MSH6 gene methylation on clinical pathology and diagnosis in retinoblastoma: a bioinformatics analysis

To explore the methylation status of MSH6 in retinoblastoma (RB) and its impact on clinicopathological features and diagnosis.

Differentially expressed genes were identified through bioinformatics screening of the GSE24673 and GSE125903 datasets, combined with GeneCards database analysis. A total of 102 RB patients and 62 trauma-enucleated controls between January 2018 and December 2023 were enrolled, with their clinicopathological data and retinal tissues collected. The mRNA and methylation levels of MSH6 in retinal tissues were detected using real-time quantitative polymerase chain reaction (PCR) and methylation-specific PCR. Western blot analysis was conducted in one pair of RB and control tissues for preliminary protein-level validation of MSH6 expression. Based on the methylation status of MSH6, RB patients were categorized into two groups: low-methylation and high-methylation. Both univariate and multivariate analyses were conducted to identify independent factors influencing the methylation levels using clinicopathological data. Receiver operating characteristic (ROC) curves were applied to evaluate the diagnostic potential of MSH6 methylation in RB.

Bioinformatics analysis of public datasets revealed that MSH6 expression was downregulated across multiple cancers, RB. Consistently, in clinical RB tissues, MSH6 mRNA expression was significantly lower than that in control retinal tissues, whereas the promoter methylation level of MSH6 was markedly higher (both P<0.001), indicating that promoter hypermethylation may contribute to transcriptional silencing of MSH6 in RB. Patients with higher MSH6 methylation levels showed more advanced pathological classification and a higher frequency of metastasis. Multivariate logistic regression confirmed that metastatic status (P=0.008, OR=3.51) and pathological classification (P=0.005, OR=3.7) were independent factors associated with MSH6 methylation. Receiver operating characteristic (ROC) analysis demonstrated that MSH6 methylation could effectively distinguish RB tissues from non-tumorous controls (AUC=0.847, sensitivity=78.43%, specificity=80.65%), suggesting that MSH6 hypermethylation may serve as a potential diagnostic biomarker for RB.

The methylation level of the MSH6 gene may be a key factor in RB pathogenesis. The methylation status of the MSH6 gene is closely associated with clinicopathological features and shows diagnostic potential.