Four distinct epithelial cell types can be sorted by flow cytometry into single populations from normal and cancerous prostate tissues. They include luminal cells, basal cells, luminal-like cancer cells, and basal cell-like cancer cells. The luminal cell-like cancer cells are considered by be less malignant than the basal cell-like cancer cells based on their gene expression. Luminal cell-like cancer cells is the predominant population in localized tumors while the basal cell-0like cancer cells is a predominant population in some metastases. Sorted cells allow us to construct cell-type specific cDNA libraries for the identification of candidate cancer-associated genes in tumorigenesis and cancer progression by the powerful techniques of genomics. The four cell type libraries are augmented by libraries of prostate cancer xenografts of different virulence as represented by histology and growth characteristics. The new gene markers discovered will be used to investigate organ specificity in cancer metastasis, the interaction between tumor cells and bone microenvironment, androgen and growth factor dependence outlined in Projects II and IV. New serine proteases and useful cell surface molecules will be used in Project III. This project will generate a cell-type expression database useful to all areas of research in prostate biology. We will also attempt to completely define the transcriptosome of prostate tumor cells using a powerful new approach termed multiple sequence signature profiles. This procedure can place 500,000 cDNA sequences each on individual beads and sequence them for 16-20 bases. We will profile LNCaP cell lines, both androgen stimulated and androgen deprived, as well as LNCaP variant that is androgen independent. In each of these three cases we will determine about 1,000,000 sequences-more than enough to virtually define the entire transcriptosome.
Showing the most recent 10 out of 157 publications
