Hepatic and lung metastases are the leading causes of mortality and major indicators of aggressiveness in colorectal cancer. The underlying molecular mechanisms contributing to the development of metastasis are still unclear. Here, we designed a novel approach to explore gene expression profiles associated with metastasis in human colorectal cancer (hCRC). A series of ten isogenic tumors from three different hCRC models were orthotopically implanted into nude mice. In these series, we analyzed the contribution of dynamic heterogeneity, independently of any intrinsic gene expression program predictive of metastasis. When screened for the presence of disseminated tumor cells in the lung and liver, as the most common host tissues for hCRC metastases, both high- and low-metastatic efficient tumors were found among these isogenic orthotopic series. The metastasis-specific cDNA macroarray analysis of 96 genes, in both tumor populations for each of the three hCRC models, characterized a common differential gene expression within a small group of genes. Our results suggest that, independently of a gene expression profile predictive of metastasis, the progressive acquisition of additional alterations occurs during hCRC tumorigenesis. This dynamic process might determine tumor progression, namely the metastasis dissemination.
Copyright 2007 S. Karger AG, Basel.