Diagnostic value and immune infiltration characterization of YTHDF2 as a critical m6A regulator in osteoarthritic synovitis

Abstract

Background: N6-methyladenosine (m6A) is a universal RNA modification pattern regulated by multiple m6A regulators. In osteoarthritis (OA), m6A regulators influence disease progression by regulating cartilage degradation. However, the function of m6A regulators in synovial tissue remains unclear. In this work, we investigated the biological significance of m6A regulators in osteoarthritic synovitis.

Methods: Datasets were acquired from Gene Expression Omnibus. Differential analysis of merged data identified the differentially expressed m6A regulators. Machine learning models were used to evaluate genetic importance. To predict disease risk, a nomogram was constructed based on above m6A regulators. Cluster analysis divided the OA sample into different subgroups. Immune infiltration revealed the immune m6A regulators, which were validated using clinical samples. Eventually, a competing endogenous RNA (ceRNA) network was constructed.

Results: We acquired five differentially expressed m6A regulators and a random forest model. The nomogram accurately predicted disease risk. We identified 122 differentially expressed genes between two m6A subgroups. The analysis of immune infiltration showed that YTHDF2 was an immune-related m6A regulator closely related with macrophages. In clinical samples, the protein and mRNA contents of YTHDF2 were consistent with the results of bioinformatic analysis. The ceRNA network based on YTHDF2 revealed 75 lncRNA nodes and 19 miRNA nodes.

Conclusion: YTHDF2 has a high diagnostic value in the synovitis of OA and significantly influences the immune status of patients. Hence, YTHDF2, a critical m6A regulator, may provide a biomarker for diagnosis and immune therapy of osteoarthritic synovitis.

Keywords: Cluster analysis; Immune infiltration; Macrophage; Osteoarthritic synovitis; ceRNA network; m6A regulators.

Fig. 1

Flow chart and data preprocessing. A Flow chart of this study. B Principal component analysis (PCA) before batch correction. C PCA after batch correction

Fig. 2

Expression differences in m6A regulators. A Boxplot of differences in m6A regulators between OA and control, *P < 0.05, **P < 0.01, ***P < 0.001. B Heat map of OA-related m6A regulators. C Chromosome circle diagram of 25 m6A regulators

Fig. 3

Construction of random forest (RF) and support vector machine (SVM). A Residual boxplots of RF and SVM. B Residual reverse cumulative distribution of RF and SVM. C Receiver operating characteristic curves of RF and SVM. D Importance score of m6A regulators in the accordance with RF

Fig. 4

Establishment of nomogram. A Nomogram of five osteoarthritis-related m6A regulators. B Calibration curve evaluates the prediction accuracy of nomogram. C Decision curve assesses the predictive ability of nomogram. D Clinical impact curve evaluates the clinical value of nomogram

Fig. 5

Cluster analysis of osteoarthritis (OA) samples based on m6A regulators. A Consistency matrix plot for K = 2–5. B Boxplot of differential expression of OA-related m6A regulators. C Heat map of differential expression of OA-related m6A regulators. D PCA of cluster A and cluster B. E Boxplot of the m6A score of the two subtypes

Fig. 6

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) between m6A clusters. A Bar plot of KEGG. B Bubble plot of KEGG. C Bar plot of GO. D Bubble plot of GO

Fig. 7

Immune infiltration of m6A regulators in osteoarthritis (OA) by ssGSEA. A Correlation of immune cells and m6A subtypes. B Association analysis of immune infiltration and five OA-related m6A regulators. C Correlation of immune cells and YTHDF2. *p < 0.05, **p < 0.01, and ***p < 0.001

Fig. 8

Immune infiltration of YTHDF2 in osteoarthritis (OA) by CIEERSORT. A Bar plot of immune cell infiltration in merged data. B Lollipop of correlation between YTHDF2 and immune cells. C Lollipop correlation between YTHDF2 and macrophage genes. Red font indicates P < 0.05

Fig. 9

Bioinformatic validation and KEGG pathways of YTHDF2. A Violin plot of YTHDF2. B ROC of YTHDF2. C High expression of YTHDF2 in GSEA. D Low expression of YTHDF2 in GSEA. E KEGG pathways of YTHDF2 according to GSVA

Fig. 10

Validation of the signature m6A regulators in clinical samples. A Protein content of YTHDF2 in osteoarthritis (OA) and normal person by western blotting. B Expression of YTHDF2 protein in OA and control groups; C Expression level of YTHDF2 mRNA in OA and control groups. ****p < 0.0001

Fig. 11

Network of ceRNA of YTHDF2. The red diamond label means mRNA, the green triangle label meas miRNA, and the blue ellipse label means lncRNA