Gene Expression Patterns in Human Tumors Identified Using Transcript Sequencing

点击翻译按钮获取中文摘要

Gene expression provides a snapshot of the cellular changes that promote tumor malignancy. Quantitative gene expression analysis, especially as implemented by DNA microarrays, has proven to be an extremely valuable tool for cancer genome characterization, and has lead to the development of new genomic-based clinical tests. Our own experience with DNA microarrays to study gene expression patterns for breast, head & neck, and lung cancers has lead to the identification of novel subtypes of tumors with distinct patient outcomes and has identified new tumor suppressor genes. In the pilot phase of The Cancer Genome Atlas (TCGA) project, multiple platforms were used including tools to study gene expression (our role), tumor genomic DNA copy number alterations, SNP genotypes, DNA methylation and gene mutational analyses. Our collaborative efforts identified new tumor subtypes of glioblastoma and painted an integrated picture linking mutations to copy number changes to expression patterns, which identified biologically distinct subtypes of disease with differences in patient outcomes. For the second phase of TCGA project, we propose to continue to perform quantitative gene expression profiling of all protein-coding genes, non-protein coding mRNAs(ncRNAs) and microRNAs, on -2000 tumors per year. This approach has proven to be one of the most informative and comprehensive cancer genome characterization tools available. In addition, we propose to generate global chromatin organization profiles of cancer to identify regions of "open" chromatin domains (nucleosome-depleted regions). We will use FAIRE (Formaldehyde-Assisted isolation of Regulatory Elements), a simple, low-cost method amenable to use on small quantities of solid tissue, coupled to next-generation DNA sequencing. Since the function of most histone modifications and chromatin remodeling activities is to regulate nucleosome occupancy, FAIRE effectively summarizes the functional output of such epigenetic mechanisms in a single robust assay. Lastly, we propose to perform integrated analyses of transcript levels with chromatin stoicture to map important regulatory elements, which can be distant to the transcript(s) that they regulate. Our study of genome-wide transcript regulation with chromatin organization will provide a critical portrait of the cancer genome that can be integrated with (and indeed can sometimes generate) other important data, including mutations and copy number events.