SCF-Mediated Protein Degradation and Cell Cycle Control

SCF-mediated protein degradation and cell cycle control

Xiaolu L Ang1 and J Wade Harper*,1

1Program in Biological and Biomedical Sciences, Department of Pathology, Harvard Medical School, Boston, MA 02115, USA

The regulatory step in ubiquitin (Ub)-mediated protein established that targeted protein degradation is a key degradation involves recognition and selection of the conduit for regulating protein levels, what is now most target substrate by an E3 Ub-ligase.E3 Ub-ligases evoke important is being able to understand the specific sophisticated mechanisms to regulate their activity mechanisms as to how this proteolysis contributes to temporally and spatially, including multiple post-transla- the regulation of cell cycle states and phase transitions. tional modifications, combinatorial E3 Ub-ligase path- This review aims to examine how various E3Ub- ways, and subcellular localization.The phosphodegrons of ligases – which are responsible for substrate recognition many substrates incorporate the activities of multiple in the regulatory step of the Ub-mediated proteolytic kinases, and ubiquitination only occurs when all necessary pathway – recognize their targets, and we focus on three phosphorylation signals have been incorporated.In this particular emerging themes in the regulation of Ub- manner, the precise timing of degradation can be mediated proteolysis. Using exemplary findings, we will controlled.Another way that the Ub pathway tightly discuss: (1) how the interaction between the E3Ub- controls the timing of proteolysis is with multiple E3 Ub- ligase and its substrate is often dependent upon ligases acting upon a single target.Lastly, subcellular phosphorylation; (2) how multiple E3Ub-ligases are localization can either promote or prevent degradation by often involved in the targeting of cell cycle proteins; and regulating the accessibility of kinases and E3 Ub-ligases. (3) how the subcellular localization of substrates and E3 This review highlights recent findings that exemplify these Ub-ligases is used to regulate proteolysis. emerging themes in the regulation of E3 Ub-ligase substrate recognition. Oncogene (2005) 24, 2860–2870. doi:10.1038/sj.onc.1208614 Ub-mediated degradation and the SCF Keywords: proteolysis; ubiquitin; SCF; APC; E3ubiquitin- (Skp1/Cullin/F-box) complex in cell cycle control ligase; phosphodegron Regulated proteolysis occurs through the Ub-mediated proteolytic pathway (reviewed in Hershko and Ciechan- over, 1998). This pathway governs the formation of a covalent multi-Ub chain, which serves as a cellular degradation signal, onto a target substrate. The Introduction ubiquitination pathway sequentially involves the E1 Ub-activating enzyme, E2 Ub-conjugating enzymes, and Cellular duplication and differentiation requires me- E3Ub-ligases. Of particular interest is the last step of chanisms that integrate cellular growth, genomic repli- the pathway where an E3Ub-ligase facilitates transfer of cation, and equal nuclear division. Misregulation of any Ub – a small molecule comprised of 76-amino-acid aspect of the cell cycle may understandably lead to residues – onto a lysine residue on the target substrate or developmental defects and diseases such as cancer. onto Lys-48 of a growing Ub chain extending from the Given that the regulation of the cell cycle is so critical substrate. Given that this multi-Ub chain is sufficient to to an organism’s well being, coordinated regulatory promote the protein’s unfolding and degradation by the mechanisms have evolved to establish a specific config- 26S proteasome, substrate ubiquitination must occur in uration of interacting proteins that define a particular a controlled and specific manner. Insight into how E3 cell cycle phase or that mediate transitions between these Ub-ligases recognize specific substrates and what con- phases. In addition to transcriptional upregulation and trols this interaction is vital for understanding the protein activation, targeted ubiquitin (Ub)-mediated regulation of the Ub pathway. proteolysis provides a swift and precise means by which The E3Ub-ligases that perform the rate-limiting to regulate the abundance of specific proteins and selectivity step of direct substrate recognition are the therefore the level of their activities (reviewed in Tyers most numerous and varied collection in the Ub path- and Jorgensen, 2000; Reed, 2003). While it has been well way. They can be classified into HECT domain, Ring- H2-finger domain, and U-box domain superfamilies (Kamura et al., 1999; Ohta et al., 1999; Seol et al., 1999; *Correspondence: JW Harper; Skowyra et al., 1999; Joazeiro and Weissman, 2000; E-mail: [email protected] Hatakeyama et al., 2001; Fang et al., 2003; Ingham SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2861 et al., 2004). The two principal E3Ub-ligases involved in of phosphorylated residues on the substrate that directly cell cycle control are the SCF complex and the APC/C interact with a protein–protein interaction domain in an (anaphase-promoting complex/cyclosome). They are E3Ub-ligase (e.g. an F-box protein), thereby linking the both members of the Ring-H2-finger-containing E3 substrate to the conjugation machinery (Winston et al., Ub-ligase family. SCF complexes are used throughout 1999c; Nash et al., 2001). Additionally, phosphorylation the cell cycle, whereas APC functions primarily during allows temporal regulation of substrate degradation, mitosis and G1 phase. even if SCF complexes are constitutively active through- The SCF complex has three primary subunits: Skp1, out the cell cycle. Cullin, and Rbx1/Roc1 (reviewed in Deshaies, 1999; The best characterized example of phosphodegron Koepp et al., 1999). This complex can interact with an utilization involves the G1/S transition in S. cerevisiae, assortment of modular F-box proteins, which all share where the WD40 propeller structure of the F-box an F-box motif that was first identified in yeast Cdc4 protein Cdc4p recognizes sequences in Sic1p after their and cyclin F and that was shown to interact directly with phosphorylation by Cln1/2-Cdc28 kinases (Nash et al., Skp1 (Bai et al., 1996). The Cul1/Rbx1 components 2001; Orlicky et al., 2003). Sic1p is a Cdk inhibitor, and form the E3Ub-ligase core that associates with E2 Ub- its degradation frees Cln5/6-Cdc28 to initiate DNA conjugating enzymes. F-box proteins directly recruit replication. Sic1p contains as many as nine Cdk ubiquitination substrates and bridge the interaction phosphorylation sites, which individually bind weakly between the E2 and the substrate (Feldman et al., to Cdc4p. However, upon phosphorylation of a mini- 1997; Skowyra et al., 1997). Thus, the identity of the mum of six Cdk sites, Sic1p is efficiently recognized by F-box protein determines the target of the SCF. There Cdc4p, ubiquitinated, and degraded (Nash et al., 2001). are a variety of F-box proteins including subfamilies In this manner, Sic1p acts as a gauge for Cln-Cdc28 with WD40 domains (Fbw) or with leucine-rich repeats activity by accumulating phosphates until a threshold (Fbl) (Cenciarelli et al., 1999; Winston et al., 1999b; Jin number of phosphorylation events promotes its turn- et al., 2004). Moreover, F-box proteins frequently over (Deshaies and Ferrell, 2001; Harper, 2002). More- recognize their substrates in a manner dependent upon over, suboptimal phosphorylation sites confer particular modifications, for example, phosphorylation ultrasensitivity to the system, as individual Sic1 proteins or glycosylation (Feldman et al., 1997; Skowyra et al., ‘resist’ ubiquitination when Cln-Cdc28 activity is 1997; Nash et al., 2001; Yoshida et al., 2002; Orlicky increasing, but still too low to signal for Sic1p to release et al., 2003; Wu et al., 2003). The SCF complex itself is Clb-Cdc28 from inhibition. Thus, the system avoids also subject to post-translational modifications in the ‘noise’ or premature Clb-Cdc28 activation, and instead form of neddylation on the Cullin subunit (Lammer creates a switch-like transition where Clb-Cdc28 will be et al., 1998). Neddylation describes the linkage of Nedd8 activated in a uniform manner. – a small 76-residue protein with sequence similarity to In the case of Sic1p, phosphorylation by kinase Cln- Ub – to a lysine residue on the substrate. Neddylation Cdc28 appears to be sufficient for recognition by Cdc4p. occurs on the winged-helix B domain of Cullins, which is In contrast, many targets of mammalian SCF complexes part of the domain that forms a globular tail binding require a complex interplay between multiple kinases or Rbx1 and the E2-conjugating enzyme (Zheng et al., kinase pathways in order to generate the requisite 2000; Kawakami et al., 2001). Although Nedd8 con- phosphodegron on the SCF target. For example, jugation is known to dramatically increase Ub transfer SCFb-TRCP substrates IkBa, b-catenin, Emi1 (early by E2’s bound to the Rbx1/Cullin complex, the precise mitotic inhibitor 1), Cdc25A, and Wee1 require the mechanism for this stimulation is unknown. combined activities of multiple kinases to create their phosphodegron motifs (Figure 1) (Winston et al., 1999c; Liu et al., 2002; Jin et al., 2003; Margottin-Goguet et al., 2003; Watanabe et al., 2004). Early studies identified a Substrate recognition by the SCF complex is dependent small peptide motif (i.e. phosphodegron) on both IkBa upon phosphorylation and b-catenin that was able to bind to b-TRCP in a phosphorylation-dependent manner (Winston et al., Most substrates require phosphorylation to interact 1999c). The motif or a closely related sequence with the F-box protein in an SCF complex. The SCF (DpSGFXpS, where F is a hydrophobic amino acid pathway, in fact, was elucidated through efforts to and pS is phosphoserine) has been found to be involved understand how phosphorylation controls the abun- in the majority of SCFb-TRCP substrates. In these cases, dance of cyclin and Cdk inhibitors in the Saccharomyces dual phosphorylation within the motif is important for cerevisiae cell cycle (Bai et al., 1996; Feldman et al., high-affinity binding. With IkK, a kinase complex 1997; Skowyra et al., 1997). Its discovery provided a containing both I-kappa kinase a and I-kappa kinase direct linkage between protein degradation and a b is required for proper IkBa phosphorylation on Ser-32 protein kinase orchestrating the cell cycle. and Ser-36 to create the phosphodegron. In contrast, Within the context of a complex cellular environment, generation of the analogous phosphorylation event in phosphorylation allows for substrate discrimination by b-catenin by GSK3b requires prior phosphorylation by the SCF complex due to its ability to specifically casein kinase 1, which serves as a priming enzyme for recognize phosphorylated motifs called phosphode- GSK3b (Figure 1). However, recent work suggests that grons. A phosphodegron is defined as one or a series b-TRCP has the capability to bind to noncanonical

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2862 residues to create the EpSTDpSGF phosphodegron, βTRCP with Ser-79 and Ser-82 being phosphorylated (Jin et al., 2003). Thus, the phosphorylation site following the P P DSG tripeptide is missing in the Cdc25A degron. DSGXXS Currently, it is unclear whether Ser-79 phosphorylation is required for interaction of Cdc25A with b-TRCP or is simply required for phosphorylation of Ser-82, although IκKα and IκKβ studies with synthetic peptides indicate that phosphorylation at neither Ser-76 (Chk1 site) nor Ser-79 is P P absolutely required for binding to b-TRCP. In addition, EDS GLDS IκBα Wee1 employs a related EDpSAF sequence but, like 32 36 Cdc25A, lacks the second phosphoserine (Watanabe et al., 2004). This phosphorylation event is carried out by polo-kinase 1 (Plk1) in a manner that appears to GSK3 CKIα depend upon prior phosphorylation of Ser-123by Cdc2. It is possible that phospho-Ser-123is directly involved in P P P P the interaction with b-TRCP or, more likely, phosphor- LDS GIHS GATT TAPS LS β-catenin ylation at this site provides a platform for binding of 33 37 41 45 Plk1 to Wee1 through its polo-box, which binds to Cdc2 consensus sites. Like Wee1, Emi1 is also targeted for degradation by the combined action of Cdc2 and Plk1, Cdk1 PLK1 as discussed below (Hansen et al., 2004; Moshe et al., 2004). Additional examples of dual phosphorylation mechanisms are discussed below with respect to targets P P P P PPP of the SCFFbw7 complex. S Emi1 T S S EDS145GYSS149 S There are many implications to having a degron undergo multiple phosphorylations by distinct kinases. First, such a phosphodegron has the capacity of Chk1 Kinase X integrating multiple input signals into a single molecule that can serve as a read out for the status of upstream P PP pathways. The cell likely has to monitor many different events and environmental cues throughout its cycle such SSES TDS GLDS Cdc25A 79 82 as the successful completion of mitotic checkpoints or whether its environment contains sufficient nutrients before entering S phase. Phosphorylation-dependent PLK1 Cdc2 protein degradation may be used to signal completion of previous actions. Another consequence of the multi- P P phosphorylated degron is that each kinase pathway now EDS AF S PVK Wee1 becomes dependent upon the actions of the other to 53 123 efficiently target the substrate for degradation. Figure 1 Multiple kinases regulate the degradation of SCF substrates, including that of b-TRCP. The priming kinase, depicted in italics, first phosphorylates the substrate to prime it for subsequent binding by the second kinase. The second kinase completes formation of the phosphodegron consensus motif (in Cdh1 and its negative regulation of the Skp2 oncogene bold) by phosphorylating residues that can now be directly recognized by b-TRCP. The absence of the priming activity results The F-box protein Skp2 (Fbl1) is required for degrada- in an inability for the second kinase to phosphorylate efficiently the phosphodegron, as well as the stabilization of the substrate. The tion of the p27 Cdk inhibitor during S-phase and is absence of the second kinase activity also prevents efficient essential for proper S-phase progression (Pagano et al., substrate turnover, as the phosphodegron motif remains unpho- 1995; Carrano et al., 1999). Moreover, Skp2 has been sphorylated and unrecognizable by b-TRCP attributed oncogenic properties including the ability to promote inappropriate T-lymphocyte proliferation in CD4-Skp2 transgenic models, where ectopic expression of Skp2 leads to a loss of p27 and collaborates with PTEN tumor suppressor downregulation (Latres et al., phosphodegrons. One such example is found with the 2001; Yang et al., 2002). Correspondingly, Skp2 is found cell cycle regulator Cdc25A. In response to DNA to be upregulated in a wide variety of human cancers, damage, Cdc25A is targeted for destruction by a including those affecting the breast, lung, colorectal, and pathway that depends upon the Chk1 kinase and b- prostate tissues (Hershko et al., 2001; Signoretti et al., TRCP (Busino et al., 2003; Jin et al., 2003). Phosphor- 2002; Yang et al., 2002; Yokoi et al., 2002, 2004). Cells ylation by Chk1 is thought to prime Cdc25A for expectedly have many ways to regulate Skp2 activity. phosphorylation by a yet unidentified kinase on nearby First, Skp2 expression is cell cycle regulated, with SKP2

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2863 transcription being activated at the G1/S transition, metaphase – signals the cell to prepare for mitotic exit possibly via E2F. Second, as part of an SCF complex, by initiating the elimination of mitotic cyclin activity Skp2 is subject to turnover through an autoubiquitina- (Shirayama et al., 1999). This is accomplished in two tion pathway (Wirbelauer et al., 2000) analogous to that phases. First, APCCdc20 targets the bulk of cyclins seen with other F-box proteins (Zhou and Howley, including Clb3and Clb5. The second phase involves 1998). Interestingly, Skp2 autoubiquitination is nega- the full activation of APCCdh1 to govern the virtually tively regulated by Cks1, a Skp2 binding protein that is complete elimination of Clb2, which is required for cells an essential cofactor for p27 ubiquitination (Wang et al., to exit mitosis (Wasch and Cross, 2002). Full activation 2004). Increased destabilization of Skp2 in response to of APCCdh1 arises by a lowering of overall Cdk activity transforming growth factor-b has been linked to loss of through APCCdc20 activity and release of Cdc14 phos- expression of Cks1, whose interaction with Skp2 phatase through the mitotic exit network and associated appears to physically protect it from autoubiquitination pathways (Visintin et al., 1999; Stegmeier et al., 2002). (Wang et al., 2004). Furthermore, to smooth the transition from APCCdc20 to Another recently identified mechanism to regulate APCCdh1, APCCdh1 is able to mediate the destruction of Skp2 directly employs APCCdh1. The APC comprises APCCdc20 in late mitosis (Prinz et al., 1998; Shirayama approximately 12 subunits, and its regulation is et al., 1998). Targeted proteolysis thus reinforces the governed by at least two activating subunits Cdc20 feedback loop that ultimately raises APCCdh1 levels over (fizzy) and Cdh1 (Hct1 and fizzy related) (reviewed in that of APCCdc20 (Harper et al., 2002; Peters, 2002). Harper et al., 2002; Peters, 2002). Each of the activating Recently, two groups reported a direct role for subunits confers a different specificity to the APC, and APCCdh1 in the degradation of Skp2 and Cks1 (Figure 2) each directly associates with the APC under different (Bashir et al., 2004; Wei et al., 2004). Targeting of cellular conditions. Cdc20 associates with the APC SCFSkp2/Cks1 by APCCdh1 serves to stabilize p27, under conditions of high Cdk activity, and Cdh1 subsequently leading to the maintenance of G1 phase activation is prevented by high Cdk activity. Similar to and blocking premature S-phase initiation. In the the role of F-box proteins in the SCF complex, these presence of a nondegradable form of Skp2, or upon activating subunits can directly engage substrates either depletion of Cdh1 by RNAi, populations of cells by recognition of destruction box motifs by Cdc20 or by released from nocodazole arrest enter S phase more KEN boxes in the case of Cdh1. rapidly, as judged by higher levels of BrdU incorpora- The activation of APCCdh1 results from the following tion. sequence of events, much of which was elucidated in An interesting question is how do SCFSkp2/Cks1 levels yeast models. Elimination of Pds1 by APCCdc20 – ever accumulate to sufficiently high levels to promote allowing for the segregation of sister chromatids at appropriate S-phase entry, given that proteolysis is such

Cdh1 P P c-myc APC Cyclin E LLTPPQS c-Myc LPTPPLS ERK Ub E2 CKB1 S62 Skp2 Cyclin E P SCF p27 c-myc Cdk2

Cdk2 GSK3

GSK3 S62 P T58 P T384 c-myc P P T380 Cyclin E T62 P Ub E2 Cdk2 Rbx1

Ub Fbw7 E2 Cyclin E SCF

Cul3 BTB? P Skp1 Notch constitutive?

Figure 2 Multiple E3Ub-ligases regulate cell cycle proteins. Free cyclin E is turned over by an SCF3complex, while cyclin E-Cdk2 regulation involves an intricate E3Ub-ligase pathway involving APC Cdh1, SCFSkp2/Cks1, and SCFFbw7. The various E3Ub-ligases are likely to be coordinated temporally and spatially. SCFFbw7 also serves to ubiquitinate cell cycle protein c-Myc and Notch, which both harbor oncogenic potential

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2864 an efficient and irreversible mechanism and G1 is an abundance comes from the finding that this region of c- intrinsically stable state? In other words, what is the cue Myc is among the most frequently mutated regions in that causes an equilibrium shift in G1 to release Skp2 the protein. Many of the mutations, such as T58I, are and Cks1 from APCCdh1-mediated degradation? The known to stabilize dramatically the protein (Bhatia answer remains unknown. One possibility lies in et al., 1993; Bahram et al., 2000; Gregory and Hann, previous work suggesting that even though SCFSkp2 2000). destroys most of the p27 population during S phase, a Recent studies have revealed multiple mechanisms small fraction of p27 actually relocalizes to the controlling c-Myc abundance and activity. Concurrent cytoplasm in G1 (Rodier et al., 2001; Ishida et al., papers by Kim et al. (2003) and von der Lehr et al. 2002). The accumulation of cytoplasmic p27 occurs (2003) first demonstrated a role for Skp2 in influencing through phosphorylation of Ser-10 by hKIS in the c-Myc stability, and also surprisingly a role for Skp2 in nucleus and phosphorylation of Thr-187 by Akt in the promoting c-Myc transcriptional activation (reviewed in cytoplasm to export and retain p27, respectively (Boehm Jin and Harper, 2003). Skp2 associates with c-Myc, but et al., 2002; Fujita et al., 2002; reviewed in Reed, 2002). in contrast to early expectations, this association The sensitivity of hKIS and Akt to environmental occurred through myc-box II (MB-II) and helix–loop– nutrients and mitogens exposes them as prime candi- helix zipper (HLH-Zip) domains in c-Myc, did not dates to be ‘liberators of Cdks’ through the export of require phosphorylation at Thr-58, and appeared to be p27 during G1. Consequently, activation of this small phosphorylation-independent. Skp2 overexpression pro- pool of nuclear Cdk’s may be able to initiate a pathway moted c-Myc ubiquitination in vitro, with a mild effect that could lead to the inactivation of Cdh1 through Cdk on c-Myc stability. However, the major effect of Skp2’s phosphorylation, and thus S phase can initiate (Ang and interaction with c-Myc appeared to be a dramatic Harper, 2004; Guardavaccaro and Pagano, 2004). activation of c-Myc activity, as assessed in tissue culture Inhibition of Cdh1 by Emi1, an E2F-regulated gene cells. This was interesting in view of the fact that Skp2 is (Hsu et al., 2002), is also likely to be involved in the an oncogene and could possibly collaborate with c-Myc switch mechanism controlling the G1/S transition, but to promote cell proliferation. Two questions arise: (1) If how Emi1 function fits into p27 destruction prior to E2F c-Myc activity requires Skp2, then why is the phenotype activation is unclear at present. of Skp2 null mice relatively mild, and how can its phenotype be largely reversed by deletion of the Skp2 target p27 (Nakayama et al., 2004)? (2) If Skp2 is not involved in the GSK3b-dependent pathway, what E3 c-Myc turnover by Fbw7/Ago Ub-ligase complex controls c-Myc turnover via the phosphorylation-dependent pathway? c-Myc is a basic-helix–loop–helix/leucine zipper (bHLH/ While the answer to the first question is still unknown, Zip) transcription factor that can form heterodimers recent work has established Fbw7/Ago as the primary with a variety of DNA binding proteins – including mechanism for phosphorylation-dependent degradation Max, Mad, and Mnt – to bind consensus DNA of c-Myc in Drosophila and human cells via the GSK3b sequences located in 50 regulatory regions within pathway (Moberg et al., 2004; Welcker et al., 2004a; promoters (reviewed in Levens, 2003). All heterodimer Yada et al., 2004). Archipelago (ago) in Drosophila and complexes in the c-Myc network are capable of binding Fbw7 in mammals are the apparent orthologs of the well-studied E-box motif (CACGTG). In general, budding yeast Cdc4p and recognize related phosphode- activation of this c-Myc transactivational network grons (Koepp et al., 2001; Nash et al., 2001). Welcker regulates expression of target genes to direct cellular et al. (2004a) and Yada et al. (2004) found that removal growth, proliferation, differentiation, and apoptosis. As of Fbw7 or overexpression of a dominant-negative form expected, c-Myc is often hijacked and upregulated in of Fbw7 leads to stabilization of c-Myc. Furthermore, it human cancers such as Burkitt’s lymphoma, myeloid was shown that Fbw7 directly binds to c-Myc and that and plasma cell leukemia, carcinomas of the breast, this binding is dependent upon Thr-58 phosphorylation cervix, lung, and colon, osteosarcoma, and glioblastoma by GSK3b. Lastly, SCFFbw7 promotes GSK3b-depen- (reviewed in Spencer and Groudine, 1991). Although c- dent c-Myc ubiquitination in vitro. In vitro, c-Myc dually Myc can be upregulated due to translocation, defects in phosphorylated on Thr-58 and Ser-62 binds to Fbw7 turnover pathways may also contribute to elevated and phosphorylation on Ser-62 is required for priming levels of c-Myc found in cancer. c-Myc is a highly the binding of GSK3b to c-Myc; however, it has recently unstable protein and is known to be degraded through been proposed that Ser-62 phosphorylation must be the proteasome pathway. Previous studies have impli- removed by PP2A prior to c-Myc ubiquitination (Yeh cated phosphorylation of two residues in myc-box I et al., 2004). Phosphorylation of Thr-58 and Ser-62 (MB-I) in its turnover (Henriksson et al., 1993; generates a phosphodegron that interacts with Fbw7. Lutterbach and Hann, 1994; Pulverer et al., 1994). This phosphodegron is similar in sequence to that found Phosphorylation on Ser-62 by Erk family kinases is in cyclin E, which is also recognized by Fbw7 (Figure 2). thought to prime c-Myc for phosphorylation by GSK3b Using clonal analysis, Moberg et al. found that on Thr-58, with Thr-58 phosphorylation being central to archipelago (ago) mutant cells have a growth advantage mitogen-induced c-Myc turnover (Sears et al., 2000). compared to wild-type (WT) cells. Ago (À/À) cells The critical importance of Thr-58 in controlling c-Myc appear to be maintained at the same size as WT, but

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2865 they overtake WT cells in tissue due to an increased Fbw7 has also been demonstrated to mediate the frequency of cell division and shortened cell cycle degradation of cyclin E (Figure 2) (Koepp et al., 2001; duration. It was also determined in these same clonal Moberg et al., 2001; Strohmaier et al., 2001). studies, as well as with RNAi, that loss of ago results in Cyclin E and Cdk2 have been implicated in G1/S increased Drosophila Myc (dMyc) levels in vivo. control, and together, they form a complex that is Furthermore, in addition to showing direct phosphor- involved in centrosome duplication, DNA replication, ylation-dependent interaction between and dMyc, con- and activation of histone transcription (Reed, 2003; Ye sistent with the work of Weckler et al. and Tada et al., et al., 2003). Ectopic expression of cyclin E in Moberg et al. demonstrated genetic interactions by mammalian cells results in a premature and unusually using an ago allele to rescue dMyc organism-size and long S phase, as well as genomic instability. Its wing compartment-size defects. oncogenic impact on the cell cycle may be witnessed How then are the activities of Skp2 and Fbw7 toward in its dramatic upregulation in many tumor types, c-Myc spatially and temporally regulated? It is quite and high cyclin E levels even correlate directly with likely, based on the available data, that Skp2 and Fbw7 cancer severity and inversely with patient survival rates play opposing roles in c-Myc activity. Skp2 may (Porter et al., 1997; Strohmaier et al., 2001; Spruck predominantly function to increase c-Myc activity and et al., 2002). this may occur primarily during S phase, when Skp2 There are at least two pools of cyclin E in the cell, and levels are the highest. Concomitant with c-Myc activa- each is turned over by a different degradation pathway. tion by Skp2 and ubiquitination is the turnover of c- Free cyclin E – not bound to Cdk2 – interacts with a Myc through the proteasome. Conceivably, c-Myc may Cullin 3-based complex in a phosphorylation-indepen- be activated by ubiquitination and then destroyed after dent manner, and may accordingly control the it has engaged in a productive transcription complex. abundance of free cyclin E throughout the cell cycle Hence, Skp2 may be targeting only the small fraction of (Singer et al., 1999; reviewed in Winston et al., 1999a). c-Myc in the cell that engages in transcription at that Cul3forms a modular E3analogous to the SCF but moment. Inherent in this model is that the Skp2–c-Myc uses BTB proteins as specificity factors (Furukawa interaction occurs on chromatin. In contrast, recent et al., 2003; Geyer et al., 2003; Pintard et al., 2003; Xu studies indicate that Fbw7 may be selectively ubiquiti- et al., 2003). BTB proteins merge the functional nating c-Myc within the nucleolus (Welcker et al., properties of Skp1 and F-box proteins into a single 2004b). The g isoform of Fbw7 is located in the polypeptide chain (Xu et al., 2003). The BTB domain nucleolus of tissue culture cells. Its ablation by RNAi adopts a conserved a/b-fold that interacts with Cul3in a leads to increased cell size and increased accumulation manner similar to that observed with the Skp1/Cul1 of c-Myc, whereas depletion of the cytoplasmic b form complex. BTB proteins often contain additional pro- of Fbw7 does not lead to these phenotypes. The tein–protein interaction domains that can directly importance of the nuclear a isoform is uncertain at the recruit substrates to the Cul3-based SCF (SCF3) moment. Interestingly, expression of c-Myc bearing a complex for ubiquitination. A role for Cul3in degrada- Thr-58 to alanine mutation abrogates the increase in cell tion of free cyclin E suggests the involvement of a BTB size seen with depletion of Fbw7g. Moreover, inhibition protein. However, such an adaptor protein has not yet of the proteasome leads to the accumulation of c-Myc in been identified for cyclin E. the nucleolus. Depletion of Fbw7g by RNAi closely The second pool of cyclin E is bound to Cdk2, and it mimics this, indicating that the pathway leading to c- is thus protected from Cul3-mediated turnover (Clur- Myc degradation in response to phosphorylation by man et al., 1996; Won and Reed, 1996). Previous studies GSK3b involves passage through the nucleolus. Thus, have implicated SCFFbw7 in turnover of Cdk2-bound the effects of Skp2 and c-Myc likely occur in different cyclin E and this occurs in a phosphorylation-dependent cellular locations, occurs in response to different signals, manner (Koepp et al., 2001; Moberg et al., 2001; and appear to have different consequences with respect Strohmaier et al., 2001). A major determinant for cyclin to c-Myc function. E ubiquitination is phosphorylation of Thr-380 (Clur- man et al., 1996; Won and Reed, 1996). Phosphoryla- tion at this site is catalysed by two kinases acting redundantly, Cdk2 and GSK3b (Welcker et al., 2003). Fbw7 functions generally in the degradation of oncogenes Mutation of Thr-380 to alanine leads to dramatic stabilization of cyclin E and reduces its interaction with Involvement of the SCFFbw7 complex in phosphoryla- Fbw7. In addition, synthetic peptides encompassing tion-dependent c-Myc turnover is consistent with Thr-380 are capable of binding tightly to Fbw7 (Koepp suggestions of Fbw7 functioning as a recessive tumor et al., 2001). Cyclin E phosphorylation, however, is suppressor gene (Moberg et al., 2001; Strohmaier et al., complex and it is currently unclear precisely how each of 2001; Spruck et al., 2002; Rajagopalan et al., 2004). these phosphorylation events contributes to cyclin E Fbw7 further functions in the ubiquitination of other regulation and the recognition by Fbw7. At least two proteins with oncogenic potential. It was first identified additional phosphorylated residues play roles in cyclin E in Caenorhabditis elegans as Sel-10, which interacts with turnover (Strohmaier et al., 2001; Welcker et al., 2003; LIN-12/Notch proteins to promote their degradation Ye et al., 2004). Mutation of T62 to alanine leads to (Hubbard et al., 1997). In addition to c-Myc and Notch, stabilization of cyclin E. This effect appears to be due

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2866 to both direct and indirect mechanisms. First, cyclin Regulation of Ub-mediated protein degradation by ET62A displays a greatly reduced extent of Thr-380 subcellular localization phosphorylation, as assessed using phospho-specific antibodies (Ye et al., 2004). Reduced phosphorylation In addition to temporal control achieved by having on Thr-380 would be expected to stabilize dramatically phosphorylation regulate the association of an E3Ub- cyclin E. However, synthetic peptides encompassing ligase with its substrate, regulation of E3Ub-ligase Thr-62 have the capacity to bind to Fbw7, and activity is similarly sophisticated with regard to its therefore, in principle, this phosphorylation event could spatial organization and coordination. Subcellular function as a phosphodegron for cyclin E. Indeed, localization is thus implemented to prevent ubiquitina- residual association of cyclin ET380 with Fbw7 is tion from occurring or to bring together components abolished when Thr-62 is simultaneously mutated that will lead to the desired degradation. to alanine, consistent with a small direct role for Thr-62 phosphorylation in Fbw7 recognition (Ye et al., 2004). Additional studies indicate a role for Ser-384 in controlling cyclin E turnover, which is Localization of Swe1p and Cdc5 to the septin collar likely phosphorylated by Cdk2 (Welcker et al., 2003; Ye et al., 2004). Its mutation to alanine stabilizes The regulation of budding yeast Swe1p (Wee1 ortho- the protein, although not to the extent seen with log), which inhibits the Clb–Cdc28 complex from Thr-380 mutation. In addition, mutation to alanine activating premature mitosis during S and G2 phase, reduces the association with both Fbw7a and involves subcellular localization. Upon hyperphosphor- Fbw7b, without affecting Thr-380 phosphorylation ylation, Swe1p is subject to Ub-mediated degradation (Ye et al., 2004). Thus, phosphorylation of at least (Sia et al., 1998; Sreenivasan and Kellogg, 1999). three residues in cyclin E contributes to its recognition However, when there are defects in the formation of by Fbw7. The kinase involved in phosphorylation of the septin collar, Swe1 remains hypophosphorylated and Thr-62 has not been identified, but would be expected to stabilized (Barral et al., 1999; Shulewitz et al., 1999). collaborate with Cdk2 and GSK3b in controlling cyclin This results in an extended G2 phase and elongated cells, E abundance. In addition, while all three isoforms of as there is a failure to switch from polarized to isotropic Fbw7 can promote cyclin E turnover in a cotransfection growth during budding. This series of events has been assay, it remains to be determined whether specific referred to as the morphogenetic checkpoint. In a recent Fbw7 isoforms normally regulate endogenous cyclin E paper, Sakchaisri et al. (2004) investigate the kinases turnover. involved in Swe1p regulation. They first screen six As mentioned above, cyclin E overexpression has kinases known to associate with septin filaments at the been linked with genomic instability in human cells bud neck, and found that both Cla4 and Cdc5 (Plk1) are (Spruck et al., 1999). Recent studies have demonstrated capable of additively phosphorylating Swe1 in an in an analogous phenotype in tissue culture cells that are vitro kinase assay. Cdc5 activity is essential for Swe1p deleted for the FBW7 gene (Rajagopalan et al., 2004). degradation in vivo, and in Cdc5-deficient cells, Swe1p Colon cancer cell lines lacking Fbw7 display increased remains stabilized at the bud neck. Moreover, a levels of cyclin E expression and genomic instability. To catalytically active Cdc5 truncation mutant (CdcDC) understand the role of Fbw7 in development, two that cannot localize to the bud neck is unable to groups created mice lacking the FBW7 gene (Tetzlaff promote Swe1p degradation, suggesting that the Cdc5 et al., 2004; Tsunematsu et al., 2004). Mice lacking specifically at the neck is responsible for Swe1p turn- Fbw7 die at Bday 10 of gestation with multiple over. Costaining with tubulin and Cdc5 demonstrated abnormalities, including defects in hematopoietic and that Cdc5 localizes to the bud neck prior to mitosis, vascular development, and heart morphogenesis (Tet- consistent with its role in local Swe1p turnover. Cla4- zlaff et al., 2004; Tsunematsu et al., 2004). Moreover, depleted cells, interestingly, harbored little alterations placentas from FBW7À/À mice displayed an increased with regard to Swe1p stability or septin filament number of giant trophoblast cells undergoing DNA assembly when arrested with nocodazole in mitosis. synthesis. These cells displayed increased levels of cyclin However, when arrested in S phase with hydroxyurea, E, as expected based on the phenotypes observed with these Cla4-depleted cells contain hypophosphorylated loss of FBW7 in tissue culture cells (Tetzlaff et al., 2004). Swe1p. Together, the data support a role for Cla4 In the embryo, evidence for altered expression of notch activity in S phase contributing to Cdc5 phosphoryla- proteins (particularly Notch-1 and Notch-4), as well as tion of Swe1p at the septin ring upon the onset of increased levels of expression of notch target genes metaphase. Since Swe1p degradation is confined and (Hes1, Herp1, and Herp2), was observed (Tetzlaff et al., localized to the septin ring by first localizing Cdc5, 2004; Tsunematsu et al., 2004). Cyclin E levels were also Sakchaisri et al. propose that the septin ring at the bud elevated in the embryo, when reduced levels of CYCLIN neck acts as a scaffold or a platform for the sequential E messenger RNA is taken into account (Tetzlaff et al., phosphorylation and targeted degradation of Swe1 to 2004). Thus, the available data indicate multiple roles occur smoothly, accurately, and efficiently (Figure 3a). for Fbw7 in controlling the abundance of a number of The identity of the E3required for Swe1 turnover is important signaling molecules including proteins that unknown, but in mammalian cells, SCFb-TRCP has act as oncogenes. been implicated in the turnover of Wee1 upon its

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2867 a activation. Hansen et al. (2004) and Moshe et al. (2004) recently demonstrate that Plk1 catalyses the formation of the Emi1 phosphodegron that is recognized by SCFb- TRCP. The resulting degradation of Emi1 releases the APC from inhibition, and in the presence of cyclin B/ Cdk1 activity, APCCdc20 accumulates. Both labs establish septin collar that mitotically active Plk1 and proper phosphorylation Hsl1 Swe1 Swe1 of sequences in Emi1 are capable of binding to the polo- Hsl7 Swe1 P Cla4 Cdc5 box in Plk1. In vitro assays further demonstrate that P P P P P E3 Plk1 directly binds and phosphorylates Emi1. Interestingly, Emi1 degradation appears to occur speciﬁcally at mitotic spindles (Figure 3b) (Hansen b et al., 2004). In early mitosis, Emi1 localizes to the interchromosomal regions. As mitosis progresses into Plk1 metaphase, Emi1 undergoes redistribution to the spindle poles. Detailed analysis using confocal immunoﬂuores- P β P TRCP spindle pole cence demonstrated that Plk1 accumulates on early Emi1 P mitotic centrosomes and spindle poles (Golsteyn et al.,

Skp1 1994, 1995). Redistribution of Emi1 in metaphase

Ub results in its colocalization with Plk1. Upon closer Cul1 E2 inspection of the interacting proteins using immuno- Rbx1 fluorescence, Hansen et al. discover that although Emi1 and Plk1 colocalize to the assembling spindle poles, Emi1 appears to be excluded from the exact regions on Figure 3 Subcellular localization regulates E3Ub-ligase activity. the centrosomes to where Plk1 localizes. Perhaps, (a) Swe1p localizes to the septin collar following Cla4 phosphorylation, and is possibly tethered there by an Hsl1–Hsl7 complex, so destruction of Emi1 is one of the principle functions of that it can be efficiently phosphorylated by Cdc5. Cdc5 localizes to Plk1 in spindle assembly, and may also provide a reason the septin ring via its C-terminal localization domain. Phosphor- for its localization. ylation of Swe1p by Cdc5 results in Swe1p hyperphosphorylation and the creation of a phosphodegron motif that is recognized by an E3Ub-ligase ( b) Emi1 degradation by SCFb-TRCP occurs at the spindle poles during mitosis. Emi1 is first phosphorylated by Cdk activity and then redistributes at metaphase from interchromoso- Summary and the impact of Ub-mediated proteolysis mal regions to the spindle poles, likely through direct interaction with the polo-box domain in Plk1. Thus, Plk1 is situated to There is no doubt that targeted Ub-mediated proteolysis perform the second round of phosphorylation and promote ubiquitination by SCFb-TRCP represents an important regulatory strategy in cell cycle control. It is swift, precise, and can support numerous regulatory inputs simultaneously. The key feature to the selectivity and timing of this degradation pathway lies in phosphorylation by cyclin B-Cdc2 and Plk1 (Watanabe the interaction between the E3Ub-ligase and its targeted et al., 2004). substrate. The molecular basis for much of this recognition is still being elucidated, nevertheless there are emerging themes in how E3Ub-ligases recognize their targets, how they autoregulate and interact with Emi1 degradation at mitotic spindle poles one another in bone fide pathways, and how their activities are controlled in the context of a complex Emi1 has emerged as a critical regulator of mitotic cellular environment. The themes discussed here include progression and the G1/S transition through its inhibi- a dependency on post-translational modifications to tion of the APC (Reimann et al., 2001). In mitosis, Ub- either activate particular E3Ub-ligases or at the mediated degradation of Emi1 by SCFb-TRCP allows the substrate level where multisite phosphorylation med- activation of APCCdc20 that leads to mitotic exit iates formation of the phosphodegron. There is also the (Margottin-Goguet et al., 2003). Later, at the end of realization that E3Ub-ligases can be networked into G1, Emi1 expression increases once again through complicated pathways to regulate the activities of other upregulation by E2F (Hsu et al., 2002). This results in E3Ub-ligases. And lastly, in a complex cellular APCCdh1 inhibition, Skp2 and cyclin A accumulation, environment, E3Ub-ligases benefit from being able to p27 degradation, and S-phase entry. compartmentalize and restrict their activities to parti- In mammalian cells undergoing mitosis, the kinase cular subcellular regions. activities of cyclin B/Cdk1 and Plk1 have been shown to The importance of the Ub degradation pathway may induce APCCdc20 activation (reviewed in Harper et al., be best witnessed in cases where components of the Ub 2002; Peters, 2002). While cyclin B/Cdk1directly phos- pathway are misregulated, usually leading to carcino- phorylates the APC to promote APCCdc20 assembly and genesis or developmental defects. Many of these activation, Plk1 serves a more indirect role in APC components had been first identified as potent

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2868 oncogenes or tumor suppressors. For example, F-box be discovered with regard to the mechanisms of how protein Skp2 is normally absent in postmitotic cells, and Ub-mediated degradation operates and interacts with as an oncogene, it accumulates to abnormally high levels other cellular pathways. A grasp of these mechanisms in a wide variety of cancers, including those found in will become imperative in understanding the underlying breast, prostate, colorectoal, and lung tissues (Hershko causes of certain diseases, and perhaps may even et al., 2001; Signoretti et al., 2002; Yang et al., 2002; provide an avenue for treatment. Yokoi et al., 2002, 2004). In contrast, FBW7 is often found mutated in ovarian and endometrial cancers Acknowledgements (Moberg et al., 2001; Strohmaier et al., 2001; Spruck The work in this laboratory is supported by NIH Grants et al., 2002). There are inevitably many more aspects to GM54137 and AG11085.

References

Ang XL and Harper JW. (2004). Sci. STKE, 13, pe31. Hershko A and Ciechanover A. (1998). Annu. Rev. Biochem., Bahram F, von der Lehr N, Cetinkaya C and Larsson LG. 67, 425–479. (2000). Blood, 95, 2104–2110. Hershko D, Bornstein G, Ben-Izhak O, Carrano A, Pagano Bai C, Sen P, Hofmann K, Ma L, Goebl M, Harper JW and M, Krausz MM and Hershko A. (2001). Cancer, 91, Elledge SJ. (1996). Cell, 86, 263–274. 1745–1751. Barral Y, Parra M, Bidlingmaier S and Snyder M. (1999). Hsu JY, Reimann JD, Sorensen CS, Lukas J and Jackson PK. Genes Dev., 13, 176–187. (2002). Nat. Cell Biol., 4, 358–366. Bashir T, Dorrello NV, Amador V, Guardavaccaro D and Hubbard EJ, Wu G, Kitajewski J and Greenwald I. (1997). Pagano M. (2004). Nature, 428, 190–193. Genes Dev., 11, 3182–3193. Bhatia K, Huppi K, Spangler G, Siwarski D, Iyer R and Ingham RJ, Gish G and Pawson T. (2004). Oncogene, 23, Magrath I. (1993). Nat. Genet., 5, 56–61. 1972–1984. Boehm M, Yoshimoto T, Crook MF, Nallamshetty S, Ishida N, Hara T, Kamura T, Yoshida M, Nakayama K and True A, Nabel GJ and Nabel EG. (2002). EMBO J., 21, Nakayama KI. (2002). J. Biol. Chem., 277, 14355–14358. 3390–3401. Jin J, Cardozo T, Lovering RC, Elledge SJ, Pagano M and Busino L, Donzelli M, Chiesa M, Guardavaccaro D, Ganoth Harper JW. (2004). Genes Dev., 18, 2573–2580. D, Dorrello NV, Hershko A, Pagano M and Draetta GF. Jin J and Harper JW. (2003). Cancer Cell, 3, 517–518. (2003). Nature, 426, 87–91. Jin J, Shirogane T, Xu L, Nalepa G, Qin J, Elledge SJ and Carrano AC, Eytan E, Hershko A and Pagano M. (1999). Nat. Harper JW. (2003). Genes Dev., 17, 3062–3074. Cell Biol., 1, 193–199. Joazeiro CA and Weissman AM. (2000). Cell, 102, Cenciarelli C, Chiaur DS, Guardavaccaro D, Parks W, Vidal 549–552. M and Pagano M. (1999). Curr. Biol., 9, 1177–1179. Kamura T, Koepp DM, Conrad MN, Skowyra D, Moreland Clurman BE, Sheaff RJ, Thress K, Groudine M and Roberts RJ, Iliopoulos O, Lane WS, Kaelin Jr WG, Elledge SJ, JM. (1996). Genes Dev., 10, 1979–1990. Conaway RC, Harper JW and Conaway JW. (1999). Deshaies RJ. (1999). Annu. Rev. Cell Dev. Biol., 15, 435–467. Science, 284, 657–661. Deshaies RJ and Ferrell Jr JE. (2001). Cell, 107, 819–822. Kawakami T, Chiba T, Suzuki T, Iwai K, Yamanaka K, Fang S, Lorick KL, Jensen JP and Weissman AM. (2003). Minato N, Suzuki H, Shimbara N, Hidaka Y, Osaka F, Semin. Cancer Biol., 13, 5–14. Omata M and Tanaka K. (2001). EMBO J., 20, Feldman RM, Correll CC, Kaplan KB and Deshaies RJ. 4003–4012. (1997). Cell, 91, 221–230. Kim SY, Herbst A, Tworkowski KA, Salghetti SE and Tansey Fujita N, Sato S, Katayama K and Tsuruo T. (2002). J. Biol. WP. (2003). Mol. Cell, 11, 1177–1188. Chem., 277, 28706–28713. Koepp DM, Harper JW and Elledge SJ. (1999). Cell, 97, Furukawa M, He YJ, Borchers C and Xiong Y. (2003). Nat. 431–434. Cell Biol., 5, 1001–1007. Koepp DM, Schaefer LK, Ye X, Keyomarsi K, Chu C, Harper Geyer R, Wee S, Anderson S, Yates J and Wolf DA. (2003). JW and Elledge SJ. (2001). Science, 294, 173–177. Mol. Cell, 12, 783–790. Lammer D, Mathias N, Laplaza JM, Jiang W, Liu Y, Callis J, Golsteyn RM, Mundt KE, Fry AM and Nigg EA. (1995). J. Goebl M and Estelle M. (1998). Genes Dev., 12, 914–926. Cell Biol., 129, 1617–1628. Latres E, Chiarle R, Schulman BA, Pavletich NP, Pellicer A, Golsteyn RM, Schultz SJ, Bartek J, Ziemiecki A, Ried T and Inghirami G and Pagano M. (2001). Proc. Natl. Acad. Sci. Nigg EA. (1994). J. Cell Sci., 107, 1509–1517. USA, 98, 2515–2520. Gregory MA and Hann SR. (2000). Mol. Cell. Biol., 20, Levens DL. (2003). Genes Dev., 17, 1071–1077. 2423–2435. Liu C, Li Y, Semenov M, Han C, Baeg GH, Tan Y, Zhang Z, Guardavaccaro D and Pagano M. (2004). Oncogene, 23, Lin X and He X. (2002). Cell, 108, 837–847. 2037–2049. Lutterbach B and Hann SR. (1994). Mol. Cell. Biol., 14, Hansen DV, Loktev AV, Ban KH and Jackson PK. (2004). 5510–5522. Mol. Biol. Cell, 15, 5623–5634. Margottin-Goguet F, Hsu JY, Loktev A, Hsieh HM, Reimann Harper JW. (2002). Trends Cell Biol., 12, 104–107. JD and Jackson PK. (2003). Dev. Cell, 4, 813–826. Harper JW, Burton JL and Solomon MJ. (2002). Genes Dev., Moberg KH, Bell DW, Wahrer DC, Haber DA and Hariharan 16, 2179–2206. IK. (2001). Nature, 413, 311–316. Hatakeyama S, Yada M, Matsumoto M, Ishida N and Moberg KH, Mukherjee A, Veraksa A, Artavanis-Tsakonas S Nakayama KI. (2001). J. Biol. Chem., 276, 33111–33120. and Hariharan IK. (2004). Curr. Biol., 14, 965–974. Henriksson M, Bakardjiev A, Klein G and Luscher B. (1993). Moshe Y, Boulaire J, Pagano M and Hershko A. (2004). Proc. Oncogene, 8, 3199–3209. Natl. Acad. Sci. USA, 101, 7937–7942.

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2869 Nakayama K, Nagahama H, Minamishima YA, Spruck CH, Strohmaier H, Sangfelt O, Muller HM, Hubalek Miyake S, Ishida N, Hatakeyama S, Kitagawa M, M, Muller-Holzner E, Marth C, Widschwendter M and Iemura S, Natsume T and Nakayama KI. (2004). Dev. Cell, Reed SI. (2002). Cancer Res., 62, 4535–4539. 6, 661–672. Sreenivasan A and Kellogg D. (1999). Mol. Cell. Biol., 19, Nash P, Tang X, Orlicky S, Chen Q, Gertler FB, Mendenhall 7983–7994. MD, Sicheri F, Pawson T and Tyers M. (2001). Nature, 414, Stegmeier F, Visintin R and Amon A. (2002). Cell, 108, 514–521. 207–220. Ohta T, Michel JJ, Schottelius AJ and Xiong Y. (1999). Mol. Strohmaier H, Spruck CH, Kaiser P, Won KA, Sangfelt O and Cell, 3, 535–541. Reed SI. (2001). Nature, 413, 316–322. Orlicky S, Tang X, Willems A, Tyers M and Sicheri F. (2003). Tetzlaff MT, Yu W, Li M, Zhang P, Finegold M, Mahon K, Cell, 112, 243–256. Harper JW, Schwartz RJ and Elledge SJ. (2004). Proc. Natl. Pagano M, Tam SW, Theodoras AM, Beer-Romero P, Del Sal Acad. Sci. USA, 101, 3338–3345. G, Chau V, Yew PR, Draetta GF and Rolfe M. (1995). Tsunematsu R, Nakayama K, Oike Y, Nishiyama M, Ishida Science, 269, 682–685. N, Hatakeyama S, Bessho Y, Kageyama R, Suda T and Peters JM. (2002). Mol. Cell, 9, 931–943. Nakayama KI. (2004). J. Biol. Chem., 279, 9417–9423. Pintard L, Willis JH, Willems A, Johnson JL, Srayko M, Kurz Tyers M and Jorgensen P. (2000). Curr. Opin. Genet. Dev., 10, T, Glaser S, Mains PE, Tyers M, Bowerman B and Peter M. 54–64. (2003). Nature, 425, 311–316. Visintin R, Hwang ES and Amon A. (1999). Nature, 398, Prinz S, Hwang ES, Visintin R and Amon A. (1998). Curr. 818–823. Biol., 8, 750–760. von der Lehr N, Johansson S, Wu S, Bahram F, Castell A, Porter PL, Malone KE, Heagerty PJ, Alexander GM, Gatti Cetinkaya C, Hydbring P, Weidung I, Nakayama K, LA, Firpo EJ, Daling JR and Roberts JM. (1997). Nat. Nakayama KI, Soderberg O, Kerppola TK and Larsson Med., 3, 222–225. LG. (2003). Mol. Cell, 11, 1189–1200. Pulverer BJ, Fisher C, Vousden K, Littlewood T, Evan G and Wang W, Ungermannova D, Jin J, Harper JW and Liu X. Woodgett JR. (1994). Oncogene, 9, 59–70. (2004). Oncogene, 23, 1064–1075. Rajagopalan H, Jallepalli PV, Rago C, Velculescu VE, Kinzler Wasch R and Cross FR. (2002). Nature, 418, 556–562. KW, Vogelstein B and Lengauer C. (2004). Nature, 428, Watanabe N, Arai H, Nishihara Y, Taniguchi M, Watanabe 77–81. N, Hunter T and Osada H. (2004). Proc. Natl. Acad. Sci. Reed SI. (2002). Cell Cycle, 1, 389–390. USA, 101, 4419–4424. Reed SI. (2003). Nat. Rev. Mol. Cell. Biol., 4, 855–864. Wei W, Ayad NG, Wan Y, Zhang GJ, Kirschner MW and Reimann JD, Freed E, Hsu JY, Kramer ER, Peters JM and Kaelin Jr WG. (2004). Nature, 428, 194–198. Jackson PK. (2001). Cell, 105, 645–655. Welcker M, Orian A, Jin J, Grim JA, Harper JW, Eisenman Rodier G, Montagnoli A, Di Marcotullio L, Coulombe P, RN and Clurman BE. (2004a). Proc. Natl. Acad. Sci. USA, Draetta GF, Pagano M and Meloche S. (2001). EMBO J., 101, 9085–9090. 20, 6672–6682. Welcker M, Singer J, Loeb KR, Grim J, Bloecher A, Gurien- Sakchaisri K, Asano S, Yu LR, Shulewitz MJ, Park CJ, West M, Clurman B and Roberts JM. (2003). Mol. Cell, 12, Park JE, Cho YW, Veenstra TD, Thorner J and Lee KS. 381–392. (2004). Proc. Natl. Acad. Sci. USA, 101, 4124–4129. Welcker M, Orian A, Grim JA, Eisenmann RN and Clurman Sears R, Nuckolls F, Haura E, Taya Y, Tamai K and Nevins BE. (2004b). Curr. Biol., 14, 1852–1857. JR. (2000). Genes Dev., 14, 2501–2514. Winston JT, Chu C and Harper JW. (1999a). Genes Dev., 13, Seol JH, Feldman RM, Zachariae W, Shevchenko A, Correll 2751–2757. CC, Lyapina S, Chi Y, Galova M, Claypool J, Sandmeyer S, Winston JT, Koepp DM, Zhu C, Elledge SJ and Harper JW. Nasmyth K, Deshaies RJ, Shevchenko A and Deshaies RJ. (1999b). Curr. Biol., 9, 1180–1182. (1999). Genes Dev., 13, 1614–1626. Winston JT, Strack P, Beer-Romero P, Chu CY, Elledge SJ Shirayama M, Zachariae W, Ciosk R and Nasmyth K. (1998). and Harper JW. (1999c). Genes Dev., 13, 270–283. EMBO J., 17, 1336–1349. Wirbelauer C, Sutterluty H, Blondel M, Gstaiger M, Shirayama M, Toth A, Galova M and Nasmyth K. (1999). Peter M, Reymond F and Krek W. (2000). EMBO J., 19, Nature, 402, 203–207. 5362–5375. Shulewitz MJ, Inouye CJ and Thorner J. (1999). Mol. Cell. Won KA and Reed SI. (1996). EMBO J., 15, 4182–4193. Biol., 19, 7123–7137. Wu G, Xu G, Schulman BA, Jeffrey PD, Harper JW and Sia RA, Bardes ES and Lew DJ. (1998). EMBO J., 17, Pavletich NP. (2003). Mol. Cell, 11, 1445–1456. 6678–6688. Xu L, Wei Y, Reboul J, Vaglio P, Shin TH, Vidal M, Elledge Signoretti S, Di Marcotullio L, Richardson A, Ramaswamy S, SJ and Harper JW. (2003). Nature, 425, 316–321. Isaac B, Rue M, Monti F, Loda M and Pagano M. (2002). J. Yada M, Hatakeyama S, Kamura T, Nishiyama M, Tsune- Clin. Invest., 110, 633–641. matsu R, Imaki H, Ishida N, Okumura F, Nakayama K and Singer JD, Gurian-West M, Clurman B and Roberts JM. Nakayama KI. (2004). EMBO J., 23, 2116–2125. (1999). Genes Dev., 13, 2375–2387. Yang G, Ayala G, De Marzo A, Tian W, Frolov A, Wheeler Skowyra D, Craig KL, Tyers M, Elledge SJ and Harper JW. TM, Thompson TC and Harper JW. (2002). Clin. Cancer (1997). Cell, 91, 209–219. Res., 8, 3419–3426. Skowyra D, Koepp DM, Kamura T, Conrad MN, Conaway Yeh E, Cunningham M, Arnold H, Chasse D, Monteith T, RC, Conaway JW, Elledge SJ and Harper JW. (1999). Ivaldi G, Hahn WC, Stukenberg PT, Shenolikar S, Uchida Science, 284, 662–665. T, Counter CM, Nevins JR, Means AR and Sears R. (2004). Spencer CA and Groudine M. (1991). Adv. Cancer Res., 56, Nat. Cell Biol., 6, 308–318. 1–48. Ye X, Nalepa G, Welcker M, Kessler BM, Spooner E, Qin J, Spruck CH, Won KA and Reed SI. (1999). Nature, 401, Elledge SJ, Clurman BE and Harper JW. (2004). J. Biol. 297–300. Chem., 279, 50110–50119.

Oncogene SCF-mediated protein degradation and cell cycle control XL Ang and J Wade Harper 2870 Ye X, Wei Y, Nalepa G and Harper JW. (2003). Mol. Cell. Yoshida Y, Chiba T, Tokunaga F, Kawasaki H, Iwai K, Biol., 23, 8586–8600. Suzuki T, Ito Y, Matsuoka K, Yoshida M, Tanaka K and Yokoi S, Yasui K, Mori M, Iizasa T, Fujisawa T and Inazawa Tai T. (2002). Nature, 418, 438–442. J. (2004). Am. J. Pathol., 165, 175–180. Zheng N, Wang P, Jeffrey PD and Pavletich NP. (2000). Cell, Yokoi S, Yasui K, Saito-Ohara F, Koshikawa K, Iizasa T, 102, 533–539. Fujisawa T, Terasaki T, Horii A, Takahashi T, Hirohashi S Zhou P and Howley PM. (1998). Mol. Cell, 2, and Inazawa J. (2002). Am. J. Pathol., 161, 207–216. 571–580.

Oncogene