Substrate Recognition by the Cdh1 Destruction Box Receptor Is a General Requirement for APC/CCdh1-mediated Proteolysis*

The anaphase-promoting complex, or cyclosome (APC/C), is a ubiquitin ligase that selectively targets proteins for degradation in mitosis and the G1 phase and is an important component of the eukaryotic cell cycle control system. How the APC/C specifically recognizes its substrates is not fully understood. Although well characterized degron motifs such as the destruction box (D-box) and KEN-box are commonly found in APC/C substrates, many substrates apparently lack these motifs. A variety of alternative APC/C degrons have been reported, suggesting either that multiple modes of substrate recognition are possible or that our definitions of degron structure are incomplete. We used an in vivo yeast assay to compare the G1 degradation rate of 15 known substrates of the APC/C co-activator Cdh1 under normal conditions and conditions that impair binding of D-box, KEN-box, and the recently identified ABBA motif degrons to Cdh1. The D-box receptor was required for efficient proteolysis of all Cdh1 substrates, despite the absence of canonical D-boxes in many. In contrast, the KEN-box receptor was only required for normal proteolysis of a subset of substrates and the ABBA motif receptor for a single substrate in our system. Our results suggest that binding to the D-box receptor may be a shared requirement for recognition and processing of all Cdh1 substrates.

The anaphase-promoting complex, or cyclosome (APC/C), 2 polyubiquitylates numerous cell cycle regulatory proteins to target them for proteasomal degradation. In doing so, the APC/C triggers anaphase chromosome segregation, the inactivation of cyclin-dependent kinases and other mitotic kinases, cytokinesis, and establishment of a stable G 1 state (1)(2)(3). Proper coordination of late mitotic events by the APC/C is critical for faithful genome transmission during cell division (4 -6). How the APC/C selectively targets the appropriate substrates at the correct time to orchestrate cell cycle events is an important question and an active area of investigation (3,(7)(8)(9).
The APC/C co-activators Cdc20 and Cdh1 contribute to substrate recognition via docking sites on their conserved WD40 repeat domains (7,9). Two substrate degron motifs that bind distinct sites shared by both co-activators have been characterized in detail. The destruction box (D-box), originally identified in sea urchin cyclin B and subsequently found in many other APC/C substrates, has the consensus sequence RXXLXX(I/V/L)XN but is often found with alternative amino acids at the I/V/L and N positions (10 -12). The KEN-box was first identified as an APC/C degron in human Cdc20 (13) and has since been found in many other APC/C substrates (9). The location and structure of the D-box and KEN-box binding sites on Cdh1 and Cdc20 were recently revealed through x-ray diffraction and high resolution cryoelectron microscopy studies, providing detailed views of how common degrons are recognized by APC/C co-activators (12, 14 -16).
Despite these significant recent advances, our overall understanding of APC/C substrate recognition determinants is still incomplete. Although mutation of D-and KEN-box degrons stabilizes substrates in vivo, these motifs are often not sufficient for degradation, suggesting that other sequence elements are important as well (9,11). Moreover, the RXXL and KEN consensus sequences are found frequently in proteins that are not APC/C substrates and thus must exist in an appropriate structural context to function as degrons. Some APC/C substrates contain both D-and KEN-boxes, whereas others appear to contain only one or the other, and there are now many examples of substrates that lack functional consensus sequences for both (7,9). The source of binding specificity and affinity in substrates lacking one or both canonical degron motifs remains unknown in most cases. Finally, although Cdc20 and Cdh1 have very similar WD40 domain structures with nearly identical D-and KEN-box binding sites (12,14), they recognize distinct sets of substrates, suggesting again that additional factors influence specificity (17).
Many alternative APC/C degrons have been identified, often in substrates lacking consensus D-or KEN-boxes. These include GXEN in Xenopus Xkid (18), the A box or DAD motif (RXLXPSN) in Xenopus Aurora A (19,20), the O-box (PASPLTEKNAK, essential residues underlined) in Drosophila ORC1 (21), LXEXXXXN in Saccharomyces cerevisiae Spo13 (22), LLK in human Claspin (23), the CRY-box in mouse Cdc20 (24), NKSEN in budding yeast Sgo1 (25), and the ABBA motif, FX(I/V/L)(F/Y/H)X(D/E), in human cyclin A, Bub1, and budding yeast Clb5 (26,27). Only the ABBA motif has a known binding site on the co-activator WD40 domains (12,28). Some alternative degrons resemble D-and KEN-boxes, including RQLF in budding yeast Ase1 (29) and GXEN and NKSEN in XKid and Sgo1, respectively. Furthermore, a KXXL sequence in budding yeast Cin8, the LXEXXXXN degron in Spo13, and the Orc1 O-box can bind the D-box docking site on Cdh1 to some extent, and XKid GXEN can similarly bind the KEN-box receptor site (12), suggesting that they are functional variants of these consensus degrons. Many of the alternative sequences, however, appear distinct from the canonical D-and KEN-box degrons. Where they bind APC/C co-activators or other subunits and how they function to promote degradation are largely unknown. Some substrates still have undefined degrons, including several (e.g. budding yeast Iqg1 (30), Cik1 (31), and Kip1 (32) and human Usp1 (33)) for which extended regions have been identified as sufficient and/or necessary for degradation but that lack consensus D-and KEN-box motifs. Nuclear localization also appears to be required for degradation of APC/C substrates in budding yeast (34), and therefore some degrons may act by promoting nuclear import and not by binding the APC/C or its co-activators.
The lack of a common degron in all APC/C substrates and the diversity of reported degrons raise the question of whether any universal determinants of substrate recognition exist. One explanation for the diversity in substrate degrons is that different substrate classes bind independent sites on the co-activators and/or core APC/C. If so, this would provide a potential mechanism for differential regulation of substrate binding and processing. There is some precedent for this idea. The Cdc20 targets cyclin A and Nek2A have alternative mechanisms for recognition by APC/C, allowing them to be degraded in prometaphase when the mitotic checkpoint complex is active and blocks degradation of securin and cyclin B, which rely on canonical degrons (35,36). Here, we used an in vivo substrate degradation assay in budding yeast to test for distinct substrate recognition modes for the Cdh1 co-activator and to explore the relative contributions of the known degron binding sites to Cdh1 substrate proteolysis. Our results argue against completely independent substrate binding modes and suggest that most, if not all, Cdh1 substrates require a functional D-box receptor on APC/C Cdh1 for efficient proteolysis, despite the common lack of canonical D-box motifs.

Acm1 Inhibits Proteolysis of Diverse Cdh1 Substrates-The
Acm1 protein is a pseudosubstrate inhibitor of APC/C Cdh1 that acts by competitively blocking substrate binding to Cdh1 using canonical D-and KEN-box degron motifs and the recently identified ABBA motif (originally called the A-motif in Acm1) (37)(38)(39). Acm1 has only been shown to inhibit the ubiquitylation and degradation of a small number of classical APC/C Cdh1 substrates that contain well defined KEN-and/or D-box degrons (12,(37)(38)(39)(40). It remains unclear whether Acm1 can also inhibit degradation of substrates with non-canonical degrons that lack consensus KEN-and/or D-box motifs. To answer this question, we established an in vivo assay in which substrate stability can be compared in the presence and absence of Acm1 (Fig. 1A). We used yeast cells harboring plasmids that conditionally express protein A-tagged substrates from the galactose-inducible GAL1 promoter and a stabilized Acm1 mutant (Acm1⌬52), which retains its central Cdh1 inhibitory domain (41), from the methionine-repressible MET25 promoter. The stable Acm1⌬52 mutant was necessary because Acm1 itself is highly unstable in the G 1 phase because of an APC/C-independent proteolytic mechanism (41). The cells were first arrested in the G 1 phase, when APC/C Cdh1 is active, using ␣-factor pheromone. Substrate expression was then induced briefly with galactose either in the absence or presence of methionine to control Acm1⌬52 expression. Finally, substrate expression was terminated with glucose and cycloheximide, and stability was monitored over time by anti-protein A immunoblotting.
We tested a panel of 15 known Cdh1 substrates (Table 1), including proteins containing functional KEN-and D-box consensus motifs, either a KEN-or D-box but not both, a noncanonical degron motif without consensus KEN-or D-boxes, or otherwise undefined degrons. The presence of Acm1 strongly stabilized all Cdh1 substrates, regardless of degron type (Fig. 1, B and C, and Table 1).
This result could be explained either by Acm1 having a general effect on APC/C catalytic function independent of competing for substrate binding or by all substrates having one or more common or overlapping docking sites required for ubiquitylation by APC/C. Currently, there is no evidence that Acm1 directly inhibits the catalytic activity of APC/C toward substrates. Instead, current models for Acm1 action suggest that it acts as a competitive inhibitor of substrate binding (38,40,42,43). However, we cannot rule out general inhibition of the catalytic function of APC/C by Acm1 in addition to its effect on substrate binding.
The ABBA Motif in Acm1 Is a Functional APC/C Cdh1 Degron-To independently test whether diverse Cdh1 substrates share a common docking site, we engineered yeast strains expressing cdh1 mutant alleles with altered degron binding sites to compare the stability of the same substrate collection with a wild-type CDH1 strain. Docking sites for the KEN-box, D-box, and ABBA motif have been identified through high resolution structural studies (12, 14 -16, 28, 44), although the ABBA motif has not been demonstrated to be a functional in vivo degron for APC/C Cdh1 as it is for APC/C Cdc20 (26,27). It was first identified as an element in Acm1 required for high affinity binding and inhibition of budding yeast APC/ C Cdh1 (38,40). Replacement of amino acids in all three of these Cdh1 docking sites disrupts binding of their respective ligand motifs in vitro (12,14,45). For strain validation we monitored endogenous Clb2 levels in cycling and G 1 -arrested cells. Clb2 remained stable in the G 1 phase in the cdh1 dbr and cdh1 kbr strains containing mutations in the D-box receptor and KENbox receptor, respectively, but was degraded in the cdh1 amr strain containing mutations in the ABBA motif receptor ( Fig.  2A), consistent with the presence of functional D-and KENboxes in Clb2 that direct its APC/C-mediated degradation (46,47). Stabilization of endogenous Clb2 was not due to differences in expression level of the mutant proteins (Fig. 2B) or efficiency of G 1 arrest (Fig. 2C). The receptor mutations also did not disrupt overall Cdh1 structure because the Cdh1 dbr and Cdh1 kbr proteins still interacted by co-immunopurification (co-IP) with Acm1 via its other interaction motifs (Fig. 2D). All three receptor mutants also interacted with the core APC/C similar to wild-type Cdh1 (Fig. 2E).
We tested whether the Acm1 ABBA motif is a functional APC/C Cdh1 degron using the same assay from Fig. 1, except with Acm1⌬52 expressed from the GAL1 promoter as the substrate. Paradoxically, Acm1 can be converted into an APC/ C Cdh1 substrate in vitro by mutation of its central KEN-and D-box motifs (37,38,40). Consistent with these in vitro results, mutation of the D-and KEN-box motifs in Acm1⌬52 (Acm1 db/kb ) destabilized the protein in vivo, dependent on Cdh1 (Fig. 2F). Additional mutation of ABBA motif residues to create an Acm1 db/kb/am triple mutant stabilized the protein (Fig. 2G). In the cdh1 amr strain Acm1⌬52 db/kb was strongly stabilized compared with a wild-type CDH1 strain and was more stable than in Time ( -Acm1 + Acm1 -Acm1 + Acm1 A FIGURE 1. Acm1 blocks in vivo proteolysis of diverse Cdh1 substrates. A, flow chart for the in vivo substrate stability assay. The methionine-repressible Acm1⌬52 expression is only relevant to Fig. 1 and is omitted from the assays used in subsequent figures. Control cells lacked the Acm1⌬52 expression plasmid to establish basal instability of substrates. CHX, cycloheximide. See "Experimental Procedures" for more details. B and C, stability of the representative Cdh1 substrates that either contain (B) or lack (C) canonical D-box and/or KEN-box degrons, expressed with C-terminal protein A epitope tags, was monitored over time by anti-protein A immunoblot in the absence or presence of the Acm1⌬52-protein A fusion protein (labeled as Acm1 in all panels). Time 0 is the point where protein expression is terminated. Both substrate and Acm1 were detected with anti-protein A antibody. G6PD is a load control in all panels. When present, an asterisk indicates that two different exposures of the same blot are compared, separated by a white bar, so that the time 0 signals are of comparable intensity. This sometimes provides a better visualization of differences in substrate half-lives between the two conditions because stabilization can lead to increased steady-state protein levels. See Table 1 for results from complete substrate set. All experiments were performed three times with equivalent results.
the cdh1 dbr and cdh1 kbr strains (Fig. 2H). Taken together, these results demonstrate that the ABBA motif in Acm1 is a functional APC/C Cdh1 degron. With this information in hand, we proceeded to test the relative contributions of the three degron docking sites to Cdh1 substrate degradation.

The D-box Receptor on Cdh1 Is Required for Efficient Proteolysis of Most, if Not All, Cdh1
Substrates-Strikingly, mutation of the D-box receptor site on Cdh1 by alteration of just 4 conserved amino acids (40,45) strongly stabilized all 15 Cdh1 substrates in our in vivo stability assay ( Fig. 3 and Table 1), although not always to the same degree as Acm1⌬52 overexpression. This result was surprising in that roughly half of the substrates tested lack recognizable, consensus D-box motifs. The variability in the stabilizing effect across substrates could reflect different contributions of other elements to Cdh1 binding or differences in the nature of interactions with the D-box receptor region of Cdh1.

KEN-Box and ABBA Motif Receptors Are Required for Normal Proteolysis of Distinct Subsets of Cdh1
Substrates-We next explored the requirement for functional KEN-box and ABBA motif receptor sites on Cdh1 for substrate proteolysis. Unlike the universal dependence of normal proteolysis rates on the D-box receptor, the proteolysis of some substrates appeared completely independent of the KEN-box receptor in the cdh1 kbr strain ( Fig. 4A and Table 1). Nonetheless, proteolysis of a subset of substrates was strongly impaired in cdh1 kbr cells (Fig.  4, B and C, and Table 1). For the most part, substrates exhibiting dependence on the KEN-box receptor contained consensus KEN-box motifs. In some cases, like Spo12, KEN sequences have not been functionally linked to APC/C-mediated proteolysis. The Spo13 and Sgo1 proteins were interesting in that proteolysis was strongly dependent on the KEN-box receptor, yet neither contains a consensus KEN-box motif. This suggests that variations on this motif can also functionally engage the KEN-box docking site on Cdh1, consistent with experiments using the GXEN motif in XKid (12). The Kip1 protein also lacks a KEN-box in the C-terminal region required for APC/C-mediated proteolysis (32) but requires the KEN-box receptor for efficient proteolysis in our assay.
We tested a subset of our Cdh1 substrate panel in the cdh1 amr strain, including those that showed no dependence on the KEN-box receptor, and found that the stability of most was unaffected by disruption of the ABBA motif binding site ( Fig.  5A and Table 1). Several substrates exhibited a very slight increase in stability that could reflect a minor overall effect of ABBA motif receptor mutations on Cdh1 function or a weak contribution to substrate binding but did not meet our criteria for stabilization. We found a single Cdh1 substrate, Cik1, whose stability was strongly dependent on the ABBA motif receptor site (Fig. 5B), even though the Cik1 N terminus that is necessary and sufficient for APC/C-mediated proteolysis (31) lacks any sequence resembling the ABBA motif consensus. Thus, in addition to the known function in promoting binding of the inhibitor Acm1 to Cdh1, the ABBA motif (or variants of it) appears to be capable of promoting substrate recognition by budding yeast APC/C Cdh1 .
Novel APC/C Substrates Can Be Identified Using in Vivo Stability Assays and Cdh1 Inhibition by Acm1-The universal stabilizing effect of Acm1⌬52 on Cdh1 substrates in our assay in Fig. 1 suggested that it could be used as a tool for identifying novel substrates. To explore this idea, we tested two proteins, Ipl1 and Swe1, whose abundances during the cell cycle were previously demonstrated to be dependent on APC/C function (48) but, to our knowledge, have not been validated as direct APC/C substrates. Ipl1 is the budding yeast ortholog of Aurora B kinase, and Swe1 is the ortholog of Wee1 kinase. The stability of both proteins in G 1 -arrested cells increased when Acm1⌬52 was co-expressed (Fig. 6A). To verify that APC/C can target Ipl1 a Sources providing evidence for functional degrons or degron-containing regions are included after each substrate name. Underlining indicates substrates that lack a consensus D-box motif required for their APC/C-mediated proteolysis. b Only degrons with experimental evidence demonstrating their requirement for substrate degradation are noted with ϫ. c Stabilization (denoted with ϫ) is defined as a minimum of 30-min increase in the time required to reach 50% of the initial protein level compared with control. The Acm1 column reports results of assays measuring inhibition of protein degradation by the Acm1⌬52 protein. The last three columns report results of assays measuring dependence of substrate degradation on the D-box, KEN-box, and ABBA motif receptor sites of the Cdh1 WD40 domain. d RQLF sequence in Ase1 was reported as a D-box but does not match the RXXL D-box consensus motif. e NT, not tested. f The N-terminal 80 amino acids of Cik1 are necessary and sufficient for degradation but lack consensus degron motifs. g The N-terminal 42 amino acids of Iqg1 are required for degradation but lack consensus degron motifs. h The C-terminal 101 amino acids of Kip1 are necessary and sufficient for degradation but lack consensus degron motifs. i Spo12 contains consensus D-box and KEN-box motifs, but mutation of both did not prevent degradation.
and Swe1 for degradation we compared their stability in CDH1 and cdh1 dbr strains. Stability of both Ipl1 and Swe1 increased in the cdh1 dbr strain, confirming their APC/C Cdh1 -mediated turnover and demonstrating its D-box receptor dependence (Fig.  6B). Although it remains unclear whether APC/C-mediated degradation of Ipl1 and Swe1 has functional significance in yeast, this work demonstrates that, in principle, our in vivo assay can be used as a tool to determine whether a protein of interest is recognized as a substrate by APC/C Cdh1 in a physiological environment. Given the apparent conservation of APC/C structure and substrate recognition (16), it could potentially be used to test candidate APC/C substrates from other species as well.

Discussion
The most significant conclusion from this study is that efficient degradation of most, if not all, APC/C Cdh1 substrates requires engagement of the D-box receptor site on Cdh1, despite the lack of a universal D-box sequence motif common to all substrates. Thus, the apparent diversity in APC/C Cdh1 substrate degron motifs may not reflect complexity in the mechanisms of substrate recognition. The D-box was the first APC/C degron identified and characterized (10), and it contains elements that bind both to the WD40 domain of the coactivator proteins and to the Apc10 subunit of the core APC/C (44, 45, 49 -54). The broad requirement of the D-box receptor for efficient APC/C Cdh1 substrate proteolysis suggests that the D-box may contribute more than just binding affinity, consistent with recent structural and mechanistic studies. D-box engagement may be important for positioning substrate lysines for ubiquitin transfer, for the co-activator-induced conformational change in the core APC/C that brings the catalytic RING module into close proximity with the substrate (16), for enhancing functional interaction with the E2 (55), or for enhancing catalysis in some other way. Despite the universal The ABBA motif is a functional APC/C Cdh1 degron in budding yeast. A, immunoblot analysis of endogenous Clb2 levels in asynchronous and ␣-factor-treated G 1 cultures of the indicated strains. WT is the parental strain from which cdh1⌬ was engineered. The remaining strains have 3FLAG-CDH1 or degron receptor mutant alleles (dbr, D-box receptor mutations; kbr, KEN-box receptor mutations; amr, ABBA motif receptor mutations) expressed from the CDH1 promoter integrated into cdh1⌬. B, anti-FLAG immunoblot from whole cell extracts of asynchronous cultures of cdh1⌬ and the indicated derivative strains demonstrating equivalent expression from the integrated 3FLAG-cdh1 alleles. C, the ability of the integrated 3FLAG-CDH1, 3FLAG-cdh1 dbr , and 3FLAG-cdh1 kbr strains used for substrate stability assays to stably arrest in the G 1 phase in response to ␣-factor treatment was determined by anti-Acm1 immunoblot and flow cytometry. Acm1 is highly unstable specifically in G 1 cells. Acm1 expression from GAL1p was terminated at time 0. Flow cytometry data are from 60-min point to illustrate maintenance of arrest throughout the experiment. D, the interaction of endogenous Acm1 with 3FLAG-Cdh1 dbr and 3FLAG-Cdh1 kbr mutants was assessed by anti-FLAG co-IP. WCE, whole cell extract. E, the interaction of 3FLAG-Cdh1 and the three degron docking site mutants with the core APC was assessed by anti-FLAG co-IP. NC, negative control using parental strain lacking 3FLAG-Cdh1. F, stability of Acm1⌬52-protein A and a variant with mutations in the central D-and KEN-boxes (Acm1 db/kb ) was evaluated in CDH1 and cdh1⌬ strain backgrounds. G, same as F comparing stability of Acm1 db/kb with a variant containing an additional mutation of the ABBA motif (Acm1 db/kb/am ) in the CDH1 strain. H, stability of Acm1⌬52-protein A with mutations in its D-and KEN-boxes was compared in strains expressing either wild-type CDH1 or one of three cdh1 degron receptor mutants. In this experiment Acm1 was detected with an anti-Acm1 antibody, and the load control is a nonspecific band from this blot.
dependence of Cdh1 substrate degradation on an intact D-box receptor, we note that turnover of the artificial Acm1 db/kb mutant was only slightly impaired in the cdh1 dbr strain (Fig.   2H), presumably because of the unusual contributions of other high affinity docking motifs required for effective Cdh1 inhibition (12,38,40).   Table 1 for results from complete substrate set. All experiments in A and B were performed three times with equivalent results. C, example flow cytometry results from cells at the 60-min point in the assays (Pds1 experiment shown) demonstrating maintenance of G 1 arrest throughout time course.   Table 1 for results from complete substrate set. Three independent experiments were performed for all substrates with equivalent results. D, example flow cytometry results from cells at the 60-min point (Cdc5 experiment shown) demonstrating maintenance of G 1 arrest throughout time course.
In contrast to the D-box, the selective requirement of other motifs, e.g. the KEN-box and ABBA motif, for substrate proteolysis is consistent with these motifs providing primarily specificity and binding affinity and not a crucial contribution to the catalytic reaction. Although it is unclear what the functional implications of having a KEN box versus an ABBA motif are, the use of a variety of short linear motifs for substrate docking ensures highly selective binding while allowing for flexibility in the substrate structural requirements. It seems likely that additional docking sites on Cdh1 remain undiscovered. For example, it is unclear how a D-box alone could provide sufficient binding affinity and specificity compared with substrates that use multiple degrons motifs, and degradation of several of the substrates examined here (e.g. Nrm1, Mps1, and Cdc20) was independent of both the KEN-box and ABBA motif receptors. Nonetheless, the widespread use of D-box receptor docking argues that all Cdh1 substrates share some aspects of recognition and processing by APC/C Cdh1 .
The ABBA motif was originally identified in Acm1 as a unique element required for full APC/C Cdh1 inhibition (40). Our results with Cik1 and the Acm1 db/kb mutant demonstrate clearly that the ABBA motif is another functional APC/C Cdh1 degron, consistent with recent reports identifying the ABBA motif as a degron for APC/C Cdc20 in both yeast and humans (26,27) and with prior in vitro work with Acm1 (40). Considering that Acm1 possesses three distinct consensus APC/C Cdh1 degrons, it remains unclear how it evades ubiquitination by APC/C Cdh1 , although a lack of suitably positioned lysines has been proposed (40). The ABBA motif that binds budding yeast Cdc20, identified in the S phase cyclin Clb5 (27), is distinct from the Acm1 ABBA motif that binds Cdh1. The two differ primarily at the beginning of the motif, where Clb5 orthologs contain a pair of conserved basic amino acids instead of the conserved phenylalanine of the ABBA motif consensus sequence FX(I/V/ L)(F/Y/H)X(D/E) (27). In principle, differences in the ABBA motif receptor of Cdh1 and Cdc20 could contribute to the distinct specificities of the two APC/C co-activators, although it does not seem to contribute significantly to recognition of most Cdh1 substrates. Interestingly, the ABBA motif recognized by human APC/C Cdc20 more closely resembles the motif recognized by APC/C Cdh1 in budding yeast, and the residues in human Cdc20 that form the ABBA motif receptor are likewise similar to those found in budding yeast Cdh1 (28).
Our results emphasize that the current definitions of the three best characterized APC/C substrate degrons (D-box, KEN-box, and ABBA motif) are still incomplete. We identified substrates that require the D-box, KEN-box, and ABBA motif receptors yet lack sequences that match the current consensuses for these degrons. Additional work is needed to identify the functional degrons in many substrates lacking the consensus motifs, for example the D-box in the kinesin Cin8, the Dand KEN-boxes in the kinesin Kip1, the D-box and ABBA motif in Cik1, and the KEN-box in Spo13. The NKSEN sequence in Sgo1 that is required for its APC/C-mediated proteolysis was reported to be part of an unconventional D-box (25)  -Acm1 + Acm1 -Acm1 + Acm1 FIGURE 6. Novel APC/C substrates can be identified using the in vivo Cdh1 inhibition assay. A, same as Fig. 1 with candidate Cdh1 substrates Ipl1 and Swe1. B, same as Fig. 3 (A and B).
viously reported to have no effect on Spo12 turnover (56). However, our results clearly demonstrate that Spo12 degradation by the APC/C is dependent on functional D-and KEN-box receptor sites, suggesting either that mutation of Spo12 degrons renders the protein unstable independent of the APC/C or that cryptic D-and KEN-boxes exist. Additional structural studies will be crucial in the future to characterize how consensus and non-consensus degron sequences can bind the same receptor sites to promote substrate ubiquitylation, and it still remains unclear which docking sites some of the alternative degrons reported for APC/C substrates bind to and which may contribute to degradation in other ways aside from binding the co-activator, for example by promoting nuclear localization (34). Given the prevalence of unconventional degron motifs in Cdh1 substrates from other species and the high conservation of APC/C core and co-activator subunits, our findings likely have general relevance to APC/C substrate recognition in all eukaryotes.
BG1805-based plasmids expressing APC/C substrates with C-terminal protein A tags are from the Yeast ORF Collection (GE Dharmacon; YSC3868). For plasmids expressing ASE1-TAP, SPO12-TAP, and SPO13-TAP (pHLP520, pHLP521, and pHLP522), the coding regions were amplified by PCR from genomic DNA and inserted into the Gateway TM entry vector pENTR/D-TOPO (Life Technologies; K240020) as directed by the manufacturer. The genes were then transferred into destination vector pAG416GAL-ccdB-TAP (Addgene; catalog no. 14267) (57). pHLP523 was constructed by moving an XbaI and XhoI fragment containing the acm1⌬52-protein A fusion from pHLP400 described previously (41) into p415MET for control by the methionine-repressible MET25 promoter. pHIP032 and pHIP042 were constructed previously (42).
Cell Growth-YP medium contained 1% yeast extract and 2% peptone and was supplemented with 2% of the appropriate carbon source (either dextrose, raffinose, or galactose). Synthetic dropout medium contained 6.7 g/liter yeast nitrogen base, the appropriate amino acid supplement mixture for maintaining plasmid selection or inducing expression from MET25p, and 2% of the appropriate carbon source. All liquid yeast cultures were grown at 30°C with shaking at 225 rpm. For G 1 arrest, ␣-factor was added to 100 g/L for 2-3 h until a uniform "schmoo" morphology was observed by microscopy. All cell cycle arrests were confirmed by flow cytometry on an Accuri C6 flow cytometer as described previously (43). All strains are derived from BY4741: MATa his3⌬1 leu2⌬0 met15⌬0 ura3⌬0. YKA898 and YKA1008 differ in the nature of the cdh1 amr mutation as described under "Experimental Procedures." YKA898 was used in the experiment in Fig. 2H  In Vivo Protein Stability Assay-Yeast were cultured at 30°C overnight until saturated in the appropriate selective synthetic dropout medium for plasmid maintenance. Saturated cultures were diluted to A 600 Ϸ 0.1 in YP with 2% raffinose. Growth was continued to A 600 Ϸ 0.2-0.3 and then ␣-factor was added. For the Acm1⌬52 inhibition assay, arrested cells were pelleted by centrifugation at 5,000 ϫ g for 3 min, washed with 10 ml of water, and pelleted again. The cell pellet was resuspended in synthetic dropout medium lacking methionine (to induce Acm1⌬52 expression) and containing 2% raffinose as carbon source and ␣-factor to maintain the arrest. After 30 min, galactose was added to 2% to induce expression of APC/C substrates. For assays in strains expressing cdh1 degron receptor mutants, no medium change was needed following G 1 arrest. In both experimental workflows, glucose (2%) and cycloheximide (0.5 mg/ml) were added to terminate expression after 45 min. 10-ml cell aliquots were harvested at regular intervals, collected by centrifugation, washed with 10% trichloroacetic acid, and pellets frozen at Ϫ80°C until processing for immunoblotting.
Exposures were selected with equivalent initial intensities (time ϭ 0), and the remaining protein at each subsequent time point was calculated as a percentage of the initial value after normalizing to the G6PDH load control signal. We considered substrates stabilized if it took a minimum of 30 min longer for the immunoblot signal to be reduced to less than 50% of the initial value relative to the control in all trials.
Co-immunopurification-Strains YKA412, YKA627, YKA628, YKA897, and YKA1006 were cultured overnight in 5 ml of synthetic dropout medium lacking histidine then back-diluted into 500 ml and either grown to A 600 Ϸ 0.3 and arrested in the G 1 phase with ␣-factor (for co-IP analysis of Cdh1 interaction with core APC/C) or to A 600 Ϸ 0.6 without G 1 arrest (for co-IP of Cdh1 with Acm1). The cells were pelleted, washed once with 1 ml water, and resuspended in 10 volumes of buffer L (25 mM HEPES, pH 7.5, 100 mM sodium acetate, 10% glycerol, 0.1% Triton X-100, 0.5 mM dithiothreitol) supplemented with 1 mM phenylmethylsulfonyl fluoride, 1 M pepstatin, 100 M leupeptin, and 5 mM EDTA and then distributed into 1.5-ml microcentrifuge tubes (maximum 900 l/tube). Acid-washed 0.5-mm glass beads (Ϸ200 l, Biospec Products) were added to each tube, and cells were agitated at 4°C using a Disruptor Genie (Scientific Industries) until Ͼ75% cell lysis was achieved (by microscopic inspection). Extracts were clarified by centrifugation at 16,000 ϫ g for 30 min and the supernatant was incubated with 20 l of anti-FLAG resin (equilibrated first in Buffer L) with gentle rotation for 2 h at 4°C. The resin was washed with 10 ml of buffer L in a 15-ml conical tube four times for 5 min each, transferred to a lowbind microtube and washed two more times with 1 ml of buffer L. Bound protein was eluted twice by incubating with 200 g/ml 3xFLAG peptide in 40 l of buffer L on a rotating ZZ is a domain of protein A and was used for immunoblot detection in this study. 3C indicates a protease 3C cleavage site, HA indicates the hemagglutinin epitope, and His 6 indicates a hexahistidine tag, none of which were used in this study. TAP is the tandem affinity purification tag, consisting of protein A and calmodulin binding domains, of which the protein A portion was used for immunoblot detection. GE indicates that all BG1805 plasmids are part of the yeast ORF collection available from GE Dharmacon. platform for 20 min at room temperature. The pooled eluates were then analyzed by SDS-PAGE and immunoblotting.
Author Contributions-L. Q. designed and conducted most of the experiments, analyzed the results, and, with M. C. H., wrote and edited the manuscript. D. S. P. S. F. G. conducted experiments characterizing the ABBA motif as a functional degron and analyzed the results. M. M. helped establish the assay system for studying Acm1 inhibition of Cdh1 substrate degradation. M. C. H. designed experiments, analyzed data, and wrote the manuscript.