Background: Breast cancer, like that of other human cancers, consists of malignant epithelial cells that are interwoven with normal stromal cells such as fibroblasts and myofibroblasts. In vitro-based and animal-based studies have suggested that stromal cells within a developing tumor aid in the maintenance and growth of malignant breast epithelial cells. The molecular mechanisms associated with stromalepithelial interactions in human breast cancer are largely unknown, and the elucidation of such mechanisms may provide for novel interventional strategies. Recently, we successfully combined laser capture microdissection with DNA microarray technologies to comprehensively interrogate the malignant breast epithelial cells that constitute the various stages of breast cancer progression. This approach is unique from other microarray-based studies of breast cancer in that it generated cell-specific, rather than tissue-specific gene expression profiles from patient-matched normal and malignant breast epithelium. With this approach, we have identified novel gene expression patterns, and such patterns have led to the generation of novel hypotheses regarding breast cancer development. Adaptation of this approach to the interrogation of tumor stromal cells is a nature extension of these technologies and will generate novel stromal cell-specific gene expression profiles. Such profiles in conjunction with the corresponding epithelial-specific profiles will provide researchers with a unique cell specific breast cancer gene expression database from which to generate novel hypotheses regarding breast cancer epithelial-stromal interactions.

Objective/Hypothesis: The objective is to generate, for the first time, gene expression profiles (fingerprints) of stromal cells that are associated with the pre-invasive and invasive stages of breast cancer progression and compare such profiles to that of normal breast stroma.

Specific Aims: (1) To generate stromal-specific gene expression profiles of tumor-associated stromal cells and normal breast stromal cells from 36 different breast cancer specimens. (2) To perform the comparative gene expression analysis of these stromal-specific profiles to the epithelial-specific profiles derived from the same cohort of patients.

Study Design: We have collected normal and cancerous breast tissue specimens from 36 women with breast cancer. We will specifically isolate stromal cells associated with the different stages of breast cancer progression and normal stromal cells (patient-matched) through the use of laser capture microdissection. Genetic material (RNA) from each sample will be amplified, labeled, and used to interrogate a gene microarray that contains 12,000 different genes. We will perform comparative analysis of the gene expression profiles of the normal stroma to that of the tumor-associated stroma, and the resulting data will be correlated with both pathological stage and tumor grade.

Relevance: Our knowledge base of the gene expression alterations that occur within human breast cancer stromal cells is virtually nonexistent. The tumor stromal-specific information gained from this study will create a novel database that will serve as invaluable resource for future studies aimed at elucidating stromal-epithelial interactions in human breast cancer. It is anticipated that this study will generate novel hypotheses regarding the role of stroma in the initiation, maintenance, and progression of breast cancer.