Advances in genomic technology have driven the collection of unprecedented amounts of data from human cancers. Critical challenges include extracting biological insights from this data and then translating these findings into improved patient care. Toward those ends, I performed a meta-analysis on 98 gene expression-linked cancer survival studies. In these studies, investigators measured gene expression in primary tumors surgically excised from patients and then correlated that data with clinical information on the patient's treatment and outcome. While data from over 20,000 patients with 20 distinct tumor types are available in public repositories, no comprehensive pan-cancer analysis has previously been reported. My analysis resulted in the discovery that the up-regulation or down regulation of hundreds of genes is highly correlated with the duration of patient survival across many common human cancers. This gene set provides unique insight into a poorly-understood facet of cancer biology: while the molecular alterations that distinguish tumor cells from normal cells have become increasingly well-characterized, the differences between highly-aggressive tumors and benign, indolent tumors remain largely unknown. My research group will utilize this bioinformatics pipeline as a discovery engine to identify and characterize the genes specifically responsible for cancer mortality.
In Aim 1, I describe several single-gene assays and targeted screens that will assess the function of genes whose expression is associated with death from cancer. These assays will determine which of these genes are bona fide oncogenes and which promote tumor metastasis.
In Aim 2, I describe experiments to assess the functions of genes whose expression correlates with prolonged survival in cancer patients. These genes could represent novel tumor suppressors or could otherwise inhibit tumor cell dissemination.
In Aim 3, I propose several mechanistic experiments to elucidate the biology that connects the expression or repression of the genes identified in this study with patient survival. These functional studies will determine the effects of these genes on known cancer-related pathways, including the epithelial-mesenchymal transition, a developmental program hypothesized to play a crucial role in tumor metastasis. Lastly, in Aim 4, I describe an expanded set of molecular data linked to cancer survival that will be analyzed in order to create a mortality map of genomic features that drive or suppress tumor progression. This broader analysis will be used to identify new cancer-associated genes for future functional studies. Collectively, these Aims will greatly expand our knowledge of the features that differentiate fatal and non-fatal human tumors. Additionally, this work will provide an abundance of data with the potential to improve the stratification of cancer patient risk, and will identify novel targets for drug development to specifically inhibit the growth of the most aggressive cancers.
This project leverages clinical data from thousands of patients to detect the genetic differences between fatal and non-fatal human tumors. By studying the genes that are responsible for these differences, this project will vastly advance our understanding of why certain cancers recur after surgery while other tumors can be effectively cured. The results from this work will enhance the ability of doctors to determine how aggressively to treat each cancer patient.
| Giuliano, Christopher J; Lin, Ann; Smith, Joan C et al. (2018) MELK expression correlates with tumor mitotic activity but is not required for cancer growth. Elife 7: |
| Lin, Ann; Giuliano, Christopher J; Sayles, Nicole M et al. (2017) CRISPR/Cas9 mutagenesis invalidates a putative cancer dependency targeted in on-going clinical trials. Elife 6: |
