BBB Seminar: Karl-Henning Kalland
Genome-wide analysis of coordinated gene expression in a cell culture model of stepwise EMT and malignant transformation
The loss of epithelial characteristics and the acquisition of a migratory phenotype, referred to as epithelial to mesenchymal transition (EMT), is a crucial event in tumor metastasis. Our group has established an EMT model based on primary prostate epithelial cells (EP156T) which gave rise to cells with a mesenchymal phenotype (EPT1) without malignant transformation (1). To achieve transformed prostate cells, EPT1 cells were kept growing in long-term saturation density cultures to select for cells overriding quiescence.
Following about 30 subconfluent passages and thereafter, foci appeared repeatedly in EPT1 monolayers following three to four weeks in confluent cultures. For comparison, in parallel experiments foci appeared neither in confluent cultures of EPT1 cells of early passage nor in confluent EP156T cells of either low or late passages.
Cells were isolated from the foci of EPT1 monolayers and formed robust colonies in soft agar suggesting full malignant transformation. The transformed cells were named EPT2. In this stepwise transformation model, EPT2 cells showed much higher abilities to proliferate at confluence, migrated better, were less dependent on exogenous growth factors and were more resistant to staurosporine-induced apoptosis than EP156T and EPT1 cells. Gene expression profiling and Western blots showed progressive down-regulation of E-cadherin from EP156T, through EPT1 to EPT2, along with coordinated switches of a number of structural and regulatory gene expression modules of the cytoskeleton, cell junctions, matrix remodeling and membrane receptors.
Epigenetic modifications have been shown to be very critical in gene regulation during carcinogenesis (2), yet the role in EMT remains to be clarified. We have done a genome-wide analysis of histone methylation (H3K4me3 and H3K27me3), DNA methylation (DNAMe) and gene expression profiles in epithelial cells (EP156T) and their progeny mesenchymal cells (EPT1) and transformed cells (EPT2) in our cell culture model of stepwise prostate carcinogenesis. We found that H3K4me3 and H3K27me3 are very clear marks of active or repressed genes, while DNA methylation is only slightly associated with global gene repression in all three cell lines. However, DNA methylation is a strong silencing mark when H3K4me3/DNAMe bivalently marked genes were excluded. Analysis of the transcription activities of genes with bivalent marks showed that H3K4me3/K27me3 is a repressive mark while H3K4me3/DNAMe is an active mark of gene expression. In individual genes, the transcription activities of genes with bivalent marks corresponded to the mark with the strongest signal intensity. Furthermore, quantitative changes of H3K4me3 or H3K27me3 during EMT correlated positively and negatively to activated and repressed transcription, respectively. Finally, a set of critical genes during EMT, including epithelial cell junction genes and the EMT inducers, FGFRs, were identified to show consistent changes concerning epigenetic modification and gene expression. All these findings provide the first blueprint of genome-wide epigenetic modification during EMT in prostate cells.
1. Ke XS, Qu Y, Goldfinger N, Rostad K, Hovland R, Akslen LA, Rotter V, Oyan AM, Kalland KH: PloS ONE 2008, 3:1-11.
2. Ke XS, Qu Y, Rostad K, Li WC, Lin B, Halvorsen OJ, Haukaas SA, Jonassen I, Petersen K, Goldfinger N, Rotter V, Akslen LA, Øyan AM, Kalland KH: PLoS ONE 2009, 4:e4687.
Host: Karl Johan Tronstad, Department of Biomedicine