Extracellular proteases of the matrix metalloproteinase (MMP) and serine protease families participate in many aspects of tumour growth and metastasis. Using quantitative real-time RT -PCR analysis, we have undertaken a comprehensive survey of the expression of these enzymes and of their natural inhibitors in 44 cases of human prostate cancer and 23 benign prostate specimens. We found increased expression of MMP10, 15, 24, 25 and 26, urokinase plasminogen activator-receptor (uPAR) and plasminogen activator inhibitor-1 (PAI1), and the newly characterised serine proteases hepsin and matriptase-1 (MTSP1) in malignant tissue compared to benign prostate tissue. In contrast, there was significantly decreased expression of MMP2 and MMP23, maspin, and the protease inhibitors tissue inhibitor of metalloproteinase 3 (TIMP3), TIMP4 and RECK (reversion-inducing cysteine-rich protein with Kazal motifs) in the cancer specimens. The expression of MMP15 and MMP26 correlated positively with Gleason score, whereas TIMP3, TIMP4 and RECK expression correlated negatively with Gleason score. The cellular localisation of the expression of the deregulated genes was evaluated using primary malignant epithelial and stromal cell cultures derived from radical prostatectomy specimens. MMP10 and 25, hepsin, MTSP1 and maspin showed predominantly epithelial expression, whereas TIMP 3 and 4, RECK, MMP2 and 23, uPAR and PAI1 were produced primarily by stromal cells. These data provide the first comprehensive and quantitative analysis of the expression and localisation of MMPs and their inhibitors in human prostate cancer, leading to the identification of several genes involved in proteolysis as potential prognostic indicators, in particular hepsin, MTSP1, MMP26, PAI1, uPAR, MMP15, TIMP3, TIMP4, maspin and RECK.

No abstract

No abstract

OBJECTIVES To describe our experience with the optimization and validation of laser‐capture microdissection (LCM) for biomarker analysis in prostate tissues. As LCM allows the separation of benign and malignant epithelial structures and stromal elements, it not only allows identification of the source of the biomarker, but might also accentuate gene or protein expression changes by reducing contamination by other cellular elements. MATERIALS AND METHODS In all, 19 fresh‐frozen prostate tissue samples were subjected to LCM, with the cDNA being analysed using quantitative polymerase chain reaction for several genes, to identify the optimum number of cells for capture, as well as gene markers assessing for the purity of the captured cells. The localization was further confirmed by in situ hybridization. RESULTS Prostate‐specific antigen (PSA) and cytokeratin 8, were expressed solely by epithelial cells, whereas hepatocyte growth factor (HGF) and tissue inhibitor of metalloproteinases‐3 (TIMP3) were expressed only by stromal cells, and the levels of transcripts of these genes were unaltered between benign and malignant tissues. CONCLUSIONS These data suggest that PSA, cytokeratin 8, HGF and TIMP3 are reliable gene markers of purity of epithelial and stromal compartments for LCM of prostate tumours. Although this technique is not new and is increasingly used in laboratories, it needs optimization and stringent validation criteria before data analysis. This applies to all tissue types subjected to LCM.