RESUMO
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.