Journ

alof

Cell

Scie

nce

Novel NEDD1 phosphorylation sites regulate c-tubulinbinding and mitotic spindle assembly

Maria Ana Gomez-Ferreria1, Mikhail Bashkurov1, Andreas O. Helbig1, Brett Larsen1, Tony Pawson1,Anne-Claude Gingras1,2 and Laurence Pelletier1,2,*1Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario M5G 1X5, Canada2Department of Molecular Genetics, University of Toronto, Toronto, Ontario M5S 1A8, Canada

*Author for correspondence (pelletier@lunenfeld.ca)

Accepted 18 April 2012Journal of Cell Science 125, 3745–3751� 2012. Published by The Company of Biologists Ltddoi: 10.1242/jcs.105130

SummaryDuring cell division, microtubules organize a bipolar spindle to drive accurate chromosome segregation to daughter cells. Microtubules arenucleated by the c-TuRC, a c-tubulin complex that acts as a template for microtubules with 13 protofilaments. Cells lacking c-TuRC corecomponents do nucleate microtubules; however, these polymers fail to form bipolar spindles. NEDD1 is a c-TuRC-interacting protein

whose depletion, although not affecting c-TuRC stability, causes spindle defects similar to the inhibition of its core subunits, including c-tubulin. Several residues of NEDD1 are phosphorylated in mitosis. However, previously identified phosphorylation sites only partiallyregulate NEDD1 function, as NEDD1 depletion has a much stronger phenotype than mutation of these residues. Using mass spectrometry,we have identified multiple novel phosphorylated sites in the serine (S)557–S574 region of NEDD1, close to its c-tubulin-binding domain.

Serine to alanine mutations in S565–S574 inhibit the binding of NEDD1 to c-tubulin and perturb NEDD1 mitotic function, yieldingmicrotubule organization defects equivalent to those observed in NEDD1-depleted cells. Interestingly, additional mutations in the S557–T560 region restore the capacity of NEDD1 to bind c-tubulin and promote bipolar spindle assembly. All together, our data suggest that the

NEDD1/c-tubulin interaction is finely tuned by multiple phosphorylation events in the S557–S574 region and is critical for spindleassembly. We also found that CEP192, a centrosomal protein similarly required for spindle formation, associates with NEDD1 andmodulates its mitotic phosphorylation. Thus CEP192 may regulate spindle assembly by modulating NEDD1 function.

Key words: NEDD1, Spindle, Phosphorylation, Y-tubulin, Mitosis, CEP192

IntroductionThe fidelity of chromosome segregation during cell division

relies on the formation of a microtubule-based bipolar spindle

(O’Connell and Khodjakov, 2007). Microtubules are polar

filaments nucleated by the c-TuRC, a ring-shaped complex

formed by c-tubulin associated with GCP-2, GCP-3, GCP-4,

GCP-5 and GCP-6 (Kollman et al., 2011). NEDD1 is a c-TuRC-

interacting protein that binds directly to c-tubulin (Manning et al.,

2010). Although NEDD1 is not necessary for the stability of the

c-TuRC, its depletion causes spindle assembly defects that

resemble the depletion of core c-TuRC components, where

microtubules fail to organize into bipolar spindles. Analyses of

mitotic and meiotic cells confirm the critical role of NEDD1 in

spindle formation in human, Xenopus and Drosophila (Haren

et al., 2006; Liu and Wiese, 2008; Luders et al., 2006; Ma et al.,

2010; Verollet et al., 2006). NEDD1 also associates with the

Augmin complex to allow c-tubulin recruitment to spindle

microtubules, thereby promoting microtubule amplification

(Goshima et al., 2008; Lawo et al., 2009). The NEDD1/

Augmin interaction depends on NEDD1 phosphorylation on

S411 (Johmura et al., 2011; Luders et al., 2006; Uehara et al.,

2009). Phosphorylation on T550 creates a priming site for PLK1,

which further phosphorylates NEDD1 to regulate microtubule

assembly in the vicinity of chromosomes (Zhang et al., 2009).

Similar to NEDD1 and c-tubulin, the centrosome protein

CEP192, plays a critical role in centrosome biogenesis and

spindle assembly. Depletion of CEP192 causes defects in

centriole duplication and pericentriolar material (PCM)

recruitment, thereby interfering with spindle pole organization

(Fig. 1A) (Gomez-Ferreria et al., 2007; Gomez-Ferreria and

Sharp, 2008; Zhu et al., 2008). Additionally, microtubules

nucleated near the chromosomes fail to reorganize into robust

bipolar spindles. CEP192 interacts with Aurora kinase A and

regulates its activation in Xenopus egg extracts (Joukov et al.,

2010). However, the mechanisms underlying the role of CEP192

in spindle formation in human cells remain poorly understood.

Here we show that CEP192 interacts with NEDD1 and

regulates its mitotic phosphorylation, which suggests that

CEP192 may control spindle assembly by modulating NEDD1

activity. Phosphorylation site mapping of mitotic NEDD1 reveals

that multiple residues in the serine (S)557–S574 region are

phosphorylated in vivo. This phosphorylation regulates the

NEDD1/c-tubulin interaction and is required for bipolar spindle

assembly.

Results and DiscussionCEP192 interacts with NEDD1 and modulates its mitoticphosphorylation

To unravel the molecular mechanisms underlying CEP192

function in spindle assembly, we performed affinity purification

followed by mass spectrometry (AP-MS). We found that CEP192

interacts with NEDD1, which was of particular interest since

Short Report 3745

Journ

alof

Cell

Scie

nce

depletion of either protein yields similar defects in mitotic

microtubule organization (Fig. 1A; supplementary material Fig.

S1) (Haren et al., 2006; Luders et al., 2006). Further analyses

showed that this interaction depends on the N-terminal region of

CEP192 (Fig. 1B).

In mitosis, NEDD1 displays a complex electrophoretic

pattern where slower migrating bands correspond to

differentially phosphorylated forms (Fig. 1C, bands b and c)

(Haren et al., 2009; Johmura et al., 2011). Interestingly,NEDD1 mitotic phosphorylation is altered in CEP192-

depleted cells, as levels of the hyperphosphorylated band are

consistently reduced (Fig. 1C, band c). These data suggest that

CEP192 may regulate spindle assembly by modulating NEDD1

phosphorylation in mitosis.

Mapping of NEDD1 phosphorylation sites in mitosis

We then mapped NEDD1 phosphorylated residues in mitotic

control and CEP192 depleted cells by mass spectrometry (MS;

Fig. 1D, upper panel; see also supplementary material Fig. S2;

Table S1). This analysis identified novel NEDD1 phosphorylation

sites and corroborated some previously described ones (Fig. 1D,

lower panel) (Johmura et al., 2011; Luders et al., 2006; Santamaria

et al., 2011; Zhang et al., 2009).

A quantitative analysis of the spectra was performed using the

MaxQuant software, which monitors abundance of phosphopeptides

by determining their ion intensity (supplementary material Fig. S3;

Tables S2, S3). Our results suggest that double phosphorylation of

the peptide 503–527 and phosphorylation at S468, S586, S460, S325

and S332 are modulated by CEP192 (peptides B–F in

supplementary material Fig. S3). The abundance of the doublephosphorylated peptide 555–570 (A) decreases to 15% after

CEP192 depletion. However, the low absolute ion intensity of

this peptide does not definitively support that this double

phosphorylation is in fact CEP192 dependent.

NEDD1 phosphorylation regulates spindle assembly

To investigate the role of global NEDD1 phosphorylation in

spindle assembly, we initially generated the mutant NEDD1-A23,

where the majority of phosphorylation sites identified in this or

previous studies were replaced with alanine (Ala; Fig. 1D,

underlined residues). We established a rescue assay where cells

are depleted of endogenous NEDD1 using RNA interference

Fig. 1. CEP192 interacts with NEDD1 and

modulates its mitotic phosphorylation.

(A) HeLa cells were transfected with esiRNA

targeting CEP192, NEDD1 or luciferase

(CT, negative control). Yellow circles indicate

arrays of microtubules not organized into a

bipolar structure, criteria that we used for

defining the phenotype of ‘disorganized

spindle’. Scale bar: 5 mm. (B) HEK293 cells

were transfected with CEP192 (1941 aa), or

deletion mutants encompassing the N-terminal

(aa 1–1058) or C-terminal (aa 979–1941)

region tagged with FLAG.

Immunoprecipitation and western blots were

performed with anti-FLAG and anti-NEDD1

antibodies. (C) HeLa cells were synchronized

with a double-thymidine block and 6 h after

release, arrested in mitosis with monastrol

(100 mM) or nocodazole (0.3 mM) for 5 h.

EsiRNA was transfected during the first

thymidine block. Extracts were analyzed using

7.5% 100:1 acrylamide:bisacrylamide gels.

AURKA levels are a loading control for

mitotic extracts. In mitotic NEDD1, slower

migrating bands correspond to differentially

phosphorylated forms (bands b and c; band a

corresponds to the faster migrating form/s of

NEDD1). (D) Diagram showing NEDD1

phosphorylation sites in the short isoform (660

aa). The upper part of the diagram shows the

sites we identified by Mascot. The lower part

shows phosphorylation sites previously

described (see text). Residues located in the

same tryptic peptide are grouped. Novel sites

identified in this study are indicated in red.

Sites functionally characterized in the literature

are in green. Residues mutated in the NEDD1-

A23 mutant (see text) are underlined.

Journal of Cell Science 125 (16)3746

Journ

alof

Cell

Scie

nce

(RNAi), and then transfected with GFP alone or RNAi-resistant

NEDD1, either wild-type or mutant, fused to GFP (Fig. 2A). In

GFP transfected cells, 90% of spindles are disorganized with

dispersed PCM or small, while most of the cells transfected with

wild-type NEDD1 have bipolar spindles (Fig. 2B,C). Cells

expressing a mutant where Ala substitutes S411 and T550

(A411-550) show fragmented centrosomes and, as previously

reported, defects in microtubule nucleation within the spindle,

resembling depletion of Augmin subunits (HAUS) (Luders et al.,

2006; Zhang et al., 2009). This ‘HAUS-like’ phenotype is caused

by the inability of this mutant to associate with the Augmin

complex (Lawo et al., 2009; Uehara et al., 2009). By contrast,

cells expressing the NEDD1-A23 mutant have more drastic

spindle defects where 70% are disorganized or abnormally small

with fragmented centrosomes.

We next analyzed the role of these residues in regulating the

well-established oligomerization properties of NEDD1 and its

ability to bind c-tubulin (Haren et al., 2006; Luders et al., 2006;

Manning et al., 2010). Immunoprecipitation assays show that

both GFP–NEDD1 and A411-550 mutant interact with c-tubulin

and endogenous NEDD1; however, mutations in NEDD1-A23

mutant clearly affect these interactions (Fig. 2D). Therefore, the

additional substitutions in NEDD1-A23 impair NEDD1 function

and ability to bind c-tubulin to a greater extent than mutation of

the previously characterized sites S411 and T550 (Zhang et al.,

2009).

Phosphorylation in the S557–S574 region regulates

spindle assembly

We then determined the function of specific NEDD1

phosphorylation sites, being initially interested in those whose

phosphorylation is reduced in CEP192-depleted cells (supplementary

material Fig. S3, peptides B–F). Mutation of S460, S468, S325,

S332, S516 or S586 to Ala has no significant effect on the ability of

NEDD1 to rescue bipolar spindle assembly in cells depleted of

endogenous NEDD1 (supplementary material Fig. S4, and data not

shown). Similarly, mutation of S493 and S516 (peptides L and M),

whose phosphorylation is higher after CEP192 depletion, to aspartic

Fig. 2. NEDD1 phosphorylation regulates

bipolar spindle assembly. (A–C) RNAi-rescue

experiments. (A) Western blot with anti-

NEDD1 antibodies detects endogenous NEDD1

and expression of the GFP–NEDD1 proteins.

(B) Representative spindles phenotypes. Scale

bar: 5 mm. (C) Quantification of the phenotypes

shown in B. Cells expressing similar levels of

GFP signal were scored. (D) Proteins expressed

in HEK293 cells were immunoprecipitated with

anti-GFP antibodies and immunoblotted.

NEDD1 regulation in spindle assembly 3747

Journ

alof

Cell

Scie

nce

and glutamic acid respectively, is compatible with NEDD1 activity

(data not shown). However, phosphorylation of these sites could

control other aspects of NEDD1 function, which may therefore

be modulated by CEP192. Additionally, CEP192 may regulate

microtubule organization by modulating NEDD1 phosphorylation at

yet unidentified residues.

Interestingly, our MS analysis shows that the region [amino

acids (aa)] 555–579 proximal to the c-tubulin-binding domain of

NEDD1 (aa 599–660) is phosphorylated on multiple residues in

vivo (Fig. 3A; supplementary material Fig. S2B, Fig. S3A,

peptides G–H). While the tryptic peptide 571–579 includes

a single phosphorylation site at S574, the peptide 555–570

includes six Ser/Thr residues. Several spectra identify single

phosphorylations at residues S565, S566 and S568 and a lower

number shows phosphorylation at S557, S558 and T560; indeed

phosphorylation of S557 and S558 was previously reported

(Santamaria et al., 2011). We also detect double phosphorylated

forms of this peptide, although with a very low intensity

(supplementary material Fig. S3A, peptide A). Sequence

alignment of this region shows that these phosphoresidues are

conserved in different species (Fig. 3A).

In order to study the functional relevance of this region we

started by mutating stretches of phosphorylated residues located

in close proximity, hypothesizing that in hyperphosphorylated

proteins, like NEDD1, functional regulation may depend on

multiple phosphorylations located in a particular region rather

than on unique sites. Using the RNAi-rescue system described

earlier, we observed that Ser to Ala mutation in the S565–S574

region abrogates NEDD1 function, as 80% of spindles are either

small or disorganized [Fig. 3B,C, mutant A4 (A565-566-568-

574)]. Single substitution to Ala of S565, S566, S568 or

S574 does not significantly interfere with NEDD1 function

(supplementary material Fig. S5A), while mutation of S565-

S566-S568 (mutant A3.b) has only a minor effect (Fig. 3C). Ser

Fig. 3. Phosphorylation in the S557–S574

region regulates bipolar spindle assembly.

(A) Sequence alignment using ClustalW.

Dotted lines show the phosphorylation sites

modified in the indicated mutants.

(B–E) RNAi-rescue experiments.

(B) Representative cells with bipolar, small or

disorganized spindles. Scale bar: 5 mm.

(C,D) Quantification of the spindle

phenotypes. (E) Ratio of GFP signals in the

centrosome/cytosol. Signal intensity was

determined as detailed in the Materials

and Methods.

Journal of Cell Science 125 (16)3748

Journ

alof

Cell

Scie

nce

to Glu mutations in the S565–S574 region are compatible with

NEDD1 function in mitotic spindle assembly, further supportingthat these residues can be phosphorylated in vivo [Fig. 3D,phosphomimetic mutant E4 (E565-566-568-574)]. Interestingly,

although mutation to Ala of S557-S558-T560 in the wild-typeprotein does not affect NEDD1 activity [mutant A3.a (A557-558-560)], these mutations in the A4 (A565-566-568-574) mutantbackground [mutant A7 (A557-558-560-565-566-568-574)]

restore NEDD1 function (Fig. 3A,C). Further mutational analysis ofthe region S557–T560 revealed that mutation to Ala of T560individually or the combination of S557-S558 partially restores A4

(A565-566-568-574) function (supplementary material Fig. S6A,mutants A560+A4 and A557-558+A4). This suggests thatphosphorylation in the S557–T560 region has a negative effect on

NEDD1 function: in mutant A3.a (A557-558-560), phosphorylationof the S565–S574 region may promote NEDD1 activity, while inmutant A4 (A565-566-568-574), phosphorylation of the S557–T560

region may inhibit NEDD1 function. In wild-type NEDD1,phosphorylation at S565–S574 would counter the negative effect ofphosphorylation at S557–T560.

Interestingly, the drastic spindle defects observed with the

mutant A4 (A565-566-568-574) are comparable to thoseobserved when S636-Y637, in the c-tubulin-binding domain ofNEDD1, are replaced with Ala (Fig. 3B,C, mutant A636-637).

Mutation of these two residues is reported to inhibit NEDD1/c-tubulin interaction (Manning et al., 2010). We notice thatmutants A4 (A565-566-568-574) and A636-637 show a weaklocalization to the centrosome (Fig. 3B,E). Consistently, these

mutants do not rescue the localization of c-tubulin tocentrosomes upon NEDD1 depletion (supplementary materialFig. S7).

Phosphorylation in the S557–S574 region regulatesc-tubulin binding

Analysis of the capacity of these mutants to bind c-tubulin and

endogenous NEDD1 show that Ser to Ala mutation of the S565–S574 region [mutant A4 (A565-566-568-574)] inhibits theseinteractions. Single mutation to Ala of S565, S566 or S574 does

not have a significant effect on c-tubulin binding, and only apartial defect is observed for S568 (supplementary material Fig.S5B,C). Mutation of S557-S558-T560 in the A4 (A565-566-568-574) mutant background [A7 (A557-558-560-565-566-568-574)]

partially rescues NEDD1/c-tubulin interaction, which couldexplain the ability of mutant A7 to assemble bipolar spindles(Fig. 4A, Fig. 3C). Consistently with its capacity to form bipolar

spindles, the phosphomimetic mutant E4 (E565-566-568-574)interacts with c-tubulin and endogenous NEDD1 (Fig. 3D,Fig. 4B). As previously described, mutant A636-637 still forms

oligomers but does not interact with c-tubulin (Fig. 4A)(Manning et al., 2010). Similarly, the mutant E636-637, whichis unable to rescue the formation of bipolar spindles, does not

interact with c-tubulin (Fig. 3D, Fig. 4B).

Mutant A7 (A557-558-560-565-566-568-574), althoughunable to interact with endogenous NEDD1, forms bipolarspindles, suggesting that NEDD1 oligomerization is not

required for spindle assembly (Fig. 4A, Fig. 3C). Nevertheless,oligomerization is predicted to be a common feature in more than35% of the proteins in the cell to better support function (Ali and

Imperiali, 2005). Therefore, it is possible that in a morephysiological condition, NEDD1 oligomerization positivelycontributes to spindle assembly.

Mutation to Ala of S565-S566-S568 [A3.b (A565-566-568)],although shows a minor effect in the RNAi-resistant rescue

assays, clearly affects c-tubulin binding (Fig. 3C, Fig. 4A). Thismutant still retains some capacity to bind c-tubulin, as comparedto mutant A4 (A565-566-568-574) and A636-637 (Fig. 4A;supplementary material Fig. S8), which do not rescue bipolar

spindle assembly. Thus, it is possible that A3.b (A565-566-568)residual capacity to bind c-tubulin is enough in the appropriatecellular context to form bipolar spindles. Additionally, it is

conceivable that the ability of A3.b (A565-566-568) to rescuespindle assembly, in contrast to A4 (A565-566-568-574), is dueto its partial capacity to bind endogenous NEDD1. In rescue

assays, depletion of endogenous NEDD1 is unlikely to becomplete. In this scenario, A3.b (A565-566-568), by interactingwith the remnant endogenous NEDD1, could form chimericoligomers wild-type/mutant NEDD1 that bind c-tubulin and

promote bipolar spindle assembly.

Taken together, these data suggest that the critical function ofNEDD1 in spindle assembly depends on its ability to bind c-

tubulin and that this interaction is regulated by phosphorylationin multiple residues at S557–S574 region. The fact that mutationsat S557–T560 and S565–S574 have different effects, suggests

that NEDD1 activity may be finely tuned through multiplephosphorylation events in the S557–S574 region. Furthermore, arecent proteomic screen reports NEDD1 ubiquitylation at

K570 (Kim et al., 2011), thus a complex interplay betweenubiquitylation and phosphorylation in this region may regulateNEDD1/c-tubulin interaction.

We propose a model where NEDD1 phosphorylation in the

region S565–S574, close to the c-tubulin-binding domain,exposes this domain for interaction with c-tubulin. This bindingmay activate the c-TuRC to nucleate microtubules capable of

assembling a bipolar spindle (Fig. 4C). Structural analyses of thec-TuRC suggest that this complex is assembled in an ‘off’ statethat requires activation to efficiently nucleate microtubules

(Guillet et al., 2011; Kollman et al., 2011). In vivo,microtubules can be generated in absence of NEDD1 and c-tubulin (Gomez-Ferreria et al., 2007; Haren et al., 2006);however, why these microtubules are unable to build a bipolar

spindle is unclear. It is possible that they are uncapped andtherefore highly prone to depolymerization (Wiese and Zheng,2000). Moreover, they may lack the 13-fold symmetry

established when the c-TuRC acts as a template for nucleation(Kollman et al., 2011). In this scenario, MAPs (microtubuleassociated proteins) and molecular motors (dynein and kinesins)

could be unable to crosslink and slide these unstable and/orasymmetric polymers to organize a functional bipolar spindle(Walczak and Heald, 2008). By contrast, phosphorylation at

S411 would only regulate the Augmin-dependent microtubuleamplification pathway (Fig. 4C) (Goshima et al., 2008; Lawoet al., 2009; Uehara et al., 2009). Finally, we notice that thismodel could explain the role of NEDD1 in spindle assembly in

systems where NEDD1 does not determine c-tubulin recruitmentto centrosomes, as in meiotic cells, Xenopus and Drosophila (Liuand Wiese, 2008; Ma et al., 2010; Verollet et al., 2006). We

cannot discard, however, that NEDD1 regulates spindle assemblyby yet unidentified c-tubulin independent pathways.

Given the phenotypic similarities between CEP192 depletion

and Ser to Ala mutations in the S565–S574 region, we expectedthat phosphorylation at these sites depends on CEP192. However,our semi-quantitative phosphoproteomic data suggest that CEP192

NEDD1 regulation in spindle assembly 3749

Journ

alof

Cell

Scie

nce

has, at best, a modulating influence on the phosphorylation of this

region (supplementary material Fig. S3, peptides A, G and H).

CEP192 may still positively regulate the NEDD1/c-tubulin

interaction by favoring phosphorylation towards the S565–S574

instead of S557–T560 region. Therefore, more detailed

phosphoproteomic studies and the use of phospho-specific

antibodies will be required to test this hypothesis. Nevertheless,

the physical interaction we describe between NEDD1 and

CEP192, the similar phenotypes observed upon depletion of

either protein, the change in the electrophoretic motility of NEDD1

upon CEP192 depletion and the different phosphorylation status of

NEDD1 in absence of CEP192 together provide overlapping lines

of evidence that suggest an interplay between CEP192 and

NEDD1 phosphorylation during mitotic spindle assembly.

Although important questions regarding the mechanism

underlying the role of NEDD1 and c-tubulin in mitotic spindle

assembly remain to be answered, we now show that the

interaction between these two proteins regulates this process

and can be spatially and temporally controlled by

phosphorylation in the S557–S574 region. Additionally, our

data suggest that CEP192 may exert its critical role in bipolar

spindle assembly by regulating NEDD1 phosphorylation.

Materials and MethodscDNA cloning

CEP192 cDNA [1941 amino acids (Gomez-Ferreria et al., 2007)] or the differentdeletion mutants were cloned into pcDNA3 fused with FLAG. NEDD1 cDNAMGC:26881 (Open Biosystems, BC027605) was cloned into pcDNA3 fused toGFP. The cDNA sequence encodes a 667-amino-acid protein which corresponds tothe long isoform of NEDD1. The mutated phosphorylated residues are namedaccording to their position in the short isoform of NEDD1 (660 aa).

RNA silencing

EsiRNA (endoribonuclease-prepared siRNA) was generated as described previously(Kittler et al., 2005). The targeted regions are encompassed by the following primers:CEP192 (NM_032142): 59-TTTTCAAGGGCTAGTATGTCTGA-39, 59-GGATGTT-ATTCTGGGGTTCCT-39; NEDD1 39UTR (NM_001135175): 59-TAAAAATTGT-ACAGTATGTCATCTACCCAATAG-39, 59-CAAAAATTGTCTTATCAAATGTA-CAATAAATAATA-39; non-targeting esiRNA (AY_015988, luciferase): 59-TGGTTT-GGTTGTTGATGGAA-39, 59-GTGCCTGGTGAAACTTGGTT-39.

NEDD1 phosphorylation sites mapping

A clonal cell line expressing FLAG–NEDD1 was generated with the Flp-In T-RExsystem (Invitrogen) in U2OS cells following the protocol from the manufacturer

Fig. 4. Phosphorylation in S557–S574

region regulates NEDD1/NEDD1 and

NEDD1/c-tubulin interactions.

(A,B) Proteins expressed in HEK293 cells

were immunoprecipitated with anti-GFP

antibodies and immunoblotted with anti-

NEDD1 and anti-c-tubulin antibodies. The

upper panel shows a shorter exposure of the

GFP–NEDD1 proteins. Numbers under the

western blots refer to the band intensity of

the immunoprecipitated protein normalized

to the intensity of the bait. Quantifications

were done using ImageJ analysis software.

Values are the average of at least three

independent experiments. Standard

deviations are included in supplementary

material Fig. S8. (C) Model explaining the

role of NEDD1 phosphorylation in spindle

assembly. (I) Phosphorylation at S565–S574

is required to expose the c-tubulin-binding

domain. The NEDD1/c-tubulin interaction

activates the c-TuRC. Microtubules

nucleated using the c-TuRC as a template

are able to build a bipolar spindle. (II) Ser to

Ala mutations in S565–S574 inhibit

the interaction NEDD1/c-tubulin and the c-

TuRC is inactive. The resulting microtubules

do not form bipolar spindles. (III) Mutation

of S411 only affects the Augmin-dependent

microtubule amplification pathway. The

model shows an inactive form of NEDD1

where intramolecular interactions bury its

c-tubulin-binding domain. However, it is

conceivable that in the inactive state of

NEDD1, an additional protein could block

the NEDD1/c-tubulin interaction by

competing with c-tubulin for its binding

domain. Phosphorylation at S565–S574 also

controls NEDD1 oligomerization. Our data

suggest that NEDD1 oligomerization is not

required for spindle assembly, however it

could affect this process in more

physiological systems.

Journal of Cell Science 125 (16)3750

Journ

alof

Cell

Scie

nce

(Malecki et al., 2006). Cells were synchronized in mitosis (supplementary materialFig. S2A). Lysis was performed using a buffer with 10 mM Tris pH 7.4, 100 mMNaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 10% glycerol, 0.1% SDS,0.5% deoxycholate, 10 mM NaF, 50 mM b-glycerophosphate, 5 nM okadaic acid,5 nM calyculin A, 1 mM DTT and protease inhibitors. FLAG–NEDD1 wasimmunoprecipitated with anti-FLAG M2 magnetic beads (Sigma, M8823) anddigested with trypsin. Phosphopeptide enrichment was performed using Ga (III)resin (SwellGel Gallium-Chelated Disc; Pierce, 89853). Peptides were loaded onZorbax C18 (Agilent ZorbaxSB, 3.5 mm) and analyzed by LC-MS/MS using aThermoFinnigan Orbitrap. Files were searched with Mascot v2.2 againstRefSeqV42. For the quantitative analysis of the data, the spectra correspondingto three experiments for control and CEP192 esiRNA-treated cells were analyzedwith MaxQuant software v1.1.1.36 (Cox and Mann, 2008). False discovery ratewas set to 1% and a retention time alignment window of 5 min was utilized. MS/MS tolerance was set to 0.5 Da. Ion intensities for oxidated, deamidated andmis-cleaved peptides were summed. The values shown in supplementary materialFig. S3A correspond to the raw ion intensities arbitrarily divided by 106.

Microscopy and automatic quantification of signal intensitiesThree-dimensional images were acquired on a DeltaVision Core System (AppliedPrecision) equipped with an IX71 microscope (Olympus), a CCD camera(CoolSNAP HQ2 102461024; Roper Scientific) and 606/1.42 NA objective(Olympus). Z-stacks (0.4 mm apart) were collected, deconvolved using theSoftWorx v4.0 (Applied Precision) and shown as maximum intensity projections.Automated analysis of fluorescence intensities was performed on 12-bit TIFFimages using Acapella v2.18 (Perkin Elmer). Cellular and centrosomal masks weredetected in the CEP192 channel using an adaptive threshold to specifically outlinecytosolic and centrosomal regions. Fluorescence intensity of GFP–NEDD1 andc-tubulin was analyzed using the detected masks.

Rescue experimentsHeLa cells were transfected with 0.4 mg of esiRNA targeting NEDD1 39UTR. 24 hlater, 1 mg of plasmids expressing wild-type NEDD1 or the different mutants weretransfected for 36 h. For immunofluorescence, cells were processed as previouslydescribed (Lawo et al., 2009). For the spindle phenotypes quantification we showthe average of at least three experiments where we counted 50 cells per condition.Error bars refer to standard deviation (s.d.).

Immunoprecipitation and western blottingHEK293 cells were transfected with FLAG- or GFP-fused proteins for 36 h. Forimmunoprecipitation we used antibodies against FLAG (Sigma, A2220), GFP(kindly provided by D. Drechsel) and NEDD1 (Abcam, ab57336). For blotting weused antibodies against FLAG (Sigma, F7425), NEDD1, c-tubulin (Sigma,T6557), AURKA (Abcam, ab13824), CEP192 (Zhu et al., 2008) or a-tubulin(Sigma, T9026).

AcknowledgementsWe thank C. Yeh, S. Lawo and J. Goncalves for critical reading ofthe manuscript, and C. Holley and A. Tagliaferro for esiRNAproduction.

FundingThis work was funded by the Canadian Cancer Society [grantnumber 019562 to L.P.]; Fundacion Caja Madrid [to M.G.]. L.P.holds a Canada Research Chair in Centrosome Biogenesis andFunction.

Supplementary material available online at

http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.105130/-/DC1

ReferencesAli, M. H. and Imperiali, B. (2005). Protein oligomerization: how and why. Bioorg.

Med. Chem. 13, 5013-5020.Cox, J. and Mann, M. (2008). MaxQuant enables high peptide identification rates,

individualized p.p.b.-range mass accuracies and proteome-wide protein quantification.Nat. Biotechnol. 26, 1367-1372.

Gomez-Ferreria, M. A. and Sharp, D. J. (2008). Cep192 and the generation of themitotic spindle. Cell Cycle 7, 1507-1510.

Gomez-Ferreria, M. A., Rath, U., Buster, D. W., Chanda, S. K., Caldwell, J. S.,Rines, D. R. and Sharp, D. J. (2007). Human Cep192 is required for mitoticcentrosome and spindle assembly. Curr. Biol. 17, 1960-1966.

Goshima, G., Mayer, M., Zhang, N., Stuurman, N. and Vale, R. D. (2008). Augmin: aprotein complex required for centrosome-independent microtubule generation withinthe spindle. J. Cell Biol. 181, 421-429.

Guillet, V., Knibiehler, M., Gregory-Pauron, L., Remy, M. H., Chemin, C.,Raynaud-Messina, B., Bon, C., Kollman, J. M., Agard, D. A., Merdes, A. et al.

(2011). Crystal structure of c-tubulin complex protein GCP4 provides insight intomicrotubule nucleation. Nat. Struct. Mol. Biol. 18, 915-919.

Haren, L., Remy, M. H., Bazin, I., Callebaut, I., Wright, M. and Merdes, A. (2006).NEDD1-dependent recruitment of the gamma-tubulin ring complex to the centrosomeis necessary for centriole duplication and spindle assembly. J. Cell Biol. 172,505-515.

Haren, L., Stearns, T. and Luders, J. (2009). Plk1-dependent recruitment of gamma-tubulin complexes to mitotic centrosomes involves multiple PCM components. PLoS

ONE 4, e5976.

Johmura, Y., Soung, N. K., Park, J. E., Yu, L. R., Zhou, M., Bang, J. K., Kim, B. Y.,

Veenstra, T. D., Erikson, R. L. and Lee, K. S. (2011). Regulation of microtubule-based microtubule nucleation by mammalian polo-like kinase 1. Proc. Natl. Acad.

Sci. USA 108, 11446-11451.

Joukov, V., De Nicolo, A., Rodriguez, A., Walter, J. C. and Livingston, D. M. (2010).Centrosomal protein of 192 kDa (Cep192) promotes centrosome-driven spindleassembly by engaging in organelle-specific Aurora A activation. Proc. Natl. Acad.

Sci. USA 107, 21022-21027.

Kim, W., Bennett, E. J., Huttlin, E. L., Guo, A., Li, J., Possemato, A., Sowa, M. E.,

Rad, R., Rush, J., Comb, M. J. et al. (2011). Systematic and quantitative assessmentof the ubiquitin-modified proteome. Mol. Cell 44, 325-340.

Kittler, R., Heninger, A. K., Franke, K., Habermann, B. and Buchholz, F. (2005).Production of endoribonuclease-prepared short interfering RNAs for gene silencing inmammalian cells. Nat. Methods 2, 779-784.

Kollman, J. M., Merdes, A., Mourey, L. and Agard, D. A. (2011). Microtubulenucleation by c-tubulin complexes. Nat. Rev. Mol. Cell Biol. 12, 709-721.

Lawo, S., Bashkurov, M., Mullin, M., Ferreria, M. G., Kittler, R., Habermann, B.,Tagliaferro, A., Poser, I., Hutchins, J. R., Hegemann, B. et al. (2009). HAUS, the8-subunit human Augmin complex, regulates centrosome and spindle integrity. Curr.

Biol. 19, 816-826.

Liu, L. and Wiese, C. (2008). Xenopus NEDD1 is required for microtubule organizationin Xenopus egg extracts. J. Cell Sci. 121, 578-589.

Luders, J., Patel, U. K. and Stearns, T. (2006). GCP-WD is a gamma-tubulin targetingfactor required for centrosomal and chromatin-mediated microtubule nucleation. Nat.

Cell Biol. 8, 137-147.

Ma, W., Baumann, C. and Viveiros, M. M. (2010). NEDD1 is crucial for meioticspindle stability and accurate chromosome segregation in mammalian oocytes. Dev.

Biol. 339, 439-450.

Malecki, M. J., Sanchez-Irizarry, C., Mitchell, J. L., Histen, G., Xu, M. L., Aster,J. C. and Blacklow, S. C. (2006). Leukemia-associated mutations within theNOTCH1 heterodimerization domain fall into at least two distinct mechanisticclasses. Mol. Cell. Biol. 26, 4642-4651.

Manning, J. A., Shalini, S., Risk, J. M., Day, C. L. and Kumar, S. (2010). A directinteraction with NEDD1 regulates gamma-tubulin recruitment to the centrosome.PLoS ONE 5, e9618.

O’Connell, C. B. and Khodjakov, A. L. (2007). Cooperative mechanisms of mitoticspindle formation. J. Cell Sci. 120, 1717-1722.

Santamaria, A., Wang, B., Elowe, S., Malik, R., Zhang, F., Bauer, M., Schmidt, A.,

Sillje, H. H., Korner, R. and Nigg, E. A. (2011). The Plk1-dependentphosphoproteome of the early mitotic spindle. Mol. Cell. Proteomics 10,M110.004457.

Uehara, R., Nozawa, R. S., Tomioka, A., Petry, S., Vale, R. D., Obuse, C. and

Goshima, G. (2009). The augmin complex plays a critical role in spindle microtubulegeneration for mitotic progression and cytokinesis in human cells. Proc. Natl. Acad.

Sci. USA 106, 6998-7003.

Verollet, C., Colombie, N., Daubon, T., Bourbon, H. M., Wright, M. and Raynaud-Messina, B. (2006). Drosophila melanogaster gamma-TuRC is dispensable fortargeting gamma-tubulin to the centrosome and microtubule nucleation. J. Cell Biol.

172, 517-528.

Walczak, C. E. and Heald, R. (2008). Mechanisms of mitotic spindle assembly andfunction. Int. Rev. Cytol. 265, 111-158.

Wiese, C. and Zheng, Y. (2000). A new function for the gamma-tubulin ring complexas a microtubule minus-end cap. Nat. Cell Biol. 2, 358-364.

Zhang, X., Chen, Q., Feng, J., Hou, J., Yang, F., Liu, J., Jiang, Q. and Zhang,

C. (2009). Sequential phosphorylation of Nedd1 by Cdk1 and Plk1 is required fortargeting of the gammaTuRC to the centrosome. J. Cell Sci. 122, 2240-2251.

Zhu, F., Lawo, S., Bird, A., Pinchev, D., Ralph, A., Richter, C., Muller-Reichert, T.,

Kittler, R., Hyman, A. A. and Pelletier, L. (2008). The mammalian SPD-2 orthologCep192 regulates centrosome biogenesis. Curr. Biol. 18, 136-141.

NEDD1 regulation in spindle assembly 3751