RB, the Conductor That Orchestrates Life, Death and Differentiation

L Khidr and P-L Chen

Department of Biological Chemistry, University of California-Irvine Med Sci 1, Irvine, CA, USA

The retinoblastoma susceptibility gene was the first tumor recombinant Rb cDNA to Rb deficient cancer cells suppressor gene identified in humans and the first tumor repressed their tumorigenic potential in nude mice suppressor gene knocked out by targeted deletion in mice. assays, ultimately led to the exciting key discovery that RB serves as a transducer between the cell cycle indeed, Rb is the first known human suppressor of machinery and promoter-specific transcription factors, tumors (Huang et al., 1988). Intense studies in the past its most documented activity being the repression of the two decades support a central theme that the retino- E2F family of transcription factors, which regulate the blastoma gene product (RBprotein) plays critical roles expression of genes involved in cell proliferation and in regulating cell cycle progression and cellular differ- survival. Recent investigations of RB function suggest that entiation. Its role in regulating the G1-phase of the cell it works as a fundamental regulator to coordinate cycle was confirmed by overexpression studies in which pathways of cellular growth and differentiation. In this cells were microinjected with purified unphosphorylated review, we unravel the novel role of an equally important RBprotein during early G1-phase and resulted in aspect of RB in downregulating the differentiation reversible G1 arrest, while injection of similar amounts inhibitor EID-1 during cellular differentiation by teasing of RBin late G1-phase or early S-phase exhibited no apart the signal, which elicit differentiation and limit cell effect on DNA synthesis (Goodrich et al., 1991). This cycle progression, since the molecular mechanisms relat- observation revealed a restriction point that is con- ing to RB activation of differentiation is much less trolled by RBduring the cell cycle transition from early understood. We review the various roles for RB in G1-phase to late G1/S-phase, through which normal differentiation of neurons, muscle, adipose tissue, and cells are committed to DNA synthesis and cell division the retina. In addition, we provide an update for the (Flemington et al., 1993; Helin et al., 1993). Thereafter, current models of the role of RB in cell cycle to entry and RBin complex with its sequence-specific DNA-binding exit, extending the view toward chromatin remodeling and trancription factor E2F (Nevins, 1992), emerged as the expose the dichotomies in the regulation of RB family key gatekeeper to restrain cell cycle progression by members. We conclude with a discussion of a novel RB repressing the transcription of genes required for regulatory network, incorporating the dynamic contribu- S-phase and DNA synthesis in addition to inhibiting tion of EID family proteins. cell cycle re-entry from quiescence in order to maintain a Oncogene (2006) 25, 5210–5219. doi:10.1038/sj.onc.1209612 terminally differentiated state. By deduction, functional RBis therefore essential for proper cell cycle withdrawal Keywords: RB; differentiation; E2F; EID-1 during cellular differentiation although this detailed aspect of regulation by RBremains unclear. Regulation of RBby cyclin-dependent kinases is conserved among various cell types (Sherr, 1996). However, the role of RB in cellular differentiation has been shown to display high Introduction tissue specificity that could involve cell type specific factors, as well as functionally redundant players like Extensive studies of the gene responsible for the p107 and p130. This review extends the current under- heritable malignant childhood disease of the retina, the standing of the Rb tumor suppressor by examining its retinoblastoma gene, provided the first evidence for the role as a regulator of cellular differentiation, presents presence of tumor suppressors in our genome in the new data suggesting its function in development, and early 1980s (Benedict et al., 1983; Cavenee et al., 1983; provides future direction for studies concerning how RB Cavenee et al., 1985). Indeed, mutation of the retino- maintains these roles in addition to safeguarding blastoma susceptibility (Rb) gene was found both in genomic integrity in the cell. familial and sporadic retinoblastoma cases (Friend et al., 1986; Lee et al., 1987). Yet the fascinating experiment in which reintroduction of a single copy of RB in development and differentiation

Correspondence: Dr P-L Chen, Department of Biological Chemistry, RB and Neurogenesis University of California-Irvine Med Sci 1, Irvine, CA 92697, USA. A role for RBin mammalian development was first E-mail: [email protected] established after the demonstration that RbÀ/À embryos RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5211 die between days 13 and 15 of gestation (Clarke et al., myotubes reenter the cell cycle, while normal myoblasts 1992; Jacks et al., 1992; Lee et al., 1992). As deregula- enter a postmitotic state before differentiating into tion of cell cycle progression and cellular differentiation myotubes (Novitch et al., 1996; Zacksenhaus et al., play a critical role in tumorigenesis, it became evident 1996). This observed phenomenon recapitulates earlier that an alternative function of RBis to remove work, which postulated RBas a prerequisite differentiation blocks that are set at certain stages of for the establishment of the postmitotic state of skeletal development to maintain the proper timing of differ- muscle cells (Schneider et al., 1994). This postulate was entiation through its specific interaction with the later supported by several animal studies demonstrating inhibitor of differentiation, Id2. Id2, a target of RB, is newborn mice lacking RBexhibit severe defects in required to maintain the timing of differentiation in skeletal muscle (Zacksenhaus et al., 1996; de Bruin et al., many processes of mammalian development and also 2003; Wu et al., 2003). In the absence of RB, skeletal serves as a positive regulator of cell cycle progression muscle cells show ectopic DNA synthesis (Norton et al., 1998). The unrestricted activation and/or apoptosis (Schneider et al., 1994; Novitch of Id2 underlies defects in differentiation of erythrocytes et al., 1996; Novitch et al., 1999). To tease apart the and neuronal cells of Rb knockout mice, presumably molecular mechanism governing the steps of Rb since the interaction between RBand Id2 is proposed to expression during myogenesis, Delehouzee et al. have sequester Id2 (Iavarone et al., 1994). The observed proposed a model involving GABP, a widely expressed defects in neurogenesis and erythropoiesis caused by loss erythroblast transformation specific transcription factor of Rb in the developing RbÀ/À mouse embryo recapitu- GA-binding protein that targets both nuclear-encoded lates the consequences of uncontrolled Id2 activity, as mitochondrial proteins and neuromuscular-specific suggested by the observation that a null mutation genes, YY1, a ubiquitous and multifunctional zinc- of Id2 rescues defects and prolongs the life span of finger transcription factor member of the Polycomb RbÀ/À embryos (Lasorella et al., 2000). These observa- Group protein family which can both initiate and tions, in amalgamation to the increased prevalence regulate transcription by activation or repression of of abnormal mitosis and apoptosis observed in the target genes, and HCF-1, the abundant chromatin- intermediate zones of the developing neural tube, associated human factor which is a heterodimeric coupled to a decrease in the expression of several complex essential for G1-phase progression and proper neuronal markers, including neurotrophin receptors cytokinesis at the end of M-phase. In their study, they TrkA, TrkB, and p75, implicated a role for RB in demonstrate before differentiation GABP and YY1 are neurogenesis (Lee et al., 1994). Later, to confirm this present on the Rb promoter limiting the level of Rb contention, using a neuronal specific promoter driving a expression. As cells advance through myogenesis, LacZ reporter gene, widespread abnormalities in the HCF-1 is recruited to, and YY1 is removed from, the developing nervous system was observed, including the promoter resulting in a sharp upregulation of Rb gene olfactory epithelium, the retina, and the neocortex expression observed at day 2 of myogenesis (Delehouzee (Slack et al., 1998). These studies suggested that RBis et al., 2005). This model for RBregulated myogenic essential immediately following commitment to a differentiation is further supported by earlier studies neuronal fate. Interestingly, there is a high incidence investigating the effect of Rb deletion before and after of apoptosis in the extraembryonic tissues of the myogenic differentiation (Huh et al., 2004). Myf5- placenta. A provocative theory suggested that some of cre;Rb floxed mice, which deletes RBin myoblasts, die the abnormalities observed in RbÀ/À embryos could be immediately at birth with a characteristic phenotype attributed to non-cell autonomous functions, since the exhibiting a high incidence of apoptosis and almost supply with normal placenta to RbÀ/À embryos demon- complete absence of myofibers. MCK-cre;Rb floxed strated that they can reach full term, but die with severe mice, which deletes Rb in differentiated fibers, were problems in the skeletomuscular system (Wu et al., viable and display normal muscle phenotype. Together, 2003). Additionally, a recent conditional mouse model these studies suggest that RBis required for progression with telencephalon-specific Rb deletions reveals a cell- of differentiation and not for the maintenance of the autonomous role of Rb in neuronal migration (Fergu- differentiated state. In contrast, differentiated mouse son et al., 2005). hair cells with a conditional deletion of Rb are shown to be able to re-enter cell cycle and divide, while continuing RB and skeletal muscle system to become highly differentiated and functional, suggest- During muscle differentiation, it has been demonstrated ing that the requirement of RBfor the maintenance of that the mRNA and protein levels of Rb increase postmitotic state might be tissue specific (Mantela et al., (Coppola et al., 1990). Rb augments the activity of 2005; Sage et al., 2005). MyoD, an early muscle basic helix-loop-helix transcription factor important for muscle differentiation and cell RB and adipogenesis cycle suppression, since the activity of MyoD-mediated RB, p107, and p130, known collectively as pocket transcriptional activation of certain myogenic genes proteins due to a shared viral oncoprotein-binding requires functional RB(Gu et al., 1993). Moreover, it domain (Ewen et al., 1991; Mayol et al., 1993), regulate was observed that inactivation of RBresults in the the cell cycle by restraining the G1–S transition through inhibition of differentiation of myoblasts and myotubes the reversible regulation of E2F-responsive genes in culture and accordingly, terminally differentiated (Flemington et al., 1993; Helin et al., 1993), by

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5212 inhibiting Cdk2, or by stabilizing p27-kip (Woo et al., minor contribution to the adult retina (Maandag et al., 1997; Rodier et al., 2005). Moreover, they are also 1994), suggesting RbÀ/À cells undergo apoptosis. To implicated in the control of G0 exit (Hannan et al., 2000; further support this observation, it was later shown that Scott et al., 2001), the spatial organization of replication expression of E7, which binds RB, p107 and p130, (Barbie et al., 2004), and genomic rereplication (Co- induces cell death in the retina (Howes et al., 1994), brinik, 2005). Additionally they contribute to silencing while expression of SV40T antigen, which binds RB of such genes in cells that are undergoing senescence or family proteins and p53, results in tumor formation differentiation (Frolov and Dyson, 2004). However, (Windle et al., 1990). Taken together, these studies during adipogenesis opposing roles for RBand p130/ suggested that RbÀ/À cells must escape apoptosis in the p107 has been demonstrated in which MEFs lacking Rb development of retinoblastoma (Classon and Harlow, are unable to undergo adipose conversion in response to 2002). More recently, it was demonstrated that the treatment with standard adipogenic inducers (Chen laminar organization of retinas from RbÀ/À embryos is et al., 1996) and have general defects in their differentia- severely disrupted which is caused by alteration in tion potential, while p107/p130-deficient cells display an retinal cell fate specification and differentiation. Rb is increased differentiation potential (Classon et al., 2000). expressed in proliferating retinal progenitor cells in the It has been shown that Rb deficient MEFs or stem cells postnatal mouse retina and in differentiating rod undergo differentiation preferentially into white adipo- photoreceptors, where it exerts two very distinct roles cytes when the cells are treated with a PPAR-gamma (Zhang et al., 2004). In replicating progenitor cells, Rb is ligand (Hansen et al., 2004). Moreover, inactivation of required for efficient cell cycle exit while in differentiat- the RBfamily by SV40 large T antigen displays similar ing rods it is required for appropriate maturation (Dyer defects in adipocyte differentiation (Higgins et al., 1996). and Bremner, 2005). In contrast to the role of Rb in Adipocyte precursor cells give rise to two major cell other regions of the developing CNS, its dual role in populations with distinct physiological roles: white and these two retinal cell populations is cell autonomous. brown. White adipose tissue stores energy, while brown RBforms a functional complex with Pax6, which is adipose tissue dissipates energy through adaptive essential for lens formation (Cvekl et al., 1999). It is thermogenesis. Mice lacking p107 displayed a replace- expressed in lens progenitor cells before the onset of ment of white adipose tissue with brown adipose tissue, crystalline expression and it serves as an important which is mediated by the regulation of PGC-1 alpha regulatory factor required for expression of c-Maf, (Scime et al., 2005). Cre-mediated deletion of Rb in MafA/L-Maf, Six3, Prox1, and retinoic acid signaling adult primary preadipocytes blocked their differentia- both in lens precursor cells and the developing lens tion into white adipocytes (Scime et al., 2005). These (Cvekl et al., 2004). It was previously demonstrated studies suggest, at least in vitro, that RBand its pocket using Pax 6 -cre, which is also active in peripheral retinal proteins serve as a molecular switch that determines progenitors at embryonic day 10, that RBwas the pathway leading to brown or white adipocyte inactivated in the developing retina and leads to a differentiation (Hansen et al., 2004). Nevertheless, reduction in both ganglion and bipolar cells (Chen et al., the underlying molecular mechanism largely remains 2004). Therefore, this result in addition with the latter unknown. supports an important role for RBat the molecular level in specific neurons critical to their differentiation or RB and retinal development survival. During retinal development, cell cycle exit and cell fate specification are coordinated to ensure that the adult retina forms appropriately, since the overall size of the The role of RB in cell cycle exit and entrance retina and the proportion of each cell type contained therein is essential for proper vision. When cell Restriction point proliferation and cell fate specification become un- The restriction point separates the G1-phase of con- coupled, visual processing is severely compromised and tinuously cycling cells into two functionally different tumor formation ensues, as exemplified in the case of parts: the first part represents the growth factor retinoblastoma. Importantly, loss of RBcauses ectopic dependent postmitotic interval from mitosis to the cell division in the retina, and even more severely in restriction point and is constant in length (3–4 h), and retinas deficient for both RBand p107. This phenotype the second part which represents the growth factor is consistent with retinoblastoma development, which independent pre-S-phase interval of G1 that lasts from requires deletion of both Rb and p107 (Robanus- the restriction point to S-phase and varies in time (1 to Maandag et al., 1998). Immunocytochemical studies 10 h) (Pardee, 1974; Dou et al., 1993). The normal suggested that retinoblastoma arises from death resis- function of RBin regulating cellular growth can be tant inner nuclear layer precursors (Chen et al., 2004), elucidated from its undulating phosphorylation pattern and implies that RbÀ/À retinal cells that fail to commit to which is tightly coupled to distinct phases of the cell terminal differentiation, and not RBdeficient cells prone cycle, and the inactivation of which may lead to to apoptosis, leads to oncogenic transformation. To unbridled cell growth. RBbecomes hyperphosphory- tease apart the mechanism by which cells that lack RB lated in late G1-phase, which is believed to cause its undergo apoptosis or develop tumors, early studies functional inactivation, and remains hyperphosphory- using chimeric mice elucidated that RbÀ/À cells provide a lated in S-, G2-, and M-phases (Buchkovich et al., 1989;

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5213 Chen et al., 1989; DeCaprio et al., 1989). Cells in G0/G1 (Trimarchi et al., 1998). Activator-E2Fs utilize RBas express RBin an un- or hypophosphorylated form. The their major binding partner, while repressor-E2Fs bind mechanism and significance of RBphosphorylation p130, while p107 only binds preferentially to E2F-4 comes from the observation that certain cyclins, which (Frolov and Dyson, 2004). The p107, p130 pocket are known to regulate the specificity and activity of proteins repress transcription by counteracting E2F- cyclin-dependent kinases (cdk), trigger phosphorylation mediated transactivation simply by binding to and of RBand, as a result inhibit RBfunction (Sherr, 1996). masking the E2F activation domain or they can recruit Injecting purified Rb protein into early G1 cells halts corepressor complexes to E2F. Several of the complexes their progression into S-phase of the cell cycle. By that have been linked to RBhave enzymatic activities contrast, late G1 cells are inert to Rb inhibition and directed toward chromatin or recognize particular continue progressing into S-phase (Goodrich et al., histone modifications suggesting that they impinge upon 1991). This phenomenon is reminiscent of the growth chromatin structure (Zhang and Dean, 2001), which is factor requirement preceding the restriction point. supported by the previous finding that these protein Indeed, growth factors stimulate the expression of complexes display a unique expression pattern during D-type cyclins which complex with Cdk 4, 6 to mediate cell cycle progression (Shirodkar et al., 1992; Dyson, the initial phosphorylation of the RBprotein (Kato 1998; Nevins, 1998). The RB/E2F complex can be et al., 1993) and mitigates the repression of cyclin E detected in G1-phase and seems to play a critical role in transcription by Rb and, its tightly bound cellular the G1 to S transition. Activator-E2Fs exist in free form transcription factor, E2F (Harbour et al., 1999). E2F during S-phase. The p130/E2F complex is most evident was originally identified on the basis of its involvement in quiescent cells and the p107/E2F complex is present in adenovirus E2 gene transcription (Nevins, 1992). It is in both G1- and S-phase. Thus the question remains as required for transcription of cellular genes that partici- to the exact physiological role of different RB/E2F pate in growth control and DNA synthesis and has been complexes and of the factors that determine which RB/ demonstrated to bind the hypophosphorylated form of E2F complex will bind to and regulate E2F response RB(Hiebert et al., 1992; Johnson et al., 1993; Schwarz genes. et al., 1993). Thus, it is the hypophosphorylated RB species that curbs cell growth by binding E2F, thereby The E2F target genes blocking its growth promoting activity. Phosphorylated E2F genes play a key role in the regulation of cyclin E RBor by complex with DNA viral oncoproteins, expression and have provided some of the first mechan- liberates E2F from RBand leads to transcriptional istic evidence of how mutation of RBleads to activation of growth promoting genes (Nevins, 1994; deregulated cell division. Deregulation of E2F family Taya, 1997). Accumulation of cyclin E together with members gives rise to a variety of severe biological Cdk2 orchestrates the continued phosphorylation on consequences (Trimarchi and Lees, 2002). According to RBprotein and leads to the release of E2F, thereby Wells et al. (2002), loss or overexpression of E2F1 using facilitating the activation of genes critical for S-phase transgenic mouse models has been linked to dysplasia progression (Obeyesekere et al., 1995; Harbour and and tumor formation in the liver, neoplasia of the skin, Dean, 2000). Cyclin E-Cdk2 and cyclin A-Cdk2 also and tumors in the reproductive tract, lung and regulates centrosome duplication concomitantly with lymphatic system. The combined use of E2F1 nullizy- nuclear DNA replication (Hinchcliffe et al., 1999; Lacey gous mice, oligonucleotide-based microarrays, and et al., 1999), whereas terminal differentiation requires chromatin immunoprecipitation revealed that E2F the permanent withdrawal from the cell cycle and target promoters were bound by multiple different involves the repression of genes critical for S-phase E2Fs, the mRNA levels of target genes decrease in entry. Therefore, the cellular response to functional Rb accordance with loss of E2F1 and, in addition, the fundamentally depends on its status in the cell cycle. observation that E2F1, and not other E2Fs, is bound to the proximal promoter regions in vivo highly suggests RB and E2F that specific genes are directly, and nonredundantly, The RB/E2F complex consists of three RB like proteins regulated by E2F1. However, several of the promoters (RB, p107, and p130), and six E2F family members identified do not have a consensus E2F site implying (E2F -1, -2, -3, -4, -5, and -6) each of which can that E2F1 can be recruited by other DNA-binding heterodimerize with two DNA-binding protein partners proteins to execute its regulatory function. Recently, (DP1 and DP2) to form 12 different DNA-binding using genome-wide microarray analysis of the proximal transcriptional regulators (Cam and Dynlacht, 2003). promoters of nearly1200 human genes, many new target The E2F family members share DNA-binding specificity genes of E2Fs have been identified (Ren et al., 2002). since they bind to and regulate the promoters for genes Based on the results, it was implicated that besides whose products are important for cell cycle progression regulating genes critical for cell cycle progression, E2F and DNA synthesis, however, they can be divided into could govern the expression of genes required for DNA two classes: activators (E2F1, E2F2, and E2F3) and repair, chromatin assembly, condensation, and segrega- repressors (E2F4 and E2F5) of transcription. E2F6 is tion, as well as multiple checkpoints that ensure genomic considered to be an independent member that lacks integrity. In another study, ChIP assays using an several functional domains, including the cyclin- and antibody to E2F4 to probe human CpG islands by pRb-binding domains and the trans-activation domain microarray analysis, revealed many genes involved in

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5214 DNA repair or recombination (Weinmann et al., 2002). structures, in senescent cells, the formation of which Collectively, these studies suggest that E2F targets requires the RBpathway (Narita et al., 2003).Impor- include genes involved in chromatin structure and tantly, the levels of me-K9-H3 increased and correlated DNA replication/repair/recombination, and deregula- with the recruitment of RBand HP1 to the E2F tion of these target genes could have profound effects on response promoters. Taken together these studies the integrity of the genome. establish a critical role for the RB/E2F repressor in dampening E2F response genes in senescent cells and RB and chromatin modifiers have identified a critical difference between resting cells RBand the E2F family genes are critical for cells to and senescent cells. In resting cells, p130/E2F4-mediated progress through G1-phase. How they control the deacetylation on the promoter of E2F response genes expression of E2F response genes is under intense can quickly respond to growth signals whereas RB- investigation. It was proposed that RBsuppresses mediated methylation events stabilize senescent cells. It activator-E2Fs by directly binding to the activation will be interesting to determine the significance of the domain (Flemington et al., 1993) (Helin et al., 1993). In role of RBtumor suppression in senescent cells. vitro transcription analysis revealed that RBblocks the Interestingly, Ait-Si-Ali et al. (2004) showed by ChIP assembly of preinitiation complexes (Ross et al., 1999). analysis that methylation of H3-K9 (me-H3-K9) can be Recently, a model has been proposed that RBfamily detected at several E2F target promoters in differentiat- members recruit various chromatin modifiers to repress ing cells but not in cycling cells. It can be ascertained E2F response genes (Brehm et al., 1998; Luo et al., 1998; from these studies that RB/E2F-regulated promoters Magnaghi-Jaulin et al., 1998; Robertson et al., 2000) undergo a permanent silencing mechanism when cells (Nielsen et al., 2001; Vandel et al., 2001; Pradhan and commit to terminal differentiation, although a more Kim, 2002). In the absence of growth stimuli, unphos- thorough appreciation of this mechanism remains to be phorylated RBfamily proteins bind E2F and recruit elucidated. corepressors to enable factors that remodel chromatin at the E2F response promoter. These factors include, Differential regulation by the RB family members but are not limited to: histone deacetylases (HDACs) E2F response promoters are occupied by E2F4 in (Brehm et al., 1998; Luo et al., 1998; Magnaghi-Jaulin complex with p107 or p130 in resting cells (Moberg et al., 1998), histone methyltransferases (Nielsen et al., et al., 1996). This biased occupancy can be partially 2001; Vandel et al., 2001), and DNA methylases explained by the observation that there is an additional (Robertson et al., 2000; Pradhan and Kim, 2002). E2F-binding domain present in the C-terminus of RB The p130/E2F4 complex binds the E2F response (Dick and Dyson, 2003). Consistently, studies in vitro promoter in resting cells (Moberg et al., 1996; see also show that the RB-E2F complex has relatively weak Macaluso et al., in this Issue). By ChIP analysis, it DNA-binding activity (Ray et al., 1992). Thus an was demonstrated that these promoters were also bound emerging model identifies RBas a harness that by HDACs, implying that genes regulated by E2F sequesters activator type E2Fs in early G1- or G0- are repressed in early G1-phase through deacetylation phase, and results in the E2F response promoter to on the nucleosomes (Takahashi et al., 2000; Wells et al., become occupied by repressor type E2Fs (Frolov and 2000; Ferreira et al., 2001; Taubert et al., 2004). Dyson, 2004). This model suggests a coregulation of When cells transit through midG1 these E2F response E2F response genes by the RBfamily proteins. promoters are occupied by activator-E2Fs. This repla- However, a recent study using array analysis to detect cement correlates with the dissociation of HDACs from gene expression profiles in cells lacking RBor p107/ the promoters. Subsequently, HATs, such as p300, p130, revealed a distinct gene expression profile suggest- CBP and p/CAF, arrive at the promoter region in ing a nonredundant regulatory role governed by the RB correlation with the level of histone hyperacetylation. family members (Black et al., 2003). Specifically, genes Although RBhas been shown to bind HDACs, encoding DNA replication and cell cycle regulatory the contribution of RBto the deacetylation state during proteins are found to be the signature of Rb-deficient G0 is not clear. Instead, the interaction between RB cells. In contrast, in cells lacking p130 and p107, genes and the histone methyltransferase SUV39H1 has been that encode proteins generally involved in the regulation shown to be critical in modifying chromatin state, since of cell growth and those encoding proteins known to be it was shown that SUV39H1 function requires associa- involved in maintaining the extracellular matrix, are tion with RB(Nielsen et al., 2001; Vandel et al., 2001). deregulated. The molecular basis for the differential SUV39H1 methylates K9 of histone H3 and facilitates expression profile remains to be explored. the binding for the HP1 protein. HP1 has been suggested to play a role in establishing a heterochromatin-like state. Moreover, methylation of histones will E2F response gene and tumorigenesis render a promoter in a more stable repression state while The E2F transcription factors, together with RBand its acetylation can be easily reversible. Recently, related proteins p107 and p130, mediate the activation SUV39H1-mediated methylation was found restricted or repression of genes involved in diverse cellular in resting cells (Nicolas et al., 2003), suggesting its role in processes, including the cell cycle and DNA repair. sustaining cell cycle arrest. Narita et al. have noted the Interestingly, RBremains to be the only gene in the presence of punctuate DNA foci, heterochromatin-like family found to be mutated in human cancers. Genetic

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5215 studies revealed the contribution of the deregulation of cells and was proposed to inactivate its activity. The E2F1 and E2F3 in apoptosis and aberrant cell unphosphorylated form of RBpersists in both differ- proliferation seen in RBdeficient embryos (Tsai et al., entiating and differentiated cells, implying its critical 1998; Ziebold et al., 2001). Recently, Lee et al. (2002) role in executing and maintaining proper differentiation. investigated how E2F4 augments RBtumor suppres- Recent advancements have led to the identification of sion. Interestingly, elimination of E2F4 suppresses the the E1A-like inhibitor of differentiation 1 (EID-1) development of both pituitary and thyroid tumors in (MacLellan et al., 2000; Miyake et al., 2000). Its Rb þ /À mice. This study, for the first time, documented potential role in RBacetylation (Nguyen et al., 2004; the importance of the RB/E2F complex in tumorigen- Krutzfeldt et al., 2005) may provide the link to the esis. Careful biochemical analysis revealed that elimina- unique role of RBin cellular differentiation in compar- tion of E2F4 rescued the deregulated activator-E2Fs in ison to its related family members. the absence of RB, suggesting that activator-E2Fs could significantly contribute to the tumorigenesis mediated CDK kinases by the loss of RB. Transition between cell cycle phases is regulated by a series of temporally coordinated phosphorylation events that target members of the pocket protein family, RB, Regulating RB protein p107, and p130. The kinases responsible for pocket protein phosphorylation are cyclin-dependent kinases, We have discussed so far the lessons learned from which are activated in part by their association with animal model systems, in vitro cell biology, and cyclins. Several species of RBwere identified by sodium biochemical analysis concerning the role of RBin dodecyl sulfate–polyacrylamide gel electropheresis cellular differentiation and cell cycle progression. (SDS-PAGE) and displayed a unique expression pattern Initially RBwas envisioned to function as a negative as cells transited through the cell cycle (Buchkovich regulator of cell cycle progression and a positive et al., 1989; Chen et al., 1989; DeCaprio et al., 1989). RB regulator for cellular differentiation. The emerging is hypo-phosphorylated in resting and early G1 cells. model concerning the role of RBin these diverse cellular Progressive phosphorylation became evident when activities is the availability of RBat G1-phase, which resting cells were challenged with mitogenic signals that determines whether cells will commit to entering the cell led to the accumulation of D cyclins, which function cycle, or prepare for differentiation once the specifica- reciprocally, to elicit the initial phosphorylation (Sherr, tion of a specific lineage is acquired. How cells 1996). Subsequently, cyclin E accumulation accounts for orchestrate the diverse signals to modulate the action further phosphorylation on RBand leads to full of RBis largely determined through post-translational activation of E2F1. The importance of these phosphor- modification on RB(Figure 1). It is widely accepted that ylation cascades was supported by observations from cells govern RBfunction through multiple CDK kinase cancer studies, since most human malignancies have an complexes that integrate environmental cues. In the aberration in the Rb pathway due to Cdk hyperactiva- absence of growth stimuli, unphosphorylated RBbinds tion. Cancer cells bearing wild-type RBprotein usually proteins such as E2F family members, as well as contain a highly deregulated CDK kinase pathway. For corepressors, and together enable chromatin-remodeling example, amplification of cyclin D and CDK4 in bypass activities. Phosphorylation of RBis specific for cycling of the requirement for growth stimuli signals, were found in a subset of cancers. Additionally, a high frequency of mutations in CDK inhibitors such as p27, p21, and INK, were found to negatively regulate CDK and promote tumor formation.

Acetylation of RB by CBP/p300 CBP and p300 are highly related proteins that serve as transcriptional co-activators for nearly every gene- speciﬁc trans-activator (Eckner, 1996). They participate in multiple regulatory networks and communicate with other co-activators and multiprotein complexes, making them co-integrators that are involved in multiple biological processes that affect cell growth, transformation and development (Goodman and Smolik, 2000). It was proposed that EID proteins regulate differentiation by antagonizing CBP/p300 or a co-regulator of SHP Figure 1 Schematic drawing and post-translational modiﬁcation (Bavner et al., 2002). EID1, E1A-like inhibitor of of RB. Progressive modulation by CDK kinases alters RB protein differentiation, was cloned on the basis of interactions structure and leads to the sequential release of HDAC, and E2F. EID-1 binds to RBin differentiating cells and leads to p300- with the retinoblastoma tumor suppressor protein and mediated acetylation of RB, inducing Mdm2 binding. The was subsequently characterized as an inhibitor of CBP/ association of Mdm2 leads to the degradation of EID-1. p300 dependent functions of differentiation (Miyake

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5216 et al., 2000) (MacLellan et al., 2000). EID2 was cloned as the second member of this family and was shown to act as an inhibitor of gene expression (Ji et al., 2003). Interestingly, it has been suggested that EID2 inhibition is a combination of two mechanisms, binding to CBP/ p300 or recruiting histone deacetylases. EID3 is specifically expressed in testis and, at the subcellular level localized to both nucleus and cytoplasm (Bavner et al., 2005). EID3 inhibits gene expression mediated by nuclear receptors (NR), most probably by interacting with the SRC-1 interacting domain (SID) of CBP that Figure 2 Comparative protein domain mapping of E1A and has been implicated in co-activator assembly, thus EID-1. Critical conserved domains of E1A are CR1 and CR2. CR2 preventing recruitment of CBP to a natural NR- contains the L Â C Â E RB-binding domain. Domain conferring regulated promoter. repression of differentiation resides in CR1. Of particular interest to the role of RBin differentiation, however, is its interaction with EID1. E1D1 mediates the repression of CBP/p300-dependent transactivation by sequestration of CBP/p300, TBP, and evolutionarily represent the cellular origin of p/CAF. It contains a L Â C Â E RB-binding motif, E1A. Moreover, since EID is quickly degraded through inhibits genes regulated by a variety of nuclear an RB-dependent mechanism when cells commit for receptors, and is a cofactor of the orphan nuclear terminal differentiation, viral E1A, another L Â C Â E receptor SHP, small heterodimer partner NR0B2 containing protein, may compete for the L Â C Â E (Miyake et al., 2000; MacLellan et al., 2000). EID1 containing molecules for RBbinding. Importantly, RB binds Rb and p300/CBP and inhibits the MyoD target defective for L Â C Â E binding and competent for E2F genes possibly by in part inhibiting the HAT activity of binding displayed intact cell cycle regulation activity, p300/CBP. EID-1 is subjected to proteasome-dependent yet, failed to induce cellular differentiation, highlighting degradation when cells exit the cell cycle possibly the importance of L Â C Â E binding activity in regulat- mediated by pRb/MDM2 and is required for normal ing differentiation. muscle differentiation (MacLellan et al., 2000; Miyake Several possibilities arise. First, many L Â C Â E et al., 2000; Krutzfeldt et al., 2005). One important containing proteins have been documented and each observation is that RBacetylation by CBP/p300 play a distinct role in various cellular processes (Riley promotes its association with MDM2 and leads to the et al., 1994). Among them are cell cycle regulator cyclin degradation of E1D-1 (Chan et al., 2001; Nguyen et al., D, transcription co-repressor CtIP, and the nucleosome 2004). CBP/p300-mediated RB acetylation is unique disruptor BRG1. However, what other L Â C Â E among its family members and specifically in differ- containing RB-associated proteins are subject to entiating cells. These observations provide a novel MDM-2 associated degradation? Second, mice defective regulation of RBactivity through acetylation. More- for both MDM2 and p53 develop normally although over, new evidence supports the binding of the RET they develop a high tumor incidence later in life finger protein (RFP)/tripartite motif protein (TRIM 27) (McDonnell et al., 1999). Are there E3 ligase(s) other to EID1 and RB, to inhibit its role as a transcriptional than MDM2 involved in EID-1 degradation in the activator but not as a repressor, and prevents the absence of MDM2? RBis known to interact with cdc27, degradation of E1D-1 (Krutzfeldt et al., 2005). The a component of E3 ligase complex APC (Chen et al., downregulation of RFP leads to an enhanced differ- 1995). It will be interesting to know whether APCs are entiation phenotype in culture cells. Thus, a new involved in the RB-dependent regulation of EID-1. regulatory network involving RB, critical for differen- The novel concept that EID-1 acts as an inhibitor of tiation, is emerging perhaps through RB-mediated E1D- cellular differentiation urges critically the examination 1 degradation regulated through CBP/p300 and RET of its participation in the RB-regulated differentiation finger proteins. process.

New RB regulatory network EID-1 was isolated through its association with RB. Concluding remarks Sequence comparison between EID-1 and E1A revealed an additional conserved domain anterior to RBseems to be a multifaceted protein with multiple the L Â C Â E motif (Figure 2). This domain resides in functions. Until recently, the role of RBin regulating the CR1 region of E1A and has been shown cell cycle progression and differentiation was categor- to be required for its repression of differentiation ized into two separate avenues: its inhibitory role on activity (Sandmoller et al., 1996). However, E2F response genes with an emphasis on control of whether this region is also critical for EID-1 activity the G1 to S transition, and its promotion of differentia- awaits further analysis. As EID-1 shares a similar tion by activating certain tissue-speciﬁc transcription structure arrangement with E1A, there remains factors. However, accumulating evidence insinuates a the intriguing possibility that EID proteins could decision-making role for RBupon encountering cellular

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5217 stimuli in preparation for cell division, gradually molds RBto release E2F. When cells are committed to eliciting the differentiation program, RBwill be required to remove or counteract differentiation inhibitors. RB deficiency then, leads to either apoptosis for cells committed to a specific differentiation program, or uncontrolled cell growth for cycling cells. It is clear that RBis utilized by so many processes, however, given the complexity of known RBassociated proteins, the in vivo RBcomplex configuration remains rudimentary. To date, it is unclear how many functional RBcomplexes exist. It will be a great challenge to determine if RBbinds E2F and EID1 simultaneously, each executing its specific function in concert with RBupon integration of cellular Figure 3 RBis a fundamental mediator of proliferation, apoptosis stimuli, or rather if the RBinteracting proteins find RB and differentiation signals. RBpromotes differentiation by down- separately and execute their function independently. It regulating differentiation inhibitors, EID and ID2. RBrestrains cell cycle progression by inhibiting activator-E2Fs, while p107 and may be possible that a RBcore complex exists while p130 modulate repressor-E2Fs to exit the cell cycle. Deregulation transiently associated members are recruited under of E2F1 leads to apoptosis. specific growth and differentiation conditions in the cell. Or, perhaps there are various RBcomplexes, each participating in distinct processes. One of the challenges signals (Figure 3). In the proposed model, RB, and its is to sort out the relative distribution of RBcomplexes family members, maintain the critical G1 juncture. in G0/G1-phases to elucidate the intricate RBnetwork Phosphorylation cascades in coordination with growth in regulating cell fate.

References

Ait-Si-Ali S, Guasconi V, Fritsch L, Yahi H, Sekhri R, Classon M, Kennedy BK, Mulloy R, Harlow E. (2000). Proc Naguibneva I et al. (2004). EMBO J 23: 605–615. Natl Acad Sci USA 97: 10826–10831. Barbie DA, Kudlow BA, Frock R, Zhao J, Johnson BR, Cobrinik D. (2005). Oncogene 24: 2796–2809. Dyson N et al. (2004). Mol Cell Biol 24: 595–607. Coppola JA, Lewis BA, Cole MD. (1990). Oncogene 5: Bavner A, Johansson L, Toresson G, Gustafsson JA, Treuter 1731–1733. E. (2002). EMBO Rep 3: 478–484. Cvekl A, Kashanchi F, Brady JN, Piatigorsky J. (1999). Invest Bavner A, Matthews J, Sanyal S, Gustafsson JA, Treuter E. Ophthalmol Vis Sci 40: 1343–1350. (2005). Nucleic Acids Res 33: 3561–3569. Cvekl A, Yang Y, Chauhan BK, Cveklova K. (2004). Int J Dev Benedict WF, Murphree AL, Banerjee A, Spina CA, Sparkes Biol 48: 829–844. MC, Sparkes RS. (1983). Science 219: 973–975. de Bruin A, Wu L, Saavedra HI, Wilson P, Yang Y, Black EP, Huang E, Dressman H, Rempel R, Laakso N, Asa Rosol TJ et al. (2003). Proc Natl Acad Sci USA 100: SL et al. (2003). Cancer Res 63: 3716–3723. 6546–6551. Brehm A, Miska EA, McCance DJ, Reid JL, Bannister AJ, DeCaprio JA, Ludlow JW, Lynch D, Furukawa Y, Grifﬁn J, Kouzarides T. (1998). Nature 391: 597–601. Piwnica-Worms H et al. (1989). Cell 58: 1085–1095. Buchkovich K, Duffy LA, Harlow E. (1989). Cell 58: Delehouzee S, Yoshikawa T, Sawa C, Sawada J, Ito T, Omori 1097–1105. M et al. (2005). Genes Cells 10: 717–731. Cam H, Dynlacht BD. (2003). Cancer Cell 3: 311–316. Dick FA, Dyson N. (2003). Mol Cell 12: 639–649. Cavenee WK, Dryja TP, Phillips RA, Benedict WF, Godbout Dou QP, Levin AH, Zhao S, Pardee AB. (1993). Cancer Res R, Gallie BL et al. (1983). Nature 305: 779–784. 53: 1493–1497. Cavenee WK, Hansen MF, Nordenskjold M, Dyer MA, Bremner R. (2005). Nat Rev Cancer 5: 91–101. Kock E, Maumenee I, Squire JA et al. (1985). Science 228: Dyson N. (1998). Genes Dev 12: 2245–2262. 501–503. Eckner R. (1996). Biol Chem 377: 685–688. Chan HM, Krstic-Demonacos M, Smith L, Demonacos C, La Ewen ME, Xing YG, Lawrence JB, Livingston DM. (1991). Thangue NB. (2001). Nat Cell Biol 3: 667–674. Cell 66: 1155–1164. Chen D, Livne-bar I, Vanderluit JL, Slack RS, Agochiya M, Ferguson KL, McClellan KA, Vanderluit JL, McIntosh WC, Bremner R. (2004). Cancer Cell 5: 539–551. Schuurmans C, Polleux F et al. (2005). EMBO J [E-pub Chen PL, Riley DJ, Chen Y, Lee WH. (1996). Genes Dev 10: ahead of print]. 2794–2804. Ferreira R, Naguibneva I, Mathieu M, Ait-Si-Ali S, Robin P, Chen PL, Scully P, Shew JY, Wang JY, Lee WH. (1989). Cell Pritchard LL et al. (2001). EMBO Rep 2: 794–799. 58: 1193–1198. Friend SH, Bernards R, Rogelj S, Weinberg RA, Rapaport Chen PL, Ueng YC, Durfee T, Chen KC, Yang-Feng T, Lee JM, Albert DM et al. (1986). Nature 323: 643–646. WH. (1995). Cell Growth Differ 6: 199–210. Frolov MV, Dyson NJ. (2004). J Cell Sci 117: 2173–2181. Clarke AR, Maandag ER, van Roon M, van der Lugt NM, Goodman RH, Smolik S. (2000). Genes Dev 14: van der Valk M, Hooper ML et al. (1992). Nature 359: 1553–1577. 328–330. Goodrich DW, Wang NP, Qian YW, Lee EY, Lee WH. Classon M, Harlow E. (2002). Nat Rev Cancer 2: 910–917. (1991). Cell 67: 293–302.

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5218 Gu W, Schneider JW, Condorelli G, Kaushal S, Mahdavi V, Nevins JR. (1994). Curr Opin Genet Dev 4: 130–134. Nadal-Ginard B. (1993). Cell 72: 309–324. Nevins JR. (1998). Cell Growth Differ 9: 585–593. Hannan KM, Kennedy BK, Cavanaugh AH, Hannan RD, Nguyen DX, Baglia LA, Huang SM, Baker CM, McCance DJ. Hirschler-Laszkiewicz I, Jefferson LS et al. (2000). Oncogene (2004). EMBO J 23: 1609–1618. 19: 3487–3497. Nicolas E, Roumillac C, Trouche D. (2003). Mol Cell Biol 23: Hansen JB, Jorgensen C, Petersen RK, Hallenborg P, De 1614–1622. Matteis R, Boye HA et al. (2004). Proc Natl Acad Sci USA Nielsen SJ, Schneider R, Bauer UM, Bannister AJ, Morrison 101: 4112–4117. A, O’Carroll D et al. (2001). Nature 412: 561–565. Harbour JW, Dean DC. (2000). Genes Dev 14: 2393–2409. Norton JD, Deed RW, Craggs G, Sablitzky F. (1998). Trends Harbour JW, Luo RX, Dei Santi A, Postigo AA, Dean DC. Cell Biol 8: 58–65. (1999). Cell 98: 859–869. Novitch BG, Mulligan GJ, Jacks T, Lassar AB. (1996). J Cell Hiebert SW, Chellappan SP, Horowitz JM, Nevins JR. (1992). Biol 135: 441–456. Genes Dev 6: 177–185. Novitch BG, Spicer DB, Kim PS, Cheung WL, Lassar AB. Higgins C, Chatterjee S, Cherington V. (1996). J Virol 70: (1999). Curr Biol 9: 449–459. 745–752. Obeyesekere MN, Herbert JR, Zimmerman SO. (1995). Hinchcliffe EH, Li C, Thompson EA, Maller JL, Sluder G. Oncogene 11: 1199–1205. (1999). Science 283: 851–854. Pardee AB. (1974). Proc Natl Acad Sci USA 71: 1286–1290. Howes KA, Ransom N, Papermaster DS, Lasudry JG, Albert Pradhan S, Kim GD. (2002). EMBO J 21: 779–788. DM, Windle JJ. (1994). Genes Dev 8: 1300–1310. Ray SK, Arroyo M, Bagchi S, Raychaudhuri P. (1992). Mol Huang HJ, Yee JK, Shew JY, Chen PL, Bookstein R, Cell Biol 12: 4327–4333. Friedmann T et al. (1988). Science 242: 1563–1566. Ren B, Cam H, Takahashi Y, Volkert T, Terragni J, Young Huh MS, Parker MH, Scime A, Parks R, Rudnicki MA. RA et al. (2002). Genes Dev 16: 245–256. (2004). J Cell Biol 166: 865–876. Riley DJ, Lee EY, Lee WH. (1994). Annu Rev Cell Biol 10: Iavarone A, Garg P, Lasorella A, Hsu J, Israel MA. (1994). 1–29. Genes Dev 8: 1270–1284. Robanus-Maandag E, Dekker M, van der Valk M, Carrozza Jacks T, Fazeli A, Schmitt EM, Bronson RT, Goodell MA, ML, Jeanny JC, Dannenberg JH et al. (1998). Genes Dev 12: Weinberg RA. (1992). Nature 359: 295–300. 1599–1609. Ji A, Dao D, Chen J, MacLellan WR. (2003). Gene 318: 35–43. Robertson KD, Ait-Si-Ali S, Yokochi T, Wade PA, Jones PL, Johnson DG, Schwarz JK, Cress WD, Nevins JR. (1993). Wolffe AP. (2000). Nat Genet 25: 338–342. Nature 365: 349–352. Rodier G, Makris C, Coulombe P, Scime A, Nakayama K, Kato J, Matsushime H, Hiebert SW, Ewen ME, Sherr CJ. Nakayama KI et al. (2005). J Cell Biol 168: 55–66. (1993). Genes Dev 7: 331–342. Ross JF, Liu X, Dynlacht BD. (1999). Mol cell 3: 195–205. Krutzfeldt M, Ellis M, Weekes DB, Bull JJ, Eilers M, Vivanco Sage C, Huang M, Karimi K, Gutierrez G, Vollrath MA, MD et al. (2005). Mol Cell 18: 213–224. Zhang DS et al. (2005). Science 307: 1114–1118. Lacey KR, Jackson PK, Stearns T. (1999). Proc Natl Acad Sci Sandmoller A, Meents H, Arnold HH. (1996). Mol Cell Biol USA 96: 2817–2822. 16: 5846–5856. Lasorella A, Noseda M, Beyna M, Yokota Y, Iavarone A. Schneider JW, Gu W, Zhu L, Mahdavi V, Nadal-Ginard B. (2000). Nature 407: 592–598. (1994). Science 264: 1467–1471. Lee EY, Cam H, Ziebold U, Rayman JB, Lees JA, Dynlacht Schwarz JK, Devoto SH, Smith EJ, Chellappan SP, Jakoi L, BD. (2002). Cancer Cell 2: 463–472. Nevins JR. (1993). EMBO J 12: 10131020. Lee EY, Chang CY, Hu N, Wang YC, Lai CC, Herrup K et al. Scime A, Grenier G, Huh MS, Gillespie MA, Bevilacqua L, (1992). Nature 359: 288–294. Harper ME et al. (2005). Cell Metab 2: 283–295. Lee EY, Hu N, Yuan SS, Cox LA, Bradley A, Lee WH et al. Scott PH, Cairns CA, Sutcliffe JE, Alzuherri HM, McLees A, (1994). Genes Dev 8: 2008–2021. Winter AG et al. (2001). J Biol Chem 276: 1005–1014. Lee WH, Bookstein R, Hong F, Young LJ, Shew JY, Lee EY. Sherr CJ. (1996). Science 274: 1672–1677. (1987). Science 235: 1394–1399. Shirodkar S, Ewen ME, DeCaprio JA, Morgan J, Livingston Luo RX, Postigo AA, Dean DC. (1998). Cell 92: 463–473. DM, Chittenden T. (1992). Cell 68: 157–166. Maandag EC, van der Valk M, Vlaar M, Feltkamp C, O’Brien Slack RS, El-Bizri H, Wong J, Belliveau DJ, Miller FD. J, van Roon M et al. (1994). EMBO J 13: 4260–4268. (1998). J Cell Biol 140: 1497–1509. MacLellan WR, Xiao G, Abdellatif M, Schneider MD. (2000). Takahashi Y, Rayman JB, Dynlacht BD. (2000). Genes Dev Mol Cell Biol 20: 8903–8915. 14: 804–816. Magnaghi-Jaulin L, Groisman R, Naguibneva I, Robin P, Taubert S, Gorrini C, Frank SR, Parisi T, Fuchs M, Chan HM Lorain S, Le Villain JP et al. (1998). Nature 391: 601–605. et al. (2004). Mol Cell Biol 24: 4546–4556. Mantela J, Jiang Z, Ylikoski J, Fritzsch B, Zacksenhaus E, Taya Y. (1997). Trends Biochem Sci 22: 14–17. Pirvola U. (2005). Development 132: 2377–2388. Trimarchi JM, Fairchild B, Verona R, Moberg K, Andon N, Mayol X, Grana X, Baldi A, Sang N, Hu Q, Giordano A. Lees JA. (1998). Proc Natl Acad Sci USA 95: 2850–2855. (1993). Oncogene 8: 2561–2566. Trimarchi JM, Lees JA. (2002). Nat Rev Mol Cell Biol 3: McDonnell TJ, Montes de Oca Luna R, Cho S, Amelse LL, 11–20. Chavez-Reyes A, Lozano G. (1999). J Pathol 188: 322–328. Tsai KY, Hu Y, Macleod KF, Crowley D, Yamasaki L, Jacks Miyake S, Sellers WR, Safran M, Li X, Zhao W, Grossman T. (1998). Mol Cell 2: 293–304. SR et al. (2000). Mol Cell Biol 20: 8889–8902. Vandel L, Nicolas E, Vaute O, Ferreira R, Ait-Si-Ali S, Moberg K, Starz MA, Lees JA. (1996). Mol Cell Biol 16: Trouche D. (2001). Mol Cell Biol 21: 64846494. 1436–1449. Weinmann AS, Yan PS, Oberley MJ, Huang TH, Farnham PJ. Narita M, Nunez S, Heard E, Narita M, Lin AW, Hearn SA (2002). Genes Dev 16: 235–244. et al. (2003). Cell 113: 703–716. Wells J, Boyd KE, Fry CJ, Bartley SM, Farnham PJ. (2000). Nevins JR. (1992). Science 258: 424–429. Mol Cell Biol 20: 5797–5807.

Oncogene RB, the conductor that orchestrates life, death and differentiation L Khidr and P-L Chen 5219 Wells J, Graveel CR, Bartley SM, Madore SJ, Farnham PJ. Zacksenhaus E, Jiang Z, Chung D, Marth JD, Phillips RA, (2002). Proc Natl Acad Sci USA 99: 3890–3895. Gallie BL. (1996). Genes Dev 10: 3051–3064. Windle JJ, Albert DM, O’Brien JM, Marcus DM, Disteche Zhang HS, Dean DC. (2001). Oncogene 20: CM, Bernards R et al. (1990). Nature 343: 665–669. 3134–3138. Woo MS, Sanchez I, Dynlacht BD. (1997). Mol Cell Biol 17: Zhang J, Gray J, Wu L, Leone G, Rowan S, Cepko CL et al. 3566–3579. (2004). Nat Genet 36: 351–360. Wu L, de Bruin A, Saavedra HI, Starovic M, Trimboli A, Ziebold U, Reza T, Caron A, Lees JA. (2001). Genes Dev 15: Yang Y et al. (2003). Nature 421: 942–947. 386–391.

Oncogene