1 Cell-cycle dependent phosphorylation of yeast pericentrin regulates -TuSC-

2 mediated microtubule nucleation

3

4 Tien-chen Lin1,3, Annett Neuner1, Yvonne Schlosser1, Annette Scharf2, Lisa Weber2 and

5 Elmar Schiebel1,*

6 1-Zentrum für Molekulare Biologie der Universität Heidelberg (ZMBH), DKFZ-ZMBH Allianz,

7 Im Neuenheimer Feld 282, 69120 Heidelberg, Germany

8 2-Zentrum für Molekulare Biologie der Universität Heidelberg (ZMBH), DKFZ-ZMBH Allianz,

9 Z-MS, Im Neuenheimer Feld 282, 69120 Heidelberg, Germany

10 3-The Hartmut Hoffmann-Berling International Graduate School of Molecular and Cellular

11 Biology (HBIGS), University of Heidelberg, Heidelberg, 69120, Germany

12

13 *Correspondence: [email protected]

14 Key words: pericentrin, Spc110, Spc98, spindle pole body, Mps1, Cdk1, gamma-tubulin

15 complex, phosphorylation, mitosis, microtubules, oligomerization

16 17 Summary

18 Budding yeast Spc110, a member of -tubulin complex receptor family (-TuCR), recruits -

19 tubulin complexes to microtubule (MT) organizing centers (MTOCs). Biochemical studies

20 suggest that Spc110 facilitates higher-order -tubulin complex assembly (Kollman et al.,

21 2010). Nevertheless the molecular basis for this activity and the regulation are unclear.

22 Here we show that Spc110 phosphorylated by Mps1 and Cdk1 activates -TuSC

23 oligomerization and MT nucleation in a cell cycle dependent manner. Interaction between

24 the N-terminus of the -TuSC subunit Spc98 and Spc110 is important for this activity.

25 Besides the conserved CM1 motif in -TuCRs (Sawin et al., 2004), a second motif that we

26 named Spc110/Pcp1 motif (SPM) is also important for MT nucleation. The activating Mps1

27 and Cdk1 sites lie between SPM and CM1 motifs. Most organisms have both SPM-CM1

28 (Spc110/Pcp1/PCNT) and CM1-only (Spc72/Mto1/Cnn/CDK5RAP2/ myomegalin) types of

29 -TuCRs. The two types of -TuCRs contain distinct but conserved C-terminal MTOC

30 targeting domains.

31 32 Highlights

33 1. Phosphorylation of Spc110 N-terminal domain (N-Spc110) regulates -TuSC 34 oligomerization.

35 2. Both SPM and CM1 are required for N-Spc110 to activate -TuSC oligomerization.

36 3. The conserved N-Spc98 is required for Spc110-mediated oligomerization.

37 4. -TuCRs can be categorized into CM1-only (Spc72/Mto1/Cnn/CDK5RAP2) and 38 SPM-CM1 (Spc110/Pcp1/PCNT) type of receptors.

39 5. While the SPM-CM1 type is targeted via the PACT domain to centrosomes/SPBs, 40 the CM1-only type has either MASC (Mto1 and Spc72 C-terminus) or a CM2 motif 41 for MTOC targeting. 42 Introduction

43 The budding yeast spindle consists of ~40 microtubules (MTs) that extend between the two

44 opposed spindle pole bodies (SPBs). Because of the closed mitosis in yeast, the SPBs

45 remain embedded in the nuclear membrane throughout mitosis. Cell-cycle regulated

46 spindle assembly begins in S-phase with nucleation of MTs onto the surface of the newly

47 assembled SPB. As soon as MTs emanate from both SPBs, they interdigitate (pole to pole

48 MTs) or attach to the kinetochores (pole to kinetochore MTs) at the end of S phase, forming

49 the bipolar spindle (O'Toole et al., 1999).

50 The -tubulin complex is the core player in MT nucleation. In budding yeast Saccharomyces

51 cerevisiae, two molecules of -tubulin (Tub4) assemble together with one molecule of

52 Spc97 (ortholog of human GCP2) and Spc98 (ortholog of human GCP3) into a tetrameric -

53 tubulin small complex (-TuSC), which is conserved in all eukaryotes (Geissler et al., 1996;

54 Guillet et al., 2011; Knop et al., 1997). The purified -TuSC of budding yeast is able to self-

55 oligomerize into symmetric ring-like structures under low salt buffer conditions. The

56 diameter and the pitch of the -TuSC ring resemble that of MT cylinder, suggesting that the

57 -TuSC ring functions as a MT nucleation template. However, the in vitro nucleation activity

58 of the -TuSC assemblies remained poor, presumably because of the suboptimal spacing

59 of every second Tub4 within the -TuSC ring that blocks interaction with tubulin in the MT

60 cylinder (Choy et al., 2009; Kollman et al., 2010; Kollman et al., 2008). The concept of-

61 TuSC oligomerization is further supported by in vivo measurements of budding yeast -

62 tubulin complex components on detached single MT nucleation sites. The -

63 tubulin:Spc97:Spc98 ratio was 2.4:1.0:1.3 with a total of ~17 -tubulin molecules per

64 nucleation site (Erlemann et al., 2012), suggesting a slight excess of -tubulin and Spc98 65 molecules over Spc97 in the assembled MT nucleation site.

66 In most eukaryotic organisms, such as fission yeast, Drosophila, Xenopus and human,

67 multiple -TuSCs further assemble with additional GCP family members (GCP4-6) into the

68 larger -tubulin ring complex (-TuRC) (Anders et al., 2006; Gunawardane et al., 2000;

69 Murphy et al., 2001; Zhang et al., 2000). However these additional GCP proteins are not

70 encoded in the budding yeast genome.

71 Various proteins are involved in the recruitment of -tubulin complexes to microtubule

72 organizing centers (MTOCs) such as centrosomes and SPBs. The small protein Mozart1 is

73 encoded in most eukaryotic genomes except the one of budding yeast (Teixido-Travesa et

74 al., 2012). In Schizosaccharomyces pombe and Arabidopsis thaliana Mozart1 interacts with

75 the GCP3 subunit of -tubulin complexes (Batzenschlager et al., 2013; Dhani et al., 2013;

76 Janski et al., 2012; Masuda et al., 2013; Nakamura et al., 2012). In S. pombe Mozart1

77 appears important for the -TuSC recruitment to SPBs (Dhani et al., 2013; Masuda et al.,

78 2013).

79 Besides Mozart1, a group of conserved proteins called -tubulin complex receptors (-

80 TuCRs) are required for targeting -tubulin complexes to MTOCs. Most of them carry a

81 highly conserved centrosomin motif 1 (CM1) that interacts with GCP subunits of -tubulin

82 complexes (Sawin et al., 2004). How Mozart1 and -TuCRs cooperate is not understood.

83 However, in budding yeast cells that lack a Mozart1 gene, -TuCRs are the sole factors

84 responsible for -TuSC recruitment to SPBs. Spc110 is the budding yeast homolog of

85 pericentrin (PCNT) and functions as -TuCRs at the nuclear side of the SPB (Knop and

86 Schiebel, 1997; Sundberg and Davis, 1997). The N-terminal Spc110 encompasses the

87 CM1 that interacts with the Spc98 subunit of -TuSC (Fong et al., 2008; Knop and Schiebel, 88 1997; Nguyen et al., 1998; Samejima et al., 2008; Sawin et al., 2004; Vinh et al., 2002;

89 Zhang and Megraw, 2007). In addition, the N-terminal region of Spc110 is phosphorylated

90 in a cell-cycle dependent manner. Phospho-Spc110 appears as cells progress from S

91 phase, continues accumulating during mitosis and vanishes at the anaphase onset

92 (Friedman et al., 1996; Stirling and Stark, 1996). Spc110 phosphorylation accounts for the

93 impact of Cdk1 and Mps1 kinases on spindle dynamics (Friedman et al., 2001; Huisman et

94 al., 2007; Liang et al., 2013). However, a clear understanding behind this observation is

95 lacking. Interestingly, when -TuSC and an N-terminal fragment of Spc110 (amino acids 1-

96 220 of Spc110; Spc1101-220) were co-expressed in insect cells, a filament-like -TuSC-

97 Spc1101-220 complex formed. The nucleation capacity of this purified -TuSC-Spc1101-220

98 complex exceeded that of the -TuSC alone (Kollman et al., 2010). Thus, Spc1101-220

99 influences -TuSC properties with yet unclear mechanism.

100 Here we have tested the possibility that phosphorylation of the -TuCR Spc110 regulates

101 MT nucleation by inducing -TuSC oligomerization. Single particle analysis of -TuSC

102 incubated with phospho-mimetic Spc110 mutant proteins showed that Mps1 and Cdk1

103 promoted MT nucleation through Spc110 phosphorylation. Phosphorylated Spc110 and the

104 interaction with the N-terminal domain of Spc98 induce -TuSC oligomerization. In addition,

105 bioinformatic analysis revealed a conserved region around T18, that we named

106 Spc110/Pcp1 motif (SPM). SPM and CM1 motifs are both important for -TuSC binding and

107 oligomerization. A comparison of -TuCRs for the presence of SPM and CM1, identified

108 SPM-CM1 (Spc110, Pcp1, PCNT) and CM1-only types of -TuCRs (Spc72, Mto1, Cnn,

109 CDK5RAP2, myomegalin) in most organisms. While the SPM-CM1 type of -TuCRs carries

110 the PACT domain and is targeted only to the centrosome or the nuclear side of the SPB,

111 the CM1-only type of -TuCRs has either a MASC (Mto1 and Spc72 C-terminus) (Samejima 112 et al., 2010) or a CM2 motif and is recruited to, centrosomes, the cytoplasmic side of the

113 SPB or acentrosomal MTOCs.

114

115 Results

116

117 Phosphorylation of N-Spc110 at Mps1 and Cdk1 sites is required for efficient

118 interaction with -TuSC

119 To test whether Spc1101-220 phosphorylation promoted -TuSC ring formation, we purified

120 GST-Spc1101-220 (named Spc1101-220) from both E. coli and the baculovirus expression

121 system. Spc1101-220 encompasses Cdk1 and Mps1 phosphorylation sites and the

122 conserved CM1 motif (Figure 1A). Because of the post-translational modification

123 machinery, Spc1101-220 purified from insect cells harboured phosphorylations on S60/T68

124 and S36/S91 (Figure 1—figure supplement 1A-D), that correspond to established Mps1 and

125 Cdk1 sites, respectively (Figure 1A) (Friedman et al., 2001; Huisman et al., 2007). In

126 contrast, Spc1101-220 was not phosphorylated when purified from E. coli.

127 We incubated purified -TuSC with either the phosphorylated Spc1101-220 from insect cells

128 (Spc1101-220-P) or the non-phosphorylated Spc1101-220 from E. coli and then analyzed the

129 protein complexes by gel filtration. Spc110 dimerizes via the coiled-coli region in the centre

130 of the protein (Kilmartin et al., 1993; Muller et al., 2005). To substitute for the lack of this

131 region, we performed the assays with GST-Spc1101-220 that dimerizes via GST-GST

132 interactions. We used TB150 buffer in our assay instead of the BRB80 buffer used by

133 Kollman et al. (2010) (Kollman et al., 2010). BRB80 induces oligomerization of -TuSC

134 without the need for addition of Spc110 (Figure 1—figure supplement 2A, B). In contrast, in 135 TB150 buffer the majority of -TuSC was monomeric (Figure 1—figure supplement 2C).

136 TB150 therefore allowed us to monitor the impact of Spc1101-220 on -TuSC oligomerization

137 by gel filtration. Addition of Spc1101-220-P shifted -TuSC into fractions that eluted earlier

138 than when -TuSC was run on the columns on its own (Figure 1B, shift from faction 10 to

139 fraction 8). The peak at the void-volume (fraction 8) likely represented -TuSC oligomers. In

140 contrast, non-phosphorylated Spc1101-220 expressed in E. coli did not change the -TuSC

141 elution profile. The -TuSC pool eluted as monomeric -TuSC (Figure 1B, fractions 10-11).

142 This result suggests that Mps1 and Cdk1 kinases regulate the interaction between Spc110

143 and -TuSC through phosphorylation of Spc110 N-terminal domain.

144

145 Phosphorylation of N-Spc110 regulates the -TuSC oligomerization promoting

146 activity

147 To further confirm that phosphorylation of Spc1101-220 alters the interaction with -TuSC,

148 phosphomimetic and non-phosphorylatable Spc1101-220 mutant proteins were purified from

149 E. coli (Figure 2—figure supplement 1A). spc1102A (Cdk1 sites: S36A, S91A) and spc1103A

150 (Mps1 sites: S60A, T64A, T68A) have been studied in vivo before and shown to behave as

151 non-phosphorylatable mutants (Friedman et al., 2001; Huisman et al., 2007; Liang et al.,

152 2013). In addition, we analyzed the phosphomimetic Spc1101-220-2D (Cdk1 sites: S36D and

153 S91D), Spc1101-220-3D (Mps1 sites: S60D, T64D, T68D) and Spc1101-220-5D (Cdk1 and Mps1

154 sites: S36D, S60D, T64D, T68D and S91D). Migration of all purified Spc1101-220 mutant

155 proteins was comparable upon gel filtration, indicating that the mutations did not alter the

156 overall structure of the protein (Figure 2—figure supplement 1B, C).

157 We used gel filtration chromatography to address whether Spc1101-220 mutant proteins 158 induce higher-order -TuSC oligomers. The phosphomimetic Spc1101-220-2D (Cdk1 sites),

159 Spc1101-220-3D (Mps1 sites) and Spc1101-220-5D (Cdk1 and Mps1 sites) induced an apparent

160 shift of -TuSC towards the void-volume compared to the -TuSC only run or -TuSC plus

161 non-phosphorylated Spc1101-220-WT from E. coli (Figure 2A-C and Figure 2—figure

162 supplement 2A, B). The behavior of the phosphomimetic Spc110 mutant proteins reflected

163 that of in vivo phosphorylated Spc1101-220-P purified from insect cells (Figure 1B), indicating

164 that the mutations to acidic residues did mimic phosphorylation. In contrast, the -TuSC

165 shift was not seen upon inclusion of non-phosphorylatable Spc1101-220-5A (Cdk1 and Mps1

166 sites) with the -TuSC (Figure 2—figure supplement 2A, B). Thus, Spc1101-220-5A behaved

167 as the non-phosphorylated Spc1101-220 purified from E. coli, suggesting that the non-

168 phosphorylatable mutations did not impair the protein. This conclusion was further

169 supported by the observation that Spc1101-220-2D or Spc1101-220-3D still possessed

170 oligomerization-promoting activity when combined with non-phosphorylatable mutations

171 (Spc1101-220-2D3A and Spc1101-220-2A3D, Figure 2—figure supplement 3). Taken together, the

172 gel filtration experiment suggests that phosphorylated Spc110 induces oligomerization of -

173 TuSC.

174 Because the upper limit of the separation range of the Superose 6 gel filtration column is

175 ~5000 kDa, the -TuSC oligomers in the void-volume contain at least 14 copies of -TuSC

176 (14-mers). This indicates the presence of ring-like complexes or even higher-order

177 oligomeric structures in the void-volume. To confirm the existence of structurally organized

178 oligomers, fractions corresponding to the void-volume peak (Figure 2—figure supplement

179 2A, fraction 7) were subjected to EM analysis (Figure 2D, E and Figure 2—figure

180 supplement 4A). Consistent with the low A280 absorbance, hardly any oligomeric -TuSC

181 species were found in the void fraction of -TuSC only or -TuSC incubated with the non- 182 phosphorylated Spc1101-220-WT (Figure 2E and Figure 2—figure supplement 4B). In

183 contrast, oligomeric -TuSC ring-like assemblies were observed when Spc1101-220-2D,

184 Spc1101-220-3D and Spc1101-220-5D (Figure 2D, E and Figure 2—figure supplement 4A, B)

185 had been added to the -TuSC preparation. Interestingly, the percentage of -TuSC spirals

186 increased steadily with increasing numbers of phospho-mimetic mutations from Spc1101-220-

187 2D, Spc1101-220-3D to Spc1101-220-5D (Figure 2E and Figure 2—figure supplement 4B), while

188 the number of rings per field was relatively constant (Figure 2—figure supplement 4C). The

189 outer diameter of the ring-like assemblies was on average around 40 nm (Figure 2—figure

190 supplement 4D). Thus, combined phosphorylation of Spc110 by Cdk1 and Mps1 kinases

191 enhances the oligomerization activity of Spc110 better than phosphorylation by either

192 kinase alone.

193

194 The SPM motif is important for Spc110 oligomerization promoting activity

195 An additional phosphorylation of N-Spc110 at T18 has been identified in phosphoproteomic

196 studies (Albuquerque et al., 2008; Keck et al., 2011; Lin et al., 2011). However, its precise

197 function remained unclear (Liang et al., 2013). To elucidate the role of Spc110T18

198 phosphorylation, Spc1101-220-T18D, Spc1101-220-2D-T18D (Cdk1 plus T18), Spc1101-220-3D-T18D

199 (Mps1 plus T18), Spc1101-220-5D-T18D (T18D in addition to Mps1 and Cdk1 sites) and

200 Spc1101-220-5D-T18A proteins were purified (Figure 2—figure supplement 1A) and tested for

201 their -TuSC oligomerization promoting activity. Spc1101-220 species with T18D or T18A

202 failed to induce -TuSC oligomerization even when combined with the activating 5D

203 mutations (Figure 2F and Figure 2—figure supplement 2C). Similar data were obtained with

204 Spc1101-220-5D-T18E and Spc1101-220-5D-T18V (Figure 2-supplement 5A). Moreover, spc110T18D,

205 spc110T18E, spc110T18A and spc110T18V showed identical growth defects at 37ºC (Figure 2- 206 supplement figure 5B). Since the effects of T18D/E and T18A/V mutations were

207 indistinguishable, we cannot attribute these mutations as phosphomimetic or non-

208 phosphorylatable. However, our results strongly suggest that these mutations affect the

209 structure of an important yet unappreciated element around T18 that is important for -

210 TuSC oligomerization. Consistently, hardly any oligomeric structures were observed when

211 -TuSC was incubated with Spc1101-220-5D-T18D (Figure 2G).

212 With the sequence alignments of Spc110 orthologs from yeast to human, we observed a

213 conserved motif upstream of CM1, which we designated as Spc110/Pcp1 motif (SPM)

214 (Figure 2H). Spc110T18 sits within this motif. To evaluate the relative importance of CM1

215 and SPM, we constructed the Spc1101-220-CM1-QA mutant protein with mutations (KE to QA)

216 in two highly conserved residues of the CM1 (Figure 2—figure supplement 6) and the

217 Spc1101-220-ΔSPM lacking the SPM motif (deletion of amino acids 1-20). Spc1101-220 proteins

218 were incubated with -TuSC in TB150 buffer to examine their oligomerization capacity.

219 Spc1101-220-5D-CM1-QA and Spc1101-220-5D-ΔSPM mutant proteins failed to activate -TuSC

220 oligomerization even when the five activating Cdk1 and Mps1 mutations were present

221 (Figure 2I). Thus, Spc1101-220-5D-ΔSPM behaved as Spc1101-220-5D-T18D (compare Figure 2I with

222 Figure 2—figure supplement 2C), further emphasizing that T18D inactivates the SPM.

223 Taken together, SPM and CM1 of Spc110 are both required for -TuSC oligomerization.

224

225 The SPM and CM1 motifs and the Cdk1 and Mps1 phosphorylations regulate the

226 affinity of Spc110 to -TuSC

227 Phosphorylation of Spc110 may regulate the -TuSC oligomerization by changing its

228 binding affinity. Therefore we performed GST-pulldown assay to measure the binding of the 229 Spc1101-220 variants to -TuSC. A constant amount of -TuSC was incubated with

230 increasing concentrations of Spc1101-220 variants (0 to 300 nM). Consistent with the -TuSC

231 oligomerization assay (Figure 2), Spc1101-220-5D showed stronger (p=0.0017 for

232 Spc97/Spc98 and p=0.0006 for Tub4) -TuSC binding than Spc1101-220-WT (Figure 3A, B).

233 Interestingly, while Spc1101-220-2D and Spc1101-220-3D induced comparable levels of -TuSC

234 oligomerzation (Figure 2B), Spc1101-220-2D showed less -TuSC binding than Spc1101-220-3D.

235 A likely explanation is the 30-fold reduction in concentration that was used in the GST-

236 pulldown assay compared to -TuSC oligomerization assays. These results suggested that

237 phosphorylation of Spc110 on S36, S60, T64, T68 and S91 promote -TuSC binding to

238 Spc110.

239 We also analyzed binding of SPM (T18D and ∆SPM) and CM1 (CM1-QA) deficient

240 Spc1101-220 mutant proteins to -TuSC. Both Spc1101-220-5D-T18D and Spc1101-220-5D-CM1-QA

241 exhibited reduced -TuSC binding relative to Spc1101-220-5D (p<0.01 for Spc97/Spc98 and

242 p<0.05 for Tub4) (Figure 3C, D). The -TuSC binding capacity of Spc1101-220-5D-T18D-CM1-QA

243 was further reduced (p<0.01, Figure 3C, D), suggesting both CM1 and SPM are important

244 for -TuSC binding. Interestingly, while Spc1101-220-5D-T18D, Spc1101-220-5D-CM1-QA and

245 Spc1101-220-ΔSPM all failed to oligomerize -TuSC (Figure 2F and 2I), they showed different

246 levels of -TuSC binding capacity (Figure 3D). The decoupling of -TuSC binding capacity

247 from -TuSC oligomerization suggests that both processes are not inevitably linked. Taken

248 together, we conclude that both CM1 and SPM motifs contribute to -TuSC binding and are

249 required for promoting -TuSC oligomerization.

250

251 SPM and the phosphorylation of N-Spc110 control MT nucleation activity in vitro 252 In vitro MT nucleation assays were performed to test the effect of Spc110 variants on -

253 TuSC-mediated MT nucleation. Since phosphorylation of Spc110 and the presence of a

254 functional SPM regulate -TuSC oligomerization into template rings, we expected to see

255 differences in MT nucleation depending on these parameters (Figures 2 and 3). Compared

256 to the buffer control and -TuSC alone, a ~3-fold increase (p<0.05) in MT nucleation activity

257 was observed for Spc1101-220-2D, Spc1101-220-3D or Spc1101-220-5D upon incubation with -

258 TuSC (Figure 4A, B). In contrast, Spc1101-220-WT showed levels of MT nucleation that were

259 comparable to the buffer control and -TuSC alone. Consistent with the proposed role of

260 SPM, the MT nucleation level was reduced to buffer control levels when Spc1101-220-5D-T18D

261 was incubated with -TuSC (p=0.99). These results are in accordance with EM particle

262 quantification assays, as Spc1101-220-2D, Spc1101-220-3D or Spc1101-220-5D induced similar

263 and significantly more -TuSC ring assemblies than Spc1101-220-WT and Spc1101-220-5D-T18D

264 (p<0.05, Figure 2E and Figure 2—figure supplement 4C). In summary, these data support

265 the model that Cdk1 and Mps1 phosphorylations of Spc110 regulate -TuSC-mediated MT

266 nucleation by controlling template assembly.

267

268 Phosphorylation of Spc110 is cell cycle dependent

269 To analyze cell cycle dependent phosphorylation of Spc110, we generated one phospho-

270 specific antibody against the two Cdk1 sites (Spc110S36 and Spc110S91) (Figure 5A) and

271 another against the Mps1 sites on Spc110 (Spc110pS60-pT64-pT68) (Figure 5B). Both anti-

272 Spc110pS36-pS91 and anti-Spc110pS60-pT64-pT68 antibodies gave specific signals in vitro and in

273 vivo (Figure 5A-C).

274 To understand by which Cdk1 kinase complex Spc110S36 and Spc110S91 were 275 phosphorylated, we co-immunoprecipitated Cdk1as1 either in complex with Clb5-TAP (S

276 phase cyclin) or Clb2-TAP (mitotic cyclin) from yeast lysates and performed in vitro kinase

277 assays using Spc1101-220-WT that had been purified from E. coli. Spc110S36-S91 was

278 phosphorylated equally by Cdk1-Clb2 and Cdk1-Clb5 as detected by Spc110pS36-pS91

279 antibodies (Figure 5A). It is important to note that both Cdk1 kinase complexes had equal

280 kinase activities toward histone H3 (Figure 5A, bottom).

281 To confirm the cell-cycle dependent phosphorylation of Spc110 in vivo, we immuno-

282 precipitated (IP) Spc110-GFP with GFP-binder from cells arrested either in G1 phase with

283 -factor, S phase with hydroxyurea (HU) or metaphase with nocodazole, and then blotted

284 with phospho-specific antibodies. At G1 phase, no signal was observed with any of the two

285 phospho-specific antibodies (Figure 5D). The phosphorylation of Mps1 was detected at S

286 phase and maintained at metaphase. The Spc110pS36-pS91 signal also appeared at S phase,

287 and continuously increased at metaphase.

288 T18 within the SPM motif has been found phosphorylated in vivo and matches the minimal

289 Cdk1 consensus sequence S/T-P (Keck et al., 2011). To elucidate the function of T18

290 phosphorylation, we raised phosphospecific antibodies against pT18. In vitro

291 phosphorylation assay showed that T18 was only phosphorylated by Cdk1-Clb2 but not by

292 Cdk1-Clb5 (Figure 5—figure supplement 1A). Due to the low sensitivity of anti-Spc110pT18,

293 we used mass spectrometry for the analysis of T18 phosphorylation in vivo. Total yeast cell

294 lysates were fractionated into supernatant and pellet that contained SPBs. The pellet

295 fraction was solublized with high salt buffer and then used for Spc110 enrichment. pT18

296 was predominately detected in the SPB-containing pellet fraction of metaphase cells but not

297 in the corresponding fraction of G1 or S phase arrested cells (Figure 5—figure supplement

298 1B, C). These data suggest that T18 of Spc110 becomes phosphorylated at SPBs in M 299 phase but not S phase, which is consistent with published data (Keck et al., 2011).

300 Finally, we addressed the effects of T18 phosphorylation on -TuSC oligomerization.

301 Because the T18A/V/D/E mutations inactivated the SPM function, we turned to an in vitro

302 phosphorylation approach. Spc1101-220-5D was incubated with Cdk1as1-Clb2 in the presence

303 of absence of the Cdk1as1 inhibitor 1NM-PP1. Only in the absence of 1NM-PP1, Spc1101-

304 220-5D was phosphorylated on T18 as shown by the phospho-specific anti-T18 antibodies.

305 We incubated Spc1101-220-5D and Spc1101-220-5D-pT18 with -TuSC and analyzed -TuSC by

306 gel filtration. The activity of Spc1101-220-5D-pT18 to promote -TuSC oligomerization was

307 significantly reduced in comparison with Spc1101-220-5D (p<0.05) (Figure 5- figure

308 supplement 1D, E). This suggests that in vitro phosphorylation of T18 by Cdk1-Clb2 inhibits

309 -TuSC oligomerization.

310 Taken together, our results are in agreement with previous reports about the timing of

311 Spc110 phospho-dependent mobility shift in SDS-PAGE (Friedman et al., 1996; Stirling and

312 Stark, 1996) and suggest that the Cdk1 and Mps1 sites become phosphorylated in S and

313 metaphase. In addition, phosphorylation of T18 within SPM occurs at metaphase and may

314 negativly regulate SPM function.

315

316 Cells with spc110 phospho-, SPM- and CM1-mutant alleles have SPBs with impaired

317 MT nucleation activity

318 To reveal the function of phosphorylation of Spc110 in vivo, we replaced SPC110 with

319 phosphomimetic and non-phosphorylatable mutant alleles. There was no significant

320 difference in the expression levels of the Spc110 variants (Figure 6—figure supplement

321 1A). spc1102A/2D, spc1103A/3D, spc1105A/5D, the SPM deficient spc110T18D, spc110∆SPM and 322 the CM1 deficient spc110CM1-QA cells grew at 23ºC comparable to wild-type SPC110 (WT)

323 cells. However, spc110T18D and spc110∆SPM cells displayed a conditional lethal growth

324 defect at 37ºC and the SPM/CM1 defective spc110T18D-CM1-QA double mutant was non-viable

325 at 23ºC and 37ºC (Figure 6A, left panel). Combining T18D with 2D (Cdk1 sites), 3D (Mps1

326 sites) and 5D mutations slightly reduced growth at 37ºC compared to the 2D, 3D and 5D

327 mutations (Figure 6A). This shows that, consistent with our in vitro -TuSC oligomerization

328 results (Figure 2—figure supplement 2C), the SPM mutation T18D is dominant over the

329 activating Cdk1 and Mps1 phosphomimetic mutations.

330 The defect of some mutant alleles involved in spindle assembly becomes apparent only

331 after SAC function has been impaired (Daum et al., 2000; Wang and Burke, 1995). In the

332 absence of the SAC gene MAD2 (Li and Murray, 1991), slow growth was observed in

333 spc1102A and spc1102D cells at 37ºC. This growth defect was even more pronounced in

334 spc1105A mad2Δ cells at 37ºC (Figure 6A, right panel). All spc110 mutant alleles with SPM

335 or CM1 inactivating mutations showed reduced growth in the mad2Δ background compared

336 to MAD2 SAC proficient cells (Figure 6A, bottom panel). Together, these growth tests

337 suggest that the absence of phospho-regulation of Spc110 or the lack of SPM or CM1

338 results in defects that are compensated by a SAC induced delay in cell cycle progression.

339 Based on our model built from in vitro experiments, phosphorylation of N-Spc110 by Cdk1

340 and Mps1 in S phase promotes the binding and oligomerization of -TuSC to form the

341 template required for MT nucleation. In addition, the SPM and the CM1 motifs of Spc110

342 are important for -TuSC oligomerization and MT nucleation (Figure 2). Thus, mutations in

343 SPC110 that affect any one of these sites/motifs should display a reduction in their SPB-

344 associated tubulin signal. To test this hypothesis, we expressed GFP-TUB1 in cells and

345 measured the intensity of the tubulin signal at the SPB at different cell cycle phases. Cells 346 were synchronized with -factor in G1 and then released into pre-warmed media at 37ºC,

347 the restrictive temperature of some of the mutant cells. In SPM defective spc110T18A and

348 spc110T18D cells, the reduction of GFP-tubulin signal at SPBs was already apparent in G1

349 (Figure 6B and Figure 6—figure supplement 1B, p<0.001). In S phase, the stage where

350 newly synthesized SPBs mature and MT nucleation is initiated, all spc110 mutants except

351 for spc1105D cells exhibited a marked reduction in MT nucleation at SPBs when compared

352 with WT cells (Figure 6B; for spc1105A, p<0.05; for spc110T18D, spc110T18A, p<0.001 and

353 Figure 6—figure supplement 1B). The reduction in tubulin signal was especially pronounced

354 in spc110T18D and spc110T18A cells, indicating that SPM mutations blocked MT formation in

355 vivo. spc110∆SPM cells also showed MT defects at 37ºC (Figure 6-supplement 1F).

356 The situation in metaphase was similar to that of S phase although the overall relative

357 differences between WT and spc110 mutants were less pronounced than in S phase

358 (Figure 6B, p<0.001). Consistently, the overall spindle MT signal was reduced in spc1105A,

359 spc110T18A, spc110T18D and spc1105D-T18D cells (Figure 6B, C). Taken together, the in vivo

360 analysis of the tubulin signal at SPBs in spc110 cells is consistent with the in vitro data on

361 -TuSC oligomerization and MT nucleation (Figures 2 and 4).

362 Reduced MT signal intensity at SPBs would be indicative of impaired -TuSC activity. In the

363 simplest interpretation, -TuSC recruitment to SPBs is impaired. We found that the SPB

364 signal from the -TuSC marker Spc97-GFP was reduced in spc1105A, spc110T18D,

365 spc1105D-T18D and spc110∆SPM cells in G1 and S phase compared to SPC110WT and

366 spc1105D cells (Figure 6—figure supplement 1C, G). However, this reduction was not

367 observed in mitotic cells. Analysis of Spc97-GFP signal intensity at SPBs in SPC110 wild

368 type, spc110T18D and spc110∆SPM cells revealed that the SPM mutation allowed Spc97 SPB

369 binding but with a delay in cell cycle timing (Figure 6—figure supplement 1D, G). After SPB 370 separation in late S phase wild type, spc110T18D and spc110∆SPM cells eventually carried

371 similar amounts of Spc97-GFP at SPBs.

372 Analysis of the GFP signal at SPBs in SPC110-GFP, spc1105A-GFP spc110T18D–GFP and

373 spc1105D-T18D-GFP cells did not reveal any differences between cell types (Figure 6—figure

374 supplement 1E). Thus, the reduction of the Spc97-GFP signal in G1 and S phase most

375 likely reflects a reduction of -TuSC binding to the SPC110 mutant molecules at SPBs. The

376 observation that all spc110 cells had similar -TuSC signals at mitotic SPBs while the MT

377 signal was only reduced in spc110T18D and spc1105A cells (Figure 6—figure supplement 1C

378 versus Figure 6B), suggests that the SPB-associated -TuSC is not fully active in MT

379 organization in these cells. The functions of Spc110 probably extend beyond -TuSC

380 binding, most likely to include -TuSC oligomerization (see Discussion). However,

381 compared with spc110T18D or SPC110 cells, Spc110-GFP at the SPB was reduced in

382 spc110∆SPM cells (Figure 6—figure supplement 1H), suggesting the SPM deletion affects

383 incorporation of Spc110 into the SPB. This may contribute to the spc110∆SPM phenotype

384 and may explain why the growth defect of spc110∆SPM cells was more pronounced than that

385 of spc110T18D cells (Figure 6A).

386 To further examine the MT nucleation defect in spc110 phospho-mutant cells, we

387 performed an in vivo MT nucleation assay. Cells were first arrested in S phase with HU,

388 then MTs were depolymerized with nocodazole, followed by nocodazole wash out to trigger

389 MT nucleation by the SPB bound -TuSC (Erlemann et al., 2012). To prevent cell cycle

390 progression into anaphase we retained the cells in the HU block (Figure 6D). The re-growth

391 of MTs was measured as a change in GFP-Tub1 intensity at SPBs. In comparison to WT

392 and spc1105D cells, MT nucleation activity was reduced in the spc1105A, spc110T18A,

393 spc110T18D spc1105D-T18D cells (Figure 6E). Again, these data are consistent with the in vitro 394 data (Figures 1 to 4) and suggest that Spc110’s ability to stimulate MT nucleation is

395 regulated through stimulatory phosphorylation and requires the SPM motif.

396

397 Spc110 interaction and -TuSC oligomerization are facilitated by the N-terminal

398 region of Spc98

399 The N-terminal region of Spc98/GCP3 orthologs is conserved. Spc98/GCP3 family

400 members contain five predicted helices followed by an unstructured linker region (Figure 7A

401 and Figure 7—figure supplement 1A). The conserved GRIP1 and 2 motifs that are common

402 to all GCP proteins follow this N-terminal region (Wiese and Zheng, 2006). Analysis of -

403 TuSC by electron microscopy has shown that the N-terminus of Spc98 is in close proximity

404 to the N-Spc110 binding site (Kollman et al., 2010). This topological arrangement raises the

405 possibility that N-Spc98 is involved in binding to N-Spc110 and oligomerization of -TuSC.

406 To test this idea, we constructed truncations of N-Spc98 with which to assess the function

407 and biochemical behaviour of -TuSCSpc98∆N. All N-terminal truncations of SPC98 supported

408 cell growth at 23ºC (Figure 7B and Figure 7—figure supplement 1B). However, Δ1-156, Δ1-

409 177 and ΔLinker deletions of SPC98 showed reduced or no growth in mad2Δ cells at 37ºC.

410 We analyzed the phenotype of spc98ΔN mutants at their restrictive temperature, 37ºC.

411 Because -TuSCSpc98∆1-177 formed aggregates already in TB150 buffer and was therefore

412 unsuitable for oligomerization assay (Figure 7-supplement figure 2A, B), we focus our

413 analysis on spc98∆1-156. Like the wild-type -TuSC, the purified recombinant-TuSCSpc98∆1-

414 156 was monomeric in TB150 (Figure 7-supplement figure 2C, D). spc98∆1-156 cells

415 incubated at 37ºC had weaker GFP-Tub1 signal at their SPBs than wild type SPC98 cells

416 (WT, Figure 7C). This is consistent with the observed mild MT nucleation defect of spc98Δ1- 417 156 cells at 37ºC (Figure 7—figure supplement 1C, D). Moreover, time-lapse analysis of -

418 factor synchronized SPC98 SPC42-Cherry and spc98∆1-156 SPC42-mCherry cells with

419 SPC97-GFP or SPC110-GFP revealed a delay in Spc97-GFP recruitment to SPBs at the

420 time of SPB duplication (Figure 7D). Interestingly, slightly more Spc110-GFP was found at

421 SPBs in spc98∆1-156 cells, which may help to compensate the spc98∆1-156 phenotype. Thus,

422 the N-terminal region of Spc98 is important for optimal MT organization and timely -TuSC

423 recruitment to SPBs.

424 We next asked whether -TuSCSpc98∆1-156 can be oligomerized by incubation with Spc1101-

425 220-5D. Interestingly, Spc1101-220-5D failed to shift -TuSCSpc98∆1-156 to the void volume of the

426 Superose 6 column, as was the case for -TuSC (Figure 7E). In agreement with the

427 oligomerization experiment, -TuSCSpc98∆1-156 showed reduced binding to Spc1101-220-5D in

428 the pull down assay (Figure 7F, G). Thus, the N-terminus of Spc98 is involved in binding to

429 Spc110. In spite of not changing our conclusion, we noticed that Spc98∆1-156-6His was

430 detected less strongly than Spc98-6His by the anti-His antibodies, although in Coomassie

431 Blue stained gels both proteins were present in 1:1 ratio (Figure 7E upper panel and F).

432 Thus, it is likely that Spc98Δ1-156-6His transferred less efficiently to the blotting membrane

433 than Spc97-6His, leading to an underestimation of the Spc98Δ1-156-6His signal relative to

434 full-length Spc98-6His.

435 Finally, we tested -TuSCSpc98∆1-156 oligomerization in the BRB80 buffer that supports

436 oligomerization without the aid of Spc110. -TuSC and -TuSCSpc98∆1-156 shifted to the void

437 volume of the Superose 6 column with equal efficiency (Figure 1—figure supplement 2B

438 and Figure 7 — figure supplement 2C). Analysis of the void fraction by EM identified

439 oligomeric -TuSCSpc98∆1-156 rings (Figure 7—figure supplement 2E). Therefore, the N- 440 terminus of Spc98 is important for Spc110-mediated oligomerization but not for Spc110-

441 independent -TuSC oligomerization.

442

443 Bioinformatic analysis of the C-terminal SPB/centrosomal binding domain of SPM-

444 CM1 and CM1-only -TuCRs

445 Our biochemical and cell biology analysis have identified the novel motif SPM and CM1

446 (Sawin et al., 2004) as important motifs in Spc110 for -TuSC binding and oligomerization.

447 Using bioinformatics approaches, we investigated the presence of SPM and CM1 motifs in

448 -TuCRs from various species (Figure 8A, B). A yet unappreciated CM1 motif was identified

449 in another budding yeast -TuCR, Spc72 (Knop and Schiebel, 1998) (Figure 8A and Figure

450 8—figure supplement 1C).

451 Based on the presence of SPM and CM1 motifs, -TuCRs can be divided into the SPM-

452 CM1 and the CM1-only types. Spc110 and fission yeast Pcp1 fall into the SPM-CM1 class.

453 Our analysis also predicts SPM and CM1 motifs, albeit with moderate sequence deviation,

454 in mammalian PCNT (Figure 8B). Another group of -TuCRs have only a CM1 motif but no

455 SPM. Human CDK2RAP2 and myomegalin, Drosophila Cnn, budding yeast Spc72 and

456 fission yeast Mto1 fall into this class (Figure 8B).

457 A further level of category resolution was achieved by the classification based on the C-

458 terminal MTOC targeting domains (Figure 8—figure supplement 1A, B). It was shown that

459 yeast Spc110 (Stirling and Stark, 2000; Sundberg et al., 1996), Pcp1 (Fong et al., 2010),

460 Drosophila D-PLP (Cp309) (Kawaguchi and Zheng, 2004) and mammalian PCNT

461 (Gillingham and Munro, 2000; Takahashi et al., 2002) are targeted to nuclear side of SPBs

462 or centrosomal MTOCs via the conserved PACT domain (Figure 9A). In case of SPBs that 463 are embedded in the nuclear envelope and therefore have a cytoplasmic and a nuclear

464 side, SPM-CM1 -TuCRs only function in organizing the nuclear MTs that separate the

465 sister chromatids in mitosis (Figure 9B). In contrast, CM1-only -TuCRs of fungi, such as

466 Spc72 in budding yeast (Knop and Schiebel, 1998), Mto1 in fission yeast (Samejima et al.,

467 2010) and ApsB in Aspergillus nidulans (Zekert et al., 2010), have a MASC motif and

468 function at the cytoplasmic side of the SPB or in the cytoplasm. In case of metazoa, CM1-

469 only type of -TuCRs such as CDK5RAP2 (Wang et al., 2010) possess a CM2 domain

470 (Figure 9B and Figure 8—figure supplement 1A), targeting -TuCRs to both centrosomal

471 and acentrosomal MTOCs.

472 In summary, -TuCR family proteins can be classified regarding the -TuSC interaction

473 motifs and the MTOC targeting domain into three groups.

474

475 Discussion

476 Spc110 is a -TuSC-mediated nucleation regulator

477 In this study we show that the -TuCR Spc110 actively participates in MT formation through

478 the SPM and CM1 elements and Cdk1 and Mps1 phosphorylation sites. Binding of -TuSC

479 to Spc110 is a prerequisite for -TuSC oligomerization into the ring-like template that

480 initiates MT assembly. Phosphorylation of Spc110 by Cdk1 and Mps1 regulates -TuSC

481 binding and thereby MT nucleation. Thus, Spc110 is a MT nucleation regulator. We thereby

482 provide a molecular understanding of the previously observed spindle length phenotypes

483 (the two Cdk1-Clb5 sites) and genetic interaction with SPC97 (three Mps1 sites combined

484 with S36) of SPC110 phosphorylation site mutants (Friedman et al., 2001; Huisman et al., 485 2007). Our data also explain why an N-terminal fragment of Spc110 influences

486 oligomerization of -TuSC when co-expressed in insect cells (Kollman et al., 2010).

487 SPBs duplicate in G1/S phase of the cell cycle in a conservative manner (Adams and

488 Kilmartin, 2000; Byers and Goetsch, 1975; Pereira et al., 2001) (Figure 10A). In late G1, a

489 SPB precursor, named the duplication plaque, assembles on the cytoplasmic side of the

490 nuclear envelope at a specialized substructure of the mother SPB, named bridge (Byers

491 and Goetsch, 1975; Jaspersen and Winey, 2004). In G1/S the duplication plaque then

492 becomes inserted into the nuclear envelope. This allows binding of the Spc110-calmodulin-

493 Spc29 complex from the nuclear side to the central Spc29-Spc42 layer of the embedded

494 duplication plaque (Bullitt et al., 1997; Elliott et al., 1999). In this cell cycle phase Mps1 and

495 Cdk1-Clb5 phosphorylate the N-terminus of Spc110 at five sites to promote -TuSC binding

496 to Spc110 (Figure 10B). Consistently, Cdk1-Clb5 associates with SPBs throughout this time

497 window (Huisman et al., 2007). These phosphorylations increase -TuSC affinity for

498 Spc110-5P and induce -TuSC oligomerization into ring-like complexes that promote MT

499 nucleation (Figure 10B). Phospho-regulation of Spc110 is not absolutely essential for

500 viability as indicated by the growth of spc1105A cells. However, spc1105A cells have growth

501 defects in the absence of the SAC gene MAD2 and fail to organize the full set of MTs in S

502 phase and mitosis (Figure 6). We therefore suggest that Mps1 and Cdk1-Clb5 coordinate

503 the timing of MT nucleation through Spc110 phosphorylation with SPB duplication. A SAC

504 induced cell cycle delay can temper any defects in this phospho-regulation as indicated by

505 the genetic interaction between spc1102A and spc1105A with mad2Δ (Figure 6A).

506 Spc110T18 is an additional Cdk1 in vivo site (Albuquerque et al., 2008; Keck et al., 2011; Lin

507 et al., 2011) (Figure 5—figure supplement 1). Analysis of T18A, T18V, T18E and T18D

508 mutations resulted in similar in vitro and in vivo phenotypes. We have to assume that all 509 these mutations inactivate the important SPM motif in which T18 resides. The fact that T18

510 of Spc110 is phosphorylated only by Cdk1-Clb2 and not by Cdk1-Clb5 and the in vivo

511 phosphorylation of SPB-associated Spc110T18 in mitosis but not in S phase (Keck et al.,

512 2011) (Figure 5—figure supplement 1A-C) indicates that this Cdk1 site is different

513 compared to the two S phase Cdk1 sites S36 and S91 in Spc110. At least in vitro, Cdk1-

514 Clb2 phopshorylation of Spc110T18 inactivates in a dominant manner the -TuSC

515 oligomerization activity (Figure 5—figure supplement 1D, E). Thus, it is well possible that

516 Spc110T18 phosphorylation by Cdk1-Clb2 limits the MT nucleation activity of the SPB

517 associated -TuSC. Most likely only the surplus of -TuSC that is not engaged in MT

518 organization is phosphorylated and affected by Spc110pT18. This model fits with the

519 observation that the yeast SPB of haploid cells organizes a constant number of 20 nuclear

520 MTs during mitosis (Winey et al., 1995).

521 Co-overexpression of SPC97, SPC98 and TUB4 does not induce MT nucleation despite of

522 -TuSC assembly (Pereira et al., 1998). Our and other studies have shown that the missing

523 factor promoting -TuSC ring assembly is Spc110 (Kollman et al., 2010; Vinh et al., 2002).

524 In vivo data indicate that the binding of -TuSC to Spc110, although a prerequisite, is

525 insufficient for MT formation. For example, although the -TuSC component Spc97 bound

526 to mitotic SPBs of spc110T18D or spc1105D-T18D cells with similar efficiency to wild type cells,

527 the -TuSC in the two mutants was insufficient in full MT organization (Figure 6B and Figure

528 6—figure supplement 1C). Consistent with this model, the addition of Spc1101-220-5D-T18D to

529 recombinant -TuSC induced the formation of -TuSC dimers at ~600 kDa (Figure—figure

530 supplement 2A; note that the protein concentration in this experiment was two-times higher

531 than in Figure – figure supplement 2B) without the formation of -TuSC oligomers. We 532 therefore suggest that -TuSC binding to Spc110 and TuSC oligomerization are

533 mechanistically distinct steps (Figure 10C).

534

535 The N-terminus of Spc98 mediates interaction with N-Spc110 and oligomerization

536 The N-terminal region of Spc98 exhibits homology to other GCP3 family members including

537 human GCP3 (hGCP3, Figure 7A and Figure 7—figure supplement 1A). All homologues

538 contain five predicted helical regions that are followed by an unstructured linker region

539 before the two GRIP domains that are common to all GCP proteins (Guillet et al., 2011).

540 Analysis of the structure of yeast -TuSC by electron microscopy localised N-Spc98

541 towards of the base of the Y shaped structure, away from the C-Spc98 that interacts with -

542 tubulin. N-Spc98 is close to the N-terminus of Spc97 and N-Spc110 (Choy et al., 2009;

543 Kollman et al., 2010). This is consistent with yeast two-hybrid studies that identified N-

544 Spc98 as an interactor for Spc110 (Knop and Schiebel, 1997; Nguyen et al., 1998). We

545 have analyzed the importance of N-Spc98 by deleting different portions of this domain

546 including the helices, the linker region and both the helices and the linker. Surprisingly, the

547 N-Spc98 fragment was not essential for the viability at 23ºC (Figure 7B and Figure 7—

548 figure supplement 1B). The important function of this region, however, was revealed at

549 elevated temperatures and when SAC function was impaired (Figure 7B).

550 Biochemical analysis of -TuSCSpc98∆1-156 showed that its binding and ability to be induced

551 to oligomerize by Spc1101-220-5D was reduced (Figure 7E, F and G), although in M phase it

552 bound with the same affinity for SPBs as WT -TuSC (Figure 7D). We suggest that the

553 special SPB arrangement of Spc110 as hexameric units (Muller et al., 2005) compensates

554 in part for the reduced in vito binding of -TuSCSpc98∆1-156 to Spc1101-220. This compensation 555 may arise from cooperative interactions of -TuSC with N-Spc110. Whatever the

556 mechanism, regions of -TuSC in addition to the N-terminal 177 amino acids of Spc98

557 (Figure 7—figure supplement 1B) have to interact with Spc110. Either Spc110 also

558 interacts with the Spc98/GCP3 core or it binds to the N-terminus of Spc97/GCP2.

559 -TuSCSpc98∆1-156 was still able to oligomerize under special buffer conditions (Figure 7—

560 figure supplement 2), which is in agreement with the observation that -TuSCSpc98∆1-156 was

561 functional in vivo and promoted MT nucleation, although defects were seen at elevated

562 temperature and in mad2Δ cells. Thus, the regions that are essential for -TuSC

563 oligomerization await identification.

564

565 -tubulin complex receptors fall into three groups: the SPM-CM1-PACT, CM1-MASC

566 and CM1-CM2 types

567 -TuCRs from yeast to Drosophila to human carry a conserved CM1 motif. This CM1 motif

568 was first identified in fission yeast Mto1 (Sawin et al., 2004). Subsequently the function of

569 CM1 in -TuSC binding was confirmed for Mto1, centrosomin (Drosophila) and CDK5RAP2

570 (human) (Choi et al., 2010; Fong et al., 2008; Samejima et al., 2008; Zhang and Megraw,

571 2007). The -TuCRs Pcp1 (fission yeast) and Spc110 (budding yeast) also contain putative

572 CM1 motifs (Dictenberg et al., 1998; Fong et al., 2010; Takahashi et al., 2002; Zimmerman

573 et al., 2004) (Figure 8B).

574 In addition, we identified a yet unappreciated CM1 motif in the yeast -TuCR Spc72 (Knop

575 and Schiebel, 1998) (Figure 8A). Although the putative CM1 motif in Spc72 appears

576 degenerated compared with the full CM1 of other -TuCRs, the conserved residues shared 577 by Spc72-CM1 and full CM1 suggest that Spc72 contains a functional CM1. Consistently,

578 Spc72 carries a C-terminal MASC, as is the case for other fungal CM1-only -TuCRs

579 (Samejima et al., 2010). Moreover, the predicted CM1 of Spc72 resides in the N-terminal

580 region of Spc72 that is essential for -TuSC binding (Knop and Schiebel, 1998; Usui et al.,

581 2003).

582 Human PCNT falls in terms of structure and function into the -TuCR class (Dictenberg et

583 al., 1998; Doxsey et al., 1994; Gillingham and Munro, 2000). Analysis of the PCNT

584 sequence for CM1 in combination with the newly identified SPM, identified a potential CM1

585 in the N-terminus of PCNT (Figure 8B). This is consistent with the -tubulin binding ability of

586 human PCNT, which lies within the region that contains the SPM-CM1 (Takahashi et al.,

587 2002). In addition, super-resolution analysis of human centrosomes has mapped the

588 localization of -tubulin towards the N-terminus of PCNT (Lawo et al., 2012; Mennella et al.,

589 2012; Sonnen et al., 2012). Experimental evidences are awaited to validate the CM1 motif

590 in Spc72 and PCNT.

591 Our study has identified SPM as a second conserved motif amongst a subgroup of -tubulin

592 complex receptors (Figure 2H), namely the yeast Spc110 homologues, fission yeast Pcp1

593 and human PCNT (Figure 10) (Dictenberg et al., 1998; Doxsey et al., 1994; Flory et al.,

594 2002). The distance between SPM and CM1 is around 90-100 amino acids in the Spc110

595 and Pcp1 type of receptors. In the mammalian PCNT subfamily it is however only around

596 70 amino acids. Whether these differences reflect functionally important differences in

597 regulation or function or are indicative of co-adaptation with alterations in the -TuSC

598 binding sites remains to be seen.

599 The SPM is important for Spc110 binding to -TuSC and induced oligomerization of -TuSC

600 (Figures 3 and 4). Based on these results we propose that SPM co-operates with CM1 to 601 form a bipartite binding element for the -TuSC (Figure 10C). Interestingly, however, other

602 -tubulin complex receptors such as Spc72, fission yeast Mto1, Drosophila Cnn and human

603 CDK5RAP2 carry only the CM1 element but no SPM motif (Figure 8). Strikingly, SPM-CM1

604 type of -TuCRs carry a C-terminal PACT domain that targets these proteins to MTOCs

605 (Figure 8B). The CM1-only type of -TuCRs either contain a MASC (fungi) or a CM2 C-

606 terminal MTOC targeting sequence (metazoa) (Figure 9B). We also noticed that in some

607 metazoa and the fungi phylum Basidiomycota the PACT domain-containing -TuCRs (such

608 as AKAP9 in human and D-PLP in Drosophila) have N-terminal regions distictive from SPM

609 and CM1 motifs, probably reflecting different binding partners in -TuRC. Alltogether, we

610 can classify -TuCRs in at least three classes: CM1-MASC, CM1-CM2 and SPM-CM1-

611 PACT.

612 In budding and fission yeast, the CM1-MASC types of -TuCRs (Spc72 and Mto1) have

613 specific functions in cytoplasmic MT organization, while Spc110 and Pcp1 (SPM-CM1-

614 PACT) organize only nuclear MTs. This functional division is accompanied by the physical

615 separation of the receptors by the nuclear envelope, which remains intact during the closed

616 mitosis of both organisms. Human CDK5RAP2, myomegalin (CM1-CM2 types) and

617 pericentrin (SPM-CM1-PACT type) are associated with the centrosome (Doxsey et al.,

618 1994; Fong et al., 2008). However, CDK5RAP2 and myomegalin also have functions in

619 acentrosomal MT nucleation for example MT nucleation by the Golgi (Choi et al., 2010;

620 Roubin et al., 2013). Thus, it may be the acentrosomal/cytoplasmic function that explains

621 the lack of SPM element in Mto1/CDK5RAP2/myomegalin.

622 Budding yeast organizes MTs only with -TuSC and -TuCRs. Other organisms have

623 additional building blocks that contribute to MT assembly, such as Mozart1 and GCP4-6

624 (Dhani et al., 2013; Masuda et al., 2013). However, GCP4 to 6 are not essential for MT 625 nucleation in fission yeast and Drosophila, while GCP2/Alp4, GCP3/Alp6 and Mozart1 are

626 essential for viability (Anders et al., 2006; Verollet et al., 2006). Puzzlingly, fission yeast

627 Mozart1 is essential for -TuSC recruitment to SPBs (Masuda et al., 2013). This raises the

628 question how budding yeast compensates for Mozart1’s function. GCP3 and -TuCRs of

629 the SPM-CM1 type are conserved between organisms with or without Mozart1. It is likely

630 that small adaptive changes in GCP3 or Spc110 have evolved to compensate for the lack of

631 Mozart1 in budding yeast.

632 What could be the function of -TuCR in organisms that are able to assemble the more

633 stable -TuRC? -TuCRs may recruit and activate the already assembled -TuRC to

634 MTOCs via GCP interactions. In addition, -TuCRs could induce -TuSC assembly into

635 rings that then either function without the help of additional GCPs or subsequently become

636 stabilized by GCP4-6. In any case, considering the conservation between the SPM and

637 CM1 binding elements of -TuRCs and the N-terminal region of Spc98/GCP3, we suggest

638 that the basic principals of MT nucleation are conserved from yeast to human.

639 640 Experimental Procedures

641

642 Plasmid and strain constructions

643 A detailed list of DNA constructs and yeast strains is described in Supplementary File 1 and

644 Supplementary File 2. SPC110 alleles were subcloned into the integration vector pRS304,

645 and SPC98 alleles were subcloned into pRS305K (Christianson et al., 1992; Sikorski and

646 Hieter, 1989; Taxis and Knop, 2006). Point mutations in genes were introduced by PCR-

647 directed mutagenesis and confirmed by DNA sequencing. GST-Spc1101-220 was cloned into

648 pGEX-5X-1 vector (GE Healthcare) and His-tagged Mps1 was cloned into pET28b for

649 expression in BL21 CodonPlus E. coli and into pFastBac1 vector for overexpression in

650 baculovirus-insect cell system. All yeast strains are derivatives of S288c. Gene deletion and

651 epitope tagging of genes at their endogenous loci performed using standard techniques

652 (Janke et al., 2004; Knop et al., 1999). The red fluorescent mCherry was used to mark

653 SPBs through a fusion with SPC42 (Donaldson and Kilmartin, 1996). For SPC97-yeGFP

654 and SPC110-yeGFP strains, the endogenous SPC97 and the integrated SPC110 alleles

655 were tagged with green fluorescent yeGFP. GFP-TUB1 strains were constructed using an

656 integration plasmid (Straight et al., 1997). To remove URA3-based plasmids, transformants

657 were tested for growth on 5-fluoroorotic acid (5-FOA) plates that select against URA3-

658 based plasmids.

659

660 Antibodies

661 Affinity-purified anti-Tub4 (1:1000) and anti-Spc110 (1:1000) antibodies were described

662 previously (Geissler et al., 1996; Spang et al., 1996a; Spang et al., 1996b). Anti-His-tag

663 (1:1000) and anti-GST (1:1000) antibodies were used to detect Spc97 and Spc98 of 664 recombinant -TuSC. Secondary antibodies used in semi-quantitative blotting were

665 IRDye800- or Alexa680-conjugated anti-goat, anti-rabbit, anti-guinea pig and anti-mouse

666 IgGs (1:50,000; Rockland Immunochemicals Inc.). Phospho-specific anti-Spc110pT18, anti-

667 Spc110pS36-pS91 and anti-Spc110pS60-pT64-pT68 antibodies were raised in guinea pigs and

668 purified with immuno-affinity chromatography (1:200; Peptide Specialty Laboratories

669 GmbH).

670

671 Protein purification

672 Subunits of -TuSC and -TuSC∆N-Spc98 were expressed with the baculovirus-insect cell

673 expression system. Purification was performed as described (Gombos et al., 2013).

674 Spc1101-220 variants with an amino-terminal GST tag were purified with Glutathione

675 Sepharose 4B (GE Healthcare) as described (Vinh et al., 2002). Proteins were eluted with 5

676 mM glutathione, concentrated and run over a Superose 6 column equilibrated in HB100 to

677 remove glutathione. Recombinant Mps1 protein was purified in lysis buffer (50 mM

678 NaH2PO4 pH 7.5, 300 mM NaCl, 10% glycerol, 1 mM DTT, 1 mM PMSF, 1X Complete

679 EDTA-free protease inhibitor cocktail (Roche) with column packed with Ni-NTA Sepharose

680 (GE Healthcare). After wash with lysis buffer, protein was eluted with lysis buffer containing

681 200 mM imidazole. Small aliquots for kinase assay were snap frozen with liquid nitrogen

682 and stored at -80°C. Cdk1as1-Clb2 and Cdk1as1-Clb5 complexes were purified from yeast

683 strains kindly given by David Morgan (see Table S1 for detail) (Loog and Morgan, 2005).

684 CLB2-TAP and CLB5-TAP was encoded on 2-micron plasmid and driven by Gal1 promoter.

685 After overexpression with 2% galactose, cells were harvested and lysed for TAP-tag

686 purification as previously described (Ubersax et al., 2003).

687 688 -TuSC oligomerization assay

689 Purified -TuSC and -TuSCSpc98∆N were mixed with or without Spc1101-220 in 1:4 molar ratio

690 (2.3 : 9.3) in TB150 (50 mM Tris-HCl, pH7.0 with 150 mM KCl) or BRB80 (80mM PIPES

691 pH6.9, 1mM EGTA, 1mM MgCl2) buffer as indicated in figure legends. After incubating for 1

692 h at 4°C, the protein mixture was applied to a gel filtration column (Superdex 200 10/300 or

693 Superose 6 10/300 (GE Healthcare)). Elution profiles were recorded as absorbance at 280

694 nm. For each fraction 2% of sample volume was analyzed by SDS-PAGE. Proteins were

695 detected by silver staining or immunoblotting.

696

697 In vitro binding assay

698 For the measurement of the binding of -TuSC to Spc1101-220, recombinant GST (300 nM)

699 or GST- Spc1101-220 proteins (from 0 to 300 nM or fixed 300 nM) were incubated with -

700 TuSC (150 nM) in TB150 buffer in a total reaction volume of 30 μl on a rocking platform for

701 0.5 h at 4°C. Glutathione-Sepharose 4B bead slurry (20 μl; GE Healthcare) was then added

702 to each reaction, followed by rocking for an additional 1 h at 4°C. Beads were washed five

703 times with TB150 buffer, 0.1% NP40 followed by heating in 20 μl SDS sample buffer. Input

704 and bound proteins were analysed by immunoblotting. The signal intensities of protein

705 bands on immunoblots were quantified with ImageJ (NIH). Signal intensities were corrected

706 against the membrane background.

707 708 In vitro MT nucleation assay

709 Microtubule nucleation assay was performed as described (Gombos et al., 2013) with some

710 modifications. 3 μM-TuSC was pre-incubated with 15 μM GST-Spc1101−220 in HB100

711 buffer plus 12.5% glycerol for 30 min on ice. An equal volume of 20 μM bovine brain tubulin

712 containing 4% Alexa568-labelled tubulin in BRB80 buffer with 25% glycerol was added and

713 samples were further incubated for 30 min on ice before being transferred to 37 °°C for

714 15 min for MT polymerization. Samples were fixed with 1% formaldehyde and diluted with 1

715 ml cold BRB80 buffer. 50 l aliquot of reaction mixture was sedimented onto poly-lysine-

716 coated coverslips with a HB-4 rotor at 25,600 g for 1 h. Samples were post-fixed with pre-

717 cooled methanol, mounted on slides with CitiFluor mounting media and imaged with a

718 fluorescence microscope as described. Microtubules were counted in 20 random fields.

719

720 In vitro kinase assay

721 For the in vitro phosphorylation of Spc1101-220 variants, 0.5 g Spc1101-220 substrate was

722 incubated in kinase buffer (50 mM Tris-HCl pH 7.5, 10 mM MgCl2, 2 mM EGTA) with

723 purified Mps1 kinase or Cdk1as1 kinase co-immunoprecipitated with Clb2-TAP or Clb5-TAP.

724 For autoradiography, 5 µCi [-32P]-ATP was added to the reaction mixture and the kinase

725 reaction was incubated for 30 min at 30°C. The reaction was stopped by the addition of

726 SDS sample buffer and analysed by SDS–PAGE, Simply Safe Blue staining (Invitrogen)

727 and autoradiography. For phospho-specific antibody detection, 0.5 mM unlabeled ATP was

728 used in the kinase reaction. For the -TuSC oligomerization assay by gel filtration, the

729 phosphorylation reaction was stopped by addition of 15 M 1NM-PP1 to inhibit Cdk1as1 or

730 10 M SP100625 (Sigma Aldrich) to inhibit Mps1 kinase.

731 732 Electron microscopic analysis of -TuSC oligomers

733 To visualize -TuSC monomers and oligomers, proteins were stained with uranyl acetate

734 and analyzed by electron microscopy. Briefly, we evaporated 300 or 400 mesh

735 copper/palladium grids with a carbon layer on the copper side. To enhance the hydrophilic

736 affinity of the carbon layer, grids were glow discharged for 30~45 sec directly before the

737 preparation started. The protein solution was incubated for 30 sec on the carbon mesh

738 grids at room temperature. The grids were washed and incubated in 2% uranyl acetate for 4

739 min and blotted on Whatman 50 paper. The images were taken in low dose modus with an

740 under-focus between 0.8 to 1.5 m. Particles were viewed with a CM120 electron

741 microscope (Philips Electronics N.V., Eindhoven, Netherlands), which was operated at 120

742 kV. Images were captured with a CCD camera (Keen View, Soft Imaging systems) and

743 viewed with Digital Micrograph Software.

744 Phosphopeptide enrichment and mass spectrometry

745 Phosphopeptides were identified according to Villen et al (Dephoure et al., 2008). In brief,

746 SDS-PAGE separated Spc110 was cut out and subjected to trypsin digestion. After digest,

747 phosphopeptides were enriched by IMAC, starting with 25 µl of PHOS-Select beads

748 (Sigma). Enriched phosphopeptides were eluted and desalted by C18 columns (Stage tips).

749 Peptides were analyzed by LC-MS/MS (Orbitrap Elite, Thermo Fisher Scientific) and data

750 was processed using Thermo proteome discoverer software (1.4). Phosphorylation site

751 localization was performed on the Mascot results using PhosphoRS.

752 753 Growth assay

754 Yeast cells in the early log phase were adjusted to an OD600 of 1 with PBS. Ten-fold serial

755 dilutions of cells were spotted onto the indicated plates and incubated as indicated in the

756 figure legends.

757

758 Quantification of the Spc97-GFP, Spc110-GFP, and GFP-Tub1 signal at SPBs

759 Asynchronous cells were grown in filter-sterilized YPD with additional 0.1 mg/l adenine

760 (YPAD) to an OD600 of 0.3 at 23°C for 3 h and then shifted to 37°C for 1 h. Cells were

761 directly sampled for image acquisition. For the image acquisition, z-stack images with 21

762 0.3 μm steps (2 × 2 binning) were acquired at 37°C with a DeltaVision RT system (Applied

763 Precision) equipped with FITC (fluorescein isothiocyanate), TRITC (tetramethyl rhodamine

764 isothiocyanate) and Cy5 filters (Chroma Technology), a 100x NA 1.4 plan Apo oil

765 immersion objective (IX70; Olympus), and a CCD camera (CoolSNAP HQ; Roper

766 Scientific). Images were processed and analyzed in ImageJ (NIH).

767 The quantification of the mean background intensity and mean fluorescence intensity

768 of Spc97-GFP and Spc110-GFP signals at SPBs was performed on planes having SPBs in

769 focus. Spc97-GFP or Spc110-GFP intensity within the 3×3 pixel-area covering a single SPB

770 or two unsplit SPBs was measured. For GFP-Tub1 at SPBs, GFP-intensity within a 3×3

771 pixel-area surrounding the SPB was measured. Unsplit SPBs were measured together,

772 whereas split SPBs were measured separately. The standard error (SEM) for each data set

773 (n=50) and level of significance was determined by one-way ANOVA with Turkey’s multiple

774 comparisons’ test.

775 776 Time-lapsed live cell imaging

777 Cells were synchronized with -factor for 1.5 cell cycles and then immobilized onto glass-

778 bottom dishes. Dishes were prepared by incubating them with 100 μl concanavalin A

779 solution (6% concanavalin A, 100 mM Tris-Cl, pH 7.0, and 100 mM MnCl2) for 5 min and

780 subsequently washed with 300 μl of distilled water. Yeast cells were allowed to attach to the

781 dishes for 5–15 min at 30°C. The -factor blocked cells were then released by two washes

782 with 2 ml of prewarmed SC medium. Image acquisition was started 10–15 min after release

783 from -factor. Conditions for imaging were as follows: 15 stacks in the FITC channel, 0.1-s

784 exposure, 0.3-μm stack distance, one reference image in bright field channel with a 0.05-s

785 exposure, and 61 frames in total every 3 min. Images were quantified by measuring the

786 integrated density of the sum of projected videos for the region of interest (ROI) around the

787 SPBs and a background region selected from the periphery of the analyzed regions. The

788 mean intensity of the background was subtracted from the ROI. To correct for acquisition,

789 bleaching signal intensities were divided by a bleaching factor. The bleaching factor was

790 determined from the mean of three very short videos that had been generated with the

791 same image acquisition conditions. The data points of bleaching factor were fitted with non-

792 linear one-phase decay equation.

793

794 Nocodazole washout assay (MT regrow assay)

795 Cells (5 × 106 cells/ml) were pre-grown at 23°C in filter-sterilized YPAD. Cells were arrested

796 in G1 by treatment with 10 μg/ml α-factor (Sigma-Aldrich) for 3 h at 23°C until >95% of cells

797 showed a mating projection. G1 arrested cells were released into media with hydroxyurea

798 to arrest cells in S-phase for 2 h. The culture was then shifted to 37°C and nocodazole was 799 added. After 1 h, nocodazole was removed by exchanging media with YPAD containing

800 hydroxyurea. After removal of nocodazole, cells were sampled at 0, 30 and 60 min and

801 fixed with 4% paraformaldehyde. Image acquisition and GFP-Tub1 quantification were

802 performed as described in previous section.

803

804 Trichloroacetic acid (TCA) extraction of yeast cells

805 To measure the expression level of Spc110 variants and GFP-Tub1 in vivo, whole cell

806 lysates were prepared for SDS-PAGE and immunoblotting (Janke et al., 2004; Knop et al.,

807 1999). 2-3 OD600 of late-log phase liquid culture were resuspended in 0.2 M NaOH and

808 incubated on ice for 10 min. 150 μl 55% (w/v) TCA was added, the solutions were mixed

809 and incubated for 10 min on ice. After centrifugation the supernatant was removed. The

810 protein pellet was resuspended in high urea (HU) buffer (8 M urea, 5% SDS, 200 mM Tris-

811 HCl pH 6.8, 0.1 mM EDTA, 100 mM DTT, bromophenol blue) and heated at 65°C for 10

812 min. One-fifth of total sample amount was loaded for SDS-PAGE and western blotting.

813

814 Pull-down of Spc110-GFP with GFP-binder protein

815 For in vivo phospho-Spc110 detection, strains with integrated SPC110-yeGFP or SPC110-

816 TAP alleles were lysed in binding buffer (25 mM HEPES, pH 7.4, 150 mM NaCl, 1 mM

817 EDTA, 1 mM DTT, 50 mM NaF, 80 mM -glycerophosphate, 0.2 mM Na3VO4, and protease

818 inhibitors) lysed with acid-washed glass beads (Sigma-Aldrich) in a FastPrep FP120 Cell

819 Disrupter (Thermo Scientific). Cell lysates were incubated with 0.1% Triton X-100 for 15 min

820 and then clarified by centrifugation at 10,000 g for 10 min. To further solubilize the pellet

821 with insoluble SPBs, binding buffer containg 0.5 M NaCl and 1% Triton X-100 was added to 822 resuspend the pellet and incubated for 40 min. GFP-binder protein (Rothbauer et al., 2008)

823 conjugated Sepharose 4B slurry or IgG-conjugated Dynabeads was added into

824 supernatants and rocked at 4oC for 2 h, followed by five times of washing with binding

825 buffer. Pull-downed Spc110-GFP or Spc110-TAP was eluted by heating with SDS sample

826 buffer and analyzed by SDS-PAGE and immunoblotting or mass spectrometry.

827

828 Bioinformatic analysis

829 Protein sequences of Spc110 and Spc98 and their homologues in selected organisms were

830 aligned with MAFFT algorism built in Jalview software (Katoh and Standley, 2013;

831 Waterhouse et al., 2009) (Figure 2H, 7A, Figure 2—figure supplement 6, Figure 8—figure

832 supplement 1). The threshold of conservation was set as 20% and highlighted. The relative

833 frequency of each amino acid in each position of SPM and CM1 motifs was visualized using

834 the WebLogo 2.0 (Crooks et al., 2004) (Figure 2H, Figure 8, Figure 9, Figure 2—figure

835 supplement 6 and Figure 8—figure supplement 1). To demonstrate the occurrence of SPM

836 and CM1 motifs in selected members of -TuSC receptor family, 47 protein sequences

837 covering the putative CM1 motifs were subjected to MEME motif scanning analysis (Bailey

838 et al., 2009) (Figure 8B).

839 Jpred 3 and Disopred predicted the secondary structure of the N-terminal domain of

840 Spc98 and human GCP3 (Cole et al., 2008; Ward et al., 2004). Domain positions and

841 protein interacting regions of Spc110 and Spc98 (Figure 1A and Figure 7—figure

842 supplement 1A) were illustrated according to -TuSC binding studies of Spc72- and

843 Spc110-truncated forms and to the yeast-two-hybrid studies (Knop and Schiebel, 1998;

844 Sundberg and Davis, 1997; Usui et al., 2003). 845 For the identification of full CM1 motif on Spc110s and degenerative CM1 on Spc72, CM1

846 motif containing -TuCR protein sequences were retrieved from Pfam database

847 (Microtub_assoc, Pfam id: PF07989) and multiple-aligned with Spc110s and Spc72s from

848 species of subphylum Saccharomycotina by MAFFT multiple sequence alignment algorism.

849 To define the CM2 motif pattern, CM2 sequences from human CDK5RAP2 and Drosophila

850 Cnn were used to retrieve protein sequences containing putative CM2 with HMMER server

851 (Finn et al., 2011). The retrieved protein sequences were multiple-aligned with CDK5RAP2

852 and Cnn by MAFFT algorism and the multiple-aligned CM2 sequences were subjected to

853 Weblogo 3.0 server to visualize the pattern.

854

855 Statistical analysis

856 Statistical analysis of fluorescent intensities, immunoblotting intensities and in vitro

857 microtubule numbers was performed with GraphPad Prism 6.1. One-way ANOVA with

858 Turkey’s multiple comparisons’ test was used to compare samples and to obtain adjusted p

859 values. The number of repeated experiments and sample size are indicated in figure

860 legends. No statistical method was used to predetermine sample size. The experiments

861 were not randomized. The investigators were not blinded to allocation during the

862 experiments and outcome assessment. In general the data showed normal distribution.

863

864 Acknowledgements

865 This work was supported by a grant of the Deutsche Forschungsgemeinschaft (Schi295-3-

866 2). Dr. David Morgen is acknowledged for yeast strains. We thank U. Jäkle and S. Heinze 867 for excellent technical support. T.L. is a member of the international graduate school HBIGS

868 and support by a fellowship of the Graduiertenkolleg Regulation of Cell Division.

869 870 References

871

872 Adams, I.R., and Kilmartin, J.V. (1999). Localization of core spindle pole body (SPB) components during SPB 873 duplication in Saccharomyces cerevisiae. J Cell Biol 145, 809-823. 874 Adams, I.R., and Kilmartin, J.V. (2000). Spindle pole body duplication: a model for centrosome duplication? 875 Trends Cell Biol 10, 329-335. 876 Albuquerque, C.P., Smolka, M.B., Payne, S.H., Bafna, V., Eng, J., and Zhou, H. (2008). A multidimensional 877 chromatography technology for in-depth phosphoproteome analysis. Mol Cell Proteomics 7, 1389-1396. 878 Anders, A., Lourenco, P.C., and Sawin, K.E. (2006). Noncore components of the fission yeast gamma-tubulin 879 complex. Mol Biol Cell 17, 5075-5093. 880 Bailey, T.L., Boden, M., Buske, F.A., Frith, M., Grant, C.E., Clementi, L., Ren, J., Li, W.W., and Noble, W.S. 881 (2009). MEME SUITE: tools for motif discovery and searching. Nucleic acids research 37, W202-208. 882 Batzenschlager, M., Masoud, K., Janski, N., Houlne, G., Herzog, E., Evrard, J.L., Baumberger, N., Erhardt, M., 883 Nomine, Y., Kieffer, B., et al. (2013). The GIP gamma-tubulin complex-associated proteins are involved in 884 nuclear architecture in Arabidopsis thaliana. Front Plant Sci 4, 480. 885 Bullitt, E., Rout, M.P., Kilmartin, J.V., and Akey, C.W. (1997). The yeast spindle pole body is assembled around 886 a central crystal of Spc42p. Cell 89, 1077-1086. 887 Byers, B., and Goetsch, L. (1975). Behavior of spindles and spindle plaques in the cell cycle and conjugation 888 of Saccharomyces cerevisiae. Journal of bacteriology 124, 511-523. 889 Choi, Y.K., Liu, P., Sze, S.K., Dai, C., and Qi, R.Z. (2010). CDK5RAP2 stimulates microtubule nucleation by the 890 gamma-tubulin ring complex. J Cell Biol 191, 1089-1095. 891 Choy, R.M., Kollman, J.M., Zelter, A., Davis, T.N., and Agard, D.A. (2009). Localization and orientation of the 892 gamma-Tubulin Small Complex components using protein tags as labels for single particle EM. J Struct Biol. 893 Christianson, T.W., Sikorski, R.S., Dante, M., Shero, J.H., and Hieter, P. (1992). Multifunctional yeast high- 894 copy-number shuttle vectors. Gene 110, 119-122. 895 Cole, C., Barber, J.D., and Barton, G.J. (2008). The Jpred 3 secondary structure prediction server. Nucleic 896 acids research 36, W197-201. 897 Crooks, G.E., Hon, G., Chandonia, J.M., and Brenner, S.E. (2004). WebLogo: a sequence logo generator. 898 Genome research 14, 1188-1190. 899 Daum, J.R., Gomez-Ospina, N., Winey, M., and Burke, D.J. (2000). The spindle checkpoint of Saccharomyces 900 cerevisiae responds to separable microtubule-dependent events. Curr Biol 10, 1375-1378. 901 Dephoure, N., Zhou, C., Villen, J., Beausoleil, S.A., Bakalarski, C.E., Elledge, S.J., and Gygi, S.P. (2008). A 902 quantitative atlas of mitotic phosphorylation. Proceedings of the National Academy of Sciences of the United 903 States of America 105, 10762-10767. 904 Dhani, D.K., Goult, B.T., George, G.M., Rogerson, D.T., Bitton, D.A., Miller, C.J., Schwabe, J.W., and Tanaka, K. 905 (2013). Mzt1/Tam4, a fission yeast MOZART1 homologue, is an essential component of the gamma-tubulin 906 complex and directly interacts with GCP3(Alp6). Molecular biology of the cell 24, 3337-3349. 907 Dictenberg, J.B., Zimmerman, W., Sparks, C.A., Young, A., Vidair, C., Zheng, Y., Carrington, W., Fay, F.S., and 908 Doxsey, S.J. (1998). Pericentrin and gamma-tubulin form a protein complex and are organized into a novel 909 lattice at the centrosome. J Cell Biol 141, 163-174. 910 Donaldson, A.D., and Kilmartin, J.V. (1996). Spc42p: a phosphorylated component of the S. cerevisiae spindle 911 pole body (SPD) with an essential function during SPB duplication. J Cell Biol 132, 887-901. 912 Doxsey, S.J., Stein, P., Evans, L., Calarco, P.D., and Kirschner, M. (1994). Pericentrin, a highly conserved 913 centrosome protein involved in microtubule organization. Cell 76, 639-650. 914 Elliott, S., Knop, M., Schlenstedt, G., and Schiebel, E. (1999). Spc29p is a component of the Spc110p 915 subcomplex and is essential for spindle pole body duplication. Proc Natl Acad Sci U S A 96, 6205-6210. 916 Erlemann, S., Neuner, A., Gombos, L., Gibeaux, R., Antony, C., and Schiebel, E. (2012). An extended γ-tubulin 917 ring functions as a stable platform in microtubule nucleation. J Cell Biol 197, 59-74. 918 Finn, R.D., Clements, J., and Eddy, S.R. (2011). HMMER web server: interactive sequence similarity searching. 919 Nucleic acids research 39, W29-37. 920 Flory, M.R., Morphew, M., Joseph, J.D., Means, A.R., and Davis, T.N. (2002). Pcp1p, an Spc110p-related 921 calmodulin target at the centrosome of the fission yeast Schizosaccharomyces pombe. Cell Growth Differ 13, 922 47-58. 923 Fong, C.S., Sato, M., and Toda, T. (2010). Fission yeast Pcp1 links polo kinase-mediated mitotic entry to 924 gamma-tubulin-dependent spindle formation. EMBO J 29, 120-130. 925 Fong, K.W., Choi, Y.K., Rattner, J.B., and Qi, R.Z. (2008). CDK5RAP2 is a pericentriolar protein that functions in 926 centrosomal attachment of the gamma-tubulin ring complex. Mol Biol Cell 19, 115-125. 927 Friedman, D.B., Kern, J.W., Huneycutt, B.J., Vinh, D.B., Crawford, D.K., Steiner, E., Scheiltz, D., Yates, J., 3rd, 928 Resing, K.A., Ahn, N.G., et al. (2001). Yeast Mps1p phosphorylates the spindle pole component Spc110p in 929 the N-terminal domain. J Biol Chem 276, 17958-17967. 930 Friedman, D.B., Sundberg, H.A., Huang, E.Y., and Davis, T.N. (1996). The 110-kD spindle pole body 931 component of Saccharomyces cerevisiae is a phosphoprotein that is modified in a cell cycle-dependent 932 manner. J Cell Biol 132, 903-914. 933 Geissler, S., Pereira, G., Spang, A., Knop, M., Souès, S., Kilmartin, J., and Schiebel, E. (1996). The spindle pole 934 body component Spc98p interacts with the γ-tubulin-like Tub4p of Saccharomyces cerevisiae at the sites of 935 microtubule attachment. EMBO J 15, 3899-3911. 936 Gillingham, A.K., and Munro, S. (2000). The PACT domain, a conserved centrosomal targeting motif in the 937 coiled-coil proteins AKAP450 and pericentrin. EMBO Rep 1, 524-529. 938 Gombos, L., Neuner, A., Berynskyy, M., Fava, L.L., Wade, R.C., Sachse, C., and Schiebel, E. (2013). GTP 939 regulates the microtubule nucleation activity of gamma-tubulin. Nat Cell Biol 15, 1317-1327. 940 Guillet, V., Knibiehler, M., Gregory-Pauron, L., Remy, M.H., Chemin, C., Raynaud-Messina, B., Bon, C., 941 Kollman, J.M., Agard, D.A., Merdes, A., et al. (2011). Crystal structure of gamma-tubulin complex protein 942 GCP4 provides insight into microtubule nucleation. Nat Struct Mol Biol. 943 Gunawardane, R.N., Martin, O.C., Cao, K., Zhang, L., Dej, K., Iwamatsu, A., and Zheng, Y. (2000). 944 Characterization and reconstitution of Drosophila gamma-tubulin ring complex subunits. J Cell Biol 151, 945 1513-1524. 946 Huisman, S.M., Smeets, M.F., and Segal, M. (2007). Phosphorylation of Spc110p by Cdc28p-Clb5p kinase 947 contributes to correct spindle morphogenesis in S. cerevisiae. J Cell Sci 120, 435-446. 948 Janke, C., Magiera, M.M., Rathfelder, N., Taxis, C., Reber, S., Maekawa, H., Moreno-Borchart, A., Doenges, 949 G., Schwob, E., Schiebel, E., et al. (2004). A versatile toolbox for PCR-based tagging of yeast genes: new 950 fluorescent proteins, more markers and promoter substitution cassettes. Yeast 21, 947-962. 951 Janski, N., Masoud, K., Batzenschlager, M., Herzog, E., Evrard, J.L., Houlne, G., Bourge, M., Chaboute, M.E., 952 and Schmit, A.C. (2012). The GCP3-interacting proteins GIP1 and GIP2 are required for gamma-tubulin 953 complex protein localization, spindle integrity, and chromosomal stability. The Plant cell 24, 1171-1187. 954 Jaspersen, S.L., and Winey, M. (2004). The budding yeast spindle pole body: structure, duplication, and 955 function. Annu Rev Cell Dev Biol 20, 1-28. 956 Katoh, K., and Standley, D. (2013). MAFFT multiple sequence alignment software version 7: improvements in 957 performance and usability. Molecular biology and evolution 30, 772-780. 958 Kawaguchi, S., and Zheng, Y. (2004). Characterization of a Drosophila centrosome protein CP309 that shares 959 homology with Kendrin and CG-NAP. Mol Biol Cell 15, 37-45. 960 Keck, J.M., Jones, M.H., Wong, C.C., Binkley, J., Chen, D., Jaspersen, S.L., Holinger, E.P., Xu, T., Niepel, M., 961 Rout, M.P., et al. (2011). A cell cycle phosphoproteome of the yeast centrosome. Science 332, 1557-1561. 962 Kilmartin, J.V., Dyos, S.L., Kershaw, D., and Finch, J.T. (1993). A spacer protein in the Saccharomyces 963 cerevisiae spindle poly body whose transcript is cell cycle-regulated. J Cell Biol 123, 1175-1184. 964 Knop, M., Pereira, G., Geissler, S., Grein, K., and Schiebel, E. (1997). The spindle pole body component 965 Spc97p interacts with the gamma-tubulin of Saccharomyces cerevisiae and functions in microtubule 966 organization and spindle pole body duplication. EMBO J 16, 1550-1564. 967 Knop, M., and Schiebel, E. (1997). Spc98p and Spc97p of the yeast gamma-tubulin complex mediate binding 968 to the spindle pole body via their interaction with Spc110p. EMBO J 16, 6985-6995. 969 Knop, M., and Schiebel, E. (1998). Receptors determine the cellular localization of a gamma-tubulin complex 970 and thereby the site of microtubule formation. The EMBO journal 17, 3952-3967. 971 Knop, M., Siegers, K., Pereira, G., Zachariae, W., Winsor, B., Nasmyth, K., and Schiebel, E. (1999). Epitope 972 tagging of yeast genes using a PCR-based strategy: more tags and improved practical routines. Yeast 15, 963- 973 972. 974 Kollman, J.M., Polka, J.K., Zelter, A., Davis, T.N., and Agard, D.A. (2010). Microtubule nucleating gamma-TuSC 975 assembles structures with 13-fold microtubule-like symmetry. Nature. 976 Kollman, J.M., Zelter, A., Muller, E.G., Fox, B., Rice, L.M., Davis, T.N., and Agard, D.A. (2008). The structure of 977 the gamma-tubulin small complex: implications of its architecture and flexibility for microtubule nucleation. 978 Mol Biol Cell 19, 207-215. 979 Lawo, S., Hasegan, M., Gupta, G., and Pelletier, L. (2012). Subdiffraction imaging of centrosomes reveals 980 higher-order organizational features of pericentriolar material. Nature cell biology 14, 1148-1158. 981 Li, R., and Murray, A.W. (1991). Feedback control of mitosis in budding yeast. Cell 66, 519-531. 982 Liang, F., Richmond, D., and Wang, Y. (2013). Coordination of Chromatid Separation and Spindle Elongation 983 by Antagonistic Activities of Mitotic and S-Phase CDKs. PLoS Genet 9, e1003319. 984 Lin, T.C., Gombos, L., Neuner, A., Sebastian, D., Olsen, J.V., Hrle, A., Benda, C., and Schiebel, E. (2011). 985 Phosphorylation of the yeast gamma-tubulin Tub4 regulates microtubule function. PLoS One 6, e19700. 986 Loog, M., and Morgan, D.O. (2005). Cyclin specificity in the phosphorylation of cyclin-dependent kinase 987 substrates. Nature 434, 104-108. 988 Masuda, H., Mori, R., Yukawa, M., and Toda, T. (2013). Fission yeast MOZART1/Mzt1 is an essential gamma- 989 tubulin complex component required for complex recruitment to the MTOC, but not its assembly. Molecular 990 biology of the cell. 991 Mennella, V., Keszthelyi, B., McDonald, K., Chhun, B., Kan, F., Rogers, G., Huang, B., and Agard, D. (2012). 992 Subdiffraction-resolution fluorescence microscopy reveals a domain of the centrosome critical for 993 pericentriolar material organization. Nature cell biology 14, 1159-1168. 994 Muller, E.G., Snydsman, B.E., Novik, I., Hailey, D.W., Gestaut, D.R., Niemann, C.A., O'Toole, E.T., Giddings, 995 T.H., Jr., Sundin, B.A., and Davis, T.N. (2005). The organization of the core proteins of the yeast spindle pole 996 body. Mol Biol Cell 16, 3341-3352. 997 Murphy, S.M., Preble, A.M., Patel, U.K., O'Connell, K.L., Dias, D.P., Moritz, M., Agard, D., Stults, J.T., and 998 Stearns, T. (2001). GCP5 and GCP6: two new members of the human gamma-tubulin complex. Mol Biol Cell 999 12, 3340-3352. 1000 Nakamura, M., Yagi, N., Kato, T., Fujita, S., Kawashima, N., Ehrhardt, D.W., and Hashimoto, T. (2012). 1001 Arabidopsis GCP3-interacting protein 1/MOZART 1 is an integral component of the gamma-tubulin- 1002 containing microtubule nucleating complex. Plant J 71, 216-225. 1003 Nguyen, T., Vinh, D.B., Crawford, D.K., and Davis, T.N. (1998). A genetic analysis of interactions with Spc110p 1004 reveals distinct functions of Spc97p and Spc98p, components of the yeast gamma-tubulin complex. Mol Biol 1005 Cell 9, 2201-2216. 1006 O'Toole, E.T., Winey, M., and McIntosh, J.R. (1999). High-voltage electron tomography of spindle pole bodies 1007 and early mitotic spindles in the yeast Saccharomyces cerevisiae. Mol Biol Cell 10, 2017-2031. 1008 Pereira, G., Knop, M., and Schiebel, E. (1998). Spc98p directs the yeast gamma-tubulin complex into the 1009 nucleus and is subject to cell cycle-dependent phosphorylation on the nuclear side of the spindle pole body. 1010 Mol Biol Cell 9, 775-793. 1011 Pereira, G., Tanaka, T.U., Nasmyth, K., and Schiebel, E. (2001). Modes of spindle pole body inheritance and 1012 segregation of the Bfa1p-Bub2p checkpoint protein complex. The EMBO journal 20, 6359-6370. 1013 Rothbauer, U., Zolghadr, K., Muyldermans, S., Schepers, A., Cardoso, M.C., and Leonhardt, H. (2008). A 1014 versatile nanotrap for biochemical and functional studies with fluorescent fusion proteins. Mol Cell 1015 Proteomics 7, 282-289. 1016 Roubin, R., Acquaviva, C., Chevrier, V., Sedjai, F., Zyss, D., Birnbaum, D., and Rosnet, O. (2013). Myomegalin 1017 is necessary for the formation of centrosomal and Golgi-derived microtubules. Biology open 2, 238-250. 1018 Samejima, I., Miller, V.J., Groocock, L.M., and Sawin, K.E. (2008). Two distinct regions of Mto1 are required 1019 for normal microtubule nucleation and efficient association with the gamma-tubulin complex in vivo. J Cell 1020 Sci 121, 3971-3980. 1021 Samejima, I., Miller, V.J., Rincon, S.A., and Sawin, K.E. (2010). Fission yeast mto1 regulates diversity of 1022 cytoplasmic microtubule organizing centers. Curr Biol 20, 1959-1965. 1023 Sawin, K.E., Lourenco, P.C., and Snaith, H.A. (2004). Microtubule nucleation at non-spindle pole body 1024 microtubule-organizing centers requires fission yeast centrosomin-related protein mod20p. Curr Biol 14, 1025 763-775. 1026 Sikorski, R.S., and Hieter, P. (1989). A system of shuttle vectors and yeast host strains designed for efficient 1027 manipulation of DNA in Saccharomyces cerevisiae. Genetics 122, 19-27. 1028 Sonnen, K., Schermelleh, L., Leonhardt, H., and Nigg, E. (2012). 3D-structured illumination microscopy 1029 provides novel insight into architecture of human centrosomes. Biology open 1, 965-976. 1030 Spang, A., Geissler, S., Grein, K., and Schiebel, E. (1996a). gamma-Tubulin-like Tub4p of Saccharomyces 1031 cerevisiae is associated with the spindle pole body substructures that organize microtubules and is required 1032 for mitotic spindle formation. J Cell Biol 134, 429-441. 1033 Spang, A., Grein, K., and Schiebel, E. (1996b). The spacer protein Spc110p targets calmodulin to the central 1034 plaque of the yeast spindle pole body. J Cell Sci 109, 2229-2237. 1035 Stirling, D.A., and Stark, M.J. (1996). The phosphorylation state of the 110 kDa component of the yeast 1036 spindle pole body shows cell cycle dependent regulation. Biochem Biophys Res Commun 222, 236-242. 1037 Stirling, D.A., and Stark, M.J. (2000). Mutations in SPC110, encoding the yeast spindle pole body calmodulin- 1038 binding protein, cause defects in cell integrity as well as spindle formation. Biochim Biophys Acta 1499, 85- 1039 100. 1040 Straight, A.F., Marschall, W.F., Sedat, J.W., and Murray, A.W. (1997). Mitosis in living budding yeast: 1041 anaphase A but no metaphase plate. Science 277, 574-578. 1042 Sundberg, H.A., and Davis, T.N. (1997). A mutational analysis identifies three functional regions of the 1043 spindle pole component Spc110p in Saccharomyces cerevisiae. Mol Biol Cell 8, 2575-2590. 1044 Sundberg, H.A., Goetsch, L., Byers, B., and Davis, T.N. (1996). Role of calmodulin and Spc110p interaction in 1045 the proper assembly of spindle pole body compenents. J Cell Biol 133, 111-124. 1046 Takahashi, M., Yamagiwa, A., Nishimura, T., Mukai, H., and Ono, Y. (2002). Centrosomal proteins CG-NAP 1047 and kendrin provide microtubule nucleation sites by anchoring gamma-tubulin ring complex. Mol Biol Cell 1048 13, 3235-3245. 1049 Taxis, C., and Knop, M. (2006). System of centromeric, episomal, and integrative vectors based on drug 1050 resistance markers for Saccharomyces cerevisiae. Biotechniques 40, 73-78. 1051 Teixido-Travesa, N., Roig, J., and Luders, J. (2012). The where, when and how of microtubule nucleation - one 1052 ring to rule them all. Journal of cell science 125, 4445-4456. 1053 Ubersax, J.A., Woodbury, E.L., Quang, P.N., Paraz, M., Blethrow, J.D., Shah, K., Shokat, K.M., and Morgan, 1054 D.O. (2003). Targets of the cyclin-dependent kinase Cdk1. Nature 425, 859-864. 1055 Usui, T., Maekawa, H., Pereira, G., and Schiebel, E. (2003). The XMAP215 homologue Stu2 at yeast spindle 1056 pole bodies regulates microtubule dynamics and anchorage. EMBO J 22, 4779-4793. 1057 Verollet, C., Colombie, N., Daubon, T., Bourbon, H.M., Wright, M., and Raynaud-Messina, B. (2006). 1058 Drosophila melanogaster gamma-TuRC is dispensable for targeting gamma-tubulin to the centrosome and 1059 microtubule nucleation. J Cell Biol 172, 517-528. 1060 Vinh, D.B., Kern, J.W., Hancock, W.O., Howard, J., and Davis, T.N. (2002). Reconstitution and characterization 1061 of budding yeast gamma-tubulin complex. Mol Biol Cell 13, 1144-1157. 1062 Wang, Y., and Burke, D.J. (1995). Checkpoint genes required to delay cell division in response to nocodazole 1063 respond to impaired kinetochore function in the yeast Saccharomyces cerevisiae. Mol Cell Biol 15, 6838- 1064 6844. 1065 Wang, Z., Wu, T., Shi, L., Zhang, L., Zheng, W., Qu, J.Y., Niu, R., and Qi, R.Z. (2010). Conserved motif of 1066 CDK5RAP2 mediates its localization to centrosomes and the Golgi complex. J Biol Chem 285, 22658-22665. 1067 Ward, J.J., McGuffin, L.J., Bryson, K., Buxton, B.F., and Jones, D.T. (2004). The DISOPRED server for the 1068 prediction of protein disorder. Bioinformatics 20, 2138-2139. 1069 Waterhouse, A.M., Procter, J.B., Martin, D.M., Clamp, M., and Barton, G.J. (2009). Jalview Version 2--a 1070 multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189-1191. 1071 Wiese, C., and Zheng, Y. (2006). Microtubule nucleation: gamma-tubulin and beyond. J Cell Sci 119, 4143- 1072 4153. 1073 Winey, M., Mamay, C.L., O'Toole, E.T., Mastronarde, D.N., Giddings, T.H., Jr., McDonald, K.L., and McIntosh, 1074 J.R. (1995). Three-dimensional ultrastructural analysis of the Saccharomyces cerevisiae mitotic spindle. J Cell 1075 Biol 129, 1601-1615. 1076 Zekert, N., Veith, D., and Fischer, R. (2010). Interaction of the Aspergillus nidulans microtubule-organizing 1077 center (MTOC) component ApsB with gamma-tubulin and evidence for a role of a subclass of peroxisomes in 1078 the formation of septal MTOCs. Eukaryotic cell 9, 795-805. 1079 Zhang, J., and Megraw, T.L. (2007). Proper recruitment of gamma-tubulin and D-TACC/Msps to embryonic 1080 Drosophila centrosomes requires Centrosomin Motif 1. Mol Biol Cell 18, 4037-4049. 1081 Zhang, L., Keating, T.J., Wilde, A., Borisy, G.G., and Zheng, Y. (2000). The role of Xgrip210 in gamma-tubulin 1082 ring complex assembly and centrosome recruitment. J Cell Biol 151, 1525-1536. 1083 Zimmerman, W.C., Sillibourne, J., Rosa, J., and Doxsey, S.J. (2004). Mitosis-specific anchoring of gamma 1084 tubulin complexes by pericentrin controls spindle organization and mitotic entry. Mol Biol Cell 15, 3642- 1085 3657.

1086

1087 1088 Figure legends

1089 Figure 1. Phosphorylation of N-Spc110 is required for the -TuSC oligomerization

1090 (A) Diagram of Spc110’s functional domain organization and the position of phospho-

1091 sites investigated in this study. These sites are located at the N-terminal domain of Spc110,

1092 which directly interacts with -TuSC (Vinh et al., 2002). The conserved centrosomin motif 1

1093 (CM1) (Sawin et al., 2004; Zhang and Megraw, 2007) is also within the N-terminal domain.

1094 The C-terminal domain of Spc110 is involved in the interaction with SPB central plaque

1095 components Spc29 and Spc42 (Adams and Kilmartin, 1999; Elliott et al., 1999). CBD:

1096 calmodulin binding domain.

1097 (B) Post-translational modifications of Spc1101-220 are required for promoting -TuSC

1098 oligomerization. Spc1101-220 was expressed and purified from insect cells (middle panel) or

1099 from E. coli (right panel). Only Spc1101-220 from insect cells carried post-translational

1100 modifications (Figure 1—figure supplement 1). Recombinant -TuSC was incubated with

1101 Spc1101-220 or TB150 buffer on ice. Oligomerization of -TuSC-Spc1101-220 was tested by

1102 gel filtration chromatography using a Superdex 200 10/300 column. Peak fractions of the

1103 chromatograms were analysed by SDS-PAGE and silver staining.

1104

1105 Figure 2. Mps1 and Cdk1 phosphorylation of N-Spc110 stimulates -TuSC

1106 oligomerization

1107 (A) Summary of combination of non-phosphorylatable and phospho-mimicking mutations

1108 in Spc110 used in this study. The indicated Spc1101-220 variants were expressed, purified

1109 from E. coli and then tested for -TuSC oligomerization. WT: wild-type; 2A/D: S36A/D, 1110 S91A/D; 3A/D: S60A/D, T64A/D, T68A/D; 5A/D: S36A/D, S91A/D, S60A/D, T64A/D,

1111 T68A/D (see Figure 2—figure supplement 1A for SDS-PAGE of purified Spc1101-220

1112 proteins and Figure 2—figure supplement 1C for gel filtration chromatograms).

1113 (B) Spc1101-220 phospho-mimicking proteins induced -TuSC oligomerization. Spc1101-

1114 220 proteins were incubated with -TuSC in TB150 buffer. The reconstituted complexes were

1115 separated according to size by gel filtration using a Superose 6 column.

1116 (C) Bar graph of area ratio of void-volume peak to total area of chromatogram of (B). **

1117 marks statistical significance at p<0.01. Error bars represent SEM. N=3 to 9 for the number

1118 of experiment performed.

1119 (D) Void-volume peak fractions of (B) were subjected to negative staining and protein

1120 complexes were analyzed with electron microscopy. Representative ring-like structures of

1121 -TuSC-Spc1101-220-2D, -TuSC-Spc1101-220-3D and -TuSC-Spc1101-220-5D. See Figure 2—

1122 figure supplement 4A for additional EM pictures. Scale bar: 50 nm.

1123 (E) Quantification of (D). Shown is the particle number per field. Particles were

1124 categorized based on the morphology. N is indicated on the figure for the number of fields

1125 analysed.

1126 (F) Quantification of void-volume fractions of -TuSC chromatograms. -TuSC was

1127 incubated with Spc1101-220-WT, Spc1101-220-5D, Spc1101-220-T18A, Spc1101-220-T18D and

1128 Spc1101-220-5D-T18D as described in (B). Bar graph of area ratio of void-volume peak to total

1129 area of chromatogram was calculated as in (C). ** marks statistical significance at p<0.01.

1130 Error bars represent SEM. N=3 to 9 for the number of experiment performed. 1131 (G) Quantification of EM. Void-volume peak fractions of (F) were subjected to negative

1132 staining and protein complexes were analyzed with electron microscopy. Particles were

1133 categorized based on the morphology. Shown is the particle number per field. Note, the

1134 Spc1101-220-WT and Spc1101-220-5D graphs are the same as in (E). N is indicated on the

1135 figure for the number of fields analysed.

1136 (H) Multiple sequence alignment of SPM element of -complex receptors from yeast to

1137 human. Residues are marked according to the ClustalX colour scheme. The occurrence of

1138 each amino acid in each position of CM1 motif is presented with Weblogo 2.0.

1139 (I) The indicated Spc1101-220 proteins (Spc1101-220-5D, Spc1101-220-5D-CM1-QA and

1140 Spc1101-220-5D-∆SPM) were incubated with -TuSC in TB150 buffer. The reconstituted

1141 complexes were separated according to size by gel filtration using a Superose 6 column.

1142

1143 Figure 3. Phosphorylation of N-Spc110 regulates the affinity to -TuSC

1144 (A) GST pull-down assays were performed between -TuSC (containing His-tagged

1145 Spc97-6His and Spc98-6His) and the GST-tagged Spc1101-220 proteins. The bound

1146 proteins were eluted with sample buffer and separated on SDS-PAGE and analyzed by

1147 immunoblotting with anti-His, anti-GST and anti-Tub4 antibodies. Infrared-dye-labelled

1148 secondary antibodies were applied and detected with Li-cor imaging system.

1149 (B) Quantification for bound Spc97, Spc98 and Tub4 from (A) at 300 nM Spc1101-220. **

1150 marks statistical significance at p<0.01. Error bars represent SEM. N=3 for the number of

1151 experiment performed. 1152 (C) Mutations in SPM or CM1 affect -TuSC binding. As (A) but performed with the

1153 indicated SPM and CM1 mutant Spc110 1-220 constructs.

1154 (D) Quantification of Spc1101-220 variants and bound Spc97, Spc98 and Tub4 from (C) at

1155 300 nM Spc1101-220. ** marks statistical significance at p<0.01 and *** at p<0.001. Error

1156 bars represent SEM. N=3 for the number of experiment performed.

1157

1158 Figure 4. Spc1101-220 phosphorylation enhances MT nucleation activity in vitro

1159 (A) Enhancement of MT nucleation activity of Spc110 by Mps1 and Cdk1

1160 phosphorylation. Representative image fields of Alexa546-labeled microtubules from the

1161 nucleation assay polymerized in the presence of buffer, -TuSC and Spc1101-220 variants

1162 (see (B)). Scale bar: 10 µm.

1163 (B) Quantification of the MT nucleation assay. For each experiment, the number of MTs

1164 was counted from 20 fields. The MTs/field was normalized with MTs/field obtained by the

1165 MT nucleation reaction in the presence of buffer only. N=6 to 9 for the number of

1166 independent experiment performed. * marks statistical significance at p<0.05 and ** at

1167 p<0.01. Error bars represent SEM.

1168

1169 Figure 5. Cell cycle dependent phosphorylation of Spc110

1170 (A, B) Two phospho-specific antibodies were generated from guinea pigs to recognize

1171 phosphorylation of Cdk1 sites (pS36-pS91) (A) and Mps1 sites (p60-p64-pT68) (B). In vitro

1172 kinase assays were performed in the presence of recombinant Mps1 or Cdk1as1 and

1173 Spc1101-220, either wild type (WT) or non-phosphorylatable variants as indicated. 3A (Cdk1)

1174 indicates T18A-S36A-S91A and 3A (Mps1) S60A-T64A-T68A. In vitro phosphorylated 1175 Spc1101-220 was subjected to SDS-PAGE and immunoblot with the corresponding phospho-

1176 specific antibodies. The specific kinase activity of Cdk1as1-Clb2 and Cdk1as1-Clb5 was

1177 compared using human histone H1 as substrate (A, bottom). The numbers in the histone

1178 H1 experiment represent the relative kinase activity. As negative control of kinase activity,

1179 Mps1 was inactivated with chemical inhibitor SP100625 (B), while Cdk1as1-Clb2 and

1180 Cdk1as1-Clb5 overexpressed and purified from budding yeast were inactivated with 1NM-

1181 PP1 (A).

1182 (C) Phosphorylation of Spc110 and Spc1105A in vivo. SPC110-GFP wild type cells (WT)

1183 and spc1105A-GFP cells were arrested in mitosis with nocodazole. Spc110-GFP was

1184 enriched using GFP binder conjugated to beads, and the bound proteins were subject to

1185 immunoblotting with the indicated antibodies.

1186 (D) SPC110-GFP wild type cells were arrested in G1, S phase and mitosis as indicated.

1187 Spc110-GFP was enriched with GFP-binder and analyzed by immunoblotting with the

1188 indicated P-specific antibodies. A non-specific band was used as loading control.

1189

1190 Figure 6. Cells with spc110 phospho-, SPM- or CM1-mutant alleles have defects in

1191 spindle formation

1192 (A) Growth of ten-fold serial dilutions of SPC110 shuffle strains with integration vector

1193 encoding SPC110 (WT) or SPC110 mutants with and without the SAC gene MAD2. Growth

1194 was tested either on synthetic complete (SC) plates containing 5-FOA or SC dropout plates.

1195 (B) The indicated SPC110 wild type cells and spc110 mutant cells carrying GFP-TUB1

1196 SPC42-mCherry were incubated at 37ºC, the restrictive temperature of some of the spc110

1197 mutants (Figure 6A). The fluorescent signal of GFP-Tub1 at SPBs was quantified in G1 1198 phase cells (no buds), early S phase cells (small bud with unsplit SPBs) and M phase cells

1199 (large bud with split SPBs and spindle length within 2 m). 50 cells were analyzed per cell

1200 cycle phase and strain. Error bars represent SEM. * marks statistical significance at p<0.05,

1201 ** at p<0.01, and *** at p<0.001. N=2 or 3 for the number of experiment performed.

1202 (C) The representative metaphase cells from (B) were scanned for the distribution of the

1203 GFP-Tub1 and the Spc42-mCherry signal along the spindle axis. Scale bar: 4.5 µm.

1204 (D) In vivo MT re-nucleation assay. Wild-type SPC110 cells (WT) and the indicated

1205 spc110 cells with GFP-TUB1 were treated as indicated in the outline. The GFP-Tub1 signal

1206 at SPBs was measured at 0, 30 and 60 min after nocodazole washout. N=40 for the

1207 number of cells analyzed per time point and strain. Error bars represent SEM.

1208 (E) Data of (E) after 30 and 60 min. Error bars represent SEM. *** marks statistical

1209 significance at p<0.001.

1210

1211 Figure 7. The N-terminus of Spc98 mediates binding to N-Spc110

1212 (A) Alignment of the amino acid sequence of GCP3 homologues from yeast to human.

1213 Shown are the putative -helical regions H1-H5. Residues are marked according to the

1214 ClustalX colour scheme.

1215 (B) Growth test of spc98∆1-156 cells with and without the SAC gene MAD2 at 23ºC and

1216 37ºC.

1217 (C) GFP-Tub1 signal at SPBs in SPC98 wild type or spc98∆1-156 cells. The experiment

1218 was performed as in Figure 6B. ** marks statistical significance at p<0.01. N=100 cells

1219 analysed for each strain. 1220 (D) SPC98 and spc98∆1-156 cells with SPC42-mCherry, SPC97-GFP or SPC110-GFP

1221 were analyzed by time lapse analysis. The fluorescent intensity was measured over time

1222 (right). Error bars represent SEM. N=14 and 19 for the number of cells analysed in SPC97-

1223 GFP and SPC-110-GFP strains respectively. The Spc97-GFP and Spc110-GFP signals of

1224 selected cells are shown on the left panel. Scale bar: 5 m.

1225 (E) Spc1105D combined with -TuSCSpc98∆1-156 fails to induce oligomerization in TB150

1226 buffer. The experiment was performed as in Figure 2B. The purified -TuSCSpc98∆1-156 after

1227 SDS-PAGE and Coomassie Blue staining is shown with relative intensity of the protein

1228 bands. The protein distribution in the Superdex 200 10/300 chromatogram was analyzed

1229 with SDS-PAGE and silver staining.

1230 (F) Binding of -TuSC and -TuSCSpc98∆1-156 to Spc1101-220 mutant proteins. The binding

1231 reaction was performed and analyzed by immunoblotting as described in Figure 3A.

1232 (G) Quantification of Figure 7F. Pull-downed Spc97, Spc98 and Tub4 proteins at 300 nM

1233 Spc1101-220-5D. Error bars represent SEM. N=3 for the number of experiment performed. ***

1234 marks statistical significance at p<0.001.

1235

1236 Figure 8. Two types of -TuCRs defined by the N-terminal -TuSC binding motifs:

1237 SPM-CM1 and CM1-only

1238 (A) Graphical representations of the patterns of CM1 motif within the multiple sequence

1239 alignment of -TuCR protein sequences. The CM1 motif sequence logos were shown for -

1240 TuCR protein sequences retrieved from Pfam database (Microtub_assoc, Pfam id:

1241 PF07989), Spc110s and Spc72s from subphylum Saccharomycotina. The Spc72

1242 sequences used to generate the short CM1 logo are shown in Figure 8—figure supplement 1243 1C. The sequence logos were generated with Weblogo 3.0 server. The conserved residues

1244 shared by CM1 defined by Pfam, Spc110s and Spc72s are marked with asterisks.

1245 (B) SPM and CM1 in PCNT, Spc110 and Pcp1 homologues (SPM-CM1 type) and in

1246 CDK5RAP2/Cnn/Spc72/Mto1 (CM1 type). The occurrences of the motif in the training set

1247 sequences was calculated and shown as coloured blocks on a line with MEME motif

1248 scanning analysis (Bailey et al., 2009). Weblogo motif diagram is shown for each sequence

1249 showing the sites contributing to that motif in that sequence. The best p-value for the

1250 sequence/motif combination is listed. The p-value of an occurrence is the probability of a

1251 single random subsequence with the length of the motif, generated according to the zero-

1252 order background model, having a score at least as high as the score of the occurrence.

1253

1254 Figure 9. -TuCRs are classified into three subgroups based on N-terminal -TuSC

1255 binding motifs and C-terminal MTOC targeting motifs

1256 (A) Graphical representations of the patterns of C-terminal MTOC targeting motifs within

1257 the multiple sequence alignment of -TuCR protein sequences. The MASC and PACT motif

1258 sequence logos were shown for -TuCR protein sequences retrieved from Pfam database

1259 (Mto2_bdg, Pfam id: PF12808; PACT_coil_coil, Pfam id: PF10495) and Spc110s from

1260 subphylum Saccharomycotina. The -TuCR and Spc110 protein sequences used to

1261 generate the CM2 and PACT motif logo are shown in Figure 8—figure supplement 1A, B.

1262 The sequence logos were generated with Weblogo 3.0 server. The conserved residues

1263 shared by PACT motif defined by Pfam and Spc110s are marked with asterisks.

1264 (B) Summary table of categorization of -TuCR protein family. Based on the presence of

1265 N-Terminal -TuSC interacting motifs (CM1 and SPM) and C-terminal MTOC targeting

1266 motifs (MASC, CM2 and PACT), -TuCR protein family can be divided into three 1267 subgroups: N-terminal CM1 motif only with either C-terminal MASC or CM2 motifs, and N-

1268 terminal CM1 and SPM motifs combined with C-terminal PACT domain. Some PACT

1269 domain proteins (AKAP9 or D-PLP) have yet undefined -TuSC binding motifs.

1270

1271 Figure 10. Role of Spc110 phosphorylation during SPB duplication

1272 (A) MT nucleation by the SPB during the cell cycle. See Discussion for details.

1273 (B) Cell cycle dependent, stimulatory phosphorylations of Spc110 by Mps1 and Cdk1-

1274 Clb5 (early S phase to early M).

1275 (C) Model for the interaction of -TuSC with Spc110. The Spc110 dimer interacts via the

1276 SPM and CM1 motifs with the N-terminus of Spc98 (GCP3) and possibly also with other

1277 regions of -TuSC.

1278 Figure 1 – figure supplement 1

1279 Phosphorylation of Spc1101-220 from insect cells

1280 (A) Table of mass-spectrometry identified phosphopeptides of Spc1101-220 purified from

1281 baculovirus-insect cell expression system.

1282 (B) Mass spectra of identified Mps1 sites (S60 and T68).

1283 (C) Mass spectra of identified Cdk1 sites (S36 and S91).

1284 (D) Comparison of investigated phospho-sites of Spc110 N-terminal domain with published data

1285 (Albuquerque et al., 2008; Friedman et al., 2001; Huisman et al., 2007; Keck et al., 2011; Lin et al.,

1286 2011).

1287

1288 Figure 1 – figure supplement 2

1289 -TuSC does not oligomerize in TB150 buffer 1290 (A-C) Purified -TuSC (A) spontaneously oligomerizes in BRB80 buffer (B) but not in TB150 buffer

1291 (C). Proteins were analyzed by Superose 6 10/300 and subsequently gel filtration fractions were

1292 subjected to SDS-PAGE and silver staining. Samples in the peak fraction were subjected to

1293 negative staining and protein complexes were analyzed by electron microscopy. Corresponding

1294 scale bars are shown on the images.

1295

1296 Figure 2 - figure supplement 1

1297 Purification of Spc1101-200 variants

1298 (A) Purified GST-Spc1101-220 variants. ~1 g of protein was loaded to each lane. The SDS-

1299 PAGE was stained with Coomassie Blue.

1300 (B) Calibration of Superose 6 10/300 gel filtration column. Following protein markers were used:

1301 thyroglobulin (667 kDa), ferritin (440 kDa), aldolase (158 kDa), conalbumin (75 kDa), ovalbumin (44

1302 kDa). Partition coefficient (Kav) was calculated using the equation: (Ve-V0)/(Vc-V0), where V0 =

1303 column void volume, Ve = elution volume, and Vc =geometric column volume. The calibration curve

1304 (bottom) of Kav versus log molecular weight was plotted in semi-logarithmic scale.

1305 (C) Gel filtration chromatograms of purified GST-Spc1101-220 variants. Proteins were analyzed in

1306 TB150 buffer by Superose 6 10/300 chromatography.

1307

1308 Figure 2 – figure supplement 2

1309 Phosphomimetic but not SPM defective N-Spc110 proteins induce oligomerization of

1310 -TuSC

1311 (A) Overlapped gel filtration chromatograms with molecular weight markers. The peak of the

1312 void-volume (V0, boxed area) corresponds to molecular weight fractions higher than 5000 kDa.

1313 Note, that the concentrations of -TuSC and Spc1101-220 variants in (A) was double as high 1314 compared to (B) and (C).

1315 (B) Gel filtration chromatograms of -TuSC incubated with phosphomimetic or non-

1316 phosphorylatable Spc1101-220 mutant proteins. -TuSC was incubated in TB150 buffer with Spc1101-

1317 220 proteins from Figure 2—figure supplement 1A. Complexes were analyzed by Superose 6 10/300

1318 chromatography. Fractions corresponding to the peak were analyzed by SDS-PAGE and silver

1319 staining.

1320 (C) As (B) but with Spc1101-220 variants that have defective SPM due to T18D or T18A.

1321

1322 Figure 2 – figure supplement 3

1323 Non-phosphorylatable mutations in Mps1 or Cdk1 sites of Spc110 are neutral to -TuSC

1324 oligomerization induced by phosphomimetic mutations

1325 Gel filtration chromatograms of -TuSC incubated with indicated Spc1101-220 variants in TB150

1326 buffer. Non-phosphorylatable mutations in addition to phosphomimetic mutations were introduced

1327 on Spc1101-220 to examine the possibility that non-phosphorylatable mutations alter overall structure

1328 and thus disrupt the activity to induce -TuSC oligomerization. Complexes were analyzed by

1329 Superose 6 10/300 chromatography.

1330

1331 Figure 2 – figure supplement 4

1332 EM single particle analysis of oligomerized -TuSC

1333 (A) Additional examples of ring-like and filament-like -TuSC-Spc1101-220 complexes. Scale bar:

1334 50 nm.

1335 (B) Summary of particle numbers of -TuSC oligomerized by Spc1101-220 phospho-mimicking

1336 variants. Both the absolute particle number and the percentage of each particle category are shown.

1337 (C) Spc1101-220-2D, Spc1101-220-3D and Spc1101-220-5D induced similar level of ring-like -TuSC- 1338 Spc1101-220 complexes. * marks statistical significance at p<0.05. Error bars represent SEM. N=15-

1339 25 for the number of field analyzed. Correlated to Figure 2E, G.

1340 (D) The diameter of the ring-like -TuSC-Spc1101-220 complexes showed no difference among

1341 Spc1101-220 phosphomimatic variants. Error bars represent SEM. N=9, 12, 18 for the number of ring-

1342 like complexes measured in 2D, 3D and 5D, respectively.

1343

1344 Figure 2 – figure supplement 5

1345 Mutations of T18 abolish the -TuSC oligomerization promoting activity by inactivating the

1346 SPM motif.

1347 (A) Gel filtration chromatograms of -TuSC incubated with phosphomimetic Spc1101-220-5D with

1348 additional T18 mutations. Complexes were analyzed by Superose 6 10/300 chromatography in

1349 TB150 buffer.

1350 (B) Growth of ten-fold serial dilutions of SPC110 shuffle strains carrying the empty integration

1351 vector (null) or SPC110 (WT) or SPC110 mutant alleles on the integration vector. Wild type and

1352 SAC deficient mad2∆ cells were analyzed. Growth was tested either on synthetic complete (SC)

1353 plates containing 5-FOA or SC dropout plates.

1354

1355 Figure 2 – figure supplement 6

1356 Multiple sequence alignment of CM1 motif-containing proteins

1357 Multiple sequence alignment of selected -TuSC receptor family members containing CM1 motifs.

1358 Two of the most conserved residues within CM1 motif are marked with asterisks and mutated to

1359 disrupt CM1 function in the spc110CM1-QA mutant. The occurrence of each amino acid in each

1360 position of CM1 motif is presented with Weblogo 2.0. 1361 Figure 5 – figure supplement 1

1362 Phosphorylation of T18 most likely affects -TuSC oligomerization promoting activity

1363 of Spc110 in a negative manner

1364 (J) Phospho-specific antibody against phospho-T18 (pT18) was generated from guinea pigs. In

1365 vitro kinase assays were performed in the presence of Cdk1as1 and Spc1101-220, either wild type

1366 (WT) or non-phosphorylatable Spc1101-220-3A (Cdk1) (T18A-S36A-S91A). In vitro phosphorylated

1367 Spc1101-220 was subjected to SDS-PAGE and immunoblot with the phospho-specific antibody. As

1368 negative control of kinase activity, Cdk1as1-Clb2 and Cdk1as1-Clb5 overexpressed and purified from

1369 budding yeast were inactivated with 1NM-PP1. Note, the same Spc1101-220 samples were analyzed

1370 in Figure 5A with the anti-pS36-pS91 antibody. The Spc1101-220 blot on the bottom is the same as in

1371 Figure 5A.

1372 (K) Quantification of tryptic phospho-T18 peptide (NMEFpTPVGFIK). Data was acquired in a

1373 data independent fashion in a pseudo MRM experiment in the Orbitrap mass spectrometer

1374 fragmenting peptides with m/z = 689.81. Extracted ion chromatogram of phosphorylated

1375 NMEFTPVGFIK was quantified. Expected retention time of the tryptic phospho-T18 peptide is 24.7

1376 min.

1377 (L) Mass spectrum of fragmented phospho-T18 peptide. The graph summarizes the identified

1378 fragmented ions.

1379 (M) Gel filtration chromatograms of -TuSC. Spc1101-220-5D was incubated with Cdk1as1-Clb2 in

1380 the presence (+1NM-PP1) and absence of the Cdk1as1 inhibitor 1NM-PP1 and ATP. This allowed

1381 Spc110T18 phosphorylation in the absence but not in the presence of 1NM-PP1. Adding 1NM-PP1

1382 stopped all Cdk1as1 kinase reactions. The so modified Spc1101-220-5D-pT18 was then incubated with -

1383 TuSC. -TuSC was analysed by gel filtration using a Superose 6 column for oligomerization. The gel

1384 filtration fractions were analyzed by immunoblotting for Spc97/Spc98, Tub4, Spc1101-220 and 1385 Spc1101-220-pT18 distribution. The anti-pT18 blot on the bottom is a 25-times of contrast enhancement

1386 of the blot on top. Fractions covering the void-volume peak were marked with a red rectangle

1387 outline.

1388 (N) Quantification of void-volume fractions shown in (D). Spc1101-220-5D-T18D was included as

1389 control. Indicated protein band intensities of fractions covering the void-volume peak (red rectangle

1390 outline in (D)) were summed and divided by the total intensity of fractions. * marks statistical

1391 significance at p<0.05, ** at p<0.01, and *** at p<0.001. Error bars represent SEM. N=4 for the

1392 number of experiment performed.

1393

1394 Figure 6 – figure supplement 1

1395 Phenotype of SPC110 mutants

1396 (A) There was no significant difference in the protein levels of Spc110 variants. Asynchronous

1397 cells grown to OD600 0.5 were harvested and subjected to TCA extraction for whole cell protein

1398 lysates. SDS-PAGE and immunoblotting with anti-Spc110 were performed. Tub2 was loading

1399 control.

1400 (B) The representative images of wild type SPC110 (WT) and spc110 cells in G1 (no bud) and S

1401 phase (small bud). Microtubules were marked with GFP-Tub1 and SPBs were marked with Spc42-

1402 mCherry. See Figure 6B for experimental set up. BF: bright field. Scale bar: 4.5 µm.

1403 (C) The indicated wild type SPC110 (WT) and spc110 cells were grown at 37ºC. The cell cycle

1404 stages were determined according to cell morphology and spindle length. Unfixed cells were

1405 analyzed for Spc97-GFP signal at SPBs. N = 30 for the number of cells analyzed for each category.

1406 * marks statistical significance at p<0.05 and ** at p<0.01. Error bars represent SEM.

1407 (D) SPC110 and spc110T18D cells with SPC97-GFP were analyzed by time lapse analysis. The

1408 fluorescent intensity was recorded and measured over time (top). Error bars represent SEM. N=17 1409 for the number of cells analysed for each strains. The Spc97-GFP signals of selected cells are

1410 shown on the bottom. Scale bar: 5 m.

1411 (E) The indicated wild type SPC110 (WT) and spc110 cells were grown at 37ºC. The cell cycle

1412 stages were determined according to cell morphology and spindle length. Unfixed cells were

1413 analyzed for Spc110-GFP signal at SPBs. Error bars represent SEM. N = 30 for the number of cells

1414 analyzed for each category.

1415 (F) Similar experiment as performed in Figure 6B. The indicated SPC110 wild type and

1416 spc110∆SPM carrying GFP-TUB1 SPC42-mCherry were incubated at 37ºC. The fluorescent signal of

1417 GFP-Tub1 at SPBs was quantified in G1 phase cells (no buds), early S phase cells (small bud with

1418 unsplit SPBs) and M phase cells (large bud with split SPBs and spindle length within 2 m). 100

1419 cells were analyzed per cell cycle phase and strain. Error bars represent SEM. ** marks statistical

1420 significance at p<0.01, and *** at p<0.001.

1421 (G) Similar experiment as performed in (D) with SPC110 and spc110∆SPM cells. SPC110 and

1422 spc110∆SPM cells with SPC97-GFP were analyzed by time lapse analysis. Error bars represent SEM.

1423 N=17 for the number of cells analysed for each strains.

1424 (H) Similar experiment as performed in (E). The wild type SPC110 (WT) and spc110SPM cells

1425 were grown at 37ºC. The cell cycle stages were determined according to cell morphology and

1426 spindle length. Unfixed cells were analyzed for Spc110-GFP signal at SPBs. Error bars represent

1427 SEM. N = 40 for the number of cells analyzed for each category.

1428

1429 Figure 7 – figure supplement 1

1430 The importance of the N-terminal region of Spc98

1431 (A) Structure and functional domain organization of budding yeast Spc98 and human GCP3.

1432 Both Spc98 and GCP3 share the conserved N-terminal domain (NTD), the GCP core body and the

1433 divergent, unstructured linker connecting NTD and GCP core body. The GRIP1 region is within the 1434 region interacting with Spc97, while the GRIP2 covers the region interacting with Tub4 (Knop et al.,

1435 1997; Kollman et al., 2010; Pereira et al., 1998; Wiese and Zheng, 2006).

1436 (B) The growth test of MAD2 or mad2∆ cells with spc98 N-terminal truncated mutant alleles.

1437 Serial diluted cells were spotted on YPAD plates and incubated at indicated temperatures for two or

1438 three days. SPC98 indicates the spc98∆ SPC98 cells while WT is the unmodified wild type strain.

1439 The design of each truncated variants is shown in the cartoon below.

1440 (C) The representative images of spc98∆1-156 cells in M phase (large budded, bipolar spindle

1441 within 2 m long). SPC98 wild type and spc98∆1-156 cells with SPC42-mCherry GFP-TUB1 were

1442 grown at 37ºC for 2 h. Metaphase-like cells were analyzed for the presence of bipolar spindles. BF:

1443 bright field. Scale bar: 4 m.

1444 (D) Quantification of MT phenotype of metaphase cells of (C). One hundred SPC98 and spc98∆1-

1445 156 cells with a large bud and a SPB distance of ~2 µm were categorized into “normal spindle” or

1446 “immature/defective spindle”.

1447

1448 Figure 7 – figure supplement 2

1449 -TuSCSpc98∆1-156 maintains self-oligomerization capability in BRB80 buffer

1450 (A) -TuSCSpc98∆1-177 was purified from insect cells, incubated in TB150 buffer and then analyzed

1451 by Superose 6 10/300. The left panel presents the Coomassie Blue stained SDS-gel of purified -

1452 TuSCSpc98∆1-177.

1453 (B) -TuSCSpc98∆1-177 in fraction 7 from (A) were subjected to negative staining and analyzed with

1454 electron microscopy. The particles showed irregular morphology, suggesting the formation of

1455 protein aggregates. Corresponding scale bars are shown in the images.

1456 (C) Similar experiment as performed in (A). Purified -TuSCSpc98∆1-156 was incubated in TB150 or

1457 BRB80 buffer and then analyzed by Superose 6 10/300.-TuSCSpc98∆1-156 oligomerized in BRB80

1458 but not in TB150 buffer. 1459 (D, E) Similar experiment as performed in (B). After the gel filtration shown in (C), the peak fraction

1460 of-TuSCSpc98∆1-156 (fraction 7 in BRB80 and fraction 13 in TB150) was subjected to negative

1461 staining and analyzed with electron microscopy. Corresponding scale bars are shown on the

1462 images.

1463 

1464 Figure 8 – figure supplement 1

1465 Raw sequence input for the sequence logo of CM2, Spc110 PACT and Spc72 CM1 motifs

1466 (A) Multiple sequence alignment of selected -TuSC receptor family members containing CM2

1467 motif. Residues are marked according to the ClustalX colour scheme. The occurrence of each

1468 amino acid in each position of CM2 motif is presented with Weblogo 3.0. The CM2 motif is present

1469 in human CDK5RAP2, myomegalin and Drosophila centrosomin (Cnn).

1470 (B) Multiple sequence alignment of PACT motif sequences of Spc110s from the subphylum

1471 Saccharomycotina. The graphic presentation was generated as in (A).

1472 (C) Multiple sequence alignment of CM1 motif sequences of Spc72s from the subphylum

1473 Saccharomycotina. The graphic presentation was generated as in (A).

1474

1475 Supplementary File 1

1476 Plasmids used in this study

1477

1478 Supplementary File 2

1479 Strains used in this study Figure 1- Lin et al

A T18 S36 S91 Spc29 BD CM1 CBD SPM Coiled-coil

1 183 773 944 S60 T64 T68 Spc42 BD

B -TuSC only Spc1101-220-P (insect cell) Spc1101-220 (E. coli) kDa 78910111213 ml kDa7 8 9 10 11 12 13 ml kDa 78910111213 ml 100 100 100 Spc97 Spc98

55 55 55 Tub4 Spc1101-220

20 15 25 20 15 10 15 (mAU) (mAU) 10 (mAU) 10 280 280 280 5 A A A 5 5 0 0 0 579111315 579111315 5 7 9 11 13 15 Elution volumn (ml) Elution volumn (ml) Elution volumn (ml) Superdex 200 Superdex 200 Superdex 200 Figure 2 - Lin et al

A B C 2A/2D Gel filtration chromatograms S36 S91 60 ** CM1 SPM 50 5A/5D 40 1 183 220

(mAU) 30

S60 T64 T68 280 A 20 3A/3D 10 1.-TuSC

2.-TuSC + Spc1101-220-WT 0 peak area to total

1-220-2D 6 Percentage of void-volume 3.-TuSC + Spc110 5 1-220-3D 4 4.-TuSC + Spc110 Sample3 16 18 2 12 14 5.-TuSC + Spc1101-220-5D 1 V 8 10 6 0 6.-TuSC + Spc1101-220-5A Spc1101-220 Elution volume (ml) Superose 6 10/300

D E Oligomer Ring F Incomplete ring Spiral 1-220-2D 1-220-5D ** -TuSC/Spc110 -TuSC/Spc110 6 5 4 3 2 -TuSC/Spc1101-220-3D Particles/field 1 0 peak area to total -TuSC WT 2D 3D 5D Percentage of void-volume only WT 5D T18A N = 20 15 25 17 22 T18D 5D-T18D

Spc1101-220

G 4 I H 3

6 bits 2 1L I Y I E Y A FQ E Q SRS F GL E N HVQ R NN P Q K S F L SQ 0 M L QVLN A R 2 1 3 4 8 5 7 6 9 K 1-220-5D 10 5 N E T 11 12 13 C -TuSC + Spc110 SPM 4 Oligomer 20 Incomplete ring PCP1_SCHPO 3 G8ZV09_TORDC 15 SP110_ZYGRC Ring SP110_LACTC SP110_ASHGO 10 2 I6NDW4_ERECY Spiral SPC110_KLULA (mAU) SPC110_CANGA Particles/field 280 G8BX70_TETPH 5 1 SP110_VANPO A SP110_YEAST SP110_SACAR 0 0 H2AW02_KAZAF 5 7 9 11 13 15 17 J7RX36_KAZNA WT 5D 5D-T18D K0KVJ5_WICCF Elution volumn (ml) G5C7V2_HETGA PCNT_MOUSE PCNT_RABIT Superose 6 PCNT_HUMAN N =15 22 16 PCNT_PONAB PCNT_PANTR F6S0J0_MACMU 1-220-5D-CM1-QA PCNT_CANFA -TuSC + Spc110

20

15

10 (mAU)

280 5 A

0 57911131517 Elution volumn (ml) Superose 6

-TuSC + Spc1101-220-5D-SPM 20

15

10 (mAU)

280 5 A

0 57911131517 Elution volumn (ml) Superose 6

- Figure 3 - Lin et al.

A C Spc1101-220-WT Spc111-2201-220-2D Spc1101-220-3D Spc1101-220-5D 1. GST kDa GST GST GST GST 2. Spc1101-220-5D His (Spc97) 100 1-220-5D-T18D His (Spc98) 3. Spc110 4. Spc1101-220-5D-CM1-QA S 55 Tub4 5. Spc1101-220-5D-T18D-CM1-QA 1-220 55 GST (Spc110 ) 6. Spc1101-220-5D-SPM

100 His (Spc97) His (Spc98) Spc1101-220 (300 nM) P 55 Tub4 55 GST (Spc1101-220) kDa 12345 6

100 His (Spc97/Spc98)

B Spc97, Spc98 Tub4 55 Tub4 ** ** 55 GST (Spc110-N)

100 His (Spc97/Spc98)

55 Tub4

55 GST (Spc110-N)

Spc110 1-220 Spc110 1-220 (300 nM) (300 nM)

D GST-Spc1101-220 Spc97/Spc98 Tub4 *** *** *** ** 140 120 100 80 60 40 20 Relative intensity (%) Relative intensity (%) 0 Relative intensity (%)

5D QA QA 5D QA QA 5D QA QA SPM SPM SPM 5D-T18D 5D- 5D-T18D 5D- 5D-T18D 5D- 5D-CM1 5D-CM1 5D-CM1

5D-T18D-CM1 5D-T18D-CM1 5D-T18D-CM1 Figure 4 - Lin et al

A

1234

5 678

* * B ** ** ** **

-TuSC --++++++ Spc1101-220 - - 5D WT 2D 3D 5D 5D-T18D

12345678 Figure 5 - Lin et al

ABCdk1as1 Mps1 Clb2-TAP Clb5-TAP Spc1101-220 Spc1101-220 WT 3A (Mps1)WT 3A (Mps1) WT 3A (Cdk1)WT 3A (Cdk1)WT 3A (Cdk1)WT 3A (Cdk1) SP100625 -- ++ 1NM-PP1 --++--++ (Mps1 inhib.) 55 pS60-pT64 55 pS36-pS91 -pT68

55 55 Spc1101-220 Spc1101-220 kDa kDa

Cdk1as1 Clb2-TAP Clb5-TAP 1NM-PP1 - + - + 32P Histone H1 CBB 1 0.94

C Noco D WT

kDa WT 5A D-factor HU Noco kDa (G1) (S) (M) 130 pS60-pT64-pT68 (Mps1) pS60-pT64-pT68 130 (Mps1) 130 pS36-pS91 (Cdk1)

IP with GFP binder 130 Spc110 pS36-pS91 130 (Cdk1)

IP with GFP binder Loading control 70 Figure 5 - figure supplement 1 - Lin et al

A B

Retention time (min): 22.98-27.01 Cdk1as1 100 G1, supernatant

50 RT: 24.72 Clb2-TAP Clb5-TAP AA: 3075 Intensity (%) 0 100 S, supernatant Spc1101-220 50 Intensity (%) 0

WT 3A (Cdk1)WT 3A (Cdk1)WT 3A (Cdk1)WT 3A (Cdk1) 100 M, supernatant RT: 24.76 AA: 6601 1NM-PP1 --++--++ 50 .

Intensity (%) . . . 0

100 55 G1, pellet pT18 50

Intensity (%) 0

100 55 S, pellet Spc1101-220 50 kDa Intensity (%) 0 100 RT: 24.74 AA: 7737 M, pellet 50 RT: 25.37 AA: 749 Intensity (%) 0 23.0 23.5 24.0 24.5 25.0 25.5 26.0 26.5 27.0 Time (min) C

#1 bƈ b²ƈ Seq. yƈ y²ƈ #2 1 115.05021 58.02874 N 11 2 262.08562 131.54645 M-Oxidation 1264.56841 632.78784 10 3 391.12822 196.06775 E 1117.53299 559.27013 9 4 538.19664 269.60196 F 988.49039 494.74883 8 5 719.21065 360.10896 T-Phospho 841.42197 421.21462 7 6 816.26342 408.63535 P 660.40796 330.70762 6 7 915.33184 458.16956 V 563.35519 282.18123 5 8 972.35331 486.68029 G 464.28677 232.64702 4 9 1119.42173 560.21450 F 407.26530 204.13629 3 10 1232.50580 616.75654 I 260.19688 130.60208 2 11 K 147.11281 74.06004 1

E D Spc97/Spc98 Tub4 Cdk1as1-Clb2 Cdk1as1-Clb2 + 1NM-PP1 * ** ** ** 45 35 20 20 35 25 15 15 25 15 10 10 15 (mAU) (mAU) Percentage of total of Percentage Percentage of total of Percentage

specified proteinspecified (%) 5 280 280 5 proteinspecified (%) 5 5 A A

0 0 PP1 - + - - PP1 - + - - 5 7 9 11 13 15 17 5 7 9 11 13 15 17 Spc1101-220 - Spc1101-220 - Elution volume (ml) Elution volume (ml) 5D 5D 5D 5D Superose 6 Superose 6 5D-T18D 5D-T18D kDa6 7 8 9 10 11 12 13 14 151617 ml kDa 6 7 8 9 10 1112 13 14 1516 17 ml 1-220 His (Spc97) Spc110 100 100 His (Spc98) 55 55 Tub4 25 *** *** 55 55 20 GST (Spc1101-220-5D) 15

55 10 55 pT18 5 Percentage of total specified protein (%) 55 55 pT18 (contrast enchanced) PP1 - + - -

1-220 - Spc110 5D 5D

5D-T18D Figure 6 - Lin et al.

B A G1: GFP-Tub1 S: GFP-Tub1 SC-Trp 5-FOA SC-Trp 5-FOA *** *** 23oC 23oC 37oC 23oC 23oC 37oC *** WT 1000 *** 1000

null 800 800 2A (Cdk1 sites) 600 600 2D (Cdk1 sites) 3A (Mps1 sites) 400 400

3D (Mps1 sites) Intensity (RU) 200 Intensity (RU) 200 5A (Cdk1 +Mps1 sites) 0 0 5D (Cdk1 +Mps1 sites) T A D T D W 5A 5D 8 8 W 5A 5D T1 T1 T18A T18 WT 5D-T18D 5D-T18D null T18D M: GFP-Tub1 QA CM1 *** T18D-CM1QA SPM 1000 *** *

800 WT 600 null T18D 400 2D-T18D Intensity (RU) 200

3D-T18D 0 T A D 5D-T18D W 5A 5D 8 T1 T18 5D-T18D MAD2 mad2

C M phase M phase Spc42- GFP- Spc42- GFP- BF mCherry Tub1 BF mCherry Tub1

1500 1500 1000 1000 WT 500 500

(RU) T18D 0 (RU) 0 01234 01234  uorescent intensity

4.5 m intensity uorescent Distance (m) Distance (m)

1500 1500 1000 1000 5A 500 5D-T18D 500 (RU) 0 (RU) 0 01234 01234 uorescent intensity intensity uorescent uorescent intensity intensity uorescent Distance (m) Distance (m)

1500 5D 1000 500 (RU) 0 01234 uorescent intensity intensity uorescent Distance (m)

D 23oC 37oC HU E -factor HU Noco HU 30 min 60 min 3 hr 2 hr 1 hr *** 0 30 60 min *** *** 200 *** 200 90 WT 150 *** *** 150 5A 100 100 5D 60 T18A 50 50 T18D 0 0 5D-T18D 30 Fluorescent Intensity (RU) -50 Fluorescent Intensity (RU) -50 T A D T A D W 5A 5D 8 8 W 5A 5D T1 T1 T18 T18 5D-T18D 5D-T18D

Fluorescent Intensity (RU) 0 03060 Spc1101-220 Spc1101-220 Time after releasing from nocodazole (min) Figure 7 - Lin et al.

A H1 H2 H3 B

GCP3_ARATH GCP3_Danio_rerio GCP3_CRIGR o o GCP3_HUMAN 23 C37C GCP3_Gallus_gallus GCP3_XENLA GCP3_CULQU YPH499 GCP3_AEDAE GCP3_CAMFO GCP3_HARSA GCP3_ACREC MAD2 SPC98 GCP3_SCHPO GCP3_DROME SPC98_YEAST spc981-156 H4 H5

GCP3_ARATH YPH499 GCP3_Danio_rerio GCP3_CRIGR GCP3_HUMAN mad2 GCP3_Gallus_gallus SPC98 GCP3_XENLA GCP3_CULQU 1-156 GCP3_AEDAE spc98 GCP3_CAMFO GCP3_HARSA GCP3_ACREC GCP3_SCHPO GCP3_DROME SPC98_YEAST

C D 200 M: GFP-Tub1 Spc97-GFP WT: Spc97-GFP 5 m WT: Spc42-mCherry 150 spc981-156: Spc97-GFP spc981-156: Spc42-mCherry ** SPC98 100 500

50

400 spc981-156 Abs intensity (RU) 0 -40 -30 -20 -10 0 10 20 30 300 -27-18 -9 0 9 18 27 Time (min)

200

Intensity (RU) Spc110-GFP 500 5 m 100 WT: Spc110-GFP 400 WT: Spc42-mCherry SPC98 spc981-156: Spc110-GFP 0 300 spc981-156: Spc42-mCherry

200 spc981-156 WT 100 Abs intensity (RU) 1-156 -27-18 -9 0 9 18 27 0 -40 -30 -20 -10 0 10 20 30 Time (min)

25 EFkDa Protein (RI) 20 -TuSC Spc981-156  1-220-5D  -TuSC 100 Spc97 (0.9) 15 -TuSC + Spc110  Spc981-156 1-220-5D (mAU) 10 -TuSC + Spc110 1-156 70 Spc98 (0.9) 280 1. Spc1101-220-5D kDa 12345 67 A 5 2. GST Tub4 (1) 0 100 55 1-220-5D His (Spc97) 6 8 10 12 14 3. Spc110 1-156 1-220-5D-T18D His (Spc98 ) Elution volumn (ml) 4. Spc110 Superdex 200 5. Spc1101-220-5D-CM1-QA 55 Tub4 6. Spc1101-220-5D-T18D-CM1-QA Input 1-220-5D-SPM Spc981-156 7. Spc110 -TuSC -TuSC 55 GST (Spc1101-220) kDa 78910111213 kDa 78910111213ml 100 100 Spc97 100 His (Spc97) 1-156  His (Spc98 ) Spc98 1-156 Tub4 55 55 55 Tub4 Spc1101-220-5D 55 Resin-bound GST (Spc1101-220)

G Spc97/Spc98 Tub4  

100 100

80 80

60 60

40 40

20 20 Relative intensity (%) Relative intensity (%) 0 0 134567 134567

-TuSC -TuSCSpc98-1-156 -TuSC -TuSCSpc98-1-156 Figure 8 - Lin et al.

A GCP receptor N-terminal CM1 motifs

4.0 3.0 Full CM1 2.0 H bits K Q D R E E M V Q E Q KK LT EST DR S EER QK S E IK V K T R L L L A KD E E E L E E E S 1.0 Q G Y L G A E E L L N EK D L A K K L KK A DL T E S M F K I A L L RR L N Q G S E R H F G T A K S N A P K L Y S A V D VR R VKS D F NHR H D K D K K W E Y SR M D Q N E N F I K S D G L R Q D D Y ED R R R S M M NI A L V Q R Q QK A R G P Y R IQ D MY S D A H SE N L E N E E F RS L L V K I A S T A Q L D I D D R Q T L Q Q I T Y V V Q Y K V Q L D V I R H A I N Q D S K A A I S S K L T G R K E Q A F H E YA Q E V Q E N N Q Q A L A Q E D Q I F N A T S Q E S N ET T V Q TV E NS V N D C R K N KM I E T L T M L N R K TN EK E T S H D A KG I Q R N L P K Q D R YS V M V L S I I A V (Pfam) 0.0 IE T I 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

***** ******************* * *

4.0 Full CM1 3.0 2.0

bits Q Q K Y K Y D L K Y L E E C S KF N A T K DEL I D Q N P T P R N N F N K I K D EL L K Q N K K E E L E N K Q IQ K EE K E D Y E 1.0 N N S L E E E K L L D Q IA E K S L Q E K E H K E D K I SQL E D V E L Q K E H Q NN L I L F N R L K L D S C W H N T R M T G R D AT L Q T L R Q A Y G Y H S Q E L Q L I S S Q QK R LR K S E G E S H I F S V FE G A S N V A I LA I E I K A K T Q Y M T K S R D N K T Q E E R V K A E D I Q T H S TR S L G M V S I Q S C D E KN V N R L I L A W T Q R D E F I S I K Y I S E E T Q K N K A L E M NA K Y N A T G G L K K S K V S H N L P S L S R MR I R G Q R Q L T E DN N N R Q R N A S S R C R EH E I K A V GRK Q IME M S Q P Q A L R V R H A I K L H L D A D N M N A S T S (Spc110s) A D A D K K D V V G N NG H Y FG M G V R D DN R R R M N S S I R N I E SN I S S M N V L Q Q N N I M V Q S D H Q N N AI Y S SN V R L K H I S KA K S T T I I LA V F M V T Q S Q S Q E Q S S Y E T V L Y N YVK M L MT Y V T R T 0.0 T I 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

*** *** ***** ** 4.0 Short CM1 3.0 2.0 bits Q Q K F KQ E K L K K M SE E T I 1.0 A L S K I L I N Y I S N Y L QEL Q N L L E TNRV A A T D V D E VT R S D D A H D FQ K G E K R GC TM K L E Q H K R H QV I L S R I R K D K V V L M V N G A K N F L N I R I N L H T S S HV G L I R R A Q Q A A N V S L M Y (Spc72s) F NF

SS D QV V Y S R N D V KTH KR L RY D VMY 0.0 V M V 5 10 15 20 25 30

B SPM CM1

E1BKZ0_BOVIN

PCNT_HUMAN

G5C7V2_HETGA

PCNT_MOUSE PCNT G3H8X7_CRIGR

SPC110_YEAST

SP110_SACAR

G0VH34_NAUCC

G8BX70_TETPH

G8ZV09_TORDC

SPC110_CANGA

SP110_VANPO

SP110_LACTC

I6NDW4_ERECY Spc110 SP110_ASHGO

SPC110_KLULA

SPC110_CANGA

J7RX36_KAZNA

K0KVJ5_WICCF

Q7S473_NEUCR

Q874Y4_PODAS

G2XH36_VERDV

G0SB24_CHATD

C8VIR1_EMENI

Q2UQD3_ASPOR Pcp1 Q0U842_PHANO

H9FD39_MACMU

PCP1_SCHPO

SPC72_YEAST

MTO1_SCHPO

APSB_EMENI

CNN_DROME

Q0IER0_AEDAE

F1MGF7_BOVIN

CK5P2_HUMAN

CK5P2_MOUSE

FINPH8_CHICK

E7F4R1_DANRE MYOME_HUMAN H2R277_PANTR

F7B2D2_CALJA Cnn/Spc72/Mto1 G1LZ30_AILME F1MQ84_BOVIN CDK5RAP2/mypmegalin/ F6VAV7_HORSE

H0XE44_OTOGA G10TX3_RABIT

MYOME_MOUSE Figure 9 - Lin et al.

A

GCP receptor C-terminal motifs

4.0 MASC 3.0 2.0 bits R A G KLWI R R L S K E EE KL N RQ LE E 1.0 H E E R Q L E KEG AN DA R RK E R L Q R Q K K Q A K Q Q A E Q L E G R S LQ R D K A V L RND K S I N Q R E F I A A Q S AS M Q K H I D S A R M R QR KR NR K ST R K T MQ L Y M T N Q QK T M V I GA K H A R V S S S H N A I T V S S E Q E M S Y AA KA A I FN T A V D E S D S D R Q E DK Q R Q K Y I K N E V M A M D V GL Q K N M (Mto2_bdg DL KW E SA 0.0 R ERE R R L in Pfam) 5 10 15 20 25 30 35 40 45

4.0 3.0 2.0 CM2 bits I R M K K QN VKR R KS V FD R D K G M T K SE T R V E V K A A QL K LK K D Q 1.0 S A H Q VS K Q E R Q A L S Q T L E M K F V K F S L KVER L R T R K EH A KR Q I R E E H L T L Q Q D N Q V D VI K C Y M S M K D A N K M A N L R Y A E A FD S K V R E G D G E K N A L H K T Q V M A I E A T A K F H S N D S S L H L T I N H K Q L H T Q M Q V Q T K R I T N A C C SR TTQ N M T E I S KR D V D S Q S I N V R H Q S Y A T Q Y Q YL V H T R G L K V V L A F N Q RT F L E K T N S QIL I RKT 0.0 E L L L E I L R N E 5 10 15 20 25 30 35 40 45 T 50 55 60

4.0 3.0 2.0

bits R K Y KY A F Q I R QIQW E R R L A F S R E A M K P Q T L E R A AC V T S E F R K G E T 1.0 Y E L N F K DK R Y L G S I A Q Q A V R R K R L F L W N Q A Q H R G R GK D R A G S L S T K T I Y YL S LQ K Q M S A V R R G H E S K T S S R A M K N I V I S K A L Q L A T R T T I A E P Q R G F V A G E K A E Y K K M L A H D K K L A H E H F N S K P M L P R S A R E I P SQ R D T R I A K WK G V S P M E F L P S L G E I A G D E T K D E I E LK RF S VL E P R L Q V Y Q S E K E K A G G L T S A L T T N R H P A E L S S F C K C Q E A K N I T M I H V P T Q K Q F Q V A G AD Q D T R D N S S E S I E K R G M H G K R F L V G L P S F L V R F R P A Q V L V A A R V H M N S E C T R D M F I D L T M A K V Q F R LT E D S A H V M N IL E T N E T S T S DS M W I C N G I A IF N S G L A K L V I R G R LA V L LV P LK PACT 0.0 R R 5 10 15 20 25 K 30 35 40 45 50 55 60 65 70 75 80 (Pfam) 4.0 3.0 2.0 bits K H VVI F R Q V I R V 1.0 R R S LM R A I KL K LA L V MAK V Y E S A AR A R AE AV VATQ L N S IS LTM I Q F A K I I F AL SGNIA W R I L TD S M V M G CSL Q YN 0.0 TA R 85 90 95 W * *** ** * **** 4.0 3.0 2.0 bits H H M F D Y N K V D L QH K Y E K G T R F Y K QD I Y R K T V G P T D N N K I S R N R E D A L E N N S R V H K K L V 1.0 E M E Q K N K I IK Q L D F D S S K R S Y R L H K R A R S T E S L L E Y L T L A L GV K L A R R R E RE L I I I F L D L G K R R I E I D KQL S S D G L E F Q N N R N A L R Y I R K E L N R S F N Y P R D I I R G A N A Q S K R F R L L F E F N A N N R G ID M P Y E N I R Q E E L N A T F K M N G Q A Y T A A Y T M Q V V E L M A Q S Y S I N L G G A K L E V R E H T P S HN K N NI E T G T D I S Q Q A L S Q H Q L S P Y T N E I D D Y A Q E A D R I S YS F V G P E H MS V K Y S K L VDA EYM I D Q N K I S S L T R N Q K G L N S D P K D H S I GK M N A N A P K V HR I E TQ L KV L A D R R K N H M A F R Q S T G R R Q K Y F S VK P P N P N S L K LT N F N E Q N GM M M IQ QK K Q D N T FR K T RS K TRY T LY ELMKVM SYTSYSQVV L S QMT Y TTT SY R L T Y V S Q Q TR V T RF S Y T V N Y F P Y P LKA I PACT 0.0 D Y I F 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 (Spc110s) 4.0 ******** 3.0 2.0 bits C M R F K RA KH R L R 1.0 L R T A I Q E K V KYS K V R F EE L Q D S NA D TR A SAF M K YK A PA D Q A M HMI IT F F N V Q T K SS M LD Q SSYILVKSVILLLVTT I 0.0 A V R 85 90 95

B C-terminal motifs

Acentrosomal/ Centrosomal/ cytoplamic side of SPB nucleus side of SPB

Fungi Metazoa Fungi and metazoa MASC CM2 PACT Acentrosomal/ CDK5RAP2, Fungi and ApsB, Spc72, cytoplamic side CM1 only myomegalin, - metazoa Mto1 of SPB Cnn Ascomycota CM1 + SPM Spc110, Pcp1 Centrosomal/ nucleus side of metazoa and PotenƟal CM1 + SPM- PCNT SPB Basidiomycota No clear CM1 or SPM AKAP9, D-PLP N-terminal motifs N-terminal Figure 10 - Lin et al.

A Late G1 Early S

J-TuSC oligomerization MTs Nucleus MT nucleation Duplicat plaque Nuclear membrane

Cytoplasm mSPBBridge mSPB dSPB mSPB dSPB

Late S M

Bipolar spindle

MT nucleation

Anaphase onset

MT nucleation inactivated

B Tub4 C Promote new template formation Tub4 Spc97 Spc98

Spc97 Spc98

5P SPM ? CM1 Mps1 Mps1 Spc110 Cdk1-Clb5 Cdk1-Clb5

Spc110

Early S

Late S and early M