Deciphering Molecular Circuitry Using High-Density DNA Arrays

The immense amount of sequence data available from expressed sequence tag (EST) databases (1,2), together with the development of technologies for the highly parallel analysis of gene expression (3–5) have created the opportunity to interrogate biochemical pathways and gene function on an unprecedented scale. We describe here a set of high-density DNA arrays containing oligonucleotides complementary to more than 6,500 human EST’s. These arrays were used to generate normal and breast cancer specific gene expression profiles. More than 1,500 expressed genes were detected in both cell types examined with 85% of all gene expression observed in the range of 1–50 copies per cell. Over 300 genes demonstrated significantly different levels of expression between normal and transformed breast cells. Increased mRNA levels were observed for the Her2/neu oncogene and genes involved in its signal transduction pathway, including Grb-7, Ras, Raf, Mek and ERK. In addition, a simple categorization of the expression changes revealed patterns characteristic of loss of wild-type p53 function. Genotyping of the p53 locus using a DNA re-sequencing array revealed inactivating mutations in the p53 DNA binding domain and loss of heterozygosity. These data demonstrate a general array-hybridization-based approach to deciphering biochemical pathways and generating testable hypotheses concerning the mechanisms of cell growth and differentiation.