Whole genomic expression profile analysis of early and advanced staged colorectal cancer biopsy samples: searching for candidate pathways involved in the carcinogenesis and pathomechanism

Background: The high-density oligonucleotide microarray analysis gives an opportunity for studying the genetic and gene expression background of the diseases, but the biological and functional interpretation of the microarray results – hundreds of differently expressed, possible diagnostic and therapeutic candidate genes – was not worked out until present days.

Aims: our aims were to develop diagnostic mRNA expression patterns for indentification of different staged colorectal carcinomas and to find altered biological pathways for explanation of the pathomechanism of the CRC using Pathway Assist software.

Materials and Methods: Total RNA was extracted from frozen colonic biopsy specimens of 12 patients with colorectal carcinoma (CRC), 12 with tubulovillous or villous ademas and 8 healthy normal controls. The mRNA fraction from the extracted total RNA was amplified by T7 RNA amplification method. Biotin-labelled cRNA probes were synthesized and fragmented. The genome-wide mRNA expression profile was evaluated by GeneChip U133 Plus 2.0 microarrays (Affymetrix Inc., US, over 47 000 transcripts and variants) and GeneChip Scanner 3000. T-test, self-organizing map clustering were done by the Data Mining Tool software. Pathway analysis was done by the Pathway Assist 2.53 software.

Results: In Dukes C/D stage colorectal adenocarcinoma, 2 clusters of genes (544 and 689 genes) were found to be down-, and 1 cluster of genes (521 genes) was found to be upregulated compared to normal sample. In Dukes B stage CRC one upregulated gene cluster (69 genes) and one down-regulated gene cluster (139 genes) was found. Genes encoding proteins without unknown functions were removed from the pathway analysis. Pathways were built and 22 selected biochemical function network were graphically visualized. 158 upregulated and 137 downregulated CRC associated genes were classified into the selected biological pathways.

Conclusions: Pathway analysis of whole genomic expression profiles is a powerful method for biological interpretation of a large amount of data and for finding the functional connections between the up- and downregulated genes involved in CRC carcinogenesis and pathomechanism.