By using hierarchical clustering of global gene expression data, we were the first to identify five subtypes of breast tumors characterized by specific expression patterns. Our existing work has generated a rich and unique resource and knowledgebase from which to continue research. We will apply state-of-the-art microarray tools, novel functional technologies and associated statistical and bioinformatics methods to profile breast tumors at the DNA, RNA and protein level in four subprojects:
1. Validate th e 5 subtypes and their clinical correlates in 2 more breast cancer cohorts, focusing particularly on patients with earlier disease stages. Identify a minimal set of informative predictor genes for the classification, and carry out a comparative analysis o f microarray-based and quantitative RT-PCR methods for these genes.
2. Identify gene expression profiles that can predict presence of micro-metastases in early breast cancer, using tumors from a unique set of 900 breast cancer patients, for which researc hers at DNR have developed methods for detecting single tumor cells in bone marrow.
3. Identify and validate genes predictive of chemotherapy response, based on gene expression and array-based copy number (aCGH) profiling of tumor samples from two unique prospectve cohorts designed to focus on chemotherapy response prediction in breast cancer. The profiles found in vivo will be tested functionally using RNA interference in breast cancer cell lines.
4. Identify gene targets undergoing somatic genetic muta tions in breast cancer using a novel NMD-aCGH method. Combination of the NMD microarray data on mutated candidate genes with the aCGH data on deletions should enable us to identify classical “two-hit” tumor suppressor genes in breast cancer. The NMD-CGH m icroarray analysis will be carried out in breast cancer cell lines, mutated candidate genes will be screened in primary breast tumors.
An application for EURYI Award is being submitted.