E-GEOD-5206 - Transcription profiling of human and mouse colon cancers reveals large-scale deployment of embryonic gene programming in human and murine colon cancer: a new target for intervention.
Submitted on 30 June 2006, released on 13 June 2008, last updated on 10 June 2011
Goal of the experiment: To identify and understand the overall transcriptional programming of human colorectal cancer tumors by evaluating gene expression profiles of tumors from four murine models, and comparing and contrasting these to the developing stages of the mouse embryonic colon. Experiment description: Colorectal cancer results from multiple genetic and epigenetic events that produce variable histologies and clinical outcomes. To identify gene regulatory programs that underlie colon tumorigenesis, we profiled gene expression in 39 mouse colon tumors from four independent mouse models and compared this to mouse colon embryonic development, as well as with 100 human colon carcinomas. Here, we report a striking recapitulation of embryonic patterns of gene expression in both mouse and human colon tumors. All four of the mouse colon tumor models exhibited large-scale activation of embryonic gene expression signatures. The two nuclear beta-catenin-positive mouse tumors (azoxymethane-treated [AOM] and ApcMin/+), exhibited strong activation of genes characteristic of those expressed in the earliest embryonic stages, while tumors from two other models (Smad3-/- and Tgfb1-/- x Rag2-/-) exhibited lower activation of early stage-specific genes but substantial expression of general embryonic colon genes. Human colon cancer cases over-expressed genes characteristic of both early and late embryonic stages. Examining tumor gene expression through the lens of development has revealed an extensive network of therapeutic targets for cancer control. Experiment Overall Design: Transcriptional signatures across human colorectal carcinomas and normal human colon tissue samples were evaluated by referencing all the samples to the median gene expression value across all the samples. A second complementary strategy compared the gene expression levels of the tumors to those of normal adult colon. Experiment Overall Design: Values were transformed from a log base 2 to linear values. Each measurement was divided by the 50.0th percentile of all measurements in that sample. Each gene was divided by the median of its measurements in all samples. If the median of the raw values was below 10 then each measurement for that gene was divided by 10 if the numerator was above 10, otherwise the measurement was thrown out.
transcription profiling by array, unknown experiment type