Regulation of a Spindle Positioning Factor at Kinetochores by SUMO-Targeted Ubiquitin Ligases

Article

Regulation of a Spindle Positioning Factor at Kinetochores by SUMO-Targeted Ubiquitin Ligases Graphical Abstract

Authors Jo¨rg Schweiggert, Lea Stevermann, Davide Panigada, Daniel Kammerer, Dimitris Liakopoulos

Correspondence [email protected]

In Brief Schweiggert et al. show that the yeast spindle positioning factor Kar9, which is known to link astral microtubules to the cytoskeleton, also localizes to kinetochore microtubules and is targeted for degradation by SUMO-targeted ubiquitin ligases (STUbLs) at kinetochores. Surprisingly, although SUMO-dependent, STUbL recruitment to kinetochores is independent of Kar9 sumoylation.

Highlights d

Kar9 shuttles between cytoplasm and nucleus

d

STUbLs localize to kinetochores in a SUMO-dependent manner

d

STUbLs ubiquitylate Kar9 at kinetochores independently of Kar9 sumoylation

d

This process ensures spindle positioning and chromosome transmission fidelity

Schweiggert et al., 2016, Developmental Cell 36, 415–427 February 22, 2016 ª2016 Elsevier Inc. http://dx.doi.org/10.1016/j.devcel.2016.01.011

Developmental Cell

Article Regulation of a Spindle Positioning Factor at Kinetochores by SUMO-Targeted Ubiquitin Ligases Jo¨rg Schweiggert,1,2,3 Lea Stevermann,2,3 Davide Panigada,1 Daniel Kammerer,2 and Dimitris Liakopoulos1,2,3,* 1Centre de Recherche de Biochimie Macromole ´ culaire (CRBM), CNRS UMR 5237, 1919 route de Mende, 34293 Montpellier Cedex 05, France 2Biochemistry

Centre Heidelberg (BZH), INF 328, 69120 Heidelberg, Germany of Molecular and Cellular Biology, University of Heidelberg, 69117 Heidelberg,

3The Hartmut Hoffmann-Berling International Graduate School

Germany *Correspondence: [email protected] http://dx.doi.org/10.1016/j.devcel.2016.01.011

SUMMARY

Correct function of the mitotic spindle requires balanced interplay of kinetochore and astral microtubules that mediate chromosome segregation and spindle positioning, respectively. Errors therein can cause severe defects ranging from aneuploidy to developmental disorders. Here, we describe a protein degradation pathway that functionally links astral microtubules to kinetochores via regulation of a microtubule-associated factor. We show that the yeast spindle positioning protein Kar9 localizes not only to astral but also to kinetochore microtubules, where it becomes targeted for proteasomal degradation by the SUMO-targeted ubiquitin ligases (STUbLs) Slx5-Slx8. Intriguingly, this process does not depend on preceding sumoylation of Kar9 but rather requires SUMO-dependent recruitment of STUbLs to kinetochores. Failure to degrade Kar9 leads to defects in both chromosome segregation and spindle positioning. We propose that kinetochores serve as platforms to recruit STUbLs in a SUMO-dependent manner in order to ensure correct spindle function by regulating levels of microtubuleassociated proteins. INTRODUCTION Correct outcome of cell division requires a coordinated interplay between different domains of the mitotic spindle. While kinetochore microtubules capture and link sister chromatids to spindle poles, astral microtubules interact with cortical factors to generate forces that place the spindle in its proper position with respect to cellular geometry and polarity. Both chromosome capture and spindle positioning depend on a multitude of proteins that directly or indirectly associate with microtubules (microtubule-associated proteins [MAPs]). Some MAPs associate with both spindle domains, where they mediate connection of microtubules to kinetochores and the cell cortex but also regulate microtubule dynamics by promoting or suppressing microtubule stability (Galjart, 2005; Hildebrandt and Hoyt, 2000). In addition, a number of MAPs function specifically either in spindle

positioning or in chromosome segregation and thus associate only with astral or kinetochore microtubules, respectively. In both cases, control of the relative abundance and the correct distribution of MAPs is crucial, since imbalances therein can lead to errors in chromosome segregation and the geometry of cell division. In higher eukaryotic cells, dynein is cortically anchored and positions the spindle by pulling on astral microtubules (reviewed in Stevermann and Liakopoulos, 2012). A second mechanism requires interaction of astral microtubules with cortical actin and involves linker proteins such as the tumor suppressor adenomatous polyposis coli (APC; Caldwell et al., 2007; Fleming et al., 2009). Apart from its function in spindle positioning, APC regulates cell proliferation but, intriguingly, localizes also to kinetochores and has been shown to be required for chromosome transmission fidelity (Fodde et al., 2001; Kaplan et al., 2001). Both a dynein- and an actin-dependent pathway orientate the spindle in the asymmetrically dividing unicellular organism Saccharomyces cerevisiae (Eshel et al., 1993; Miller and Rose, 1998). While dynein forces become predominant after anaphase onset, the actin-dependent pathway acts from G1 to metaphase (Palmer et al., 1992). Herein, the protein Kar9 connects astral microtubules (in yeast called cytoplasmic microtubules [cMTs] due to the absence of nuclear envelope breakdown in mitosis) to actin cables by forming a complex with the plus-end tracking protein Bim1 (homolog of human EB1) and myosin V (Korinek et al., 2000; Lee et al., 2000; Miller et al., 2000; Yin et al., 2000). Consequently, forces generated by myosin V pull the spindle to the predetermined site of cell cleavage (i.e., the bud neck). Phosphorylation by Cdk1 and Dbf2/Dbf20, sumoylation, as well as yet unidentified signals from the kinetochore, limit loading of Kar9 to cMTs of the old spindle pole to allow concomitant spindle alignment with the mother-daughter axis (Hotz et al., 2012; Liakopoulos et al., 2003; Maekawa et al., 2003). In addition, ubiquitylation of Kar9 by the ubiquitin E2 enzymes Ubc1-Ubc4 leads to subsequent proteasomal degradation of the protein and seems to affect spindle positioning (Kammerer et al., 2010). However, the mechanisms linking the latter process to Kar9 proteolysis remain unidentified. SUMO-targeted ubiquitin ligases (STUbLs) are a conserved class of ubiquitin E3 ligases (reviewed in Sriramachandran and Dohmen, 2014) that have been shown to specifically recognize sumoylated proteins via SUMO interaction motifs (SIMs). They are required in many cellular processes including DNA damage repair and transcription (van Hagen et al., 2009; Prudden et al.,

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 415

2007; Wang and Prelich, 2009). Moreover, they are implicated in spindle positioning, spindle elongation, and chromosome segregation (Hirota et al., 2014; van de Pasch et al., 2013). However, besides the kinetochore protein CENP-I in human cells, the dynein regulator Pac1 in yeast and the kleisin subunit Mcd1, no other STUbL substrates involved in the latter processes have been identified so far (Alonso et al., 2012; D’Ambrosio and Lavoie, 2014; Mukhopadhyay et al., 2010). By identifying the mechanism for proteolysis of the spindle positioning protein Kar9, we describe here an unexpected mechanism that links spindle positioning to kinetochore function. We show that the STUbLs Slx5-Slx8 ubiquitylate and target Kar9 for proteasomal degradation through a circuit that involves nuclear transport of Kar9 and SUMO-dependent recruitment of STUbLs to kinetochores. This process maintains Kar9 levels and proper distribution of Kar9 and Bim1 between astral and kinetochore microtubules thereby ensuring correct chromosome segregation and spindle positioning. RESULTS Kar9 Interacts with STUbLs and Is Ubiquitylated In Vivo and In Vitro Independently of Its Sumoylation We identified Slx5, a subunit of the heterodimeric STUbL Slx5Slx8 (Uzunova et al., 2007; Xie et al., 2007) as a yeast two-hybrid (Y2H) interacting partner of Kar9 and confirmed a previously reported Y2H interaction between Kar9 and the STUbL Uls1 (Figure S1A; Meednu et al., 2008). To test whether STUbLs are the E3s involved in Kar9 degradation, we examined ubiquitylation of C-terminally tagged Kar9 (Kar9-TAP; Leisner et al., 2008) and Kar9-HA6 in different combinations of STUbL null mutants (Figure S1B) in ubiquitin-pulldown assays (Sacher et al., 2006). Ubiquitylation of Kar9 was reduced in slx5D and in slx8D cells but did not significantly change in the uls1D mutant, while the double deletion mutants slx5Dslx8D, slx5Duls1D, slx8Duls1D displayed Kar9 ubiquitylation levels comparable with the single slx5D/slx8D mutants. Finally, Kar9 ubiquitylation was almost absent in a deletion strain lacking Slx5, Slx8, and Uls1 (Figures 1A and S1B), and Kar9-TAP was stabilized in this triple mutant strain in cycloheximide (CHX) chase experiments (Figure 1B). Since Uls1 seemed to become relevant for Kar9 ubiquitylation only in the absence of Slx5-Slx8, we mainly focused our attention on the latter. To test whether Kar9 ubiquitylation requires prior modification of the protein by SUMO, we tested ubiquitylation of the hypo-sumoylated variants Kar9-3R-TAP and Kar9-4R-TAP as well as degradation of Kar9-4R-TAP (Leisner et al., 2008). Surprisingly, neither was the Kar9-4R-TAP variant stabilized in CHX chase (Figure 1C), nor were the levels of Kar9-3R-TAP and Kar9-4RTAP ubiquitylation diminished (Figure 1D). In addition, we did not observe an increase in the levels of sumoylated Kar9 in STUbL knockout strains (Figure 1E), contrary to other STUbL substrates (Uzunova et al., 2007). We therefore tested whether interaction between Kar9 and STUbLs is independent of Kar9 sumoylation. Indeed, the Y2H interaction between the hyposumoylated Kar9-4R and STUbLs was not reduced (Figure S1A). Furthermore, in an in vitro binding assay using purified recombinant proteins, MBP-Slx5, but not MBP-Slx8 or MBP-GFP, stoichiometrically co-precipitated with His6-Kar9-GFP (Figure 1F),

showing that the interaction between Kar9 and Slx5 does not require prior Kar9 sumoylation. We found that the sequences of Kar9 that interact with Slx5 lie within the acidic, a-helical part, since a fragment comprising amino acids (aa) 78–396 was still able to bind Slx5, while interaction of Slx5 with C-terminal Kar9 sequences was nearly absent (Figure S1D). Finally, we reconstituted the ubiquitylation reaction of Kar9 with Slx5-Slx8 in vitro and detected slower migrating isoforms of His6-Kar9-GFP (but not GFP-His6) via western blotting (Figure 1G, lane 11). These bands correspond to ubiquitylated Kar9, as they were not detectable upon removal of Kar9, ATP, ubiquitin or Ubc4 (lane 4–7). Furthermore, the reaction required the presence of both Slx5 and Slx8 (lanes 9 and 10) and did not require previous conjugation of SUMO to Kar9, and was not significantly enhanced by fusion of SUMO to Kar9 (Figure S1E). The N-terminal fragment of Kar9 (aa 1–396) that includes the Slx5 interaction sequence was ubiquitylated, while a C-terminal fragment (aa 471–644) was not (Figure 1H). Taken together, these experiments strongly suggest that Slx5-Slx8 (and in their absence Uls1) ubiquitylate Kar9 and that this process does not require preceding sumoylation of Kar9. Kar9 Interacts with Slx8 and Uls1 in the Nucleus We carried out a bimolecular fluorescence complementation (BiFC) assay to decipher where Kar9 ubiquitylation takes place (Hu and Kerppola, 2003; Kodama and Hu, 2010; Sung and Huh, 2007). For this, we used the venus YFP fragments (VN, venusYFP N-terminal fragment; VC, venusYFP C-terminal fragment), bearing the I152L mutation in VN that reduces the affinity of the fragments toward each other and thus decreases falsepositive signals (Kodama and Hu, 2010). After mild overexpression under the inducible GAL1-10 promoter using 0.2% galactose, we observed strong BiFC signals between VN-Kar9 and VC-Slx8 or VC-Uls1, respectively (Figure 2A and S2B). All other fusion combinations (including VC-Kar9 and VN-Slx8 or VNUls1) remained unproductive, suggesting that the BiFC signals were specific and not due to affinity of the venusYFP fragments toward each other (Figure S2A). We further tested the specificity of the BiFC assay in competition experiments (Kodama and Hu, 2012). We found that the BiFC signals were significantly reduced upon overexpression of untagged wild-type Kar9 (Figure 2A), indicating that the BiFC signals indeed disclosed specific interactions between Kar9 and Slx8/Uls1. We observed that the mean BiFC signal intensity per cell between Kar9 and Slx8 peaked in G2/M cells and was nearly absent in anaphase (Figure 2B), while cell cycle dependency between Kar9 and Uls1 was less pronounced. Finally, labeling the spindle pole body component Spc72 and the nucleoporin Nup49 with cyan fluorescent protein revealed that the Kar9/Slx8 and Kar9/Uls1 BiFC signals localize to the nucleus (Figure 2C). Kar9 and Slx8 produced bilobed signals present on the intranuclear spindle, whereas two to three nuclear BiFC spots appeared after combining Kar9 and Uls1, indicating that the ubiquitylation reaction could take place in the nucleus. Kar9 Traffics between Cytoplasm and Nucleus In contrast to Uls1 and Slx5-Slx8, Kar9 has never been detected in the nucleus so far (Juanes et al., 2013; Liakopoulos et al., 2003; Miller and Rose, 1998; Figure S2C). To test for a

416 Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc.

Figure 1. STUbLs Ubiquitylate Kar9 In Vivo and In Vitro Independently of Kar9 Sumoylation (A) Ubiquitylation of Kar9 is reduced in a uls1Dslx5Dslx8D deletion strain. Expression of His6-tagged ubiquitin was induced (+) and ubiquitylated proteins were isolated by Ni2+ affinity purification under denaturing conditions and analyzed for Kar9-TAP via western blotting. Strains without induction of His6-ubiquitin expression served as a control ( ). See also Figures S1B and S1C. (B) Kar9 is stabilized in a uls1Dslx5Dslx8D deletion strain in CHX chase analysis. Time points after the addition of CHX are indicated. Note that samples are loaded on a 6% SDS polyacrylamide gel. (C) Degradation of the sumoylation-defective variant Kar9-4R-TAP is not reduced in CHX chase. Left: western blot; time points after the addition of CHX are indicated. Arc1, loading control. Right: Quantification shows ratio of Kar9-TAP intensities over intensities of the loading control. Error bars indicate SD of three independent experiments; n, number of experiments. Note that samples were loaded on a 15% SDS polyacrylamide gel to obtain one single band for Kar9. These specifications apply for all CHX chases and their quantifications in the paper. See also Figure S1C. (D) Ubiquitylation of Kar9 sumoylation-defective variants is not reduced. Ubiquitylated proteins were isolated by Ni2+ affinity purification under denaturing conditions and analyzed for Kar9-TAP via western blotting, as in (A). (E) SUMO-conjugated isoforms of Kar9 are not enriched upon knockout of SLX5, SLX8, and ULS1. SUMO conjugates were isolated via Ni2+ affinity purification under denaturing conditions and analyzed for Kar9-TAP via western blotting. (F) Kar9 interacts with Slx5 in vitro. Coomassiestained gel of pull-downs of the indicated purified proteins together with competing Escherichia coli lysate in the input. See also Figure S1A. (G) Slx5-Slx8 ubiquitylate Kar9 in vitro. Lane 11, His6-Kar9-GFP was incubated with ATP, ubiquitin, His6-Uba1, His6-Ubc4, MBP-Slx5, and MBP-Slx8 at 30 C for 3 hr; lanes 4–10, same reaction as in lane 11 but lacking the indicated proteins; lanes 1–3, control reactions with GFP-His6 as substrate lacking the indicated proteins. (H) Slx5-Slx8 ubiquitylate in vitro the Kar9 fragment comprising the Slx5 interaction domain, but not the C-terminal third of Kar9. Mono-, di-, and tri-ubiquitylated forms are indicated with *, **, and ***, respectively. Ubiquitylation reaction as in (G). Ubiquitylated forms of the C-terminal fragment were not observed, even in a longer exposure. See also Figures S1D and S1E.

nuclear Kar9 fraction, we expressed Kar9-GFP in xpo1-1 cells carrying a temperature-sensitive mutation in yeast exportin-1 (Stade et al., 1997). Indeed, after blocking nuclear export by shifting these cells to 37 C for 1.5 hr, we observed strong accumulation of Kar9-GFP signals on the intranuclear spindle (Figures 3A and 3D). Since localization of Kar9 on cMTs depends on Bim1 (Miller et al., 2000; Lee et al., 2000), we tested Kar9 localization after BIM1 deletion. This led to the absence of Kar9 on both the spindle and cMTs in all cells (Figure S2D). Moreover, the BiFC signals between Kar9 and Slx8 or Uls1 were also lost in bim1D cells (Figure S2E), suggesting that the localization of Kar9 to the nuclear spindle and its interaction with STUbLs depends on Bim1.

We next sought to identify possible nuclear export (NES) and nuclear localization signals (NLS) in Kar9. Using the Minimotif Miner 3.0 software (Mi et al., 2012), we identified a putative NES in the N terminus of Kar9. Mutating three leucine residues therein to alanine (Kar9nes*, Figure 3B; see also Figure S2G for functionality test) increased the percentage of cells displaying nuclear Kar9 signals from 2.1% ± 0.3% for wild-type Kar9-YFP to 20.3% ± 4.1% for Kar9nes*-YFP (Figures 3C and 3D). Utilizing the Eukaryotic Linear Motif Finder software (Dinkel et al., 2014), we also found a putative monopartite NLS at the C terminus of Kar9. Mutating its key lysine residues to alanine (Kar9nls*) had no obvious effect on Kar9 localization. However, whereas Kar9nes*-YFP accumulated in the nucleus, the introduction of

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 417

Figure 2. Kar9 Interacts with STUbLs in a BiFC Assay (A) Images show panels from cells with visible BiFC signals between the denoted Kar9 and Slx8/Uls1 fusions. The BiFC interactions can be outcompeted by overexpression of untagged, wildtype Kar9 (Kar9[). Scale bar, 2 mm (for all images). Expression levels of the fusion proteins are shown in the western blot below the images; the protein Arc1 running at 40 kDa was used as a loading control. The column diagram on the right shows quantification of the intensity of the BiFC signal per cell (made using a developed ImageJ macro; see Supplemental Experimental Procedures for details). Error bars show the SD of mean intensities per cell for three independent experiments, each covering more than 100 cells. Asterisks indicate p values obtained by Student’s t test (*p < 0.05, **p < 0.01 ***p < 0.001). See also Figure S2B. (B) BiFC signal intensity per cell between Kar9 and the STUbLs peaks in G2/M (measured as described above). Error bars show the SD of mean intensities per cell for three independent experiments, each covering more than 150 cells in total. (C) Kar9 interacts with Uls1 and Slx8 in the nucleus in BiFC and displays bilobed localization. Images were deconvoluted using Huygens Essential 3.4. Nup49 marks the nuclear envelope, Spc72 the cytoplasmic face of the yeast microtubule organizing center. See also Figures S2A, S2B, and S2E.

the nls* mutation (Kar9nes*nls*-YFP) reversed this effect to wildtype levels (1.5 ± 1.4%), indicating that this sequence functions indeed as a nuclear transport signal. Overall, we conclude that Kar9 traffics between the nucleus and cytoplasm and is thus able to interact with STUbLs inside the nucleus. Kar9 Is Targeted for Degradation in the Nucleus To verify that ubiquitylation of Kar9 by STUbLs takes place in the nucleus, we examined turnover of Kar9 nuclear transport mutants by CHX chase. Kar9nls*-TAP was clearly stabilized (p = 0.0450 in Student’s t test) and displayed higher steady state levels, while fusion of the strong NLS from the SV40 large T antigen to Kar9 (Kar9-NLSSV40) led to a significantly faster degradation (p = 0.0015; Figures 3E and S2F and S2H). Consistently, Kar9nes*-TAP showed a tendency for faster degradation compared with the wild-type protein. Deletion of ULS1, SLX5, and SLX8 stabilized all mutant and fusion proteins to the same extent as Kar9nls*-TAP (Figure 3E), showing that nuclear targeting promotes, whereas nuclear exclusion impairs STUbL-mediated degradation of Kar9. Kinetochores Recruit Slx5-Slx8 in a SUMO-Dependent Manner to Ubiquitylate Kar9 Our in vivo and in vitro data show that sumoylation of Kar9 is not necessary for its ubiquitylation by STUbLs (Figure 1). Nevertheless, we found it surprising that STUbLs mediate this process, without SUMO being involved. Thus, we investigated whether sumoylation of other proteins is required for Kar9 degradation in vivo by examining the role of the SIMs of Slx5 and Slx8.

Overexpression of KAR9 led to a growth defect of wild-type cells (Leisner et al., 2008), while simultaneous knockout of STUbLs rendered cells almost inviable (Figure 4A). This phenotype could be rescued by expression of wild-type Slx5 and Slx8 but not of their variants bearing SIM mutations (Slx5SIM1234, 4 SIMs mutated, Xie et al., 2010; and Slx8SIM1, 1 SIM mutated, Uzunova et al., 2007), suggesting that SIMs may be required for Kar9 degradation. In agreement with this, Kar9-TAP was stabilized in the SIM-deficient cells (Figure 4B). In addition, the BiFC interaction between Slx8 and Kar9 was nearly abolished upon mutation of the SIM on Slx8 (Figure 4C). Since MBP-Slx5SIM1234 is still able to bind Kar9 directly in vitro as well as wild-type MBPSlx5 (Figure 1F), these results suggest that the SIMs are indirectly required for Kar9 ubiquitylation in vivo. Interestingly, one of the Kar9-Uls1 spots and the majority of the Kar9-Slx8 signals in the BiFC assay co-localized with the kinetochore component Spc25 (Figure 5A), suggesting that the interaction between Kar9 and the STUbLs may occur at kinetochores. Consistently, Slx5 has been reported previously to interact with various kinetochore proteins in Y2H, such as Ndc10, Spc25, and Bir1 (Montpetit et al., 2006; Wong et al., 2007). Since several of them have been shown to be sumoylated, we tested whether these interactions depend on SUMO. Indeed, the Y2H interaction between Slx5 and Ndc10 was lost upon mutation of the SIMs of Slx5 or upon mutation of four sumoylation sites of Ndc10 (ndc104R; Figure S3B; Montpetit et al., 2006), whereas Slx5SIM1234 was still able to interact with Slx8 (Figure S3A). Furthermore, we discovered a yet unidentified Y2H interaction between Slx5 and Ndc80 (Figure S3B). This

418 Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc.

Figure 3. Kar9 Becomes Degraded in the Nucleus (A) Kar9 accumulates on the intranuclear spindle in xpo1-1 strains. Strains were shifted to 37 C for 1.5 hr before imaging. Images were deconvoluted using Huygens Essential 3.4. See also Figure S2D. (B) Putative nuclear export (NES) and import sequences (NLS) of Kar9, key residues in green, mutations in red. (C) Mutation of the NES in Kar9 leads to nuclear enrichment of the protein, which can be reversed by additionally mutating the NLS of Kar9. See also Figures S2F and S2G. (D) Quantification of cells displaying nuclear Kar9 in (A) and (C). Error bars show SD of three different clones, each covering 30–60 cells in G2/M. Asterisks indicate p values as described in Figure 2A. (E) Upper panel: mutation of Kar9 NLS leads to Kar9 stabilization in CHX chase analysis, whereas mutation of the NES or fusion to the NLS of SV40 large T antigen (PKKKRKV) leads to faster Kar9 degradation. Lower panel: quantification of the CHX chase, as in Figure 1C. Compared with Kar9-TAP, the rate of degradation of the Kar9-TAP-NLSSV40 fusion is significantly faster (*p = 0.0015 in Student’s t test), while the degradation rate of Kar9nls*TAP is significantly slower (*p = 0.0450). See also Figures S2F and S2G for steady state levels and functionality of different Kar9 mutants and S2H for single/simplified quantification graphs.

interaction was also dependent on the SIMs of Slx5 but not, however, on the one identified sumoylation site of Ndc80. We repeated the BiFC assay in an ndc10-1 mutant to examine whether interaction between STUbLs and Kar9 takes place at kinetochores. Shifting ndc10-1 cells to the restrictive temperature disrupts kinetochores (He et al., 2001) without activation of the spindle assembly checkpoint (Fraschini et al., 2001). The BiFC signal was not detectable at 37 C, even in control cells. However, we observed a significant reduction in the BiFC interaction between Kar9 and Slx8 upon incubation at 37 C for 40 min followed by 2-hr incubation at 28 C (Figure S3E). Accordingly, Kar9 ubiquitylation and degradation were significantly reduced in ndc10-1 cells at 37 C (Figures 5C and S3C). In contrast, the ndc10-1 mutation did not impair degradation of the spindleassociated protein Ase1, suggesting that its effect on Kar9 turnover is specific (Figure S3D). Stabilization of Kar9 was also

evident in SAC-proficient cells bearing a temperature-sensitive allele of NDC80 (ndc80-1; Figure 5C). We next tested whether abrogation of Ndc10 and Ndc80 sumoylation leads to Kar9 stabilization. In CHX chase experiments, we found that Kar9-TAP was stabilized in ndc104R ndc80R cells lacking sumoylation of both Ndc10 and Ndc80 (Figure 5D). This effect was not as pronounced as in ndc10-1 or ndc80-1 mutants, possibly because additional sumoylated proteins participate in recruitment of STUbLs to kinetochores. Importantly, the BiFC signal between Kar9 and Slx8 was nearly absent in ndc104R ndc80R cells (Figure 5B). Together, these data indicate that STUbLs are recruited by sumoylated Ndc10, Ndc80, and possibly other kinetochore proteins to kinetochores, where they ubiquitylate and target Kar9 for degradation. Finally, we examined whether artificially increasing the local concentration of Kar9 at known Slx5-Slx8 localization sites other than kinetochores would restore Kar9 degradation in kinetochore mutants. Slx5-Slx8 are recruited to nuclear pore complexes (NPCs) through their interaction with the nucleoporin Nup84 (Nagai et al., 2008). We thus used the anchor-away technique (Haruki et al., 2008) and conditionally targeted Kar92xFKBP12 fusion proteins to Nup84-FRB at the NPC by addition of rapamycin (Figure 5E; FKBP12 [FK506 binding protein 12], FRB [FKBP12-rapamycin-binding domain of the human mTOR

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 419

Figure 4. The SIMs of Slx5 and Slx8 Are Indirectly Required for Kar9 Degradation (A) The SIMs are required to rescue cell viability upon KAR9 overexpression from the inducible GAL1-10 promoter (KAR9[). See also Figures S3A and S3B. (B) Kar9 degradation depends on the SIMs (CHX chase); quantification as in Figure 1C. Lower panel: SIM1 mutant of Slx8; key residues in green, mutations in red. slx5SIM1234 mutant of Slx5 as described in Xie et al. (2010). (C) The BiFC interaction between Kar9 and Slx8 depends on the SIM of Slx8. Signal quantification and loading controls as in Figure 2, except that error bars show the SD of mean intensities per cell for six independent clones, each covering more than 100 cells.

kinase]). Indeed, Kar9-2xFKBP12-TAP was degraded in ndc10-1 cells in the presence of rapamycin (Figure 5E). Moreover, tethering Kar9 to NPCs by fusing it to Nup84 (Nup84-Kar9-TAP) restored its degradation in the ndc10-1 background as well (Figure S3G). Summarizing, we conclude that sumoylated kinetochores recruit Slx5-Slx8 through a SUMO-SIM interaction and promote Kar9 degradation through increasing the local concentration of E3 and substrate with respect to each other.

Control of Kar9 by STUbLs Ensures Chromosome Transmission Fidelity and Correct Spindle Positioning Since Kar9 protein levels are regulated at kinetochores, we investigated whether lack of Kar9 degradation would affect the fidelity of chromosome segregation. For this, we used a sectoring assay in which chromosomal loss rates can be obtained by counting half-sectoring colonies (Spencer et al., 1990). Whereas wild-type cells exhibited a loss rate of 0.02% ± 0.02%, simultaneous deletion of all three STUbLs led to a ten

420 Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc.

(legend on next page)

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 421

times higher rate of chromosome missegregation events (Figure 6A; 0.24% ± 0.04%), consistent with previous data for deletion of SLX5 and SLX8 alone (van de Pasch et al., 2013). To examine whether this was due to stabilization of Kar9, we raised the amount of Kar9 by placing KAR9 under the control of the GAL1-10 promoter. Already mild overexpression of KAR9 led to a drastic increase in half-sectoring colonies (Figure 6B, 1.33% ± 0.13%), contrary to overexpression of DYN1 (0.08% ± 0.01%) or the dynein regulator PAC1 (0.03% ± 0.03%). This phenotype clearly correlated with the degree of overexpression, since stronger KAR9 overexpression led to red sectors in almost all colonies. Interestingly, expression of an additional copy of KAR9 from a centromeric plasmid (1–2 copies per cell) led to a significant increase in chromosome missegregation events. The amount of Kar9 protein in these cells was comparable with the amount of Kar9 in the uls1Dslx5Dslx8D deletion strain (Figure 6A). This indicates that increased levels of Kar9 can account for the observed defects in chromosome segregation upon STUbL deletion, but does not exclude that stabilization of other proteins contributes to these defects as well. Intriguingly, we could not observe an obvious enrichment of Kar9-YFP at kinetochores upon STUbL deletion, but rather an increase of the Kar9-YFP amount on cMTs (Figure 6C), similar to KAR9-YFP overexpression (see later and Miller et al., 1998). Since Kar9 interacts with Bim1 and other MAPs, the chromosome segregation defects observed upon elevated Kar9 levels could be due to disruption of the equilibrium of MAPs between kinetochore microtubules and cMTs. Strikingly, Kar9 overexpression led to nearly complete redistribution of YFP-Bim1 from nuclear to cytoplasmic microtubules (Figure 6D). Since BIM1 deletion but also overexpression (Figure 6B) provokes strong defects in chromosome segregation, these data suggest that control of Kar9 levels is important to avoid perturbation of the Bim1 equilibrium and chromosome segregation errors. Interestingly, deletion of KAR9 led also to a more than ten times higher chromosome missegregation rate (0.26% ± 0.02%) compared with wild-type. Deleting BNI1 and DYN1, which encode essential components of the Kar9- and the dynein-dependent spindle positioning pathway, respectively, did not significantly increase missegregation events (0.04% ± 0.01% and 0.05% ± 0.02%, respectively), indicating that the observed chromosome loss in kar9D cells is not due to defects in spindle positioning. Plasmids containing kar9nls*, kar9nes*, and KAR9-NLSSV40 did not recue this phenotype (0.15% ± 0.01% for kar9nls*, 0.21% ± 0.06% for kar9nes*, and 0.29% ± 0.05% for Kar9-NLSSV40), indicating that perturbation of Kar9 distribution impairs chromosome segregation. The mechanism involved here is not clear (see Discussion), however these data

show that precise control of Kar9 levels and subcellular distribution is important to avoid chromosome segregation errors. Lastly, we examined the effects of Kar9 overexpression in spindle positioning. In agreement with previous reports (van de Pasch et al., 2013), the metaphase spindle in the STUbL deletion cells often translocated completely into the bud, similarly to KAR9 overexpressing cells and cells carrying an additional copy of KAR9 (Figure 6E), whereas concomitant knockout of KAR9 reversed this phenotype. In addition, cells expressing the nuclear import defective mutant kar9nls* (but not kar9nes* or KAR9-NLSSV40) displayed an increased frequency of bud-drawn spindles, comparable with the phenotype of STUbL deletion. We thus conclude that nuclear transport of Kar9 and its subsequent degradation are required to hold the levels of Kar9 at bay in order to prevent errors in chromosome segregation and spindle positioning. DISCUSSION In this study, we unravel an unanticipated degradation pathway affecting the yeast spindle positioning factor Kar9 mainly by the STUbLs Slx5-Slx8 and, in their absence, also Uls1. The latter assumption is based on our finding that uls1D deletion does not reduce Kar9 ubiquitylation, while it is required to eliminate the residual Kar9 ubiquitylation in slx5D slx8D cells, as well as on reports suggesting that Slx5-Slx8 and Uls1 may have indeed overlapping functions (Sriramachandran and Dohmen, 2014). The protein Kar9, so far detected exclusively on cytoplasmic microtubules, traffics between cytoplasm and nucleus and becomes ubiquitylated by kinetochore-recruited STUbLs. Previously, we had shown that Kar9 ubiquitylation depends on the integrity of the septin ring, the scaffold of the cytokinetic apparatus at the bud neck, which prompted us to speculate that Kar9 could be ubiquitylated there (Kammerer et al., 2010). However, we found that forcing Kar9 into the nucleus in septin mutants, using the Kar9-NLSSV40 fusion, results in degradation of the fusion protein (Figure S4A). Thus, we favor the idea that septin mutations lead to Kar9 stabilization by interfering with Kar9 import into the nucleus and indirectly abolish its ubiquitylation and degradation at kinetochores. The presence of Kar9 at kinetochores but also the intriguing finding that KAR9 deletion impairs chromosome transmission fidelity independently of its spindle positioning function are reminiscent of the tumor suppressor APC, the proposed functional Kar9 homolog in mammalian cells (Fuchs and Yang, 1999). Like Kar9, APC mediates spindle positioning on astral microtubules, but loss of APC has been shown to provoke chromosomal missegregation as well (Fodde et al., 2001; Kaplan et al., 2001). It

Figure 5. Kinetochores Recruit STUbLs to Regulate Kar9 (A) Kar9 interacts with Slx8 and Uls1 at kinetochores in BiFC analysis. Images were deconvoluted using Huygens Essential 3.4. (B) The BiFC interaction of Kar9 and Slx8 depends on sumoylation of Ndc10 and Ndc80. ndc104R and ndc80R encode proteins with mutations in sumoylation sites (Montpetit et al., 2006). Signal quantification and loading controls as in Figure 2, except that error bars show the SD of the mean intensities per cell for six independent experiments, each covering more than 100 cells. (C) Kar9 degradation depends upon Ndc10 and Ndc80 (CHX chase). Cells were shifted to 37 C 30 min prior to the addition of CHX; quantification as in Figure 1C. Note that Kar9 degradation is accelerated at 37 C. See also Figures S3C–S3E. (D) Kar9 degradation depends on sumoylation of Ndc10 and Ndc80 (CHX chase). Quantification as in Figure 1C. (E) Recruitment of Kar9 to the Slx5-Slx8 docking site at nuclear pores restores its degradation in kinetochore mutants. Left, up: Kar9-2xFKBP-GFP is recruited to the Nup84 complex after the addition of rapamycin in cells expressing a Nup84-FRB fusion. Left, down: CHX chase of Kar9-2xFKBP-TAP in ndc10-1 Nup84-FRBexpressing cells, as in Figure 1C. Right: quantification of the CHX chase as in Figure 1C. See also Figure S3F and S3G.

422 Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc.

(legend on next page)

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 423

will be interesting to examine whether elevated levels of APC display defects similar to Kar9 overexpression and whether cells need to control the amounts of APC to avert missegregation events. Naturally, the question arises how proteins like Kar9 and APC are able to contribute to both spindle positioning and chromosome segregation in detail. One possibility is that interfering with their protein levels affects microtubule dynamics indirectly through their binding to microtubule regulators such as Bim1 (yeast EB1), Stu2 (yeast XMAP215), or Bik1 (yeast CLIP-170) (Miller et al., 2000; Moore et al., 2006; Wolyniak et al., 2006). Hence, perturbing the equilibrium of other MAPs could impair microtubule function and affect chromosome capturing. Consistent with this idea, YFP-Bim1 almost completely mislocalized to cytoplasmic microtubules upon Kar9 overexpression. In addition, the ratio of nuclear YFP-Bim1 compared with its total signal intensity was significantly reduced upon STUbL deletion (Figure S4C). However, this effect was difficult to assess due to the aberrant spindle morphology and the increase in total intensity of YFP-Bim1 in this strain (data not shown). Nevertheless, it seems that nuclear control of Kar9 levels prevents perturbation of microtubule regulators such as Bim1 and thus chromosomal missegregation in mitosis. Data from BiFC, Y2H, ubiquitin pull-down assays, and CHX chase analysis indicate that kinetochores are the cellular site where STUbL-dependent Kar9 ubiquitylation takes place. This is in line with previous observations showing that the asymmetry of Kar9 on cMTs is in part controlled by kinetochore-derived signals that depend on the spindle assembly checkpoint (Leisner et al., 2008). However, Kar9 is not symmetric on cMTs either in STUbL-deleted cells (Figure S4B) or after overexpression (Kammerer et al., 2010), suggesting that a still unidentified signal from kinetochores controls Kar9 distribution. In any case, the finding that kinetochore-dependent ubiquitylation of Kar9 by STUbLs regulates its abundance in the cytoplasm is an example of a mechanism that mediates cross-talk between different spindle domains. We think that Kar9 degradation mainly reduces its

cytoplasmic levels because the nuclear export rate of Kar9 is higher than its rate of nuclear import (Figure 7). Hence, stabilization of Kar9 upon blocking its nuclear import or upon deletion of STUbLs leads to a significant increase of cMT-bound Kar9 and, thus, mispositioned spindles. The role of sumoylation in the STUbL-dependent degradation is particularly interesting; despite the fact that Kar9 is sumoylated in vivo (Leisner et al., 2008), its sumoylation is not prerequisite for the ubiquitylation process. However, the SIMs of Slx5 are not necessary for binding Kar9 in vitro but become essential for Kar9 regulation in vivo, because they are required for the recruitment of STUbLs to kinetochores in a SUMO-dependent manner (Figure 7). Why do cells use this complicated pathway to regulate a spindle positioning factor at kinetochores and not in the cytoplasm? To position the mitotic spindle, cMTs contact the cell cortex at many different sites in a highly dynamic fashion. Hence, it may be difficult to establish a control mechanism to collectively regulate proteins on cMTs plus-ends in the cytoplasm or at the cell cortex. In contrast, kinetochores may serve as a more defined location to regulate proteins at cMT plus-ends. Many MAPs are common between kinetochore and astral microtubules and their relative abundance must be controlled to maintain proper microtubule dynamics of each microtubule type. However, not all MAPs present on astral microtubules are also present on kinetochore microtubules, especially during mitosis of higher eukaryotic cells, when astral and kinetochore microtubules share the same cytoplasm. Thus, the described degradation pathway may also serve as a quality control mechanism that degrades MAPs at kinetochores in order to generate microtubule plus-end diversity. EXPERIMENTAL PROCEDURES Yeast and Bacterial Growth Conditions, Strains, and Plasmids Yeast strains are listed in Table S1; Media and genetic manipulations were done as already described (Guthrie and Fink, 1991; Janke et al., 2004; Longtine et al., 1998); Y2H was done as in (Kammerer et al., 2010). Standard molecular biology methods were used (Ausubel et al., 1989). The colony sectoring assay was performed as described in Spencer et al. (1990).

Figure 6. Defined Levels of Kar9 Are Required for Faithful Chromosome Segregation and Spindle Positioning (A) Deletion of KAR9 causes chromosome segregation defects. Up, examples of half-sectoring colonies; down, left, quantification. Error bars indicate the SD of the mean percentage of half-sectoring colonies of three clones, each covering 1,000–3,000 colonies. Asterisks and p values as described in Figure 2A. Loss rate of wild-type (WT) cells, 0.02% ± 0.02% (in total 2/7,994 colonies); of wild-type cells transformed with an empty centromeric plasmid, 0.05% ± 0.04% (in total 2/ 4,092 colonies); of wild-type cells transformed with a centromeric plasmid bearing KAR9 wild-type, 0.18% ± 0.03% (in total 8/4309 colonies); of kar9D cells, 0.26% ± 0.02% (in total 17/6,516 colonies); of bni1D and dyn1D cells, 0.04% ± 0.01% (in total 4/9,393 colonies) and 0.05% ± 0.02% (3/7,381 colonies), respectively. Loss rate of uls1Dslx5Dslx8D cells, 0.24% ± 0.04% (in total 24/10,477 colonies); of kar9D cells transformed with an empty centromeric plasmid, 0.24% ± 0.01% (in total 11/4,566 colonies); of kar9D cells transformed with a centromeric plasmid bearing KAR9 wild-type, 0.04% ± 0.01% (in total 3/8,319 colonies); of kar9D cells transformed with a centromeric plasmid bearing kar9nls*, 0.15% ± 0.01% (in total 13/8,858 colonies); kar9nes*, 0.21% ± 0.06% (in total 8/3,844 colonies); or KAR9-NLSSV40, 0.29% ± 0.05% (in total 12/4,083 colonies). Down, right, control western blot for the strains used for the chromosome segregation assays, showing the levels of Kar9-TAP (endogenously tagged) in wild-type and uls1Dslx5/Dslx8D cells or in the wild-type cells with an additional plasmid-born Kar9-TAP copy (+). (B) Experiment as in (A), showing that overexpression of KAR9 and BIM1 impairs chromosome segregation. 0.2% (upper panel) or 2% galactose (lower panel) was used for mild or strong overexpression ([), respectively. Loss rate of wild-type cells on 0.2% galactose, 0.03% ± 0.01% (in total 3/11,620 colonies). Loss rate of KAR9[ cells on 0.2% galactose, 1.33% ± 0.13% (in total 115/8,606 colonies). Loss rate of DYN1[ cells on 0.2% galactose, 0.08% ± 0.01% (in total 6/7,453 colonies). Loss rate of PAC1[ cells on 0.2% galactose, 0.03% ± 0.03% (in total 2/6,923 colonies). Loss rate of BIM1[ cells on 0.2% galactose, 3.99% ± 0.38% (in total 116/2,929 colonies). Loss rate of bim1D cells on glucose, 4.99% ± 0.73% (in total 45/953 colonies). (C) The amount of Kar9 on cytoplasmic microtubules increases upon STUbL deletion. Left, image examples. Right, quantification. Error bars indicate the SEM for 57 (wild-type) or 95 (uls1Dslx5Dslx8D) cells. Asterisks and p values as described in Figure 2A. See also Figure S4B. (D) Overexpression of KAR9 perturbs localization of Bim1. Left, characteristic images; right, quantification. Error bars show the SD for three different experiments, each covering >20 cells. Images were deconvoluted using Huygens Essential 3.4. See also Figure S4C. (E) Increased levels of Kar9 pull the spindle completely into the bud. KAR9[ depicts KAR9 overexpression. Right, quantification. Error bars indicate the SD of three different experiments, covering 75–200 cells in G2/M.

424 Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc.

Figure 7. Model STUbLs Are Loaded on Kinetochores to Regulate Kar9 Utilizing nuclear export and import pathways, Kar9 distributes between cytoplasm and nucleus. In both compartments, Bim1 loads Kar9 to microtubule plus-ends. The global protein levels of Kar9 are regulated through ubiquitylation at kinetochores. SUMO-targeted ubiquitin ligases are recruited through their SUMO interaction motifs to sumoylated kinetochores, where they ubiquitylate Kar9 for subsequent proteasomal degradation. Defects in this process lead to elevated levels of Kar9 on cytoplasmic microtubules, which provoke spindle translocation into the bud but also disturb the distribution of Bim1 between the different spindle domains, ultimately leading to defects in chromosome segregation.

3 hr. Reaction buffer contained 50 mM HEPES, 50 mM NaCl, 2.5 mM MgCl2, and 2.5 mM DTT at pH 7.3.

Plasmids are listed in Table S2. Point mutations were introduced by sitedirected mutagenesis (Phusion polymerase, Thermo Scientific). Recombinant Protein Expression and Purification Proteins were expressed in Rosetta cells (Invitrogen). Slx5 and Slx8 were purified via an N-terminal MBP tag as described in Xie et al. (2007). Yeast Ubc4 and Kar9-GFP were purified via an N-terminal GFP with a C-terminal His6 tag after induction with 1 mM isopropyl b-D-1-thiogalactopyranoside at 18 C overnight using Talon beads (Clontech). Proteins were eluted with 50 mM HEPES (pH 7.5), 150 mM imidazole (and 250 mM NaCl for Kar9-GFP). Human recombinant Uba1 and yeast ubiquitin were purchased from Boston Biochem. In Vitro Interaction, SUMO, and Ubiquitin Pull-Down Assays In vitro pull-down (interaction) assays and in vitro ubiquitylation assays were performed as in Xie et al. (2007). MBP-GFP, MBP-Slx5, MBP-Slx5SIM1234, or MBP-Slx8 were incubated with His6-Kar9-GFP in a molar ratio of 3:1 in 1 ml of buffer containing 50 mM HEPES (pH 7.5), 150 mM NaCl, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 13 protease inhibitor mixture (Roche), 0.1 mM BSA, 1% Tween 20, and approximately 250 ml of Rosetta lysate corresponding to 1 OD600 cells. After incubation with Talon beads for 1 hr, beads were washed and proteins were eluted with SDS loading buffer and analyzed via SDS-PAGE. SUMO, ubiquitin pull-down, and CHX chase experiments were performed as in Kammerer et al. (2010), Leisner et al. (2008), and Sacher et al. (2006). For SUMO and ubiquitin pull-down assays, Kar9-TAP was expressed from a centromeric plasmid from its own promoter in cells deleted for endogenous KAR9. Expression of His6-Smt3 (yeast SUMO) or His6-ubiquitin was induced with 1 mM CuSO4 for 2–4 hr. For western blots, peroxidase anti-peroxidase (DakoCytomation), a-Arc1 (polyclonal, rabbit; E. Hurt; Simos et al., 1996), a-Nop1 (polyclonal, rabbit; E. Hurt; Schimmang et al., 1989), and a-GFP (polyclonal, rabbit; J. Lechner) were used. In Vitro Ubiquitylation Assays In vitro ubiquitylation assays were performed as in Xie et al. (2007). 0.5 mM yeast recombinant His6-Kar9-GFP was incubated with 0.2 mM human recombinant Uba1 (Boston Biochem), 7 mM yeast recombinant ubiquitin (Boston Biochem), 0.6 mM yeast recombinant Ubc4, 10 mM ATP, 0.4 mM yeast recombinant Slx5, and 0.4 mM yeast recombinant Slx8 in 20 ml at 30 C for

Fluorescence Microscopy and Bimolecular Fluorescence Complementation Proteins were genomically tagged with VN (aa 1– 155 of Venus) bearing the I152L mutation or VC (aa 156–238) (Kodama and Hu, 2010). Protein expression under the GAL110 promoter was induced with 0.2% galactose for 2 hr. Cells were visualized using an Olympus IX81 microscope (Tokyo, Japan) and the CellR imaging system and an APO-plan 1003 objective or a Zeiss Axioimager Z1 microscope (Oberkochen, Germany) with the Metamorph imaging software and a 1003 Plan Apochromat objective. BiFC signal intensities were measured with an ImageJ macro; see Supplemental Experimental Procedures for a description of the macro. SUPPLEMENTAL INFORMATION Supplemental Information includes Supplemental Experimental Procedures, four figures, and two tables and can be found with this article online at http://dx.doi.org/10.1016/j.devcel.2016.01.011. AUTHOR CONTRIBUTIONS J.S. designed and performed the experiments and wrote the manuscript; L.S., D.P., and D.K. designed and performed the experiments; D.L. designed the experiments and wrote the manuscript. ACKNOWLEDGMENTS We are grateful to R.J. Dohmen, S. Jentsch, J. Lechner, M. Knop, E. Hurt, M. Radman-Livaja, and R. Ziane for plasmids and strains; E. Hurt, M. Kos, F. Melchior, A. Werner, and S. Piatti for their support and reagents; M. Hochstrasser for providing the SIM1234 mutant of Slx5; D. Xirodimas, S. Piatti, M. Granata, J. Mansfeld, and C. Norden for reading the manuscript; R. Tasakis for technical support; and the Nikon Imaging Center (University of Heidelberg) and the Montpellier RIO imaging platform for support with microscopy and deconvolution. J.S., L.S., and D.L. were supported by the German Research Foundation (DFG Priority Program SPP1365 grant LI 1552 3-1). D.L. and D.P are supported by the Fondation pour la Recherche Me´dicale, code dossier FRM: AJE20131128973). Received: September 17, 2015 Revised: December 4, 2015 Accepted: January 14, 2016 Published: February 22, 2016

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 425

REFERENCES Alonso, A., D’Silva, S., Rahman, M., Meluh, P.B., Keeling, J., Meednu, N., Hoops, H.J., and Miller, R.K. (2012). The yeast homologue of the microtubule-associated protein Lis1 interacts with the sumoylation machinery and a SUMO-targeted ubiquitin ligase. Mol. Biol. Cell 23, 4552–4566. Ausubel, F., Brent, R., Kingston, R., Moore, D., Seidman, J., Smith, J., and Struhl, K. (1989). Current Protocols in Molecular Biology (Wiley). Caldwell, C.M., Green, R.A., and Kaplan, K.B. (2007). APC mutations lead to cytokinetic failures in vitro and tetraploid genotypes in Min mice. J. Cell Biol. 178, 1109–1120. D’Ambrosio, L.M., and Lavoie, B.D. (2014). Pds5 prevents the PolySUMOdependent separation of sister chromatids. Curr. Biol. 24, 1–11. Dinkel, H., Van Roey, K., Michael, S., Davey, N.E., Weatheritt, R.J., Born, D., Speck, T., Kru¨ger, D., Grebnev, G., Kuban, M., et al. (2014). The eukaryotic linear motif resource ELM: 10 years and counting. Nucleic Acids Res. 42, D259–D266. Eshel, D., Urrestarazu, L.A., Vissers, S., Jauniaux, J.C., van Vliet-Reedijk, J.C., Planta, R.J., and Gibbons, R. (1993). Cytoplasmic dynein is required for normal nuclear segregation in yeast. Proc. Natl. Acad. Sci. USA 90, 11172–11176. Fleming, E.S., Temchin, M., Wu, Q., Maggio-Price, L., and Tirnauer, J.S. (2009). Spindle misorientation in tumors from APC(min/+) mice. Mol. Carcinog. 48, 592–598. Fodde, R., Kuipers, J., Rosenberg, C., Smits, R., Kielman, M., Gaspar, C., van Es, J.H., Breukel, C., Wiegant, J., Giles, R.H., et al. (2001). Mutations in the APC tumour suppressor gene cause chromosomal instability. Nat. Cell Biol. 3, 433–438. Fraschini, R., Beretta, A., Lucchini, G., and Piatti, S. (2001). Role of the kinetochore protein Ndc10 in mitotic checkpoint activation in Saccharomyces cerevisiae. Mol. Genet. Genomics 266, 115–125. Fuchs, E., and Yang, Y. (1999). Crossroads on cytoskeletal highways. Cell 98, 547–550. Galjart, N. (2005). CLIPs and CLASPs and cellular dynamics. Nat. Rev. Mol. Cell Biol. 6, 487–498. Guthrie, C., and Fink, G.R. (1991). Guide to yeast genetics and molecular biology. Methods Enzymol. 194, 3–933.

Kaplan, K.B., Burds, A.A., Swedlow, J.R., Bekir, S.S., Sorger, P.K., and Na¨thke, I.S. (2001). A role for the adenomatous polyposis coli protein in chromosome segregation. Nat. Cell Biol. 3, 429–432. Kodama, Y., and Hu, C.-D. (2010). An improved bimolecular fluorescence complementation assay with a high signal-to-noise ratio. Biotechniques 49, 793–805. Kodama, Y., and Hu, C.-D. (2012). Bimolecular fluorescence complementation (BiFC): a 5-year update and future perspectives. Biotechniques 53, 285–298. Korinek, W.S., Copeland, M.J., Chaudhuri, A., and Chant, J. (2000). Molecular linkage underlying microtubule orientation toward cortical sites in yeast. Science 287, 2257–2259. Lee, L., Tirnauer, J.S., Li, J., Schuyler, S.C., Liu, J.Y., and Pellman, D. (2000). Positioning of the mitotic spindle by a cortical-microtubule capture mechanism. Science 287, 2260–2262. Leisner, C., Kammerer, D., Denoth, A., Britschi, M., Barral, Y., and Liakopoulos, D. (2008). Regulation of mitotic spindle asymmetry by SUMO and the spindle-assembly checkpoint in yeast. Curr. Biol. 18, 1249–1255. Liakopoulos, D., Kusch, J., Grava, S., Vogel, J., and Barral, Y. (2003). Asymmetric loading of Kar9 onto spindle poles and microtubules ensures proper spindle alignment. Cell 112, 561–574. Longtine, M.S., McKenzie, A., Demarini, D.J., Shah, N.G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J.R. (1998). Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14, 953–961. Maekawa, H., Usui, T., Knop, M., and Schiebel, E. (2003). Yeast Cdk1 translocates to the plus end of cytoplasmic microtubules to regulate bud cortex interactions. EMBO J. 22, 438–449. Meednu, N., Hoops, H., D’Silva, S., Pogorzala, L., Wood, S., Farkas, D., Sorrentino, M., Sia, E., Meluh, P., and Miller, R.K. (2008). The spindle positioning protein Kar9p interacts with the sumoylation machinery in Saccharomyces cerevisiae. Genetics 180, 2033–2055. Mi, T., Merlin, J.C., Deverasetty, S., Gryk, M.R., Bill, T.J., Brooks, A.W., Lee, L.Y., Rathnayake, V., Ross, C.A., Sargeant, D.P., et al. (2012). Minimotif Miner 3.0: database expansion and significantly improved reduction of falsepositive predictions from consensus sequences. Nucleic Acids Res. 40, D252–D260.

Haruki, H., Nishikawa, J., and Laemmli, U.K. (2008). The anchor-away technique: rapid, conditional establishment of yeast mutant phenotypes. Mol. Cell 31, 925–932.

Miller, R.K., and Rose, M.D. (1998). Kar9p is a novel cortical protein required for cytoplasmic microtubule orientation in yeast. J. Cell Biol. 140, 377–390.

He, X., Rines, D.R., Espelin, C.W., and Sorger, P.K. (2001). Molecular analysis of kinetochore-microtubule attachment in budding yeast. Cell 106, 195–206.

Miller, R.K., Heller, K.K., Frise, L., Wallack, D.L., Loayza, D., Gammie, A.E., and Rose, M.D. (1998). The kinesin-related proteins, Kip2p and Kip3p function differently in nuclear migration in yeast. Mol. Biol. Cell 9, 2051–2068.

Hildebrandt, E.R., and Hoyt, M.A. (2000). Mitotic motors in Saccharomyces cerevisiae. Biochim. Biophys. Acta 1496, 99–116. Hirota, K., Tsuda, M., Murai, J., Takagi, T., Keka, I.S., Narita, T., Fujita, M., Sasanuma, H., Kobayashi, J., and Takeda, S. (2014). SUMO-targeted ubiquitin ligase RNF4 plays a critical role in preventing chromosome loss. Genes Cells 19, 743–754. Hotz, M., Leisner, C., Chen, D., Manatschal, C., Wegleiter, T., Ouellet, J., Lindstrom, D., Gottschling, D.E., Vogel, J., and Barral, Y. (2012). Spindle pole bodies exploit the mitotic exit network in metaphase to drive their agedependent segregation. Cell 148, 958–972. Hu, C.-D., and Kerppola, T.K. (2003). Simultaneous visualization of multiple protein interactions in living cells using multicolor fluorescence complementation analysis. Nat. Biotechnol. 21, 539–545.

Miller, R.K., Cheng, S.C., and Rose, M.D. (2000). Bim1p/Yeb1p mediates the Kar9p-dependent cortical attachment of cytoplasmic microtubules. Mol. Biol. Cell 11, 2949–2959. Montpetit, B., Hazbun, T.R., Fields, S., and Hieter, P. (2006). Sumoylation of the budding yeast kinetochore protein Ndc10 is required for Ndc10 spindle localization and regulation of anaphase spindle elongation. J. Cell Biol. 174, 653–663. Moore, J.K., Silva, S.D., and Miller, R.K. (2006). The CLIP-170 homologue Bik1p promotes the phosphorylation and asymmetric localization of Kar9p. Mol. Biol. Cell 17, 178–191. Mukhopadhyay, D., Arnaoutov, A., and Dasso, M. (2010). The SUMO protease SENP6 is essential for inner kinetochore assembly. J. Cell Biol. 188, 681–692.

Janke, C., Magiera, M.M., Rathfelder, N., Taxis, C., Reber, S., Maekawa, H., Moreno-Borchart, A., Doenges, G., Schwob, E., Schiebel, E., et al. (2004). A versatile toolbox for PCR-based tagging of yeast genes: new fluorescent proteins, more markers and promoter substitution cassettes. Yeast 21, 947–962.

Nagai, S., Dubrana, K., Tsai-Pflugfelder, M., Davidson, M.B., Roberts, T.M., Brown, G.W., Varela, E., Hediger, F., Gasser, S.M., and Krogan, N.J. (2008). Functional targeting of DNA damage to a nuclear pore-associated SUMOdependent ubiquitin ligase. Science 322, 597–602.

Juanes, M.A., Twyman, H., Tunnacliffe, E., Guo, Z., ten Hoopen, R., and Segal, M. (2013). Spindle pole body history intrinsically links pole identity with asymmetric fate in budding yeast. Curr. Biol. 23, 1–10.

Palmer, R.E., Sullivan, D.S., and Koshland, D. (1992). Role of astral microtubules and actin in spindle orientation and migration in the budding yeast, Saccharomyces cerevisiae. J. Cell Biol. 119, 583–593.

Kammerer, D., Stevermann, L., and Liakopoulos, D. (2010). Ubiquitylation regulates interactions of astral microtubules with the cleavage apparatus. Curr. Biol. 20, 1233–1243.

Prudden, J., Pebernard, S., Raffa, G., Slavin, D.A., Perry, J.J.P., Tainer, J.A., McGowan, C.H., and Boddy, M.N. (2007). SUMO-targeted ubiquitin ligases in genome stability. EMBO J. 26, 4089–4101.

426 Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc.

Sacher, M., Pfander, B., Hoege, C., and Jentsch, S. (2006). Control of Rad52 recombination activity by double-strand break-induced SUMO modification. Nat. Cell Biol. 8, 1284–1290. Schimmang, T., Tollervey, D., Kern, H., Frank, R., and Hurt, E.C. (1989). A yeast nucleolar protein related to mammalian fibrillarin is associated with small nucleolar RNA and is essential for viability. EMBO J. 8, 4015–4024. Simos, G., Segref, A., Fasiolo, F., Hellmuth, K., Shevchenko, A., Mann, M., and Hurt, E.C. (1996). The yeast protein Arc1p binds to tRNA and functions as a cofactor for the methionyl- and glutamyl-tRNA synthetases. EMBO J. 15, 5437–5448. Spencer, F., Gerring, S.L., Connelly, C., and Hieter, P. (1990). Mitotic chromosome transmission fidelity mutants in Saccharomyces cerevisiae. Genetics 124, 237–249. Sriramachandran, A.M., and Dohmen, R.J. (2014). SUMO-targeted ubiquitin ligases. Biochim. Biophys. Acta 1843, 75–85. Stade, K., Ford, C.S., Guthrie, C., and Weis, K. (1997). Exportin 1 (Crm1p) is an essential nuclear export factor. Cell 90, 1041–1050. Stevermann, L., and Liakopoulos, D. (2012). Molecular mechanisms in spindle positioning: structures and new concepts. Curr. Opin. Cell Biol. 24, 816–824. Sung, M., and Huh, W. (2007). Bimolecular fluorescence complementation analysis system for in vivo detection of protein – protein interaction in Saccharomyces cerevisiae. Yeast 24, 767–775. Uzunova, K., Go¨ttsche, K., Miteva, M., Weisshaar, S.R., Glanemann, C., Schnellhardt, M., Niessen, M., Scheel, H., Hofmann, K., Johnson, E.S., et al. (2007). Ubiquitin-dependent proteolytic control of SUMO conjugates. J. Biol. Chem. 282, 34167–34175.

van de Pasch, L.A.L., Miles, A.J., Nijenhuis, W., Brabers, N.A.C.H., van Leenen, D., Lijnzaad, P., Brown, M.K., Ouellet, J., Barral, Y., Kops, G.J.P.L., et al. (2013). Centromere binding and a conserved role in chromosome stability for SUMO-dependent ubiquitin ligases. PLoS One 8, e65628. van Hagen, M., Overmeer, R.M., Abolvardi, S.S., and Vertegaal, A.C.O. (2009). RNF4 and VHL regulate the proteasomal degradation of SUMO-conjugated Hypoxia-Inducible Factor-2a. Nucleic Acids Res. 38, 1922–1931. Wang, Z., and Prelich, G. (2009). Quality control of a transcriptional regulator by SUMO-targeted degradation. Mol. Cell. Biol. 29, 1694–1706. Wolyniak, M.J., Blake-hodek, K., Kosco, K., Hwang, E., You, L., and Huffaker, T.C. (2006). The regulation of microtubule dynamics in saccharomyces cerevisiae by three interacting plus-end tracking proteins. Mol. Biol. Cell 17, 2789– 2798. Wong, J., Nakajima, Y., Westermann, S., Shang, C., Kang, J., Goodner, C., Houshmand, P., Fields, S., Chan, C.S.M., Drubin, D., et al. (2007). A protein interaction map of the mitotic spindle. Mol. Biol. Cell 18, 3800–3809. Xie, Y., Kerscher, O., Kroetz, M.B., McConchie, H.F., Sung, P., and Hochstrasser, M. (2007). The yeast Hex3.Slx8 heterodimer is a ubiquitin ligase stimulated by substrate sumoylation. J. Biol. Chem. 282, 34176–34184. Xie, Y., Rubenstein, E.M., Matt, T., and Hochstrasser, M. (2010). SUMO-independent in vivo activity of a SUMO-targeted ubiquitin ligase toward a shortlived transcription factor. Genes Dev. 24, 893–903. Yin, H., Pruyne, D., Huffaker, T.C., and Bretscher, A. (2000). Myosin V orientates the mitotic spindle in yeast. Nature 406, 1013–1015.

Developmental Cell 36, 415–427, February 22, 2016 ª2016 Elsevier Inc. 427