CENP-C Is a Structural Platform for Kinetochore Assembly

Current Biology 21, 399–405, March 8, 2011 ª2011 Elsevier Ltd All rights reserved

DOI 10.1016/j.cub.2011.02.005

Report CENP-C Is a Structural Platform for Kinetochore Assembly Marcin R. Przewloka,1,* Zsolt Venkei,1 Victor M. Bolanos-Garcia,2 Janusz Debski,3 Michal Dadlez,3 and David M. Glover1,* 1Department of Genetics 2Department of Biochemistry University of Cambridge, Cambridge CB2 3EH, UK 3Laboratory of Mass Spectrometry, Institute of Biochemistry and Biophysics, Polish Academy of Sciences, 02-106 Warsaw, Poland

Summary Centromeres provide a region of chromatin upon which kinetochores are assembled in mitosis [1, 2]. Centromeric protein C (CENP-C) is a core component of this centromeric chromatin [3, 4] that, when depleted, prevents the proper formation of both centromeres and kinetochores [5–10]. CENP-C localizes to centromeres throughout the cell cycle via its C-terminal part [6, 8], whereas its N-terminal part appears necessary for recruitment of some but not all components of the Mis12 complex of the kinetochore [8]. We now find that all kinetochore proteins belonging to the KMN (KNL1/Spc105, the Mis12 complex, and the Ndc80 complex) network [1] bind to the N-terminal part of Drosophila CENP-C. Moreover, we show that the Mis12 complex component Nnf1 interacts directly with CENP-C in vitro. To test whether CENP-C’s N-terminal part was sufficient to recruit KMN proteins, we targeted it to the centrosome by fusing it to a domain of Plk4 kinase [11]. The Mis12 and Ndc80 complexes and Spc105 protein were then all recruited to centrosomes at the expense of centromeres, leading to mitotic abnormalities typical of cells with defective kinetochores. Thus, the N-terminal part of Drosophila CENP-C is sufficient to recruit core kinetochore components and acts as the principal linkage between centromere and kinetochore during mitosis. Results and Discussion Centromeres are patches of chromatin defined by the presence of the specific histone H3 variant called centromeric protein A (CENP-A)/CenH3 [12], also known as CID in Drosophila, Cse4 in S. cerevisiae, and Cnp1 in S. pombe (reviewed in [13]). Shortly before each mitosis, kinetochores are assembled on centromeres in a process whose mechanism and regulation are largely unknown. Recently published studies have pointed toward the centromeric protein CENP-C as a potential candidate for the regulatory role in kinetochore formation [7, 8]. CENP-C is localized at the centromeric chromatin throughout the cell cycle [4, 6]. It is embedded in the centromeric chromatin and, together with CENP-A, forms part of the core of the centromere. The carboxy-terminal part of CENP-C is known to be responsible

*Correspondence: [email protected] (M.R.P.), [email protected] (D.M.G.)

for the binding to centromeres but by itself is insufficient to rescue the lack of CENP-C in vertebrate cells or Drosophila embryos [6, 8, 14]. However, the exact function of CENP-C’s N-terminal part has not been defined [8]. In vertebrates, CENP-C is also a component of the so-called constitutive centromere-associated network (CCAN; see [1, 2] for reviews) that comprises at least 14 proteins assembled in several subcomplexes. Whereas CENP-C seems to be a component of the centromere in all organisms, the associated CCAN proteins appear to be absent from the centromeres in C. elegans and Drosophila. Because these organisms thus provide a fundamentally simplified centromere whose functions may be easier to dissect, we were led to study the roles of CENP-C in kinetochore assembly using D. melanogaster as a model. In order to test the localization properties of the two functional parts of CENP-C, we first arbitrarily divided the open reading frame and expressed both portions in Drosophila Dmel-2 culture cells (Figure 1A). We generated stably transformed cell lines expressing a transgene encoding a fusion of EGFP to the C-terminal 623 amino acids of the 1411 amino acid CENP-C (which we refer to as EGFP::CENPC-C). This transgene was under the control of the inducible metallothionein promoter to avoid potential dominant-negative phenotypes that might arise from constitutive expression. As expected [6], this CENPC-C fusion colocalized with the centromeric marker CENP-A/CID during the entire cell cycle (Figure 1B). In contrast, a fusion of EGFP with the N-terminal 788 amino acids of CENP-C, EGFP::CENPC-N, localized to nuclei but not to centromeres when expressed from the metallothionein promoter (Figure 1C). This suggests that this half of the protein must be transported actively into the nucleus, although a nuclear localization signal within this part was not identified previously [6]. Moreover, its lack of centromeric localization suggests that this part of CENP-C is incapable of either dimerizing with full-length endogenous protein [15, 16] or interacting with other centromeric components. This raised the question of what other proteins the N-terminal part of CENP-C might interact with. To address this, we built a stably transformed Drosophila cell line able to express a fusion of protein A::CENPC-N from the metallothionein promoter. After overnight induction of protein A::CENPC-N expression, we affinity purified the product of the transgene from cell extracts using IgG-coupled magnetic beads for analysis by mass spectrometry [17]. No peptides from the centromeric proteins CENP-A/CID [18], CAL1 [19], or the C-terminal part of CENP-C were identified, in accordance with the observation that CENPC-N does not associate with centromeres. Surprisingly, however, we were able to identify the entire set of kinetochore proteins belonging to the KMN (KNL1/Spc105, the Mis12 complex, and the Ndc80 complex) network [20], including Spc105, Mis12, Nnf1, Spc25/Mitch, Nuf2, and Ndc80 (Figure 1D). In addition, we identified proteins belonging to the 14-3-3 family with high Mascot scores [21]. The high levels of phosphorylation of CENP-C in cultured Drosophila cells and in Drosophila embryos [22, 23] would accord with being able to associate with 14-3-3 proteins that interact with phosphorylated targets [21]. None of the abovementioned proteins copurified with protein A alone (see Figure S1A available online).

Current Biology Vol 21 No 5 400

Figure 1. The Amino-Terminal Part of Drosophila CENP-C Does Not Localize to Centromeres but Binds the Entire KMN Network of Core Kinetochore Proteins (A) Schematic representation of the N- and C-terminal parts of CENP-C used in this study. (B) Localization of the EGFP-tagged C-terminal part of CENP-C (EGFP::CENPC-C) during interphase and mitosis. Cells were counterstained with anti-a-tubulin (MT), anti-CENP-A/CID antibodies, and DAPI (DNA). Scale bars in (B) and (C) represent 5 mm. (C) Localization of the EGFP-tagged N-terminal part of CENP-C (EGFP::CENPC-N) during interphase and mitosis. Cells were counterstained as in (B). (D) Proteins copurifying with the fusion protein A::CENPC-N with Mascot scores (Score) greater than 200. For CG numbers, refer to FlyBase entries (http://flybase.org/). The bait is indicated in red; KMN network components are indicated in blue. Spectral counts reflect the number of peptides identified for each protein. 1Note that coverage for the bait was scored as 35%. This figure refers to the full-length protein and corresponds to 63% coverage of the 788 amino acid fragment used as bait. (E) Nnf1a binds CENPC-N in vitro, whereas MBP and Nsl1 do not. Upper panels show autoradiography of the indicated S35-labeled proteins. Coomassie-stained gels are shown in the lower panels to confirm equal loading of N- or C-terminal parts of CENP-C. Lanes are labeled as follows: M, marker proteins; In, input from the in vitro transcription-translation reaction not incubated with beads; N, in vitro-synthesized product incubated with GST::CENPC-N on beads; C, in vitro-synthesized product incubated with GST::CENPC-C on beads. See also Figure S1.

Because we were not able to find any good candidate proteins that might provide a bridge between CENPC-N and KMN components, we speculated that the interaction might

be a direct one involving the innermost part of the kinetochore, the Mis12 complex. To test this possibility, we synthesized the 35S-labeled Mis12 complex subunits Nnf1a and Nsl1 and also maltose-binding protein (MBP) as a negative control in an in vitro transcription-translation system and investigated whether they would interact with bacterially expressed CENP-C fragments. (Mis12 appeared to be either insoluble or very poorly expressed in this system, and hence we could not test its interaction.) We did not detect any binding of MBP, Nnf1a, or Nsl1 to GST::CENPC-C. However, we did detect binding of Nnf1a to the N-terminal part of CENP-C, whereas Nsl1 binding was poorly detectable (Figure 1E). This suggests that this component of the Mis12 complex may be responsible for the association of the KMN network with CENP-C. It would accord with a recent study that places Nnf1, of all the Mis12 complex components, as lying most proximal to the centromeric chromatin [24].

KMN Proteins Assemble on N-Terminal Part of CENP-C 401

Our finding that the N-terminal part of CENP-C can associate with all proteins of the KMN network led us to speculate about its potential role in kinetochore assembly and test whether it would be sufficient to recruit the KMN network and other kinetochore proteins to ectopic sites. We chose to target this part of CENP-C to centrosomes and determine whether the KMN network would be recruited to the same site. To this end, we fused a centrosome-targeting domain from Plk4 kinase [11] in frame with the EGFP::CENPC-N construct (we termed this entire construct ect-KTR, for ectopic kinetochore recruitment). As a control, we also examined the localization and recruitment capabilities of the Plk4 centrosome localization domain (CLD) alone fused with EGFP (Figure 2A). Expression of either construct in Dmel-2 cells from an inducible promoter resulted in the localization of EGFP at the centrosome, confirmed by costaining with the centrosomal markers dPLP/Cp309 (the Drosophila counterpart of pericentrin) and dSpd2/Cep192 (Figures 2B and 2C). The ect-KTR fusion protein also localized to the nuclei of a substantial subset of cells during interphase, most likely reflecting the activity of its own strong nuclear localization signal (Figure S1B). However, we chose to focus on localization of these proteins in mitotic cells when kinetochores have normally assembled at the centromeres, thus allowing us to examine the extent to which kinetochore proteins localized to either centromeres or centrosomes. Because expression of either CENPC-N or ect-KTR caused frequent and severe mitotic phenotypes in which chromosomes became scattered so as often to lie in the vicinity of the poles (see below), we focused on examining the localization of kinetochore proteins in cells where centrosomes and centromeres were well separated (Figure 2; Figure 3). We then stained cells with several antibodies raised against kinetochore proteins (see Supplemental Experimental Procedures and Figure S2 for details). Expression of the CLD alone did not lead to any change of the localization of kinetochore markers. Kinetochore proteins belonging to the KMN network colocalized exclusively with CENP-A/CID in such cells (upper panels in Figures 2B–2E). Expression of ect-KTR, on the other hand, led to the strong recruitment of all KMN network components tested to the centrosome and the frequent diminution of the levels of these proteins at the kinetochore. Thus, immunostaining clearly revealed Mis12, Nsl1 (data not shown), Nnf1a, and Spc105 to colocalize with ect-KTR and the centrosomal markers dPlp and Spd2 (lower panels in Figures 2B–2D). To determine the localization of the outer kinetochore plate Ndc80 complex, we cotransfected Dmel-2 cells with constructs expressing RFP-tagged ect-KTR and either EGFP-tagged Ndc80 (Figure 2E) or Nuf2 (data not shown). We found in both cases that these components of the Ndc80 complex localized to centrosomes in ect-KTR-expressing cells. In contrast, neither of the centromeric markers CAL1 or CENP-A/CID colocalized with ect-KTR; they were found only at centromeres (Figure 2F). Thus, the N-terminal part of CENP-C is not able to interact with centromeric proteins but is able to recruit the KMN network to an ectopic site in the cell. We also asked whether, in addition to the core KMN network proteins, other ancillary proteins known to associate temporarily with kinetochores might be relocated onto ectopically positioned CENPC-N. We chose to examine the localization of the kinetochore-associated motor protein CENP-meta (the Drosophila counterpart of CENP-E) [25] and the spindle assembly checkpoint proteins Mad2 and BubR1. We found that, unlike the core KMN network proteins that were associated with centrosomes in all cells expressing ect-KTR,

ancillary kinetochore proteins were only associated with centrosomes in a subset of ect-KTR-expressing cells. Thus, CENP-meta was predominantly centrosome associated in 78.8% of ect-KTR cells (Figure 3A). Mad2 also showed similar localization to ectopic sites (Figure 3B) regardless of whether or not the checkpoint was activated by colchicine treatment (data not shown). We did not see strong signals of Mad2 on scattered chromosomes in ect-KTR cells. This could be due to, for example, either depletion of a proper platform for binding of Mad2 on kinetochores such as the Ndc80 complex or the titration of Mad2 to centrosomes. This observation supports the idea that Mad2 binds to kinetochores, not centromeres. On the contrary, in the case of BubR1, its association with centrosomes in a subset of ect-KTR-expressing cells was much weaker than its association with centromeres in the same cells (arrows in the lower panel of Figure 3C). This is in agreement with a recent report [26] and our own data [27] suggesting that BubR1 may have at least two independent binding sites: one on centromeres, which might be more exposed in cells expressing ect-KTR, and the other on kinetochores, which in cells expressing ect-KTR would be associated with centrosomes. In fact, this result supports the idea of BubR1 being recruited to centromeres rather than kinetochores. Because none of these ancillary kinetochore proteins associated with centrosomes in CLD-expressing cells, we concluded that the ectopic localization of the N-terminal part of CENP-C was responsible for localizing the core KMN network to centrosomes, and that this then permitted recruitment of other kinetochore-associated proteins. As indicated above, we found frequent mitotic defects in cells expressing CENPC-N and ect-KTR, of which improperly congressed and scattered chromosomes were the most commonly observed aberrations (Figures 4A and 4B). The scattering of chromosomes on elongated spindles was reminiscent of the phenotype we observed after knockdown of components of the KMN network [9]. We observed that whereas spindles in CLD-expressing cells had a mean length of 6.83 6 0.78 mm (n = 25) that is typical of the parent Dmel-2 cell line [9], spindles in ect-KTR and CENPC-N cells were significantly elongated and had mean lengths of 8.7 6 1.93 mm (n = 45) and 8.04 6 1.27 mm (n = 41), respectively (Figure 4C). To determine whether cells with elongated spindles were in anaphase, we costained cells with anti-CENP-A/CID and anti-cyclin B antibodies. This revealed that ect-KTR cells showing the mitotic phenotype had high levels of cyclin B and that the chromosomes had two centromeres and so were conjoined chromatids (Figure S3). This strongly suggests that these cells were blocked in prometaphase with improperly congressed chromosomes and had not entered anaphase. Therefore, the chromosome scattering and the spindle elongation phenotype may be a consequence of the very long prometaphase block occurring as a result of the primary defect: severe congression problems. We hypothesized that the high incidence of mitotic defects present after ect-KTR or CENPC-N expression was caused by the titration out of one or more KMN network proteins by the ectopic construct. This interpretation was supported by the observation that immunolocalization signals of kinetochore proteins ectopically localized at centrosomes were usually stronger than the signals at centromeres in ect-KTR cells (Figures 2B–2E). To address this directly, we quantified fluorescence intensities of endogenous kinetochores after staining with anti-Mis12 and anti-Spc105 antibodies in relation to the intensities of the centromeric CENP-A/CID in CLD-, ect-KTR-, and CENPC-N-expressing cells. In ect-KTR- and

Current Biology Vol 21 No 5 402

Figure 2. The CLD-Tagged N-Terminal Part of CENP-C Localizes to Centrosomes and Recruits KMN Network Kinetochore Components (A) Schematic representation of the CLD construct, a fusion of EGFP with the centrosome-binding domain of SAK/Plk4, and the ect-KTR construct, a fusion of EGFP, the N-terminal part of CENP-C, and the same centrosome-binding domain. Both constructs localize to centrosomes when expressed in Dmel-2 cells. (B) Cells expressing CLD or ect-KTR (EGFP, green) stained to reveal the dPlp centrosomal marker (red) and Mis12 (blue). DNA staining (at bottom, not merged) is shown only for ect-KTR. (C) Cells expressing CLD or ect-KTR stained to reveal the centrosomal marker Spd2 (blue), the kinetochore proteins Nnf1a and Nnf1b (red), and the centromeric marker CENP-A/CID (green). (D) Cells expressing CLD or ect-KTR stained to reveal the centrosomal marker Spd2 (blue), the kinetochore protein Spc105 (red), and the centromeric marker CENP-A/CID (green). (E) Cells transfected with RFP-CLD (red) or RFP-ect-KTR (red) were cotransfected with EGFP::Ndc80 (green) and stained to reveal a-tubulin (MT; blue) and DNA (unmerged). (F) Cells transfected with RFP-CLD (red) or RFP-ect-KTR (red) were cotransfected with CAL1::EGFP (green). CAL1::EGFP always colocalizes with CENP-A/ CID (blue) on DNA (not merged) and does not colocalize with red signal of centrosomally targeted fusion proteins. Scale bars represent 5 mm. Note that in this part of the study, we focused upon ect-KTR-expressing cells that did not show excessive scattering of chromosomes (see Figure 4 and associated text). This allowed us to ensure that any signal in the vicinity of the spindle poles was truly centrosome associated (and not centromere associated). See also Figure S2.

KMN Proteins Assemble on N-Terminal Part of CENP-C 403

Figure 3. Localization of CENP-meta and the Spindle Assembly Checkpoint Proteins Mad2 and BubR1 in CLD- and ect-KTR-Expressing Cells (A) Cells expressing CLD (green) or ect-KTR (green) and stained to reveal CENP-A/CID (red) show localization of CENP-meta (blue) to ectopic sites in ect-KTR-expressing cells but not in CLDexpressing cells. (B) Cells expressing CLD (green) or ect-KTR (green) and stained to reveal DNA (blue) and Mad2 (red) show localization of Mad2 to ectopic sites in ect-KTR-expressing cells but not in CLD-expressing cells. (C) Cells expressing CLD or ect-KTR stained to reveal the kinetochore protein Spc105 (blue), the centromeric protein CENP-A/CID (green), and BubR1 (red). Staining of phosphohistone H3 is shown unmerged. Arrows indicate spindle poles in ect-KTR-expressing cells that have strong staining of Spc105 (blue) and weak staining of BubR1 (red). Stronger BubR1 staining is present on centromeres of such cells. Note that EGFP fluorescence from the CLD construct is much weaker than the signal from the fluorophore of the secondary antibody used for CENP-A/CID staining (the slide was bleached for some time to assure this effect). Arrowheads indicate such weak signals (‘‘+ GFP’’ panel), which assure that the imaged cell is in fact transgenic. Scale bars represent 5 mm.

CENPC-N-expressing cells, Mis12 levels at the kinetochore were reduced respectively to an average of 46% and 26.8% of the level in cells expressing the CLD construct alone. Similarly, levels of Spc105 were reduced to 29.2% and to 39.1% in ect-KTR and CENPC-N cells respectively (see distribution of signal intensities in Figure 4D). Thus, we conclude that overexpression of ect-KTR or nontargeted CENPC-N indeed results in a depletion of kinetochore proteins from endogenous kinetochores, and that this in turn leads to mitotic defects. Although the vast majority of CLD-expressing cells looked normal (Figure 4A), we noticed a low level of background aberrations that may be attributable to the expression of the Plk4 fragment, which is able to drive production of extra centrosomes (Q.D. Yu, N. Dzhindzhev, and D.M.G., unpublished data) and which may affect the efficiency of spindle assembly (Figure 4B). However, the finding of chromosomes scattered along elongated spindles was unique to CENPC-N- or ectKTR-expressing cells and was not seen in CLD-expressing cells. This strong dominant-negative phenotype and the

ectopic localization pattern of KMN network proteins in cells expressing ect-KTR lead to the unequivocal conclusion that the N-terminal domain of CENP-C is sufficient to recruit all of the major core kinetochore proteins, regardless of its localization in the cell. Recent studies from a number of laboratories have provided a detailed description of the interaction network between subunits of the KMN supercomplex [20, 24, 28, 29]. This knowledge is key for reaching a full understanding of how kinetochores are formed early in mitosis and how they interact with microtubules of the mitotic spindle. Considerable progress has also been made in studies of centromere inheritance, composition, and function (e.g., [30–33]). However, little is known yet about the interaction between these two complex structures. Previously, we reported that kinetochore formation in Drosophila cells is absolutely dependent upon the presence of the centromeric protein CENP-C [9], a finding recently confirmed by others [10]. The results we present here indicate that it is the N-terminal part of this protein that not only is responsible for binding to the KMN network but also is sufficient to recruit it to ectopic sites. The consequential depletion of KMN proteins from the centromeres of mitotic chromosomes to centrosomes leads to strong mitotic defects. Together, our data support a model in which one face of CENP-C binds CENP-A-rich centromeric chromatin and the other binds components of the KMN network, providing a platform for kinetochore assembly at the onset of mitosis. That such a function of CENP-C may be conserved is strongly suggested from a contemporaneous

Current Biology Vol 21 No 5 404

Figure 4. Sequestering of KMN Network Components to the Centrosomally Localized N-Terminal Part of CENP-C Leads to Chromosome Alignment Defects in Mitosis (A) Cells expressing CLD ect-KTR or EGFP::CENPC-N (green), counterstained with anti-a-tubulin antibody (red) and anti-phosphohistone H3 (pH3) or DAPI to reveal DNA (blue). Scale bars represent 5 mm. (B) Quantification of improperly congressed or scattered chromosomes in CLD-, ect-KTR-, and CENPC-N-expressing cells. 18% of the CLD-expressing cells (n = 130), 74.9% of the ect-KTR expressing cells (n = 177), and 75.6% of the CENPC-N-expressing cells (n = 41) showed chromosome positioning or spindle defects. (C) Correlation of spindle length with chromosome congression and/or scattering defects in CLD, ect-KTR, and CENPC-N cells. Mean spindle lengths are CLD cells, 6.83 6 0.78 mm (n = 25); ect-KTR cells, 8.7 6 1.93 mm (n = 45); CENPC-N cells, 8.04 6 1.27 mm (n = 41). The trend shows that longer spindles correlate with a higher frequency of chromosome congression defects. (D) Depletion of Mis12 and Spc105 proteins from endogenous kinetochores. Kinetochore protein staining intensity values (background subtracted) were divided by the intensity values of CENP-A/CID staining at the corresponding centromere (background subtracted). Mean values for Mis12 are CLD, 0.92 6 0.4 (n = 64); ect-KTR, 0.42 6 0.34 (n = 78); CENPC-N, 0.25 6 0.2 (n = 85). Mean values for Spc105 are CLD, 0.92 6 0.33 (n = 61); ect-KTR, 0.27 6 0.24 (n = 83); CENPC-N, 0.36 6 0.27 (n = 114). See text for description and interpretation. See also Figure S3.

study of its interacting partners in human cells, also published in this issue of Current Biology [34]. The length of the CENP-C molecule, the bifunctionality of its binding, and its potential for posttranslational modification together suggest that this protein might not only provide structural support for kinetochore assembly but also be a major hub for regulating

centromere and kinetochore function. It will be important to learn more about such potential regulatory mechanisms, and also to determine whether CENP-C is the only component of the centromere to provide an interface between centromeres and kinetochores in either Drosophila or vertebrates.

KMN Proteins Assemble on N-Terminal Part of CENP-C 405

Supplemental Information Supplemental Information includes three figures, one table, and Supplemental Experimental Procedures and can be found with this article online at doi:10.1016/j.cub.2011.02.005. Acknowledgments We would like to thank members of D.M.G.’s laboratory for helpful discussions. We are grateful to Nikola Dzhindzhev for the data from the ‘‘protein A alone’’ control experiment and Monica Bettencourt-Dias for the antiSpd2 antibody. We thank Don Cleveland and David Sharp for kind gifts of antibodies and Ronald Vale for the cDNA clone. This project was financially supported by grants from Cancer Research UK and the Biotechnology and Biological Sciences Research Council. Received: November 11, 2010 Revised: January 7, 2011 Accepted: February 3, 2011 Published online: February 24, 2011 References 1. Przewloka, M.R., and Glover, D.M. (2009). The kinetochore and the centromere: A working long distance relationship. Annu. Rev. Genet. 43, 439–465. 2. Santaguida, S., and Musacchio, A. (2009). The life and miracles of kinetochores. EMBO J. 28, 2511–2531. 3. Saitoh, H., Tomkiel, J., Cooke, C.A., Ratrie, H., 3rd, Maurer, M., Rothfield, N.F., and Earnshaw, W.C. (1992). CENP-C, an autoantigen in scleroderma, is a component of the human inner kinetochore plate. Cell 70, 115–125. 4. Kwon, M.S., Hori, T., Okada, M., and Fukagawa, T. (2007). CENP-C is involved in chromosome segregation, mitotic checkpoint function, and kinetochore assembly. Mol. Biol. Cell 18, 2155–2168. 5. Fukagawa, T., and Brown, W.R. (1997). Efficient conditional mutation of the vertebrate CENP-C gene. Hum. Mol. Genet. 6, 2301–2308. 6. Heeger, S., Leismann, O., Schittenhelm, R., Schraidt, O., Heidmann, S., and Lehner, C.F. (2005). Genetic interactions of separase regulatory subunits reveal the diverged Drosophila Cenp-C homolog. Genes Dev. 19, 2041–2053. 7. Tanaka, K., Chang, H.L., Kagami, A., and Watanabe, Y. (2009). CENP-C functions as a scaffold for effectors with essential kinetochore functions in mitosis and meiosis. Dev. Cell 17, 334–343. 8. Milks, K.J., Moree, B., and Straight, A.F. (2009). Dissection of CENP-Cdirected centromere and kinetochore assembly. Mol. Biol. Cell 20, 4246–4255. 9. Przewloka, M.R., Zhang, W., Costa, P., Archambault, V., D’Avino, P.P., Lilley, K.S., Laue, E.D., McAinsh, A.D., and Glover, D.M. (2007). Molecular analysis of core kinetochore composition and assembly in Drosophila melanogaster. PLoS ONE 2, e478. 10. Orr, B., and Sunkel, C.E. (2011). Drosophila CENP-C is essential for centromere identity. Chromosoma 120, 83–96. 11. Dzhindzhev, N.S., Yu, Q.D., Weiskopf, K., Tzolovsky, G., Cunha-Ferreira, I., Riparbelli, M., Rodrigues-Martins, A., Bettencourt-Dias, M., Callaini, G., and Glover, D.M. (2010). Asterless is a scaffold for the onset of centriole assembly. Nature 467, 714–718. 12. Marshall, O.J., Marshall, A.T., and Choo, K.H. (2008). Three-dimensional localization of CENP-A suggests a complex higher order structure of centromeric chromatin. J. Cell Biol. 183, 1193–1202. 13. Allshire, R.C., and Karpen, G.H. (2008). Epigenetic regulation of centromeric chromatin: Old dogs, new tricks? Nat. Rev. Genet. 9, 923–937. 14. Trazzi, S., Perini, G., Bernardoni, R., Zoli, M., Reese, J.C., Musacchio, A., and Della Valle, G. (2009). The C-terminal domain of CENP-C displays multiple and critical functions for mammalian centromere formation. PLoS ONE 4, e5832. 15. Sugimoto, K., Kuriyama, K., Shibata, A., and Himeno, M. (1997). Characterization of internal DNA-binding and C-terminal dimerization domains of human centromere/kinetochore autoantigen CENP-C in vitro: Role of DNA-binding and self-associating activities in kinetochore organization. Chromosome Res. 5, 132–141. 16. Carroll, C.W., Milks, K.J., and Straight, A.F. (2010). Dual recognition of CENP-A nucleosomes is required for centromere assembly. J. Cell Biol. 189, 1143–1155.

17. D’Avino, P.P., Archambault, V., Przewloka, M.R., Zhang, W., Laue, E.D., and Glover, D.M. (2009). Isolation of protein complexes involved in mitosis and cytokinesis from Drosophila cultured cells. Methods Mol. Biol. 545, 99–112. 18. Blower, M.D., and Karpen, G.H. (2001). The role of Drosophila CID in kinetochore formation, cell-cycle progression and heterochromatin interactions. Nat. Cell Biol. 3, 730–739. 19. Erhardt, S., Mellone, B.G., Betts, C.M., Zhang, W., Karpen, G.H., and Straight, A.F. (2008). Genome-wide analysis reveals a cell cycle-dependent mechanism controlling centromere propagation. J. Cell Biol. 183, 805–818. 20. Cheeseman, I.M., Chappie, J.S., Wilson-Kubalek, E.M., and Desai, A. (2006). The conserved KMN network constitutes the core microtubulebinding site of the kinetochore. Cell 127, 983–997. 21. Dougherty, M.K., and Morrison, D.K. (2004). Unlocking the code of 14-33. J. Cell Sci. 117, 1875–1884. 22. Bodenmiller, B., Malmstrom, J., Gerrits, B., Campbell, D., Lam, H., Schmidt, A., Rinner, O., Mueller, L.N., Shannon, P.T., Pedrioli, P.G., et al. (2007). PhosphoPep—a phosphoproteome resource for systems biology research in Drosophila Kc167 cells. Mol. Syst. Biol. 3, 139. 23. Zhai, B., Ville´n, J., Beausoleil, S.A., Mintseris, J., and Gygi, S.P. (2008). Phosphoproteome analysis of Drosophila melanogaster embryos. J. Proteome Res. 7, 1675–1682. 24. Petrovic, A., Pasqualato, S., Dube, P., Krenn, V., Santaguida, S., Cittaro, D., Monzani, S., Massimiliano, L., Keller, J., Tarricone, A., et al. (2010). The MIS12 complex is a protein interaction hub for outer kinetochore assembly. J. Cell Biol. 190, 835–852. 25. Yucel, J.K., Marszalek, J.D., McIntosh, J.R., Goldstein, L.S., Cleveland, D.W., and Philp, A.V. (2000). CENP-meta, an essential kinetochore kinesin required for the maintenance of metaphase chromosome alignment in Drosophila. J. Cell Biol. 150, 1–11. 26. Schittenhelm, R.B., Chaleckis, R., and Lehner, C.F. (2009). Intrakinetochore localization and essential functional domains of Drosophila Spc105. EMBO J. 28, 2374–2386. 27. Venkei, Z., Przewloka, M.R., and Glover, D.M. (2011). Drosophila Mis12 complex acts as a single functional unit essential for anaphase chromosome movement and a robust spindle assembly checkpoint. Genetics 187, 131–140. 28. Ciferri, C., Pasqualato, S., Screpanti, E., Varetti, G., Santaguida, S., Dos Reis, G., Maiolica, A., Polka, J., De Luca, J.G., De Wulf, P., et al. (2008). Implications for kinetochore-microtubule attachment from the structure of an engineered Ndc80 complex. Cell 133, 427–439. 29. Maskell, D.P., Hu, X.W., and Singleton, M.R. (2010). Molecular architecture and assembly of the yeast kinetochore MIND complex. J. Cell Biol. 190, 823–834. 30. Foltz, D.R., Jansen, L.E., Black, B.E., Bailey, A.O., Yates, J.R., 3rd, and Cleveland, D.W. (2006). The human CENP-A centromeric nucleosomeassociated complex. Nat. Cell Biol. 8, 458–469. 31. Hori, T., Amano, M., Suzuki, A., Backer, C.B., Welburn, J.P., Dong, Y., McEwen, B.F., Shang, W.H., Suzuki, E., Okawa, K., et al. (2008). CCAN makes multiple contacts with centromeric DNA to provide distinct pathways to the outer kinetochore. Cell 135, 1039–1052. 32. Amaro, A.C., Samora, C.P., Holtackers, R., Wang, E., Kingston, I.J., Alonso, M., Lampson, M., McAinsh, A.D., and Meraldi, P. (2010). Molecular control of kinetochore-microtubule dynamics and chromosome oscillations. Nat. Cell Biol. 12, 319–329. 33. Ohta, S., Bukowski-Wills, J.C., Sanchez-Pulido, L., Alves, Fde.L., Wood, L., Chen, Z.A., Platani, M., Fischer, L., Hudson, D.F., Ponting, C.P., et al. (2010). The protein composition of mitotic chromosomes determined using multiclassifier combinatorial proteomics. Cell 142, 810–821. 34. Screpanti, E., De Antoni, A., Alushin, G.M., Petrovic, A., Melis, T., Nogales, E., and Musacchio, A. (2011). Direct binding of Cenp-C to the Mis12 complex joins the inner and outer kinetochore. Curr. Biol. 20, this issue, 391–398.