CENP-T Proteins Are Conserved Centromere Receptors of the Ndc80 Complex

ARTICLES

CENP-T proteins are conserved centromere receptors of the Ndc80 complex

Alexander Schleiffer1, Michael Maier2, Gabriele Litos2, Fabienne Lampert2, Peter Hornung2, Karl Mechtler2 and Stefan Westermann2,3

Centromeres direct the assembly of kinetochores, microtubule-attachment sites that allow chromosome segregation on the mitotic spindle. Fundamental differences in size and organization between evolutionarily distant eukaryotic centromeres have in many cases obscured general principles of their function. Here we demonstrate that centromere-binding proteins are highly conserved between budding yeast and humans. We identify the histone-fold protein Cnn1CENP-T as a direct centromere receptor of the microtubule-binding Ndc80 complex. The amino terminus of Cnn1 contains a conserved peptide motif that mediates stoichiometric binding to the Spc24–25 domain of the Ndc80 complex. Consistent with the critical role of this interaction, artiﬁcial tethering of the Ndc80 complex through Cnn1 allows mini-chromosomes to segregate in the absence of a natural centromere. Our results reveal the molecular function of CENP-T proteins and demonstrate how the Ndc80 complex is anchored to centromeres in a manner that couples chromosome movement to spindle dynamics.

Kinetochores are large assemblies of multiple-protein complexes (Sim4 complex). Although the evolutionary conservation of the KMN exceeding 60 proteins in budding yeast and more than 100 components network has been well documented15, phylogenetic relationships in humans1–3. As essential cell-division organelles, kinetochores allow between centromere proteins have remained uncertain. Centromere the mitotic spindle to drive chromosome segregation by productively DNA is highly divergent between species, bearing no resemblance coupling microtubule plus-end dynamics to the separation of in length or sequence between evolutionarily distant yeasts such as sister chromatids. In addition, kinetochores contain error-correction Saccharomyces cerevisiae and Schizosaccharomyces pombe and between and signalling mechanisms that relay the establishment of correct yeast and humans. The simplest centromeres are found in S. cerevisiae, kinetochore bi-orientation on the spindle to the onset of anaphase4,5. comprising only 125 base pairs of DNA (point centromere) with three Kinetochores have the following general building plan: the KMN sequence elements that are conserved between the 16 chromosomes15,16. network (consisting of the four-protein Mtw1 complex, the four- The centromeres of S. pombe are regional, built on more complex protein Ndc80 complex and KNL-1 in complex with Zwint) is a DNA that contains repeated sequence elements and has the hallmarks supramolecular assembly that forms the architectural core of the of heterochromatin. Centromeric DNA in human cells spans over kinetochore and contacts spindle microtubules directly6,7. By means many megabases and contains α-satellite repeats that are, however, that are not well understood, the KMN network is anchored to neither necessary nor sufficient to initiate kinetochore assembly17. centromeric chromatin that is characterized by the presence of a These marked differences in centromere organization have led to specialized nucleosome that contains the histone variant CENP-A the notion that budding yeast may employ many proteins dedicated replacing the canonical histone H3. The conserved DNA-binding to the sequence-determined point centromere and thus may have protein CENP-C closely associates with and probably directly a fundamentally different kinetochore architecture when compared recognizes CENP-A nucleosomes8,9 to provide an essential step in with higher eukaryotes. kinetochore assembly. In addition, a group of at least 15 polypeptides, Here we reveal a number of previously undetected homology in humans termed the centromere constitutive-associated network relationships between yeast and human centromere proteins leading (CCAN) co-purify specifically with CENP-A nucleosomes10,11. Similar to the conclusion that structurally similar kinetochores assemble proteomic approaches have identified centromere proteins in budding on small point and large regional centromeres. We show that the yeast12,13 (collectively called the Ctf19 complex) and fission yeast14 histone-fold protein Cnn1CENP-T provides an evolutionarily conserved

1IMP/IMBA Bioinformatics Core Facility, Research Institute of Molecular Pathology (IMP), Dr. Bohr Gasse 7, 1030 Vienna, Austria. 2Research Institute of Molecular Pathology (IMP), Dr. Bohr Gasse 7, 1030 Vienna, Austria. 3Correspondence should be addressed to S.W. (e-mail: [email protected])

Received 29 September 2011; accepted 30 March 2012; published online 6 May 2012; DOI: 10.1038/ncb2493

ARTICLES

a b BLASTP Homo sapiens CENP-T 2 × 10–21 Xenopus laevis 4 × 10–20 Danio rerio Homo sapiens 561 aa 4 × 10–16 Ciona intestinalis 1 × 10–7 Ajellomyces capsulatus 4 × 10–5 Schizosaccharomyces pombe Cnp20 8 × 10–6 Yarrowia lipolytica Schizosaccharomyces pombe 479 aa HMMer 1.5 × 10–14 Pichia (Komagataella) pastoris HMMer

Fungi–Ascomycota 3.2 × 10–12 Candida parapsilosis HMMer Saccharomyces cerevisiae 361 aa 2.5 × 10–10 Saccharomyces kluyveri Saccharomycetes Chordata BLASTP 5 × 10–12 Saccharomyces cerevisiae Cnn1

c Histone fold CENP-T extension

2D structure

Homo s. 457 553 Mus m. 411 507 Gallus g. 534 631 Xenopus l. 662 759 Danio r. 807 904 Ciona i. 438 535 Ustilago m. 909 1031 Schizosaccharomyces p. 384 479 Talaromyces s. 365 468 Ajellomyces c. 362 464 Magnaporthe o. 448 551 Tuber m. 408 505 Pyrenophora t. 427 529 Yarrowia l. 327 424 Pichia p. 269 374 Kluyveromyces l. 147 254 Candida g. 229 333 Lachancea t. 203 305 Candida a. 624 746 Debaryomyces h. 491 617 Candida p. 682 801 Saccharomyces k. 249 351 Zygosaccharomyces r. 221 321 Ashbya g. 200 300 Vanderwaltozyma p. 254 351 Saccharomyces c. 271 361 2D structure

Figure 1 Identiﬁcation of Cnn1 as a budding yeast CENP-T homologue. regions are indicated in grey. (c) Multiple alignment of the histone fold and (a) Summary of the search strategy to deduce phylogenetic relationships the C-terminal extension of the Cnn1/CENP-T protein family. H. sapiens between human CENP-T and S. cerevisiae Cnn1. A series of BLASTP and (above) and S. cerevisiae (below) secondary-structure predictions are derived HMM searches using sequences of the conserved domain (see below) were from the Jpred 3 server (α-helical regions are in red; β-strands in green46). applied. (b) Graphical representation of Cnn1/CENP-T sequence features, Species and accession numbers are listed in Supplementary Table S2. showing the location of the N-terminal motif (orange), the histone fold (blue) Background colouring of the residues is based on the clustalx colouring and the C-terminal CENP-T family speciﬁc extension (green). Low-complexity scheme. Brackets indicate the number of omitted residues. direct link to the Ndc80 complex, the key microtubule-binding activity Ctf19 complex subunit Mcm22 is orthologous to human CENP-K of the kinetochore. (Supplementary Fig. S3 and Methods).

RESULTS Identiﬁcation of fungal orthologues of the human kinetochore Evolutionary conservation of centromere-proximal proteins protein CENP-T between yeast and humans Recent studies have demonstrated a critical role for the DNA-binding To examine whether centromere-binding proteins of evolutionarily activity of the human CENP-T/W complex in kinetochore function20 distant organisms share common ancestry, we performed a comprehen- and together with CENP-C, CENP-T plays an instructive role in sive bioinformatic analysis using sensitive remote homology searches kinetochore assembly21. Human CENP-T is 561 residues long, but in combination with secondary-structure predictions. Our analysis besides a histone domain at the carboxy terminus20,22, most of indicates a remarkable conservation of the centromere–kinetochore the protein consists of regions of compositional bias, rich in polar interface across eukaryotes (Table 1). We detected previously un- residues. On the sequence level, generally the histone fold is highly recognized homology relationships between the essential budding variable—pairwise identities between archaeal and eukaryotic histones yeast proteins Ame1 and Okp1 and the human CCAN subunits range from 15 to 20%, but the structure is highly defined, consisting of CENP-U and CENP-Q, respectively (Supplementary Figs S1 and three α-helices separated by two β-strand loops23–26. A single BLASTP S2 and Methods). Importantly, this indicates that the budding search identified orthologues in chordates and fungi with significant yeast COMA (Ctf19–Okp1–Mcm21–Ame1) complex may have a E values (<4 × 10−5), among those, the known S. pombe CENP-T biochemical counterpart in the human CENP-O/P/Q/U complex, homologue Cnp20 (SPBC800.13; ref. 27) and the uncharacterized which also forms a stable entity18,19. In addition, we find that the yeast protein YALI0A19162p in the Saccharomycetes Yarrowia lipolytica.

ARTICLES

followed by mass spectrometry to identify associated proteins (Fig. 2a). Table 1 Homology relationships between centromere proteins from human, CENP-U CENP-N CENP-H S. pombe and S. cerevisiae based on sequence analysis. For Ame1 , Chl4 and Mcm16 , this approach yielded H. sapiens S. pombe S. cerevisiae CCAN module a similar interaction profile with each protein co-purifying all other members of the Ctf19 complex and all four subunits of the Mtw1 CENP-C Cnp3 Mif2 C complex, demonstrating that the Mtw1 complex provides an important CENP-O Mal2 Mcm21 O/P/Q/U/R attachment point between CCAN and KMN components. Remarkably, CENP-P Fta2 Ctf19 CENP-T CENP-Q Fta7 Okp1 purification of Cnn1 resulted in a different set of co-purifying CENP-U Mis17 Ame1 polypeptides: whereas most members of the Ctf19 complex could CENP-R be detected, assigning Cnn1 as a CCAN component, Mtw1 complex CENP-H Fta3 Mcm16 H/I/K subunits were absent from this purification and instead all four CENP-I Mis6 Ctf3 CENP-K Sim4 Mcm22 subunits of the Ndc80 complex were co-purified. This indicated that Cnn1CENP-T might form a unique point of contact between the inner CENP-T SPBC800.13/Cnp20 Cnn1 CCAN-HF CENP-H CENP-T CENP-W SPAC17G8.15 YDR374W-A/Wip1 and outer kinetochore. In Mcm16 and Cnn1 purifications CENP-S SPBC2D10.16 YOL086W-A/Mhf1 we additionally detected the product of an uncharacterized open CENP-X SPCC576.12c YDL160C-A/Mhf2 reading frame, YDR374W-A, encoding an 89-amino-acid protein with CENP-M MNL a predicted high α-helical content. A GFP-fusion to YDR374W-A CENP-N Mis15 Chl4 CENP-L Fta1 Iml3 (Mcm19) co-localized with the kinetochore marker Nuf2 and this localization was abolished in a cnn1-deletion mutant (Fig. 2b). Recombinant Fta4 Nkp1 Cnl2 Nkp2 YDR374W-A expressed in bacteria formed a complex with the Fta5 C terminus of Cnn1, which contains the predicted histone fold (Fig. 2c). Fta6 Furthermore, chromatin immunoprecipitation (ChIP) in combination The grouping reﬂects biochemically distinct subcomplexes. For a comprehensive overview with quantitative real-time PCR showed both Cnn1 and YDR374W- of the CCAN subunits and their taxonomic distribution, see Supplementary Table S1. A to be enriched at centromeric chromatin (Fig. 2d). Sensitive homology searches could not enlarge the YDR374W-A protein family Reciprocal BLASTP searches confirmed the orthologous relationship outside the Saccharomycetes class, with the Debaryomyces hansenii between fungal and mammalian CENP-T proteins. Iterative PSI-BLAST protein DEHA2G10538g being the most distant homologue. However, searches with the conserved C-terminal domain first collected CENP-T DEHA2G10538g has also been identified in a CENP-W-derived proteins, but in further iterations included histone H4 and other HMM search (Methods). Taken together, these results indicate that histone-fold proteins. To search for a S. cerevisiae homologue, we YDR374W-A is the budding yeast homologue of human CENP- applied a series of sensitive hidden Markov model (HMM) profile W and we will refer to this protein as Wip1 (W-like protein 1, searches within specified proteomes, where we first identified a candi- Supplementary Fig. S4). date in Pichia pastoris (Pipas_chr3_0909, E value 1.5×10−14), then in In mammals, two further histone-fold proteins are known to Candida parapsilosis (CPAG_01066, E value 3.2×10−12) and then in interact with the core CCAN complex: CENP-S and CENP-X (refs 20, Saccharomyces kluyveri (SAKL0F01012g, E value 2.5×10−10; Fig. 1a). 28), forming CENP-S/CENP-X (ref. 29) and CENP-T/CENP-W In all proteomes, the second best hit was to histone H4 with highly heterodimeric complexes or SXTW heterotetrameric complexes22. significant E values as well (7.1×10−12, 7.1×10−9 and 1.2×10−8 for P. The yeast CENP-S and -X homologues Mhf1 and Mhf2 (refs 29,30) pastoris Pipas_c034_0035, C. parapsilosis CPAG_04945 and S. kluyveri co-purified each other, Mhf2 exhibited further associations with a SAKL0C10714g, respectively). The S. kluyveri CENP-T candidate subset of CCAN subunits and Mhf1 interacted with the DEAH box SAKL0F01012g was annotated to be weakly similar to S. cerevisiae Cnn1, helicase Mph1 and the DNA repair proteins Msh2 and 3, indicating and the homology was confirmed by BLASTP (E value 5×10−12). that Mhf1 and Mhf2 might have dual functions at centromeres and in Although the sequence identity on the primary amino-acid level DNA repair (Supplementary Table S3). is low, CENP-T proteins share a conserved architecture. At the To characterize the protein Wip1 further, we examined its C terminus, there is the conserved CENP-T domain, consisting of localization in deletion mutants of other CCAN subunits. Whereas a histone fold and two extra helices (Fig. 1b,c). The highly conserved the typical bi-lobed localization to kinetochore clusters was unchanged central helix of the histone fold spans around 29 residues. In S. cerevisiae in deletion mutants of mcm16CENP-H, iml3CENP-L and mcm21CENP-O, it Cnn1, this helix is considerably shorter (20 residues), and less conserved was abolished in a chl4CENP - N-deletion mutant, demonstrating that in in comparison with the adjacent non-histone-fold helices. Although addition to Cnn1, the protein critically depends on this CCAN subunit no direct phylogenetic relationship between CENP-T and Cnn1 can for its localization to kinetochores (Fig. 3a). In a reciprocal experiment be deduced, a common sequence architecture and experimental we investigated whether any other CCAN subunit depended on data, such as the biological significance of a conserved N-terminal the presence of Wip1 for kinetochore localization. We found that motif (see below), lead to the conclusion that both proteins are the localization of Mcm16CENP-H, Ctf19CENP-P and Chl4CENP-N was indeed orthologous. maintained in a wip1-deletion mutant (data not shown), but the intensity of the fluorescent signal of Cnn1 was severely reduced Proteomic analysis of the budding yeast CCAN by about 80%, consistent with the observed physical interaction To gain insight into the molecular organization of the budding yeast between Cnn1 and Wip1 (Fig. 3b). Cnn1 localization was disrupted CCAN network, we performed systematic tandem-affinity purifications to a similar extent in a chl4-deletion mutant (data not shown), as

ARTICLES a b YOL086W-A YDL160C-A Ame1–TAP Chl4–TAP Mcm16–TAP Cnn1–TAP –TAP –TAP (CENP-U) (CENP-N) (CENP-H) (CENP-T) (CENP-S) (CENP-X)

High Low High Low High Low High Low High Low High Low Ndc80 + Nuf2 + cnn1Δ NDC80-C Spc24 + Spc25 +

Mtw1 ++ ++ ++ KMN MTW1-C Dsn1 ++ ++ ++ Nnf1 ++ ++ ++ Nsl1 ++ ++ ++ Spc105 + Wip1–GFP Spc105-C YDR532c

Ctf19 ++ ++ ++ ++ Mcm21 + + + +++ Okp1 ++ ++ ++ + + Ame1 ++ ++ ++++ + + Ctf3 ++ ++ ++ Mcm16 ++ ++ ++++ + Mcm22 ++ ++ ++ + Chl4 ++ ++ ++ + + Nuf2–mCherry Iml3 ++ ++ ++ + + CCAN Nkp1 ++ ++ ++ + + Nkp2 ++ ++ ++ +

Cnn1 +++ + YOL086W-A + +++ + YDL160C-A +++++ YDR374W-A + + Cse4 ++ Mif2 ++

c d Fraction M (K)

r Ni-NTA A8 A9 A10 A11 A12 B1 B2 B3 0.15 CEN3 Superose 6 PC3.2/30 70 GAL2 55 35 0.10 25

15 IP/total input 0.05 Cnn1264–361–6×His 10 Wip1

0 Cse 4 Mif2 Cnn1 Wip1 Elution volume

Figure 2 Proteomic analysis of the yeast CCAN and identification of the Analytical size-exclusion chromatography of Wip1 co-expressed with Cnn1 subunit Wip1. (a) Summary of the systematic tandem-affinity purifications (residues 264–361) in bacteria after initial purification using Ni-NTA (TAPs) of CCAN components. ‘+’ denotes that a protein was co-purified agarose. (d) ChIP with quantitative PCR was performed for Cse4–13xMyc, with the respective bait; ‘high’ and ‘low’ denote stringency conditions Mif2–13xMyc, Cnn1–13xMyc and Wip1–13xMyc using primers spanning a during purification of 300 mM or 100 mM KCl, respectively. (b) Live-cell centromeric (CEN3) and a non-centromeric region (GAL2, chromosome XII). microscopy of Wip–GFP in wild-type cells or in a cnn1-deletion background. The no-antibody control/total-input control has been subtracted for each Scale bar, 5 µm. (c) Wip1 forms a complex with the histone fold of Cnn1. experiment. Data represent the mean of ≥ 2 independent experiments. predicted from the observation that Wip1 depended on Chl4 for observed between Cnn1–Flag and the Mtw1 complex (Supplementary kinetochore localization. Fig. S5a) or between Cnn1 and the purified Spc105/YDR532c complex (data not shown). We next examined whether the interaction of Cnn1/CENP-T is a centromere receptor for the Ndc80 complex Cnn1 with the Ndc80 complex is compatible with simultaneous Its unique protein interaction profile prompted us to investigate binding of the Mtw1 complex and thus would allow interaction with the role of Cnn1CENP-T more closely. Physical interactions between the KMN network. In three-component gel filtration experiments CENP-T and KMN components have been reported21, but the with an excess of Mtw1 complex, Ndc80–Mtw1 complex formation specificity and molecular basis of these associations have remained was readily observed as reported previously31–33. Interestingly, under uncharacterized. To rigorously probe potential interactions, we these conditions, Cnn1 failed to co-elute effectively with the purified full-length Cnn1CENP-T from yeast extracts and examined Mtw1–Ndc80 supercomplex, indicating that the Mtw1 complex and its association with reconstituted recombinant budding yeast Ndc80 Cnn1CENP-T are competing binding partners of the Ndc80 complex and Mtw1 kinetochore complexes by size-exclusion chromatography. (Supplementary Fig. S5b). Supporting this, Cnn1–Flag immobilized The appearance of a fast-eluting species indicated complex formation on beads efficiently pulled down the free recombinant Ndc80 complex, between the Ndc80 complex and Cnn1CENP-T (Fig. 4a). The complex but the amount of interacting Ndc80 complex was reduced in a seemed stoichiometric on Coomassie-stained gels and formed under dose-dependent manner with increasing concentration of Mtw1 submicromolar input concentrations, indicating a high-affinity complex included in the reaction. Cnn1–Flag failed to pull down interaction. Under the same conditions, no complex formation was the Mtw1 complex, either alone or in the presence of the Ndc80

ARTICLES

a Wip1–GFP Bright field b Cnn1–3GFP Cnn1–3GFP mcm16Δ wip1Δ

chl4Δ

1.0 iml3Δ 0.8

0.6

0.4 intensity (AU)

mcm21Δ 0.2 Normalized Cnn1–3GFP

Wild type wip1Δ

Figure 3 Localization dependencies of yeast CCAN histone-fold proteins. or wip1-deletion background. Lower panel: Quantification of Cnn1–3GFP (a) Live-cell microscopy of Wip1–GFP in various CCAN-deletion mutants. fluorescence intensity in a wild-type or wip1-deletion strain. The intensity of Note the absence of kinetochore localization in the chl4-deletion mutant. Cnn1–3GFP in a wild-type strain was normalized to 1; kinetochore clusters (b) Live-cell microscopy of yeast strains expressing Cnn1–3GFP in a wild-type in 10 anaphase cells were quantified. Scale bar, 2 µm. complex (Fig. 4b,c). Taken together, these experiments indicate that localization of the complex34. Size-exclusion chromatography indicated Cnn1CENP-T and the Mtw1 complex are competing binding partners the formation of a tight complex between the Spc24–25 dimer and for the Ndc80 complex. Cnn1, which was abolished in the Cnn1 mutant lacking the conserved To probe the interaction between the Ndc80 complex and N-terminal motif (Fig. 6d). We also established that the Ndc80 complex Cnn1 in the cell, we performed co-immunoprecipitation exper- can recruit Cnn1 to taxol-stabilized microtubules in co-sedimentation iments using myc-tagged Nuf2. Cnn1–Flag could be specifically experiments (Supplementary Fig. S7). To determine whether the co-immunoprecipitated from log-phase extracts, confirming an conserved motif is also sufficient for Ndc80 binding, we synthesized association between these proteins in vivo (Fig. 5a). Notably, when a 25-amino-acid peptide including the 15-residue binding motif. We immunoprecipitations were performed from extracts that had been characterized the association using isothermal titration calorimetry synchronized by release from an α-factor arrest, the interaction and found that this peptide binds in a 1:1 stoichiometry to the was detectable only in a time window of about 90–105 min after Spc24–25 heterodimer with an apparent dissociation constant of the release. Immunofluorescence microscopy indicated that these 3.2 µM(Fig. 6e). This affinity seems to be lower than observed in gel time points corresponded to the time of mitotic spindle elongation filtration experiments with full-length Cnn1, indicating that additional (Fig. 5b). In contrast, co-immunoprecipitation experiments indicated parts of the N terminus may contribute to the high-affinity interaction a relatively constant affinity between the Ndc80 complex and the Mtw1 with the Ndc80 complex. As the Spc24–25 domain has previously been complex over the cell cycle (Fig. 5c). We conclude that the interaction identified as the binding site for the Mtw1 complex31–33, this result between Cnn1 and Ndc80 occurs in a cell-cycle-specific manner and is can explain the observed competition for Ndc80-complex binding and predominantly detectable in anaphase. implies that the Spc24–25 domain has at least two distinct, mutually exclusive binding partners within the yeast kinetochore, the N terminus Identification of an Ndc80 receptor motif in the of Cnn1CENP-T and the Mtw1 complex. N terminus of Cnn1 Using the motif search program MEME, we could identify a short Artificial tethering of the Ndc80 complex through Cnn1 conserved motif in the otherwise unstructured N terminus of CENP-T promotes mini-chromosome segregation in the absence proteins, predicted to fold into an α-helix with several conserved of a centromere hydrophobic residues (Fig. 6a). We reasoned that this N-terminal Budding yeast strains with a deletion of Cnn1CENP-T and/or motif might be involved in the interaction with the Ndc80 complex. Wip1CENP-W did not exhibit an obvious growth or chromosome Consistent with our hypothesis, a mutant Cnn1 protein lacking the segregation defect (data not shown), indicating that in principle 15-amino-acid conserved motif (Cnn1165–79, Fig. 6b) failed to interact the Mtw1-based Ndc80 recruitment pathway is sufficient to support with the recombinant Ndc80 complex (Fig. 6c). We next established the viability in this organism. In addition, by fluorescence microscopy binding site on the Ndc80 complex: the Ndc80 heterotetramer contains we could not detect a substantial reduction of Ndc80 signal two functionally distinct ends, the microtubule-binding Ndc80–Nuf2 intensity in a cnn1-deletion mutant (Supplementary Fig. S8). In dimer and the Spc24–25 dimer, which is required for centromere contrast, the putative S. pombe CENP-T homologue Cnp20 is an

ARTICLES

abNdc80 complex ++– – + + – Mr (K) 670 158 Mtw1 complex –++ – – ++ M (K) A3 A4 A5 A6 A7 A8 A9 A10 A11 r Mr (K) 130 130 Nuf2–eGFP 70 Ndc80 70 55 Cnn1–Flag 55 Cnn1–Flag 35 Mtw1p 35 25 Spc24–6×His 25 Spc25 Cnn1–Flag Input Pulldown 130 Nuf2–eGFP 70 Ndc80 55 c + Mtw1 complex 35 – 25 Spc24–6×His Spc25 Mr (K) Ndc80 complex 130 Nuf2–eGFP 130 70 Ndc80 Nuf2–eGFP 55 Cnn1–Flag 70 55 Ndc80 Cnn1–Flag 35 35 25 Spc24–6×His Spc25 25 Cnn1–Flag Spc24–6×His + Spc25 Ndc80 complex Ndc80 pulldown Elution volume

Figure 4 Cnn1CENP-T is an exclusive and direct interaction partner of proteins are indicated at the top. (b) Pulldown experiment with recombinant the Ndc80 complex. (a) Analytical size-exclusion chromatography of Ndc80 and Mtw1 kinetochore complexes and Cnn1–Flag immobilized recombinant Cnn1–Flag (upper panel), reconstituted yeast Ndc80 complex on beads. (c) Pulldown experiment with Cnn1–Flag, Ndc80 complex and (middle) or a combination of both (lower panel). Elution positions of standard increasing concentrations of the Mtw1 complex. essential protein and mutants exhibit very severe chromosome In similar experiments, LacI fusions to CENP-T have been shown segregation phenotypes27. to recruit KMN components to ectopic chromosomal lacO arrays in To probe the biological function of Ndc80 recruitment by human cells and they can partially replace the activity of an endogenous Cnn1CENP-T in budding yeast we therefore employed a synthetic gain-of centromere in chicken DT40 cells21. Thus, the ability of Cnn1CENP-T function approach (Fig. 7a). As described previously, artificial tethering to anchor Ndc80 complexes to DNA in a manner that allows them to of proteins to a replication-competent mini-chromosome that lacks productively interact with the mitotic spindle seems to be a conserved an authentic centromere can be used to determine whether candidate feature of these proteins. proteins are able to provide the function of an artificial kinetochore35,36. Direct tethering of the Ndc80 complex through a C-terminal TetR A Wip1-deletion allele exhibits severe synthetic phenotypes in fusion to the Spc24 subunit failed to induce a mitotic stabilization combination with other kinetochore mutants of mini-chromosomes35. As a result of our biochemical experiments, The experiments so far indicate that different receptors for Ndc80 we reasoned that it might be essential to recruit the Ndc80 complex complexes are present at yeast kinetochores. This could explain through the elongated, flexible Cnn1 N terminus. We constructed a why Ndc80 loss-of-function mutants exhibit severe chromosome- strain in which the histone-fold domain of Cnn1 was replaced with detachment phenotypes38 whereas mutants of Mtw1 complex subunits the TetR repressor and compared the stability of an acentric tetO or Mif2CENP-C do not39. To address this issue, we combined mini-chromosome in this strain with that of a number of control temperature-sensitive mif2 or mtw1 mutants with cnn1 or wip1 strains (Fig. 7b). Notably, the Cnn11HF–TetR fusion, but not the deletions and investigated their phenotypes. mif2 wip1 double Spc24–TetR fusion, strongly enhanced mini-chromosome stability over mutants showed a severe growth defect (Fig. 7d) and similar synthetic multiple generations. The stabilizing effect exceeded that of Ask1–TetR, interactions were detected with okp1 and mtw1 alleles (data not shown). a known force-transducing component of the budding yeast Dam1 Inspection of chromosome segregation by immunofluorescence kinetochore complex37. Importantly, the artificial kinetochore function microscopy showed that the mif2 wip1 double mutant contained supplied by Cnn11HF–TetR was abolished in a mutant lacking amino elongated mitotic spindles with a single mass of unsegregated DNA, acids 65–79 (Fig. 7c), demonstrating that it strictly depended on the phenotypically similar to the defects seen in an ndc80-1 mutant at the ability of Cnn1 to interact with the Ndc80 complex. On the other restrictive temperature, but distinct from the mif2-3 and wip1 single hand, recruiting subunits of the Mtw1 complex (Mtw1–TetR and mutants (Fig. 7e). Thus, simultaneously interfering with both Ndc80 Nnf1–TetR) to the tetO array did not lead to plasmid stabilization. recruitment modes leads to severe defects in chromosome segregation.

ARTICLES

a b Time after α-factor release (min)

0 15 30 45 60 75 90 105 120 ×myc ×Flag ×Flag Nuf2–13×myc Cnn1–6 Cnn1–6 Nuf2–13 Input Input Cnn1–6×Flag Cnn1–6×Flag

IP: anti-Myc Cnn1–6×Flag × IP: anti-Myc Cnn1–6 Flag Log phase

0 min 75 min 90 min 105 min 120 min

Tubulin

Time after α-factor release (min)

c 0 15 30 45 60 75 90 105

myc × ×Flag ×Flag × Nuf2–13 myc –6 13 Input Mtw1–6 Mtw1 Nuf2– Input Mtw1–6×Flag Mtw1–6×Flag

IP: anti-Myc × Mtw1–6 Flag × IP: anti-Myc Mtw1–6 Flag Log phase

Figure 5 Cell-cycle-dependent interaction between Ndc80 and microscopy of samples taken at the indicated time points from the Cnn1CENP-T.(a) Co-immunoprecipitation between Nuf2–13xMyc cell-cycle experiment. Note the presence of long anaphase spindles in and Cnn1–6xFlag from yeast log-phase extracts. (b) Cell-cycle the 90 and 105 min time points. Scale bar, 5 µm. (c) Corresponding co-immunoprecipitation experiment (top panel): after release from an experiment probing the interaction between the Ndc80 and the Mtw1 α-factor arrest, extracts were prepared at the indicated times and Nuf2 complex. Nuf2–13xMyc was immunoprecipitated and the association was immunoprecipitated using anti-myc beads. Cnn1 was detected by with Mtw1–6xFlag was detected by western blot. Uncropped images of anti-Flag western blot. Lower panel: anti-tubulin immunoﬂuorescence western blots are shown in Supplementary Fig. S6.

DISCUSSION linked to signalling molecules of the mitotic checkpoint machinery44 Our investigations into budding yeast centromere proteins have and thus play important roles in sensing the presence and quality of uncovered general organization principles of kinetochores (Fig. 7f). kinetochore–microtubule attachments. The observation that Cnn1 The conserved role of CENP-T proteins is to act as direct, and the Mtw1 complex compete for Ndc80 binding and that the KMN-independent centromere receptors of the Ndc80 complex. Cnn1–Ndc80 affinity is increased during anaphase is indicative of This provides a molecular understanding for the requirement of the significant changes in kinetochore architecture that occur at the CENP-T N terminus for kinetochore assembly in human cells21 and metaphase–anaphase transition. Future experiments will have to can explain the observation that it undergoes a tension-dependent reveal how these structural rearrangements are established and what conformational change in response to microtubule attachments40. The their precise functional consequence is. CENP-T proteins and CCAN original identification of Cnn1 as an interaction partner of the Mtw1 subunits other than CENP-C seem to be absent from the kinetochores complex19 was probably due to the use of a single-step purification of Caenorhabditis elegans and Drosophila melanogaster. Consistent with protocol, which has been shown to co-purify a large set of kinetochore the existence of different Ndc80 recruitment pathways, mis12 mutants proteins41. Our results raise the question as to why kinetochores contain in these organisms exhibit a true kinetochore-null phenotype45 because multiple different Ndc80 recruiters. In budding yeast, the interaction here Ndc80 is exclusively recruited as part of the KMN network. between Ndc80 and Cnn1 is detectable predominantly in anaphase, the METHODS mitotic phase in which cells have irreversibly committed to segregate their chromosomes. Consistent with this observation, the levels of Methods and any associated references are available in the online Cnn1 at kinetochores seem to be increased in anaphase, as judged by version of the paper at www.nature.com/naturecellbiology 42 live-cell microsocopy . Recruitment through CENP-T incorporates Note: Supplementary Information is available on the Nature Cell Biology website the Ndc80 complex into a flexible attachment filament that might be especially suited for force transduction to move chromosomes, a ACKNOWLEDGEMENTS 43 The authors thank all members of the Westermann laboratory for discussions property that was reflected in our artificial kinetochore experiment . and J. M. Peters for critical reading of the manuscript. We thank T. Burkard As part of the KMN network, on the other hand, Ndc80 complexes are for suggestions, W. Lugmayr for support with high-performance computing,

ARTICLES

a Cnn1p bc

Mr (K) 670 158 HF 65-79 HF Δ HF A2 A3 A4 A5 A6 A7 A8 A9 A10 Mr (K) 130 Mr (K) HSS Cnn1HSS Cnn1 2D structure 70 250 55 Cnn1Δ65–79–Flag Homo s. 85 TP NI 130 Mus m. 86 TP NI 100 Ornithorhynchus a. 105 TPRTI L IM 70 Cnn1–Flag 35 55 Gallus g. 72 TPRV IIQN Cnn1Δ65–79–Flag 25 Xenopus l. 115 TPRS KIIQ PE Danio r. 99 RG GIIHME 35 Ciona i. 81 TPRTAL VI 25 Schizosaccharomyces p. 70 TPRDI L LS 15 Cnn1Δ65–79–Flag Ajellomyces c. 106 TPMDVL LGKV P Magnaporthe o. 144 GPFDAL LS 15 Tuber m. 122 TPRD AL 130 Nuf2–eGFP 70 Pyrenophora t. 109 TPRDI LRDLA R 55 Ndc80 Yarrowia l. 35 TPRDI LRALS MVR Cnn1Δ65–79–Flag Pichia p. 134 TPTKDLRSLSR Flag peptide eluate 35 Kluyveromyces l. 10 GLKGVLSEL LVH Candida g. 54 ESRMFLKELD S 25 Spc24–6×His Lachancea t. 8 ALKQQLRELSF K Spc25 Candida a. 142 IT YLKFFC Q Candida p. 121 IT YLRYVY S 15 Cnn1Δ65–79–Flag Saccharomyces k. 59 EL YLRELS + Zygosaccharomyces r. 60 EV YLRDL LVS Ndc80 complex Ashbya g. 54 KLRHYLREL LVN Saccharomyces c. 65 EVRSFLQDLSQVLAR Elution volume 2D structure

Time (min) M (K) 158 d r e 0 102030405060 0.2 Mr (K) A5 A6 A7 A8 A9 A10 A11 A12 B1 0.1 130 Superose 6 PC3.2/30 0 70 55 –0.1 –1 35 Spc24 –0.2 25 Spc24–6×His cal s –0.3

Spc25 μ Spc25 –0.4 15 –0.5 60NKDPNEVRSFLQDLSQVLARKSQGN84 Spc25–Spc24–6×His –0.6 + Cnn1p SPC24–25 130 –0.7 70 HF 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 55 Cnn1–Flag 35 + Molar ratio 25 Spc24–6×His Spc25 Spc24 15 Cnn1–Flag + Spc25 0 Spc25–Spc24–6×His

130 70 –2 55 Δ65–79 Δ Cnn1 –Flag Cnn1 65–79 35 HF 25 Spc24–6×His Spc25 + –4 kcal per mole of injectant 15 Cnn1Δ65–79–Flag + 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Spc25–Spc24–6×His Molar ratio

Elution volume

Figure 6 The N terminus of Cnn1CENP-T contains a conserved binding motif with reconstituted yeast Ndc80 complex (lower panel). (d) Analytical for the Spc24/25 domain of the Ndc80 complex. (a) Multiple sequence size-exclusion chromatography of the recombinant Spc24–25 heterodimer alignment of a conserved motif present in the N termini of CENP-T proteins. alone (upper panel), or in combination with full-length Cnn1–Flag (middle Secondary-structure predictions identify a potential α-helix in human (top) panel) or Cnn1165–79–Flag (lower panel). (e) Isothermal titration calorimetry and budding yeast (bottom) CENP-T proteins. (b) Puriﬁcation of overex- with a Cnn1-derived peptide (residues 60–84) encompassing the conserved pressed Cnn1–Flag and Cnn1165–79–Flag lacking the conserved N-terminal N-terminal motif and the Spc24–25 heterodimer. Upper panel shows raw motif from yeast extracts. HSS: high-speed supernatant. (c) Size-exclusion data of heat change on binding; lower panel depicts binding isotherm derived chromatography of Cnn1165–79–Flag alone (top panel) or in combination from the data.

O. Hudecz for help with presentation of the mass spectrometry results and AUTHOR CONTRIBUTIONS M. Madalinski for peptide synthesis. We thank P. De Wulf (European Institute of A.S. performed bioinformatic sequence analysis. M.M. purified CCAN proteins and Oncology, Milan, Italy) for the Cnn1–3GFP strain and communicating results before performed characterization of proteins. P.H. performed interaction studies. F.L. publication. Research in the Westermann laboratory receives funding from the conducted ChIP and biochemical experiments. K.M. guided the mass spectrometry European Research Council under the European Community’s Seventh Framework analysis. S.W. guided the study and performed biochemical and genetic experiments Programme (S.W. FP7/2007-2013)/ERC grant agreement no. 203499, and from the supported by G.L. All authors discussed results and analysed data. S.W. and A.S. Austrian Science Fund FWF (S.W., SFB F34-B03). wrote the manuscript.

ARTICLES

a bd

70 wip1Δ:URA3 60 mif2-3

50 1234

30 ARS 20 Percentage of plasmid stability Percentage URA3 tetO 10

Acentric R R R R R R Tetrads –Tet Tet –Tet –Tet –Tet Tet Mini-chromosome segregation? Control ΔHF Δ65–79 Ask1 Mtw1 Nnf1– Δ Spc24– HF wip1 mif2-3 Cnn1 Δ

Cnn1 c

YPD

doURA

Control Ask1–TetR Spc24–TetR Cnn1ΔHF–TetRCnn1ΔHFΔ65–79–TetR Mtw1–TetR Nnf1–TetR

e Tubulin DAPI f 100

80 Normal Kinetochore microtubule 60

Unsegregated Ndc80 complex 40 + Number of cells Spc105/KNL-1 Mis-segregated 20 Ndc80 receptor Nsl1 Mtw1 complex Dsn1 motif Mtw1 Mif2/CENP-C Nnf1 0 HF HF Cnn1/CENP-T Short wip1Δ mif2-3 ndc80-1 wip1Δ mif2-3 CEN DNA Wip1/CENP-W CSE4 nucleosome Figure 7 Artificial tethering of the Ndc80 complex through Cnn1CENP-T hours after shift to the restrictive temperature of the temperature-sensitive allows mini-chromosomes to segregate in the absence of a centromere. mutants. DNA is stained with DAPI; microtubules are stained with an (a) Scheme of the mini-chromosome stability assay. The segregation of a anti-tubulin antibody. Right panel: Phenotypic categories are indicated and URA3-containing mini-chromosome containing a tetO array but lacking the respective phenotypes were quantified for the individual strains. Scale a centromere is assayed in different strain backgrounds. (b) Stability bar, 5 µm. (f) Two distinct receptors for Ndc80 molecules are present at of URA3 mini-chromosomes in different strain backgrounds. Error bars yeast centromeres, one is the Mtw1 complex, which in turn may require the denote s.e.m. n = 3. (c) Examples of plating assays for fusion strains conserved protein Mif2/CENP-C for association with centromeric DNA. In Ask1–TetR, Spc24–TetR, Cnn11HF–TetR, Cnn11HF165–79–TetR, Mtw1–TetR a second pathway, Ndc80 molecules directly interact with the histone-fold and Nnf1–TetR on YPD and doURA. (d) Double deficiency of the Ndc80 complex Cnn1–Wip1 through a conserved receptor motif in the N-terminal recruitment pathways: tetrad dissection of wip11 mif2-3 double mutants. extension of Cnn1. In a normal budding yeast cell cycle this linkage is not (e) Chromosome segregation defects in wip11 mif2-3 double mutants. required for viability, but it becomes essential when Mtw1 subunits or Mif2 Left panel: Cells were processed for immunofluorescence microscopy two are functionally compromised.

COMPETING FINANCIAL INTERESTS 1. Cheeseman, I. M. & Desai, A. Molecular architecture of the kinetochore-microtubule The authors declare no competing financial interests. interface. Nat. Rev. Mol. Cell Biol. 9, 33–46 (2008). 2. Lampert, F. & Westermann, S. A blueprint for kinetochores—new insights Published online at www.nature.com/naturecellbiology into the molecular mechanics of cell division. Nat. Rev. Mol. Cell Biol. 12, Reprints and permissions information is available online at www.nature.com/reprints 407–412 (2011).

ARTICLES

3. Santaguida, S. & Musacchio, A. The life and miracles of kinetochores. EMBO J. 28, 26. Talbert, P. B. & Henikoff, S. Histone variants–ancient wrap artists of the epigenome. 2511–2531 (2009). Nat. Rev. Mol. Cell Biol. 11, 264–275 (2010). 4. Liu, D., Vader, G., Vromans, M. J., Lampson, M. A. & Lens, S. M. Sensing 27. Tanaka, K., Chang, H. L., Kagami, A. & Watanabe, Y. CENP-C functions as a scaffold chromosome bi-orientation by spatial separation of aurora B kinase from kinetochore for effectors with essential kinetochore functions in mitosis and meiosis. Dev. Cell substrates. Science 323, 1350–1353 (2009). 17, 334–343 (2009). 5. Musacchio, A. & Salmon, E. D. The spindle-assembly checkpoint in space and time. 28. Amano, M. et al. The CENP-S complex is essential for the stable assembly of outer Nat. Rev. Mol. Cell Biol. 8, 379–393 (2007). kinetochore structure. J. Cell Biol. 186, 173–182 (2009). 6. Cheeseman, I. M., Chappie, J. S., Wilson-Kubalek, E. M. & Desai, A. The conserved 29. Yan, Z. et al. A histone-fold complex and FANCM form a conserved DNA-remodeling KMN network constitutes the core microtubule-binding site of the kinetochore. Cell complex to maintain genome stability. Mol. Cell 37, 865–878 (2010). 127, 983–997 (2006). 30. McAinsh, A. D. & Meraldi, P. The CCAN complex: linking centromere specification 7. Alushin, G. M. et al. The Ndc80 kinetochore complex forms oligomeric arrays along to control of kinetochore-microtubule dynamics. Semin. Cell Dev. Biol. 22, microtubules. Nature 467, 805–810 (2010). 946–952 (2011). 8. Guse, A., Carroll, C. W., Moree, B., Fuller, C. J. & Straight, A. F. In vitro 31. Petrovic, A. et al. The MIS12 complex is a protein interaction hub for outer centromere and kinetochore assembly on defined chromatin templates. Nature 477, kinetochore assembly. J. Cell Biol. 190, 835–852 (2010). 354–358 (2011). 32. Maskell, D. P., Hu, X. W. & Singleton, M. R. Molecular architecture and assembly of 9. Carroll, C. W., Milks, K. J. & Straight, A. F. Dual recognition of CENP-A nucleosomes the yeast kinetochore MIND complex. J. Cell Biol. 190, 823–834 (2010). is required for centromere assembly. J. Cell Biol. 189, 1143–1155 (2010). 33. Hornung, P. et al. Molecular architecture and connectivity of the budding yeast Mtw1 10. Foltz, D. R. et al. The human CENP-A centromeric nucleosome-associated complex. kinetochore complex. J. Mol. Biol. 405, 548–559 (2011). Nat. Cell Biol. 8, 458–469 (2006). 34. Ciferri, C. et al. Implications for kinetochore-microtubule attachment from the 11. Okada, M. et al. The CENP-H-I complex is required for the efficient incorporation of structure of an engineered Ndc80 complex. Cell 133, 427–439 (2008). newly synthesized CENP-A into centromeres. Nat. Cell Biol. 8, 446–457 (2006). 35. Kiermaier, E., Woehrer, S., Peng, Y., Mechtler, K. & Westermann, S. A Dam1-based 12. Cheeseman, I. M. et al. Phospho-regulation of kinetochore-microtubule attachments artificial kinetochore is sufficient to promote chromosome segregation in budding by the Aurora kinase Ipl1p. Cell 111, 163–172 (2002). yeast. Nat. Cell Biol. 11, 1109–1115 (2009). 13. Westermann, S. et al. Architecture of the budding yeast kinetochore reveals a 36. Lacefield, S., Lau, D. T. & Murray, A. W. Recruiting a microtubule-binding complex conserved molecular core. J. Cell Biol. 163, 215–222 (2003). to DNA directs chromosome segregation in budding yeast. Nat. Cell Biol. 11, 14. Liu, X., McLeod, I., Anderson, S., Yates, J. R. 3rd & He, X. Molecular analysis of 1116–1120 (2009). kinetochore architecture in fission yeast. EMBO J. 24, 2919–2930 (2005). 37. Westermann, S. et al. The Dam1 kinetochore ring complex moves processively on 15. Meraldi, P., McAinsh, A. D., Rheinbay, E. & Sorger, P. K. Phylogenetic and structural depolymerizing microtubule ends. Nature 440, 565–569 (2006). analysis of centromeric DNA and kinetochore proteins. Genome Biol. 7, R23 (2006). 38. Wigge, P. A. et al. Analysis of the Saccharomyces spindle pole by matrix- 16. Westermann, S., Drubin, D. G. & Barnes, G. Structures and functions of yeast assisted laser desorption/ionization (MALDI) mass spectrometry. J. Cell Biol. 141, kinetochore complexes. Annu. Rev. Biochem. 76, 563–591 (2007). 967–977 (1998). 17. Malik, H. S. & Henikoff, S. Major evolutionary transitions in centromere complexity. 39. Screpanti, E. et al. Direct binding of Cenp-C to the Mis12 complex joins the inner Cell 138, 1067–1082 (2009). and outer kinetochore. Curr. Biol. 21, 391–398 (2011). 18. Hori, T., Okada, M., Maenaka, K. & Fukagawa, T. CENP-O class proteins form a 40. Suzuki, A. et al. Spindle microtubules generate tension-dependent stable complex and are required for proper kinetochore function. Mol. Biol. Cell 19, changes in the distribution of inner kinetochore proteins. J. Cell Biol. 193, 843–854 (2008). 125–140 (2011). 19. De Wulf, P., McAinsh, A. D. & Sorger, P. K. Hierarchical assembly of the budding 41. Akiyoshi, B. et al. Tension directly stabilizes reconstituted kinetochore-microtubule yeast kinetochore from multiple subcomplexes. Genes Dev. 17, 2902–2921 (2003). attachments. Nature 468, 576–579 (2010). 20. Hori, T. et al. CCAN makes multiple contacts with centromeric DNA to provide 42. Bock, L. J. et al. Cnn1 inhibits the interactions between the KMN complexes of the distinct pathways to the outer kinetochore. Cell 135, 1039–1052 (2008). yeast kinetochore. Nat. Cell Biol. http://dx.doi.org/10.1038/ncb2495 (2012). 21. Gascoigne, K. E. et al. Induced ectopic kinetochore assembly bypasses the 43. Powers, A. F. et al. The Ndc80 kinetochore complex forms load-bearing requirement for CENP-A nucleosomes. Cell 145, 410–422 (2011). attachments to dynamic microtubule tips via biased diffusion. Cell 136, 22. Nishino, T. et al. CENP-T-W-S-X forms a unique centromeric chromatin structure 865–875 (2009). with a histone-like fold. Cell 148, 487–501 (2012). 44. Kiyomitsu, T., Obuse, C. & Yanagida, M. Human Blinkin/AF15q14 is required for 23. Burley, S. K., Xie, X., Clark, K. L. & Shu, F. Histone-like transcription factors in chromosome alignment and the mitotic checkpoint through direct interaction with eukaryotes. Curr. Opin. Struct. Biol. 7, 94–102 (1997). Bub1 and BubR1. Dev. Cell 13, 663–676 (2007). 24. Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F. & Richmond, T. J. Crystal 45. Cheeseman, I. M. et al. A conserved protein network controls assembly of structure of the nucleosome core particle at 2.8 A resolution. Nature 389, the outer kinetochore and its ability to sustain tension. Genes Dev. 18, 251–260 (1997). 2255–2268 (2004). 25. Sandman, K. & Reeve, J. N. Archaeal histones and the origin of the histone fold. 46. Cole, C., Barber, J. D. & Barton, G. J. The Jpred 3 secondary structure prediction Curr. Opin. Microbiol. 9, 520–525 (2006). server. Nucleic Acids Res. 36, W197–W201 (2008).

METHODS DOI: 10.1038/ncb2493

METHODS kowalevskii (E value 4×10−4) or the house mouse Mus musculus (E value 8×10−3; Homology searches and protein annotation. All searches were performed in the Supplementary Fig. S2). NCBI non-redundant protein database47, or within specified proteomes, listed in Mcm22 and CENP-K are orthologous. S. cerevisiae Mcm22 is 239 residues Supplementary Table S4. Species and accession numbers are listed in Supplementary long and has a predicted coiled coil from 113–134. A BLASTP search using the Table S2. We used NCBI BLASTP and PSI-BLAST version 2.2.24 (ref. 48), and full-length Mcm22 (masking the coiled-coil region) identifies orthologues in the HMMer v.3.0 (ref. 49) with significant E values below 0.01. Fold predictions were closest Saccharomycetes organisms, such as Candida glabrata (E value: 3 × 10−8) executed with HHpred in the PDB protein structure database50,51. Domains were or A. gossypii (E value: 5 × 10−11). The conserved domain is spanning nearly the annotated with Pfam and Smart databases52,53, coiled coils were predicted using whole protein length (S. cerevisiae Mcm22: 7–233) and the coiled-coil predictions COILS (window 21; ref. 54), low-complexity regions with SEG (window 12; ref. 55) of the orthologues overlap with the one of S. cerevisiae. We constructed a HMM and secondary structure with Jpred346. Sequences were aligned with MAFFT version out of the conserved domain (the regions corresponding to the S. cerevisiae coiled 6 (L-INS-I method; ref. 56), and visualized and processed with Jalview (ref. 57) coil were masked) and searched in the proteomes of Candida dubliniensis, Pichia (Clustalx colouring scheme). To search for short motifs outside conserved domains, guilliermondii, Candida albicans and Aspergillus niger, where in each case one protein we used MEME (one occurrence per sequence, motif width 10–25; ref. 58). was found below the inclusion value (E value < 0.01; Cd36_87730, PGUT_03964, Ame1 homology search. The budding yeast protein Ame1 is 324 residues in CAWT_03094, jgi|Aspni1|178612|). In the following iterative PSI-BLAST search length and has a predicted coiled-coil region between amino acids 199 to 263. In within the NCBI non-redundant protein database, more Pezizomycotina proteins homology searches, it is a common practice to disregard compositionally biased were collected in round 1 with significant E values (<2 × 10−6), among those regions such as low-complexity, transmembrane or coiled-coil regions. When Gibberella zeae and Tuber melanosporum. The subsequent HMM search identified masking the coiled-coil region, an iterative PSI-BLAST search within the NCBI proteins from Strongylocentrotus purpuratus (E value: 2.1 × 10−6), Gallus gallus non-redundant protein database identified only Saccharomycetes strains closely (0.0019), Pan troglodytes (0.0031), Xenopus laevis (0.0055) and six more animal related to S. cerevisiae, such as Vanderwaltozyma polyspora, Zygosaccharomyces rouxii, proteins below the significant inclusion value (0.01). They all belong to the CENP-K Kluyveromyces lactis or Ashbya gossypii, with significant E values <5×10−4 (round family, where S. pombe Sim4 is included as well52 (Supplementary Fig. S3). In a 1). The search converged in round 2 without incorporation of further hits. A reciprocal approach, the Pfam CENP-K HMM identified the Giberella zeae and other multiple alignment of these orthologues revealed a conserved short region at the Pezizomycotina orthologues, and a subsequent search with a HMM including the N terminus (rich in positively charged residues—corresponding to positions 2–22 Pezizomycotina proteins hit C. glabrata Mcm22 (E values < 0.01). in S. cerevisiae Ame1) and a conserved, α-helical domain (residues 115–314), where CENP-W homology search. In mammals, besides CENP-T, three further histone- the coiled-coil region is a central part of it. We included the coiled-coil region in fold proteins are known to interact with the core CCAN complex: CENP-S, CENP-W further searches, allowing for the incorporation of unrelated coiled-coil proteins as and CENP-X (refs 20,28), forming CENP-S/CENP-X (ref. 29) and CENP-T/CENP- false positive hits. A search with a more sensitive HMM profile using sequences W heterodimeric complexes. Unlike CENP-T, these three proteins are much shorter comprising the conserved domain detected significant hits in the more distantly (81–138 residues), hardly extending the histone domain (CENP-S has an extension related Penicillium marneffei (E value 1.3×10−5) and to other Pezizomycotina fungi, of around 30 amino acids at the C terminus rich in positively charged residues). such as Uncinocarpus reesii and Arthroderma gypseum (E values 0.0029 and 0.0046, CENP-X and CENP-S are well conserved and homologues in S. pombe and S. respectively). We used BLASTP searches to enlarge the sequence family within the cerevisiae were identified with BLASTP. CENP-W family members can be detected Pezizomycotina fungi (E values < 1×10−5). They all belong to a group of conserved in Pezizomycotina fungi, such as A. niger, and in Schizosaccharomyces japonicus with hypothetical proteins, without known function. In the subsequent HMM search iterative PSI-BLAST searches, but the model iterated into histone-fold proteins, such within the S. pombe proteome, only one significant hit was identified, namely Mis17 as the transcription factors NC2β or Hap3. To find Saccharomycetes homologues, (E value 5.1×10−5; Supplementary Fig. S1). a HMM model containing the CENP-W histone domain identified the orphan Intriguingly, Mis17 has been identified as a centromere protein in genetic screens protein DEHA2G10538g in Debaryomyces hansenii with a highly significant E value and is involved in the stable deposition of CENP-A to centromeres; thus, functionally (1.2×10−11), followed by the NC2β-like protein DEHA2E16126g (E value 7×10−07), it acts like a CCAN subunit59. In addition to other Ascomycota, we enlarged our whereas we failed to detect any significant S. cerevisiae proteins. search to the proteome of the Basidiomycota Ustilago maydis. Using a HMM In a biochemical isolation of CCAN-complex-binding proteins from budding profile covering the α-helical Ame1/Mis17 domain, we identified one significant yeast cell extracts, several hypothetical proteins were characterized by mass protein (UM03572, E value 0.00035). Applying the expanded model to the NCBI spectrometry, using Mcm16 and Cnn1 as bait (Supplementary Table S3). Among non-redundant protein database resulted in 11 newly identified proteins below those, the protein YDR374W-A was of particular interest, owing to its short length the inclusion threshold (E value 0.01, apparent isoforms were not counted). Out (89 amino acids) and a high α-helical content. YDR374W-A is very weakly conserved of these, 5 proteins belong to the mammalian CENP-U gene family, with the and when applying sensitive HMM searches we found orthologues in Z. rouxii, CENP-U protein from Bos taurus as the best hit (E value 4.3×10−7). CENP-U (also K. lactis and A. gossypii. In the proteome of Debaryomyces hansenii, we identified called CENP-50) is a CENP-A co-purifying protein10 that forms a biochemically one protein DEHA2G10538g with a significant E value (3.4×10−5). Interestingly, D. stable complex with CENP-O, -P and -Q (ref. 18). The six remaining proteins hansenii DEHA2G10538g had been previously identified in the search for CENP-W were unrelated coiled-coil proteins from Trypanosoma, Nematoda, Ascomycetes, candidates. Thus, it is very likely that budding yeast YDR374W-A is the orthologue Bacteria (2 proteins) and Archaea. Searching with the Ame1/Mis17 domain within of mammalian CENP-W (Supplementary Fig. S4). the Mus musculus or Homo sapiens proteome results in CENP-U proteins as the only significant hits (E value 1.2 × 10−8 and 0.0051, respectively). In an Yeast strains and plasmids. Yeast strains are based on S288C and were generated alternative approach, a HMM–HMM comparison with the Ame1/Mis17 alignment by standard procedures. A list of used yeast strains and plasmids can be found in identified a model built from mammalian CENP-U proteins as the next best hit Supplementary Table S5. (besides the known fungi ones) with a sub-significant E value (0.016, UniProt20 database; ref. 60). CENP-U orthologues can be easily collected from mammals Protein biochemistry. Ndc80 and Mtw1 complexes were expressed and purified as to fish with BLASTP searches applying highly significant E values (<1 × 10−15). described previously33,61 with the following modifications: for the Mtw1 complex an As is the case for Ame1 and Mis17, they all share a coiled-coil region at the N-terminal 6×His-tagged Dsn1 subunit was cloned into a pST39 expression plasmid C terminus that is part of a highly conserved domain and predicted to be harbouring the Nnn1, Nsl1 and Mtw1 subunits. α-helical. Overexpression and purification of proteins from yeast extracts were performed Okp1 homology search. Okp1 is similar in architecture to Ame1, with a length as described previously62 with the following modifications. Yeast pellets (60 g) were of 406 residues and a coiled-coil region from 239 to 285. A BLASTP search with lysed with a freezer mill (Spex Centriprep) and the resulting yeast powder was stored a coiled-coil masked sequence identified proteins in the closest Saccharomycetes at −80 ◦C. For purification of Flag-tagged proteins, 10 g yeast powder was combined organisms, such as Lachancea thermotolerans, K. lactis or A. gossypii, with significant with 50 ml lysis buffer (25 mM HEPES at pH 8.0, 2 mM MgCl2, 0.1% NP-40, 150 mM E values (< 9 × 10−9). As in Ame1, the conserved domain is located at the KCl, 15 (v/v) % glycerol, 1 mM dithiothreitol and protease inhibitors), left on ice C-terminal region that is predicted to be α-helical and includes the coiled-coil for 30 min and centrifuged at 60,000g for 30 min at 4 ◦C. The supernatant was region. The sequence conservation score is slightly higher than for Ame1 and so further ultracentrifuged (100,000g, 45 min, 4 ◦C), cleared by filtration through a we could establish a linkage between Okp1 and the fission yeast Fta7 using a series 0.2 µm filter and collected. A 300 µl solution of anti-Flag M2 agarose beads was of PSI-BLAST searches: the conserved region of the Lachancea Okp1 orthologue added and rotated for 3 h at 4 ◦C. Beads were washed five times with wash buffer (including the coiled-coil region) hit one protein in Sclerotinia sclerotiorum with (lysis buffer with 0.5 M KCl, 1 mM ATP and 2 mM MgCl2) and eluted three times −5 −1 a significant E value (4 × 10 ) in round 2, and the Sclerotinia Okp1 domain in with 1 mg ml 3×Flag peptide in buffer H. Proteins were snap-frozen in liquid N2 turn found one protein in Tuber melanosporum (E value 5×10−21, round 2), which and stored at −80 ◦C. finally led to the identification of S. pombe Fta7 (E value 8×10−4, round 1). A HMM profile search with the fungi Okp1/Fta7 domain included 5 non-fungal proteins, all Interaction studies. Interaction studies were performed by size-exclusion chro- CENP-Q orthologues (inclusion value 0.01), such as the acorn worm Saccoglossus matography on Superose 6 PC3.2/30 columns. The indicated proteins or complexes

DOI: 10.1038/ncb2493 METHODS were used at 1 µM concentration, mixed and left on ice for 30 min before injecting on YPD or doURA plates, leading to the formation of approximately 600 colonies 50 µl onto the column. Fractions of 100 µl were collected, and proteins were on YPD. Experiments were performed in triplicates. precipitated using chloroform/methanol and subjected to SDS–PAGE. ChIP. Cultures were grown in YPD to mid-log phase. Cells were crosslinked for Isothermal titration calorimetry. Thermodynamic values of the interaction 30 min with formaldehyde (1%). Chromatin extract preparation and ChIP were between Cnn1-derived peptides and Spc24–25 were determined by ITC (MCS-ITC, performed as previously described63. Samples were analysed by quantitative PCR Microcal). All experiments were conducted at 25 ◦C. A 1.8 ml solution of Spc24–25 using the Biorad iQ5 real-time PCR system and software. (30 µM) was placed in the temperature-controlled cell and titrated with Cnn1 peptide at 500 µM in the syringe. The buffer used was 25 mM HEPES at pH 8 47. Sayers, E. W. et al. Database resources of the National Center for Biotechnology with 150 mM KCl. Data were analysed using ORIGIN following the instructions of Information. Nucleic Acids Res. 39, D38–D51 (2011). the manufacturer. 48. Altschul, S. F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402 (1997). Mass spectrometry. The nano-HPLC system used in all experiments was an 49. Johnson, L. S., Eddy, S. R. & Portugaly, E. Hidden Markov model speed heuristic and iterative HMM search procedure. BMC Bioinformatics 11, 431 (2010). UltiMate 3000 Dual Gradient HPLC system (Dionex), equipped with a Proxeon 50. Berman, H. M. et al. The Protein Data Bank. Nucleic Acids Res. 28, nanospray source (Proxeon), coupled to an LTQ Velos Orbitrap mass spectrometer 235–242 (2000). (Thermo Fisher Scientific), operated in data-dependent mode using a full scan in the 51. Soding, J., Biegert, A. & Lupas, A. N. The HHpred interactive server for ion cyclotron resonance cell followed by tandem mass spectrometry (MS/MS) scans protein homology detection and structure prediction. Nucleic Acids Res. 33, of the 12 most abundant ions in the linear ion trap. MS/MS spectra were acquired W244–W248 (2005). in the multistage activation mode, where subsequent activation was performed on 52. Finn, R. D. et al. The Pfam protein families database. Nucleic Acids Res. 38, fragment ions resulting from the neutral loss of −98, −49 or −32.6 m/z for potential D211–D222 (2010). follow-up phosphorylation site analysis. Precursor ions selected for fragmentation 53. Letunic, I., Doerks, T. & Bork, P. SMART 6: recent updates and new developments. were put on a dynamic exclusion list for 180 s. Monoisotopic precursor selection was Nucleic Acids Res. 37, D229–D232 (2009). enabled. 54. Lupas, A., Van Dyke, M. & Stock, J. Predicting coiled coils from protein sequences. Analysis of MS data. For peptide identification, all MS/MS spectra were searched Science 252, 1162–1164 (1991). using Mascot 2.2.04 (Matrix Science) and Sequest (Thermo Scientific) against 55. Wootton, J. C. & Federhen, S. Analysis of compositionally biased regions in sequence databases. Methods Enzymol. 266, 554–571 (1996). the yeast SGD protein sequence database (6,717 sequences; 3,020,761 residues). 56. Katoh, K. & Toh, H. Recent developments in the MAFFT multiple sequence alignment The following search parameters were used: carbamidomethylation or methyl program. Brief Bioinform. 9, 286–298 (2008). methylthiomethyl sulphoxide respectively on cysteine was set as a fixed modification; 57. Waterhouse, A. M., Procter, J. B., Martin, D. M., Clamp, M. & Barton, G. J. oxidation on methionine was set as variable modification. Monoisotopic masses Jalview Version 2–a multiple sequence alignment editor and analysis workbench. were searched within unrestricted protein masses for tryptic, chymotryptic and Bioinformatics 25, 1189–1191 (2009). unspecific (subtilisin digest) peptides. The peptide mass tolerance was set to ±5 ppm 58. Bailey, T. L. et al. MEME SUITE: tools for motif discovery and searching. Nucleic and the fragment mass tolerance to ±0.5 Da. The maximal number of missed Acids Res. 37, W202–W208 (2009). cleavages was set to 2. Using a Thermo Proteome Discoverer 1.3.0.339 (Thermo 59. Hayashi, T. et al. Mis16 and Mis18 are required for CENP-A loading and histone Scientific), the results of both search engines were combined and filtered to 1% false deacetylation at centromeres. Cell 118, 715–729 (2004). discovery rate using an integrated Percolator algorithm. Furthermore, high-quality 60. Remmert, M., Biegert, A., Hauser, A. & Soding, J. HHblits: lightning-fast criteria filters such as peptide rank 1 and minimum 2 peptides per protein were iterative protein sequence searching by HMM–HMM alignment. Nat. Methods 9, applied. 173–175 (2012). 61. Lampert, F., Hornung, P. & Westermann, S. The Dam1 complex confers microtubule plus end-tracking activity to the Ndc80 kinetochore complex. J. Cell Biol. 189, Artificial kinetochore experiments. Experiments to examine artificial kineto- 35 641–649 (2010). chore function were performed as described previously with the following mod- 62. Rodal, A. A., Manning, A. L., Goode, B. L. & Drubin, D. G. Negative regulation ifications. TetR-fusion proteins were integrated at the LEU2 locus and expressed as of yeast WASp by two SH3 domain-containing proteins. Curr. Biol. 13, extra copies. Expression was verified by western blotting of whole-cell extracts with 1000–1008 (2003). an anti-TetR antibody. Strains were transformed with mini-chromosomes, selected 63. Mendoza, M. A., Panizza, S. & Klein, F. Analysis of protein-DNA interactions during on drop-out uracil (doURA) and grown non-selectively for 24 h in YPD medium. An meiosis by quantitative chromatin immunoprecipitation (qChIP). Methods Mol. Biol. attenuance of 0.6 at 600 nm was adjusted and 70 µl of a 1:1,000 dilution was plated 557, 267–283 (2009).

SUPPLEMENTARY INFORMATION

DOI: 10.1038/ncb2493

A B

BLASTP Saccharomyces cerevisiae Ame1 4e-09 Zygosaccharomyces rouxii Ame1 Saccharomyces cerevisiae 324 AA 5e-05 Lachancea thermotolerans Saccharo- mycetes ccccccccccc HMMer

Pezizomycotina 1.3e-05 Penicillium marne ei 0.0046 Arthroderma gypseum Mis17 Schizosaccharomyces pombe HMMer 441AA Ascomycota 5.1e-05 Schizosaccharomyces pombe Mis17 ccccccc HMMer

Fungi 0.00035 Ustilago maydis HMMer CENP-U Homo sapiens 418 AA 7.8e-09 Bos taurus CENP-U N-term motif LCR conserved domain 1.2e-08 Mus musculus ccccccccccc ccc 1.5e-07 Canis familiaris coiled coil

Mammalia 0.0051 Homo sapiens

2D structure coiled coil −−−−−−−−−−−−−−−−− −−−−−−−−−−−−−−−−−−−−−−c c c c c c cCCCCCCCCCCCCCCCCCCCCCCCCCCCC−−−−−− −−−−−−−−−−−−−−−−− −−−−−−− −−−−−−− −−−−−−−−−−−− Saccharomyces c. 131 SVTT I DVLSS LF I NLFE [30] LFQT I SDQMTRDLKDI LDI NVSNNE LCYQLKQVLARKEDLNQQI I SVRNE IQE LKAGKDWHDL−−QNEQAK LNDKVKLNKR L−−−NDLTST L [39] NGL LKK I −NK I NE NLS NE LQ 321 Vanderwaltozyma p. 132 SVTT I DVL I S LFTDLFE [29] IFTMVMKSMQNDLKDI LDINVSNNE LCYQLRQ IVNAKEEMNENLVT IRSNLSK IKNGHDWYS L−−QKEHTKLKER IQLNNKL−−−NQLNQNL [26] NGT I AK I −DR I NS I LQKQ L I 308 Lachancea t. 178 SLSSIEVVASLINSLFE[29] LYSTFVQAVLADLRDLLDI NVSNNE LLTQLKQT I KRKSALTQT LLDVRHETADWEANAGA−−−−−DEQLEE LRQKSAVNAK L−−−QALAQDV [27] NGVLAQL−QTFNQS LEK LLE 352 Kluyveromyces l. 127 SMAT I DV L R R L L E T V I L [28] I LTWFLHQSKEDL I T I SE LNLSNNDLLYR LKQ I S LLKNS LSRS LLDVRSA I SSYEGPNNAL−−−−SEEQVALQNFKR I NE L I −−−−−LSQR I [20] KGLLPT L−KT LNES LQNT I G 292 Candida a. 166 RARMLDVLRY LVSSF−−[10] LHQRFQKHLLSS I DNVRDAATSVEDVAKNI ASVQKAKKQLRQMI LDLRKDHTDI GNQLNE LR LQVTNKK LETER LDV I YNQL−−−REVKAYT [31] SGLRPT L−ER I NNKLET LDN 329 Debaryomyces h. 233 RINTLDVIKHLIENFN−[13] I HNE F KNHV LHHLGY LF DAHAN I ND I SQE I NE IQKKKTQ LRNK I F E LKKS HT D I GS E LNKMRGNYQNE RS KF KNF NKMT NK L−−−QS LS S T V [39] SGMFKK L− I L I NKK LSK I EN 408 Penicillium m. 542 GVNAADV LSQ I CKE I LE [29] S I EAYGAQLEGRLFE LSE LLDSNFALGVQLKKAKREAAEMRNRLLEVRRQRQE I T LRMDAVRRKHGEEENAKMSRNV I NNT L−−−HNLDLT L [40] GGL LRQV−KAFNAQLEMTAR 734 Arthroderma g. 460 GVNAADV LSQ I CRET LE [28] AVEAFGAELEGRLFEMSELLDSNYVLAMRLKKEKREVLALRNQLMDIRKEREE IAIQADEVRRRFSADESAKTEHGTINNAL−−−HDLQLAV [41] GGL LNQV−KSFNSQLQKAAA 652 Neurospora c. 540 GVNA I DV LS E LCE GL I A[28] TMETFQEE LRTRLLERT IALDTLHALRKRVRVVQKEKLGLREE I LR IRAERDQVS LRMDAIR IKHEVDSKEALRHIS LSSAM−−−HDIDLAV [39] GGT LKQ I −QDF NAF LE RAAS 730 Magnaporthe g. 484 GVNSVDVLSQ ICEEVTA[28] AVEAFREE LRTNMLEHT I ALDNLQT LRRR LRSAQREKR LLRDQI LNVRKEREQVALRRDALRMKHEKESQKVMNRLHI SSTM−−−QDI E LAV [39] LGV LDQV−KSFNAY LERAAA 674 Pyrenophora t. 516 GITVADTTRTVLEET ID[28] LS LSFKES LNEK LLDLQDANDVLSTNFKK I K LFKRDNAE LRKE I LGLQNSREE I A I EQDS IQAHFEAEKAKVEARNT LSDNM−−−FDI EAA I [36] GGL LANV−KSFNGL LEKAAG 703 Schizosaccharomyces p. 273 T LNE LDAVQQF FQE F I K [12] VKNAFVEQVS IRLLEL IDLLDANR I LSTACKKAANEKLAVQRDLSKLREDRLSVQRKK IQLRNEY IKLSHQQNFLDDIDDFFSQCETVKNEL [26] YGLESQL−VE LQS LLTQFYQ 437 Ustilago m. 404 Y LNDVD I IWS I VDDE LR [14] ALKS LRKSVRSNFLNLSEKTDNRTT L ISQL IRARRQKRLLRKEVFAKRSE LAKLT LEAKERQKEMDDWAKEVAEVRRVNAF LLNLRHQATAW [1] −−−−−−−−−−−−−−−−−−−− 526 Danio r. 223 NPTDLDVVRDAFQEFVT[14] SVEAFSRSFEEQLAE I ITATKELKNVQRENNK INRAINQKRTRLVEANNEL IKGKTQLQKLQKDLDEMEQRFKALTEGSSLLSNLKELNRKY [29] MGAEHQL−KTVNDHLQQVLG 392 Xenopus t. 354 DLNELDVVLAEFEDIVE[14] AVDMFFNSVKEQ ITET IEESQKLKNLKRKNTKLQLE IGKKRKEL ISRREE L IENESKLKKLQKEYAEQEEQKSCLNKAQSF LNNLRE LQNNY [29] FRAESHF−QNINAKLQCFVD 523 Gallus g. 166 DITELDVVLAEFEKIAA[14] AVSAFCSAFEDQVTDL ITEVQE LKNTKKKNAKVVADIKKKRQRLMQVREKLSRTEPQL IKLQKEYAEVEERRSS LRQVVQF LTDLKE LQQDY [29] LQAERHF−QNI NRK LEYALE 335 Bos t. 237 E IMELDVF LSAFENI LL[14] A INKFHSNLKEE L IKMVQE IQMLKT LKRKNAK I ISNIEKKRQRL IEVQDELLQVEPE LKQLQIKYEE LQERKAS LRKAAYF LSNLKQLHQDY [29] LGAEHHL−QNI NYQLENLLD 406 Ailuropoda m. 244 DI TE LDVVLSAFEKT I L[14] A I KKFHSNMKEE L I KMLKEVQVSKT LKRKNTK I I SDI EKKRQR LLE LQDE LLGLDPQLKQLQ I KYDE LKERRAS LRNAACF LSNLKQLHQDY [29] LGAGNHL−QNI NHQLEK LLD 413 Canis f. 244 DI TE LDVVLSAFEKT I L[14] A I KEFHSDVKEE LLKMLKEFQMSKT LKRNNMKV I SDI EKKRQHLLE LQNE LLR LEPQMKQLQ I KHDE LTERKSS LRDAAGF LSNLKQLHQDY [29] LGAENHL−QNI NHRLEK LLD 413 Mus m. 239 DI TE LDV I LSVFEKTF L[14] A INKFYFKMKGEL IRMLKEAQMLKALKMKNTK I IANMEKKRQRL IEVQDEL IRLEPQLKQLQTKYDDLKERKSS LKKSKHFLSNLKQLCQDY [29] LGAENHL−RT I NYQLGKLLE 408 Homo s. 247 DIKELNIVLPEFEKTHL[14] A I ATFYVNVKEQF I KMLKESQMLTNLKRKNAKMI SDI EKKRQRMI EVQDE LLR LEPQLKQLQTKYDE LKERKSS LRNAAYF LSNLKQLYQDY [29] LGAESHL−RNI NHQLEK LLD 416 coiled coil −−−−−−−−−−−−−−−−− −−−−−−−−−−−−−−−−−−−−−−c c c c c c c c c c c c c c c c c c c c cCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCc c c c c c c−−−−−−− CCCCCCC CCCCCCCCCCCC 2D structure

Supplementary Figure 1

Figure S1 Ame1, Mis17 and CENP-U are homologous. (A) Summary Polar regions of compositional bias (low complexity) are in grey. Coiled of the search strategy to deduce phylogenetic relationships between coil regions are marked with a blue bar. (C) Multiple alignment of the Saccharomyces cerevisiae Ame1, Schizosaccharomyces pombe Mis17 C-terminal conserved domain. Species and accession numbers are listed and human CENP-U. A series of BLAST and Hidden Markov Model in Supplementary Table 2. The numbers of residues omitted in two long, searches using sequences of the conserved domain were applied (see mostly unaligned, polar regions are indicated in brackets. S. cerevisiae Methods). (B) All three proteins share a similar architecture: a short (above) and H. sapiens (below) coiled coil and secondary structure motif close to the N-terminus (yellow) and a conserved domain at predictions are derived from the Jpred 3 server (alpha helical regions the C-terminus (green). Regions within the conserved domain, that are in red). The alignment was calculated using MAFFT and graphically are not shown in the multiple alignment are indicated in light green. processed with Clustalx colors using Jalview.

SUPPLEMENTARY INFORMATION

Saccharomyces cerevisiae Okp1 cccccccc 406 AA

Schizosaccharomyces pombe Fta7 cccccc 244 AA

Homo sapiens CENP-Q cccccc 268 AA

2D structure coiled coil ------ccccccccccccccccccccc------

Saccharomyces c. 149 KHWKKPSK* IMIGS ILRLLE TNTVSALDSVFEKYEKEMNQMTHGDNNEVKR IYSKKERLLE I I ------L TK IKKKLRQ- AKFPS ---- R ISERDLDIEY IYSKRQF IQNRYSQELQNNERLEAILSREQNLLEETRKLCMNLKT- - -NNKKRLTEKL- - IQKDL- -HPVL 300 Kluyveromyces l. 149 RNWKRPSKDMLTNFIDLIENNIELASEEVFKQYHQELNQLYPDSE-EQAKLKDSLENKVFDT------TYNIKKRLKR-TKVPS ---- K IWVDNLNMEY IYAKGES IKKRYKSELDRAEAIERQL IREEEHLKSLQEKGETQAK- - -KRKQVLSKELSELSKTM- -HPSL 301 Lachancea t. 142 KAWQRPSKRLLNS I TDILENNIDLALEATFEKYEAECERLAIGR- -SVKHIRRDKERMLHKT------IGK IERQLQS-SKFPS ---- RVKDRDFD IEY IYAKRTF IQNSFSQELEHTDSVEHQVLREQKALEDLKHVQAKLSNRALRQTKLL TEQL - - - SHNL - -HPAL 293 Ashbya g. 126 KTWRRPPRALLRSLADLLSANADNAVEGALSAYHRELTRTLGPN- -ELAAFRHDREQL IRS I ------LHKLRTRLKR- THFPA ---- RVSEHVLD I ED I YARR TF IQQRH I AELNAAEDL ERQL L AEQQAL AAR TAERSD ------ASAKVL LDRL - - - TANL - -HPAL 271 Candida g. 93 KRWRKPSR ILVDSVMQLLE TNSANSVDIVFEKYDDELMRLLRGKRQE IEL IKEDKEKMLGD I ------LQR IEKKLRF - SKFPS ---- RL TENDLN IEY IYEKRRFLQERYVQELRKAE ILEQE IAKERKLLQDAKEL TEN IQQ- - - TNDKRL TDKL - - VRNE I - -HPVM 244 Candida a. 43 KFWQPLDSVNSQSMDNIMN----VALLKTLDESH------SKVSRSITNVWFNSANPESFKSRLSQ-TKLPP [13] NEVADVLSYDSLKRKRTVLETYLSAELKQLQELENHYNQSLLAYKSDLEYLKKFK- - - -STVKNNESKYNEELERK- -REEL 192 Pichia s. 35 RLWEPLDPQTLDSIDRLLN----IALTMTIERYG------NTAETQKVIAKTWLNVLDSKSFRSRLEV-TKVPA [9] NE ISDILSYDQLSRRKTFLETYLNAELKQLLDLQKHYQGMEVAYQLDLEYLEEFR- - - -KTTASHAKDVNKDI TNI - -RESL 182 Debaromyces h. 279 KFWEPLDAQTTSSLEKILN----MCVNRTIEKYKGGS---IVSK------KMIEAQHLLSKNWLNEVNRKSFRSRLNV-TNVPL [11] DNSLDILDYDQLTRRTQFLETYLLAELKQLSDLETYYSDLEMIYNLDLNYLNEFK- - - -KTTEINKSKMIKETIMK- -REKL 435 Ajellomyces c. 42 TKWTTLPDSVQGMIKDLFR----SIERPVVMRHRDER------KRIEAQTAVVAV------RKNLGRRLPR-MPFPP ----G TKDADFDYEAALDENRAL ELQL AAVVNSADL L RAE I TREEAHL AEDKAQL EEL E - - - - KNARAAQTERKRQTRNA- -HPV I 177 Aspergillus n. 41 SKWSTLPEPMQDKIRDMLQ----ALERPVIVRQQNER------KRYEAQAAVQAV------VKNLGKRLPR-MPFPP ----MTKDSVFEYEAAL KEHSAL EASL S TMNDS I VL LNNE I AKEEAFL AKE TKQLQEME - - - - KNAKRAEAERRRLMKNE - -HPAL 176 Arthroderma b. 46 SKWTTLPDSTQEKVKLLFQ----SVELPVITRNRDEK------KRVEAQSALATV------RKNLGKRLPK-MPFPA ----G TRDGSFDYEGALGESRAL ESQL AAATSS I RL LQAE I KREEAEL AKDKL KL EEL E - - - - RNAKAAEAERKRLNKNV - -HPVL 181 Tuber m. 94 AKWGVLTDKGRDEVMEVVR----AAERPVLMTFRREN------RRVEAQEVLHRL------VRKIGNSLTK-VPVPP ---- LGKDAYFNYENLLDKNRALEAILEPDLRQIAGLEAEIEKEQKLLEKEEDYLQELK- - - -KNAIAQENIRRQKSRNM--HPIL 229 Sclerotinia s. 357 TKWEALPPGCLDKISELLI----DVQRPVVLQIHDER------RKTQASTAIQMI------SRRLLSRIRKGMPFPK [3] TNREDDFDFEK ILDYNRALESQLTPVLHSNDLLEAELRKEEVLLENEERYLAELE- - - -ANAKDEAISRRQAARKA- -HPLL 496 Pyrenophora t. 95 DTWPQVSAQVLEQMTSVIR----NAKKDIAETQRDER------RIMAAHNALNPL------VRRLARQLAV-SRIPP ----QAKDVHFN IDKL TERNAQVSRQVTTARHAKQLLSEQVNVAQHLLKKEEDNLEELK- - - - KNAKKWRAEWKHQKKHGRLHPLL 232 Magnaporthe g. 272 SKWFPLDINSVTAVHDILL----NAQRPVLYHLRHGR------RRQEHAASIMTAV------TNRLQRKLVRGLPFPP [19] SGHAEELDYERAVDAAQALEDQLNPTLHSVALLRAELKKQEKALEDDYKSLRTLE- - - -SNARDEARSWKKSLRTT- -H ILA 427 Neurospora c. 402 SKWSPLDPPSIAAVDSIIA----DAHRPILFSLRETR------SDSSRHAHAQDILATF------ASRLHKKLEKGMPFPP [20] PSHEAEFDFEKTVNAIQALERTLDPLLHSVALLKREKEKEEQELERAYRRLRTLE- - - - TNARTQSRGWRERGKRD- -HVLA 561 Schizosaccharomyces p. 17 KRWRPLEDKQRQEIIIIFR----TCSRLVLNTIKSET------RKSLAEEWFMNI------LLKIEAPLRN-LPVPR ---- -KRKES ILFSQLLSSNMQLEQQLYSDLDHINVLQQELKVETARLENEQKSYEEMK- - - -QNMAINNSRLADLKSKL- -HPYL 151 Saccoglossus k. 60 GKWKPVQKSTKDVVVNMID----GAIMSTLNDTSAV------SYSNVQEHLNIM------RARLLSRLTQ-LKVPP ---- - VRYN- -DYKRMDNQCKYMDEVLLQGSNQMMVMEGQMKKLDECLKKKEEELVNVD- - - - ES IKQFESNGAKWKKKC- -HP IL 191 Strongylocentrotus p. 69 SHWKPITKSSQALLVQTID----SAIISVLSETSSV------AYSDVRDVLNSL------KKRILDKSRN-LKAPA ---- -TKRPKF--NTLEAACQKLESEVVEALNQEEQLSEMVEQMTRTVQEKEEILEELE----KAASKTKKKSQ------190 Branchiostoma f. 2 KKWKPISSETKLMTWNIMD----AAVTSVLSSIKPGR------RGFNEVQDQLNNL------RDGISERLST-AKAPA ------KKVDHSAIHAKRKMIREDLKMQEEQLAVLEAEIKRLEK------100 Tetraodon n. 59 ENWQPISRSFITEVENIID----LAILTTLALEKK------EKKESQEHLNAI------KKGFLADCAK-LKVPH ---- -RRQKHV- -EGLFQHHQEETKRLNVAKTTLSALEGDLRAVAIALEKAEEQATSLQ- - - -QECSLLRDQVEEEEEKA- -KQML 189 Xenopus t. 61 NVYKPLANSILEHIEMCLE----AAILSVSHDLR------TPPYAQDSLKCL------KQKLLCHCKK-MKAPT ---- -TKMSIL--KNLKKDVFGELQRMEASEETITCLEDNIENALDASDRIKEQIVDFE----Q------KMDSRKKVI--HFNT 183 Gallus g. 87 ESWQPLSESSRKFLESVMD----SGILSILCQQSK------GKEDVQKHLNLL------KERVLRVFKT-LKVPL ---- -GKLSNL - -NNVLSFQVAQKQMLEMNETALVQLQEE INEAEKSAERTEETILRLQ- - - -HD IQVLKSQLEEEEKKA- -RKLC 217 Mus m. 74 TTWQPLPKDTEEYLQSMME----SVILGILFN-IK------RKEEIQCHLNQL------KKRLLQQCAT-LKVPP ---- -RKLNYL- -KDVSKMLKMEKAQERANEESLASLQEE IDK IVETTESMTES IQSLK- - - -NK IQILTSEVEEEEQEV- -KQVF 203 Homo s. 71 KTWQPLSKSTRDHLQTMME----SVIMTILSNSIK------EKEEIQYHLNFL------KKRLLQQCET-LKVPP ---- -KKMEDL- - TNVSSLLNMERARDKANEEGLALLQEE IDKMVETTELMTGNIQSLK- - - -NK IQILASEVEEEEERV- -KQMH 201 coiled coil ------CCCCC CCCCCCCCCCCCCCCCc cccc 2D structure

Supplementary Figure 2

Figure S2 Multiple alignment of the Okp1/Fta7/CENP-Q conserved domain. A highly conserved tryptophan at the beginning of the conserved domain is marked with an asterisk. Scheme as in Supplementary Figure 1.

SUPPLEMENTARY INFORMATION

Saccharomyces cerevisiae Mcm22 Schizosaccharomyces pombe Sim4 Homo sapiens CENP-K

2D structure coiled coil ------cccccccccccccccccccccc cccccc------

Saccharomyces c. 8 LDVYIKNLENQIGNKRYFLKQAQGAIDEITKRSLDT------EGKPVNSEVFTELLRKP--MFFSE-RADPIGFSLTSNFLSLRAQSSSEWLSLMNDQSVDQKAMLLLQNNINSDLKELLRKLQHQM---TIMDSKKQDHAH 137 Zygosaccharomyces r. 11 LEVYINTIEQQIDNKKFFVKQARDAIGSLQTGGMDV------HSISSEQWQNFMKRP--MFFPE-RSDPIGLGLASTGFVSRQQSSEQWLEHMEVQLNDMQTMIRNQQQMNHEMTVLLELLLHKL------ETPSEDNTI 135 Lachancea t. 7 LSIYVGSVEREIENKGYFLKQVLDAIEVLKLKRH------ADTEESHGSWHEFSRKV--MFTSE-RSDPIGLCLAFCEERLRLKTSKEYLTM--EPLDALKEALKFQKSLNRGLESFMELMKVRN------KGSGPVF 127 Kluyveromyces l. 17 QLAYIAALKRELTAKEQLLRQTIQACNRAEANL------NTDVVNYRVWDQLAQRQ--MYTSS-RSDPIGIASLKQSLQVDNKQ-----SIFNETEQFLVEMVNKQKQFNADMSSLIVMLQERL------RDENQSPEN 135 Ashbya g. 25 ADPLEQQLRQELENKRFFLEQAERALQNLDIEALE------AEGNVNAEAWNKLLETQ--LFIPE-RSDPLGIAVAIDGETKAVNSGAG--AVLSRQIEELEEMVQDQRKLCDDMQALNSILKKRV------NASFVQEEA 148 Candida g. 8 IDEYTAVLEREIENKRYFLKESHDALRDLIESKAERLNGAGSVQG------RRSAINKDVWQKFMEKP--MYLPE-RQDPIGLNLVSARLREKTESMGPWLEV------EKEIVHVEETYLNSLRQLNAAMQDTI---AEFRKNPPKPRE 139 Pichia g. 30 DEVYERRLKKDIEKMEAFLAQAETAEALLQRRVQQTN------VDPKIDTIALHHAASKL--SYIPE-KGDVLGMATASVKTQELVDESRDAYVHLKTFAKDRHEISETYRQLLDDYGEVTQAIDYEL------KNERLQHAE 157 Candida a. 14 PNTYTHVLQKDIARVETFIKECDKAIAQLDESAPVG------TQIIALYETLGVI---PYTPD-KNDTIGTAATTVVLQSMINRYTP----QSTTPIDFSEIIADLNHLRAKKQTALADLQSRN------123 Candida d. 15 PNTYTQVLQKDIARVEGFIKECDKAISMLDESASVE------AQIIALYHTLGAI---PYIPD-KNDTITTAATTVVLEKLVNQYKT----QPTSPTDYLEMITSLKALRSEKLNHISDLQLKL------124 Aspergillus n. 35 LQAYNRRLDETLRELQDQVKRQEDDLRRLREVNAFDLSK- - - - IGSDTWSRV- -AQVQRAKRAYDSLLKSE- - TEFPA-SGSPLPSLLAIEETSRLVKETKVSVSMTAENISSNRQRLKLEEGSLRDAQVIRDGLRKR I - - -QTISSEKTTKEK 176 Gibberella z. 3 LQAYTVSLDQTLRELQQQVRQHEEELERLRSNELRL------PESVA--GQTQVIQAALKEVTESD--PFLPS-PGSLLPALLALRRTHQTIQESNTYLDSQRSTHEQLSRQLEADEARLKDQKLLSDALTARI---QSLRDEVDASTN 137 Magnaporthe g. 12 SERHAASMNRSAVELERMIKGHEESLHR ILQSSGTAAAV------VPSAV- -GSFE IMI TAYQDVTASS- -P ILPS-RDSLLPALLATRRTHRTAAESRAFLEAQEAAVERARRKLEAEESGLEDQR ILRKCLEER I - - -QSLHGELESRTE 149 Neurospora c. 23 DPVHEANLERTIRELKRMTAEAEATLNELKQNIAQNGS------PPPLNPL- -SSMHLLTRAYTSLAESDPTPLLPF-PDSVLPALLAIRKTSQTI TQTNSYLSSSSTLLDQQKRHLENLKTEVREQKELQEALQARV- - -ELLRKGMEAKKE 163 Botryotinia f.. 10 S AAY TTRLDQTLQDLQDRVQEQEAALAKLRAKSSLP ISSG- - -PSPDPQTHL - -RQLRALKPVYDTLSTTP- -PYLPP-AKFALPALLDHRTNTSC IRETQEC I TSSSHDLSTTSSLLQKETSSHQDALSMHSALQTR I - - -SAL TTQL TQQTQ 152 Chaetomium g. 11 SRVYEAQLERTIQELNRRKGELEDALQQLRASAASPGA------AFSAE- -DSLDIMTKAYQDVTEAE- -P ILPP-LGSPLPALLAMRRAHQTIAETNEYADSQAASLEQLKRR IEAEQTSLREQQALQTALEGRI - - -ESLRQGLENREE 147 Glomerella g. 9 GDRYALNLEQSLQELQQNIQQHKEKLAALQGNDGHINK------NLSGAGREPYEVMKSAFEQVASTP- -PFLPF-SESVLPALLALRKTHQTI TESQAYLMSQGASLDRTKRRLE IEQANLADQKLLQQSLEKR I - - -QSLKGEAESRME 147 Pyrenophora t. 29 ASS TEARLAQTVKELQARVQEQQAALDKLRSESDVDLDTAAY-ASEDPRQKL - -QQL IAVKDAYRCLKPTA- - TYLPG-KGSPLPSLLAARTLQQNVQGTKEAIANVKSQLDTKE TTLRREEANLHDANQL TQAMEAR I - - -ERLRAQHADRSQ 173 Paracoccidioides b. 32 HHAYNRQLDQTLKQLQDQVKRQEYALQELRLSTATTLPS- - - -PGLSPHARL- -AQTRRATKAYNSLSQTD- -PVLPP-PDSLLSPLLALRETFRQILDLKSS ISTVAHQLSTDRERLKAEEAALRDATL ISTGLDER IRDLQTMRAGDREKEA 176 Tuber m. 46 QTEYEASLVQATTELRRRLMDEEMALSELSSQI TIPDDIA- - -SRESPHRKL- -RRLKAEKGAYDVAFQTI - -DYLPG-RDSPLNLLLAYRNVGRV IEQTKVAIP IVNERLRSVQAQIKAEEGNLEEQNRLTEALRKR I - - -GELEEE- - -EQR 185 Schizosaccharomyces p. 29 RLALFESLVQSDQLLVEKIEKWEKKVSKFTENSKS------EKRKLQLELFQEIIDLK------KCIAFKPIIGKNAIDNNISDLKKNLHSNKKLEAVLKEELHQIKKFSSDLQSLKSSMGERQKQQAAMSRRGKK--- 152 Ustilago m. 60 LFAVYEDEKAKNDALARTIDDRLQALASIKKSQAD------EVGTDDRKRLELEWQQLEFAIRALEQEK- -QHLPALDSATVLDIRTRAELYDLNLQYTSALAMLRADLSSERDALARETQLRSDLDVVSAGLSKRF- - -VKLQRKSQQDLT 200 Nematostella v. 12 KTSYMQDLKDQCEGLWLHIHQTHARLRSRAENMTCF- -LPGEENLEK IHEE ILKAKEKRLQAEYAAIESSG- -VEVPS-MKEEYLDASLKQQLAQNTSQLDDTLL ILKSKRKE IAEQLEREKE ILEQHKDIE ISLLNKT- - -KKIEDTKENFQS 157 Camponotus f. 2 ----EYTLEDAFETLKDKKIEREKNLEDLINKISNNEIFTLKSSNVSSFDDVKLKQEKLYQSLLKEIQQIE--PEDYPILRTSDLRVEVITEMEEEICNMQELLNSLKCDFSDIKEDITYLKNKKDGLEKMRGAYLDMA------134 Nasonia v. 3 --SVEFTLEDTYEHLKDVKCEKEQEVKELIDIVQNNKVRTLQ-PNVTP-EEILLKRYKFAQSLAKFVAQEQ--PKDLSIHNTPDFYSGALKGLEDEVQAMEELDKITDEEIAELESDIAYMEKKQTALELIKEASLNSG------135 Strongylocentrotus p. 7 MQVSEL ILKKECEESWKQVNRNQAVLLSQPRHGPVG------DSQNALAAISQMKARCLQGE IDAVKAKK- -VKVLP-NDPVILESLLKEDLERNVSQLEECLSVVEGRRRELEDDLARQKELLSQHQDMNRALESRL- - -REAEEGGNTSRP 147 Branchiostoma f. 26 VASAKCEMKQECVEVWRDVNVLQQRLQEE TGDASPT------DRPSPLLAVSRAKQLQLQAQVEVLESQQ- - PQLMS- THKKVVEAVLVNHLKES IEQLSELLPAMQAQQKELSASLELEEQVLHQRQELGRMLEARL - - - AEVPEAHE TSV- 165 Danio r. 24 MSAAKAVLLDECEEQFGLLQKLQNEIIMADTEAYDD------QSNEAVNRLIAITSELEQWQEIE--PKLLT-TNTEVLLAVGKKELQRLHSQLKMVLSCSQAKLDAQKKILKREQTWQSEKQEALNSVTEKI---ARLQQEKQKSD- 158 Xenopus l. 19 PEEAREELLQQCEAIWKQMEECQSKLTMSGPETLPE------TDVQLYLLMMQVKALTAEYEQWQKRT- -PE I IS-DNQDVLLAVGKEELEK IAQELEMVLSSVQAKNKKLKKDLEKEQNWLDGQQKIVDALTSRQ- - -DELKNQLTAFSE 157 Gallus g. 7 P IDAKEELLRECENIWKEMEKCQSKLTLLAAEPVPE------SDAKVSLLLTRMQALRAEYHQWQKRN- -PEL IS- TNPEVLLLLGEEELQKVKKELEMVLSAVQLKNEKLKEDLEREQQWHDEQVQILNAFKE IE- - -EEMKKEVVTDSE 145 Ornithorhynchus a. 19 AAEAKEAL INECEELWKQMEECQNKLSSLGPETLTD------SDAQLSLL IMRVKGLTAERNQWQKRT- -PE IVP-NNQDILSALGKEELQKLDQDLEMVLSAVQAQNVKLKEDLRREQQWLNEQEQL IESLKVTH- - -DELKTQEVHFSE 157 Mus m. 54 VIDTEEELIKECEEMWKDMEDCQNKLSLIGTETLTN------ADAQLSLLIMQMKCLTAELGQWKKRK--PEIIP-LNEDVLLTLGKEEFQKLRCDLEMVLSTIQSKNEKLKEDLEREQQWLDEQQQILDTLNVLN---SDVENQVVTLTE 192 Homo s. 17 VTNTEEEL IRECEEMWKDMEECQNKLSL IGTETLTD------SNAQLSLL IMQVKCLTAELSQWQKKT- -PETIP-LTEDVL I TLGKEEFQKLRQDLEMVLSTKESKNEKLKEDLEREQRWLDEQQQIMESLNVLH- - -SELKNKVETFSE 155 coiled coil ------ccccccccccccccccccccc------ccccccccccccccccccccccccccccccccccccccccccc cccccc------2D structure

2D structure coiled coil ------ccccccccccccccccccccc------Saccharomyces c. 138 ------IRTRKARNKELWDSLADFL-KGYLVP----NLDDNDESIDSLTNEVMLLMKRLIEHDLNLTLNDFSSKT------IPIYRLLLRANIITVIEGSTNPGTKYIKLIDFNE 235 Zygosaccharomyces r. 136 QETPV------QRNHTLRNELKNFI-RDFLSL----DLADSQNTAEQVCSDVMVVIERLINYDTNLTTTDFPPST------KGLFRLLLRGNLITLNEVGDKR---YVKLTDFAS 230 Lachancea t. 128 DQETGDLT------KHQNSVLWRKLQIFV-ENFLAI----DTSSESTEADIVAGKIFEVLSHLIKGQKAVQKDCVQQQL------PGFYRLLIKTKCTYTDNEGC------VRLIFHES 223 Kluyveromyces l. 136 AGDPI------GMIQEMKSKIFDLI-RNHMAV----DISTPIYSADIAAADMTNLMRRLINGDSTLTVHDFQPDC------MNLFRILDTAYLLQKIDMNGTI---YVKLEDLT- 229 Ashbya g. 149 TETPC------SSEEVEQIKDSFL-SEFLVP----DLCRDASDWPATNTAMEDVLRRLVKLDPQLRLTDFAPHC------MGLYRMLSKANVIQEIFTDGHD-NPYIRLYNLLE 244 Candida g. 140 ELVSKDYS------LSSLKTQHESLHKELKEFV-TRYLEP------NAPENTSAEEMLQLISTLVQGK-TLDKDQFKNS------QSLFRLLMKGMLLENTDTNS------YKLIDLVS 232 Pichia g. 158 EEPSLDIKHQLDILDRKLVGASEHEQTLTKHLKRLVVTSLMIE----GWDRDSIEFKEGSSMCMQLVQQLVQAAVDSQITEQPQWVNVRPDAVL------QKFITQLMLGDMIYVEEVNGVQ---RLRLREYGL 278 Candida a. 124 ------FASPLPEKLAEARELEKLLNSYI------AKINNQ-- 152 Candida d. 125 ------AGEFASPLHAKLAEASSLHDLLSSYI------AKINSK-- 156 Aspergillus n. 177 KTASQLARELAEQEEEKQKELDKAAEEMKTTLYNFV-DETLAPMLAAGDNQPANRREAAALEMRNLLDALLEAG------[61] 306 Gibberella z. 138 VTPEEGAKERLQELRTKKKTFDRDTTKLMKVLLRFI-ANHLAPMLAAEGSGRQNEVTAAAAEMRQLTEELLNTLSEAQGDNSASYVQLSRE------SAAARFLVRSKIAQFHPRDAMR----LRLVDFGR[59] 312 Magnaporthe g. 150 PTPQEISHARIQELRLKKRTYDRETSALLRTFNKFI-DEHLAPLLAAKTGQNWDETAAAASDMRELTEQLLQKLVESEGDHWASYVHIPKE------SAAARFLIRSRVAQFHPKDATR----LRLIDFGK[62] 327 Neurospora c. 164 MTPEELAKEELEKMEGERNRYDEETKKLVRELNWFI-EEYLGPLLAAKRKRERDEASAAGAEMRSLIEKLLNKSMDAGGDSSAAYVKIERE------TAAARFLVRSKVAVFHPRDATR----LRLVDFGR[72] 351 Botryotinia f.. 153 KSPSQIFSELQSSFQAGKTHYDTETGNLIRAFNTFI-DSHLAPMLAAEEEEEWDETRAAAAEMRELAEQLLNGLLEADETGSDGYVTLERD------SAAARFLVRSKVAQYHFKDARK----LKLIDFEK[67] 335 Chaetomium g. 148 K TEEQ I AKER I AEL KKQKDLWDKQTSSLMKQLDWF I -DEHLGPMLAARRKQKGNEAAAASAEMRDL[62] TEQLMNRLMEADGD TSAAYVE I ARE ------SAAARFL VRSKVAMFHPRDARK - - - - L RL VDFGK 325 Glomerella g. 148 LTPQQVVQERLDELRQKKKRYDTETSRLLKALRKFI-DDHLAAQLAAQKRGEADEAAAAGREMRALTEQLLNQLVESGGDSTTAYVKLSKE------TAAARFLVRSKVAQFHPRDATK----LKLIDFGR[61] 324 Pyrenophora t. 174 KTPAQLARELIAAKRAQKDAYDADMQRLGQAMNDFI-NDYLSNMLAAEEKDPPTEAEAADAEMRQLIENLFATLVGPGG--GKAYFQLERD------SAASRFLVRAKIAQFHPRDARK----LRLIDFGR[63] 350 Paracoccidioides b. 177 KAPTQLAKEVMQQQRRRKAELEKKTTALREALTGFV-DEHLASMLAAQGGRESNKREAAGAEMHALLDSLLAVA------ETSSYIELEQE------TAASRFLVKAKIAQFHPRDARR----LRLIDFAR[88] 374 Tuber m. 186 VDVGVVHDDAIKAMNKKRKELLKRTSKLLRELLAFLKDGGLARMLAA--QGEIGPEESMVEEFKILLEELMNASLETT-----QYVNLKKD------GAAARFLVRAKVAAYHPKDAGK----LKLLDFGS[42] 337 Schizosaccharomyces p. 153 ------SIIHSAIIQEKERYEKESQELFTFLDGFL-DEHAPSLLEGSLNFENISKKDLTEEFKQVIENLMNNSIIPGS----PYIEVKN------ERIVSFLVLASLCTVDPQDPSK----VKLIPFSD[19] 275 Ustilago m. 201 S------ESAIRELNRKFKREEQRFKELLAQLIDMG-TSLFGP------DSRKVVTLRHYLDEFMNQAWDKPLD---PWVSSTKLAMRRTGGEVDDAMIEFFIRANIIVPHPKDSRR----WRLVGFHK---- 309 Nematostella v. 158 G------TSKVRELERKQRIAETQLKQVMRKLGVFL-REHFPL------EQTLDPTINYVTLQEMLEELMNTQHDRPHD---PFMQLQPHHW------PPYVELLLRCGLVLRHNQDSNQ----IKLESFNV[16] 279 Camponotus f. 135 ------ETFANETYEKELVITKRIFQKVKNDLYTVV-DTIFPD------NEGFKELLAALTSAYMKGGDD---VYVDVIPDV------LYYVNFLIENDIVQYHRNNKTR----IRMTELL----- 229 Nasonia v. 136 ------DKEAATNFEAEMIVTKQIFGDVKKDLAFVV-DTLFPN------NGAFQNLLFVLTRALNKGGDD---LYITVTPDV------LDFVNFLEEADIIQFHPNDKTK----IKLRNL------229 Strongylocentrotus p. 148 NE-----DSQIKILKRKQKKAGEIQTVLLKQIGDFV-GNHFPL------AGQINYISLLEIIELLMNKGMDGDTD---PYVSIHEAFW------PPYIELLLRCDIASRHPQDASR----IRLVPLHL-[8] 258 Branchiostoma f. 166 ------DSRVKTFRRKLKQAEQFHTDLVQRLAEFL-DTHFPP----PVDKTLDVDDPQWTPLNTMIETLINKSVDSPHD---PYIPMEPRYW------PPYIEMLLRCGIAQRHPDDSNR----LKLAPFNL[15] 288 Danio r. 159 ------HSVSQEMKRKIHKLKEYHSSLMEMLSDML-AEHFPL------AAYDAAEINLISLSEIIERLTNKTLETPHD---PYVTIDDTFW------PPYTEMLLRNGIAMRHPEDCNK----IRLENFF-[13] 275 Xenopus l. 158 ------KRVYQDIAGKLYKVRAHKEELLSALGDFL-EEHFPL------GQSGKKSVQLMTLHEILEILINKLMTTPHD---PYLVLEPHHW------PPYIEMLLRYGIALRHPEDPKR----IRLEAFHQ[12] 274 Gallus g. 146 ------KRVFQELKNQMLELKEYKKKLMNALGEFL-EEHFPL------RYSEEPPEQVITIHEILEILLNQLMSTPHE---PYVTVDDSFW------PPYLELLLRSGIVLRHPEDPNR----IRLEAFHE[10] 260 Ornithorhynchus a. 158 ------ERAVRELKSKMIKIKAYKEELLSALGEFL-DNHFPL------KKCAGPTVQLISLHEILEILISKLLTAPHE---PYVAIQDSFW------PPYIELLLRSGIALRHPEDPNR----IRLEAFHQ[12] 274 Mus m. 193 ------SRIFNELTTKIRGIKEFKEKLLLTLGAFL-DNHFPL------KNIQDSNAQLITLNEILEMLINRMFDVPHD---PYVKIRDSFW------PPYIELLLRYGIALRHPEDPSQ----IRLEAFHQ-[9] 306 Homo s. 156 ------SRIFNELKTKMLNIKEYKEKLLSTLGEFL-EDHFPL------KNIQESSVNLITLHEMLEILINRLFDVPHD---PYVKISDSFW------PPYVELLLRNGIALRHPEDPTR----IRLEAFHQ-[9] 269 coiled coil ------2D structure

Supplementary Figure 3

Figure S3 Multiple alignment of the Mcm22/Sim4/CENP-K conserved domain. Scheme as in Supplementary Figure 1.

SUPPLEMENTARY INFORMATION

Homo sapiens CENP-W Schizosaccharomyces pombe C17G8.15 Saccharomyces cerevisiae YDR374W-A

2D structure

Homo s. 1 -MALSTIVSQRKQIKRKAPRGFLKRVFKRKKPQ-----LRLEKS-GDLLVHLNCLLFVHRLAEESRTNACASKCRVINKEHVLAAAKVILKKSRG------88 Mus m. 1 -MAPSTTVTRR--VKRKAPRAFLKRTLKQKKPH-----LGLGRC-CDLLIHLNCLLFIQRLAEESRTNACESKSRVIKKDHVLAAGKVILKKSRG------86 Gallus g. 1 ------MRRTVPRGTLRKIIKKHKPH-----LRLAAN-TDLLVHLSFLLFLHRLAEEARTNAFENKSKIIKPEHTIAAAKVILKKSRG------76 Xenopus t. 1 ------MKGAIPRGTLRSILKKHQPT-----MRREAE-LDVLVHLNCLLFVRRLVEEAQLKALENKSSVIKPEHIKAVAKTALKKSKG------76 Danio r. 1 ------MSKKAPRAALKHHMKK-NAN-----IRIGKN-ADLMAQLNLLMVLHRLAEECRVKAFEEKSATIKTHHVRAVAKKLLKSTRG------75 Saccoglossus k. 1 ------MKTLKRKFPRTTVRTSMRKMMSEKAGAPIHLNKN-VELLTYLMYMKFIQRLAEKTDEVARENKETVITEEHVNKVMKGVLKSCQG------84 Schizosaccharomyces p. 1 ------MSYPKARIRRFFRQ-HCQ-----RSLESS-ALDLVYLDYCLFLQSLLKEANIEAAKTGERRVQPEHVRSIQRKILAQFKG------73 Pyrenophora t. 1 ------MPQTLYPRATVKKIVKA-HSN-----RGVSRN-VDILIYLNYIVFMQDLLKEANIKSKQHGEKGVSAKSIKRVSEGVLRKYKG------76 Ajellomyces c. 1 ------MVVTPKLYPRATVKRIVKA-HSK-----RNISKN-ADILIFLDYMLFMQELMREASIRARRAGEKVISARSVRKVTEGTLRKFKG------78 Gibberella z. 1 ------MALGKKPYPKATVKKIIKA-HSN-----HNIKKN-ADVTIFLDYILFMETLVKEAAIHAKQSGERGLSARSVKKVTRDTLTKFKG------78 Debaryomyces h. 1 ------MVSTNKASYPKSTFKKVLNS-KTK-----YKFKNDESDLLIYLLYVDYVNKLMNKGRDIQERAGSAEISTNHLELANNELSKHYRG------80 Ashbya g. 1 ------MYEPVSNYVLEELREQ-IPK-----ISVRGEPESILLSCLYQELLSRVIEESKRFADRDSTKHITAEHLDEAVEALLGDVDRGADGAWP83 Kluyveromyces l. 1 ------MSEHLRRIVVDGLLGRIRDNTLGT-----LDEPSEAEETLIGCLFADLIRKLMDEAKLQAEKDGTRTISVGHLRDAKDLILESSPDNNED--- 85 Zygosaccharomyces r. 1 ------MGFDLGQYLLDQWRKR-YEF-----VEEPSESERLILSSGFQEMLRKLLVEAQSNAHRDGFNEVRPAHLEAALDELLDA------73 Saccharomyces c. 1 MDTEALANYLLRQLSLDAEENKLEDLLQR-QNE- - -DQESSQEYNKKLLLACGFQAILRK------ILLDARTRATAEGLREVYPYHIEAATQAFLDSQ------89 2D structure

Supplementary Figure 4

Figure S4 Multiple alignment of the CENP-W conserved domain. Scheme as in Supplementary Figure 1.

SUPPLEMENTARY INFORMATION

A 670 kDa 158 kDa

MW A3 A4 A5 A6 A7 A8 A9 A10 Superose 6 PC3.2/30 130 70 55 Cnn1-Flag 35 25 HF Cnn1p

Cnn1-Flag

130 6xHis-Dsn1 70 55 Nsl1 35 Mtw1p Dsn1 Mtw1 Mtw1 complex 25 Nnf1p Nnf1 Nsl1p

15 Homo sapiens CENP-W Mtw1 complex

130 Schizosaccharomyces pombe C17G8.15 6xHis-Dsn1 70 HF Cnn1p Saccharomyces cerevisiae YDR374W-A 55 Cnn1-Flag + Mtw1p 35 + Nsl1 2D structure Nnf1p 25 Dsn1 Nsl1p Mtw1 Mtw1 complex Homo s. 1 -MALSTIVSQRKQIKRKAPRGFLKRVFKRKKPQ-----LRLEKS-GDLLVHLNCLLFVHRLAEESRTNACASKCRVINKEHVLAAAKVILKKSRG------88 Nnf1 Mus m. 1 -MAPSTTVTRR--VKRKAPRAFLKRTLKQKKPH-----LGLGRC-CDLLIHLNCLLFIQRLAEESRTNACESKSRVIKKDHVLAAGKVILKKSRG------86 Gallus g. 1 ------MRRTVPRGTLRKIIKKHKPH-----LRLAAN-TDLLVHLSFLLFLHRLAEEARTNAFENKSKIIKPEHTIAAAKVILKKSRG------76 15 Cnn1-Flag Xenopus t. 1 ------MKGAIPRGTLRSILKKHQPT-----MRREAE-LDVLVHLNCLLFVRRLVEEAQLKALENKSSVIKPEHIKAVAKTALKKSKG------76 + Danio r. 1 ------MSKKAPRAALKHHMKK-NAN-----IRIGKN-ADLMAQLNLLMVLHRLAEECRVKAFEEKSATIKTHHVRAVAKKLLKSTRG------75 Mtw1 complex Saccoglossus k. 1 ------MKTLKRKFPRTTVRTSMRKMMSEKAGAPIHLNKN-VELLTYLMYMKFIQRLAEKTDEVARENKETVITEEHVNKVMKGVLKSCQG------84 Schizosaccharomyces p. 1 ------MSYPKARIRRFFRQ-HCQ-----RSLESS-ALDLVYLDYCLFLQSLLKEANIEAAKTGERRVQPEHVRSIQRKILAQFKG------73 Pyrenophora t. 1 ------MPQTLYPRATVKKIVKA-HSN-----RGVSRN-VDILIYLNYIVFMQDLLKEANIKSKQHGEKGVSAKSIKRVSEGVLRKYKG------76 elution volume Ajellomyces c. 1 ------MVVTPKLYPRATVKRIVKA-HSK-----RNISKN-ADILIFLDYMLFMQELMREASIRARRAGEKVISARSVRKVTEGTLRKFKG------78 Gibberella z. 1 ------MALGKKPYPKATVKKIIKA-HSN-----HNIKKN-ADVTIFLDYILFMETLVKEAAIHAKQSGERGLSARSVKKVTRDTLTKFKG------78 Debaryomyces h. 1 ------MVSTNKASYPKSTFKKVLNS-KTK-----YKFKNDESDLLIYLLYVDYVNKLMNKGRDIQERAGSAEISTNHLELANNELSKHYRG------80 Ashbya g. 1 ------MYEPVSNYVLEELREQ-IPK-----ISVRGEPESILLSCLYQELLSRVIEESKRFADRDSTKHITAEHLDEAVEALLGDVDRGADGAWP83 B Kluyveromyces l. 1 ------MSEHLRRIVVDGLLGRIRDNTLGT-----LDEPSEAEETLIGCLFADLIRKLMDEAKLQAEKDGTRTISVGHLRDAKDLILESSPDNNED--- 85 Zygosaccharomyces r. 1 ------MGFDLGQYLLDQWRKR-YEF-----VEEPSESERLILSSGFQEMLRKLLVEAQSNAHRDGFNEVRPAHLEAALDELLDA------73 fraction Superose 6 PC3.2/30 Saccharomyces c. 1 MDTEALANYLLRQLSLDAEENKLEDLLQR-QNE- - -DQESSQEYNKKLLLACGFQAILRK------ILLDARTRATAEGLREVYPYHIEAATQAFLDSQ------89 A3 A4 A5 A6 A7 A8 A9 A10 2D structure Spc24 130 6xHis-Dsn1 70 Nuf2 Ndc80 55 Nuf2-eGFP Spc25 Ndc80 35 + Mtw1p 25 Spc25 Nsl1 Nsl1pNnf1p Dsn1 Spc24-6xHis Nnf1 Mtw1 15 Mtw1 complex Supplementary Figure 4 Ndc80 complex + Mtw1 complex

130 Ndc80 6xHis-Dsn1 HF 70 Nuf2-eGFP 55 Cnn1-Flag Spc24 + 35 Mtw1p Nuf2

25 Spc25 Spc25 Spc24-6xHis Ndc80 + Cnn1-Flag Nsl1p Nnf1p 15 + Nsl1 Ndc80 complex Dsn1 + Mtw1 Mtw1 complex Nnf1

Mtw1 complex elution volume

Figure S5 Cnn1 and the Mtw1 complex are competing binding partners combined sample. B. Size exclusion chromatography of Ndc80 complex and for the Ndc80 complex. A. Size exclusion chromatography of Cnn1-Flag Mtw1 complex in the absence (top panel) or presence (lower panel) of Cnn1- alone (upper panel), recombinant Mtw1 complex alone (middle panel), or in Flag. Note that Cnn1-Flag fails to efficiently co-elute with the Ndc80-Mtw1 combination (lower panel).Supplementary Note that the elution position isFigure unchanged in5 the complex assembly.

SUPPLEMENTARY INFORMATION

* non-specific

0 120 min

Nuf2-13xmyc

Input

Mtw1-6xFlag

FigureSupplementary S6 Uncropped Western blots Figure corresponding 6 to Figure 5b,c.

SUPPLEMENTARY INFORMATION

0 10 5 2.5 1.25 0.625 0.312 μM Tubulin kDa SN P SN P SN P SN P SN P SN P SN P

100 Nuf2-EGFP Ndc80

35 Spc24-6xHIS 25 Spc25

100

55 Cnn1-FLAG

25 20 mM KCl

100 Nuf2-EGFP Ndc80 70

55 Cnn1-FLAG

35 Spc24-6xHIS

25 Spc25 20 mM KCl

Supplementary Figure 7

Figure S7 Microtubule cosedimentation experiment using Ndc80 microtubules. Note that Cnn1-Flag fails to cosediment with microtubules complex alone (upper panel), Cnn1-Flag (middle panel) or a combination on its own, but is recruited into the pellet fraction in the presence of the of both (lower panel) with different concentrations of taxol-stabilized Ndc80 complex.

SUPPLEMENTARY INFORMATION

a b

Wip1-GFP Nuf2-mCherry

14000

12000 cnn1∆ 10000

8000 WT 6000

4000 Nuf2-mCherry intensity (AU)

2000

wild-type cnn1∆

Supplementary Figure 8

Figure S8 Wild-type and cnn1 deletion cells (identified by the lack anaphase cells was quantified. Anaphase kinetochore clusters in ten of Wip1 signal at kinetochores) were mixed, imaged on the same cells were analyzed and error bars represent standard error of the microscopy slide, and the Nuf2-mCherry signal in large budded mean.