These results consequently show that NPM1 is involved in the increase of proliferation and clonogenic capacities of prostate tumour cells. The above results demonstrate that NPM1 exerts a control on the clonogenic capacities of prostate cancer cells and suggests that, besides its effect on the cell proliferative rate, it could also contribute to both tumour growth and aggressiveness. To further evaluate the role of NPM1, we examined the invasion and migration capacities of shNPM1 LNCaP cells. The knockdown of NPM1 significantly decreased the invasion of LNCaP cells through matrigel-coated filters by 40%. We further assessed the LNCaP cells migration ability by physically wounding cells plated on the cell culture plates. As shown in Figure 2b, at 72 h after being scrubbed, shNPM1 cells were unable to recolonize denuded zone as faster as did the control cells. To test whether the decrease of migration and invasion capacities is accompanied by a decrease of the three-dimensional growth abilities of shNPM1 LNCaP cells, in vitro tests in soft agar were performed. Decrease in NPM1 in LNCaP cells almost abolishes their ability to form 3-D colonies. These results from in vitro assays strongly suggest that NPM1 regulates the migratory and invasive properties of prostate cancer cells. To further determine the role of NPM1 in prostate cancer, we analysed tumourigenesis of the shNPM1 LNCaP and shScr LNCaP prostate cancer cell lines in vivo following subcutaneous injection in the flank of male nude mice. ShNPM1 cells produce smaller tumours and, contrary to shScr LNCaP cells, are more likely to produce no tumour at all when grafted. In detail, tumours initiated from LNCaP control cells appear at 22 days post-injection whereas tumours originating from LNCaP shNPM1 cells are noticeable only at 24 days postinjection. Furthermore, growth curves never get paralleled even after 29 days. Thus, tumours initiated from NPM1 knocked-down cells remained smaller than control tumours with 85% decrease of the average tumour volume. This important decrease of tumour volume results in a 95% decrease of tumour weight from LNCaP shNPM1 cells. This decrease is unlikely due to a loss of expression of the anti-NPM1 shRNA since all shNPM1 tumours preserve a 90% decrease in NPM1 accumulation. Taken together, these data clearly demonstrate that NPM1 is involved in prostate tumour growth in vitro and in vivo.