPK8 2.36E–01 0.6499273 Kaempferol NFKB 2.46E–01 0.6415523 Kaempferol [100]Open in a new tab Molecular docking Three crucial genes, MAPK8, PPARG, and NFKB, were chosen for molecular docking studies based on microarray data analysis and cellular validation. These key genes were then subjected to docking with the active compounds of kaempferol. MAPK8 protein exhibited the most stable binding to kaempferol, with a binding affinity of -9.43 kJ/mol (Fig. [101]5A). PPARG protein formed a binding complex with kaempferol and showed a binding affinity of -7.9 kcal/mol (Fig. [102]5B). NFKB protein interacted with kaempferol and demonstrated a binding affinity of -6.8 (Fig. [103]5C). The docking binding energy results for MAPK8, PPARG, and NFKB with kaempferol are presented in Table [104]5. The molecular docking results involve amino acid residues, number of hydrogen bonds, and length are presented in Table [105]6. Fig. 5. [106]Fig. 5 [107]Open in a new tab Molecular docking. (A) MAPK8 molecular docking. (B) PPARG molecular docking. (C) NFKB molecular docking. These images were provided by AutoDocktools (version 1.5.6), AutoDock Vina (version1.1.2), PyMol (version 2.5.7) software. Table 5. Binding energies of kaempferol and each key target. Key target PDBID Binding energy (kcal/mol) MAPK8 4L7F  − 9.3 MAPK8 4L7F  − 8.7 MAPK8 4L7F  − 8.5 PPARG 1NYX  − 7.9 PPARG 1NYX  − 7.9 PPARG 1NYX  − 7.6 NFKB 1SVC  − 6.8 NFKB 1SVC  − 6.4 NFKB 1SVC  − 6.3 [108]Open in a new tab Table 6. The molecular docking results involve amino acid residues, number of hydrogen bonds, and length. Key target Hydrogen bond number Hydrogen bond length(A) Amino acids MAPK8 3 2 MET-111 2.4 MET-111 2.4 GLN-117 PPARG 3 2.2 SER-225 2.3 GLU-471 2.6 ILE-472 NFKB 2 2.1 ASN-103 2.4 ASP-209 [109]Open in a new tab Inhibition proliferation and invasion of RA-FLS by kaempferol RA-FLS cells were exposed to various concentrations of kaempferol (0, 2, 5, 10, 20, and 40 μM). The impact of kaempferol on RA-FLS invasion was assessed through qRT-PCR, revealing that, with increasing concentration, kaempferol suppressed the RNA expression of matrix metallopeptidase 2 (MMP2), matrix metallopeptidase 9 (MMP9), N-cadherin, and Vimentin, particularly at 10 μM (Fig. [110]6A,B,C,D). The scratch assay demonstrated that kaempferol hindered the migration of RA-FLS cells at 10 μM (Fig. [111]6E). Immunofluorescence results for PCNA revealed that kaempferol suppressed the proliferation of RA-FLS cells at 10 μM (Fig. [112]6F). Additionally, CCK8 assay results indicated that kaempferol markedly reduced the cell viability of RA-FLS cells at 10 μM (Fig. [113]6G). Fig. 6. [114]Fig. 6 [115]Open in a new tab Inhibition of proliferation and invasion of RA-FLS by kaempferol. (A) RA-FLS cells were treated with different concentrations of kaempferol and RNA expression of MMP2 was detected by qRT-PCR. (B) RA-FLS cells were treated with different concentrations of kaempferol and RNA expression of MMP9 was detected by qRT-PCR. (C) Different concentrations of kaempferol-treated RA-FLS cells, qRT-PCR to detect RNA expression of N-cadherin. (D) Different concentrations of kaempferol treated RA-FLS cells, qRT-PCR to detect RNA expression of Vimentin. (E) Different concentrations of kaempferol treated RA-FLS cells, scratch detection of cell migration, Scale bar is 100 μm. (F) Different concentrations of kaempferol treated RA-FLS cells, immunofluorescence detection of PCNA protein expression, Scale bar is 50 μm. (G) Different concentrations of kaempferol-treated RA-FLS cells, cell viability detected by CCK8. *P < 0.05; **P < 0.01; ***P < 0.001. Kaempferol inhibits autophagy and attenuates oxidative stress in RA-FLS cells The study revealed that kaempferol notably decreased reactive oxygen species in RA-FLS and reduced the mitochondrial membrane potential level of the cells (Fig. [116]7A). Additionally, it inhibited the expression of SOD2 protein (Fig. [117]7B). Furthermore, kaempferol suppressed cellular autophagy in RA-FLS, as evidenced by Western blot results showing P62 protein accumulation and decrease in LC3B protein levels (Fig. [118]7C). Fig. 7. [119]Fig. 7 [120]Open in a new tab Kaempferol inhibits autophagy and attenuates oxidative stress in RA-FLS cells. (A) RA-FLS cells were treated with different concentrations of kaempferol, cellular reactive oxygen species were detected by DCFH-DA, and cell membrane potential was detected by JC-1, Scale bar is 100 μm. (B) RA-FLS cells were treated with different concentrations of kaempferol and protein expression of SOD2 was detected by Western blot. (C) In different concentrations of kaempferol-treated RA-FLS cells, Western blot detected protein expression of P62, LC3B. *P < 0.05; **P < 0.01; ***P < 0.001. Inhibition of cellular autophagy by kaempferol attenuates aberrant proliferation and inflammation in RA-FLS via the MAPK8/NLRP3 pathway To explore the intrinsic mechanism by which kaempferol inhibits the abnormal proliferation and inflammation of RA-FLS, RA-FLS cells were initially treated with varying concentrations of kaempferol. The results indicated that, as the kaempferol concentration increased, the expression of MAPK8 protein was significantly suppressed (Fig. [121]8A). Furthermore, the protein expression of NLRP3 and interleukin-1β (IL-1β) was markedly reduced (Fig. [122]8B). To further confirm that kaempferol inhibits the abnormal proliferation and inflammation of RA-FLS by targeting MAPK8, we conducted Western blot analysis of MAPK8 overexpression and the protein expression of NLRP3, PCNA, SOD2, and IL-1β. The results indicated that MAPK8 overexpression reversed the therapeutic effects of kaempferol, promoting cellular proliferation, inflammation, and oxidative stress in RA-FLS (Fig. [123]8C). Moreover, immunofluorescence results indicated that the overexpression of MAPK8 activated cellular autophagy, enhanced LC3B protein expression and diminished P62 protein accumulation (Fig. [124]8D,E). Fig. 8. [125]Fig. 8 [126]Open in a new tab Kaempferol inhibits cellular autophagy through the MAPK8/NLRP3 pathway to attenuate the abnormal proliferation and inflammation of RA-FLS. (A) Different concentrations of kaempferol were treated with RA-FLS cells, and the protein expression of P-MAPK8 and MAPK8 were detected by Western blot. (B) In different concentrations of kaempferol-treated RA-FLS cells, Western blot detected protein expression of NLRP3, IL-1β. (C) Overexpression of MAPK8 and kaempferol treated RA-FLS cells, Western blot detected protein expression of NLRP3, IL-1β, PCNA, SOD2. (D) Overexpression of MAPK8 with kaempferol-treated RA-FLS cells and immunofluorescence detection of LC3B protein expression, Scale bar is 50 μm. (E) Overexpression of MAPK8 with kaempferol-treated RA-FLS cells and immunofluorescence detection of P62 protein expression, Scale bar is 50 μm. *P < 0.05; **P < 0.01; ***P < 0.001. Discussion RA is the most prevalent autoimmune disease,which lacks a complete cure. Hence, the development of new therapeutic options is imperative^[127]21. Network pharmacology, molecular docking, and in vitro cellular experiments can offer theoretical support for drug discovery, drug target screening, and preclinical drug validation.In this study, we explored the potential target genes of kaempferol, an active component of the traditional Chinese medicine Compound Fuzi Decoction, for the treatment of RA. We also investigated the inherent mechanism of kaempferol in alleviating RA by inhibiting cellular autophagy through the MAPK8/NLRP3 pathway. Our results demonstrated that kaempferol by targeting and suppressing the expression of MAPK8/NLRP3 inhibited cellular autophagy and reactive oxygen species production in RA-FLS. This led to the alleviation of abnormal proliferation and invasion of RA-FLS cells and reduction in the level of inflammation in RA-FLS. Utilizing network pharmacology, we conducted screening and identifing 10 protein targets that potentially serve as key targets for the action of kaempferol against RA. Among them, three genes, MAPK8, PPARG, and NFKB, exhibited stable and differential expression in the microarrays. MAPK8 is activated under cellular stress conditions and participates in the regulation of apoptosis^[128]22. It can mediate apoptosis through multiple pathways, including the regulation of apoptosis-related proteins and the promotion of apoptosis via the mitochondrial pathway^[129]23,[130]24. MAPK8 is involved in the regulation of inflammatory cell migration, cytokine release, and the expression of inflammation-associated genes^[131]25–[132]27. MAPK8 is a crucial regulator of cellular stress response. It can be activated by various external stimuli (e.g., oxidative stress, heat stress, etc.), subsequently regulating cell survival or death^[133]28,[134]29. Overall, MAPK8, as a crucial signaling molecule, plays a pivotal role in cell physiology and pathology. PPARG is a nuclear receptor belonging to the nuclear-activated receptor (NR) superfamily, and it plays a key role in regulating a variety of biological processes. PPARG is one of the key transcription factors for adipocyte differentiation. It promotes adipocyte differentiation and fat storage, thus regulating overall lipid metabolism^[135]30,[136]31. PPARG also plays a crucial role in the regulation of glucose metabolism^[137]32,[138]33. PPARG has anti-inflammatory effects and can inhibit inflammatory responses by regulating the expression of inflammation-related genes^[139]34. Recent studies have indicated that PPARG also plays a crucial role in regulating the function of immune cells. It can influence the polarization state of monocytes and macrophages and regulate the activity of T cells^[140]35. NFKB is a crucial transcription factor that plays a pivotal role in various biological processes, including immunity, inflammation, apoptosis, and cell proliferation^[141]36. Recent studies have shown that NF-κB is involved in the regulation of innate and adaptive immune responses by regulating the expression of a variety of immune-related genes, including immunoglobulins, cytokines, and inflammatory mediators^[142]37. NF-κB regulates the expression of inflammation-related genes, such as tumor necrosis factor (TNF), interleukin (IL) family members, inflammatory mediators, and cell adhesion molecules, thereby regulating the occurrence and development of inflammatory responses^[143]38. Meanwhile, NF-κB is involved in regulating cell proliferation and tumor cell transformation by modulating the expression of genes such as cell cycle proteins, pro-proliferative factors, oncogenes, and tumor suppressors^[144]39,[145]40. These results suggest that the abnormal expression of MAPK8, PPARG, and NFKB proteins may be associated with RA. Therefore, the effects of kaempferol on these protein targets are worthy of further exploration. In this study, GO analysis revealed that therapeutic targets were enriched in biological processes (epithelial cell apoptotic processes, response to reactive oxygen species, regulation of DNA-binding transcription factor activity, intracellular receptor signaling pathways, negative regulation of lipid storage), cellular components (membrane rafts, membrane microdomains, pore complexes, immune synapses, myelin sheaths), and molecular functions (nuclear receptor activity, ligand-activated transcription factor activity, DNA-binding type transcription factor binding, transcriptional coregulator binding). The enriched KEGG results primarily involved inflammatory, oxidative, cellular autophagy, and cell cycle pathways. The top 10 enriched pathways comprised the TNF signaling pathway, IL-17 signaling pathway, apoptosis signaling pathway, NOD-like receptor signaling pathway, cell membrane DNA sensing pathway, NF-kappa B signaling pathway, Toll-like receptor signaling pathway, HIF-1 signaling pathway, cAMP signaling pathway, and p53 signaling pathway^[146]18–[147]20. Studies have reported that the TNF signaling pathway promotes the progression of RA by inducing the activation of fibroblast-like synoviocytes^[148]41. The IL-17 signaling pathway mitigates the progression of rheumatoid arthritis by inhibiting the inflammatory response triggered by TNF-α or LPS^[149]42. The hypoxia inducible factor-1 (HIF-1) signaling pathway and NF-kB signaling pathway are involved in RA progression by regulating ROS clearance through a cascade^[150]43. Additionally, the NOD-like receptor signaling pathway, Toll-like receptor signaling pathway, and p53 signaling pathway are also implicated in the development of RA^[151]44–[152]46. The molecular mechanism of kaempferol against RA was validated through molecular docking and in vitro cellular experiments. Molecular docking analysis revealed that kaempferol exhibited varying affinities with the screened key protein targets, with MAPK8 showing the highest affinity. This suggests that MAPK8 plays a pivotal role in kaempferol’s ability to counteract RA. It was found that in RA, MAPK8 may influence cell proliferation and invasion. Additionally, MAPK8 is involved in mitigating oxidative stress in RA-FLS, among other functions. In vitro cellular experiments further demonstrated that kaempferol inhibited cellular autophagy and reactive oxygen species production in RA-FLS by suppressing the expression of MAPK8/NLRP3, thereby alleviating the abnormal proliferation and invasion of RA-FLS cells and reducing the level of inflammation in RA-FLS. It can be observed that the MAPK8-mediated signaling pathway plays a pivotal role in RA. In this study, we employed a comprehensive analysis combining network pharmacology, molecular docking, and in vitro cellular experiments to explore the potential mechanism of kaempferol in RA. However, there are still some limitations, and further experimental validation of the signaling pathway mediated by MAPK8 is necessary. The current study offers a research direction and theoretical support for further elucidating the specific mechanism of kaempferol against RA. Conclusion Our study offers a theoretical foundation for utilizing herbal compounds as effective drugs in the treatment of RA. Our findings suggest that kaempferol participates in the progression of RA by targeting MAPK8. Furthermore, we elucidated the intrinsic mechanism through which kaempferol mitigates RA by inhibiting cellular autophagy via the MAPK8/NLRP3 pathway. This provides theoretical support for a more effective exploration of therapeutic targets for RA. Supplementary Information [153]Supplementary Information 1.^ (1MB, pdf) [154]Supplementary Information 2.^ (5.8MB, ppt) Author contributions X.C. and W. G. conceived and designed the study, H. L. completed the experiments, H. L. and X. F. wrote the paper, and S.C. assisted with some of the experiments. All authors read and agreed on the final manuscript. Funding This work was supported by the Natural Science Foundation of Anhui Province (No.2018085MH264), the Scientific Research Fund Project of Anhui Medical University (No. 2020xkj181), the Basic and Clinical Cooperative Research Promotion Plan of Anhui Medical University (No. 2021xkjT003). Data availability The research data used and analysed during the current study will be available upon reasonable request from the corresponding authors. Declarations Competing interests The authors declare no competing interests. Ethics approval and consent to participate Study protocol obeyed the Declaration of Helsinki for all human or animal experimental investigations and was approved and Filed by the Biomedical Ethic committee board of Anhui Medical University. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Haomiao Liu, Huangying Lu and Xuefei Fan contributed equally to this work. Contributor Information Xiaoyu Chen, Email: cxyayd@163.com. Weilu Gao, Email: weiqiang83@163.com. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-025-91311-6. References