Calycosin ameliorates osteoarthritis by regulating the imbalance between chondrocyte synthesis and catabolism

Statistical analysis was processed with GraphPad Prism 9.2 software. Data were analyzed by using one-way analysis of variance (ANOVA) followed by post hoc Dunnett’s t-test for multiple comparisons and presented as the mean ± SD. A value of P < 0.05 was statistically significant.

The animal experimental design is illustrated in Fig. 2a. To evaluate the efficacy of calycosin (PubChem CID: 5,280,448, Fig. 2b), we performed pathological staining of cartilage tissue, including Safranin O/Fast green and toluidine blue. Our results indicate that, in the sham group, sections were uniformly stained. Cartilage surface was smooth and continuous, with oval or round-shaped chondrocytes. The cartilage layer was thick and intact. However, the cartilage thickness and chondrocyte populations in the model group were lower than those in the sham group, which resulted in higher OARSI scores (Fig. 2c, d). It is noteworthy that these phenomena were reversed when treated with calycosin. The calycosin-treated group shows superficial zone restoration, uniform staining, and regular distribution of chondrocytes (Fig. 2d). Col-2 and Sox-9 are the main proteins in cartilage. Immunofluorescence staining showed that Col-2 and Sox-9 were reduced in the OA group but increased in the cartilage matrix after being treated with calycosin (Fig. 3).

Our screening strategy identified 234 calycosin-related and 1225 OA-related targets, respectively. Each predicted target occurs at least two times in different databases. According to the Venn diagram, 58 overlapping genes were identified (Fig. 4a). Protein–protein interaction (PPI) network of 55 nodes (3 isolated nodes are removed), 367 edges were established from the STRING database (confidence > 0.4, medium confidence) (Fig. 4b). To find the hub genes and clusters, the Cytohubba and MCODE plugins in Cytoscape were used. Accordingly, we identified two core clusters (Figure A1a-b, Supplementary figure) and 13 hub genes (Ccnd1, Mapk3, Tp53, Egfr, Hsp90aa1, Ptgs2, Tnf, Esr1, Vegfa, Mapk14, Igf1r, Ar, Alb) (Fig. 4c). Among them, 13 genes overlapped between the two plugins, including COX2, EGFR, and mitogen-activated protein kinase (MAPK14).

We performed GO functional enrichment and KEGG pathway analysis using the DAVID database (a P < 0.01 was considered Statistical significance). The GO terms were classified into three categories: biological process (BP), cellular component (CC); and molecular function (MF). Finally, 372 GO Terms (260 in GO_BP, 45 in GO_CC, and 67 in GO_MF) and 100 KEGG pathways were identified. The top 15 GO functional annotations were presented in Fig. 5a. The categories of GO_BP were response to positive regulation of transcription from RNA polymerase IIpromoter, signal transduction, negative regulation of apoptotic process, cytokine-mediated signaling pathway, response to drug, response to xenobiotic stimulus, positive regulation of MAPK cascade, and positive regulation of peptidyl-serine phosphorylation. The critical GO_CC were related to extracellular region, extracellular exosome, macromolecular complex, endoplasmic reticulum lumen and fcolin-1-rich granule lumen. The critical GO_MF were mainly enriched in the identical protein binding and enzyme binging. The Sankey plot indicates the related genes contained in the top 10 GO terms, in which COX2, TNF, and AR occur six times (Fig. 5b). The KEGG annotations are mainly associated with pathways in cancer, prostate cancer, endocrine resistance, proteoglycans in cancer, chemical carcinogenesis-receptor activation, and the PI3K-Akt signaling pathway, etc. (Fig. 6a, b).

The 12 hub targets (ALB was removed) and calycosin were used as receptors and ligand, respectively, for further molecular docking to investigate the affinity between ligand and protein targets. The stability of the binding of the receptors and ligand depends on the binding affinity. The lower the binding affinity, the more stable the binding conformation of the receptor and the ligand. Therefore, we selected the receptor with the lowest affinity that binds to calycosin. The screening results are shown in Table 2, which revealed that COX-2 (Binding sites: GLY45, HIS39, CYS47, CYS36, GLY136, LEN153, GLN452, and PRO154), MAPK14 (Binding sites: SER293, LEU291, HIS199, ASP294, LEU246, ILE250, and ALA255) and EGFR (Binding sites: SER14, SER201, and LYS207) could be representative proteins (Fig. 7). The interactions between calycosin and other potential targets are shown in Supplementary Figure A2. These evidences suggest that calycosin has the ability to co-target COX-2, EGFR and MAPK14.

Calycosin alleviates IL-1β induced inflammation in chondrogenic ADTC5 cells.

In chondrogenic ADTC5 cells, the proinflammatory cytokine IL-1β is markedly overexpressed in OA and mimics the chondrocyte inflammation model. ADTC5 cells were exposed to IL-1β (10 ng/mL), which caused a considerable decrease in COX-2 and phospho-EGFR (p-EGFR) expression (Fig. 8a-c), while no significant changes of MAPK14 or p-MAPK14 (Figure A1, Supplementary figure) were observed. In addition, compared to the OA group, ADTC5 chondrogenic cells showed a moderate decrease of MMP-9 when co-cultured with 32 μM calycosin (Fig. 8d). In agreement with in vivo experiments, immunofluorescence staining (Fig. 8e) demonstrated that Col-2 and Sox-9 were remarkably increased after calycosin treatment. The quantification analysis results were shown in Fig. 8 f-j.

OA is a typical chronic inflammatory disease, and there is still no drug that can prevent the degenerative cascade reaction. With the aggravation of OA, the cracks on the surface of the cartilage deepened, ulcers formed, and gradually developed into deep cartilage tissue [65]. This is closely related to an imbalance between chondrocyte synthesis and catabolism, including excessive degradation of structural proteins such as Col-2 and increased synthesis of MMPs [66, 67]. Currently, the first-line drugs approved by the FDA are mainly NSAIDs, among which the COX-2 inhibitors can effectively relieve the pain. However, both traditional NSAIDs and selective COX-2 inhibitors have a certain degree of adverse reactions, which results in limited clinical application. Calycosin has been demonstrated to have various pharmacological activities. It especially exhibits a strong anti-inflammatory effect. On the one hand, since low-grade inflammation plays a critical role in the pathogenesis of OA, calycosin can inhibit inflammatory factors such as COX-2 to relieve OA. On the other hand, MMPs play a crucial role in the destruction of cartilage, calycosin decreases the MMP-9 produced by chondrocytes [66, 68]. After being treated with calycosin, the expressions of Col-2 and Sox-9 are also increased. This contributes to cartilage collagen synthesis. Mice lacking EGFR in cartilage (Col2-Cre) exhibited reduced chondrocyte and synovial fluid secretion compared to their wild-type siblings [67], demonstrating that EGFR is essential for maintaining the homeostasis of articular cartilage. However, like some other growth factor signaling pathways, EGFR plays a dual role in articular cartilage. On the one hand, EGFR signaling can promote articular surface lubrication by increasing proteoglycan 4 (PRG4) [69]. On the other hand, EGFR signaling also plays a catabolic action by inhibiting the chondrogenic master transcription factor Sox-9 and stimulating MMP9 expression [70]. Therefore, further study is needed to determine the specific role that EGFR signaling plays in the OA process. Though calycosin can decrease the phosphorylation levels of SRC, EGFR, ERK1/2, and Akt in breast cancer lines and meningitis [63, 71], its role in chondrocytes remains unclear. Our results show that the binding capacity of calycosin with EGFR is consistent with previous studies [71, 72]. In this study, we report the beneficial aspects of inhibiting EGFR, the effects of calycosin did not cause negative impacts. This phenomenon may be due to the various biological activities of calycosin. We have only discussed the anti-inflammatory effects here. Other.

pharmacological effects of OA are still worth exploring. For example, calycosin has been reported to have a strong osteogenic activity, it can inhibit bone resorption and stimulate bone formation [73, 74]. Another study shows that calycosin plays an anti-osteoporosis effect through the IGF1R/PI3K/Akt signaling pathway [30]. In addition, calycosin exerts estrogenic properties, which are also controversial in OA, probably related to interaction with EGFR. Therefore, the various physiological activities and mechanisms of calycosin need to be further studied.

Our findings provide insights into the relationship between OA and calycosin, the hub genes, and significant GO and KEGG pathways, which deserve further study.

Within hub genes, vascular endothelial growth factor A (VEGFA) has been shown to be involved in the OA process [75]. Increased expression of VEGFA was positively correlated with the secretion of ECM, including collagen II and aggrecan [76]. In addition, age and gender occupied an unignored position among the well-established epidemiological risk factors of OA [1, 77]. This phenomenon has attracted the attention of researchers who are interested in the role of estrogen in OA [78]. An investigation of OA combining transcriptomic and proteomic characterization data revealed that the estrogen receptor agonists diethylstilbestrol and alpha-estradiol hold promise as therapeutic candidates and drug targets [79]. Calycosin, is a member of 7-hydroxyisoflavones and a member of 4'-methoxyisoflavones, showing a similar structure to estrogen estradiol and acts as a plant-derived phytoestrogen, it is functionally related to an isoflavone. In addition, it possesses estrogen-like activity [36]. Our results also indicate that calycosin has the capacity to bind with estrogen receptors ESR1, which makes it widely studied in cancers like breast cancer. Therefore, seeking alternative solutions to estrogen replacement therapy may benefit from phytoestrogen therapy due to its fewer side effects [80, 81]. In the present study, the phosphoinositide-3-kinase (PI3K)/Akt signaling pathway was significantly enriched. The PI3K-Akt signaling pathway plays a crucial role in OA, as does the mitogen-activated protein kinase (MAPK) signaling pathway. Studies have reported that the increased expression level of P-Akt can induce chondrocyte proliferation [82] and inhibition of phosphorylation of MAPK and NF-κB can inhibit the expression of COX-2, MMPs, iNOS in chondrocytes [83, 84]. It is reported that calycosin induces apoptosis in osteosarcoma cell lines via ERβ-mediated PI3K/Akt signaling pathways. Recent studies have shown that NF-κB and PI3K/AKT are inhibited by calycosin in OA chondrocytes, suggesting that it may act as a protective agent against OA. [85]. The pathways and genes enriched in this paper are still valuable to be explored. Nevertheless, this paper does not verify the intersection targets involved one by one, it only provides a map of possible targets, which is the limitation of our study.

In this paper, we explored the possible targets of calycosin in the treatment of OA through bioinformatics mining. Furthermore, experiments indicated that calycosin has therapeutic potential by targeting COX-2/EGFR. However, further research is needed before calycosin can be considered a therapeutic option for OA. In summary, by integrating network pharmacology and molecular experiments, we proposed a map of possible targets and a relatively low-risk approach to OA. These findings provide strong support for calycosin as a potential therapeutic, in response to the global clinical challenge of OA.