Identification of several inflammation-related genes based on bioinformatics and experiments

Abstract

Background: Osteoarthritis (OA) is a common disease of elderly individuals, with an unclear pathogenesis and limited treatment options to date. Inflammation occurs prominently in osteoarthritis, thereby making anti-inflammatory treatments promising in clinical outcomes. Therefore, it is of diagnostic and therapeutic significance to explore more inflammatory genes.

Method: In this study, appropriate datasets were first acquired through gene set enrichment analysis (GSEA), followed by inflammation-related genes through weighted gene coexpression network analysis (WGCNA). Two machine learning algorithms (random forest-RF and support vector machine-recursive feature elimination, SVM-RFE) were used to capture the hub genes. In addition, two genes negatively associated with inflammation and osteoarthritis were identified. Afterwards, these genes were verified through experiments and network pharmacology. Due to the association between inflammation and many diseases, the expression levels of the above genes in various inflammatory diseases were determined through literature and experiments.

Result: Two hub genes closely related to osteoarthritis and inflammation were obtained, namely, lysyl oxidase-like 1 (LOXL1) and pituitary tumour-transforming gene (PTTG1), which were shown to be highly expressed in osteoarthritis according to the literature and experiments. However, the expression levels of receptor expression-enhancing protein (REEP5) and cell division cycle protein 14B (CDC14B) remained unchanged in osteoarthritis. This finding was consistent with our verification from the literature and experiments that some genes were highly expressed in numerous inflammation-related diseases, while REEP5 and CDC14B were almost unchanged. Meanwhile, taking PTTG1 as an example, we found that inhibition of PTTG1 expression could suppress the expression of inflammatory factors and protect the extracellular matrix through the microtubule-associated protein kinase (MAPK) signalling pathway.

Conclusions: LOXL1 and PTTG1 were highly expressed in some inflammation-related diseases, while that of REEP5 and CDC14B were almost unchanged. PTTG1 may be a potential target for the treatment of osteoarthritis.

Keywords: GSEA; Inflammation; Negatively related genes; Osteoarthritis; WGCNA.