Identification of proteins for breast cancer risk: an integrative epidemiologic and genomic study

Project Summary Genome-wide association studies (GWAS) have identified over 200 genetic loci for breast cancer. To translate these findings into an improved understanding of cancer biology, and thus, disease prevention and treatment, it is essential to identify the target genes of these loci. However, causal genes (and the underlying

biological mechanisms) remain unknown for most of the GWAS-identified breast cancer risk loci despite recent publications of transcriptome-wide association studies that have discovered a few putative causal genes. Proteins, the final gene products, are key working molecules for cellular functions. Thus, proteins have a more

direct impact on disease risk than mRNAs. In addition, protein levels and mRNA levels are only moderately correlated, with the median correlation being only 0.46, and for many genes, the correlation is close to zero. Therefore, compared with investigating RNAs, directly investigating proteins may be more likely to reveal

causal genes and underlying biological mechanisms. We propose a novel proteogenomics study to identify putative target proteins for breast cancer by capitalizing on multiple existing unique resources, 1) normal fresh- frozen breast tissue samples from over 5,000 cancer-free women in the Susan G. Komen Tissue Bank (KTB);

2) proteomics and genomics data in 125 breast cancer tissue samples in the Clinical Proteomic Tumor Analysis Consortium (CPTAC); and 3) GWAS data of ~ 427,000 breast cancer cases and controls from three large consortia. We will conduct a well-powered proteome-wide association study (PWAS) by integrating

genomics and proteomics data in breast tissues in three racial groups (Aim 1). We will prospectively validate the promising breast cancer risk proteins in pre-diagnostic normal breast tissues in a nested case-control study including 120 incidence breast cancer cases and 480 individually matched controls (Aim 2). We will evaluate

the biological functions of the top breast cancer risk proteins by conducting a series of bioinformatics analyses and in vitro functional assays (Aim 3). The proposed study will be conducted by an interdisciplinary collaborative team of experts in breast cancer, epidemiology, genomics, proteomics, bioinformatics, and

biostatistics, with almost 20-years of collaboration. This proposed project is extremely cost-efficient because breast tissue samples and the genomics data for Aim 1 will be available at no cost to this proposed study. The proposed study is highly innovative as it is the first prospective breast tissue-based proteomics study and also

the first proteogenomics study by integrating genomics and proteomics data. Our study will be highly significant and impactful by 1) unveiling the target proteins for future mechanistic research; 2) improving our understanding of the genetic and biological basis for breast cancer; and 3) having important implications for

disease prevention and treatment for this common malignancy.