p120E4F-1: A Novel Candidate Factor for Mediating Bmi-1 Function in Hematopoietic Stem Cells.

We have recently demonstrated that the Polycomb group (Pc-G) gene bmi-1 is essential for the self-renewal/proliferation of both normal and leukemic hematopoietic stem cells (HSCs). Interestingly, none of the gene products with which Bmi-1 interacts are expressed in this cellular compartment. Therefore, it has been postulated that the proliferative function of Bmi-1 in HSCs is mediated through its interaction with non-identified partners. A yeast two-hybrid assay, using Bmi-1 as bait to screen a fetal liver cDNA library, led to the isolation of p120E4F-1, previously identified as a negative regulator of cell proliferation. Mutational analysis showed that this interaction occurs through the HTH motif of Bmi-1, shown to be essential for its ability to extend the replicative lifespan of fibroblasts. The interaction between endogenous Bmi-1 and p120E4F-1 has been confirmed in mammalian cells by co-immunoprecipitation experiments. Curiously, Bmi-1 and p120E4F-1 interact specifically in the cytoplasmic compartment. In addition, increasing the protein levels of Bmi-1 leads to a reduction in p120E4F-1 protein levels, without affecting cellular localization or mRNA levels. Using overexpression and small interfering RNA (siRNA) -mediated knock-down experiments, we show that the overexpression of Bmi-1 into NIH3T3 cells enhances cell growth, whereas reduction of endogenous Bmi-1 results in significant growth suppression. Conversely, overexpression of p120E4F-1 led to an inhibition of cell proliferation, whereas reduction of p120E4F-1 level led to an hyperproliferation. Complementation studies also demonstrates that the proliferative defect induced by the overexpression of p120E4F-1 can be partially rescued by overexpression of Bmi-1 and that the proliferative defect induced by Bmi-1 knock-down is suppressed by the reduction of p120E4F-1 protein level. Moreover, we show that both the HTH and the RING1 domain of Bmi-1 are necessary to abrogate the anti-proliferative effect of p120E4F-1. Taken together, these data demonstrate that Bmi-1 and p120E4F-1 interact physically and genetically and suggest that Bmi-1 down regulates p120E4F-1 through a post transcriptional mechanism. As Bmi-1 is a RING finger domain containing protein, we postulate that it might be acting as an E3 ubiquitin ligase on p120E4F-1. To investigate this possibility we used the ts20 mutant cell line expressing a thermolabile ubiquitin-activating enzyme (E1) that is inactivated at elevated temperature, preventing ubiquitination and subsequent degradation. We found that endogenous p120E4F-1 accumulates markedly when cells are shifted to the restrictive temperature. Moreover, introduction of Bmi-1 specific siRNA prevents accumulation p120E4F-1 at the nonpermissive temperature, suggesting that p120E4F-1 is subject to ubiquitination and that Bmi-1 is required for its degradation. Altogether our data reveal an unsuspected cytoplasmic function for Bmi-1 in the regulation of cell cycle progression through its interaction with p120E4F-1.