Binding of Activated Cyclosome to p13 suc1

Previous studies have indicated that a ∼1,500-kDa complex, designated the cyclosome or anaphase-promoting complex, has a regulated cyclin-ubiquitin ligase activity that targets cyclin B for degradation at the end of mitosis. The cyclosome is inactive in the interphase of the embryonic cell cycle and is converted to the active form in late mitosis in a phosphorylation-dependent process initiated by protein kinase Cdc2-cyclin B. We show here that the active, phosphorylated form of the cyclosome from clam oocytes binds to p13 suc1 , a protein known to associate with Cdc2. The following evidence indicates that the binding of the cyclosome to p13 suc1 is not mediated via the Cdc2-cyclin B complex: (a) activated cyclosome binds to p13 suc1 -Sepharose following its separation from Cdc2-cyclin B by gel filtration chromatography; (b) cyclosome from interphase extracts, activated by a kinase in which cyclin B has been replaced by an N-terminally truncated derivative fused to glutathione S-transferase, binds well to p13 suc1 -Sepharose but not to glutathione-agarose. An alternative possibility, that the phosphorylated cyclosome binds directly to a phosphate-binding site of p13 suc1 , is supported by the observation that the cyclosome is efficiently eluted from p13 suc1 -Sepharose by phosphate-containing compounds. This information was utilized to develop a procedure for the affinity purification of the cyclosome. A factor abundant in the fraction not adsorbed to p13 suc1 -Sepharose stimulates the activity of purified cyclosome. It is suggested that binding of Suc1 may have a role in the regulation of cyclosome activity.

subunit of protein kinase Cdc2. The activation of Cdc2 is required for cells to undergo mitosis, while the inactivation of Cdc2, caused by the specific and regulated proteolysis of its cyclin B subunit, is essential for exit from mitosis (reviewed in Ref. 3). Studies in cell-free extracts that reproduce embryonic cell cycles showed that cyclin B is degraded by the ubiquitin pathway (4, 5), a system in which proteins are targeted for degradation by ligation to ubiquitin (reviewed in Ref. 6). Cyclin ubiquitinylation and degradation depends on a partially conserved 9-amino acid motif, the "destruction box," which is usually located ϳ40 -50 amino acid residues from the N terminus of mitotic cyclins (4).
We have been studying the mechanisms of cyclin degradation by fractionation of a clam oocyte cell-free system (7). Three components were identified to be involved in the ligation of ubiquitin to cyclin B: the ubiquitin-activating enzyme E1, a specific ubiquitin-carrier protein E2-C (7,8), and a cyclin ubiquitin ligase activity associated with particulate material. E1 and E2-C are constitutively active, but the particle-associated ligase activity is cell cycle-regulated; it is inactive in the interphase, but becomes activated at the end of mitosis (7). The particle-associated component was extracted with salt, partially purified, and found to be a ϳ1,500-kDa complex, the cyclosome (1). The cyclosome acts on both cyclin A and cyclin B and requires intact destruction box sequences of both cyclins. The activity of the cyclosome is regulated by reversible phosphorylation, as indicated by the findings that the active, mitotic form of the cyclosome can be converted to the inactive form by treatment with an okadaic acid-sensitive phosphatase (9) and that the inactive, interphase form of the cyclosome can be converted to the active form by incubation with protein kinase Cdc2-cyclin B (1,9). Activation of the cyclosome by protein kinase Cdc2-cyclin B includes a time lag (1,9), which may serve to prevent the premature inactivation of the kinase in the cell cycle. Based on these findings, we suggested the cyclosome has a regulated cyclin-ubiquitin ligase activity, which targets cyclin B for destruction at the end of mitosis (1).
A similar particle, termed the APC, was identified in Xenopus egg extracts by King et al. (2). By the use of antibodies that cross-react with Xenopus proteins, these investigators furthermore identified two subunits of the complex as homologues of the products of budding yeast CDC16 and CDC27 genes, which are required for exit from mitosis and the degradation of mitotic cyclins (10). Other subunits of the cyclosome/APC particle have been identified in a variety of organisms (11)(12)(13)(14)(15)(16) and appear to be strongly conserved in evolution. The cyclosome is also involved in the degradation of anaphase inhibitor proteins, which contain destruction box sequences (17)(18)(19). The subunits of the complex responsible for its specific actions, as well as those responsible for the regulation of its activity, have not yet been identified.
To define the mode of action and mechanisms of regulation of the cyclosome, its extensive purification is essential. The Xenopus cyclosome/APC particle was purified by immunoprecipitation (2,15), but such preparations are not suitable for biochemical studies. In the present report we describe the binding of the cyclosome to p13 suc1 and its use for affinity purification. p13 suc1 was originally identified in the fission yeast by its ability to suppress certain temperature-sensitive mutations of Cdc2 (20). A homologous protein in budding yeast, designated Cks, was found to bind strongly to the cyclin-dependent kinase (21). The Suc1/Cks family of proteins is essential for viability and is highly conserved in evolution (reviewed in Refs. 22 and 23), but its exact functions remained unknown. Genetic evidence in yeasts indicates multiple roles in the cell cycle, including entry into mitosis, exit from mitosis, and transition between G 1 and S phases of the cell cycle (24,25). Recent biochemical studies with immunodepleted extracts of Xenopus eggs further indicated that Suc1/Cks is required in at least two stages of the embryonic cell cycle: in the activation of the Cdc2-cyclin B complex by tyrosine dephosphorylation of Cdc2, and in exit from mitosis due to cyclin B degradation (26). The crystal structures of Suc1/Cks proteins (27,28) and of their complex with Cdc2 (29) have been solved. In addition to the Cdk binding site they contain a highly conserved phosphate-binding site. It has been suggested (but not yet demonstrated) that the latter site may bind to some phosphorylated proteins, and thus Suc1/ Cks may have a role in targeting Cdk-cyclin kinases to certain phosphorylated proteins (26,29).
We show here that p13 suc1 selectively binds the active, phosphorylated form of the cyclosome. Our evidence suggests that this binding is due to interaction of the cyclosome with the phosphate-binding site of p13 suc1 . This information was utilized to develop an affinity procedure for the purification of the cyclosome.

EXPERIMENTAL PROCEDURES
Ubiquitin from bovine erythrocytes, rcm-BSA, STI, and p-nitrophenylphosphate (pNPP) were obtained from Sigma, and okadaic acid was obtained from Boehringer Mannheim. Ubiquitin aldehyde was prepared as described (30). E1 was purified from human erythrocytes (31). Recombinant clam E2-C was expressed in bacteria as described (8) and purified by gel filtration on Superdex 50. p13 suc1 was expressed in E. coli and purified by gel filtration as described (32). p13 suc1 was coupled to cyanogen bromide-activated Sepharose-4B at a concentration of 11-13 mg of protein/ml of swollen gel. For control beads, BSA was coupled to Sepharose at a similar concentration. Glutathione-agarose was purchased from Sigma (G-4510). Recombinant human GST-⌬88cyclin B was expressed as described (33) and purified by affinity chromatography on glutathione-agarose. The expression and purification of clam cyclin A and its N-terminally truncated derivative (1,34) and of N-terminal fragments of sea urchin cyclin B fused to protein A (4, 17) have been described previously. Proteins were radioiodinated as described (4).
Extracts of M phase and interphase clam oocytes were prepared as described previously (7). Both types of extracts were subjected to fractionation on DEAE-cellulose (7), and the fraction not adsorbed to the column, fraction 1, was extracted with 0.25 M KCl, as described (1). This salt extract of fraction 1, which contains the cyclosome (1), served as the source for affinity purification.
Affinity Purification of Cyclosome-Prior to affinity purification, the cyclosome was converted to the active form by preincubation with ATP (1,9) as follows. The incubation mixture contained the following in a volume of 1.2 ml: 50 mM Hepes-KOH (pH 7.2), 1 mM DTT, 1 mM MgCl 2 , 0.5 mM ATP, 10 mM phosphocreatine, 100 g/ml creatine phosphokinase, 12 mg of protein of salt extract of fraction 1 from M phase clam oocytes, and 1 M okadaic acid. Following incubation at 18°C for 60 min, the sample was mixed with an equal volume of a solution consisting of 500 mM KCl, 40 mM Tris-HCl (pH 7.2) and 1 mM DTT and then added to 1.2 ml of p13 suc1 -Sepharose beads. The beads had been washed previously three times with 10-ml portions of buffer A (50 mM Tris-HCl, pH 7.2, 0.25 M KCl, and 1 mM DTT). The sample was mixed with beads at 0°C for 60 min and then transferred to a column (0.7 cm diameter) at 4°C. The flow-through fraction was collected, together with a 5-ml wash with buffer A. The column was washed with 30 ml of buffer B (50 mM Tris-HCl (pH 7.2), 250 mM KCl, 20% (v/v) glycerol, 1 mM DTT, and 0.5% (v/v) Nonidet P-40) and then with 10 ml of buffer C (50 mM Tris-HCl, pH 7.2, 1 mM DTT, and 20% (v/v) glycerol). The cyclosome was eluted either with 30 ml of 50 mM Tris-HCl, pH 9.0, containing 1 mM DTT and 0.2 mg/ml STI, or with 20 ml of 50 mM pNPP in 50 mM Tris-HCl, pH 7.2, 1 mM DTT, and 0.2 mg/ml STI. The addition of STI was necessary to prevent the adsorption of the dilute enzyme to surfaces. STI was chosen as the carrier protein, since it has a relatively low molecular mass (20 kDa), and thus it does not interfere with the detection of cyclosome subunits, which are of larger size, in SDSpolyacrylamide gel electrophoresis. Elution was at a flow rate of 1-1.5 ml/min. When the pH 9 buffer was used for elution, the eluate was collected to a tube containing Tris-HCl, pH 7.2 (0.1 M, final concentration), to decrease the pH. Both eluate and flow-through fractions were concentrated with Centriprep-10 concentrators (Amicon), diluted at least 10-fold with 50 mM Tris-HCl (pH 7.2) containing 1 mM DTT and concentrated again to ϳ0.5 ml. Glycerol was added to 20% (v/v), and samples were stored at Ϫ70°C. p13 suc1 -Sepharose beads could be regenerated by washing with 30 ml of 50 mM Tris-HCl (pH 9.0), followed by a wash with 30 ml of 1 M KCl. The beads were stored in 50 mM Tris-HCl (pH 7.2) containing 0.02% sodium azide.
Preparation of GST-⌬88-cyclin B-Cdc2-GST-⌬88-cyclin B-Cdc2 was prepared by the incubation of interphase extracts of clam oocytes (in which cyclin B is absent) with recombinant GST-⌬88-cyclin B and ATP, followed by the purification of the active kinase. The reaction mixture contained the following in a volume of 3.6 ml: 15 mg/ml protein of extract of interphase clam oocytes, 0.2 mg/ml GST-⌬88-cyclin B, and other ingredients similar to those described above for preincubation with ATP and okadaic acid, except that the concentration of MgCl 2 was 5 mM. Following incubation at 18°C for 60 min, the formation of active kinase was verified by an assay of histone H1 kinase activity. It was found that H1 kinase activity rose from very low initial levels (ϳ0.5 units/g of protein) to levels comparable with those in mitotic extracts (ϳ60 units/g of protein). The kinase was then purified by adsorption to p13 suc1 -Sepharose, elution with free p13 suc1 , and removal of free p13 suc1 by gel filtration, as described (7).
Assay of Cyclin-Ubiquitin Ligase Activity-Reaction mixtures contained the following in a volume of 10 l: 40 mM Tris-HCl (pH 7.6), 1 mg/ml rcm-BSA, 1 mM DTT, 5 mM MgCl 2 , 10 mM phosphocreatine, 50 mg/ml creatine phosphokinase, 50 M ubiquitin, 1 M ubiquitin aldehyde, 1 pmol of E1, 5 pmol of E2-C, 1 M okadaic acid, enzyme source as specified, and 1-2 pmol (ϳ10 5 cpm) of 125 I-labeled cyclin B-(13-91)/ protein A. This N-terminal fragment of cyclin B has been shown to be a suitable substrate for destruction box-specific, cell cycle-regulated ubiquitinylation and degradation of mitotic cyclins (1,4,7). When enzyme activity associated with Sepharose beads was determined, samples were agitated during incubation. Following incubation at 18°C for 1 h, the samples were subjected to electrophoresis on a 12.5% polyacrylamide-SDS gel. Results were quantified with a PhosphorImager (Molecular Dynamics). The amount of radioactivity in all cyclin-ubiquitin conjugates was expressed as the percentage of total radioactivity in each lane (1). Reactions were conducted in the range linear with enzyme concentration, which was 5-40% 125 I-cyclin ligated to ubiquitin. One unit of ligase activity was defined as that converting 1% 125 I-cyclin to ubiquitin conjugates under the conditions described above.
Miscellaneous Assays-Activity of protein kinase cyclin B-Cdc2 was measured by the phosphorylation of histone H1 following adsorption to p13 suc1 -Sepharose. Virtually all such activity in extracts of meiotic M phase clam oocytes is due to cyclin B-Cdc2 complexes (data not shown). Histone H1 kinase assays were conducted as described (35), except that the concentration of ATP was 300 M, and reactions were conducted at 18°C. One unit of enzyme activity is defined as that causing the incorporation of 1 pmol of phosphate into histone H1 under these conditions. Protein concentration was determined with the Bio-Rad assay, using bovine serum albumin as standard. To estimate the amount of protein bound to beads, proteins were first eluted from beads by mixing with 4 M guanidine hydrochloride for 1 h at room temperature.

RESULTS
Binding of Active Cyclosome to p13 suc1 -Sepharose-p13 suc1 from fission yeast and homologous Cks proteins from other organisms bind strongly to Cdc2 and to some other members of the Cdk family of proteins (22,23). p13 suc1 -Sepharose beads are therefore commonly used to isolate Cdk-cyclin complexes (36).
In preliminary experiments, we found that the active form of the cyclosome 2 bound tightly to p13 suc1 -Sepharose. We then tried to examine the nature of the interaction between p13 suc1 and the cyclosome and to exploit this binding for the affinity purification of the cyclosome.
In the experiment shown in Fig. 1, a crude fraction from M phase clam oocytes (1) was preincubated with ATP, or not preincubated, and then was applied to p13 suc1 -Sepharose beads. Such M phase extracts are made about 10 -15 min before the anaphase of meiosis I (7), when protein kinase Cdc2cyclin B is already active but the cyclosome is still inactive. When such extracts are incubated in the presence of ATP and okadaic acid, the cyclosome becomes activated by endogenous protein kinase Cdc2-cyclin B (1,9). Following mixing of preincubated extracts with p13 suc1 -Sepharose or control beads, the beads were thoroughly washed, and cyclin-ubiquitin ligation activity associated with the beads or remaining in the supernatants was assayed as described under "Experimental Procedures." Cyclin-ubiquitin conjugates are the ladder of bands of molecular size higher than free 125 I-cyclin. As shown in Fig. 1, without preincubation, very little cyclin-ubiquitin ligation activity was associated with p13 suc1 -Sepharose beads (lane 1), and most of it remained in the supernatant (lane 4). Following preincubation of extracts with ATP, considerable activity of cyclin-ubiquitin ligation was associated with p13 suc1 -Sepharose beads (lane 2), and it disappeared from the supernatant almost completely (lane 5). These findings suggest that the conversion of the cyclosome to the active form increases its affinity to p13 suc1 . Although almost all ligase activity disappeared from the supernatants of preincubated extracts, only 15-25% of the activity was recovered associated with p13 suc1 -Sepharose beads. Binding of cyclosome was apparently specific for p13 suc1 -Sepharose, because under similar conditions, there was no significant binding to control (BSA-Sepharose) beads (Fig. 1,  lane 3), and all activity remained in the supernatant (lane 6).
Mode of Binding of Cyclosome to p13 suc1 -We have considered several alternative possibilities to account for the binding of the cyclosome to p13 suc1 . It is possible that binding is not direct but that the cyclosome binds Cdc2-cyclin B, which in turn is bound to p13 suc1 . The cyclin B subunit of the Cdc2-cyclin B protein kinase complex is a substrate of the cyclosome for ubiquitin ligation, and thus cyclin B-Cdc2 may be tightly bound to the active site of the ligase. The Cdc2-cyclin B protein kinase is also an activator (although not necessarily a direct one) of the cyclosome, and it is possible that the kinase is tightly bound to site(s) of the cyclosome that it may phosphorylate. In both of these cases, p13 suc1 would bind to the Cdc2 subunit of the Cdc2-cyclin B complex, which is tightly associated with the cyclosome. A third possibility is that p13 suc1 may bind directly to the active, phosphorylated form of the cyclosome, possibly by its phosphate-binding site (27)(28)(29).
We have first examined the possibility that the cyclosome is bound to p13 suc1 via Cdc2-cyclin B by asking the question whether, following activation of the cyclosome, the presence of Cdc2-cyclin B is still required for binding to p13 suc1 -Sepharose. In the experiment shown in Fig. 2, the cyclosome was first converted to the active form by preincubation of M phase extract with ATP and then was separated from Cdc2-cyclin B by gel filtration on Superose-6 in the presence of salt. As shown previously (9), this procedure separates the active cyclosome (ϳ1,500 kDa) from most of Cdc2-cyclin B (ϳ100 kDa). Following gel filtration, the binding of each fraction to p13 suc1 -Sepharose beads was examined. It may be seen in Fig. 2 that the profile of cyclin-ubiquitin ligation activity adsorbed to p13 suc1 -Sepharose closely followed that before adsorption. The recovery of activity was about 20% in each fraction, regardless of the amount of residual Cdc2-cyclin B. Thus, for example, cyclinubiquitin ligase activity bound to p13 suc1 beads in fractions 20 -22, which had no detectable Cdc2-cyclin B kinase activity, to an extent similar to that in fractions 24 -26, which contained a low amount of residual kinase activity. In all fractions, cyclinubiquitin ligase activity was completely removed from the supernatants following adsorption to p13 suc1 beads (data not shown). These findings suggest that following the activation of the cyclosome, the continued presence of Cdc2-cyclin B is not required for the binding of the cyclosome to p13 suc1 .
It is possible that the active form of the cyclosome is associated with a tightly bound molecule of Cdc2-cyclin B, which is below the detection limit of our kinase assay and is not dissociated during gel filtration. We therefore continued to examine this problem by another approach, using interphase extracts. In interphase extracts (prepared from emetine-arrested twocell clam embryos; Ref. 7) cyclin B is mostly degraded, levels of protein kinase Cdc2-cyclin B are about 100-fold lower than in M phase extracts, and the cyclosome is inactive. The cyclosome can be converted to the active form by incubation of interphase extracts with protein kinase Cdc2-cyclin B and ATP (1,7). In the experiment shown in Table I, interphase extract was incubated with a derivative of Cdc2 kinase containing recombinant cyclin B that lacked the N-terminal 88 amino acid residues and was fused to glutathione S-transferase (GST). This truncated derivative of cyclin B can form active protein kinase with Cdc2, which can activate the cyclosome, but it cannot be ubiquitinyl- and then was applied to p13 suc1 -Sepharose or BSA-Sepharose beads as described under "Experimental Procedures," except that samples of 200 g of protein were added to 20 l of beads. Following adsorption at 0°C for 1 h, the supernatants were collected, and the beads were washed four times with 1-ml portions of buffer B, followed by two washes with buffer C (see "Experimental Procedures"). Cyclin-ubiquitin ligation was determined in samples of 5% of the washed beads (lanes 1-3) and 2.5% of the supernatants (lanes 4 -6), as described under "Experimental Procedures." Cyc, position of free 125 I-cyclin. Numbers on the right indicate the position of molecular mass markers (kDa). ated by the cyclosome because it lacks the "destruction box" region at the N-terminal part (4,34). N-terminally truncated cyclin B does not interfere with cyclin-ubiquitin ligation, even at high concentrations (Ref. 4 and see below), so it presumably cannot bind to the active site of the cyclosome involved in cyclin-ubiquitin ligation. Following incubation of interphase extract with ATP and without added protein kinase Cdc2, only a small amount of cyclin-ubiquitin ligase activity was found to be bound to p13 suc1 -Sepharose beads, which was presumably due to the action of the low amounts of Cdc2-cyclin B (assayed as H1 kinase) that remain in such extracts. After incubation of interphase extracts with protein kinase GST-cyclin B-Cdc2, which activates the cyclosome, there was a marked increase in the amount of cyclin-ubiquitin ligation activity associated with p13 suc1 -Sepharose beads (Table I). This was accompanied by a complete disappearance of cyclin-ubiquitin ligase activity from the supernatant (data not shown). This finding confirms the conclusion that the active form of the cyclosome binds preferentially to p13 suc1 . This observation also appears to be incompatible with the notion that the cyclosome is bound to p13 suc1 via cyclin B bound to the active ubiquitination site, since ⌬88cyclin B is neither a substrate nor a competitive inhibitor, and thus it cannot be bound to the ubiquitination active site of the cyclosome.
In this experiment, we have also taken advantage of the fact that GST-⌬88-cyclin B-Cdc2 can be bound to glutathione-agarose beads via its GST moiety. If binding of the cyclosome to p13 suc1 is mediated by Cdc2-cyclin B bound to any site or subunit of the cyclosome, it would be expected that, following incubation of interphase extracts with GST-⌬88-cyclin B-Cdc2, the cyclosome would be bound to GSH-agarose beads along with the kinase. However, no significant binding of cyclinubiquitin ligase activity to GSH beads could be detected under conditions identical to those promoting cyclosome binding to p13 suc1 -Sepharose (Table I). By contrast, the binding of GST-⌬88-cyclin B-Cdc2 to GSH beads (assayed by H1 kinase activity) was as efficient as its binding to p13 suc1 beads. The cumulative evidence from the above experiments does not support the notion that Cdc2-cyclin B mediates an indirect binding of the cyclosome to p13 suc1 . A direct binding of active, phosphorylated cyclosome to the phosphate-binding site of p13 suc1 is suggested by the nature of compounds that promote the elution of the cyclosome from p13 suc1 beads, as described below.
Elution of Cyclosome from p13 suc1 -Sepharose Beads-We have next examined different experimental conditions for the elution of the cyclosome from p13 suc1 beads. The aim of these FIG. 2. Activated cyclosome binds to p13 suc1 -Sepharose following separation from Cdc2-cyclin B by gel filtration chromatography. Salt extract of fraction 1 from M phase clam oocytes (6 mg of protein) was preincubated with ATP, as described under "Experimental Procedures," and then was applied to a Superose-6 HR 10/30 column (Pharmacia Biotech Inc.) equilibrated with 50 mM Tris-HCl (pH 7.2), 250 mM KCl, 0.2 mg/ml rcm-BSA, and 1 mM DTT. Fractions of 0.5 ml were collected at a flow rate of 0.4 ml/min. Column fractions were concentrated by ultrafiltration with Centricon-30 concentrators (Amicon), diluted 20-fold with buffer C (see "Experimental Procedures"), and concentrated again to final volume of 45-60 l. 70% of every second fraction was adsorbed onto 30 l of p13 suc1 -Sepharose beads, and then the beads were washed as described in the legend to Fig. 1. Cyclin-ubiquitin ligase activity was assayed in samples of column fractions before adsorption (E) and adsorbed to beads (q) and was expressed as the total activity in each fraction. Cdc2-cyclin B kinase activity associated with p13 suc1 -Sepharose beads (Ⅺ) was assayed by the phosphorylation of histone H1, as described under "Experimental Procedures."

TABLE I Effects of incubation of interphase extract with GST-⌬88 cyclin B/ Cdc2 on binding of cyclin-ubiquitin ligase to immobilized p13 or GSH
Samples of salt extract of fraction 1 from interphase oocytes (150 g of protein) were preincubated with ATP as described under "Experimental Procedures," except that the time of preincubation was 90 min. Where indicated, 4,000 units of GST-⌬88 cyclin B/Cdc2 were added in preincubation. Subsequently, samples were brought to 250 mM KCl, as described under "Experimental Procedures," and absorbed onto 40 l of beads indicated in the table. Cyclin-ubiquitin ligase and histone H1 kinase activities associated with beads were assayed as described under "Experimental Procedures." experiments was 2-fold: to gain an insight into the mode of the interaction of p13 suc1 with the cyclosome and to establish experimental conditions for the affinity purification of the cyclosome. In the experiments shown in Table II, activated cyclosome was first bound to p13 suc1 -Sepharose, and then samples of the beads were suspended in various solutions. The elution of the enzyme was monitored both by the decrease of activity associated with beads and by its appearance in the eluate. At pH 7.2 and at low ionic strength, the enzyme was strongly bound to p13 suc1 -Sepharose beads, and no significant activity was eluted in this solution. Treatment with 300 mM KCl caused a partial loss of enzyme activity from beads, while 600 mM KCl caused a drastic loss of cyclin-ubiquitin ligase activity associated with p13 suc1 -Sepharose. However, this was not accompanied by a corresponding increase in enzyme activity in the eluate (Table II, experiment 1). It rather seems that high salt concentration causes the inactivation of cyclin-ubiquitin ligase. In contrast to the effects of KCl, 150 mM potassium phosphate caused significant elution of enzymatically active cyclosome from p13 suc1 -Sepharose (Table II, experiment 1). A similar elution was obtained with 150 mM potassium sulfate (data not shown). Phosphate and sulfate are known to bind to the anionbinding site of Cks/Suc1 proteins (27)(28)(29), and thus the results suggest that elution by these compounds may be due to competition with the cyclosome on binding to the anionic site of p13 suc1 . Under the conditions described above, elution of the cyclosome by phosphate and sulfate was only partial, and thus was not suitable for affinity purification. Higher concentrations of phosphate or sulfate caused the loss of enzyme activity, presumably due to inactivation by high ionic strength. We therefore searched for other phosphate compounds that may elute the cyclosome and found that pNPP at significantly lower concentration (50 mM), caused an almost complete loss from p13 suc1 beads and a relatively high yield in the eluate (Table II, experiment 1). The role of the phosphate moiety of pNPP in elution was indicated by the lack of influence of p-nitrophenylglycerol (50 mM) on this process (data not shown). Efficient elution of the enzyme from p13 suc1 beads was also obtained by raising the pH to 9.0 (Table II, experiment 2). In subsequent experiments we therefore used elution by either pH 9 or pNPP for affinity purification of the cyclosome.
It is notable that free p13 suc1 , at concentrations of 1-5 mg/ml, did not cause significant elution of cyclosome from p13 suc1 -Sepharose beads, although it eluted protein kinase Cdc2-cyclin B at high efficiency (data not shown). On the other hand, free p13 suc1 did prevent the binding of the cyclosome to p13 suc1 -Sepharose beads. It might be that the binding of the cyclosome particle to the beads sterically hinders the access of free p13 suc1 to the binding site.
The Fraction Not Adsorbed to p13 suc1 -Sepharose Contains Factor(s) That Stimulate the Activity of Affinity-purified Cyclosome-As noted above, although cyclin-ubiquitin ligase activity disappeared from the supernatants following adsorption of the cyclosome to p13 suc1 -Sepharose beads (Fig. 1), only 15-25% of the activity was recovered associated with beads, and around 10% of initial activity was recovered following elution with pNPP or pH 9 (Table I). We wondered, therefore, whether the enzyme was separated from a stimulatory factor during affinity purification. Indeed, we found that the addition of small amounts of the flow-through fraction to the affinity-purified enzyme strongly stimulated cyclin-ubiquitin ligation activity (Fig. 3A). The extent of the stimulation varied between 3-and 6-fold, depending on the preparation of affinity-purified cyclosome. Similar stimulation of activity by the flow-through fraction was observed with enzyme eluted from beads by pH 9 buffer (Fig. 3A) or with pNPP or phosphate (data not shown). When the flow-through fraction was subjected to gel filtration chromatography on Superose-6, stimulatory activity eluted in two peaks: a sharp peak at about 100 kDa and a higher molecular mass broad peak at about 400 -800 kDa (Fig. 3B). At least part of the higher molecular mass peak may be an aggregate of the lower molecular mass factor, since when the higher molecular mass peak was subjected to repeated separation on Superose-6 in the presence of 0.25 M KCl, part of activity was converted to the lower molecular mass form (data not shown).
The mode of action of the stimulatory factor from the flowthrough fraction is not known. It does not seem to be involved in the phosphorylation process responsible in the conversion of the interphase form of the cyclosome to the mitotic form (see "Discussion"). The purification and characterization of this factor is the subject of continued work. In the present study, the stimulatory activity of this factor was used for the detection of purified cyclosome at high sensitivity.
Purification of the Cyclosome by Affinity Chromatography and Gel Filtration-The purification obtained by the affinity procedure is summarized in Table III. Affinity chromatography was carried out as described under "Experimental Procedures," using elution at pH 9. Cyclin-ubiquitin ligase activity of the cyclosome was followed in the various fractions in the presence or absence of the flow-through stimulatory factor. In addition, the amounts of active protein kinase Cdc2-cyclin B (assayed as H1 kinase) and of total protein were also estimated in the various fractions. As expected, the flow-through fraction contained most of protein and lacked appreciable cyclin-ubiquitin ligase and H1 kinase activities. About 5% of the initial protein was bound to p13 suc1 -Sepharose beads (before elution at pH 9), along with 26% of the cyclin-ubiquitin ligase, and essentially all H1 kinase activities were recovered. Most of the protein bound to p13 suc1 -Sepharose is probably phosphorylated, since without preincubation with ATP and okadaic acid, only about 2% of the initial protein was bound (data not shown). Elution

TABLE II
Elution of cyclosome from p13 suc1 -Sepharose by phosphate compounds or high pH Salt extract of fraction 1 from M phase oocytes was preincubated with ATP as described under "Experimental Procedures." Samples of 300 g of protein were adsorbed onto 30 l of p13 suc1 -Sepharose beads, for 60 min at 0°C. Subsequently, beads were washed twice with 1-ml portions of buffer A, followed by two washes with buffer C (see "Experimental Procedures"). The beads were then suspended in 0.5-ml portions of the solutions indicated in the table (first column), which also contained 1 mg/ml rcm-BSA and 1 mM DTT. In experiment 1, all solutions contained 50 mM Tris-HCl, pH 7.2. In addition, solutions of potassium phosphate and p-nitrophenyl phosphate were also adjusted to pH 7.2. Elution was carried out at 0°C for 30 min, with stirring every 5 min. Subsequently, beads were separated from eluates by centrifugation (700 rpm, 3 min). The beads were further washed three times with 1-ml portions of buffer C and were suspended in 50 l of buffer C containing 2 mg/ml rcm-BSA. Eluates were concentrated with Centricon-30 concentrators, diluted 20-fold with buffer C, and concentrated again to 50 l. Cyclin-ubiquitin ligase activity associated with beads or eluates was expressed as the percentage of the initial activity of salt extract. with pH 9 buffer released most of the cyclin-ubiquitin ligase activity from p13 suc1 -Sepharose. However, only about one-third of the total protein that was adsorbed to the beads and a small fraction of the H1 kinase were eluted at pH 9, and the rest remained adsorbed to p13 suc1 -Sepharose beads. This relative selectivity in the elution of the cyclosome at pH 9 caused its enrichment in the eluate and its separation from most of protein kinase Cdc2-cyclin B. The recovery of cyclin-ubiquitin ligation activity in the pH 9 eluate was about 13% when assayed without the flow-through and 66% when assayed in the presence of the flow-through. Thus, the overall purification achieved by this procedure was around 30-fold when assayed in the presence of the flow-through stimulatory factor (Table III). It is notable that while the flow-through stimulated the activity of the cyclosome in the pH 9 eluate about 5-fold, it stimulated only slightly the activity of the enzyme bound to p13 suc1 beads (Table III and see "Discussion"). It is also noteworthy that when assayed in the presence of the flow-through, the amount of cyclin-ubiquitin ligation activity in the pH 9 eluate was 2-fold higher than that associated with the p13 suc1 -Sepharose beads prior to elution (Table III). It might be that the catalytic efficiency of immobilized enzyme associated with beads is lower than that of the free cyclosome in solution or that immobilized enzyme is not accessible to added flow-through factor.
To further purify the cyclosome, the affinity-purified preparation was subjected to gel filtration chromatography on Superose-6 (Fig. 4). The activity of cyclin-ubiquitin ligation was assayed in the presence or absence of the flow-through stimulatory factor. Following gel filtration, activity without the flowthrough was very low and was stimulated more than 10-fold by the flow-through (Fig. 4A). The elution position of affinitypurified cyclosome was similar to that observed previously with cruder preparations (1,9), with an apparent molecular size of about 1,500 kDa. This similarity in size indicates that most of the cyclosome is bound to p13 suc1 and is eluted as an entire complex, although the possible loss of some loosely bound subunits cannot be ruled out (see "Discussion"). The purity of the preparation was examined by silver staining. As expected, the pH 9 eluate (Fig. 4B, OR lane) contained numerous protein bands. However, many of these were removed from the region at the cyclosome by the gel filtration procedure. While the preparation following gel filtration is not homogenous, at least nine protein bands in the subunit region of 60 -200 kDa (marked by dots in Fig. 4B) coincided with cyclosome activity centered in fractions 22-24. Some of these protein bands may be subunits of the clam cyclosome. We could not determine the extent of further purification in the gel filtration procedure due to the very low amounts of material obtained in this step.
Selectivity of Action of Affinity-purified Cyclosome-It was important to examine whether the affinity-purified cyclosome preparation retains selectivity for the destruction box of cyclin, because a subunit of the complex responsible for the selectivity of cyclin recognition may be lost or inactivated during purification. For example, immunopurified APC/cyclosome from Xenopus eggs has lost part of its destruction box selectivity (2). In the experiment shown in Fig. 5, the action of affinity-purified cyclosome in the ligation to ubiquitin of different derivatives of cyclin was compared with that of a crude preparation. As may be seen, an N-terminal fragment of cyclin B consisting of amino acid residues 13-66 (that contains the destruction box; Ref. 4) was as effectively ligated to ubiquitin by the pure enzyme as by the crude preparation (Fig. 5, lanes 2 and 3). A similar construct, in which the RAAL sequence in the destruction box has been scrambled to AARL (17) is very poorly ligated to ubiquitin by both crude and pure cyclosome preparations (Fig. 5, lanes 5   FIG. 3. The activity of affinity-purified cyclosome is stimulated by factor(s) present in the fraction not adsorbed to p13 suc1 -Sepharose. Affinity purification was carried out as described under "Experimental Procedures," using pH 9 buffer for elution. A, influence of increasing concentrations of flow-through fraction. Cyclin-ubiquitin ligation was assayed with 0.5 l of affinity-purified cyclosome preparation in the presence of the indicated concentrations of the flow-through fraction. B, gel filtration chromatography of the flow-through fraction. The flow-through fraction (7.3 mg of protein) was separated on a Superose-6 column under conditions identical to those described for Fig. 2. Cyclin-ubiquitin ligation was determined with 0.5% of the column fractions in the presence of a constant amount affinity-purified cyclosome preparation. The amount of cyclin-ubiquitin ligation obtained by cyclosome alone (19.2 units) was subtracted from all results. The elution positions of marker proteins (kDa) are indicated by arrows. and 6). In the latter case, only a small amount of the monoubiquitinylated derivative was found, indicating a low affinity of the destruction box mutant for the enzyme, resulting in poor processivity in the addition of multiple ubiquitin molecules. The selectivity of affinity-purified cyclosome for destruction box-containing cyclins was further examined by the competi-tion of different unlabeled cyclin derivatives on the ligation to ubiquitin of 125 I-labeled cyclin B-(13-91) fragment (Fig. 5,  lanes 7-13). If the purified enzyme retains destruction box selectivity, it is expected that only cyclins that contain intact destruction box sequences would compete. It was indeed found that while unlabeled cyclin B-(13-66) fragment strongly inhibited cyclin-ubiquitin ligation by the purified cyclosome, the corresponding AARL destruction box mutant did not (Fig. 5,  lanes 9 and 10). Similarly, N-terminally truncated derivatives of cyclin B and cyclin A, which lack the destruction box regions, did not compete on cyclin-ubiquitin ligation by the purified enzyme (Fig. 5, lanes 11 and 13), while full-length cyclin A competed strongly (lane 12). These data indicate that the af-TABLE III Affinity purification of cyclosome on p13 suc1 -Sepharose Affinity purification was carried out as described under "Experimental Procedures," using elution with pH 9 buffer. Cyclin-ubiquitin ligase activity was determined in the absence or presence of flow-through fraction (100 g protein/ml). a A sample of washed beads was withdrawn before the elution step. b Protein in the pH 9 eluate could not be determined directly because of the presence of carrier STI (see "Experimental Procedures"). It was estimated by the decrease in protein associated with p13 suc1 beads following elution. The actual amount of protein recovery is probably lower, and thus the estimate of purification is a minimal value.

FIG. 4. Gel filtration chromatography of affinity-purified cyclosome.
A sample of affinity-purified cyclosome, prepared from 12 mg of protein of salt extract, was separated on Superose-6 under conditions similar to those described for Fig. 2, except that rcm-BSA was replaced by STI (0.2 mg/ml) as the carrier protein. A, cyclin-ubiquitin ligation activity was estimated in samples of column fractions in the presence (q) or absence (E) of 100 g/ml of flow-through fraction. The elution position of a 670-kDa marker protein is indicated by the arrow. B, samples of 10% of the column fractions were subjected to electrophoresis on an 8% polyacrylamide-SDS gel and stained with silver. OR, a sample of 0.5% of the original affinity-purified preparation before gel filtration. Numbers at the top indicate fraction numbers, and numbers on the right indicate the positions of molecular mass marker proteins (kDa). Protein bands that co-migrate with cyclosome activity (see Fig.  4A) are indicated by dots between fractions 20 and 21.
FIG. 5. Selectivity of affinity-purified cyclin-ubiquitin ligase for destruction box-containing mitotic cyclins. Cyclin-ubiquitin ligation was carried out as described under "Experimental Procedures," except that 125 I-labeled substrate was cyclin B-(13-66)/protein A in lanes 1-3 and its corresponding AARL destruction box mutant in lanes 4 -6. Where indicated, 2.5 g of protein of crude mitotic salt extract of fraction 1 (Crude) or 5 l of affinity-purified cyclosome preparation (Pure) were added as the source of enzyme. In lanes 9 -13, the indicated unlabeled cyclin derivatives were added at the following concentrations: 200 g/ml (lanes 9 and 10), 50 g/ml (lanes 11-13). Cyc, position of the free labeled cyclin fragment; Contam., contamination in the preparations of labeled cyclin fragments. finity-purified cyclosome retains selectivity for destruction boxcontaining cyclins, as observed previously with a partially purified preparation (1). DISCUSSION This study was initiated by the observation that the active form of the cyclosome binds strongly to p13 suc1 -Sepharose beads (Fig. 1). Since the best known property of p13 suc1 and of homologous Cks proteins is their ability to bind Cdks (22,23,29), we have first examined the possibility that cyclosome binding is mediated via Cdc2-cyclin B, which is both a substrate of the enzyme and its activator. We did not find evidence for such an indirect binding, since activated cyclosome bound to p13 suc1 -Sepharose following its separation from Cdc2-cyclin B (Fig. 2), and cyclosome from interphase extracts activated by GST-⌬88cyclin B bound well to p13 suc1 -Sepharose, but not to GSH-Sepharose (Table I). An alternative possibility, that the cyclosome is bound directly to the phosphate-binding site of p13 suc1 (27)(28)(29) is suggested by its elution by anions such as phosphate or sulfate and more effectively by the phosphate ester p-nitrophenyl phosphate (Table II). Since activation of the cyclosome is due to its phosphorylation (9,15), the latter suggestion provides a straightforward explanation for the observation that only the active form of the cyclosome binds to p13 suc1 -Sepharose. A possible role of the anionic-binding site of Cks proteins in their binding to phosphoproteins has been suggested (26,29), but we are not aware of any previous instance in which this has been shown to take place. It appears that many other phosphorylated proteins may also bind to p13 suc1 -Sepharose, since incubation of M phase extracts with ATP and okadaic acid, which promotes the accumulation of phosphorylated proteins, caused a more than 2-fold increase in the binding of total proteins to p13 suc1 -Sepharose beads (data not shown).
Although many different (presumably phosphorylated) proteins bind to p13 suc1 -Sepharose, considerable purification of the cyclosome is obtained by the present affinity procedure. This is aided by the fact that only a part of the proteins bound to p13 suc1 are eluted at pH 9, along with the cyclosome (Table III). Furthermore, the large size of the cyclosome allows its efficient separation from many other proteins in a subsequent step of gel filtration chromatography (Fig. 4). The final preparation, although not homogenous, appears to be highly purified (Fig.  4B). It does not contain significant amounts of Cdc2-cyclin B protein kinase activity, and no Cdk was detectable by immunoblotting with anti-PSTAIRE antibody (data not shown). It thus appears that this purification procedure may be valuable for future studies on the mode of action of the cyclosome and its regulation by protein kinase Cdc2-cyclin B.
In the course of this study, we have observed that the activity of affinity-purified preparations of the cyclosome is greatly stimulated by a factor abundant in the fraction not adsorbed to p13 suc1 -Sepharose. In gel filtration chromatography of this flow-through fraction, stimulatory activity eluted in two peaks, the higher molecular size of which may be an aggregate of the smaller factor (Fig. 3). At present, we do not know the mode of action of this stimulatory factor; its purification and characterization are being pursued in our laboratory. The stimulatory factor is apparently not a protein kinase based on the following observations: (a) the activity of the factor is not inhibited by staurosporin, an agent that inhibits completely the activation of the interphase form of the cyclosome by protein kinase Cdc2cyclin B; (b) the factor stimulates cyclosome activity also when ATP is replaced by the nonhydrolyzable ␤-␥ analog AMPPNP (data not shown). This ATP analog is not a substrate for protein kinases, but it can be used in the E1 reaction (which involves the scission of the ␣-␤ bond of ATP) that is needed for the cyclin-ubiquitin ligation assay. We have furthermore found that the activity of the stimulatory factor was not affected by phosphatase treatment, while affinity-purified cyclosome was inactivated by phosphatase treatment (data not shown), as observed previously with partially purified cyclosome (9). It thus seems that the stimulatory factor is not involved in the protein kinase pathway that converts the interphase form of the cyclosome to the mitotic phosphorylated form; rather, it stimulates the activity of phosphorylated cyclosome. The factor is also not involved in the selective recognition of destruction box-containing cyclins, since similar selectivity was observed in the absence of the factor (Fig. 5) as in its presence (data not shown). The stimulatory factor may be an easily dissociable subunit of the cyclosome, which is dissociated or inactivated during enzyme purification. Such a subunit may be present in both cyclosome-associated and free forms, as reported previously for some subunits of the cyclosome from fission yeast (16) and budding yeast (14). If the factor is a dissociable subunit, it may be still present in cyclosome associated with p13 suc1 -Sepharose, the activity of which is stimulated only slightly by the flow-through fraction (Table III). Alternatively, it is possible that cyclosome immobilized on beads is not accessible to added factor. Further work is required to characterize this stimulatory factor. In the present study, the use of the flowthrough factor was essential to detect highly purified cyclosome following gel filtration, and the activity was almost completely dependent on the addition of the stimulatory factor (Fig. 4A).
Although the present data bear directly only on the characterization of the binding of the cyclosome to p13 suc1 and its use for affinity purification, they may also give a clue as to the possible role of p13 suc1 in cyclin degradation. As noted above, genetic evidence in yeasts indicated that p13 suc1 and other homologous Cks proteins are required at multiple stages of the cell cycle, including exit from mitosis and the degradation of cyclin B (24,25). Biochemical experiments in immunodepleted extracts of Xenopus eggs also indicated that Cks is required for cyclin degradation (26), but these experiments could not distinguish whether Cks is required for the activity of this system or for the activation of the cyclin degradation machinery. We found no evidence to indicate that p13 suc1 is required for the activity of the mitotic, phosphorylated form of the cyclosome (data not shown), but it may be involved in the activation process. Since p13 suc1 can bind both protein kinase Cdc2 and the phosphorylated cyclosome, it is possible that it may direct the kinase to the phosphorylated cyclosome. Assuming that multiple phosphorylations are required for full activation of the cyclosome, initial phosphorylations may cause tighter binding of protein kinase Cdc2 to the cyclosome via p13 suc1 and thus may accelerate the rate of further phosphorylations. Such a model may also account, at least in part, for the lag kinetics of cyclosome activation by protein kinase Cdc2-cyclin B. Work in our laboratory is now aimed at examining the possible role of Suc1 protein in the regulation of cyclosome activity.