[Revealing the role of Hippo pathway in osteoarthritis based on transcriptomic analysis].

Abstract

Objective: To explore the role of Hippo signaling pathway-related genes in osteoarthritis (OA), identify potential diagnostic biomarkers, and provide a basis for precision diagnosis and targeted therapy.

Methods: Transcriptomic data from GSE114007 and GSE57218 in GEO database were analyzed. Differentially expressed genes (DEGs) were identified using DESeq2 and intersected with Hippo pathway genes and weighted gene co-expression network analysis (WGCNA) results to obtain differentially expressed Hippo-related genes (DE-HRGs). Candidate genes were screened using the least absolute shrinkage and selection operator (LASSO) regression, support vector machine-recursive feature elimination (SVM-RFE), and the Boruta algorithm. Their diagnostic value was assessed with receiver operating characteristic (ROC) curves and neural network model. To validate the findings, a rat OA model was established by using Hulth method with control and model groups, 8 rats in each group. The control group received no treatment, while OA group were underwent surgical modeling. Four weeks after modeling, X-ray examinations of knee joints were conducted to observe joint space, the shape of joint surfaces, and the formation of bone spurs. The expression levels of epidermal growth factor receptor (EGFR) and protein kinase AMP-activated non-catalytic subunit beta 2 (PRKAB2) in cartilage tissue were detected by real-time fluorescence quantitative polymerase chain reaction (RT-qPCR). Gene set enrichment analysis (GSEA) was employed to analyze functional pathways of EGFR and PRKAB2, and single-sample gene enrichment analysis (ssGSEA) was used to evaluate the relationship between key genes and immune cell infiltration.

Results: A total of 3, 996 DEGs were identified, and 11 DE-HRGs were obtained. LASSO regression, SVM-RFE and Boruta algorithms jointly selected 5 candidate genes. Among them, EGFR and PRKAB2 had an area under the curve (AUC) greater than 0.8 in both training set and validation set (0.823 and 0.811 respectively), demonstrating excellent diagnostic performance. Neural network model further demonstrated its discriminative ability (with an AUC of 0.835). In rabbit OA model, X-ray images showed control group had narrowed knee joint space accompanied by bone spurs formation. The results of RT-qPCR showed compared with control group, the expressions of EGFR (1.5±0.8) and PRKAB2 (1.2±0.7) in OA group were lower than those in control group (3.8±1.5), (2.9±1.3), and the differences were statistically significant (P<0.05). The functional enrichment analysis revealed both EGFR and PRKAB2 were involved in signaling pathways such as Hippo and adenosine monophosphate-activated protein kinase (AMPK). The analysis of immune infiltration results showed PRKAB2 was positively correlated with Th17 cells (r=0.47, P<0.01), while EGFR was negatively correlated with natural killer T cells (r=-0.61, P<0.001).

Conclusion: EGFR and PRKAB2 can serve as potential biomarkers for OA. They may participate in the occurrence and development of OA by regulating Hippo pathway and immune microenvironment, providing new ideas for molecular mechanism research and targeted therapy of OA.