Multiomics Analysis Reveals Neuroblastoma Molecular Signature Predicting Risk Stratification and Tumor Microenvironment Differences.

Abstract

Neuroblastoma (NB) remains associated with high mortality and low initial response rate, especially for high-risk patients, thus warranting exploration of molecular markers for precision risk classifiers. Through integrating multiomics profiling, we identified a range of hub genes involved in cell cycle and associated with dismal prognosis and malignant cells. Single-cell transcriptome sequencing revealed that a subset of malignant cells, subcluster 1, characterized by high proliferation and dedifferentiation, was strongly correlated with the hub gene signature and orchestrated an immunosuppressive tumor microenvironment (TME). Furthermore, we constructed a robust malignant subcluster 1 related signature (MSRS), which was an independent prognostic factor and superior to other clinical characteristics and published signatures. Besides, TME differences conferred remarkably distinct therapeutic responses between high and low MSRS groups. Notably, polo-like kinase-1 (PLK1) was one of the most crucial contributors to MSRS and remarkably correlated with malignant subcluster 1, and PLK1 inhibition was effective for NB treatment as demonstrated by in silico analysis and in vitro experiments. Overall, our study constructs a novel molecular model to further guide the clinical classification and individualized treatment of NB.