DEVELOPMENT OF ADVANCED TECHNOLOGIES FOR COMPLETE GENOMIC AND PROTEOMIC CHARACTERIZATION OF QUANTIZED HUMAN TUMOR CELLS

Background: Cancers are composed of heterogeneous cell populations. Objective: We propose to study individual cells from freshly excised glioblastomas to determine whether there are quantized populations in tumors. From these studies, we will identify disease-perturbed networks, be able to identify driver mutations in cancer, identify candidate biomarker proteins, and assess responses of quantized cancer population to drugs. We will then employ newly developed SRM (selected reaction monitoring) targeted proteomics techniques to search for these candidate biomarkers in blood and later validate them against 100 bloods from glioblastoma patients. We are looking for biomarkers that will be able to stratify disease (identify different types of glioblastoma), assess progression, enable early detection of reoccurrence, and finally provide prognostic insights. This work will be achieved by the following aims.Specific Aims: (1) Isolate up to 1000 cells from each of five human glioblastomas and quantify initially 500 different transcripts from each cell (transcription factors, CD [cluster of differentiation] molecules, relevant signal transduction pathways, etc.). Determine whether computational analyses can classify these cells into discrete quantized cell types. (2) Sort the disassociated tumor cells from several glioblastomas into their quantized cell populations using cell sorting/CD antibodies to each quantized cell type for genomic and proteomic analyses, functional analyses, and establish primary cell lines. (3) Assess 20-40 candidate blood biomarkers in the bloods of 100 glioblastoma patients with regard to their ability to stratify disease, assess disease progression, and predict at an early stage the reoccurrence of the glioblastoma (early detection). Eventually we will use these biomarkers to assess the effectiveness of therapy. (4) Ten to 20 cells from each major quantized glioblastoma cell type from two patients will be used to determine the complete genome sequences. We will also determine the normal genome sequences of each patient and their family members to enable the 70% error correction. The mutations will be analyzed against quantitative changes in the transcriptomes, miRNAomes, and proteomes and against the relevant biological networks. (5) Analyze the quantized cell populations for their responses (transcriptome, miRNAome, etc.) to the perturbations of key glioblastoma-relevant molecules (e.g., nodal points in networks) by RNAi perturbations as well as their responses to drugs and natural ligands.Study Design: (1) We will analyze single cells from two tumors and from normal brain tissue resected along with each tumor. We expect to see astrocytes, endothelial cells, stromal cells, etc., and these will be differentiated by transcriptome analyses. On the single-cell analyses of the glioblastoma cell line, we have been able to demonstrate distinct cell-surface molecules that could be used with cell sorting to separate each of the three quantized cell types, and then analyze each for genome sequence and at the mRNA, miRNA, and targeted protein levels to identify disease-perturbed networks. (2) The biological experiments using natural ligands, drugs, and selected RNAis are routine on tumor tissues and should present no problem with regard to the quantized cell populations. (3) The genomic sequencing of the families and of the quantized cell populations are experiments that our collaborator Complete Genomics has already carried out (including a complete genome sequence from just 10 cells). (4) The SRM assays and proteomic analyses of blood have been pioneered at Institute for Systems Biology. The comparisons of normal bloods and glioblastoma-stratified bloods should provide insights into biomarkers that can be used for the various aspects of diagnosis, including stratification, assessment of progression, early identification of reoccurrence, etc.Innovation: We are using four emerging or relatively new technologies or strategies to attack fundamental problems of cancer biology: (1) single-cell analyses of cancer cells to assess population heterogeneity; (2) complete genome sequence analyses of families to provide highly accurate non-cancer and cancer sequences for comparison against the genomes of cancer cells; (3) the complete genomic sequence being determined from as few as 10 cells if necessary (if the quantized populations are small); and (4) the use of SRM proteomics assays to identify tissue and blood biomarkers for diagnosis and cancer mechanisms.Impact: The direct benefits for patients include the ability to stratify cancers into their discrete types to enable impedance matching against effective drugs. Stratification will be achieved by (1) analysis of quantized cell populations, (2) identification of unique combinations of disease-perturbed networks, and (3) identification of unique nodes in relevant networks for each stratified disease type that can be detected in the blood by SRM assays. The important outcome of this proposal is that all of these approaches can be applied to almost any cancer.