Characterization of a gene expression biomarker predictive of hypoxia-inducible factor-1 modulation.

Abstract

We describe here a gene expression biomarker that can accurately identify chemicals and genetic conditions that perturb hypoxia-inducible factor-1 (HIF-1), a transcription factor critical for the cellular response to hypoxia involved in the pathophysiology of cancer, inflammation, and ischemia. The HIF-1 biomarker genes were identified from transcript profiles in a variety of human cell lines after hypoxia as well as genetic knockdown of the HIF1A gene. The HIF-1 biomarker of 122 genes was found to be enriched for: genes bound by HIF-1 using ChIP-Seq, metabolic pathways regulated by HIF-1 (e.g., gluconeogenesis). Using Ingenuity Pathway Analysis, HIF-1 was the top predicted regulator of the genes. The biomarker could identify activation of HIF-1 by overexpression of the HIF1A gene or suppression of the negative regulator, von-Hipple Lindau (VHL) gene; suppression of HIF-1 occurred by inhibition of the expression of HIF1A or the heterodimer partner aryl hydrocarbon receptor nuclear translocator (ARNT). The biomarker was specific for HIF-1 activation, as activation or suppression of the HIF-2 gene was not detected by the biomarker. Using a reference chemical set with 180 positives and 45 negatives, the balanced accuracy of the HIF-1 biomarker to identify activation was 94.1%. An in-silico screen of a gene expression compendium identified hundreds of potential activators. Eleven of these were selected for verification and all were found to activate HIF-1 biomarker genes in wild-type but not HIF1A-null cells. The HIF-1 biomarker will be a useful tool to identify environmentally relevant chemicals that affect HIF-1 in high-throughput transcriptomics screening studies.