Pocket Proteins and Cell Cycle Control

Pocket proteins and cell cycle control

David Cobrinik*,1

1Dyson Vision Research Institute and Department of Ophthalmology, Weill Medical College of Cornell University, 1300 York Avenue, LC303, New York, NY 10021, USA

The retinoblastoma protein (pRB) and the pRB-related hit hypothesis and validated the tumor suppressor gene p107 and p130 comprise the ‘pocket protein’ family of cell concept. cycle regulators. These proteins are best known for their While the cloning of RB concluded the search for the roles in restrainingthe G1–S transition throughthe retinoblastoma suppressor, it opened a new chapter in regulation of E2F-responsive genes. pRB and the p107/ cancer biology focusing on the new gene’s function. p130 pair are required for the repression of distinct sets of Unexpectedly, RB was found to be mutated in many genes, potentially due to their selective interactions with additional cancers (Harbour et al., 1988; Lee et al., 1988; E2Fs that are engaged at specific promoter elements. In Horowitz et al., 1989), and its encoded protein, pRB, addition to regulating E2F-responsive genes in a reversible was widely expressed and regulated in a manner manner, pocket proteins contribute to silencingof such consistent with its having a general cell cycle role (Lee genes in cells that are undergoing senescence or differ- et al., 1987; Buchkovich et al., 1989; DeCaprio et al., entiation. Pocket proteins also affect the G1–S transition 1989; reviewed in Cobrinik et al., 1992). Additional through E2F-independent mechanisms, such as by inhibit- studies revealed that pRB acted together with two ingCdk2 or by stabilizingp27 Kip1, and they are implicated related proteins, p107 and p130, to regulate prolifera- in the control of G0 exit, the spatial organization of tion in most if not all cell types (Cobrinik, 1996; Dyson, replication, and genomic rereplication. New insights into 1998; Nevins, 1998). It is now evident that pRB, p107, pocket protein regulation have also been obtained. Kinases and p130, together known as the ‘pocket proteins,’ are previously thought to be crucial to pocket protein central participants in a gene regulatory network that phosphorylation have been shown to be redundant, and governs the cellular response to antimitogenic signals, new modes of phosphorylation and dephosphorylation and whose deregulation constitutes one of the hallmarks have been identified. Despite these advances, much of cancer (Hanahan and Weinberg, 2000). remains to be learned about the pocket proteins, particularly with regard to their developmental and tumor suppressor functions. Thus continues the story of the pocket proteins and the cell cycle. Pocket proteins, E2Fs, and the G1–S transition Oncogene (2005) 24, 2796–2809. doi:10.1038/sj.onc.1208619 Many aspects of pRB’s cell cycle function were outlined Keywords: pocket proteins; pRB; p107; p130; E2F; cell about 10 years ago (Weinberg, 1995). By then, pRB was cycle known to be a nuclear protein that is minimally (or hypo-) phosphorylated in resting or G0 cells, to be increasingly phosphorylated during progression through G1, and to be maintained in a hyperphosphorylated More than 50 yeas ago, a children’s tumor was realized state until late in mitosis (Figure 1). pRB phosphoryla- to occur sporadically in some patients, but to be tion was recognized to be directed by mitogenic signals inherited in others (Falls and Neel, 1951). The tumor that converge on the cell cycle machinery, composed of was retinoblastoma, and although it was quite rare, its D cyclins acting with Cdk4 or Cdk6 in early and mid- uncommon genetics offered a handle with which to G1, and composed of E cyclins acting with Cdk2 in late graspthe cause of one form of human cancer. It was G1. Hypophosphorylated pRB was further recognized hypothesized that retinoblastomas begin as a result of to inhibit proliferation through its association with two mutations, or ‘hits’ (Knudson, 1971), which were other proteins. A principal target appeared to be the later surmized to disable the two alleles of a tumor E2F transcription factor, which was known to regulate suppressor gene (Cavenee et al., 1983). The cloning of proliferation-associated genes (Nevins, 1992). the retinoblastoma (RB) gene (Friend et al., 1986) and Early studies suggested that hypophosphorylated the identification of biallelic RB mutations in retino- pRB binds and inhibits the E2F transactivation domain, blastoma tumors (Dunn et al., 1988) supported the two and that pRB phosphorylation releases E2F to allow the expression of genes that mediate S phase entry (Flemington et al., 1993; Helin et al., 1993). This model illustrated how loss or mutation of RB leads to *Correspondence: D Cobrinik; E-mail: [email protected] deregulation of E2F and inappropriate proliferation. It Pocket proteins and cell cycle control D Cobrinik 2797 2002). This demonstrated that pocket protein-mediated pRB pRB repression of such genes is important for inhibiting the X E7 G1–S transition. More recently, pocket proteins and E2Fs were found to govern the expression of a far wider network of genes than those controlling S phase entry and progression. Specifically, they also regulate genes that control entry pRB p16 Cyc D into and progression through mitosis, and they coordi- E2F Cdk4/6 nate these cell cycle programs with transcriptional programs mediating checkpoint control, the DNA G0 damage response, apoptosis, differentiation, and development (Ishida et al., 2001; Muller et al., 2001; Polager M G1 p p et al., 2002; Ren et al., 2002; Weinmann et al., 2002; E2F pRB Dimova et al., 2003). Thus, the deregulation of E2Fs has p p far wider consequences than simply increased prolifera- pRB tion, and may also account for the propensity of tumor cells to undergo apoptosis, to have checkpoint defects, to have an altered DNA damage response, and to resist G2 S the effects of differentiation agents. The expanding influence of the pocket protein–E2F network has been the subject of several recent reviews p p (Cam and Dynlacht, 2003; Blais and Dynlacht, 2004; pRB Bracken et al., 2004; Liu et al., 2004; Dimova and Dyson, 2005), and is not further detailed here. Instead, Figure 1 A simplified view of the pRB–E2F pathway. In this this review focuses on mechanisms through which rendition, pRB binds and inhibits E2F in G0 and early G1. In specific pocket proteins regulate different E2F-respon- proliferating cells, pRB phosphorylation by cyclin D-Cdk4/Cdk6 sive genes, particularly the cyclin E1 and cyclin A2 releases E2F, which then induces genes that mediate S phase entry. (hereafter, ‘cyclin E’ and ‘cyclin A’) effectors of the In tumor cells, the pRB–E2F interaction is disrupted by mutation of the RB gene (X), by pRB binding to DNA tumor virus G1–S transition. E2F-independent functions of pocket oncoproteins such as human papilloma virus E7, or by inappropri- proteins, roles of pocket proteins in other cell cycle ate pRB phosphorylation due to overexpression of D cyclins, loss transitions, new means of pocket protein regulation, and of the p16INK4A inhibitor of Cdk4/Cdk6, or mutation or over- the partially redundant behaviors of pocket proteins in expression of the Cdk4 or Cdk6 genes cell cycle control and development are also discussed. also illustrated how other oncogenic insults deregulate A dynamic pocket protein–E2F network E2F through their effects on pRB (Figure 1). For example, this could be accomplished by DNA tumor In mammalian cells, the pocket protein and E2F families virus proteins that bind to pRB and displace E2F, or by can be divided into subgroups whose members have inappropriate pRB phosphorylation due to overexpres- similar functions (Figure 2). Among the pocket proteins, sion of D cyclins, loss of the p16INK4A inhibitor of Cdk4 p107 and p130 are more related to one another than and Cdk6, or mutation or amplification of the Cdk4 or either is related to pRB, in terms of both their structure Cdk6 genes (Weinberg, 1995; Sherr, 1996). While this and their function (Classon and Dyson, 2001). Like model remains useful for conceptualizing pRB’s core cell pRB, the p107 and p130 proteins bind E2Fs, and their cycle function, we now understand the pRB–E2F phosphorylation by G1 cyclins and Cdks results in E2F pathway to be far more elaborate. dissociation. However, p107 and p130 preferentially A fundamental addition to the early model came with bind different E2Fs from those bound by pRB, and are the recognition that pRB and E2F are members of required for the regulation of distinct E2F-responsive protein families, consisting of the pRB, p107, and p130 genes. Moreover, whereas pRB is commonly expressed pocket proteins, E2F1–7, and the DP1 and DP2-binding in both proliferating and nonproliferating cells, p107 is partners of E2F1–6 (Cobrinik, 1996; Dyson, 1998; most prominent in proliferating cells and p130 is most Nevins, 1998; Dimova and Dyson, 2005). In addition, prominent in arrested cells (Classon and Dyson, 2001). it was realized that pocket proteins not only inhibit E2F- Among the E2F family, E2F1, E2F2, and E2F3a are mediated transactivation but also function together with strong transcriptional activators that are preferentially E2Fs to actively repress transcription and inhibit G1–S bound and inhibited by pRB (Figure 2). E2F1-3a are progression (Zhang et al., 1999; He et al., 2000). generally absent or expressed at low levels in quiescent Moreover, the ectopic expression of E2F-responsive cells, and are induced to high levels and stimulate gene genes such as cyclin E1, cyclin A2, and Emi1 (which expression in late G1 (Dyson, 1998). A distinct E2F3b stabilizes cyclin A) was found to overcome pocket isoform is expressed throughout the cell cycle, and is protein-dependent G0/G1 arrests (Alevizopoulos et al., implicated in forming transcriptional repressor com- 1997; Leng et al., 1997; Lukas et al., 1997; Hsu et al., plexes with Rb in G0 (Leone et al., 2000), and in

Oncogene Pocket proteins and cell cycle control D Cobrinik 2798

pRB p107 p130

E2F1 E2F2 E2F3a E2F3b E2F4 E2F5 E2F6 E2F7

Activator E2Fs Repressor E2Fs

Figure 2 Interactions among pocket proteins and E2F transcription factors. Pocket proteins can be subdivided into the pRB and p107/p130 groups. pRB preferentially binds to the activator E2Fs, E2F1, E2F2, and E2F3a; as well as to E2F3b, which may function mainly as a repressor. p107 and p130 preferentially bind the repressor E2Fs, E2F4 and E2F5. E2F6 and E2F7 form transcriptional repressor complexes but do not bind pocket proteins. The DP1 and DP2 proteins form heterodimers with E2F1–6 to allow binding to DNA (not shown), whereas E2F7 binds as a homodimer

forming pocket protein-independent repressor com- provide for several different states of E2F-responsive plexes in both resting and normally proliferating cells promoters in the different cell cycle phases (Figure 3). (Aslanian et al., 2004; Ginsberg, 2004). In contrast, Thus, in G0 and early G1 (Figure 3A1), p107 and p130 E2F4 and the generally less abundant E2F5 are poor form repressor complexes in conjunction with E2F4 or transcriptional activators, in part owing to their lack of E2F5 at most if not all E2F-responsive promoters. a nuclear localization signal. E2F4 and 5 are expressed Meanwhile, pRB is thought to bind E2F1–3 either at, or throughout the cell cycle, but in G0 and early G1 they sequestered away from, such promoters. A third are bound and recruited to the nucleus by p107 and arrangement may involve binding of E2F6 together p130, and form transcriptional repressor complexes. with polycomb group and other repressor proteins Two other family members, E2F6 and E2F7, do not (Figure 3A2) (Ogawa et al., 2002). In late G1 bind pocket proteins, and repress E2F-responsive genes (Figure 3B), pocket proteins are generally phosphory- through other means (Dimova and Dyson, 2005). lated and dissociate from E2Fs. E2F4 and 5 then As noted above, pocket proteins appear to mediate relocate to the cytoplasm, and the promoters generally G0/G1 cell cycle arrest through their repression of E2F- bind E2F1–3. Interestingly, at some promoters, E2F1–3 regulated genes. They are believed to accomplish this by may bind to sites different from those vacated by E2F4 recruiting a variety of chromatin-modifying activities to and 5 (Araki et al., 2003; Zhu et al., 2004). For many E2F-responsive promoters (Frolov and Dyson, 2004). promoters, the binding of E2F1–3 correlates with For example, chromatin immunoprecipitation (ChIP) recruitment of histone acetyltransferases (HATs), in- analyses demonstrated that pocket proteins recruit type creased histone acetylation, and transcriptional activa- I histone deacetylases (HDACs), resulting in removal of tion (Figure 3B1) (Takahashi et al., 2000; Rayman et al., acetyl groups from histones H3 and H4 and in a 2002; Taubert et al., 2004). However, expression of compacted chromatin structure that is refractory to genes such as cyclin A2, cdc2,andcyclin B1 is delayed, transcription initiation (Takahashi et al., 2000; Ferreira and may require the subsequent binding of other et al., 2001; Morrison et al., 2002; Rayman et al., 2002). transcription factors, such as B-myb (Figure 3B2 and The recruitment of HDACs is indirect (Lai et al., 2001) C4) (Zhu et al., 2004). After entry into S phase, E2F1–3 and accompanied by recruitment of other corepressors, continue to bind and probably activate many promoters chromatin remodeling factors, and histone methyltrans- (Figure 3C1). However, E2F1–3 bind other promoters ferases (Frolov and Dyson, 2004). Pocket proteins may only until the G1-S transition. These promoters may also repress transcription through HDAC-independent then be activated by other factors, as in the case of the mechanisms, both in Drosophila (Taylor-Harding B-myb gene (Figure 3C2) (Takahashi et al., 2000), or et al., 2004) and in mammalian cells (Strobeck et al., may be bound and repressed by E2F7, as for Cdc6 and 2000; Zhang et al., 2000), as discussed for cyclin A E2F1 (Figure 3C3) (Di Stefano et al., 2003). States or below. ChIP analyses further showed that in late G1, timing other than those displayed in Figure 3 are also pocket protein repressor complexes dissociate from possible (Pediconi et al., 2003; Aslanian et al., 2004), E2F-responsive promoters, whereas activating E2Fs and the situation is further embellished by transcription bind such promoters and restore histone acetylation factors that bind to promoters in cooperation with (Takahashi et al., 2000; Rayman et al., 2002; Taubert specific E2Fs or pocket proteins (Karlseder et al., 1996; et al., 2004). Lin et al., 1996; Le Cam et al., 1999; Chen et al., 2002; The specific interactions of E2Fs and pocket proteins, Schlisio et al., 2002; Giangrande et al., 2004; Zhu et al., together with their distinct expression, subcellular 2004). The various contemporaneous states of different localization, and promoter-binding patterns (as defined E2F-responsive promoters may confer individualized by ChIP analyses), have suggested several related regulation of promoter activity. models of the pocket protein–E2F network (Stevaux An important unresolved aspect of this system is and Dyson, 2002; Trimarchi and Lees, 2002; Cam and whether E2F1–3 and pRB participate in promoter Dynlacht, 2003; Bracken et al., 2004). The models repression in G0 and early G1. In ChIP analyses

Oncogene Pocket proteins and cell cycle control D Cobrinik 2799 a HDAC pRB E2F1-3 HDAC p107 PcG p130

G0/early G1 E2F4/5 E2F6 A1 A2 b p p pRB p p HAT p107 Late G1 p130 E2F1-3 E2F1-3 B1 B2

c HAT HAT HAT

S/G2 E2F1-3 TF E2F7 E2F1-3 TF C1 C2 C3 C4

Figure 3 Alternative pocket protein–E2F complexes at E2F-responsive promoters. The ﬁgure shows pocket protein–E2F complexes that have been detected at promoters by ChIP analyses in G0 and early G1, in late G1, and in S or G2 phase cells. Additional components of the complexes, such as DP proteins and various cofactors and chromatin modifying proteins, have been omitted for clarity. The protein arrangements do not necessarily reﬂect actual protein–protein interactions. (a) G0 and early G1.A1: E2F4, p130, and HDAC1 (or their counterparts E2F5, p107, and HDAC2) have been detected at many E2F-responsive promoters, and HDAC1 recruitment required p107/p130 for cyclin A, E2F1, cdc2, and B-myb (Takahashi et al., 2000; Wells et al., 2000; Rayman et al., 2002). E2F1–3 and pRB have been detected at few promoters, and recruitment of HDAC1 required pRB in the case of cyclin E (Wells et al., 2000; Morrison et al., 2002). A2: E2F6 and a polycomb group (PcG) protein have also been detected at promoters in G0 cells (Ogawa et al., 2002). (b) Late G1: E2F1–3 have been detected at all E2F-responsive promoters examined to date. B1: For several promoters, E2F1–3 were shown to recruit histone acetyltransferases, (HATs) in conjunction with promoter activation (Taubert et al., 2004). B2: Other promoters, such as cyclin A2, cdc2, and cyclin B1, bound E2F1–3 in late G1 but were not active until S or G2 (Rayman et al., 2002; Zhu et al., 2004). (c) S and G2: Among promoters that were active in late G1, some continue to bind E2F1–3 and HATs (C1); others no longer bind E2F1–3 but may be activated by other transcription factors (TFs) (C2); and others are bound and repressed by E2F7 (C3) (Takahashi et al., 2000; Di Stefano et al., 2003). E2F-responsive promoters that bound E2F1–3 but were inactive in late G1 may be induced through the appearance of other transcription factors, such as B-myb, in the case of cdc2 and CyclinB1 (C4) (Zhu et al., 2004)

of mitogen-deprived cells, all cell cycle-regulated repressor complexes by p107 and p130. However, this E2F-responsive promoters examined to date bound arrangement does not result in redundant control of E2F4 and either p107 or p130. However, relatively E2F-responsive genes by pRB and the p107/p130 pair. few promoters detectably bound E2F1–3 or pRB, Indeed, many – and possibly most – E2F-responsive and, with the exception of cyclin E, this binding genes are deregulated in p107/p130-deficient cells seemed to occur at low levels (Takahashi et al., (Hurford et al., 1997; Mulligan et al., 1998; Ren et al., 2000; Wells et al., 2000; Morrison et al., 2002; Rayman 2002). A distinct set consisting of cyclin E and p107 was et al., 2002). These findings are consistent with the deregulated in pRB-deficient cells in one study (Hurford notion that pRB directly participates in the repression of et al., 1997), whereas overlapping sets of genes were only a limited set of E2F-responsive genes in G0 and deregulated in pRB-deficient versus p107/p130-deficient early G1 cells. cells in a more recent report (Black et al., 2003). The basis for this discrepancy is unclear, although subtle differences in the growth arrest regimens, and hence Roles for specific pocket proteins in the regulation differences in the G0 states, may be responsible. of E2F-responsive genes Interestingly, despite the overlap in pRB- and p107/ p130-regulated genes in the Black et al. (2003) study, a As shown in Figure 3, E2F-responsive genes are gene expression set that was characteristic of pRB- generally regulated in G0/G1 through the sequestration deficient fibroblasts was also characteristic of pRB- of activating E2Fs by pRB, and through formation of deficient pituitary and thyroid tumor cells, suggesting

Oncogene Pocket proteins and cell cycle control D Cobrinik 2800 that pRB has similar transcriptional effects in diverse complexes. However, it is less evident how the cyclin E cell types. and p107 promoters are deregulated due to the loss of Currently, the features that dictate whether a pRB, despite that they, too, bind E2F4-p107/p130 promoter is deregulated by loss of pRB, or by the complexes. Nevertheless, recent analyses have provided loss of p107/p130, have not been defined. Clearly, clues to the role of pRB in the regulation of cyclin E, the loss of E2F4–p107/p130 repressor complexes which is arguably one of the more crucial E2F- contributes to the deregulation of E2F-responsive responsive genes. Cyclin E is normally expressed only promoters in p107/p130-deficient cells (Stevaux and in late G1, is correctly regulated in p107/p130 deficient Dyson, 2002; Cam and Dynlacht, 2003). However, it cells, but is derepressed in G0 and early G1 in pRB- is less clear why E2F4–pRB repressor complexes, deficient cells (Herrera et al., 1996; Hurford et al., 1997). which are often detected in vitro, fail to regulate such The periodic expression of cyclin E depends upon two promoters in place of the p107/p130 repressor com- E2F elements located within 40 base pairs of the major plexes in vivo. Indirect evidence suggests that E2F4–pRB transcription start site (Figure 4). One of these, termed complexes may have low affinity for E2F-responsive CERM, is necessary for repression of cyclin E in G0 and promoters, since sites that bound E2F4–p130–HDAC1 early G1. CERM consists of a variant E2F-binding site complexes in wild-type cells failed to bind E2F4, and an AT-rich region, and cooperatively binds E2F4 pRB, or HDAC1 in p107/p130-deficient cells (Rayman (or 5) and an unidentified AT-rich DNA-binding protein et al., 2002). (Le Cam et al., 1999). E2F4 (or 5) binds to CERM as In some cases, a specific pocket protein may be needed part of a complex termed CERC, which includes p107 or to regulate a promoter due to its selective recruitment by p130, an HDAC activity, and other proteins. Upon p107 other transcription factors. For example, p107 was and p130 phosphorylation in late G1, CERM is vacated found to mediate the TGFb-induced repression of c- coinciding with promoter activation. In p107/p130- Myc due to its recruitment by Smad3 (Chen et al., 2002). deficient cells, pRB can substitute for p107/p130 in the Here, the involvement of p107 was dictated by a CERC to maintain repression in G0 and early G1 (Le bipartite Smad-E2F promoter element, by the specific Cam et al., 1999; Polanowska et al., 2001; Fabbrizio binding of Smad3 to E2F4, and by the specific binding et al., 2002). of both Smad3 and E2F4 to p107. p107 or p130 might A second element that regulates cyclin E is a bipartite be recruited to other E2F-responsive promoters in a ‘E2F-Sp1’ site. This element is crucial to cyclin E similar way, particularly in light of evidence that E2F4 expression and is constitutively occupied during the often binds to promoters in cooperation with other cell cycle (Botz et al., 1996; Le Cam et al., 1999). transcription factors (Le Cam et al., 1999; Araki et al., Because Sp1 binds E2F1–3, but not E2F4 or 5 2003; Zhu et al., 2004). (Karlseder et al., 1996), this element may cooperatively bind Sp1 (or a related protein) and E2F1–3, and may A crucial role for pRB in the regulation of cyclin E thereby attract pRB in G0 and early G1 cells. ChIP analyses support this idea, as E2F1–3 and pRB The above illustrates how many promoters may be de- were readily detected at the cyclin E promoter regulated due to the loss of E2F4–p107/p130 repressor (Wells et al., 2000; Morrison et al., 2002). ChIP assays

HP1 HDAC HDAC PRMT5 pRB p107 G0 /early G1 SUV39H p130 CERC E2F1-3 Sp1 E2F4/5

"E2F-Sp1" AT-rich E2F

CERM

Late G1 E2F1-3 Sp1

"E2F-Sp1" AT-rich E2F

Figure 4 Model for pRB-dependent repression of cyclin E. The repression of cyclin E in G0 and early G1 is mediated by a site termed ‘CERM,’ which binds an AT-rich DNA-binding protein, E2F4 or 5, p130 or p107, and other proteins that together constitute the CERC. A second regulatory site, termed the ‘E2F-Sp1’ element, is required for cyclin E expression and may cooperatively bind Sp1 (or a related protein) and E2F1–3, and thereby attract pRB and associated repressor proteins (Le Cam et al., 1999; Nielsen et al., 2001; Polanowska et al., 2001; Fabbrizio et al., 2002; Morrison et al., 2002). pRB can substitute for p107/p130 in the CERC, and may there by maintain normal regulation in p107/p130-deficient cells (Le Cam et al., 1999). In contrast, it is suggested that p107 and p130 might not substitute for pRB at the E2F-Sp1 element (except under unusual circumstances, Lee et al., 2002), leading to E2F derepression in pRB-deficient as well as in E2F1–3-deficient cells (Herrera et al., 1996; Wu et al., 2001; Saavedra et al., 2003)

Oncogene Pocket proteins and cell cycle control D Cobrinik 2801 further showed that pRB is required for the recruitment Pocket protein-mediated repression of cyclin A through of HDAC1 as well as for histone deacetylation within HDAC-independent mechanisms the cyclin E transcription initiation region (Morrison et al., 2002). Whereas pRB is expected to dissociate from In addition to being regulated by covalent histone the cyclin E promoter when phosphorylated in late G1, modifications, transcription is regulated by nucleosome the E2F-Sp1 element was found to remain occupied and sliding and assembly processes known as chromatin to be crucial to transcriptional activation (Botz et al., remodeling (Narlikar et al., 2002). Pocket proteins may 1996; Le Cam et al., 1999), suggesting that activating coordinate histone deacetylation with chromatin remo- E2Fs remain. deling by concurrently binding an HDAC complex and Interestingly, pRB-mediated histone deacetylation a chromatin remodeling machine termed the hSWI/SNF may prepare the cyclin E promoter for additional complex (Zhang et al., 2000). hSWI/SNF complexes modifications. For example, pRB also recruits the contain either the Brahma (Brm) ATPase or the Brm- SUV39H1 histone H3 methyltransferase (or a related related gene product (BRG1) (Narlikar et al., 2002), methyltransferase), promotes methylation of histone H3 both of which may interact with each of the pocket lysine 9 (H3K9), and attracts the repressor protein, HP1 proteins (Dunaief et al., 1994; Strober et al., 1996). (Nielsen et al., 2001). However, it is currently unclear In keeping with the notion that pocket proteins whether pRB promotes H3K9 methylation of the cyclin coordinate histone deacetylation with chromatin remo- E promoter in a reversible G0 state (quiescence), or does deling, both HDAC and Brm/BRG1 activities were so only in cells that have undergone permanent cell cycle required for a constitutively active pRB to repress many exit (see ‘Pocket proteins and silencing of E2F-respon- E2F-responsive genes. However, HDAC activity was sive genes,’ below). dispensable, but Brm/BRG1 were still required, for pRB Together, the above observations imply that pRB to repress other genes such as cyclin A, and to induce a mediates the cyclical repression of cyclin E due to cell cycle arrest mediated in part by cyclin A repression its recruitment of HDACs and potentially histone (Strobeck et al., 2000; Zhang et al., 2000; Sever- methyltransferases. However, since p107 and p130 Chroneos et al., 2001; Siddiqui et al., 2003). This can also bind and recruit such proteins (Ferreira suggests that pRB may more generally elicit repression et al., 2001; Rayman et al., 2002; Nicolas et al., 2003), through Brm/BRG1 than through HDACs. these functions alone seem unlikely to underlie pRB may elicit HDAC-independent, but Brm/BRG1- pRB’s special role in cyclin E repression. Several dependent, repression of cyclin A through two mechan- lines of evidence suggest that pRB may be crucial isms. In fibroblasts, pRB-mediated repression of cyclin to this repression due to the cooperative binding of A was associated with an altered chromatin structure at E2F1–3 and Sp1 (or a related protein) to this promoter, the cyclin A promoter (Siddiqui et al., 2003). Con- and the need for pRB to bind E2F1–3 and form a cordantly, the repression of cyclin A following serum repressor complex. For example, cyclin E was dere- starvation required both pRB and Brm, and was pressed not only by loss of pRB but also by the associated with Brm-dependent formation of nucleo- combined loss of E2F1–3, suggesting that these E2Fs somes in the transcription initiation region (Philips participate in a repressor complex (Wu et al., 2001). et al., 1998; Coisy et al., 2004). These findings suggest Cyclin E was also derepressed in cells lacking only that pRB represses cyclin A through Brm-mediated E2F3, perhaps reflecting that E2F3b is the main E2F1–3 nucleosome remodeling in fibroblasts. member that is expressed after mitogen deprivation In another study, p16 was used to induce pocket (Leone et al., 2000; Saavedra et al., 2003). Moreover, in protein activity in osteosarcoma cells (Dahiya et al., pRB-deficient cells, the additional loss of E2F4 restored 2001). Here, the repression of cyclin A was associated cyclin E repression, while inducing the formation of with recruitment of Brm to the promoter, consistent noncanonical p107–E2F3 and p130–E2F1 complexes with a role for chromatin remodeling in this setting as (apparently by increasing p107 and p130 availability) well. In addition, however, cyclin A repression required (Lee et al., 2002). This striking result implies that the polycomb group protein HPC2, and was accom- p107/p130 can substitute for pRB at the cyclin panied by HPC2 binding to the cyclin A promoter. E promoter, but only when available in sufficient Overexpression experiments suggested that HPC2 might quantities to bind E2F1–3. bind the promoter through a complex of E2F, pRB, and In summary, a variety of data suggests that pRB a corepressor called CtBP (Dahiya et al., 2001, 2003), may be crucial to the repression of cyclin E due to although it is unclear whether the endogenous HPC2 the specific binding of E2F1–3 to the promoter is recruited to the promoter in this way. E2F-Sp1 element, and the preferential binding of While the above findings show that Brm is crucial to pRB to these E2F species. Underlying this model is the repression of cyclin A, further information is needed the more general notion that distinct E2F–pocket to understand the HDAC-independent effects of pocket protein complexes may be positioned at particular proteins at this promoter. For example, cyclin A E2F recognition sites through the cooperative interac- repression has not been shown to require direct binding tions of specific E2Fs with other transcription factors. of pRB to Brm, and it remains possible that pocket In turn, the unique arrangements of such factors may proteins indirectly engage Brm activity. Moreover, the dictate the distinctive regulatory features of individual specific roles of pocket proteins and E2Fs at specific E2F-responsive genes. cyclin A promoter sites remain undefined. p107/p130

Oncogene Pocket proteins and cell cycle control D Cobrinik 2802 were needed for cyclin A repression in one series of under both circumstances, and it is possible that the experiments (Hurford et al., 1997; Rayman et al., 2002); pRB binding reflected increased expression of pRB or its pRB was needed in another series (Philips et al., 1998; cognate E2Fs during senescence. Likewise, an E1A Coisy et al., 2004); and both pRB and the p107/p130 mutant that inactivates the entire pocket protein family pair were needed in a third analysis (Black et al., 2003). blocked the silencing of cyclin A, whereas knockdown of Perhaps these disparate results reflect the utilization of only pRB had a comparatively small effect. These different pocket proteins and cyclin A repression findings appear consistent with the notion that each of mechanisms under different experimental conditions. the pocket proteins can contribute to silencing, in Notably, pocket protein-mediated repression in Droso- keeping with each of their abilities to bind to SUV39H1 phila generally did not require HDAC activity (Taylor- (Nielsen et al., 2001, 2003; Vandel et al., 2001; Nicolas Harding et al., 2004). Further study of this system may et al., 2003). Nonetheless, specific pocket proteins may provide insight into HDAC-independent repression of be required to silence specific promoters in specific E2F-responsive genes in mammals, as well as in flies. settings, depending on the architecture of the promoters and on the relative pocket protein expression levels. A role for pRB in the silencing of cyclin E, perhaps Pocket proteins and silencingof E2F-responsive genes analogous to pRB’s role in the repression of cyclin E, could be an instance in which a specific pocket protein is The finding that pRB elicits H3K9 methylation at the utilized. cyclin E promoter (Nielsen et al., 2001) has raised the Given that pocket proteins mediate both the repres- issue of whether this modification is used to repress sion of E2F-responsive genes during quiescence and the E2F-responsive genes in a reversible manner, or is only silencing of such genes in senescence and differentiation, used to silence such genes in cells that have undergone a major challenge will be to define the contextual signals permanent cell cycle exit. Of note, methylated H3K9 has that dictate whether the ‘repression’ or the ‘silencing’ been detected at some promoters that were reversibly program is engaged. The evidence to date suggests that repressed (Ghosh and Harter, 2003; Nicolas et al., limited and potentially reversible H3K9 methylation 2003), but was detected at higher levels, in some cases in might occur at E2F-regulated promoters during quies- a heterochromatin-like trimethylated state, and asso- cence, whereas more extensive H3K9 methylation and ciated with HP1, only in irreversibly arrested cells recruitment of HP1 occurs during senescence and (Narita et al., 2003; Ait-Si-Ali et al., 2004). On this basis, differentiation. Thus, contextual cues in cells that are it appears that pocket protein-mediated methylation of entering senescence or undergoing differentiation may H3K9 may indeed induce silencing of E2F-responsive induce the transition between these chromatin states. promoters in some contexts. One aspect that might influence this transition is the One context in which pocket protein-mediated silen- identity of the histone methyltransferase that is recruited cing may occur is in cells that are entering senescence. to a promoter, since p107 was found to bind not only Accordingly, higher levels of methylated H3K9 were SUV39H1 but also Eu-HMTase1 (Nicolas et al., 2003). detected at the cyclin A and PCNA promoters when Whereas SUV39H1 is a H3K9 trimethylase that is human fibroblasts entered senescence versus when implicated in transcriptional silencing, Eu-HMTase1 is they were quiescent. In senescent cells, these promoters homologous to the ‘G9a’ H3K9 mono- and dimethylase also associated with HP1 and were resistant to E2F- (Peters et al., 2003; Rice et al., 2003) and is implicated in mediated activation. Finally, the pocket protein family reversible promoter repression (Ogawa et al., 2002). was required for H3K9 methylation, HP1 recruitment, and promoter silencing (Narita et al., 2003). These findings suggest that pocket protein-dependent H3K9 Pocket protein regulation of the G1–S transition, methylation, HP1 recruitment, and silencing of E2F- independent of E2F responsive promoters distinguishes senescence from a quiescent state. While pocket proteins have wide-ranging effects Similar results were obtained in an analysis of mediated through E2Fs, they also regulate cell cycle H3K9 methylation during myoblast differentiation. transitions through E2F-independent mechanisms. For Upon differentiation, several E2F-responsive genes example, p107 and p130 (but not pRB) bind and inhibit acquired trimethylated H3K9 in a process that depended the cyclin E/Cdk2 and cyclin A/Cdk2 kinases (Zhu upon SUV39H (including SUV39H1 and SUV39H2). et al., 1995). This activity is mechanistically similar to This is consistent with the possibility that pocket that of the ‘Cip/Kip’ family of Cdk inhibitors (CKIs), protein–SUV39H complexes mediate silencing of E2F- including p21Cip1,p27Kip1, and p57Kip2, and functions responsive genes during cell differentiation (Ait-Si-Ali redundantly to these CKIs in some contexts. In this et al., 2004). regard, p130 was required to inhibit Cdk2 and to One issue raised by the above findings is whether prevent S phase entry in mitogen-deprived p27À/À particular pocket proteins have special roles in mediat- fibroblasts (Coats et al., 1999). Similarly, p107 was ing H3K9 methylation and transcriptional silencing. In needed to inhibit Cdk2 and permit mitotic exit in the study by Narita et al. (2003), pRB was recruited to p21À/Àp27À/À cells that express a stabilized form of cyclin the cyclin A and PCNA promoters in senescent, but not A (Chibazakura et al., 2004). However, apart from quiescent, cells. However, p130 bound the promoters these synthetic effects, a physiological role for the CKI

Oncogene Pocket proteins and cell cycle control D Cobrinik 2803 function of p107/p130 has not been documented. ished production of ribosomal RNAs and tRNAs Perhaps the CKI function is important in contexts (Lajtha, 1963; Ladd et al., 1997; Ren and Rollins, where p107/p130 are needed for normal cell cycle 2004). Pocket proteins may contribute to the G0 state by control, such as in chondrocytes (Laplantine et al., repressing ribosomal RNA and tRNA gene expression, 2002; Rossi et al., 2002), epidermal cells (Ruiz et al., whereas pocket protein phosphorylation may restore 2003), neural stem cells (Vanderluit et al., 2004), or cells ribosomal RNA and tRNA levels to effect the G0–G1 that have inactivated pRB (see below). transition (Cavanaugh et al., 1995; White, 1997; Hannan More recently, pRB was found to inhibit Cdk activity et al., 2000; Scott et al., 2001). and G1–S progression by increasing the expression of In keeping with these considerations, the phosphor- p27 (Alexander and Hinds, 2001; Ji et al., 2004). In ylation of pRB by cyclin C-Cdk3 was recently found to osteosarcoma cells, pRB rapidly repressed E2F-respon- increase RNA content while mediating G0 exit (Ren and sive genes at the RNA level, but induced cell cycle exit Rollins, 2004). Notably, whereas pRB phosphorylation well before the protein products of such genes declined by cyclin C-Cdk3 mediated the G0–G1 transition, it did (Ji et al., 2004). In contrast, a pRB-induced increase in not supplant the need for the cyclin D-Cdk4/Cdk6 or p27 preceded, and was required for, cell cycle exit. cyclin E/Cdk2 for progression from G1 into S. Mechanistically, pRB stabilized p27 by binding to the F Conversely, these ‘G1 Cdks’ were able to mediate box protein Skp2, interfering with the Skp2–p27 pocket protein phosphorylation during the G0–G1 interaction, and inhibiting p27 ubiquitination (Ji et al., transition, albeit in a delayed fashion, and seem likely 2004). As p27 has widespread cell cycle effects, its to do so in mouse cells that have a mutant Cdk3 gene increased expression may be important in many contexts (Ye et al., 2001; Ren and Rollins, 2004). where pRB regulates G1–S progression. In addition to regulating the G1–S transition through p27, pRB has other E2F-independent functions through Pocket proteins and DNA replication y and rereplication which it may inhibit proliferation. For example, pRB can induce the formation of PML nuclear bodies (Fang Pocket proteins are known to regulate replication et al., 2002), which have widespread effects on cell indirectly, by repressing genes that mediate S phase proliferation, differentiation, and survival (Salomoni entry as well as genes that encode components of the and Pandolfi, 2002). pRB can also suppress Ras replication machinery. In addition, after S phase has signaling (Lee et al., 1999) and augment expression of begun, pRB and possibly p107 can inhibit S phase differentiation-associated genes (Sellers et al., 1998; progression by repressing cyclin A, and thereby inhibit- Thomas et al., 2001). Additionally, pRB may have ing Cdk2 and PCNA function (Karantza et al., 1993; E2F-independent effects through interactions with one Kondo et al., 2001; Sever-Chroneos et al., 2001). These of the many pRB-binding proteins (Morris and Dyson, transcriptional effects may be utilized to induce an intra- 2001). The recently-described hLin-9 protein may be S phase block in response to DNA damage (Harrington particularly relevant, in that it cooperated with pRB to et al., 1998; Knudsen et al., 2000; Kondo et al., 2001; induce a differentiation-related process, and fulfilled a Lan et al., 2002). role similar to pRB to suppress transformation, but did However, studies in Drosophila have shown that not affect the regulation of E2F (Gagrica et al., 2004). pocket proteins can also directly affect replication. One context in which the E2F-independent effects of These studies showed that the E2F1 and pRB homologs pRB may be important is in the suppression of (dE2F1 and Rbf) bind the Drosophila origin recognition retinoblastoma. This notion has come from evidence complex (DmORC) protein and the chorion gene cluster that several ‘low-penetrance’ pRB mutants retain a origin of replication, and thereby limit the physiological partial ability to suppress retinoblastoma despite that amplification of this cluster in ovarian follicle cells they are defective in binding to E2F (Otterson et al., (Royzman et al., 1999; Bosco et al., 2001). dE2F1 and 1997; Harbour, 2001). Various low penetrance pRB Rbf were also found to suppress genomic rereplication mutants retain abilities to induce p27, to induce PML in Drosophila, yet later work indicated that this was nuclear bodies, to augment differentiation-associated mediated through transcriptional effects (Cayirlioglu gene expression, to suppress Ras signaling, and to et al., 2003). Recent studies suggest that pocket proteins cooperate with hLin-9 in differentiation (Sellers et al., also have replicative functions in vertebrates. 1998; Lee et al., 1999; Fang et al., 2002; Ji et al., 2004; In a surprising development, pocket proteins were Gagrica et al., 2004), suggesting that one or more of found to affect the spatial organization of replication in these effects might contribute to retinoblastoma sup- primary mammalian cells. Initially, pocket proteins and pression. HDACs were found to colocalize with the small number of perinucleolar foci in which replication takes place at the start of S phase. In rapidly proliferating cells, these Pocket protein regulation of the G0–G1 transition few foci subsequently redistributed to form hundreds of replication foci that no longer localized with pocket In addition to regulating the G1–S transition, pocket proteins (Kennedy et al., 2000). However, in cells that proteins may regulate the earlier transition from G0 to received antimitogenic signals and were destined for cell G1. G0 cells can be distinguished from those in G1 by cycle exit, the replication foci did not redistribute, and their diminished RNA content, which reflects dimin- remained associated with pocket proteins throughout

Oncogene Pocket proteins and cell cycle control D Cobrinik 2804 S phase. Remarkably, the presence of pRB, p107, and/or from M phase, rather than S phase, abnormalities. p130 was crucial for the production of these focal Notably, prolongation of the mitotic checkpoint can replication structures (Barbie et al., 2004). While the cause tetraploidy through an ‘adaptation’ process, in function of these structures is not known, they seem to which chromosomes decondense and cells enter a provide pocket proteins the opportunity to modify state resembling G1 but with 4N DNA content (Lanni chromatin concurrent with replication. Thus, they might and Jacks, 1998). From this perspective, polyploidy in prepare certain promoters for repression or modify pRB-deficient cells may result from increased expression replication origins in cells that are preparing for cell of mitotic checkpoint proteins, a prolonged mitotic cycle arrest. block, and then this adaptation response. pRB is also connected to replication through its role in preventing genomic rereplication. pRB inhibits the spontaneous production of polyploid cells, and sup- New ways to control pocket protein phosphorylation presses the increased polyploidy that occurs in cells that experience S phase DNA damage, in cells that are Cyclin D-Cdk4/Cdk6 and cyclin E-Cdk2 have long been blocked at G2/M by ectopically expressed CKIs, and implicated in pocket protein phosphorylation (Wein- in cells that are blocked in M phase by microtubule- berg, 1995). These Cdks have also been suggested to inhibiting drugs (Di Leonardo et al., 1997; Harrington mediate phosphorylation of distinct pRB residues et al., 1998; Niculescu et al., 1998). Several pRB (Mittnacht, 1998) and to inactivate pRB in a step-wise functions could plausibly account for this capability. fashion during G1 (Hatakeyama et al., 1994; Lundberg For example, in cells that were irradiated in early and Weinberg, 1998; Harbour et al., 1999). However, S phase, pRB associated with an early firing origin of recent studies challenge these notions, and instead replication during the S phase block, and then bound to highlight the functional redundancy of the cyclin-Cdk additional replication origins in the order in which they network in pocket protein phosphorylation. fired. The presence of pRB at origins, at about the time As noted earlier, pRB phosphorylation during the that they replicated, suggested that pRB might modify G0–G1 transition appears to be mediated by cyclin C- the origins in a manner that prevents rereplication (Avni Cdk3, but can be mediated by other Cdks in Cdk3- et al., 2003). Notably, pRB’s association with late firing deficient cells (Ren and Rollins, 2004). An additional replication origins after irradiation is reminiscent of its level of redundancy was revealed in finding that D association with late S phase replication foci in cells that cyclins, E cyclins, and the Cdk4/Cdk6 or Cdk2 kinases receive antimitogenic signals (Barbie et al., 2004), were not required for pRB or p107 phosphorylation or suggesting that the two processes may be related. for G1–S progression in mouse fibroblasts (Berthet A second means by which pRB might suppress et al., 2003; Geng et al., 2003; Ortega et al., 2003; Kozar rereplication could be through effects on the replication et al., 2004; Malumbres et al., 2004). Concordantly, in licensing machinery, composed of MCM proteins, cell types that normally require cyclin D1, the mis- CDC6, and others. This apparatus binds to replication expression of cyclin E1 was able to substitute for cyclin origins in late M and G1, and prepares – or licenses – D1 function (Geng et al., 1999). These observations them to function after entry into S (Blow and Hodgson, have a number of far reaching implications that are 2002). Licensing activity is normally suppressed in discussed elsewhere (Pagano and Jackson, 2004; Sherr S, G2, and early M phase by cyclin A-Cdk activity and Roberts, 2004). However, with regard to pocket (Yam et al., 2002). pRB might further suppress proteins, the findings mainly imply that either Cdk4/ relicensing by interacting with MCM7 (Sterner et al., Cdk6 or Cdk2 can mediate the phosphorylation events 1998), or by repressing genes that encode components that are needed for S phase entry. of the licensing apparatus. In this regard, it is notable In another unexpected development, pocket proteins that pRB-deficient cells have increased expression of have been found to be regulated by kinases other than MCM2, 4, 5, 6, and 7 (Black et al., 2003). pRB-deficient the canonical cyclin-Cdks. Thus, whereas mitogen- cells also have increased cyclin E, which could further induced pocket protein phosphorylation and release promote relicensing and rereplication (Spruck et al 1999; from E2F generally occurs over a period of hours and Coverley et al., 2002). depends on the induction of cyclin D or cyclin E genes, Finally, pRB may prevent polyploidy by enforcing the in several instances this response is far more rapid and normal expression of mitotic checkpoint proteins such occurs through other mechanisms. For example, angio- as Emi1 and Mad2. The Emi1 and Mad2 genes are tensin II and serotonin treatment of vascular smooth regulated by E2F, and the proteins are overexpressed in muscle cells resulted in an ERK-dependent and CDK4- pRB-deficient cells or tumors (Hsu et al., 2002; mediated phosphorylation of pRBSer795, and pRB Hernando et al., 2004). In turn, overexpression of dissociation from E2F, within 10 min (Garnovskaya Emi1 or Mad2 delays passage through mitosis, and et al., 2004). Angiotensin II and serotonin are strong induces the production of binucleate, aneuploid, and mitogens for these cells, suggesting that the phosphor- polyploid cells (Margottin-Goguet et al., 2003; Hernan- ylation initiates cell cycle entry. Also, stimulation of the do et al., 2004). Importantly, the overexpression of Fas ‘death receptor’ elicited pRB phosphorylation and Mad2 was also required for deregulated E2F to induce release from E2F within 30 min. Here, pRB appeared to aneuploidy and polyploidy (Hernando et al., 2004), be phosphorylated by p38MAPK on residues distinct implying that E2F-induced chromosomal defects result from those targeted by Cdks (Wang et al., 1999; Nath

Oncogene Pocket proteins and cell cycle control D Cobrinik 2805 et al., 2003). This p38-dependent pRB phosphorylation personal communication). This might result in a resulted in the release of E2F1 and is implicated in Fas- paradoxical increase in tumor suppressor activity induced apoptosis (Hou et al., 2002). following pRB loss, and could account for the general Conversely, pocket protein dephosphorylation was absence of RB mutations, in tumors that have oncogenic known to occur in the period from anaphase to G1 and Ras signaling. to depend upon protein phosphatase 1 (Ludlow et al., Currently, the basis for the compensatory antiproli- 1993), but is now also known to occur in other cell cycle ferative effect of p107 is unknown. As one possibility, phases in response to growth inhibitory signals. For p107 may partially substitute (though it clearly does not example, protein phosphatase 2A (PP2A) is implicated fully substitute) for pRB in the regulation of E2F1–3, in producing limited pRB hypophosphorylation in and it may thereby repress some E2F-responsive genes S phase cells and more extensive pRB hypophosphor- that are normally controlled by pRB (Lee et al., 2002). ylation in response to g irradiation (Avni et al., 2003). Alternatively, the p107 CKI activity may contribute to Similarly, oxidative stress elicited a rapid PP2A-depen- G0/G1 arrest, perhaps compensating for diminished p27 dent dephosphorylation of pRB, p107, and p130 expression in pRB-deficient cells (Ji et al., 2004). (Cicchillitti et al., 2003), and the differentiation agent all-trans-retinoic acid elicited rapid PP2A-dependent dephosphorylation of p130 (Vuocolo et al., 2003). Rapid Focal deficits in pocket protein redundancy in dephosphorylation may also occur in a developmental development and tumorigenesis context. For example, in chondrocytes, FGF induced rapid dephosphorylation of p107, but not p130 or pRB, Whereas the redundant (or compensatory) actions of preceding a p107-dependent G0/G1 arrest (Laplantine pocket proteins can mediate a G0/G1 arrest in pRB- et al., 2002; Dailey et al., 2003). deficient and p107/p130-deficient mouse fibroblasts in Whereas pocket protein phosphorylation is often vitro, they apparently fail to do so in a number of cell reversed by dephosphorylation, in some situations, types in vivo. This, in turn, may lead to focal phosphorylation may cause permanent inactivation. proliferative defects that impede normal development For example, p130 phosphorylation by Cdk4/Cdk6 and homeostasis following pRB or p107/p130 loss. For permits binding to Skp2, polyubiquitination, and example, pRB-deficient mouse embryos display aberrant proteasome-mediated degradation in late G1 (Tedesco proliferation, often with apoptosis and differentiation et al., 2002). Phosphorylation of pRB may also lead to defects, in fetal liver, lens, brain, muscle, skin, and degradation, but through a different mechanism. Speci- placenta (reviewed in Liu et al., 2004). A subset of the fically, high levels of Cdk activity may result in apoptotic responses resulted from defective placental phosphorylation of pRB-Ser567, and thus destabilization function and the ensuing hypoxia (Wu et al., 2003), but of a pRB intramolecular interaction, exposure of a the proliferation and differentiation defects were gen- proteolytic cleavage site, and degradation by an un- erally cell autonomous. Additional proliferation and indentified protease. This phosphorylation-induced differentiation defects have been observed postnatally in degradation of pRB may increase the availability of tissue-specific Rb knockouts in the pituitary, lung, E2F1 and promote apoptosis in cells that have elevated cerebellum, retina, prostate, and skin (Vooijs et al., Cdk activity (Ma et al., 2003). 1998; Marino et al., 2003; Chen et al., 2004; MacPher- son et al., 2004; Maddison et al., 2004; Ruiz et al., 2004; Wikenheiser-Brokamp, 2004; Zhang et al., 2004a). p107 Functional redundancy and compensation amongpRB and and p130 are needed to control proliferation and p107/p130 differentiation in a more limited spectrum of tissues, including cartilage, epidermis, and brain (Rossi et al., Whereas pRB and the p107/p130 pair do not regulate 2002; Ruiz et al., 2003; Vanderluit et al., 2004). E2F-responsive genes in a fully redundant manner (see Currently, the reason why pocket proteins function above), these proteins show considerable redundancy at redundantly in some settings, but fail to do so in others, the cell cycle level. Indeed, fibroblasts lacking either is unknown. Certain cell types might rely upon pRB or p107/p130 maintain the ability to arrest in G0/ particular pocket proteins because the other family G1, whereas fibroblasts lacking all three pocket proteins members are expressed at low levels or fail to increase, in (as well as those lacking only pRB and p107) fail to do a compensatory manner, after an initial pocket protein so under a variety of circumstances (Dannenberg et al., is lost. Alternatively, the affected cell types might 2000; Sage et al., 2000). require differentiation functions that are uniquely Interestingly, in pRB-deficient fibroblasts, the ability fulfilled by one or another pocket protein, such as to arrest in G0/G1 requires a compensatory increase in pRB-specific functions in osteogenic and myogenic p107. The compensatory increase in p107 was delayed differentiation (Thomas et al., 2001; Huh et al., 2004; after an acute loss of pRB, and this provided an Nguyen et al., 2004) and p107/p130-specific functions in opportunity for further proliferation and immortaliza- chondrocyte differentiation (Laplantine et al., 2002; tion (Sage et al., 2003). The compensatory increase in Rossi et al., 2002). Defining the roles of specific pocket p107 also occurs in pRB-deficient human tumor cells proteins in different developmental contexts may and may have a particularly potent antiproliferative provide fundamental insight into the underlying differ- effect in cells that have activated Ras (M Classon, entiation pathways.

Oncogene Pocket proteins and cell cycle control D Cobrinik 2806 Interestingly, most cell types that exhibit deregulated Among the six general criteria that are needed for a proliferation in constitutive or tissue-specific Rb and normal cell to produce malignant progeny (Hanahan p107/p130 knockouts do not go on to form tumors. and Weinberg, 2000), it is important to recognize that Indeed, in Rb þ /À mice, somatic mutagenesis is expected the loss of pocket protein function may generally impart to produce RbÀ/À cells in most if not all cell types. only the capacity to ignore antimitogenic signals. However, tumors form only from pituitary and thyroid Nevertheless, in some contexts, loss of specific pocket cells (Williams et al., 1994). Similarly, RB þ /À humans proteins may have more pleiotropic effects, potentially are expected to generate diverse types of RBÀ/À cells, yet relating to their multiple biochemical functions. A tumors arise from retinal cells in >90% of such potential example of this is in the developing human individuals but arise from all other cell types in only retina, where loss of pRB is sufficient to initiate the B1% of RB þ /À individuals per year (Abramson, formation of retinoblastoma tumors. This suggests that 1999). Thus, in addition to the aberrant prolifera- in the retinoblastoma cell of origin, other pocket tion that ensues from the loss of pRB in many cell proteins do not compensate for the loss of pRB, and types, other criteria must be met for tumorigenesis p53 and other proteins that normally guard against the to proceed. effects of pocket protein loss are not effectively engaged. In some settings, the reason why pRB loss results in Moreover, the rapid onset of retinoblastoma suggests proliferation but not tumorigenesis may be that p107 that the loss of pRB may concurrently fulfill several of fails to compensate for the immediate proliferative the requirements for tumorigenesis, or that the retino- effects of pRB loss, but nonetheless provides a ‘backup’ blastoma cell of origin is inclined to fulfill the other tumor suppressor effect. This scenario may apply to the requirements within a short time frame, and perhaps mouse retina, in which the loss of pRB elicited with relatively few additional genetic changes. However, deregulated proliferation whereas combined loss of we currently have little understanding of the circuitry pRB and p107 elicited retinal tumors (Robanus- that underlies this unusual susceptibility. Defining how Maandag et al., 1998; Chen et al., 2004; MacPherson the loss of pocket protein function coalesces with other et al., 2004; Zhang et al., 2004a, b). For other cell types gene expression programs to bring about retinoblasto- that aberrantly proliferate after pRB loss, tumorigenesis ma, as well as other tumors, will constitute a pivotal new may be prevented by p53 (Williams et al., 1994). For chapter in the biology of cancer. example, loss of pRB in the lung resulted in aberrant proliferation and mild hyperplasia, whereas combined loss of pRB and p53 resulted in small cell lung Acknowledgements I thank Nick Dyson for helpful discussions and Marie Classon, carcinomas (Meuwissen et al., 2003; Wikenheiser- Peter K Jackson, and Liang Zhu for discussing work prior to Brokamp, 2004). Presumably, additional genetic and publication and critical reading of the manuscript. My epigenetic changes will be found to complement pocket apologies are given to those whose work could not be cited protein defects in a way that converts aberrantly due to space considerations. I was supported by the CV Starr proliferating cells into tumor cells. Foundation during the preparation of this review.

References

Abramson DH. (1999). Ophthalmic Genet., 20, 193–204. Blais A and Dynlacht BD. (2004). Curr. Opin. Genet. Dev., 14, Ait-Si-Ali S, Guasconi V, Fritsch L, Yahi H, 527–532. Sekhri R, Naguibneva I, Robin P, Cabon F, Polesskaya Blow JJ and Hodgson B. (2002). Trends Cell Biol., 12, 72–78. A and Harel-Bellan A. (2004). EMBO J., 23, Bosco G, Du W and Orr-Weaver TL. (2001). Nat. Cell Biol., 3, 605–615. 289–295. Alevizopoulos K, Vlach J, Hennecke S and Amati B. (1997). Botz J, Zerfass-Thome K, Spitkovsky D, Delius H, Vogt B, EMBO J., 16, 5322–5333. Eilers M, Hatzigeorgiou A and Jansen-Durr P. (1996). Mol. Alexander K and Hinds PW. (2001). Mol. Cell. Biol., 21, Cell. Biol., 16, 3401–3409. 3616–3631. Bracken AP, Ciro M, Cocito A and Helin K. (2004). Trends Araki K, Nakajima Y, Eto K and Ikeda MA. (2003). Biochem. Sci., 29, 409–417. Oncogene, 22, 7632–7641. Buchkovich K, Duffy LA and Harlow E. (1989). Cell, 58, Aslanian A, Iaquinta PJ, Verona R and Lees JA. (2004). Genes 1097–1105. Dev., 18, 1413–1422. Cam H and Dynlacht BD. (2003). Cancer Cell, 3, 311–316. Avni D, Yang H, Martelli F, Hofmann F, ElShamy WM, Cavanaugh AH, Hempel WM, Taylor LJ, Rogalsky V, Ganesan S, Scully R and Livingston DM. (2003). Mol. Cell, Todorov G and Rothblum LI. (1995). Nature, 374, 177–180. 12, 735–746. Cavenee WK, Dryja TP, Phillips RA, Benedict WF, Godbout Barbie DA, Kudlow BA, Frock R, Zhao J, Johnson BR, R, Gallie BL, Murphree AL, Strong LC and White RL. Dyson N, Harlow E and Kennedy BK. (2004). Mol. Cell. (1983). Nature, 305, 779–784. Biol., 24, 595–607. Cayirlioglu P, Ward WO, Silver Key SC and Duronio RJ. Berthet C, Aleem E, Coppola V, Tessarollo L and Kaldis P. (2003). Mol. Cell. Biol., 23, 2123–2134. (2003). Curr. Biol., 13, 1775–1785. Chen CR, Kang Y, Siegel PM and Massague J. (2002). Cell, Black EP, Huang E, Dressman H, Rempel R, Laakso N, Asa 110, 19–32. SL, Ishida S, West M and Nevins JR. (2003). Cancer Res., Chen D, Livne-bar I, Vanderluit JL, Slack RS, Agochiya M 63, 3716–3723. and Bremner R. (2004). Cancer Cell, 5, 539–551.

Oncogene Pocket proteins and cell cycle control D Cobrinik 2807 Chibazakura T, McGrew SG, Cooper JA, Yoshikawa H Ghosh MK and Harter ML. (2003). Mol. Cell, 12, 255–260. and Roberts JM. (2004). Proc. Natl. Acad. Sci. USA, 101, Giangrande PH, Zhu W, Rempel RE, Laakso N and Nevins 4465–4470. JR. (2004). EMBO J., 23, 1336–1347. Cicchillitti L, Fasanaro P, Biglioli P, Capogrossi MC and Ginsberg D. (2004). Dev. Cell, 6, 742–743. Martelli F. (2003). J. Biol. Chem., 278, 19509–19517. Hanahan D and Weinberg RA. (2000). Cell, 100, 57–70. Classon M and Dyson N. (2001). Exp. Cell Res., 264, 135–147. Hannan KM, Kennedy BK, Cavanaugh AH, Hannan RD, Coats S, Whyte P, Fero ML, Lacy S, Chung G, Randel E, Hirschler-Laszkiewicz I, Jefferson LS and Rothblum LI. Firpo E and Roberts JM. (1999). Curr. Biol., 9, 163–173. (2000). Oncogene, 19, 3487–3497. Cobrinik D. (1996). Curr. Top. Microbiol. Immunol., 208, Harbour JW. (2001). Arch. Ophthalmol., 119, 1699–1704. 31–60. Harbour JW, Lai S-L, Whang-Peng J, Gazder AF, Minna JD Cobrinik D, Dowdy SF, Hinds PW, Mittnacht S and and Kaye FJ. (1988). Science, 241, 353–357. Weinberg RA. (1992). Trends Biochem. Sci., 17, 312–315. Harbour JW, Luo RX, Dei Santi A, Postigo AA and Dean Coisy M, Roure V, Ribot M, Philips A, Muchardt C, DC. (1999). Cell, 98, 859–869. Blanchard JM and Dantonel JC. (2004). Mol. Cell, 15, 43–56. Harrington EA, Bruce JL, Harlow E and Dyson N. (1998). Coverley D, Laman H and Laskey RA. (2002). Nat. Cell Biol., Proc. Natl. Acad. Sci. USA, 95, 11945–11950. 4, 523–528. Hatakeyama M, Brill JA, Fink GR and Weinberg RA. (1994). Dahiya A, Wong S, Gonzalo S, Gavin M and Dean DC. Genes Dev., 8, 1759–1771. (2001). Mol. Cell, 8, 557–569. He S, Cook BL, Deverman BE, Weihe U, Zhang F, Prachand Dahiya A, Wong S, Gonzalo S, Gavin M and Dean DC. V, Zheng J and Weintraub SJ. (2000). Mol. Cell. Biol., 20, (2003). Mol. Cell, 11, 1693–1696. 363–371. Dailey L, Laplantine E, Priore R and Basilico C. (2003). J. Cell Helin K, Harlow E and Fattaey AR. (1993). Mol. Cell. Biol., Biol., 161, 1053–1066. 13, 6501–6508. Dannenberg JH, van Rossum A, Schuijff L and te Riele H. Hernando E, Nahle Z, Juan G, Diaz-Rodriguez E, (2000). Genes Dev., 14, 3051–3064. Alaminos M, Hemann M, Michel L, Mittal V, Gerald W, DeCaprio JA, Ludlow JW, Lynch D, Furukawa Y, Grifﬁn J, Benezra R, Lowe SW and Cordon-Cardo C. (2004). Nature, Piwnica-Worms H, Huang C-M and Livingston DM. (1989). 430, 797–802. Cell, 58, 1085–1095. Herrera RE, Tan VP, Williams BO, Makela T, Weinberg RA Dimova DK and Dyson N. (2005). Oncogene, 24, 2810–2826. and Jacks T. (1996). Mol. Cell. Biol., 16, 2402–2407. Dimova DK, Stevaux O, Frolov MV and Dyson NJ. (2003). Horowitz JM, Yandell DW, Park S-H, Canning S, Whyte P, Genes Dev., 17, 2308–2320. Buchkovich K, Harlow E, Weinberg RA and Dryja TP. Di Leonardo A, Khan SH, Linke SP, Greco V, Seidita G and (1989). Science, 243, 937–940. Wahl GM. (1997). Cancer Res., 57, 1013–1019. Hou ST, Xie X, Baggley A, Park DS, Chen G and Walker T. Di Stefano L, Jensen MR and Helin K. (2003). EMBO J., 22, (2002). J. Biol. Chem., 277, 48764–48770. 6289–6298. Hsu JY, Reimann JD, Sorensen CS, Lukas J and Jackson PK. Dunaief JL, Strober BE, Guha S, Khavari PA, Alin K, Luban (2002). Nat. Cell Biol., 4, 358–366. J, Begemann M, Crabtree GR and Goff SP. (1994). Cell, 79, Huh MS, Parker MH, Scime A, Parks R and Rudnicki MA. 119–130. (2004). J. Cell Biol., 166, 865–876. Dunn JM, Phillips RA, Becker AJ and Gallie BL. (1988). Hurford RK, Cobrinik D, Lee MH and Dyson N. (1997). Science, 241, 1797–1800. Genes Dev., 11, 1447–1463. Dyson N. (1998). Genes Dev., 12, 2245–2262. Ishida S, Huang E, Zuzan H, Spang R, Leone G, West M and Fabbrizio E, El Messaoudi S, Polanowska J, Paul C, Cook JR, Nevins JR. (2001). Mol. Cell. Biol., 21, 4684–4699. Lee JH, Negre V, Rousset M, Pestka S, Le Cam A and Ji P, Jiang H, Rekhtman K, Bloom J, Ichetovkin M, Pagano Sardet C. (2002). EMBO Rep., 3, 641–645. M and Zhu L. (2004). Mol. Cell, 16, 47–58. Falls H and Neel J. (1951). Arch. Ophthalmol., 46, 367–389. Karantza V, Maroo A, Fay D and Sedivy JM. (1993). Mol. Fang W, Mori T and Cobrinik D. (2002). Oncogene, 21, 5557– Cell. Biol., 13, 6640–6652. 5565. Karlseder J, Rotheneder H and Wintersberger E. (1996). Mol. Ferreira R, Naguibneva I, Mathieu M, Ait-Si-Ali S, Robin P, Cell. Biol., 16, 1659–1667. Pritchard LL and Harel-Bellan A. (2001). EMBO Rep., 2, Kennedy BK, Barbie DA, Classon M, Dyson N and Harlow 794–799. E. (2000). Genes Dev., 14, 2855–2868. Flemington EK, Speck SH and Kaelin Jr WG. (1993). Proc. Knudson AG. (1971). Proc. Natl. Acad. Sci. USA, 68, Natl. Acad. Sci. USA, 90, 6914–6918. 820–823. Friend SH, Bernards R, Rogelj S, Weinberg RA, Rapaport Knudsen KE, Booth D, Naderi S, Sever-Chroneos Z, Fribourg JM, Alberts D and Dryja TP. (1986). Nature, 323, 643–646. AF, Hunton IC, Feramisco JR, Wang JY and Knudsen ES. Frolov MV and Dyson NJ. (2004). J. Cell Sci., 117, (2000). Mol. Cell. Biol., 20, 7751–7763. 2173–2181. Kondo T, Higashi H, Nishizawa H, Ishikawa S, Ashizawa S, Gagrica S, Hauser S, Kolfschoten I, Osterloh L, Agami R and Yamada M, Makita Z, Koike T and Hatakeyama M. (2001). Gaubatz S. (2004). EMBO J., 23, 4627–4638. J. Biol. Chem., 276, 17559–17567. Garnovskaya MN, Mukhin YV, Vlasova TM, Grewal JS, Kozar K, Ciemerych MA, Rebel VI, Shigematsu H, Zagozd- Ullian ME, Tholanikunnel BG and Raymond JR. (2004). zon A, Sicinska E, Geng Y, Yu Q, Bhattacharya S, Bronson J. Biol. Chem., 279, 24899–24905. RT, Akashi K and Sicinski P. (2004). Cell, 118, 477–491. Geng Y, Whoriskey W, Park MY, Bronson RT, Medema RH, Ladd AC, Pyatt R, Gothot A, Rice S, McMahel J, Traycoff Li T, Weinberg RA and Sicinski P. (1999). Cell, 97, CM and Srour EF. (1997). Blood, 90, 658–668. 767–777. Lai A, Kennedy BK, Barbie DA, Bertos NR, Yang XJ, Geng Y, Yu Q, Sicinska E, Das M, Schneider JE, Bhattachar- Theberge MC, Tsai SC, Seto E, Zhang Y, Kuzmichev A, ya S, Rideout WM, Bronson RT, Gardner H and Sicinski P. Lane WS, Reinberg D, Harlow E and Branton PE. (2001). (2003). Cell, 114, 431–443. Mol. Cell. Biol., 21, 2918–2932.

Oncogene Pocket proteins and cell cycle control D Cobrinik 2808 Lajtha LG. (1963). J. Cell Physiol., 62 (Suppl. 1), 143–145. Nguyen DX, Baglia LA, Huang SM, Baker CM and McCance Lan Z, Sever-Chroneos Z, Strobeck MW, Park CH, Baskaran DJ. (2004). EMBO J., 23, 1609–1618. R, Edelmann W, Leone G and Knudsen ES. (2002). J. Biol. Nicolas E, Roumillac C and Trouche D. (2003). Mol. Cell. Chem., 277, 8372–8381. Biol., 23, 1614–1622. Lanni JS and Jacks T. (1998). Mol. Cell. Biol., 18, 1055–1064. Niculescu III AB, Chen X, Smeets M, Hengst L, Prives C and Laplantine E, Rossi F, Sahni M, Basilico C and Cobrinik D. Reed SI. (1998). Mol. Cell. Biol., 18, 629–643. (2002). J. Cell Biol., 158, 741–750. Nielsen SJ, Schneider R, Bauer UM, Bannister AJ, Morrison Le Cam L, Polanowska J, Fabbrizio E, Olivier M, Philips A, A, O’Carroll D, Firestein R, Cleary M, Jenuwein T, Herrera Ng Eaton E, Classon M, Geng Y and Sardet C. (1999). RE and Kouzarides T. (2001). Nature, 412, 561–565. EMBO J., 18, 1878–1890. Ogawa H, Ishiguro K, Gaubatz S, Livingston DM and Lee EY, Cam H, Ziebold U, Rayman JB, Lees JA and Nakatani Y. (2002). Science, 296, 1132–1136. Dynlacht BD. (2002). Cancer Cell, 2, 463–472. Ortega S, Prieto I, Odajima J, Martin A, Dubus P, Sotillo R, Lee EY-HP, To H, Shew J-Y, Bookstein R, Scully P and Lee Barbero JL, Malumbres M and Barbacid M. (2003). Nat. W-H. (1988). Science, 241, 218–221. Genet., 35, 25–31. Lee KY, Ladha MH, McMahon C and Ewen ME. (1999). Otterson GA, Chen W, Coxon AB, Khleif SN and Kaye FJ. Mol. Cell. Biol., 19, 7724–7732. (1997). Proc. Natl. Acad. Sci. USA, 94, 12036–12040. Lee W-H, Shew J-Y, Hong FD, Sery TW, Donoso LA, Pagano M and Jackson PK. (2004). Cell, 118, 535–538. Young L-J, Bookstein R and Lee EY-H. (1987). Nature, 329, Pediconi N, Ianari A, Costanzo A, Belloni L, Gallo R, Cimino 642–645. L, Porcellini A, Screpanti I, Balsano C, Alesse E, Gulino A Leng X, Connell-Crowley L, Goodrich D and Harper JW. and Levrero M. (2003). Nat. Cell Biol., 5, 552–558. (1997). Curr. Biol., 7, 709–712. Peters AH, Kubicek S, Mechtler K, O’Sullivan RJ, Derijck Leone G, Nuckolls F, Ishida S, Adams M, Sears R, AA, Perez-Burgos L, Kohlmaier A, Opravil S, Tachibana Jakoi L, Miron A and Nevins JR. (2000). Mol. Cell. Biol., M, Shinkai Y, Martens JH and Jenuwein T. (2003). Mol. 20, 3626–3632. Cell, 12, 1577–1589. Lin SY, Black AR, Kostic D, Pajovic S, Hoover CN and Philips A, Huet X, Plet A, Le Cam L, Vie A and Blanchard Azizkhan JC. (1996). Mol. Cell. Biol., 16, 1668–1675. JM. (1998). Oncogene, 16, 1373–1381. Liu H, Dibling B, Spike B, Dirlam A and Macleod K. (2004). Polager S, Kalma Y, Berkovich E and Ginsberg D. (2002). Curr. Opin. Genet. Dev., 14, 55–64. Oncogene, 21, 437–446. Ludlow JW, Glendening CL, Livingston DM and DeCarprio Polanowska J, Fabbrizio E, Le Cam L, Trouche D, Emiliani S, JA. (1993). Mol. Cell. Biol., 13, 367–372. Herrera R and Sardet C. (2001). Oncogene, 20, 4115–4127. Lukas J, Herzinger T, Hansen K, Moroni MC, Resnitzky D, Rayman JB, Takahashi Y, Indjeian VB, Dannenberg JH, Helin K, Reed SI and Bartek J. (1997). Genes Dev., 11, Catchpole S, Watson RJ, te Riele H and Dynlacht BD. 1479–1492. (2002). Genes Dev., 16, 933–947. Lundberg AS and Weinberg RA. (1998). Mol. Cell. Biol., 18, Ren B, Cam H, Takahashi Y, Volkert T, Terragni J, Young 753–761. RA and Dynlacht BD. (2002). Genes Dev., 16, 245–256. Ma D, Zhou P and Harbour JW. (2003). J. Biol. Chem., 278, Ren S and Rollins BJ. (2004). Cell, 117, 239–251. 19358–19366. Rice JC, Briggs SD, Ueberheide B, Barber CM, Shabanowitz MacPherson D, Sage J, Kim T, Ho D, McLaughlin ME and J, Hunt DF, Shinkai Y and Allis CD. (2003). Mol. Cell, 12, Jacks T. (2004). Genes Dev., 18, 1681–1694. 1591–1598. Maddison LA, Sutherland BW, Barrios RJ and Greenberg Robanus-Maandag E, Dekker M, van der Valk M, Carrozza NM. (2004). Cancer Res., 64, 6018–6025. ML, Jeanny JC, Dannenberg JH, Berns A and te Riele H. Malumbres M, Sotillo R, Santamaria D, Galan J, Cerezo A, (1998). Genes Dev., 12, 1599–1609. Ortega S, Dubus P and Barbacid M. (2004). Cell, 118, 493–504. Rossi F, MacLean HE, Yuan W, Francis RO, Semenova E, Margottin-Goguet F, Hsu JY, Loktev A, Hsieh HM, Reimann Lin CS, Kronenberg HM and Cobrinik D. (2002). Dev. Biol., JD and Jackson PK. (2003). Dev. Cell, 4, 813–826. 247, 271–285. Marino S, Hoogervoorst D, Brandner S and Berns A. (2003). Royzman I, Austin RJ, Bosco G, Bell SP and Orr-Weaver TL. Development, 130, 3359–3368. (1999). Genes Dev., 13, 827–840. Meuwissen R, Linn SC, Linnoila RI, Zevenhoven J, Mooi WJ Ruiz S, Santos M, Segrelles C, Leis H, Jorcano JL, and Berns A. (2003). Cancer Cell, 4, 181–189. Berns A, Paramio JM and Vooijs M. (2004). Development, Mittnacht S. (1998). Curr. Opin. Genet. Dev., 8, 21–27. 131, 2737–2748. Morris EJ and Dyson NJ. (2001). Adv. Cancer Res., 82, 1–54. Ruiz S, Segrelles C, Bravo A, Santos M, Perez P, Leis H, Morrison AJ, Sardet C and Herrera RE. (2002). Mol. Cell. Jorcano JL and Paramio JM. (2003). Development, 130, Biol., 22, 856–865. 2341–2353. Muller H, Bracken AP, Vernell R, Moroni MC, Christians F, Saavedra HI, Maiti B, Timmers C, Altura R, Tokuyama Y, Grassilli E, Prosperini E, Vigo E, Oliner JD and Helin K. Fukasawa K and Leone G. (2003). Cancer Cell, 3, 333–346. (2001). Genes Dev., 15, 267–285. Sage J, Miller AL, Perez-Mancera PA, Wysocki JM and Jacks Mulligan GJ, Wong J and Jacks T. (1998). Mol. Cell. Biol., 18, T. (2003). Nature, 424, 223–228. 206–220. Sage J, Mulligan GJ, Attardi LD, Miller A, Chen S, Narita M, Nunez S, Heard E, Lin AW, Hearn SA, Spector Williams B, Theodorou E and Jacks T. (2000). Genes Dev., DL, Hannon GJ and Lowe SW. (2003). Cell, 113, 703–716. 14, 3037–3050. Narlikar GJ, Fan HY and Kingston RE. (2002). Cell, 108, Salomoni P and Pandolﬁ PP. (2002). Cell, 108, 165–170. 475–487. Schlisio S, Halperin T, Vidal M and Nevins JR. (2002). EMBO Nath N, Wang S, Betts V, Knudsen E and Chellappan S. J., 21, 5775–5786. (2003). Oncogene, 22, 5986–5994. Scott PH, Cairns CA, Sutcliffe JE, Alzuherri HM, McLees A, Nevins JR. (1992). Science, 258, 424–429. Winter AG and White RJ. (2001). J. Biol. Chem., 276, Nevins JR. (1998). Cell Growth Differ., 9, 585–593. 1005–1014.

Oncogene Pocket proteins and cell cycle control D Cobrinik 2809 Sellers WR, Novitch BG, Miyake S, Heith A, Otterson GA, Vooijs M, van der Valk M, te Riele H and Berns A. (1998). Kaye FJ, Lassar AB and Kaelin WG. (1998). Genes Dev., 12, Oncogene, 17, 1–12. 95–106. Vuocolo S, Purev E, Zhang D, Bartek J, Hansen K, Soprano Sever-Chroneos Z, Angus SP, Fribourg AF, Wan H, Todorov DR and Soprano KJ. (2003). J. Biol. Chem., 278, 41881– I, Knudsen KE and Knudsen ES. (2001). Mol. Cell. Biol., 21, 41889. 4032–4045. Wang S, Nath N, Minden A and Chellappan S. (1999). EMBO Sherr CJ. (1996). Science, 274, 1672–1677. J., 18, 1559–1570. Sherr CJ and Roberts JM. (2004). Genes Dev., 18, 2699–2711. Weinberg RA. (1995). Cell, 81, 323–330. Siddiqui H, Solomon DA, Gunawardena RW, Wang Y Weinmann AS, Yan PS, Oberley MJ, Huang TH and and Knudsen ES. (2003). Mol. Cell. Biol., 23, 7719–7731. Farnham PJ. (2002). Genes Dev., 16, 235–244. Spruck CH, Wong K-A and Reed SI. (1999). Nature, 401, Wells J, Boyd KE, Fry CJ, Bartley SM and Farnham PJ. 297–300. (2000). Mol. Cell. Biol., 20, 5797–5807. Sterner JM, Dew-Knight S, Musahl C, Kornbluth S and White RJ. (1997). Trends Biochem. Sci., 22, 77–80. Horowitz JM. (1998). Mol. Cell. Biol., 18, 2748–2757. Wikenheiser-BrokampKA. (2004). Development, 131, Stevaux O and Dyson NJ. (2002). Curr. Opin. Cell Biol., 14, 4299–4310. 684–691. Williams BO, Remington L, Albert DM, Mukai S, Bronson Strobeck MW, Knudsen KE, Fribourg AF, DeCristofaro MF, RT and Jacks T. (1994). Nat. Genet., 7, 480–484. Weissman BE, Imbalzano AN and Knudsen ES. (2000). Wu L, de Bruin A, Saavedra HI, Starovic M, Trimboli A, Proc. Natl. Acad. Sci. USA, 97, 7748–7753. Yang Y, Opavska J, Wilson P, Thompson JC, Ostrowski Strober BE, Dunaief JL and Guha Goff SP. (1996). Mol. Cell. MC, Rosol TJ, Woollett LA, Weinstein M, Cross JC, Biol., 16, 1576–1583. Robinson ML and Leone G. (2003). Nature, 421, 942–947. Takahashi Y, Rayman JB and Dynlacht BD. (2000). Genes Wu L, Timmers C, Maiti B, Saavedra HI, Sang L, Chong GT, Dev., 14, 804–816. Nuckolls F, Giangrande P, Wright FA, Field SJ, Greenberg Taubert S, Gorrini C, Frank SR, Parisi T, Fuchs M, Chan ME, Orkin S, Nevins JR, Robinson ML and Leone G. HM, Livingston DM and Amati B. (2004). Mol. Cell. Biol., (2001). Nature, 414, 457–462. 24, 4546–4556. Yam CH, Fung TK and Poon RY. (2002). Cell Mol. Life Sci., Taylor-Harding B, Binne UK, Korenjak M, Brehm A and 59, 1317–1326. Dyson NJ. (2004). Mol. Cell. Biol., 24, 9124–9136. Ye X, Zhu C and Harper JW. (2001). Proc. Natl. Acad. Sci. Tedesco D, Lukas J and Reed SI. (2002). Genes Dev., 16, USA, 98, 1682–1686. 2946–2957. Zhang HS, Gavin M, Dahiya A, Postigo AA, Ma D, Luo RX, Thomas DM, Carty SA, Piscopo DM, Lee J, Wang W, Harbour JW and Dean DC. (2000). Cell, 101, 79–89. Forrester WC and Hinds PW. (2001). Mol. Cell, 8, Zhang HS, Postigo AA and Dean DC. (1999). Cell, 97, 53–61. 303–316. Zhang J, Gray J, Wu L, Leone G, Rowan S, Cepko CL, Trimarchi JM and Lees JA. (2002). Nat. Rev. Mol. Cell Biol., Zhu X, Craft CM and Dyer MA. (2004a). Nat. Genet., 36, 3, 11–20. 351–360. Vandel L, Nicolas E, Vaute O, Ferreira R, Ait-Si-Ali S and Zhang J, Schweers B and Dyer MA. (2004b). Cell Cycle, 3, Trouche D. (2001). Mol. Cell. Biol., 21, 6484–6494. 952–959. Vanderluit JL, Ferguson KL, Nikoletopoulou V, Parker M, Zhu L, Harlow E and Dynlacht BD. (1995). Genes Dev., 9, Ruzhynsky V, Alexson T, McNamara SM, Park DS, 1740–1752. Rudnicki M and Slack RS. (2004). J. Cell Biol., 166, 853– Zhu W, Giangrande PH and Nevins JR. (2004). EMBO J., 23, 863. 4615–4626.

Oncogene