Publications

pRB protein promotes differentiation through inhibition of RBP2 protein

Inactivation of the retinoblastoma gene (RB) is viewed as a necessary step in the development of human cancers. It might be a result of mutation in the RB gene or, more common, due to pRB hyperphosphorylation. Both types of inactivation impair the ability of pRB to interact with some cellular proteins. These protein-protein interactions have been suggested to be responsible for several processes deregulated in cancer. The view on the pRB as a tumor suppressor has been dominated by its role in negative regulation of cell cycle progression and linked to the regulation of E2F family of transcription factors. Binding to pRB results in repression of E2F on promoters important for cells to pass from quiescence through G1 to growth.

One of the pRB tumor suppressor functions has been connected to its ability to promote differentiation and senescence, a process during which cell stops dividing and attributes a certain cell fate, and there is growing appreciation that this might be primary tumor suppressor function of pRB. The role of pRB is not only in the initiation and maintenance of the permanent cell cycle exit, but also in concomitant induction of gene expression required for terminal differentiation. Importantly, the pRB function in differentiation does not involve E2F repression and associated acute cell cycle arrest. pRB has been shown to positively regulate several proteins required for cell fate determination. However, underlying mechanisms are unknown. We have shown that the ability of pRB to promote differentiation correlates with its ability to bind and inhibit the RBP2 protein (Figure 1A). RBP2 is a critical player in differentiation and that to accomplish this, it regulates some genes required for maintenance and progression of the cell-fate program.

Initially, we were focused on the search for proteins that mediate the ability of pRB to induce differentiation. We identified a critical downstream target of pRB in promoting differentiation, RBP2. The results of my work further support the emerging view that cancer is not a disease of proliferation but rather a disease of differentiation. Some pRB mutations exhibit lower frequency for tumor development in patients with familial retinoblastoma than others. These pRB mutant proteins were found to retain their interaction with the RBP2 protein, while losing many other protein interactions. Our epistasis experiments showed that downregulation of RBP2 by siRNA phenocopies restoring pRB function in Rb-/- cell lines (Figure 1B).

Figure 1. Knockdown of RBP2 Cooperates with MyoD to Promote Differentiation

(A) Domain structure of RBP2

(B) Photomicrographs of RB-/- Mouse Embryo Fibroblasts transfected with a plasmid encoding pRB (WT or D22) or RBP2 siRNA (‘2’) or scrambled ‘2’(‘2sc’). Cells were fixed and stained with anti-MHC (red), which is a marker of myogenic differentiation, and counterstained with the nuclear stain DAPI (blue).

Surprisingly, we found that RBP2, E2F, pRB and other members of RB family, such as p130, have some common targets (Figure 2), suggesting that regulation of these genes during cell cycle is coupled to their regulation during differentiation (Figure 3). pRB prevents RBP2 from repressing certain genes, such as CBFA1-responsive genes, while augmenting RBP2’s ability to activate others, such as BRD2 and BRD8; these observations are consistent with our biochemical data on the redistribution of RBP2 from transcriptionally inactive to active chromatin during cell differentiation.

Figure 2. RBP2 and pRB Bind to the Same Target Genes

(A) ChIP with RBP2 antibodies, anti-Flag or control IgG on the OC and control HOXA5 promoter. Experiments were performed in SAOS-2 cells transfected to produce, where indicated, Flag-tagged pRB.

(B) ChIP with the indicated antibodies performed in U937 cells treated for differentiation for 27 or 96 hours, on the BRD2 and BRD8 promoter regions.

Figure 3. Current models.

To get a better idea about the biological function of RBP2, we identified, in an unbiased way, direct RBP2 target genes. For these purposes we applied a genome-scale analysis called human location analysis, or ChIP-on-chip. First, we enriched for the RBP2 bound chromatin fragments using ChIP and then, hybridized them to a microarray containing sequences of human promoters. The microarray contained the ~19,000 proximal promoter regions (i.e., the region spanning 700 base pairs upstream and 200 base pairs downstream of the transcription start site) of ~10,000 human genes. To investigate how RBP2 changes during differentiation, we performed ChIP-on-chip in cells at different stages of differentiation and compared them to undifferentiated cells. Our ongoing collaboration with the institutions that pioneered this technology, Richard Young’s lab in the MIT and NimbleGen Inc., allowed us to generate high-quality data that unambiguously link RBP2 to regulation of distinct sets of genes (Figure 4).


Figure 4. RBP2 target genes. Venn diagram showing number of RBP2 target promoters in differentiating cells. Pie chart representing GO of RBP2 targets.