Development and validation of biomarkers related to PANoptosis in osteoarthritis

Y.-Y. Zhang, H.-S. Zhao, Y.-F. Sun, B.-W. Lu, L. Sun

First Clinical Medical College, Shandong University of Traditional Chinese Medicine, Jinan, China. Sunludr@163.com

---

• Bioinformatics

OBJECTIVE: Osteoarthritis (OA) is a high-incidence disease of the orthopedic system. However, studies on the molecular mechanisms of OA and pyroptosis, apoptosis, and necroptosis (PANoptosis) at the transcriptome level remain scarce. Therefore, this study purposed to detect biomarkers in OA and explore their relationship to the immune microenvironment.

MATERIALS AND METHODS: OA-related expression data was sourced from the Gene Expression Omnibus (GEO) database. Subsequently, differentially expressed analysis and a Venn diagram were performed to obtain differentially expressed PANoptosis-related genes (DEPGs). Furthermore, the least absolute shrinkage and selection operator (LASSO), Support Vector Machine-Recursive Feature Elimination (SVM-RFE), and random forest (RF) were implemented to screen diagnostic genes. Receiver operating characteristic (ROC) curves were performed to verify the diagnostic ability of the diagnostic genes. Next, immune infiltration analysis was performed to find the relationships between differential immune cells (OA vs. normal) and diagnostic genes. Finally, drug prediction analysis was also carried out, and the expression of diagnostic genes was verified in external datasets.

RESULTS: A total of 62 DEPGs were identified, which were enriched for regulating apoptotic signaling pathways, tumor necrosis factor (TNF) signaling pathways, and other related pathways. Three feature genes, nuclear factor-kappa-B inhibitor-alpha (NFKBIA), RING finger protein 34 (RNF34), and serine incorporator 3 (SERINC3) were obtained by intersecting genes obtained by the LASSO regression algorithm, SVM algorithm, and RF algorithm and showed excellent diagnostic efficacy with the Area under the curve (AUC) values of individual genes were all greater than 0.7, indicating that the model was more effective. Immuno-infiltration analysis showed that RNF34 was positively correlated with CD56dim natural killer cells, type 17 helper T cells, and NFKBIA was positively correlated with plasmacytoid dendritic cells. Additionally, 12 drugs were predicted by NFKBIA, such as gambogic acid and dioscin. In addition, NFKBIA and SERINC3 were significantly downregulated, and RNF34 was upregulated in OA samples.

CONCLUSIONS: Three genes (NFKBIA, RNF34, and SERINC3) related to PANoptosis, were obtained by bioinformatics analysis, which would provide a new direction for the diagnosis and treatment of OA.

Free PDF Download