Postmortem vs. neoplastic gene expression: Clues to cancer development and therapy.

Organismal death does not immediately end gene expression. Studies of postmortem gene expression in zebrafish and mice and in the myocardium, liver, prostate, pericardial fluid, and blood of human cadavers have identified genes whose expression is increased after organismal death. Cancer can be considered a form of "un-death" since excessively proliferating cells are typically unusually resistant to apoptosis (programmed cell death), and are subject to strong selective pressure for "uncontrolled life." The changes in gene expression observed in organismal death, particularly in mammals (mice and humans), can be compared to that observed in human neoplasia, and the comparison of these expression patterns can inform us about human cancer. Here we present a hypothesis based on the following three tenets: (a) there will be distinct and opposing patterns of gene expression between the postmortem state and cancer with respect to key physiological outputs such as growth, apoptosis, invasion, and prognosis; (b) cancer cells considered more aggressive (e.g., derived from a metastasis and/or resistant to agents that suppress growth or induce apoptosis) will exhibit expression of relevant genes more unlike that of the postmortem condition while less aggressive neoplastic cells will exhibit gene expression more similar to the postmortem condition; and (c) targeting gene expression in cancer to produce a more postmortem-like pattern will promote less tumorigenic and less aggressive cell phenotypes. To evaluate components (a) and (b) of our hypothesis, we focus on previously published gene expression data from colorectal cancer (CRC) and colonic adenoma cells and compare that to postmortem expression data. This preliminary analysis in general supports our hypothesis, with more aggressive neoplastic cell types exhibiting gene expression patterns most unlike that found in the postmortem condition; this suggests that cancer and the postmortem condition represent opposing ends of a gene expression spectrum in the balance between life and death. Subsequently, we discuss the possibilities for further testing of the hypothesis, particularly for part (c), and we also discuss the possible implications of the hypothesis for cancer therapeutics.