Abstract
OBJECTIVE: Subtypes of medulloblastoma (MB), WNT, SHH, Group 3 and Group 4, have different prognoses and the impact of abnormal nucleotide metabolism remains unclear. Multi-omics data was integrated to analyze the effect of nucleotide metabolism genes on MB molecular characteristics, prognosis and drug sensitivity. The aim was to identify subtype-specific therapeutic targets to inform treatment. METHODS: A total of 132 MB samples datasets were accessed, UMAP and hierarchical clustering analysis were performed using the expression profiles of 1804 nucleotide metabolism genes and association with molecular subtype was evaluated. Genes were screened and prognostic signatures constructed by univariate Cox regression, cross-validation and LASSO-Cox regression and predictive efficacy was verified in training and independent validation sets. Pharmacogenomic data were combined to predict differences in drug sensitivity between high- and low-risk groups. RESULTS: Nucleotide metabolism gene expression profiles were distinct among the four major MB subtypes, indicating coupling of metabolic reprogramming and tumor lineage. 51 prognostic genes were screened and were involved in RNA splicing, anatomical structure maintenance and purine compound metabolism. A signature was constructed from 17 nucleotide metabolism genes which distinguished high from low-risk (p < 0.001) groups and gave an independent prognosis in multivariate analysis. Drug sensitivity analysis showed the high-risk group to be more sensitive to MEK/ERK inhibitors and the low-risk group to PLK1, IGF1R/IR, ROCK and mTORC1/2 inhibitors. CONCLUSION: Nucleotide metabolism-transcription coupling endows MB subtypes with heterogeneity and affects prognosis. The signature is a quantitative tool for individualized risk assessment and metabolism targeted therapy.