Molecular Cell, Vol. 13, 137–147, January 16, 2004, Copyright 2004 by Cell Press -Regulated Recognition of the Destruction Box of by the APC/C in Xenopus Egg Extracts

Hiroyuki Yamano,1 Julian Gannon, Hiro Mahbubani, ubiquitylates other during the cell cycle, such and Tim Hunt* as Securin/Cut2/Pds1, Nek2A, Xkid1, Cdc20, Cdc5, Cancer Research UK Ase1, and Geminin and targets them for destruction by Clare Hall Laboratories the 26S proteasome (Charles et al., 1998; Cohen-Fix et South Mimms al., 1996; Funabiki and Murray, 2000; Funabiki et al., Herts EN6 3LD 1996; Hames et al., 2001; Juang et al., 1997; McGarry United Kingdom and Kirschner, 1998; Shirayama et al., 1998). The APC/C thus regulates at least two key events in mitosis: sister chromatid separation and the inactivation of CDKs Summary (cyclin-dependent kinases) to exit from mitosis (Harper et al., 2002; Morgan, 1999; Peters, 2002; Zachariae and Substrates for mitotic proteolysis such as cyclin B Nasmyth, 1999). This ordered ubiquitylation/proteolysis have a 9 residue destruction motif, the destruction box is regulated by activation of the APC/C via phosphoryla- (D-box). To identify the receptor that specifically binds tion of its subunits (Golan et al., 2002; Kotani et al., the D-box, we used affinity chromatography with im- 1998; Lahav-Baratz et al., 1995; Patra and Dunphy, 1998; mobilized D-box matrices. We find that the APC/C Rudner and Murray, 2000; Yamada et al., 1997) and also from Xenopus egg extracts binds to the D-box of cyclin by the recruitment of activators, comprising two related B, whereas Fizzy (Cdc20) does not. Mutations in the WD40-repeat proteins, Fizzy (Cdc20) and Fizzy-related D-box abolished this interaction. We show that this (Cdh1/Hct1). Genetic experiments have demonstrated binding is regulated in the cell cycle, such that the that the Fizzy family proteins are essential for the APC/ APC/C from egg extracts in interphase does not bind C-dependent ubiquitylation, and that they somehow or- to the D-box matrix. Our results suggest that the chestrate temporal differences in substrate ubiquityla- APC/C forms a stable interaction with the D-box of its tion during the progression of mitosis and G1-phase substrates in a cell cycle-dependent manner. (Dawson et al., 1995; Fang et al., 1998; Kramer et al., 2000; Lim et al., 1998; Schwab et al., 1997; Shirayama Introduction et al., 1998; Sigrist et al., 1995; Sigrist and Lehner, 1997; Visintin et al., 1997; Zachariae et al., 1998). Nonetheless, The pathway is an ATP-dependent tagging sys- the roles of Fizzy family proteins for the APC/C-depen- dent proteolysis are not well defined. The favored cur- tem for degradation. It is an essential system in rent model for the role of Fizzy family proteins (Cdc20, eukaryotic cells, where it is used for degrading damaged Cdh1/Hct1) is to recognize the substrates (Burton and and misfolded proteins and also for degrading short- Solomon, 2001; Hilioti et al., 2001; Ohtoshi et al., 2000; lived regulatory proteins during processes such as the Pfleger et al., 2001; Schwab et al., 2001) and, presum- cell cycle, transcription, signal transduction, and devel- ably, deliver them to the APC/C, although they might opment (Hershko and Ciechanover, 1998; Jackson et have additional roles in the regulation of APC-dependent al., 2000; Peters, 2002; Zachariae and Nasmyth, 1999). proteolysis. On the other hand, two recently published The covalent attachment of ubiquitin chains to proteins reports claim that the APC/C, rather than Fizzy, is in- can be highly selective and delicately regulated. It is volved in substrate recognition (Meyn et al., 2002; Pass- achieved by the sequential action of three enzymes, E1, more et al., 2003). Thus, it has become less clear how E2, and E3; ubiquitin is first attached by a thioester APC/C substrates are recognized, and by what. Notably, bond to a ubiquitin-activating enzyme (E1) in an ATP- those APC/C substrates that are destroyed in anaphase, dependent manner, then transferred to a small and with no exception, have a destruction box (D-box) as somewhat specific ubiquitin-carrying protein (E2), and their destruction signal. Thus, it is important to find what finally the ubiquitin is conjugated onto the target sub- binds to the D-box and how the recognition is coordi- strates either by E2 alone or, more often, in conjunction nated with ubiquitylation. We may ask, for example, with a substrate-specific ubiquitin ligase (E3). Proteins whether it is the recognition or the ubiquitylation that is carrying long chains or trees of ubiquitin are then deliv- regulated? Here, we directly explore such questions in ered to the proteasome for degradation (Hershko and Xenopus egg extracts, where the picture is simplified Ciechanover, 1998; Hochstrasser, 1996; Varshavsky, by the absence of Fizzy-related (Lorca et al., 1998). We 1997). Since this proteolysis pathway is a precise and used affinity chromatography with immobilized D-box irreversible process, it allows a rapid switch to control matrices to search for the D-box “receptor.” In frog egg biological transitions. The anaphase-promoting com- extracts we find that very little Fizzy binds to the D-box, plex/cyclosome (APC/C) is a large E3 ubiquitin ligase whereas most of the APC/C is specifically retained by complex, which was originally identified as cyclin B ubi- a D-box matrix. We also present data suggesting that quitin ligase (Irniger et al., 1995; King et al., 1995; Su- purified APC/C can bind to the D-box matrix without dakin et al., 1995; Tugendreich et al., 1995), but it also Fizzy protein. Moreover, the interaction between the D-box and the APC/C appears to be regulated in the *Correspondence: [email protected] cell cycle. Although these experiments do not define 1Present address: Marie Curie Research Institute, The Chart, Oxted, the role of Fizzy, our data imply that the APC/C can bind Surrey RH8 0TL, United Kingdom. to the D-box without Fizzy protein, once it has been Molecular Cell 138

Figure 1. Affinity Chromatography with Im- mobilized D-Boxes (A) Schematic diagram of the D-box affinity columns; wild-type D-boxes are indicated by open and mutant D-boxes by filled rect- angles. (B) Flow-through of the various D-box col- umns were used for testing cyclin proteolysis in frog egg extracts. Full-length fission yeast cyclin B (Cdc13) and a version lacking 67 N-terminal residues (⌬67, stable control) were

used as substrates. CaCl2 was added to initi- ate proteolysis.

activated in mitosis, a process that somehow involves even in the presence of Ca2ϩ, whereas the flow-through Fizzy. from the column carrying the mutated tandem D-box efficiently destroyed Cdc13. The single N70 affinity resin Results was slightly less effective than N70-2X, although its FT showed impaired destruction for Cdc13 compared to Destruction Box Affinity Columns Deplete Essential the FT from the mutated D-box column (Dm) or GST Elements of Cyclin Proteolysis alone. These results indicate that the D-box affinity col- We constructed a variety of affinity columns to explore umns could deplete elements that are required for com- how destruction boxes (D-box) are recognized by the plete cyclin proteolysis. ubiquitylation machinery. Most experiments used the N-terminal 70 residues of fission yeast cyclin B (N70) D-Boxes Deplete APC/C Rather than Fizzy fused to GST, since we previously showed that this in Xenopus Egg Extracts N-terminal fragment allowed efficient APC/C-dependent Next, we investigated what factor(s) were depleted from ubiquitylation and proteolysis in Xenopus egg extracts egg extracts, rendering the extracts incompetent for (Yamano et al., 1996, 1998). Four constructs were made: cyclin destruction. Since we previously found that the N70 fused to GST, a tandem N70 construct (N70-2X), D-box receptor was immunoprecipitated by anti-Apc3 and the two equivalent constructs with the R and L of antibodies (Yamano et al., 1998), and there are several the D-box(es) mutated to A residues (Figure 1A). To test lines of evidence that the Fizzy family of proteins are whether these constructs worked as bait for the D-box involved in anaphase-promoting complex/cyclosome receptor, we used Xenopus egg extracts in which the (APC/C) substrate recognition (Burton and Solomon, APC/C-dependent cyclin destruction assay can be per- 2001; Hilioti et al., 2001; Ohtoshi et al., 2000; Pfleger et formed. Cytostatic factor (CSF) arrested Xenopus egg al., 2001; Schwab et al., 1997, 2001; Visintin et al., 1997), extracts are arrested at metaphase in meiosis II, and we examined which fractions contain the APC/C, Fizzy, addition of Ca2ϩ releases this arrest and triggers cyclin or Cdc2 (a negative control), after affinity chromatogra- proteolysis via APC/C activation (King et al., 1995; Masui phy. CSF-arrested metaphase egg extracts (input, I), and Markert, 1971; Murray et al., 1989; Tunquist and flow-through (F), and bound (B) fractions were immu- Maller, 2003; Watanabe et al., 1991; Yamano et al., 1998). noblotted with anti-Apc3, anti-Fizzy, or anti-Cdc2 anti- If essential elements such as a D-box receptor are de- bodies (Figure 2A). The flow-through from the tandem pleted by these D-box affinity columns, one would ex- N70 column contained only about 5% of the input pect that the flow-through (FT) would fail to destroy APC/C, indicating that the N70 column was able to de- cyclin even in the presence of Ca2ϩ. To assay D-box plete up to 95% of APC/C from egg extracts (lanes or D-box-independent cyclin proteolysis, we used two 1and 2). The APC/C retained by the D-box matrix was radiolabeled substrates; a full-length fission yeast cyclin successfully eluted by GSH (lane 3), whereas the same B (Cdc13) and an N-terminal 67 residue truncated ver- tandem construct with the mutated D-box (Dm-2X) sion of Cdc13 (⌬67) that lacks a D-box and serves as a hardly depleted the APC/C at all, and the flow-through control. As shown in Figure 1B, the flow-through from (F) contained as much APC/C as the input (lanes 4–6). the tandem N70 column was unable to destroy Cdc13 Consistent with this result, the single N70-GST depleted Destruction Box Receptor of Cyclin B 139

Figure 2. The APC/C, but Not Fizzy, Is Retained by a D-Box Affinity Matrix (A) CSF-arrested frog egg extracts (input, I) were applied to the indicated affinity matrices, and separated into flow-through (F) and bound (B) fractions followed by immunoblotting with anti-Apc3, anti-Fizzy, or anti-PSTAIRE antibodies. (B) Same as (A) using N70-2X column, but CSF or APC/C depleted CSF extracts were used. (C) Titration of N70. Aliquots of biotinylated N70: 30, 10, 5, 1, 0.5, 0.25, or 0 ␮g, were attached to Streptavidin-magnetic beads. Next, 20 ␮l of CSF extract were applied onto the beads. After incubation at 23ЊC for 20 min, the beads were isolated by a magnetic stand. The supernatants were analyzed by immunoblotting with anti-Apc3, anti-Fizzy, anti-PSTAIRE, or anti-UbcP4 (Ubcx homolog in fission yeast) antibodies.

(D) Cyclin destruction assay using the supernatants from (C). CaCl2 was added to trigger cyclin proteolysis. (E) Purified APC/C with or without additional ubiquitin was added to the supernatant of the N70-2X column showed in (A), followed by cyclin destruction assays. (F) Same as (A), but interphase extracts were used as input. more than 90% of the APC/C (lanes 7 and 8), whereas a small fraction of Fizzy was bound to the APC/C and the Dm or plain GST columns did not appreciably de- not because it interacted directly with the D-box. plete the APC/C (lanes 10–15). Note that the single N70 In a slightly different approach, we synthesized N70 column only retained a trace amount of the APC/C in the peptides with an N-terminal biotin residue and attached bound fraction, presumably due to a weaker interaction, them to streptavidin-coated magnetic beads to investi- and the APC/C leaching off during the washes (lane 9). gate the relationship between depletion of the APC/C The same membranes were also probed with an anti- (Apc3) and the inability to destroy cyclin in egg extracts. Fizzy antibody, which revealed that most of Fizzy was Supernatants of egg extracts were examined after addi- in the flow-through (FT) in all columns tested, suggesting tion of the biotinylated N70 peptides and harvesting that the Fizzy in egg extracts does not directly (or only by magnetic beads. Immunoblots of the supernatants transiently) interact with the D-box of cyclin B. However, showed that the APC/C was depleted by N70 peptides in we reproducibly detected a small fraction, typically an N70 dose-dependent manner (Figure 2C). In contrast, about 2%–3% of the total Fizzy, in the bound fraction Fizzy, Cdc2, or Ubcx (the relevant E2 ubiquitin-conjugat- (lane 3), so we wished to test whether the Fizzy bound ing enzyme) were not depleted by N70 peptide. Cyclin directly to the cyclin N terminus or indirectly via the destruction in the supernatants was well correlated with APC/C. To test this point, we first immunodepleted the the residual amount of the APC/C in the egg extracts APC/C from egg extracts, then applied them to the N70 (Figure 2D). We tested if adding back purified APC/C column. Using magnetic beads that were covalently into the depleted supernatants (Figure 2E) could rescue coupled to a monoclonal antibody of the Apc3 subunit the destruction of cyclin B, and found that addition of (AF3.1), we could show that more than 95% of endoge- purified APC/C (see below) or APC/C together with ubi- nous APC/C could be removed (data not shown). Figure quitin rescued cyclin B destruction (lanes 5–12). These 2B shows that once the APC/C had been depleted from data all support the view that the APC/C specifically egg extracts, Fizzy was no longer in the bound fraction interacts with N70. (compare lanes 3 and 6). Thus, we conclude that the We also used affinity chromatography with the GST Fizzy in the bound fraction (lane3) was retained because fusion proteins to examine interphase egg extracts. Fig- Molecular Cell 140

Figure 3. Interaction between the D-Box and the APC/C in Egg Extracts (A) Large volumes (2 ml) of anaphase extracts (⌬90-arrested egg extracts) were applied to 100 ␮l N70-GST or Dm-GST columns, and the appearance of proteins in successive drops ␮l) of the flow-through was monitored 30ف) by immunoblotting with anti-Apc3, anti-Fizzy, and anti-PSTAIRE antibodies. Lane E, bound proteins eluted by reduced glutathione. (B) Quantitation of (A). For each protein, inten- sities are plotted relative to the level after saturation of the column. (C) Sucrose gradient centrifugation of ana- phase egg extracts. Anaphase HSS was over- laid on a 15%–40% sucrose density gradient and spun at 40,000 rpm for 12 hr. Fractions were analyzed by immunoblotting with anti- Apc3, anti-Fizzy, and anti-PSTAIRE anti- bodies.

ure 2F shows that neither APC/C nor Fizzy were retained extracts: APC/C was 365 nM and Fizzy was 348 nM in by the D-box affinity matrix when interphase egg extract the extract (see Experimental Procedures). Previously was loaded on the columns, suggesting that the interac- the concentration of Fizzy was estimated as only 50–100 tion between the D-box and the APC/C is cell cycle regu- nM (Lorca et al., 1998). lated. Next, we investigated the size distribution of the APC/C and Fizzy in egg extracts by sucrose density The Role of Fizzy in the Interaction between gradient separation. Figure 3C shows that more than the D-Box and the APC/C in Egg Extracts 90% of the APC/C in anaphase egg extracts is in the To further explore the interactions between the D-box 20S fractions, whereas only a small fraction (less than and the APC/C in egg extracts, we applied large volumes 10%) of Fizzy is in the same 20S fraction and most of (relative to the column volume) of anaphase extracts the Fizzy is in between 4.5S and 11.3S. These results (⌬90-arrested egg extracts) to N70-GST or control col- were the same in interphase egg extracts (data not umns and analyzed the appearance of proteins in the shown). We conclude that 90%–95% of the APC/C in flow-through (FT). The expectation was that if protein(s) egg extracts is not tightly bound to Fizzy. Equally clearly, specifically interacted with the D-box, the appearance however, a small but probably significant fraction— of these proteins in the flow-through would be delayed 5%–10%—of the APC/C is associated with Fizzy. The compared to the mutant D-box column. Eventually, the significance of this association is unclear, but consider- affinity matrix will saturate, and the ligand will appear ing that far more APC/C binds to the destruction box in the effluent. Fractions of the effluent were immu- column than does Fizzy, it seems that while Fizzy is noblotted by anti-Apc3, anti-Fizzy, or anti-Cdc2. Both required to activate the APC/C, its continuing presence Fizzy and Cdc2 appeared in the flow-through at the is not required for the maintenance of a stable complex same time regardless of whether N70 or Dm-GST col- between the N terminus of cyclin B and the APC/C. umns were used (Figures 3A and 3B). In contrast, ap- pearance of the APC/C (Apc3) was retarded in the flow- Fizzy-free APC/C Can Bind the D-Box of Cyclin B through of the N70-GST column but not in the control We next investigated whether Fizzy-free APC/C could column. These data support the previous finding that interact with the D-box of cyclin B. To do that, the APC/C the APC/C, but not Fizzy, interacts stably with the D-box was purified from CSF-arrested egg extracts (see Exper- in egg extracts. We eliminated an alternative possibility, imental Procedures), and the Fizzy-bound APC/C sub- that Fizzy is more abundant than the APC/C by measur- sequently removed by anti-Fizzy antibodies (Figure 4A, of the (%5ف) ing the concentrations of the APC/C and Fizzy in egg lanes 1 and 2). Since only a small portion Destruction Box Receptor of Cyclin B 141

Figure 4. Fizzy-free APC/C Binds to the D-box, but Not Mutant D-Box Affinity Matrix (A) Purified APC/C or “Fizzy-free” APC/C from CSF extracts was analyzed by immunoblot- ting with anti-Apc3 or anti-Fizzy antibodies. (B) Top, Fizzy-free APC (input, I) was applied onto biotinylated N70 or Dm columns, and after incubation at 23ЊC for 20 min, flow- through (F) and bound (B) fractions were sep- arated. The fractions analyzed by immu- noblotting with anti-Apc3 antibody. Bottom, quantitation of the immunoblots with relative value against input (I). (C) Same as (B), but using securin-GST or securin-D-box/KEN-box double mutant-GST matrices.

APC/C was bound to Fizzy in egg extracts, identical an asterisk. Fizzy could not be detected as a stained subunits of the APC/C were detected by silver staining band, but a small amount of Fizzy could be detected the APC/C and the Fizzy-free APC/C (data not shown). by immunoblotting the gel (Figure 5B). Significantly, ana- The Fizzy-free APC/C was applied to N70 or Dm beads, phase APC/C contained about 2.5 times more Fizzy than and separated into flow-through (F) and bound (B) frac- interphase APC/C. tions. As shown in Figure 4B, about 80% of the APC/C To study the interaction between the D-box and the bound to the D-box matrix (N70) whereas the D-box APC/C, we set up surface plasmon resonance (SPR) mutant column (Dm) hardly retained any APC/C. Thus, using N-terminal biotinylated peptides (N70 or the equiv- we conclude that the APC/C, but not Fizzy, recognizes alent peptide with D-box mutations) attached to the the D-box of cyclin B. surface of streptavidin-coated sensor chips. Purified Up to now, all experiments were conducted with the APC/C was applied as analyte. Figure 5C shows that N-terminal fragment of cyclin B. To check our conclu- the anaphase APC/C specifically interacted with N70 sions with another D-box-containing substrate of the (D-box), but not the D-box mutant peptide. The dissocia- APC/C, we examined whether securin also bound the tion rate was calculated as 0.0123, corresponding to a APC/C. Constructs comprising full-length securin fused half-life of interaction between anaphase APC/C and to GST or an equivalent chimera lacking both KEN box the D-box of approximately 1 min. Both N70 and Dm and D-box (Hagting et al., 2002) were purified from bac- peptide were overloaded onto the sensor chip, and the teria and used to make affinity columns. Fizzy-free dissociation rate was calculated from the initial dissocia- APC/C was applied to wild-type securin or the mutant tion curve. In contrast, interphase APC/C did not signifi- securin beads, and separated into flow-through (F) and cantly interact with either N70 or D-box mutant peptides bound (B) fractions. As shown in Figure 4C, the flow- (Figure 5C, right panel). through from the securin column contained less than We used N70 or Dm affinity columns to compare the 5% of the input APC/C and about 40% of the APC/C binding affinities of anaphase and interphase APC/C. could be recovered from the matrix (lanes 1–3). In con- The APC/C from metaphase (i.e., CSF-arrested extracts) trast, about 80% of the input APC/C failed to bind to and anaphase extracts bound equally well to the N70 the mutant securin column, and only 13% of the APC/C matrix, and not to the D-box mutant (Dm) matrix (lanes was retained by the “mutant” matrix (lanes 4 and 5). 1–6). In contrast, the APC/C from interphase extracts These data indicate that Fizzy-free APC/C could interact was not retained by the D-box matrix (lanes 7–9). These with the D-box of securin, suggesting that the APC/C data confirm that interactions between the D-box and must contain a universal D-box receptor. the APC/C are regulated in the cell cycle, and are only activated in mitosis. Interphase and Mitotic APC/C Differ in Their Affinity for the D-Box Comparing the APC/C from CSF-Arrested Extracts Finally, we compared the abilities of purified APC/C from with Anaphase or Interphase APC/C in Terms mitotic extracts with the APC/C from interphase extracts of their Ability to Initiate Cyclin Proteolysis in terms of their ability to bind to the D-box. We used We also compared the abilities of the APC/C from CSF- immunoaffinity chromatography to purify APC/C from arrested extracts with the APC/C from anaphase and anaphase (⌬90 arrested) or interphase extracts (see Ex- interphase extracts in terms of their ability to restore perimental Procedures). The anaphase APC/C showed cyclin B destruction in an APC/C-depleted extract. We active ubiquitylating activity, whereas the interphase used AF3.1 coupled beads to deplete the APC/C from APC/C was inactive (data not shown). Figure 5A com- the CSF extracts, and added back various forms of pares the protein content of the two preparations. All APC/C that had been purified from egg extracts in differ- the known APC/C subunits were clearly seen on the ent cell cycle stages. As showed in Figure 6A, we could gel, indicated by their numbers. Subunits 1, 3, 6, and specifically deplete the APC/C (Ͼ95%) from the CSF 8 showed mobility shifts, presumably due to altered extracts, but the levels of Fizzy, Cdc2, and Ubcx were phosphorylation, and the interphase preparation con- essentially unchanged. Then, using the APC/C-depleted tained three prominent unidentified bands, marked with extracts, we performed cyclin destruction assay (Figure Molecular Cell 142

Figure 5. Interaction with the D-Box and the APC/C Is Cell Cycle Regulated (A) Purification of the APC/C from anaphase or interphase extracts. Using AF3.1-immunoaffinity chromatography and specific elution peptides, the soluble APC/C was purified, and visualized by silver staining. The numbers in figure refer to subunit numbers of the APC/C. The bands marked with asterisks are unknown. (B) Parallel samples from (A) were analyzed by immunoblotting with anti-Apc3 and anti-Fizzy antibodies. (C) APC/C purified from either anaphase or interphase extracts was applied onto streptavidin-coated sensor chips of a Biacore apparatus, loaded with biotinylated N70 or Dm in surface plasmon resonance (SPR). The APC/C was injected at time 0, and injection continued for 100 s. (D) Metaphase (CSF-arrested; Meta), anaphase (⌬90 arrested; Ana), or interphase (Int) extracts were subjected to biotinylated N70 or Dm affinity chromatography. After incubation at 23ЊC for 20 min, the bound fractions were isolated by a magnetic stand, washed several times, and analyzed by immunoblotting with anti-Apc3, anti-Fizzy, anti-PSTAIRE, or anti-UbcP4 (Ubcx homolog in fission yeast) antibodies.

6B). Without further addition, the APC/C-depleted ex- the APC/C purified from CSF extracts could rescue tracts were unable to destroy Cdc13 even in the pres- cyclin destruction (ϩAPCCSF; lanes 9–12) in a calcium- ence of Ca2ϩ (ϩmock; lanes 1–4), whereas adding back dependent manner. The APC/C purified from anaphase

Figure 6. Qualitative Difference of the APC/C Isolated from Different Cell Cycle Stages in Egg Extracts (A) The APC/C was depleted from CSF extracts using AF3.1 coupled beads. The APC/C depletion process was repeated; first depletion for single depletion (lane 2) and second depletion for sequential double depletion (lane 3). Untreated extract (lane 1) and both depleted extracts were analyzed by immunoblotting with anti-Apc3, anti-Fizzy, anti-PSTAIRE, or anti-UbcP4 (Ubcx homolog in fission yeast) antibodies.

(B) The APC/C-depleted CSF extracts were assayed for cyclin destruction with or without added CaCl2 as indicated and the following additions: lanes 1–8, buffer; lanes 9–16, APC/C purified from CSF extracts; lanes 17–24, APC/C from anaphase extract; and lanes 25–32, APC/C from interphase extract. Destruction Box Receptor of Cyclin B 143

extracts (APCAna) could rescue cyclin destruction, but To strengthen these conclusions, we performed paral- surprisingly, could initiate cyclin destruction even in the lel studies using securin-GST affinity columns with simi- absence of Ca2ϩ (lanes 17–24), although there was a lar results, although we detected a small amount of short lag compared to the reactions performed in the interaction between securin and the APC/C that did not presence of Ca2ϩ. This suggests that the APCAna is quali- appear to depend on the D-box. This may indicate that tatively different from the APCCSF. The APC/C purified there are additional sites for interaction between these from interphase extracts (APCInt) was also able to rescue proteins, and the same may also be true for cyclins and cyclin destruction although it displayed a short lag other D-box-containing substrates; we would not claim phase before initiating cyclin destruction (compared and do not imagine that the D-box provides the only lanes 25–28 with lanes 9–12). This lag could be explained interaction site between substrates and the APC/C (see by the activation of APCInt in the egg extracts, probably Carroll and Morgan, 2002). by phosphorylation of its subunits (Figure 5A), which is Previous studies of how Fizzy family members recog- thought to be necessary for the APC/C to be fully acti- nize substrates are somewhat confusing. For example, vated (King et al., 1995; Lahav-Baratz et al., 1995). the simple question of which parts of the Fizzy family protein actually recognize which domain or structure of Discussion substrates is debated. Schwab et al. (2001) reported that WD-repeat domain of Hct1/Fizzy-related was es- These results indicate that the APC/C, but not Fizzy sential to bind to Clb2/cyclin B, but the D-box of Clb2/ (Cdc20) bind stably to the destruction box (D-box) of cyclin B was not essential for this interaction. Since cyclin B in Xenopus egg extracts. They are inconsistent Clb2 was shown to contain a functional KEN-box, the with the idea that substrates are bound to the APC/C interaction might be through the KEN-box. The D-box in stoichiometric complexes with Fizzy. We selected the mutant of Clb2 is stable in vivo, however, indicating that N-terminal 70 residues of fission yeast cyclin B (N70) interaction with Hct1 is not sufficient for Clb2 proteolysis for these studies, based on previous work showing that and the D-box might be recognized by something else. this was a sufficient fragment to confer efficient APC/ Using Pds1 as bait in a yeast two-hybrid screen, Hilioti C-dependent ubiquitylation and proteolysis in ana- et al. (2001) isolated C-terminal fragments of Cdc20/ phase, and that these properties required an intact de- Fizzy containing the WD-repeat domain. In contrast, struction box (Yamano et al., 1996, 1998). Indeed, the however, Pfleger et al. (2001) reported that N-terminal fragments of Cdc20 and Cdh1, instead of the WD-repeat N70-GST fusion protein is efficiently degraded in ana- domain, were important for substrate specificity al- phase extracts. Other investigators have reported diffi- though the interactions were weak, since only up to 10% culty using full-length cyclins as substrates for the of the input Xkid1 or about 2% of cyclin bound to the APC/C, because of problems with protein aggregation Cdc20/Cdh1 affinity beads. They mention that binding (Pfleger et al., 2001). In addition, we used crude frog of the WD-repeats to substrates is observed, but that egg extracts for many of our experiments, where APC/C, it was not a D-box-specific interaction. Finally, Burton Fizzy, and other components are present at physiologi- and Solomon (2001) studied the interactions between cal concentration. Although the idea and execution of Cdc20/Fizzy or Cdh1/Fizzy-related and the protein Hsl1. these experiments is simple, as far as we know, no report Binding could be dependent on both the D-box and has been previously published that took this approach to KEN-box, and they found that the D-box is apparently investigate D-box-interacting proteins under physiologi- more important than KEN-box even in the interaction of cal conditions. Cdh1 with Hsl1. In summary, although previous investi- Using this approach, we find that the D-box of cyclin gators have provided evidence for D-box/KEN-box- B forms relatively stable interactions with the APC/C, dependent interactions between Fizzy or Fizzy-related and not with Fizzy. Moreover, only active forms of the and APC/C substrates, it is not clear which region of APC/C bind the D-box, for essentially no retention of the Fizzy family proteins interacts with the substrates. the APC/C occurred when interphase frog egg extracts Moreover, the role of the D-box of substrates is poorly were used. Yet it is clear that Fizzy family proteins (Fizzy/ understood in terms of whether or not Fizzy directly Cdc20 or Fizzy-related/Cdh1) are essential for activation interacts with it. of the APC/C-dependent proteolysis (Fang et al., 1998; We have also observed that recombinant Fizzy or Kramer et al., 2000; Lim et al., 1998; Schwab et al., 1997; Fizzy-related proteins purified from insect cells could Shirayama et al., 1998; Sigrist and Lehner, 1997; Visintin bind to our N70 constructs, but significantly, such inter- et al., 1997; Zachariae et al., 1998). Several previous actions did not require an intact D-box. Moreover, addi- reports have suggested specific interactions between tion of only 20% by volume of crude frog egg extracts substrates for degradation by this system and members into the binding reaction largely eliminated the interac- of the Fizzy family (Burton and Solomon, 2001; Hilioti et tion between the Fizzy and the D-box, and the APC/C al., 2001; Ohtoshi et al., 2000; Pfleger et al., 2001; from the egg extracts was now bound to the D-box (H.Y. Schwab et al., 2001). Our failure to find significant reten- and T.H., unpublished data). It is worth recalling that tion of Fizzy by the D-box affinity columns is most easily when we investigated the minimum size of peptides as explained by a model in which Fizzy proteins activate competitors in the simple destruction assay, we found the APC/C to recognize and bind the D-box. We were that removing the surroundings of the D-box strongly even able to detect strong and specific binding of Fizzy- reduced the potency of the peptides (Yamano et al., free APC/C to D-box-containing peptides. The nature 1998). Almost certainly, some of the binding energy for of this activation process is unclear, and our experi- these interactions comes from flanking regions of ments do not address this problem. these substrates. Molecular Cell 144

Figure 7. Model for Activation of the D-Box Receptor and the APC/C-Dependent Prote- olysis The APC/C purified from interphase, meta- phase (CSF-arrested frog egg extract), or anaphase has different activity in terms of recognition and proteolysis of cyclin B. In in- terphase, cyclin B is neither recognized or ubiquitylated. During mitosis, mitotic kinase(s) phosphorylate the APC/C, priming it for acti- vation by Fizzy and turning on recognition of the D-box. This form of the APC/C is not fully active, however, owing to inhibitory signals emanating from the spindle assembly check- point (perhaps acting to sequester Fizzy), and cyclin B is consequently stable. In anaphase, the APC/C is fully activated by as yet unknown mechanisms (here indicated by a further conformational change), and cyclin B is rapidly ubiquitylated, leading to its proteolysis. Note that Fizzy is not a permanent member of the complex, and that tight interactions with E2 ubiquitin- conjugating enzymes have never been observed. The phosphorylation sites are indicated in purely cartoon form; there is no direct evidence that polo kinase must phosphorylate the APC/C to activate it fully, although depletion of polo precludes such activation.

Our data do not exclude a model in which recruitment tween APC/C, Fizzy-related, and substrate. It seems of substrates to the APC/C occur through Fizzy. For rather unlikely, however, that Fizzy and Fizzy-related example, Fizzy might be involved in a kind of preselec- should act in completely different ways. Our data con- tion, or ushering, process for the substrates, roughly cerning Fizzy disfavor stoichiometric models. selecting and presenting them to the APC/C, which then performs a final check for the presence of D-boxes or Cell Cycle Dependency of the KEN-boxes. But our experiments with purified APC/C APC/C-Dependent Proteolysis free of Fizzy are more consistent with the idea that con- Our data indicate that, comparing interphase and mitotic formational changes occur in subunit(s) of the APC/C APC/C, it is the recognition of the D-box by the APC/C (D-box receptor) (in conjunction with phosphorylation that is cell cycle regulated, rather than its ability to ubiq- of subunit[s] of the APC/C), and these changes activate uitylate already-bound substrates. We have repeatedly the D-box receptor, leading to the formation of a “D-box tried, and failed, to detect any stable interaction with pocket.” Then, the activated D-box receptor specifically E2 ubiquitin conjugating enzymes to the APC/C. But it recognizes the D-box of substrates, and binds them may be that several mechanisms to ensure that the with a long enough half-life to allow the processive addi- APC/C substrates are ubiquitylated and degraded at the tion of a poly-ubiquitin chain (Figure 7). Possibly, in the right time and in the right place. For example, what is case of Fizzy-related, it can activate the (nonmitotically the difference between the APC/C’s activity in the CSF- phosphorylated) APC/C to have a broader range of sub- arrested frog egg extract (perhaps equivalent to a mi- strate specificity including both the D-box and the totic checkpoint arrested cell) and its full activity after KEN-box. Ca2ϩ addition? According to affinity chromatography, Recently, Passmore et al. (2003) reported that Doc1/ both forms of the APC/C bind equally well to the destruc- Apc10 contributes to APC/C-substrate recognition. tion box, yet cyclin B is stable in CSF extracts. That They showed that although APC⌬doc1 binds to Fizzy nor- would seem to suggest that the ability of the APC/C mally, it cannot interact with the APC/C substrates such to bind ubiquitin-loaded E2, or its ability to elongate as Clb2 and Hsl1. Adding back purified Doc1/Apc10 to ubiquitin chains is also regulated. Yet we know that the APC⌬doc1 restores the APC/C-substrate interactions and APC/C is capable of degrading cyclin A in a D-box- E3 ligase activity. Thus it is possible that subunit(s) in- dependent manner in CSF extracts. One of us (T.H.) cluding Doc1/Apc10 form a substrate recognition pocket prefers to interpret the difference in activity as a differ- that could be only activated after Fizzy interacts with ence in the half-life of the D-box:APC/C complexes—a the APC/C in mitosis. This is not the only interpretation of difference that would presumably be reflected in the these data, however. For example, one could envisage a processivity of the E3 ligase. H.Y. prefers the idea that model in which Doc1 is essential to the proper confor- the intrinsic ubiquitin ligase activity of the APC/C is regu- mation of the APC/C, and the role of Fizzy is to guide lated, as well as its D-box affinity. It is not really possible mitotic protein kinases to phosphorylate a certain con- to decide between these two extremes on the basis of stellation of sites on the APC/C in order to activate it the available evidence (see Figure 7). fully. Other possible models include the idea that both The D-box was identified as a degradation signal in the APC/C and Fizzy/Fizzy-related bind (different parts substrates more than 10 years ago (Glotzer et al., 1991), of) substrates, thereby increasing the strength of the and it is surprising, given its importance, how little is APC/C-substrate interactions. Thus, Carroll and Morgan known about how it is recognized. This study provides (2002) show Fizzy-related as binding the substrate to evidence that the APC/C directly binds the N terminus the APC/C while suggesting that Doc1 acts as a “pro- of cyclin B in a D-box-dependent manner, although we cessivity” factor that strongly enhances the affinity of have not identified the subunit(s) of the APC/C that form the APC/C for its cyclin substrate. Their model strongly the crucial receptor. And we have not investigated how implies the formation of stoichiometric complexes be- much of the surroundings of the D-box are required for Destruction Box Receptor of Cyclin B 145

retention by affinity matrices. The identification of the MAb AF3.1 (Yamano et al. 1998) was used to purify the APC/C from actual D-box receptor will be obvious targets for future egg extracts. projects. Understanding the role of the Fizzy family of proteins is similarly intriguing; in our view it is very poorly Quantitation of APC/C, Fizzy, and Cdc2 in Xenopus Egg Extracts understood, although crucial to the overall process, and To measure the concentration of APC/C in frog egg extracts, the central to its regulation. C-terminal 76 residues of Apc3 were isolated from a Xenopus cDNA library and expressed as GST-Apc3-C76 in bacteria. The GST-Apc3- Experimental Procedures C76 was purified and its concentration measured by Coomassie blue staining using an albumin standard curve. This allowed us Expression and Purification of Recombinant Proteins to determine the concentration of Apc3 in frog egg extracts by The N-terminal 70 residues of fission yeast cyclin B, referred to as quantitative immunoblotting with MAb AF3.1, which recognizes the N70 or N70 with the mutated D-box (Dm); created by changing the C-terminal 10 residues. To quantify Fizzy and Cdc2, we translated wild-type D-box, RHALDDVSN to AHAADDVSN were used to make Fizzy, Cdc2, and cyclin A mRNAs in a coupled in vitro transcription- D-box affinity columns (Yamano et al., 1996). Tandem N70 (N70- translation system (TNT, Promega) in the presence of [35S]methio- 2X), N70, tandem Dm (Dm-2X), and Dm were subcloned into nine. This allowed us to measure the specific activity of [35S]methio- pET16bp-GST. These four proteins and GST alone (without insert) nine by reference to a cyclin A standard and immunoblotting with were expressed in bacteria, purified on GSH-Sepharose (Amer- MAb E23. Knowing the number of methionine residues in Fizzy and sham-Biotech), and eluted with reduced glutathione (20 mM) fol- Cdc2 allowed the concentration of Fizzy and Cdc2 to be calculated. lowed by dialysis against PBS at 4ЊC. Human securin and securin The translation mix then served as a standard for the unknown egg D-box/KEN-box double mutant constructs were subcloned into extracts We estimated the concentration of the APC/C, Fizzy, and pET16bp-GST and purified as described above. Both N70-GST and Cdc2 as 365, 348, and 470 nM, respectively. securin-GST were rapidly and completely degraded in frog egg ex- ϩ tracts induced to enter anaphase by added Ca2 , whereas the mu- Purification of the APC/C from Egg Extracts tant forms were stable (data not shown). The APC/C was purified from untreated extracts or high-speed su- pernatants (HSS) of CSF, anaphase, or interphase egg extracts using D-Box Affinity Chromatography AF3.1-coupled beads for immunoaffinity chromatography. HSS Purified GST fusion proteins (50 ␮g) were bound to 20 ␮l of GSH- were prepared by centrifugation (100,000 ϫ g for 1 hr) after 2-fold Sepharose beads, and incubated with 20 ␮l of egg extracts at 23ЊC dilution of egg extracts with XB buffer. We checked that all of the for 20 min, then centrifuged to separate flow-through (FT) and bound APC/C remained in the supernatant after this centrifugation. AF3.1- (B). 10 ␮l of FT was used for the cyclin destruction assay, and the coupled beads were prepared by mixing an appropriate amount remaining was separated by SDS-PAGE and analyzed by immu- of AF3.1 producing hybridoma cell supernatants with Dynabead- noblotting. Bound fractions were washed five times with binding Protein A beads (Dynal), then washed three times with PBS and buffer A (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5% NP40, 50 mM coupled with dimethyl pimelimidate (Sigma). For Figure 5, 10 ml of NaF, 60 mM ␤-glycerophosphate, 5 mM EDTA, 0.1 mM orthovana- HSS of anaphase or interphase egg extracts were incubated with date). The bound proteins were eluted with 20 mM GSH, 20 mM AF3.1-coupled Dynabeads-Protein A (20 ml AF3.1 supernatant plus Tris-Cl, pH 8.1. To make affinity columns with biotinylated D-box 400 ␮l of Dynabead suspension) at 23ЊC for 30 min, then washed peptides, N-terminal biotinylated N70 or Dm was chemically synthe- twice with a binding buffer A (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, sized and the indicated amounts (30, 10, 5, 1, 0.5, 0.25 ␮g) of pep- 0.5% NP40, 50 mM NaF, 60 mM ␤-glycerophosphate, 5 mM EDTA, tides bound to 50 ␮l of Dynabeads M-280 streptavidin (Dynal Bio- 0.1 mM orthovanadate), twice with 0.4 M NaCl in XB and twice with tech, UK). The beads were incubated with excess of biotin to occupy the binding buffer A again. Finally, the APC/C was eluted from the all of the active streptavidin sites on the beads. Incubation with egg beads by incubating with 100 ␮l of 3 mg/ml of the triple AF3.1 extracts was basically the same as above; a magnetic stand (Dynal) epitope peptides for 5 min. More than 80% of the APC/C was eluted was used to separate “flow-through” and bound fractions. Bound by this method. The purified anaphase APC/C, but not interphase fractions were washed five times with the binding buffer A and boiled APC/C, showed active ubiquitylating activity for cyclin B. Subunits in the presence of SDS-PAGE sample buffer to elute bound proteins. of the APC/C (Figure 5A) were identified by either immunoblotting with antibodies or by MALDI/TOF peptide mass finger printing. To Xenopus Egg Extracts and Cyclin Destruction Assay prepare Fizzy-free APC/C (Figure 4), the APC/C was first purified CSF extracts were prepared from unfertilized Xenopus eggs as de- from egg extracts using only 500 ␮l of CSF extract and 5 ml of AF3.1 scribed by Murray et al. (1989). Interphase extracts were prepared supernatant coupled to Dynabeads-protein A. Fizzy-bound APC/C by adding CaCl2 (0.4 mM final) whereas anaphase extracts were was removed by anti-Xenopus Fizzy-coupled Dynabeads-protein A prepared by adding nondegradable sea urchin cyclin B (⌬90) and immunoprecipitation. The supernatant, which contained Fizzy-free

CaCl2. Cyclin destruction assays were performed in CSF-arrested APC/C, was subject to affinity chromatography using the appro- extracts by adding CaCl2 to initiate anaphase. To prepare substrates, priate affinity matrix: streptavidin-coated Dynabeads loaded with fission yeast cyclin B (Cdc13) and the N-terminal 67 residues trun- biotinylated N70 or Dm-N70, or GSH-Sepharose loaded with sec- cated version of Cdc13 (⌬67) were labeled with [35S]methionine plus urin-GST or D-box/KEN box mutant securin-GST. cysteine (Promix; Amersham Biosciences, UK) in a coupled in vitro transcription-translation system (TNT, Promega). Basically, 1 ␮lof APC/C Depletion from Egg Extracts the substrates was mixed with 9 ␮l of CSF extracts; samples were We found that 0.5 ml of AF3.1 supernatant coupled to Dynabeads- taken at the indicated intervals after adding CaCl2, and analyzed by protein A beads could deplete more than 95% of the APC/C from SDS-PAGE. 100 ␮l of egg extracts by mixing and incubating at 23ЊC for 20 min and using a magnetic stand (Dynal) to remove the beads. This Antibodies process was repeated to ensure more than 95% depletion, which Rabbit polyclonal antibodies were raised against the synthetic pep- was confirmed by immunoblotting with anti-Apc3 (Transduction lab). tide CFEVDPVTKKEKEKARSSKSIIHQSIR, corresponding to the C terminus of Xenopus Fizzy, after conjugating the peptides to keyhole Sucrose Gradient Centrifugation limpet hemocyanin (KLH). Polyclonal antibodies were also raised 200 ␮l of anaphase or interphase HSS was overlaid on a 5 ml sucrose against recombinant fission yeast UbcP4 (Xenopus Ubcx homolog), gradient (15%–40%) made in XB buffer, which was then spun at Xenopus Ubcx, and a Xenopus Fizzy C-terminal fragment. Anti- 40,000 rpm for 12 hr at 4ЊC in SW55 rotor (Beckman). 27 fractions UbcP4 recognized Ubcx in Xenopus egg extracts. Other antibodies were taken manually, separated by SDS-PAGE, and analyzed by used in this study were as follows: anti-Apc3 (Transduction labs; immunoblotting. The standard proteins (S value) used for the estima- C40920), anti-human Fizzy, anti-Apc2, anti-Apc6, and anti-Apc8 (a tion were thyroglobulin (19.4 S), catalase (11.3 S) and bovine serum kind gift from Jan-Michael Peters) for immunoblotting. Anti-Apc3 albumin (BSA; 4.5 S). Molecular Cell 146

Surface Plasmon Resonance Hershko, A., and Ciechanover, A. (1998). The ubiquitin system. Annu. Biotinylated N70 and Dm were attached onto a SA sensor chip Rev. Biochem. 67, 425–479. (Biacore). Anaphase or interphase APC/C (50 ␮l) purified from 10 Hilioti, Z., Chung, Y.S., Mochizuki, Y., Hardy, C.F., and Cohen-Fix, O. ml of HSS was injected onto the sensor chip at a flow rate 30 ␮l/min, (2001). The anaphase inhibitor Pds1 binds to the APC/C-associated and analyzed by surface plasmon resonance (SPR) using a Biacore protein Cdc20 in a destruction box-dependent manner. Curr. Biol. 3000 machine. 11, 1347–1352. Hochstrasser, M. (1996). Ubiquitin-dependent protein degradation. Acknowledgments Annu. Rev. Genet. 30, 405–439. Irniger, S., Piatti, S., Michaelis, C., and Nasmyth, K. (1995). We thank Jan-Michael Peters for antibodies against several APC/C involved in sister chromatid separation are needed for B-type cyclin subunits and against human Fizzy, Chizuko Tsurumi for anti-Xeno- proteolysis in budding yeast. Cell 81, 269–278. pus Fizzy antibodies, and Jon Pines for human securin and the D-box/KEN-box mutant constructs. We are also grateful to Cancer Jackson, P.K., Eldridge, A.G., Freed, E., Furstenthal, L., Hsu, J.Y., Research UK’s central services, in particular, Nicola O’Reilly for Kaiser, B.K., and Reimann, J.D. (2000). The lore of the RINGs: sub- peptide synthesis and Nick Totty for mass spectrometry. H.Y. was strate recognition and catalysis by ubiquitin ligases. Trends Cell supported by the “Unit Process and Combined Circuit,” PRESTO, Biol. 10, 429–439. JST, Japan. Juang, Y.L., Huang, J., Peters, J.M., McLaughlin, M.E., Tai, C.Y., and Pellman, D. (1997). APC-mediated proteolysis of Ase1 and the Received: August 20, 2003 morphogenesis of the mitotic spindle. Science 275, 1311–1314. Revised: November 11, 2003 King, R.W., Peters, J.M., Tugendreich, S., Rolfe, M., Hieter, P., and Accepted: November 11, 2003 Kirschner, M.W. (1995). A 20S complex containing CDC27 and Published: January 15, 2004 CDC16 catalyzes the mitosis-specific conjugation of ubiquitin to cyclin B. Cell 81, 279–288. References Kotani, S., Tugendreich, S., Fujii, M., Jorgensen, P.M., Watanabe, N., Hoog, C., Hieter, P., and Todokoro, K. (1998). PKA and MPF- Burton, J.L., and Solomon, M.J. (2001). D box and KEN box motifs activated polo-like kinase regulate anaphase-promoting complex in budding yeast Hsl1p are required for APC-mediated degradation activity and mitosis progression. Mol. Cell 1, 371–380. and direct binding to Cdc20p and Cdh1p. Genes Dev. 15, 2381–2395. Kramer, E.R., Scheuringer, N., Podtelejnikov, A.V., Mann, M., and Carroll, C.W., and Morgan, D.O. (2002). The Doc1 subunit is a pro- Peters, J.M. (2000). Mitotic regulation of the APC activator proteins cessivity factor for the anaphase-promoting complex. Nat. Cell Biol. CDC20 and CDH1. Mol. Biol. Cell 11, 1555–1569. 4, 880–887. Lahav-Baratz, S., Sudakin, V., Ruderman, J.V., and Hershko, A. Charles, J.F., Jaspersen, S.L., Tinker-Kulberg, R.L., Hwang, L., Szi- (1995). Reversible phosphorylation controls the activity of cyclo- don, A., and Morgan, D.O. (1998). The Polo-related kinase Cdc5 some-associated cyclin-ubiquitin ligase. Proc. Natl. Acad. Sci. USA activates and is destroyed by the mitotic cyclin destruction machin- 92, 9303–9307. ery in S. cerevisiae. Curr. Biol. 8, 497–507. Lim, H.H., Goh, P.Y., and Surana, U. (1998). Cdc20 is essential for Cohen-Fix, O., Peters, J.M., Kirschner, M.W., and Koshland, D. the cyclosome-mediated proteolysis of both Pds1 and Clb2 during (1996). Anaphase initiation in Saccharomyces cerevisiae is con- M phase in budding yeast. Curr. Biol. 8, 231–234. trolled by the APC-dependent degradation of the anaphase inhibitor Lorca, T., Castro, A., Martinez, A.M., Vigneron, S., Morin, N., Sigrist, Pds1p. Genes Dev. 10, 3081–3093. S., Lehner, C., Doree, M., and Labbe, J.C. (1998). Fizzy is required Dawson, I.A., Roth, S., and Artavanis-Tsakonas, S. (1995). The Dro- for activation of the APC/cyclosome in Xenopus egg extracts. EMBO sophila cell cycle fizzy is required for normal degradation of J. 17, 3565–3575. cyclins A and B during mitosis and has homology to the CDC20 Masui, Y., and Markert, C.L. (1971). Cytoplasmic control of nuclear gene of Saccharomyces cerevisiae. J. Cell Biol. 129, 725–737. behavior during meiotic maturation of frog oocytes. J. Exp. Zool. Fang, G., Yu, H., and Kirschner, M.W. (1998). Direct binding of CDC20 177, 129–146. protein family members activates the anaphase-promoting complex McGarry, T.J., and Kirschner, M.W. (1998). Geminin, an inhibitor of in mitosis and G1. Mol. Cell 2, 163–171. DNA replication, is degraded during mitosis. Cell 93, 1043–1053. Funabiki, H., and Murray, A.W. (2000). The Xenopus chromokinesin Meyn, M.A., III, Melloy, P.G., Li, J., and Holloway, S.L. (2002). The Xkid is essential for metaphase alignment and must destruction box of the cyclin Clb2 binds the anaphase-promoting be degraded to allow anaphase chromosome movement. Cell complex/cyclosome subunit Cdc23. Arch. Biochem. Biophys. 407, 102, 411–424. 189–195. Funabiki, H., Yamano, H., Kumada, K., Nagao, K., Hunt, T., and Morgan, D.O. (1999). Regulation of the APC and the exit from mitosis. Yanagida, M. (1996). Cut2 proteolysis required for sister-chromatid Nat. Cell Biol. 1, E47–E53. seperation in fission yeast. Nature 381, 438–441. Murray, A.W., Solomon, M.J., and Kirschner, M.W. (1989). The role Glotzer, M., Murray, A.W., and Kirschner, M.W. (1991). Cyclin is of cyclin synthesis and degradation in the control of maturation degraded by the ubiquitin pathway. Nature 349, 132–138. promoting factor activity. Nature 339, 280–286. Golan, A., Yudkovsky, Y., and Hershko, A. (2002). The cyclin-ubiqui- Ohtoshi, A., Maeda, T., Higashi, H., Ashizawa, S., and Hatakeyama, tin ligase activity of cyclosome/APC is jointly activated by protein M. (2000). Human p55(CDC)/Cdc20 associates with cyclin A and is kinases Cdk1-cyclin B and Plk. J. Biol. Chem. 277, 15552–15557. phosphorylated by the cyclin A-Cdk2 complex. Biochem. Biophys. Hagting, A., den Elzen, N., Vodermaier, H.C., Waizenegger, I.C., Res. Commun. 268, 530–534. Peters, J.M., and Pines, J. (2002). Human securin proteolysis is Passmore, L.A., McCormack, E.A., Au, S.W., Paul, A., Willison, K.R., controlled by the spindle checkpoint and reveals when the APC/C Harper, J.W., and Barford, D. (2003). Doc1 mediates the activity switches from activation by Cdc20 to Cdh1. J. Cell Biol. 157, 1125– of the anaphase-promoting complex by contributing to substrate 1137. recognition. EMBO J. 22, 786–796. Hames, R.S., Wattam, S.L., Yamano, H., Bacchieri, R., and Fry, A.M. Patra, D., and Dunphy, W.G. (1998). Xe-p9, a Xenopus Suc1/Cks (2001). APC/C-mediated destruction of the centrosomal kinase protein, is essential for the Cdc2-dependent phosphorylation of the Nek2A occurs in early mitosis and depends upon a cyclin A-type anaphase- promoting complex at mitosis. Genes Dev. 12, 2549– D-box. EMBO J. 20, 7117–7127. 2559. Harper, J.W., Burton, J.L., and Solomon, M.J. (2002). The anaphase- Peters, J.M. (2002). The anaphase-promoting complex: proteolysis promoting complex: it’s not just for mitosis any more. Genes Dev. in mitosis and beyond. Mol. Cell 9, 931–943. 16, 2179–2206. Pfleger, C.M., Lee, E., and Kirschner, M.W. (2001). Substrate recog- Destruction Box Receptor of Cyclin B 147

nition by the Cdc20 and Cdh1 components of the anaphase-promot- ing complex. Genes Dev. 15, 2396–2407. Rudner, A.D., and Murray, A.W. (2000). Phosphorylation by Cdc28 activates the Cdc20-dependent activity of the anaphase-promoting complex. J. Cell Biol. 149, 1377–1390. Schwab, M., Lutum, A.S., and Seufert, W. (1997). Yeast Hct1 is a regulator of Clb2 cyclin proteolysis. Cell 90, 683–693. Schwab, M., Neutzner, M., Mocker, D., and Seufert, W. (2001). Yeast Hct1 recognizes the mitotic cyclin Clb2 and other substrates of the ubiquitin ligase APC. EMBO J. 20, 5165–5175. Shirayama, M., Zachariae, W., Ciosk, R., and Nasmyth, K. (1998). The Polo-like kinase Cdc5p and the WD-repeat protein Cdc20p/fizzy are regulators and substrates of the anaphase promoting complex in Saccharomyces cerevisiae. EMBO J. 17, 1336–1349. Sigrist, S., Jacobs, H., Stratmann, R., and Lehner, C.F. (1995). Exit from mitosis is regulated by Drosophila fizzy and the sequential destruction of cyclins A, B and B3. EMBO J. 14, 4827–4838. Sudakin, V., Ganoth, D., Dahan, A., Heller, H., Hershko, J., Luca, F.C., Ruderman, J.V., and Hershko, A. (1995). The cyclosome, a large complex containing cyclin-selective ubiquitin ligase activity, targets cyclins for destruction at the end of mitosis. Mol. Biol. Cell 6, 185–197. Tugendreich, S., Tomkiel, J., Earnshaw, W., and Hieter, P. (1995). CDC27Hs colocalizes with CDC16Hs to the centrosome and mitotic spindle and is essential for the metaphase to anaphase transition. Cell 81, 261–268. Tunquist, B.J., and Maller, J.L. (2003). Under arrest: cytostatic factor (CSF)-mediated metaphase arrest in vertebrate eggs. Genes Dev. 17, 683–710. Varshavsky, A. (1997). The ubiquitin system. Trends Biochem. Sci. 22, 383–387. Visintin, R., Prinz, S., and Amon, A. (1997). CDC20 and CDH1: a family of substrate-specific activators of APC-dependent proteoly- sis. Science 278, 460–463. Watanabe, N., Hunt, T., Ikawa, Y., and Sagata, N. (1991). Indepen- dent inactivation of MPF and cytostatic factor (Mos) upon fertiliza- tion of Xenopus eggs. Nature 352, 247–248. Yamano, H., Gannon, J., and Hunt, T. (1996). The role of proteolysis in cell cycle progression in Schizosaccharomyces pombe. EMBO J. 15, 5268–5279. Yamada, H., Kumada, K., and Yanagida, M. (1997). Distinct subunit functions and cell cycle regulated phosphorylation of 20S APC/ cyclosome required for anaphase in fission yeast. J. Cell Sci. 110, 1793–1804. Yamano, H., Tsurumi, C., Gannon, J., and Hunt, T. (1998). The role of the destruction box and its neighbouring lysine residues in cyclin B for anaphase ubiquitin-dependent proteolysis in fission yeast: defining the D-box receptor. EMBO J. 17, 5670–5678. Zachariae, W., and Nasmyth, K. (1999). Whose end is destruction: cell division and the anaphase-promoting complex. Genes Dev. 13, 2039–2058. Zachariae, W., Schwab, M., Nasmyth, K., and Seufert, W. (1998). Control of cyclin ubiquitination by CDK-regulated binding of Hct1 to the anaphase promoting complex. Science 282, 1721–1724.