Developmental Cell Short Article - core.ac.uk · 2Dipartimento di Genetica e Biologia Molecolare and IBPM/CNR, Sapienza ... Sloan-Kettering Institute, Memorial Sloan-Kettering Cancer

Developmental Cell

Short Article

Morgana/chp-1, a ROCK Inhibitor Involvedin Centrosome Duplication and TumorigenesisRoberta Ferretti,1,6 Valeria Palumbo,2,6 Augusta Di Savino,1 Silvia Velasco,1 Mauro Sbroggio,1 Paolo Sportoletti,3

Lucia Micale,4 Emilia Turco,1 Lorenzo Silengo,1 Gioacchino Palumbo,4 Emilio Hirsch,1 Julie Teruya-Feldstein,5

Silvia Bonaccorsi,2 Pier Paolo Pandolfi,1,3 Maurizio Gatti,2 Guido Tarone,1 and Mara Brancaccio1,*1Dipartimento di Genetica, Biologia e Biochimica, Universita di Torino, 10126 Torino, Italy2Dipartimento di Genetica e Biologia Molecolare and IBPM/CNR, Sapienza, Universita di Roma, 00185 Roma, Italy3Cancer Genetics Program, Beth Israel Deaconess Cancer Center, Department of Medicine and Pathology, Harvard Medical School, Boston,

MA 02215, USA4Dipartimento di Genetica e Microbiologia, Universita di Bari, Bari, Italy5Department of Pathology, Sloan-Kettering Institute, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA6These authors contributed equally to this work

*Correspondence: [email protected]

DOI 10.1016/j.devcel.2009.12.020

SUMMARY

Centrosome abnormalities lead to genomic insta-bility and are a common feature of many cancer cells.Here we show that mutations in morgana/chp-1result in centrosome amplification and lethality inboth Drosophila and mouse, and that the fly centro-some phenotype is fully rescued by the human ortho-log of morgana. In mouse cells, morgana formsa complex with Hsp90 and ROCK I and II, and directlybinds ROCK II. Morgana downregulation promotesthe interaction between ROCK II and nucleophosmin(NPM), leading to an increased ROCK II kinaseactivity, which results in centrosome amplification.morgana+/� primary cells and mice display an in-creased susceptibility to neoplastic transformation.In addition, tumor tissue array histochemical analysisrevealed that morgana is underexpressed in a largefraction of breast and lung human cancers. Thus,morgana/chp-1 appears to prevent both centrosomeamplification and tumorigenesis.

INTRODUCTION

Centrosomes are the major microtubule organizing centers in

most animal cells. They comprise a pair of centrioles surrounded

by pericentriolar material (PCM), which nucleates and organizes

microtubules (MTs), contributing to the formation of a bipolar

spindle (Bettencourt-Dias and Glover, 2007; Bornens, 2002). In

normal cells, centrosome duplication is tightly regulated and

occurs only once in every cell cycle. When this regulatory mech-

anism is eluded, multiple rounds of centrosome duplication

within the same cell cycle can result in diploid cells with supernu-

merary centrosomes, a phenomenon known as centrosome

amplification. Cells with supernumerary centrosomes can form

aberrant spindles that are unable to mediate proper chromo-

some segregation, leading to aneuploid and polyploid cells

that are thought to promote tumor progression (Fukasawa,

486 Developmental Cell 18, 486–495, March 16, 2010 ª2010 Elsevier

2007; Nigg, 2006; Ganem et al., 2009). Consistent with these

results, having supernumerary centrosomes is one of the most

common phenotypes associated with solid human tumors

(Fukasawa, 2007; Nigg, 2006). Recent work has shown that

centrosome amplification can lead to tumors also in Drosophila,

suggesting a conserved mechanism for centrosome-related

carcinogenesis (Basto et al., 2008).

Factors controlling proper centrosome duplication include

several proteins implicated in tumor onset and progression,

such as p53, the protein kinases PLK1, Aurora A, Nek2, ROCK

I and II, the polo-like kinases PLK2 and PLK4, the E7 oncoprotein

of the papilloma virus, nucleophosmin (NPM), and BRCA1

(Fukasawa, 2007; Nigg, 2006). In this report, we show that the

CHORD-containing (Shirasu et al., 1999) protein chp-1 regulates

centrosome duplication. Here, we rename chp-1 ‘‘morgana’’

to avoid confusion with the chromodomain protein Chp1

(NP_593666). While nonvertebrate genomes contain a single

chp-coding gene, vertebrates harbor two chp genes, morgana/

chp-1 and melusin (Brancaccio et al., 2003b). Melusin is exclu-

sively expressed in muscle and heart (Brancaccio et al., 1999)

and is involved in triggering cardiomyocyte hypertrophy in

response to mechanical stress (Brancaccio et al., 2003a; De

Acetis et al., 2005; Sbroggio et al., 2008). Morgana/chp-1 has

also been described as a stress protein (Hahn, 2005; Wu et al.,

2005), and its function, as an essential component of the innate

immune resistance to pathogens, has been investigated in detail

in barley and Arabidopsis (Shirasu et al., 1999). In this report, we

demonstrate that morgana/chp-1 is required for both Drosophila

and mouse development and to repress centrosome overdupli-

cation in both organisms. In addition, our results suggest that

morgana downregulation promotes carcinogenesis both in

mice and humans.

RESULTS

Morgana/chp-1 Is an Essential ProteinMorgana function was first investigated by targeted disruption

of the morgana gene in mouse (see Figures S1A and S1B avail-

able online). morgana+/� mice were fertile and did not exhibit

developmental problems or visible pathology. At 3.5 days pc,

Inc.

mailto:[email protected]

Figure 1. Morgana/chp-1 Is Required for Mouse Embryogenesis and Mitotic Division

(A) Offspring from morgana+/� intercrosses.

(B) morgana+/+ and morgana�/� blastocysts collected at 3.5 days of gestation and cultured in vitro for 3 days. Bars, 100 mm.

(C) The TUNEL assay reveals apoptotic cells in morgana�/� blastocysts cultured in vitro for 3 days. Bars, 100 mm.

(D) Western blots showing that the morgana level in morgana+/+ MEFs is approximately twice than in morgana+/�MEFs. Cells were at the second in vitro passage

(P2). Loading control, vinculin.

(E) morgana+/� MEFs (P2) exhibit higher frequencies of polyploid cells than wild-type MEFs (four MEF preparations per genotype; p < 0.005).

(F) Percentages of diploid cells with more than two centrosomes in morgana+/+ and morgana+/� MEFs at P2 (data from five preparations per genotype).

Only interphase cells with a nuclear DNA content not exceeding 4C were analyzed (see Experimental Procedures). MEFs were stained for g-tubulin (red), a-tubulin

(green), and DAPI (blue). Inserts show enlargements of the centrosomes. Bar, 10 mm.

(G and H) A centrosome duplication assay in HU-treated NIH 3T3 (G) and U2OS (H) cells. Cells were infected with a morgana RNAi construct (black column),

or a nonsilencing RNAi construct (gray column), or were noninfected (NT). Bars in graphs represent standard errors. (**p < 0.01; ***p < 0.001).

Developmental Cell

Role of Morgana in Centrosome Duplication and Cancer

morphologically normal morgana�/� embryos were present at

a normal Mendelian frequency, but after this stage embryos

were no longer detectable (Figure 1A). Strikingly, after 3 days

in culture, the inner cell mass of morgana�/� blastocysts iso-

lated from morgana+/� intercrosses stopped proliferating and

Develo

died by apoptosis (Figures 1B and 1C). Interestingly, tropho-

blastic cells, which undergo DNA endoreplication without cell

division, survived in the absence of the wild-type morgana

function (data not shown), suggesting a mitotic role for this

protein.

pmental Cell 18, 486–495, March 16, 2010 ª2010 Elsevier Inc. 487

Developmental Cell


To determine the possible mitotic function of morgana, we

took advantage of the Drosophila model system. In this

organism, embryos homozygous for a mitotic mutation usually

survive until late larval stages, exploiting the material stored

into the egg by their heterozygous mothers. The brains of these

larvae undergo progressive depletion of the maternal product,

allowing phenotypic characterization of the mitotic mutation

(Gatti and Goldberg, 1991). We have previously shown that the

Drosophila genome harbors a single gene encoding a CHORD-

containing protein (Brancaccio et al., 2003b): the CHORD/

CG6198 gene that here we rename morgana (mora). We identi-

fied a putative mora mutant allele, mora1, in the Exelixis’ collec-

tion of PiggyBac element insertions (Thibault et al., 2004).

Sequencing of mora1 revealed that it carries a PiggyBac insertion

in the second exon of CG6198 (Figure 2A). The mutant pheno-

type elicited by mora1 was rescued by a wild-type mora/

CG6198 transgene (see below).

While flies heterozygous for mora1 were fully viable and

morphologically normal, mora1 homozygotes died as third

instar larvae. Mutant larvae were devoid of imaginal discs

and displayed strongly reduced testes and salivary glands,

suggesting an underlying defect in cell proliferation (Gatti and

Goldberg, 1991). Examination of brain preparations from

mora1 mutant larvae revealed that in 95% of dividing cells

the chromosomes are elongated and irregularly condensed

(Figure 2B). In many of these cells, the chromosomes were

so poorly condensed that they were difficult to be recognized

as such. However, mitotic figures could be unambiguously

identified by immunostaining with an anti-phospho histone H3

antibody that marks mitotic chromatin (Wei et al., 1999)

(Figure 2C; Figure S2A). The mitotic index (MI) in mora1 mutant

brains (3.95%; n = 1579) was approximately twice the control

MI (2.07%; n = 3809) (see also Figure S2B), indicating that

mutant cells spend more time in mitosis than their wild-type

counterparts. Imaging of dividing larval neuroblasts expressing

the GFP-tagged histone H2AV (Clarkson and Saint, 1999) re-

vealed that wild-type cells (n = 10) complete cell division within

18 min of the chromosomes becoming detectable within the

nucleus (Figure 2D; Movie S1). In contrast, mora1 mutant cells

(n = 10) remained arrested in a prophase/prometaphase-like

state for the duration of imaging (60 min) (Figure 2D; Movie

S2). However, the analysis of fixed preparations revealed that

a few cells manage to escape the prophase/prometaphase

block. In control cells, the frequency of anaphases relative to

all mitotic figures was approximately 20% (n = 400). In mora1

mutant brains anaphase frequencies ranged from 0.5 to 1.6%

(n = 800), and all mutant anaphases displayed strong defects

in chromosome segregation (Figure 2E; Figure S2C). Examina-

tion of mitotic figures in which at least the heterochromatic

regions of the chromosomes were clearly recognizable (Fig-

ure 2B) revealed that 30% (n = 400) of mora neuroblasts

are polyploid; in control brains the frequency of polyploid

cells is approximately 0.2% (n = 800) (see also Figure S2D).

In addition, mora brains displayed 7.7% (n = 1828) apoptotic

cells, while the frequency of these cells in control brains

was only 0.5% (n = 1730) (Figure S2E and data not shown).

These results suggest that cells unable to undergo anaphase

either revert to interphase, becoming polyploid, or undergo

apoptosis.


morgana/chp-1 Mutations Cause CentrosomeAmplificationExamination of mora1 brain preparations stained for DNA,

tubulin, and the centrosomal marker centrosomin (Cnn; Megraw

et al., 1999) revealed a striking centrosome amplification pheno-

type (Figure 2E; Figure S2F). Cells that were unambiguously

diploid (n = 57) displayed a number of centrosomes ranging

from 2 to 18, with an average of 5.9 centrosomes per cell. Immu-

nostaining for both g-tubulin and the centriole marker DSas-4

(Basto et al., 2006) showed that the two proteins colocalize

(Figure S2G). This suggests that the supernumerary centro-

somes do not result from PCM fragmentation but are instead

generated by multiple rounds of centrosome duplication.

Drosophila Mora shows 42% identity and 63% similarity with

its human ortholog, and it exhibits a similar domain organization

(Figure 2A). We constructed flies homozygous for mora1 and

simultaneously expressing a human morgana transgene (Fig-

ure 2F). Third instar larvae of this genetic constitution showed

imaginal discs and testes of normal appearance, and most

mitotic divisions observed in their brains were completely normal

(Figure 2G). For example, none of the observed mitotic figures

was polyploid and 98% of dividing cells displayed two centro-

somes (n = 200). However, neither the human morgana nor the

Drosophila mora+ transgene rescued the lethal phenotype of

mora1 mutants, which died during metamorphosis (data not

shown). This may be attributable to the fact that both the human

and the Drosophila transgenes were expressed under the control

of an actin promoter rather than the endogenous promoter.

Thus, we conclude that human morgana fully rescues the mitotic

phenotype of Drosophila mora mutants; our results do not permit

us to assess whether the human gene can substitute its fly ortho-

log during metamorphosis.

We next analyzed cultured mouse embryonic fibroblasts

(MEFs) from morgana+/� mice to disclose possible phenotypes

caused by haploinsufficiency. Using an antibody raised against

morgana, we found that heterozygous MEFs display the ex-

pected reduction of morgana to 50% of the wild-type level

(Figure 1D). Analysis of metaphase spreads revealed that the

frequency of polyploid metaphases was significantly higher in

morgana+/� than in morgana+/+ MEFs (Figure 1E). In interphase

cells with a nuclear DNA content not exceeding 4C (G1, S, and

G2 nuclei of diploid cells, and G1 nuclei of tetraploid cells), im-

munostaining for g-tubulin revealed that the frequency of cells

with supernumerary centrosomes (R3 centrosomes/cell) is

significantly higher in morgana+/� than in morgana+/+ MEFs

(9.3% versus 3.5%; p < 0.005) (Figure 1F).

To determine whether the supernumerary g-tubulin signals

observed in morgana+/� MEFs are centriole-containing centro-

somes or acentriolar PCM aggregates, we immunostained these

cells for centriolar markers such as centrin, Nap-1, or Sas-6.

Centrin is enriched in the distal lumen of all centrioles (Paoletti

et al., 1996), Nap1 associates with the free proximal ends of

both engaged and disengaged centrioles (Mayor et al., 2000),

while Sas-6 marks the growing procentriole (Strnad et al.,

2007). In cells containing multiple g-tubulin signals, all of them

colocalized with both centrin (Figures S1C–S1E) and Nap1

(Figures S1F–S1H). In addition, when Sas-6 signals were clearly

visible, they invariably colocalized with those of g-tubulin

(Figure S1I–S1M). These results indicate that all supernumerary

Inc.

Figure 2. Mutations in mora Cause Mitotic Defects in Drosophila Neuroblasts

(A) Organization of the Drosophila (D-Mora) and human morgana (H-Mora) proteins. The vertical line indicates the position of the PiggyBac element. Numbers

indicate percentages of amino acid identity. CHORD: cysteine and histidine rich domain; CS: CHORD and Sgt1 domain.

(B) DAPI-stained chromosomes from mora1 (mora) and wild-type larval brains. Mutant chromosomes are irregularly condensed.

(C) Larval brain mitotic figures stained for both DNA (DAPI) and Phospho-histone H3.

(D) Mitotic division in live larval neuroblasts (NBs) expressing GFP-histone H2AV. Numbers indicate the minutes elapsed from the beginning of imaging.

(E) mora1 NBs stained for DNA (blue), centrosomin (Cnn, red), and a-tubulin (a-tub, green) showing supernumerary centrosomes and disorganized spindles.

(F) An anti-H-morgana antibody recognizes a band in extracts from mora1 homozygotes carrying the human morgana transgene (trsg), but not in extracts from

either wild-type or mora1/mora1 flies. Loading control, a-tubulin (a-tub).

(G) mora1/mora1 larval NBs expressing the human morgana transgene exhibit normal mitotic figures; centrosomin (Cnn, red), a-tubulin (green), and DNA (blue).

Bars, 5 mm in all panels.

Developmental Cell


Developmental Cell 18, 486–495, March 16, 2010 ª2010 Elsevier Inc. 489

Figure 3. Morgana/chp-1 Inhibits ROCK Kinase Activity

(A) 293 cell lysates precipitated with an anti-morgana antibody or unrelated IgG (Ig), resolved by SDS-PAGE, and stained with Coomassie blue. The bands specif-

ically present in the morgana lane were identified as ROCK I, ROCK II, and Hsp90 by mass spectrometry.

(B) 293 cell lysates precipitated with anti-morgana or control Ig, blotted and stained with antibodies against ROCK I, ROCK II, Hsp90, citron kinase, and morgana.

TE, total extract.

(C) Lysates from 293 cells transfected with myc-Rock I and myc-Rock II, precipitated with an anti-Myc antibody or control Ig, and blotted and stained with specific

antibodies. TE, total extract.

(D) Western blots of extracts from morgana+/+ or morgana+/� MEFs at P2 (three preparations per genotype) immunostained for morgana, phospho-MLC2

(P-MLC2), total MLC2, and vinculin as loading control. morgana+/+ and morgana+/� MEFs exhibit identical levels of ROCK I and ROCK II (data not shown).

The graph shows the average intensity of the P-MLC2 bands normalized on total MLC2 (data from eight preparations per genotype).

(E) In morgana+/� MEFs, ROCK II kinase activity is higher than in morgana+/+ MEFs. ROCK II was immunoprecipitated from MEF extracts; ROCK II, phospho-

MYPT1 (P-MYPT1), and total MYPT1 were detected by western blotting with specific antibodies. Y-27632 is a selective ROCK inhibitor. The graph shows the

average intensity of the P-MYPT1 bands normalized on total MYPT1 (data from four preparations per genotype).

(F) Treatment with 2 mM Y-27632 lowers the ROCK kinase activity of morgana+/� MEFs to the wild-type level.

(G) A 48 hr treatment with 2 mM Y-27632 rescues the centrosome amplification phenotype of morgana+/� MEFs (data from three preparations per genotype).

(H) morgana+/�MEFs treated for 48 with 2 mM Y-27632 and untreated morgana+/+ MEFs exhibit comparable frequencies of polyploid metaphases (data from three

preparations per genotype; untreated +/+ versus +/�, p < 0.005; untreated +/� versus treated +/�, p < 0.05; +/+ and treated +/� are not significantly different).

(I) The amount of ROCK II-associated NPM in morgana+/� MEFs is significantly higher than in morgana+/+ MEFs. Extracts precipitated with anti-ROCK II or

control Ig were blotted and stained with specific antibodies. In the graph, the amount of ROCK II bound to NPM is normalized to the total amount of immuno-

precipitated ROCK II. TE, total extract.

Developmental Cell


490 Developmental Cell 18, 486–495, March 16, 2010 ª2010 Elsevier Inc.

Developmental Cell


centrosomes of morgana+/� diploid cells contain either a pair of

engaged centrioles or a single nonengaged centriole, and are

thus genuine centrosomes. The colocalization of g-tubulin and

Sas-6 signals further suggests that the centrioles of supernu-

merary centrosomes duplicate synchronously.

To ascertain whether centrosome amplification observed in

morgana+/� cells is the consequence of multiple rounds of cen-

triole duplication during interphase, we performed a centrosome

duplication assay (Meraldi et al., 1999) in both NIH 3T3 mouse

fibroblasts and U2OS human osteosarcoma cells. Cells infected

with lentivirus carrying either a morgana or a nonsilencing inter-

fering sequence were arrested in S phase by hydroxyurea (HU).

Following 48 hr HU treatment, cells were fixed, immunostained

for g-tubulin, and then examined for centrosome number. Both

NIH and U2OS morgana RNAi cultures, in which morgana

expression was reduced to 20% of the control level (data not

shown), showed a significant increase in cells with more than

two centrosomes compared to controls (Figures 1G and 1H).

We thus conclude that morgana downregulation results in

centriole overduplication in S phase arrested cells.

Morgana/chp-1 Interacts with Hsp90 and ROCK I and II,and Inhibits ROCK Kinase ActivityTo define the role of morgana in the control of centrosome

duplication, we searched for morgana binding partners in

mammals by coimmunoprecipitation analysis followed by

mass spectrometry. Two major bands of 160 and 90 kDa copre-

cipitated with morgana (Figure 3A). MALDI-TOF and LC-nano-

spray-IT analysis showed that the 160 kDa band contains Rho

kinase I and II (ROCK I and ROCK II), while the 90 kDa band is

Hsp90, which has been previously shown to interact with mor-

gana (Hahn, 2005; Wu et al., 2005). Identities of ROCK I, ROCK

II, and Hsp90 in coimmunoprecipitates were confirmed by

western blotting with specific antibodies (Figure 3B). Citron

kinase, another Rho-activated kinase related to ROCK, was

not present in the complex, demonstrating the specificity of

the interactions (Figure 3B). Association of morgana with Rho

kinases was further confirmed by immunoprecipitation experi-

ments from cells transfected with Myc-tagged ROCK I and

ROCK II (Figure 3C).

To detect possible functional relationships between morgana

and its interacting partners, we measured ROCK kinase activity

on a well-known substrate, myosin light chain 2 (MLC2). Western

blotting with an anti-phospho-MLC2 antibody revealed that the

MLC2 phosphorylation level is significantly higher in morgana+/�

MEFs than in wild-type MEFs (Figure 3D). Since ROCK II has

been implicated in the control of centrosome duplication

(Ma et al., 2006), we focused on this kinase. We first asked

whether morgana directly binds ROCK II. We performed a far

western blotting assay using ROCK II immunoprecipitated from

293 cell extracts and different morgana fragments fused to the

maltose binding protein (MBP). Full-length morgana and its

CHORD domains, but not its CS domain or MPB alone, reacted

with immobilized ROCK II (Figures S3A–S3C), suggesting a direct

(J) Morgana inhibits NPM-dependent ROCK II activation. The kinase assay was p

as a substrate; precipitation with unrelated IgG is a negative control (Ig). MBP, H

reactions as indicated. The graph shows the average intensities of P-MYPT1 lev

represent standard errors. (*p < 0.05; **p < 0.01).

Develo

morgana-ROCK II interaction. We next immunoprecipitated

ROCK II from morgana+/� and morgana+/+ MEFs and determined

kinase activity in vitro using myosin phosphatase targeting

subunit 1 (MYPT1) as a ROCK substrate (Figure 3E). Western

blotting with an anti-phospho-MYPT1 antibody showed that

ROCK II from morgana+/� MEFs has a higher kinase activity

than ROCK II from wild-type cells (Figure 3E). We thus conclude

that morgana binds ROCK II and inhibits its kinase activity.

To ascertain whether ROCK kinase activity correlates with

centrosome amplification, we used the ROCK kinase inhibitor

Y-27632. morgana+/� MEFs were incubated for 48 hr with

2 mM Y-27632, an inhibitor concentration that lowers MLC2

phosphorylation to the wild-type level (Figure 3F; Figure S3D).

A comparison between treated and untreated morgana+/�

MEFs revealed that the inhibitor rescues the mitotic phenotypes

elicited by morgana haploinsufficiency: the frequency of cells

(with a nuclear DNA content %4C) showing more than two

centrosomes went back to the wild-type frequency (Figure 3G),

and the frequency of polyploid metaphases was significantly

reduced (Figure 3H). These results strongly suggest that the

increased ROCK kinase activity observed in morgana+/� MEFs

is responsible for centrosome amplification and polyploidy.

Morgana/chp-1 Downregulates ROCK II Kinase Activityby Inhibiting ROCK-NPM InteractionTo investigate the mechanism of morgana-mediated downregu-

lation of ROCK kinase activity, we asked whether morgana

directly inhibits ROCK II. A kinase assay on ROCK II immunopre-

cipitated from 293 cell extracts showed that addition of recombi-

nant morgana does not affect ROCK kinase activity (data not

shown and Figure 3L). We thus investigated whether morgana

affects the ability of ROCK II to bind proteins that modulate its

activity. We focused on nucleophosmin (NPM), a protein that

binds and activates ROCK II and positively regulates centro-

some duplication (Ma et al., 2006). We found that in morgana+/�

MEFs the amount of NPM bound to ROCK II is significantly

higher than in controls (Figure 3I). Moreover, the ability of

recombinant NPM to activate ROCK II precipitated from 293

cell extracts (Figure 3L) was abolished by the addition of

recombinant morgana (Figure 3L). Collectively, these experi-

ments indicate that morgana inhibits ROCK II kinase activity by

interfering with NPM-mediated ROCK activation.

Morgana/chp-1 Reduction Predisposes Cellsto Oncogenic Transformation and EnhancesTumor SusceptibilityWe next asked whether morgana plays roles in cell proliferation

and tumorigenesis. We found that morgana+/� MEFs grow

slightly slower than their wild-type counterparts at early in vitro

passages (P2–P5) (Figure 4A). However, at later in vitro pas-

sages, morgana+/� MEFs showed a higher doubling rate than

morgana+/+ MEFs (Figure 4B). The enhanced growth potential

of morgana+/�MEFs was confirmed by a colony formation assay

at low density seeding. As shown in Figure 4C, morgana+/�MEFs

erformed on ROCK II immunoprecipitated from 293 cell extracts using MYPT1

is-tagged NPM, MBP-tagged morgana, or Y-27632 was added to the kinase

els normalized to total MYPT1 from two different experiments. Bars in graphs


Figure 4. Role and Status of Morgana/chp-1 in Tumorigenesis

(A) Growth curves of morgana+/+ and morgana+/� MEFs at P2 (two MEF preparations per genotype).

(B) Population doublings of morgana+/+ and morgana+/� MEFs (four MEF preparations per genotype).

(C) Low density seeding focus forming assay on morgana+/+ and morgana+/� MEFs at different passages (P5, P8, and P9). The picture shows a representative

result obtained at P9 (two MEF preparations per genotype).

(D) Population doublings of morgana+/+ and morgana+/� MEFs infected with RasV12 or with the empty vector (two MEF preparations per genotype).

(E) Soft agar growth assays of morgana+/+ and morgana+/� MEFs infected with either RasV12 or the empty vector (two MEF preparations per genotype).

(F) morgana+/� mice are sensitive to DMBA-induced tumor formation. The graph shows the average number of DMBA-induced lung tumors per mouse.

The number of tumors observed in morgana+/� mice was normalized to the mean of tumors arising in wild-type mice of the same litter.

Developmental Cell


492 Developmental Cell 18, 486–495, March 16, 2010 ª2010 Elsevier Inc.

Developmental Cell


at P9 displayed a strikingly increased ability to form colonies

compared to wild-type. When we infected morgana+/� and

wild-type MEFs with oncogenic Ras (Ras V12), we found that

RasV12 led to growth arrest and premature senescence in

wild-type MEFs, whereas Ras-infected morgana+/� cells were

able to proliferate at high rate (Figure 4D) and formed colonies

in both low seeding (data not shown) and soft agar assays

(Figure 4E).

To extend these observations to an in vivo model, we tested

the susceptibility of morgana+/� mice to tumor development in

response to chemical mutagens. morgana+/� and morgana+/+

newborn mice were exposed to the carcinogen 7,12-dimethyl-

benz[a]anthracene (DMBA) and examined 5 months later for

the presence of lung tumors (Serrano et al., 1996). morgana+/�

mice displayed a 2.5-fold increase in the frequency of lung

tumors compared to their wild-type littermates exposed to the

same treatment (Figure 4F). Thus, morgana downregulation

results in an increased MEF proliferation rate, predisposes cells

to Ras-driven oncogenic transformation, and increases suscep-

tibility to the chemical induction of lung tumors.

Morgana/chp-1 Expression Is Reducedin Most Human CancersWe finally investigated morgana expression in two prevalent

human cancers. We generated an anti-morgana monoclonal

antibody (P1/PP0), which recognizes a single band of the ex-

pected size (37 kDa) in western blots from both human and

mouse cell extracts (Figure 1D and data not shown). This anti-

body decorates control cells but not cells in which morgana

was silenced by siRNA (Figure S4A). Immunostaining of tumor

tissue arrays with P1/PP0 showed that morgana expression is

strongly reduced in 67.3% (37/55) of breast and 57.7% (45/78)

of lung cancer samples compared with control tissues (Figures

4G and 4H; Figures S4B and S4C). Interestingly we also found

that morgana is overexpressed in a minority of breast (5.4%)

and lung (10.3%) cancer samples. This latter finding may reflect

a phenomenon, called genomic convergence, that tends to

counteract genetic instability in advanced cancer cells, allowing

clonal expansion of cells with chromosome compositions that

confer a proliferative advantage (Heim et al., 1988) (Chiba

et al., 2000). Morgana overexpression might be one of the mech-

anisms through which some tumors achieve genetic stabilization

via suppression of centrosome amplification.

DISCUSSION

We have shown that impairment of the morgana function causes

centrosome amplification in both Drosophila and mice, and that

the activity of the human morgana gene rescues the centrosome

phenotype elicited by mutations in Drosophila mora. We found

that morgana forms a complex with Hsp90, ROCK I, and

ROCK II, three proteins that have been implicated in centrosome

behavior (Chevrier et al., 2002; Du and Hannon, 2004; Lange

et al., 2000; Ma et al., 2006). In particular, previous studies

(G and H) Bar graphs showing levels of morgana expression in tumor tissue array

breast and lung cancer samples compared with control tissues (p < 0.001 and p <

positive, normal or low for morgana expression are also shown. Bars in graphs r

Develo

showed that ROCK II kinase activity promotes centrosome dupli-

cation and is enhanced by NPM binding (Ma et al., 2006). We

found that morgana binds ROCK II, inhibits the association of

ROCK II with its activator NPM, and abolishes NPM-dependent

ROCK II activation. We thus propose that morgana is a highly

conserved suppressor of centrosome overduplication that acts

by interfering with ROCK II-NPM binding. Our results suggest

that morgana and NPM might compete for ROCK II binding;

however, a precise definition of morgana-ROCK II-NPM interac-

tions will require further studies.

We found that the doubling rate of morgana+/� MEFs at early

in vitro passages is lower than that of wild-type MEFs. In

contrast, morgana+/� MEFs at late passages display doubling

rates that are much higher than those observed in wild-type

MEFs. The growth pattern of the morgana+/� MEFs is strikingly

similar to the growth pattern of nucleophosmin+/� MEFs, which

also exhibit centrosome amplification (Grisendi et al., 2005).

These findings are consistent with the hypothesis that the

genomic instability caused by centrosome amplification initially

hampers cell proliferation, while favoring, over time, the emer-

gence of overgrowing clones. It is also conceivable that the

emergence of overgrowing clones from morgana+/� cells would

promote both Ras-mediated oncogenic transformation and

development of DMBA-induced lung tumors.

Our results indicate that morgana downregulation leads to

centrosome amplification and promotes tumorigenesis in mice.

Moreover, we have shown that morgana expression is lost in a

large fraction of human breast and lung cancers. Together, these

results suggest that morgana might be an important factor in

human tumor etiology and progression, and they encourage

further studies on the biological role of this protein.

EXPERIMENTAL PROCEDURES

Targeted Disruption of the Mouse Morgana/chp-1 Gene

The morgana gene (chordc1) was inactivated by homologous recombination in

mouse ES cells. The knockout construct consisted of a 50 arm (a 3.5 kb

genomic fragment), a b-galactosidase gene with a neomycin resistance

cassette (6 kb), and a 30 arm (a 6.3 kb genomic fragment). ES recombinant

clones lacking 30 bp including the morgana ATG translation start site (see

Figure S1A) were identified by Southern blot analysis and injected into

C57BL/6 blastocysts to generate chimeras.

Drosophila Genetics and Cytology

The PBac{PB}CHORDc02881 (mora1) allele, the lines expressing the

GFP-tagged histone H2AV, genetic markers, and special chromosomes are

described in FlyBase (http://www.flybase.org/). Germline transformation

experiments with full-length cDNAs cloned into a pP[UAST] vector were

performed by standard methods. Human and Drosophila transgenes were

induced by the P{Act5C-GAL4}25FO1 insertion (FlyBase). Chromosome

preparations were obtained as previously described (Gatti and Goldberg,

1991). For immunofluorescence, brains were fixed and stained according to

Bonaccorsi et al. (2000). Time-lapse imaging of living neuroblasts was carried

out as described (Savoian and Rieder, 2002).

Mouse Blastocyst Isolation

Mouse embryos from morgana+/� intercrosses were collected at E3.5 by

flushing the uterus with HEPES-buffered medium 2 (M2; Sigma). After 3 days

s of breast (G) and lung (H) cancers. Morgana expression is strongly reduced in

0.01, respectively). Representative pictures of samples that are either strongly

epresent standard errors (*p < 0.05; **p < 0.01; ***p < 0.001).


http://www.flybase.org/

Developmental Cell


of culture, blastocysts were photographed and then genotyped by nested

PCR. TUNEL assays were performed with the Roche detection kit, according

to manufacturer’s instructions.

MEF Preparation and Analysis

MEFs were prepared from E13.5 embryos and passaged every 3 days (3T9

protocol). Growth curves and plating efficiency at low seeding density were

determined as described in Supplemental Information. Metaphase prepara-

tions from exponentially growing MEFs were prepared as described (Babu

et al., 2003). For immunostaining, cells were fixed and stained using standard

methods. To identify MEFs with a nuclear DNA content %4C, the intensity of

DAPI fluorescence was quantified using the AxioVision (Zeiss) image analysis

software. We tuned the system by examining wild-type MEFs and identifying

two DAPI intensity ranges associated with G1 and G2 DNA contents.

Antibody Preparation, Coimmunoprecipitation, Far Western,

and ROCK Kinase Assay

The P1/PP0 monoclonal antibody was obtained by immunizing mice with a

GST-mouse morgana fusion protein. Antibody purification, western blotting,

immunoprecipitation, and mass spectrometry were performed by standard

methods (Brancaccio et al., 1999; Supplemental Information). Far western blot-

ting and ROCK kinase assay were performed as described by Wu et al. (2007)

and Rubenstein et al. (2007), respectively (see Supplemental Information).

Carcinogen Treatment

DMBA treatments were performed as described in Serrano et al. (1996). Five

months after DMBA treatment, mice were sacrificed and screened for tumors

formation.

Human Tumor Histochemistry

Tumor tissue arrays were stained with our P1/PP0 anti-morgana monoclonal

antibody, after a pretreatment with 10 mM citrate buffer (pH 6.0) and micro-

wave (see Supplemental Information).

Statistical Analysis

In statistical analyses, significance was tested using a two-tailed Student’s

t test or, when required, one-way ANOVA followed by Newman Kleus’ test.

SUPPLEMENTAL INFORMATION

Supplemental Information includes four figures, two movies, Supplemental

Experimental Procedures, and Supplemental References and can be found

with this article online at doi:10.1016/j.devcel.2009.12.020.

ACKNOWLEDGMENTS

We thank F. Fusella, F. Damilano, M. Asher, P. Scaravaglio, O. Azzolino, and

T. Cravero for skillful technical assistance; F. Di Cunto, K. Kaibuchi, S. Naru-

miya, T. Megraw, and J. Raff for clones and antibodies, and R. Srinivasan

for comments on the manuscript. This work was supported by PRIN grants

2005 to G.T. and 2006 to M.B., Progetto di Ricerca Sanitaria Finalizzata

2007 and 2008 to M.B., AIRC 2008 to M.B., Centro di Eccellenza di Biologia

e Medicina Molecolare (BEMM) and AIRC 2008 to M.G., and an NIH grant

to P.P.P.

Received: May 29, 2009

Revised: November 4, 2009

Accepted: December 28, 2009

Published: March 15, 2010

REFERENCES

Babu, J.R., Jeganathan, K.B., Baker, D.J., Wu, X., Kang-Decker, N., and

van Deursen, J.M. (2003). Rae1 is an essential mitotic checkpoint regulator

that cooperates with Bub3 to prevent chromosome missegregation. J. Cell

Biol. 160, 341–353.


Basto, R., Lau, J., Vinogradova, T., Gardiol, A., Woods, C.G., Khodjakov, A.,

and Raff, J.W. (2006). Flies without centrioles. Cell 125, 1375–1386.

Basto, R., Brunk, K., Vinadogrova, T., Peel, N., Franz, A., Khodjakov, A., and

Raff, J.W. (2008). Centrosome amplification can initiate tumorigenesis in flies.

Cell 133, 1032–1042.

Bettencourt-Dias, M., and Glover, D.M. (2007). Centrosome biogenesis and

function: centrosomics brings new understanding. Nat. Rev. Mol. Cell Biol.

8, 451–463.

Bonaccorsi, S., Giansanti, M.G., and Gatti, M. (2000). Spindle assembly in

Drosophila neuroblasts and ganglion mother cells. Nat. Cell Biol. 2, 54–56.

Bornens, M. (2002). Centrosome composition and microtubule anchoring

mechanisms. Curr. Opin. Cell Biol. 14, 25–34.

Brancaccio, M., Guazzone, S., Menini, N., Sibona, E., Hirsch, E., De Andrea,

M., Rocchi, M., Altruda, F., Tarone, G., and Silengo, L. (1999). Melusin is a

new muscle-specific interactor for beta(1) integrin cytoplasmic domain.

J. Biol. Chem. 274, 29282–29288.

Brancaccio, M., Fratta, L., Notte, A., Hirsch, E., Poulet, R., Guazzone, S.,

De Acetis, M., Vecchione, C., Marino, G., Altruda, F., et al. (2003a). Melusin,

a muscle-specific integrin beta1-interacting protein, is required to prevent

cardiac failure in response to chronic pressure overload. Nat. Med. 9, 68–75.

Brancaccio, M., Menini, N., Bongioanni, D., Ferretti, R., De Acetis, M., Silengo,

L., and Tarone, G. (2003b). Chp-1 and melusin, two CHORD containing

proteins in vertebrates. FEBS Lett. 551, 47–52.

Chevrier, V., Piel, M., Collomb, N., Saoudi, Y., Frank, R., Paintrand, M.,

Narumiya, S., Bornens, M., and Job, D. (2002). The Rho-associated protein

kinase p160ROCK is required for centrosome positioning. J. Cell Biol. 157,

807–817.

Chiba, S., Okuda, M., Mussman, J.G., and Fukasawa, K. (2000). Genomic

convergence and suppression of centrosome hyperamplification in primary

p53�/� cells in prolonged culture. Exp. Cell Res. 258, 310–321.

Clarkson, M., and Saint, R. (1999). A His2AvDGFP fusion gene complements

a lethal His2AvD mutant allele and provides an in vivo marker for Drosophila

chromosome behavior. DNA Cell Biol. 18, 457–462.

De Acetis, M., Notte, A., Accornero, F., Selvetella, G., Brancaccio, M.,

Vecchione, C., Sbroggio, M., Collino, F., Pacchioni, B., Lanfranchi, G., et al.

(2005). Cardiac overexpression of melusin protects from dilated cardiomyop-

athy due to long-standing pressure overload. Circ. Res. 96, 1087–1094.

Du, J., and Hannon, G.J. (2004). Suppression of p160ROCK bypasses cell

cycle arrest after Aurora-A/STK15 depletion. Proc. Natl. Acad. Sci. USA 101,

8975–8980.

Fukasawa, K. (2007). Oncogenes and tumour suppressors take on centro-

somes. Nat. Rev. Cancer 7, 911–924.

Ganem, N.J., Godinho, S.A., and Pellman, D. (2009). A mechanism linking

extra centrosomes to chromosomal instability. Nature 460, 278–282.

Gatti, M., and Goldberg, M.L. (1991). Mutations affecting cell division in

Drosophila. Methods Cell Biol. 35, 543–586.

Grisendi, S., Bernardi, R., Rossi, M., Cheng, K., Khandker, L., Manova, K., and

Pandolfi, P.P. (2005). Role of nucleophosmin in embryonic development and

tumorigenesis. Nature 437, 147–153.

Hahn, J.S. (2005). Regulation of Nod1 by Hsp90 chaperone complex. FEBS

Lett. 579, 4513–4519.

Heim, S., Mandahl, N., and Mitelman, F. (1988). Genetic convergence and

divergence in tumor progression. Cancer Res. 48, 5911–5916.

Lange, B.M., Bachi, A., Wilm, M., and Gonzalez, C. (2000). Hsp90 is a core

centrosomal component and is required at different stages of the centrosome

cycle in Drosophila and vertebrates. EMBO J. 19, 1252–1262.

Ma, Z., Kanai, M., Kawamura, K., Kaibuchi, K., Ye, K., and Fukasawa, K.

(2006). Interaction between ROCK II and nucleophosmin/B23 in the regulation

of centrosome duplication. Mol. Cell. Biol. 26, 9016–9034.

Mayor, T., Stierhof, Y.D., Tanaka, K., Fry, A.M., and Nigg, E.A. (2000). The

centrosomal protein C-Nap1 is required for cell cycle-regulated centrosome

cohesion. J. Cell Biol. 151, 837–846.

Inc.

http://dx.doi.org/doi:10.1016/j.devcel.2009.12.020

Developmental Cell


Megraw, T.L., Li, K., Kao, L.R., and Kaufman, T.C. (1999). The centrosomin

protein is required for centrosome assembly and function during cleavage in

Drosophila. Development 126, 2829–2839.

Meraldi, P., Lukas, J., Fry, A.M., Bartek, J., and Nigg, E.A. (1999). Centrosome

duplication in mammalian somatic cells requires E2F and Cdk2-cyclin A. Nat.

Cell Biol. 1, 88–93.

Nigg, E.A. (2006). Origins and consequences of centrosome aberrations in

human cancers. Int. J. Cancer 119, 2717–2723.

Paoletti, A., Moudjou, M., Paintrand, M., Salisbury, J.L., and Bornens, M.

(1996). Most of centrin in animal cells is not centrosome-associated and

centrosomal centrin is confined to the distal lumen of centrioles. J. Cell Sci.

109, 3089–3102.

Rubenstein, N.M., Callahan, J.A., Lo, D.H., and Firestone, G.L. (2007).

Selective glucocorticoid control of Rho kinase isoforms regulate cell-cell

interactions. Biochem. Biophys. Res. Commun. 354, 603–607.

Savoian, M.S., and Rieder, C.L. (2002). Mitosis in primary cultures of

Drosophila melanogaster larval neuroblasts. J. Cell Sci. 115, 3061–3072.

Sbroggio, M., Ferretti, R., Percivalle, E., Gutkowska, M., Zylicz, A., Michowski,

W., Kuznicki, J., Accornero, F., Pacchioni, B., Lanfranchi, G., et al. (2008). The

mammalian CHORD-containing protein melusin is a stress response protein

interacting with Hsp90 and Sgt1. FEBS Lett. 582, 1788–1794.

Develo

Serrano, M., Lee, H., Chin, L., Cordon-Cardo, C., Beach, D., and DePinho, R.A.

(1996). Role of the INK4a locus in tumor suppression and cell mortality. Cell 85,

27–37.

Shirasu, K., Lahaye, T., Tan, M.W., Zhou, F., Azevedo, C., and Schulze-Lefert,

P. (1999). A novel class of eukaryotic zinc-binding proteins is required for

disease resistance signaling in barley and development in C. elegans. Cell

99, 355–366.

Strnad, P., Leidel, S., Vinogradova, T., Euteneuer, U., Khodjakov, A., and

Gonczy, P. (2007). Regulated HsSAS-6 levels ensure formation of a single

procentriole per centriole during the centrosome duplication cycle. Dev. Cell

13, 203–213.

Thibault, S.T., Singer, M.A., Miyazaki, W.Y., Milash, B., Dompe, N.A., Singh,

C.M., Buchholz, R., Demsky, M., Fawcett, R., Francis-Lang, H.L., et al.

(2004). A complementary transposon tool kit for Drosophila melanogaster

using P and piggyBac. Nat. Genet. 36, 283–287.

Wei, Y., Yu, L., Bowen, J., Gorovsky, M.A., and Allis, C.D. (1999). Phosphory-

lation of histone H3 is required for proper chromosome condensation and

segregation. Cell 97, 99–109.

Wu, J., Luo, S., Jiang, H., and Li, H. (2005). Mammalian CHORD-containing

protein 1 is a novel heat shock protein 90-interacting protein. FEBS Lett.

579, 421–426.

Wu, Y., Li, Q., and Chen, X.Z. (2007). Detecting protein-protein interactions by

far western blotting. Nat. Protoc. 2, 3278–3284.


Documents

Developmental Cell Short Article - core.ac.uk · 2Dipartimento di Genetica e Biologia Molecolare and IBPM/CNR, Sapienza ... Sloan-Kettering Institute, Memorial Sloan-Kettering Cancer