Creation of the Human Cancer Metabolome Atlas

Project Summary/Abstract Metabolic reprogramming is a hallmark of cancer, prompting molecular mechanisms to promote tumor survival, growth, and resistance to treatment. Therapeutic targeting of tumor metabolism therefore is believed to be a promising avenue of cancer treatment. To further the development of metabolism-targeted cancer therapy,

more information is needed concerning the heterogeneity of tumor metabolism and how tumor type, stage, and other tumor/patient characteristics determine the metabotype of tumors. Although publicly available resources exist for genomics and proteomics data, such as The Cancer Genome Atlas (TCGA) and the Human Proteome

Atlas (HPA), no such database exists for the cancer metabolome. Resources like TCGA and HPA have become cornerstones of cancer research, making the lack of a cancer metabolome resource a glaring omission, particularly since metabolomics is considered to be the omics closest to phenotype. We propose to

create the Human Cancer Metabolome Atlas (HCMA) by collecting metabolomics data on numerous patient- derived tumors and making an open-access resource to the cancer community to better understand the role of metabolism in cancers. In Aim 1,we will analyze a large-scale collection (>2000) patient-derived xenografts

(PDXs) using mass spectrometry-based metabolomics. We will obtain these PDXs from NCI’s patient derived model repository (PDMR), Charles River Laboratory, and Crown Bioscience. Each PDX model has already been characterized by RNA-sequencing, whole exome sequencing, and comes with a wealth of patient

information/metadata. In Aim 2, we will evaluate how metabotypes of PDXs vary based on gene expression patterns, presence of metastatic disease, and survival outcomes. Using a combination of univariate, multivariate, and pathway analyses, we will uncover key roles of cancer metabolism in these processes and

which aspects of cancer metabolism contribute to heterogeneity across patients. Finally, in Aim 3 we will create a public resource for the HCMA using the National Metabolomics Data Repository (NMDR, also known as Metabolomics Workbench). Raw and processed metabolomics data will be uploaded to this resource along

with associated patient metadata and existing genetic data to facilitate data mining and discovery efforts to be performed by outside investigators. Existing data analysis modules of the NMDR including univariate analysis, multivariate analysis, pathway analysis, and metabolite class analysis will be integrated with the HCMA to

make the metabolomics data accessible and easily analyzable by other investigators. This resource will continually be built upon in future proposals to expand the HCMA and provide a powerful resource for new discoveries in cancer biology and therapeutics.