Comment on ‘Updated investigations of cancer excesses in individuals born or resident in the vicinity of Sellafield and Dounreay’: premature all-clear for nuclear power

We read with great interest the paper by Bao et al (2014) entitled 'COUP-TFII regulates metastasis of colorectal adenocarcinoma cells by modulating Snail1', showing upregulation of Snail and downregulation of E-cadherin in the human epithelial intestinal cell line HIEC overexpressing COUP-TFII upon transfection (Bao et al, 2014; Figure 5) and converse effects in COUP-TFII-KO loVo cells (Bao et al, 2014; Figure 4).We previously explored the role of COUP-TFII as a regulator of E-cadherin expression because we also had noticed a correlation between abnormally high expression of COUP-TFII and lack of E-cadherin in a large panel of carcinoma cell lines of various origins including 22 colon cancer cell lines.We transfected several cell lines with an inducible COUP-TFII expression vector and, as shown in Figure 1, did not observe a pronounced effect on E-cadherin expression.In the human colon carcinoma line HCT116 in particular, E-cadherin remained perfectly stable.HCT116 cells are not invasive in Matrigel assays and invasion was not induced by COUP-TFII.Importantly, in the lung carcinoma line A549 that is invasive straightaway, E-cadherin expression was moderately increased together with, as previously reported (Navab et al, 2004), the invasion capacity of the cells.We also extinguished COUP-TFII expression in E-cadherin-negative cell lines (SW800, MDA-MB 231) using RNA interference and saw no E-cadherin switch-on.Together with the data from Bao et al, this shows that the effect of COUP-TFII is highly dependent on the cellular context.This is actually not very surprising since transcriptional regulation by COUP-TFII is quite complex, involving dimerisation and recruitment of cofactors, and can encompass not only activation but also direct and indirect repression (Park et al, 2003).