Gene expression profiling of renal cell carcinoma: a DNA macroarray analysis

OBJECTIVES To examine differences in gene expression levels between renal cell carcinoma (RCC) tissue and 'normal' appearing renal tissue using a commercially available DNA macroarray. MATERIALS AND METHODS Tissue was obtained from 47 consecutive radical nephrectomies, 29 of which were eligible. DNA macroarrays were analysed on the tumour and normal-appearing control tissue to measure the expression of 1185 cancer-related genes. The group of samples was also stratified according to the presence or absence of granular cells and according to tumour grade. Quantitative real-time polymerase-chain reaction (PCR) was also performed on seven key genes present on the macroarray. RESULTS In all, 444 genes were over-expressed and 33 genes were under-expressed. Using selection criteria reduced the list to nine that were significantly over-expressed and 23 that were under-expressed. These significant genes belonged to the families of oncogenes, growth factors, interleukins, receptors, immune system components, cytoskeleton, matrix proteins and intracellular modulators, or they coded for proteins involved in DNA transcription and RNA translation, DNA repair, protein turnover, and metabolism of carbohydrates and lipids. There were differences in gene expression according to the presence or absence of granular cells and according to tumour grade. Using quantitative real-time PCR there was overexpression of epidermal growth factor receptor, c-myc, transforming growth factor-alpha, vascular endothelial growth factor and vimentin, and under-expression of TYRO3 protein tyrosine kinase. The von Hippel-Lindau gene was under-expressed but not significantly. CONCLUSIONS A procedure for collecting and storing fresh renal tissue and subsequent gene expression profiling of RCC and normal renal tissue was established. A commercially available DNA macroarray coupled With the significance analysis of macroarrays allowed the identification of sets of differentially expressed cancer-related genes that were characteristic of RCC, compared with apparently normal renal tissue, and which distinguished among subgroups divided according to tumour grade and histological subtype. Quantitative PCR is important to validate the results of macroarray experiments.