OA is one of the most common joint diseases in the elderly, with a high public health burden and no cure. Articular cartilage has the function of buffering and reducing friction, and it is also the most severely degenerated part of OA. Therefore, rebuilding the integrity of articular cartilage is expected to replace joint replacement as a new method for radical OA.
Oxidative stress and reactive oxygen species (ROS) have been shown to be strongly associated with the occurrence of OA. When chondrocytes, synoviocytes, and osteoblasts are continuously subjected to external mechanical stress, they can produce excessive pro-inflammatory mediators to break down the pro-inflammatory mediators. Oxidative/antioxidant balance, which in turn degrades the extracellular matrix. As a member of the KLFs family, Kruppel-like factor 9 (KLF9) plays an important role in the oxidative stress response. Studies have shown that Nrf2 can stimulate the expression of KLF9, and KLF9 inhibits the expression of several important antioxidant enzymes such as thioredoxin reductase 2, resulting in a further increase in Klf9-dependent ROS and ultimately the degradation of cartilage. Through the KEGG pathway enrichment analysis, we learned that the differentially expressed genes in OA are mainly involved in the IL-17 signaling pathway, and IL-17 can also promote the process of oxidative stress. Whether there is a potential connection between the two needs further research.
Epiphycan (EPYC) is a protein-coding gene. It is a member of the leucine-rich small repeat proteoglycan (SLRP) family. This gene consists of seven exons and regulates fibrillation by interacting with collagen fibrils and other extracellular matrix proteins. EPYC is involved in cartilage formation in normal synovial tissue. EPYC knockout mice develop osteoarthritis with age[39–40]. In the present study, EPYC is overexpressed in OA, and we speculate that this is most likely because the destruction of articular cartilage causes chondrocytes to increase EPYC production in an attempt to repair the damaged extracellular matrix (ECM). Since EPYC belongs to the SLRPS family, the effects of the SLRP family on cartilage and the mechanisms involved in the occurrence of OA are numerous and complex, including changes in the extracellular collagen network and TGF-b signaling pathway. Therefore, the mechanism of EPYC regulation in OA needs to be further elucidated. In addition, studies have confirmed that NSAIDS drugs can reduce the expression of EPYC in prostate cancer cells. As the first-line treatment of OA, NSAIDS drugs need to be further explored.
The CIBERSORT score is widely used in gene expression profiling to quantify immune cell scores with high accuracy. The infiltration of immune cells in OA synovial tissue has become the consensus of many scholars. Among them, CD4 + T cells, mast cells, and macrophages play an important role in synovial inflammation. IgE-dependent mast cell activation and the pathogenic role of mast cell-mediated tryptase in osteoarthritis have been demonstrated, but mast cells themselves are not differentially expressed in OA synovial tissue. In this study, the immune infiltration analysis showed that resting mast cells were highly expressed in OA synovial tissue, while activated mast cells were lowly expressed. We speculate that this is probably because mast cells are not directly involved in the pathogenic process of OA, but mediate Other proteases or histamine indirectly lead to the occurrence of OA, but the specific mechanism of its role in OA needs to be further elucidated. Regulatory T cells (Tregs) play an important immunomodulatory role in many inflammatory and autoimmune diseases, but they are more inhibiting osteoclasts and helper T cells to protect local articular cartilage from destruction[43–45]. Our experimental results suggest that regulatory T cells infiltrate the OA synovium, which is likely related to synovial tissue destruction leading to reactive proliferation of regulatory T cells to suppress local inflammatory responses. Of course, this requires further verification by experiments.
It is not the first time that machine learning algorithms have been used in gene screening for OA. In past experiments, we believed that the threshold of difference was too low, which resulted in too many differential genes after filtering, which affected the accuracy of enrichment analysis and machine learning algorithms. On this basis, we increased the threshold of the difference analysis to 3 times, that is, set the LogFC to 1.5. We believe that the results of this analysis are more accurate and meaningful.
In short, we processed the chip expression data by computer, screened the characteristic diagnostic genes of OA by using machine learning algorithm, and explored the relationship between them and immune cells, in order to provide reference direction for the early diagnosis and treatment of OA.