Abstract 119: Pan-cancer analysis of HIF pathways defines a robust molecular signature that reflects tissue and cellular hypoxia in bulk and single-cell RNA-seq analyses

Abstract

Abstract Low levels of tissue oxygen, or intra-tumor hypoxia, is a common feature of many solid cancers in which dysregulated cell proliferation often outstrips the ability to develop new functional blood vessels. Furthermore, hypoxia pathways may also be activated in response to mutation in key tumor suppressor genes or oncogenes. Importantly, in many types of cancer, activation of hypoxia pathways is associated with a clinically aggressive phenotype as well as with resistance to therapy. Indeed, hypoxia drives many of the “hallmarks” of cancer, including cell proliferation, apoptosis, metabolism, immune responses, genomic instability, vascularization, invasion and metastasis. However, the full-range of genes activated by hypoxia varies greatly between cell-types and for many this remains uncharacterized. The key mediator of the cellular transcriptional response to hypoxia is the HIF (hypoxia-inducible factor) family of transcription factors. Here, we generate pan-genomic datasets encompassing HIF ChIP-seq and RNA-seq across cancer cell lines from diverse common tumor types. We then employ integrated analysis of our 72 transcriptome and whole-genome-binding datasets to define a core set of HIF-target genes that are conserved across multiple cell types. We show that these genes can be used to identify HIF activation and tumor hypoxia in bulk RNA-seq analysis of solid tumors, and that this can be applied to large tumor databases (TCGA). We then leverage this analysis to identify tumor-type-specific HIF-associated genes in 32 different cancer types, including those lacking cell-culture models. We further show that these conserved HIF-target genes can be applied to single cell RNA-seq datasets from solid tumors and used to examine intra-tumor heterogeneity in HIF-pathway activation within individual cancer samples. Importantly, using a panel of genes reduces the phenomenon of “drop-out” typically observed in gene-level expression in these datasets. Lastly, applying this gene signature to single cell data allows the hypoxia response to be deconvoluted across diverse cell-types in heterogenous tumor samples. This includes not only the identification of HIF-activity in cancer cells, but also in non-cancer cells within the tumor, thereby allowing the effects of intra-tumor hypoxia on stromal responses (such as tumor angiogenesis) to be studied in vivo in solid tumors. Citation Format: Olivia Lombardi, Ran Li, Silvia Halim, Peter J. Ratcliffe, David R. Mole. Pan-cancer analysis of HIF pathways defines a robust molecular signature that reflects tissue and cellular hypoxia in bulk and single-cell RNA-seq analyses [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 119.