Background: Hypoxia is well known to confer a more aggressive phenotype in neuroblastoma. We sought to integrate gene expression data from diagnostic tumors and hypoxic neuroblastoma cell lines to identify genes that correlate with poor patient outcome and are involved in the hypoxia response which may represent novel therapeutic targets.
Methods: Linear models identified DEG from tumors of two cohorts of 567 total neuroblastoma patients who survived compared to those that did not. Mixed linear models identified differentially expressed genes (DEG) between 12 neuroblastoma cell lines grown in hypoxia or normoxia. qPCR validated expression differences in multiple neuroblastoma cell lines. shRNA was used to knock down expression of HK2, the most overexpressed of the identified genes in hypoxia.
Results: Eight genes, SLCO4A1, ENO1, HK2, PGK1, MTFP1, HILPDA, VKORC1, and HIST1H1C, were differentially expressed in all datasets and significantly associated with outcome on Kaplan-Meier analysis. Six of these eight had hypoxia inducible factor (HIF) binding sites on CHIP-seq analysis verifying that they are indeed hypoxia targets. Three of these genes are part of the glycolytic pathway and three more are directly involved in cellular metabolism. We functionally validated our computational findings and showed that six of the eight genes are significantly up regulated in five independent neuroblastoma derived cell lines. shRNA knockdown of the HK2 gene leads to decreased cell growth by MTT and a G0/G1 arrest in flow cytometry analysis in the La1-55n cell line and may represent an important potential drug target for this disease.
Conclusions: Our findings highlight the utility of combining diverse data sets to identify key genetic drivers of disease that may be otherwise be overlooked as we have identified a compact and robustly validated group of eight genes which may represent new therapeutic and prognostic targets for children with neuroblastoma.