Joint analysis of the metabolomics and transcriptomics uncovers the dysregulated network and develops the diagnostic model of high-risk neuroblastoma

作者：Bang Du, Fei Zhang, Qiumei Zhou, Weyland Cheng, Zhidan Yu, Lifeng Li, Jianwei Yang, Xianwei Zhang, Chongchen Zhou, Wancun Zhang · 发表于：Scientific Reports · 年份：2023 · DOI：10.1038/s41598-023-43988-w · 被引用次数：13 · 研究领域：Cancer, Hypoxia, and Metabolism、Neuroblastoma Research and Treatments、Ubiquitin and proteasome pathways

Abstract High-risk neuroblastoma (HR-NB) has a significantly lower survival rate compared to low- and intermediate-risk NB (LIR-NB) due to the lack of risk classification diagnostic models and effective therapeutic targets. The present study aims to characterize the differences between neuroblastomas with different risks through transcriptomic and metabolomic, and establish an early diagnostic model for risk classification of neuroblastoma.Plasma samples from 58 HR-NB and 38 LIR-NB patients were used for metabolomics analysis. Meanwhile, NB tissue samples from 32 HR-NB and 23 LIR-NB patients were used for transcriptomics analysis. In particular, integrative metabolomics and transcriptomic analysis was performed between HR-NB and LIR-NB. A total of 44 metabolites ( P < 0.05 and fold change > 1.5) were altered, including 12 that increased and 32 that decreased in HR-NB. A total of 1,408 mRNAs ( P < 0.05 and |log 2 (fold change)|> 1) showed significantly altered in HR-NB, of which 1,116 were upregulated and 292 were downregulated. Joint analysis of both omic data identified 4 aberrant pathways ( P < 0.05 and impact ≥ 0.5) consisting of glycerolipid metabolism, retinol metabolism, arginine biosynthesis and linoleic acid metabolism. Importantly, a HR-NB risk classification diagnostic model was developed using plasma circulating-free S100A9, CDK2 , and UNC5D , with an area under receiver operating characteristic curve of 0.837 where the sensitivity and specificity in...