Differential gene and protein expression in primary breast malignancies and their lymph node metastases as revealed by combined cDNA microarray and tissue microarray analysis

BACKGROUND. Metastatic disease is a major adverse prognostic factor in breast carcinoma. Lymph node metastases often represent the first step in the metastatic process. METHODS. To gain insight into the molecular events that underlie breast carcinoma metastasis, the authors compared gene expression profiles, obtained by cDNA microarray analysis, of nine matched primary tumors and metastases after screening for enrichment of tumor cells. Statistical analysis identified genes that are expressed at elevated or decreased levels in metastases relative to the corresponding primary tumors. Multidimensional scaling analysis indicated that in terms of expression levels, primary tumors were tightly clustered, whereas metastases exhibited a greater spread; this finding points to the more heterogeneous nature of metastases. Among the differentially expressed entities were the invasion- and tissue modeling-related genes IGFBP5, fibronectin, and MMP2, the cell cycle regulatory gene cyclin D1; other genes, such as enolase 2; and an expressed sequence tag similar to angiopoietin 1. To validate and extend these initial findings, the authors constructed a tissue microarray consisting of 100 primary malignancies paired with their lymph node metastases. Antibodies for the IGFBP-5, fibronectin, MMP-2, cyclin D1, and MDM-2 proteins were used to stain tissue array sections. RESULTS. Consistent with microarray data, statistically significant overexpression of IGFBP-5, down-regulation of cyclin D1, and unchanged MDM-2 levels were observed in metastatic tumor cells. Nonetheless, although fibronectin and MMP2 mRNA expression levels were decreased in many metastasis specimens, expression levels of the corresponding proteins in the extracellulay matrix were elevated in most metastases. Decreased expression of fibronectin and MMP2 in lymph node metastases was further confirmed by real-time polymerase chain reaction assays performed on five additional specimen pairs. CONCLUSIONS. The results of the current study suggest that extracellular matrix protein expression and nuclear gene expression are associated via a negative-feedback regulatory mechanism. Therefore, gene expression profiling and tissue array validation should be combined to elucidate molecular events associated with the metastatic process. (C) 2004 American Cancer Society.