Oncogene (2010) 29, 5103–5112 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc SHORT COMMUNICATION The Ras oncogene signals centrosome amplification in mammary epithelial cells through D1/Cdk4 and Nek2

X Zeng1, FY Shaikh2,5, MK Harrison1,5, AM Adon1, AJ Trimboli2, KA Carroll3, N Sharma2, C Timmers2, LA Chodosh4, G Leone2 and HI Saavedra1

1Department of Radiation Oncology, Emory University School of Medicine, and Emory Winship Cancer Institute, Atlanta, GA, USA; 2Department of Molecular Virology, Immunology and Medical Genetics, Program of Human Cancer Genetics, Columbus, OH, USA; 3Department of Pathology and Laboratory Medicine, Atlanta, GA, USA and 4Department of Cancer Biology and Abramson Family Cancer Research Institute, University of Pennsylvania School of Medicine, Philadelphia, PA, USA

Centrosome amplification (CA) contributes to carcinogen- (Janes et al., 1994; Hoadley et al., 2007), overexpression esis by generating aneuploidy. Elevated frequencies of of H-Ras, K-Ras and N-Ras in 69% breast cancers or CA in most benign breast lesions and primary tumors mutational activation of K-Ras in 6.5% breast cancers suggest a causative role for CA in breast cancers. Clearly, (Clair et al., 1987; Miyakis et al., 1998). Amplification of identifying which and how altered signal transduction c-Myc results in its overexpression in 15–70% breast pathways contribute to CA is crucial to breast cancer cancers (Chrzan et al., 2001; Rummukainen et al., 2001; control. Although a causative and cooperative role for Blancato et al., 2004). Direct evidence that Ras and c-Myc and Ras in mammary tumorigenesis is well c-Myc are involved in mammary cancers was obtained documented, their ability to generate CA during mammary by the coexpression of c-Myc, H-RasG12V, telomerase tumor initiation remains unexplored. To answer that and SV40 T antigens in primary human mammary question, K-RasG12D and c-Myc were induced in mouse epithelial cells, resulting in transformation (Elenbaas mammary glands. Although CA was observed in mammary et al., 2001). In addition, MMTV (mouse mammary tumors initiated by c-Myc or K-RasG12D, it was detected tumor virus) transgenic mice expressing c-Myc (Sinn only in premalignant mammary lesions expressing et al., 1987; D’Cruz et al., 2001), H-RasG12V (Sinn et al., K-RasG12D. CA, both in vivo and in vitro, was associated 1987; Sarkisian et al., 2007), N-Ras (Mangues et al., with increased expression of the centrosome-regulatory 1990) and K-RasG12D (Omer et al., 2000; Podsypanina , and Nek2. Abolishing the expression et al., 2008) developed mammary tumors. of cyclin D1, Cdk4 or Nek2 in MCF10A human Concurrently, and cell-cycle arrest are mammary epithelial cells expressing H-RasG12V abrogated transient barriers to Myc and Ras mammary carcinogen- Ras-induced CA, whereas silencing or B2 had esis (D’Cruz et al., 2001; Sarkisian et al.,2007). no effect. Thus, we conclude that CA precedes mammary For example, overexpression of H-RasG12V in mouse tumorigenesis, and interfering with centrosome-regulatory mammary epithelial cells results in transient cell-cycle targets suppresses CA. arrest between 14 and 32 days after induction (Sarkisian Oncogene (2010) 29, 5103–5112; doi:10.1038/onc.2010.253; et al., 2007). Indeed, an active pathway is a major published online 28 June 2010 obstacle to H-RasG12V-initiated mammary tumors (Hund- ley et al.,1997b;Sarkisianet al., 2007), and c-Myc Keywords: Ras; centrosome amplification; mammary triggers activating K-Ras mutations to induce nonregres- cancers; cyclin D1; Cdk4; Nek2 sing mammary tumors (D’Cruz et al., 2001). Chromo- some instability (CIN) is a potential mechanism used by oncogenes to abrogate transient barriers to mammary cancers; consistent with this, mammary tumors expres- Introduction sing H-RasG12V and c-Myc are genomically unstable (Hundley et al.,1997b;Weaveret al.,1999).However, Overexpression of Ras and Myc proto-oncogenes the source, timing and relevance of oncogene-dependent in breast cancers is associated with poor prognosis CIN to mammary tumorigenesis are unknown. (Berns et al., 1992; Shackney et al., 2004). Ras is Centrosome amplification (CA), the acquisition of three constitutively active in breast cancers through deregu- or more centrosomes within a cell, is one of the major lated Her2, Erb4 and Egfr tyrosine receptors contributors to CIN in human cancers (Doxsey, 2002). CA is frequently observed in human cancers—including Correspondence: Dr HI Saavedra, Department of Radiation Onco- prostate, colon, breast and cervical cancer—which sug- logy, Emory University School of Medicine, 1365C Clifton Road, gests an involvement in tumorigenesis (Pihan et al., 1998; Room C3084, Atlanta, GA 30322, USA. Carroll et al., 1999; Ghadimi et al., 2000; Lingle et al., E-mail: [email protected] 5These authors contributed equally to this work. 2002). CA is a potential initiator of mammary tumori- Received 23 October 2009; revised 22 April 2010; accepted 20 May 2010; genesis, as a significant fraction of benign breast lesions published online 28 June 2010 (Berman et al., 2005; Guo et al., 2007) and most breast Centrosome amplification in mammary precursor lesions X Zeng et al 5104 cancers display CA (Lingle et al., 2002; Schneeweiss transgenic Ras and c-Myc (Supplementary Figures S1c et al., 2003; Guo et al., 2007). Centrosomes ensure equal and d) showed that levels of K-RasG12D (sevenfold over segregation of by directing the bipolarity the control) are within the average Ras expression in of the mitotic spindle (Fukasawa, 2005). Thus, CA human breast tumors, which are 2–10-fold relative to generates multipolar spindles, merotelic attachments the nonaffected mammary epithelium (Clair et al., 1987; (attachment of single kinetochores to microtubules Miyakis et al., 1998). In contrast, c-Myc levels emanating from different poles), chromosomal lagging are much higher (50–70-fold over controls) than the and aneuploidy, a major type of CIN (Ganem et al., average c-Myc levels in human breast tumors, which 2009). As CA and multipolar mitoses are potentially are 1.8–4-fold relative to the nonaffected mammary transforming, they are suppressed by various mechani- epithelium (Chrzan et al., 2001; Rummukainen et al., sms, including mitotic catastrophe, centrosomal cluster- 2001; Blancato et al., 2004). ing during , and genomic convergence (Oikawa As reported previously for H-RasG12V and K-RasG12D et al., 2005; Quintyne et al., 2005; Ganem et al., 2009). (Sinn et al., 1987; Podsypanina et al., 2008), mammary The most direct evidence showing the involvement tumors initiated by K-RasG12D occurred much faster of CA in tumorigenesis is that ectopic expression relative to c-Myc (Supplementary Figure S1e), and of centrosome-regulatory proteins in transplanted coexpression of K-RasG12D and c-Myc induced mam- Drosophila neuronal stem cells resulted in tumors (Basto mary tumors faster than either transgene did separately. et al., 2008; Castellanos et al., 2008). In mammalian These results allowed us to select a time point of 5 days cancers, aneuploidy is ubiquitous (Duesberg and Rasnick, to investigate events associated with early prema- 2000). In contrast to mammalian cells, in which lignancy, as it precedes tumorigenesis by a few weeks. low-level aneuploidy initiates and sustains various We assessed various abnormal phenotypes associated mouse tumors (Weaver and Cleveland, 2007; Schliekel- with the expression of K-RasG12D and c-Myc that have man et al., 2009), and in contrast to mammalian tumors, been thoroughly studied in tumors, but are poorly which are aneuploid, tumorigenesis in Drosophila was understood in early premalignancy; these include not accompanied by CIN. In fact, of the five indepen- histological changes, ectopic proliferation and apoptosis dent genetic alterations required to transform primary (Hundley et al., 1997a; D’Cruz et al., 2001; Bearss et al., human mammary epithelial cells (Elenbaas et al., 2001), 2002; Blakely et al., 2005; Podsypanina et al., 2008). three (namely H-RasG12V, inactive Rb and p53) trigger Mammary glands expressing oncogenes displayed dis- CA and CIN (Fukasawa et al., 1996; Saavedra et al., tinct histopathological changes at premalignancy: c-Myc 1999; Iovino et al., 2006), whereas c-Myc triggers led to mild hyperplasia of ducts and lobules, with single- aneuploidy and recombinations (Weaver layered acini adjacent to each other. In contrast, et al., 1999); this suggests a close relationship between K-RasG12D, or K-RasG12D and c-Myc severely altered CA, CIN and mammary tumor initiation. the normal structure of the mammary gland; specifically, Showing that CA is involved in mammary tumor ducts and lobules were hyperplastic, epithelial cells initiation requires establishing that oncogene-driven CA occupied the lumen of the acini, and had invaded into occurs during premalignancy and identifying single or the stroma (Figure 1a). Such distinctions were obscured cooperating oncogenes responsible for CA. The identi- in tumors caused by K-RasG12D or c-Myc, as both fication of centrosome-regulatory proteins deregulated harbored numerous malignant epithelial cells and scanty by oncogenes would allow future therapeutic inter- stroma (Figure 1a). ventions to abrogate CA and aneuploidy that drive Ki-67 immunostaining showed that all oncogenes breast tumors. We show the ability of Ras to signal enhanced proliferation of mammary epithelial cells, and CA in premalignant mouse mammary lesions and that K-RasG12D and c-Myc cooperated to increase those human mammary epithelial cells through cyclin frequencies in premalignant lesions (Figures 1a and b). D1/Cdk4 and Nek2. Cleaved caspase-3 showed that although K-RasG12D and c-Myc increased cellular apoptosis during prema- lignancy, only c-Myc signaled apoptosis in tumors. Results K-RasG12D suppressed c-Myc-signaled apoptosis at both stages (Figures 1a and c). Oncogene expression results in dysplasia, ectopic Thus, K-RasG12D and c-Myc are triggering malignant proliferation and apoptosis during premalignancy phenotypes during premalignancy, and their synergistic and tumorigenesis nature is obvious as soon as premalignancy because they Doxycycline-inducible MMTV transgenic mice expres- cooperate to modulate frequencies of proliferation and sing K-RasG12D and/or c-Myc for 5 days, or until apoptosis, as well as to accelerate mammary tumor mammary tumors developed, were used to address formation. various abnormal phenotypes involved in mammary tumor initiation (D’Cruz et al., 2001). Real-time PCR CA during premalignancy is specific to K-RasG12D using transgene-specific primers (Supplementary Table Frequencies of CA were assessed in mammary S1) showed that K-RasG12D and/or c-Myc expressed premalignant lesions and tumors initiated by K-RasG12D robustly in the corresponding transgenic groups, which and/or c-Myc using immunostainings against g-tubulin was undetectable in controls (Supplementary Figures and pericentrin, proteins within the pericentriolar S1a and b). Western blots detecting endogenous and material of centrosomes essential to the nucleation of

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5105 H&E (20 X) Ki-67 (63 X) Caspase-3 (63 X)

5 days Chronic/Tumor 5 days Chronic/Tumor 5 days Chronic/Tumor

Control

MMTV-rtTA; tetO-c-Myc

MMTV-rtTA; tetO-K-RasG12D

MMTV-rtTA; tetO-c-Myc; tetO-K-RasG12D

§ 90 30 5 days * 5 days * * * 75 Chronic/ 25 Chronic/ Tumor Tumor * 60 * 20 * * 45 15 * 30 * 10 positive cells § %Cleaved Caspase-3 % Ki-67-positive cells 15 5 § 0 0 MMTV-rtTA; MMTV-rtTA; MMTV-rtTA MMTV-rtTA MMTV-rtTA MMTV-rtTA MMTV-rtTA tetO-c-Myc MMTV-rtTA tetO-c-Myc tetO-c-Myc tetO-K-RasG12D tetO-c-Myc tetO-K-RasG12D tetO-K-RasG12D tetO-K-RasG12D Figure 1 Inducible expression of K-RasG12D and c-Myc in mouse mammary glands results in distinct histopathology, ectopic proliferation, and apoptosis. (a) Hematoxylin–eosin (H&E) staining of mammary gland cross-sections. The first two columns show that expressions of K-RasG12D or K-RasG12D and c-Myc result in dysplasia in mice treated with doxycycline for 5 days or until tumors developed (chronic means that controls were treated chronically for 10 months but did not develop tumors). The first column was from H&E performed in paraffin-embedded 10 mm sections; the second was performed in 10 mm frozen sections. All immunostainings (columns 3–6) were performed in 7–10 mm frozen mammary cross-sections. The second pair of columns show mammary epithelial cells immunostained with antibodies against Ki-67 in red (Abcam, Cambridge, MA, USA, ab15580; the secondary antibody is conjugated with Alexa Fluor 555). The third pair of columns show mammary epithelial cells immunostained with antibodies against cleaved-caspase-3 in green (Cell Signaling, Danvers, MA, USA, 9661; the secondary antibody is conjugated with Alexa Fluor 488). The nuclei were stained with DAPI in blue. (b, c) The percentages of proliferating cells (with positive Ki-67 staining) and apoptotic cells (with positive caspase-3 staining) were calculated at 5 days or in chronically induced mice. Each group included three independent mice. Averages and s.d. were calculated as 200 epithelial cells per mouse. Significance was assessed using an unequal variance t-test, calculated in Excel (Microsoft, Redmond, WA, USA) (significance: *Po0.05 as compared with MMTV-rtTA controls, yPo0.05 when comparing the same transgenic group 5 days with long-term treatment). microtubules (Figure 2a). In spite of the universal CA However, K-RasG12D and c-Myc did not significantly found in tumors (Figure 2c), only mammary glands increase frequencies of CA compared with K-RasG12D expressing K-RasG12D or K-RasG12D and c-Myc displayed alone. Thus, CA occurs during tumor initiation and is elevated frequencies of CA at premalignancy (Figure 2b). specific to the K-RasG12D pathway.

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5106 MMTV-rtTA; MMTV-rtTA; MMTV-rtTA; tetO-c-Myc; MMTV-rtTA tetO-c-Myc tetO-K-RasG12D tetO-K-RasG12D

Combined (100 X)

5 days

Chronic/ Tumor

80 80 MMTV-rtTA MMTV-rtTA 70 70 MMTV-rtTA-tetO-c-Myc MMTV-rtTA-tetO-c-Myc MMTV-rtTA; tetO-K-RasG12D 60 60 MMTV-rtTA; tetO-K-RasG12D MMTV-rtTA; tetO-c-Myc; tetO-K-RasG12D MMTV-rtTA; tetO-c-Myc; tetO-K-RasG12D 50 50

40 40

30 30 * * 20 * 20 * * number of centrosomes number of centrosomes % Cells with the indicated % Cells with the indicated 10 10

0 0 1 2 ≥3 1 2 ≥3 Centrosome Number (5 days) Centrosome Number (Chronic or Tumor) Figure 2 Expression of K-RasG12D results in centrosome amplification in premalignant mammary lesions, whereas c-Myc-induced centrosome amplification is only detected in tumors. (a) Coimmunostaining with antibodies against g-tubulin (Abcam, ab11317; the secondary antibody is conjugated with Alexa Fluor 555) and pericentrin (BD Biosciences, San Jose, CA, USA, 611814; the secondary antibody is conjugated with Alexa Fluor 488). Pictures presented as merged from the green and red channels, resulting in yellow signals. Arrows indicate cells with CA. Insets of the indicated area are presented for easier visualization. Chronic means that controls were treated chronically for 10 months but did not develop tumors. (b, c) Frequencies of mammary epithelial cells with 1, 2, or X3 centrosomes. Each group included three independent mice. Averages and s.d. were calculated as 200 epithelial cells per mouse. Significance was assessed using an unequal variance t-test, calculated in Excel (significance: *Po0.05 as compared with MMTV-rtTA controls).

K-RasG12D and c-Myc result in different expression levels 2003). Various checkpoint controls activated in response of products governing cell and centrosome to overexpressed H-RasG12V are also involved in the duplication cycles negative regulation of the centrosome cycle; for exa- One of the major mechanisms generating CA is the mple, expression of p21WafÀ1,p16INK4A and p19ARF deregulation of the centrosome duplication cycle in plateau at 8 days after induction, whereas p53 is acti- the late G1/. Deregulation may arise as a vated at 4 days after induction (Sarkisian et al., 2007). consequence of the downregulation of negative regu- A second major mechanism leading to the deregulation of lators of cell and centrosome cycles, including p53 the centrosome cycle is the overexpression of cell and (Fukasawa et al., 1996), NPM (Grisendi et al., 2005), centrosome-regulatory molecules; these include E2F2 p21WafÀ1 (Duensing et al., 2006), p16INK4A (Berman et al., and E2F3 (Meraldi et al., 1999), cyclin D1 (Nelsen 2005; McDermott et al., 2006), Brca1 (Xu et al., 1999), et al., 2005), E and A in p53-null cells (Hanashiro Brca2 (Tutt et al., 1999) and E2F3 (Saavedra et al., et al., 2008), Plk4 (Kleylein-Sohn et al., 2007),

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5107 Mps-1 (Fisk and Winey, 2001) or Nek-2 (Hayward D D ; ; ; G12 et al., 2004). G12 rtTA rtTA Taking these mechanisms into account, we screened rtTA rtTA c-Myc c-Myc; -K-Ras the steady-state transcriptional levels of various mole- -K-Ras tetO- MMTV- MMTV- tetO- tetO MMTV- MMTV- cules involved in cell and centrosome cycles using tetO quantitative real-time PCR (Table 1). Supplementary 1 2 1 2 1 2 1 2 Table S1 lists the primer sequences used to amplify Nek2a Centrosome differentially regulated . None of the CKIs cyclin D1 were significantly downregulated; rather, we observed cyclin E1 p27Kip1 significant overexpression of some of those transcripts, Centrosome cycle including p16INK4A, p19ARF and p27Kip1. Similarly, Nek2, p21Waf1 E2F2, E2F3a, cyclin D1, and Plk4 were p16INK4a phospho-ser15 upregulated. p53 We selected a subset of differentially expressed gene Total products from Table 1 and assessed their steady-state phospho-ser780 protein levels with western blots (Figure 3). In general, phospho-ser807/811 Rb K-RasG12D and c-Myc led to a more robust deregulation Total of most target genes relative to either single oncogene. β-actin The results from real-time PCR were not always Figure 3 Expression of K-RasG12D and c-Myc during prema- consistent with western blots, perhaps because real-time lignancy results in differential expression of proteins governing the PCR can detect minuscule amounts of mRNAs. For cell and centrosome cycles. Lysates were extracted from the mammary glands of 3-month-old mice treated with doxycycline example, western blots did not detect upregulated for 5 days. Proteins were detected by western blots. Antibodies p19ARF or cyclin B2. In addition, even though p16INK4A from Cell Signaling were cyclin D1 (#2922), p16Ink4a (#4824), mRNA was upregulated by K-RasG12D, the p16INK4A phospho-Ser15 p53 (#9284), phospho-Rb (Ser780, #9307; Ser807/ protein was only robustly upregulated in mammary 811, #9308), Rb (#9313) and b-actin (#4970). Antibodies from G12D BD Biosciences included Nek2 (#610593); those from Santa Cruz glands coexpressing K-Ras and c-Myc. In other Biotechnology (Santa Cruz, CA, USA) included cyclin E1 (sc-481), instances, upregulated mRNA corresponded to upregu- p27Kip1 (sc-528), p21Waf1 (sc-397) and p53 (sc-6243). Duplicates (1, 2) lated proteins; for example, p27Kip1 was upregulated represent two independent mice from each group.

Table 1 Five-day induction of K-RasG12D and c-Myc results in the differential expression of various genes Gene name MMTV-rtTA; tetO-c-Myc MMTV-rtTA; tetO-K-RasG12D MMTV-rtTA; tetO-c-Myc; tetO-K-RasG12D

Aurora kinase A 0.97 (0.25) 0.90 (0.25) 2.08 (0.13) Brca1 3.17 (0.28)a 3.55 (1.95) 18.90 (0.43)a Brca2 1.11 (0.19) 0.74 (0.42) 0.86 (0.24) Cdc2 1.35 (0.10) 1.31 (0.14) 3.25 (0.30)a Cdk2 0.60 (0.34) 0.48 (0.11) 0.51 (0.27) Cdk4 0.86 (0.15) 0.64 (0.14) 1.41 (0.07) c-Nap-1 0.81 (0.23) 0.54 (0.38) 0.34 (0.18) (CCNA2) 2.12 (0.17)a 1.74 (0.14) 3.53 (0.31)a (CCNB1) 2.12 (0.27) 1.48 (0.24) 5.71 (0.46)a Cyclin B2 (CCNB2) 6.51 (0.08)a 4.45 (0.16)a 29.79 (0.26)a Cyclin D1 (CCND1) 1.11 (0.27) 4.91 (0.28)a 4.03 (0.46)a Cyclin E1 0.69 (0.23) 0.47 (0.45) 0.51 (0.48) 0.77 (0.22) 0.57 (0.57) 1.10 (0.31) E2f1 0.99 (0.33) 0.56 (0.27) 1.14 (0.10) E2f2 2.60 (0.62)a 4.52 (0.33)a 14.06 (0.31)a E2f3a 2.41 (0.05)a 1.53 (0.35) 3.35 (0.35)a E2f3b 0.52 (0.41) 0.47 (0.42) 0.70 (0.40) E2f4 1.10 (0.42) 1.08 (0.19) 1.86 (0.35) E2f5 0.70 (0.54) 0.53 (0.27) 1.18 (0.44) Mps-1 1.21 (0.55) 1.00 (0.63) 1.62 (0.24) Nek2 0.69 (1.20) 3.88 (2.87) 9.80 (0.61)a NPM 1.94 (0.12) 0.75 (0.11) 1.39 (0.28) p15INK4B (CDKN2B) 0.63 (0.07) 1.73 (0.31) 2.54 (0.14)a p16INK4A (CDKN2A) 0.47 (0.69) 19.34 (0.33)a 48.84 (0.17)a p18INK4C (CDKN2C) 0.61 (0.11) 0.02 (0.32)a 0.01 (0.11)a p19INK4D (CDKN2D) 1.30 (0.50) 1.19 (0.14) 3.98 (0.33)a p19ARF 0.48 (0.03) 7.84 (0.63)a 76.64 (0.67)a p21Waf1 (CDKN1A) 0.54 (0.16) 1.48 (1.45) 4.13 (0.30)a p27Kip1 (CDKN1B) 58.76 (0.22)a 29.79 (0.47)a 50.10 (0.72)a Plk4 3.05 (0.69) 2.33 (1.03) 11.18 (0.40)a

Data is presented as fold change and s.d. (in parentheses), stemming from three independent mice in each genetic group. aDifference is significant by an unequal variance t-test (Po0.05) as compared with the control group (MMTV-rtTA).

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5108 by all oncogenes, and p21WafÀ1 was upregulated by duplication (Adon et al., 2010). is localized K-RasG12D and c-Myc. Another important checkpoint, in the centrosome (Bailly et al., 1992). Thus, it is p53, was hyperphosphorylated in mammary glands reasonable to assume that the upregulation of Nek2, expressing K-RasG12D, or K-RasG12D and c-Myc. cyclin D1 cyclin E1 or B2 may mediate Ras and Various gene products associated with CA were Ras- and c-Myc-dependent CA. Hence, MCF10A cell upregulated; for example, Nek2 was equally upregulated lines stably expressing H-RasG12V, or H-RasG12V and by all combinations of oncogenes, and cyclin E1 c-Myc were generated. MCF10A is a nontransformed was only upregulated by K-RasG12D and c-Myc. More human mammary epithelial cell line with intact p53 importantly, cyclin D1 was upregulated at the mRNA (Neve et al., 2006). The MCF10A system showed minor and protein levels in mammary epithelial cells expres- differences relative to the transgenic mice; for example, sing K-RasG12D, or K-RasG12D and c-Myc. Consistent although H-RasG12V, or H-RasG12V and c-Myc caused with upregulated cyclins D1 or E1 was the increased upregulation of Nek2 and cyclin D1 as observed in vivo, phosphorylation of Rb in mammary glands expressing the expression of cyclins E1 or B2 was unchanged, K-RasG12D, or K-RasG12D and c-Myc. but nevertheless highly expressed (Figure 4a). As ectopic Taken together, the data indicated that rather than expression of H-RasG12V, or H-RasG12V and c-Myc results downmodulating CKIs or p53, K-RasG12D or K-RasG12D in CA (Figure 4c), Nek2, cyclinD1, Cdk4, cyclin E1 and and c-Myc-dependent CA may arise from their ability to upregulate targets that are critical in regulating both the centrosome cycle (such as Nek2) and the cell cycle (such as cyclins D1 and E1). pBH(vector) + - - pBP-Myc - - + G12V - + + pBH-HRas control siRNA Cyclin D1/Cdk4 and Nek2 contribute to Ras- or Ras- and cyclin B2 cyclin B2 Myc-triggered CA We have described the ability of K-RasG12D, either cyclin E alone or coexpressed with c-Myc, to deregulate various Cdk4 Cdk4 key regulators of cell and centrosome duplication cycles. Of those, Nek2 is a centrosome separase normally active cyclin D1 cyclin D1 at mitosis (Fry et al., 1998). Nek2 is overexpressed in Nek2 Nek2 breast cancers and exhibits centriole-splitting activity when expressed in (Hayward et al., 2004). β-actin β-actin Deregulated cyclin E/Cdk2 signals CA (Tarapore et al., 2001; Saavedra et al., 2003; Hanashiro et al., 2008). 20 control siRNA cyclin B2 siRNA Similarly, cyclinD1/Cdk4, is associated with CA (Nelsen * * et al., 2005) and is a key regulator of centrosome cyclin E siRNA * 15 Cdk4 siRNA * cyclin D1 siRNA * * Nek2 siRNA 10 Figure 4 Oncogene-induced centrosome amplification is sup- § § § pressed by siRNA-mediated silencing of Cdk4, cyclin D1 or § § amplification Nek2. MCF10A cells were stably transfected with plasmids 5 § encoding empty vector (pBABE-hygro or pBABE-puro), H-RasG12V G12V

(pBABE-hygro-H-Ras ), c-Myc (pBABE-puro-c-Myc) and H- % Cells with centrosome RasG12V and c-Myc (pBABE-hygro-H-RasG12V and pBABE-puro-c- Myc). These established cell lines were then transfected with 0 Vector G12V G12V control siRNAs duplexes (Ambion, Austin, TX, USA, #4611) or H-Ras H-Ras & against cyclins B2, D1, E1, Cdk4 and Nek2 (siRNA duplexes c-Myc are presented in Supplementary Table S2). (a) Western blots of cyclin B2 (Abcam, ab82287), cyclin E (Santa Cruz Biotechnology, control siRNA cyclin B2 siRNA 60 sc-480), Cdk4 (Abcam, ab7955), cyclin D1 (Cell Signaling, 2922) cyclin E siRNA Cdk4 siRNA and Nek2 (BD Biosciences, 610593) protein levels in MCF10A cyclin D1 siRNA Nek2 siRNA cells stably expressing vector control, H-RasG12V or H-RasG12V 50 and c-Myc, which were starved in 0.2% FBS for 60 h. (b) Western blots of parental MCF10A transfected with control siRNAs, or 40 siRNAs targeting cyclin B2, cyclin E1, Cdk4, cyclin D1 or Nek2, showing knockdown efficiencies. (c, d) Frequencies of CA (double 30 § § immunostaining with g-tubulin and pericentrin) and proliferation § (immunostaining with BrdU, BD Pharmingen, San Jose, CA, USA, § 20 NA61) in MCF10A cells ectopically expressing empty vector, G12V G12V

H-Ras , or H-Ras and c-Myc in the presence of control % Replicating cells or targeted siRNAs. The average and s.d. were calculated from 10 triplicate experiments (*Po0.05 as compared with MCF10A cells transfected with empty vector and control siRNA; yPo0.05 as 0 G12V compared with MCF10A cells transfected with H-Ras ,or Vector H-RasG12V H-RasG12V& G12V H-Ras and c-Myc vectors together with control siRNA). c-Myc For each experiment, we counted at least 200 cells per group.

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5109 Table 2 Cell-cycle distribution of MCF10A cells transfected with unable to induce CA in early premalignancy, but induced empty vector, H-RasG12V and/or c-Myc, and the indicated siRNAs CA in tumors. These findings place K-RasG12D among a Vector siRNA G1 S G2/M group of oncogenes, including Aurora A and Pin-1, (%) (%) (%) causing CA in premalignant mammary lesions (Suizu et al., 2006; Wang et al., 2006). In contrast, c-Myc falls Control Control 60.2±0.02 13.8±0.55 25.9±0.53 in the category of genetic alterations, including ablated Cdk4 50.3±1.32a 18.1±1.15a 31.6±2.46 Cyclin D1 63.9±0.71a 11.8±0.18a 24.3±0.89 p53, that does not display CA during premalignancy Cyclin E1 78.8±0.10a 6.0±0.06a 15.2±0.16a (Goepfert et al., 2000; Fleisch et al., 2006). As some Cyclin B2 68.3±0.55a 13.2±1.99 18.5±1.44a oncogenic stimuli lead to CA at premalignancy, whereas Nek2 71.1±1.46a 9.4±0.26a 19.5±1.21a others do so later, suggests that some oncogenes directly H-RasG12V Control 63.2±0.03 11.3±0.17 25.5±0.14 trigger CA to rapidly initiate tumors, whereas others Cdk4 51.2±0.46a 17.0±2.26 31.8±1.80a require additional genetic or epigenetic changes to Cyclin D1 65.9±1.95 11.8±0.23 22.2±1.72 induce CA. The capacity of K-RasG12D to induce CA Cyclin E1 76.1±2.76a 6.8±1.25a 17.1±1.51a and lower apoptotic frequencies may contribute to faster a a Cyclin B2 75.7±0.28 8.2±1.02 16.2±1.30 times to tumors relative to c-Myc, as both oncogenes are Nek2 74.6±0.05a 7.5±0.10a 17.9±0.05a similarly efficient in triggering ectopic proliferation in H-RasG12V Control 54.9±0.01 10.4±1.23 34.7±1.22 premalignant lesions and tumors. Another major finding and c-Myc was the identification of a subset of K-RasG12D-specific Cdk4 47.6±0.38a 15.0±1.76 37.4±1.38 centrosome-regulatory targets mediating CA in prema- Cyclin D1 58.8±0.95a 13.7±0.78 27.5±1.73a Cyclin E1 64.4±0.30a 7.6±0.04 28.0±0.26a lignant mammary lesions, including Nek2 and cyclin Cyclin B2 62.8±1.96a 14.0±1.38 23.3±0.58a D1/Cdk4, and that their silencing abrogated CA in Nek2 65.4±2.01a 8.2±1.42 26.4±0.58a human mammary epithelial cells. Nevertheless, the question remains as to whether they mediate CA in vivo. Abbreviation: siRNA, small interfering RNA. Interestingly, c-Myc resulted in Nek2 upregulation a Differences are significant (Po0.05) as compared with control without causing CA. An explanation for this is that siRNA. A total of 10 000 cells were collected per experiment using flow cytometry. The columns represent the percentage of cells in Nek2 is necessary, but not sufficient to trigger CA the G1, S or G2/M phase, gated using FlowJo (Ashland, OR, USA), without cooperating with other altered centrosome- and presented as mean±s.d. The results stem from triplicates. regulatory molecules. There is precedent for the cooperative nature of Nek2: For example, the ectopic cyclin B2 were silenced (Figure 4b) using small inter- expression of Nek2 cannot induce CA unless mammary fering RNAs presented in Supplementary Table S2. epithelial cells are preimmortalized with the SV40 Silencing Cdk4, cyclin D1 or Nek2 abrogated oncogene- T-antigen (Hayward et al., 2004). In addition, we triggered CA (Figure 4b). These data showed that showed that c-Myc enhanced proliferation without Nek2 and cyclin D1/Cdk4 are critical to oncogene- causing CA during premalignancy, showing that triggered CA. CA and ectopic proliferation arise independently. An explanation for the ability of silenced cyclin D1, In contrast, c-Myc mammary tumors harbored CA. Cdk4 and Nek2 to suppress CA is that their down- This suggests that c-Myc cooperates with secondary regulation causes cell-cycle arrest. Cell-cycle analysis by alterations to cause CA; one of the alterations might be flow cytometry (Table 2) showed that silencing cyclin K-Ras, as it is a common hotspot in nonregressing E1, cyclin B2 and Nek2 in MCF10A cells expressing mouse mammary tumors initiated by c-Myc (D’Cruz H-RasG12V significantly increased cell population in the et al., 2001). and correspondingly decreased cell population Evidence suggests that ablated E2F3 and p53 upre- in the G2/M phases, whereas inhibition of Cdk4 or gulate Cdk2 to trigger CA (Fukasawa et al., 1996; cyclin D1 did not. In MCF10A cells expressing vector Saavedra et al., 2003). In fact, our recent work showed control, or coexpressing H-RasG12V and c-Myc, silencing that Cdk2 and Cdk4 are key mediators of CA in p53- Nek2, cyclins D1, B2 or E1 significantly elevated cells null cells (Adon et al., 2010). In contrast, the ectopic accumulating in the G1 phase. 5-Bromo-2-deoxyuridine expression of cyclin E (the catalytic partner of Cdk2) in incorporation assays (Figure 4d) showed that knock- hepatocytes only results in mild CA, whereas over- downs of cyclin D1 and Nek2 significantly inhibited expression of cyclin D1 (the catalytic partner of Cdk4) cell proliferation induced by ectopically expressing results in more severe CA (Nelsen et al., 2005), H-RasG12V, or H-RasG12V and c-Myc. Thus, deregulation suggesting that in some cell/tissue types, cyclin D1/ of the cell cycle is not the only cause of CA. Cdk4 is more potent than cyclin E/Cdk2 in signaling CA. This is also evidenced by our observations that silencing cyclin D1/Cdk4 significantly inhibits Discussion H-RasG12V, or H-RasG12V and c-Myc-dependent CA. In contrast, inhibition of cyclin E1 or cyclin B2 severely This study addresses some important questions regard- alters cell-cycle profiles without affecting CA. ing the relationship between CA and mammary tumori- Although our studies clearly showed that Cdk4 is genesis. First, we show that CA precedes tumorigenesis involved in Ras-induced CA, it is unknown how it leads and is oncogene specific, as K-RasG12D initiated CA in to CA. One explanation is that Cdk4 phosphorylates mammary precursor lesions. In contrast, c-Myc was targets required for regulating the centrosome cycle.

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5110 For example, there is a strong correlation between potential of breast cancer cells (Wu et al., 2008; Tsunoda hyperactive Cdk2, hyperphosphorylation and inactiva- et al., 2009), we propose that centrosomal-regulatory tion of NPM—a major negative regulator of the targets downstream of Ras would represent important centrosome cycle (Saavedra et al., 2003). Our recent future targets for intervening with breast tumorigenesis. work showed that NPM is phosphorylated by Cdk4 during the G1 phase, and that expressing NPMT199A,a mutant lacking the Cdk2/Cdk4 phosphorylation site prevented CA in p53-null cells (Adon et al., 2010). Conflict of interest Similarly, deregulated Cdk4 may also use canonical Cdk2 phosphorylation sites in molecules involved in The authors declare no conflict of interest. other steps in the centrosome cycle, including CP110, Mps-1 and Plk4—regulators of centriole duplication that are direct Cdk2 targets or that require Cdk2 Acknowledgements for their optimal activity (Chen et al., 2002; Fisk et al., 2003; Habedanck et al., 2005; Kleylein-Sohn et al., We thank Drs Rene Opavsky, Paul W Doetsch, Ya Wang and 2007). Another explanation is that hyperactive cyclin Hui Wang for manuscript discussions. We also thank D/Cdk4 hyperphosphorylates Rb, leading to increased Ms Carla G Saavedra and Meredith Roberts for editing; activity and the deregulation of centrosome- Dr Harold Varmus for providing tetO-K-RasG12D mice; regulatory targets. We will identify transcriptional Dr J Brugge for nontransformed MCF10A cells; and Jana and posttranscriptional targets of cyclin D1/Cdk4 in Opavska, Joi Carmichael and Stacy Sannem for technical the future. assistance. We thank Dr Adam Marcus (from the Emory As ablation of cyclin D1 or Cdk4 abrogates mam- Imaging Core) and Mr Alan Bakaletz, for imaging advice. Lewis A Chodosh was funded by NIH R01CA98371, DOD mary tumorigenesis in MMTV-Ras or MMTV-Neu BCRP W81XWH-05-1-0405 and NIH U01 CA105490, Gus- (Her2) mice (Yu et al., 2001; Reddy et al., 2005), Cdk4/ tavo Leone by R01CA85619, R01HD042619, R01CA121275, Cdk6-specific inhibitors reduce ectopic proliferation in R01HD047470 and P01CA097189, Harold Saavedra by human Her2 þ breast cancer cells (Finn et al., 2009), K01CA104079, and a Georgia Cancer Coalition Distinguished and inhibitors of Nek2 decrease the tumorigenic Scholar Award.

References

Adon AM, Zeng X, Harrison MK, Sannem S, Kiyokawa H, Kaldis P cancer: chromosome instability induced by p53 mutation and/or et al. (2010). Cdk2 and Cdk4 regulate the centrosome cycle and Mdm2 overexpression. Oncogene 18: 1935–1944. are critical mediators of centrosome amplification in p53-null cells. Castellanos E, Dominguez P, Gonzalez C. (2008). Centrosome Mol Cell Biol 30: 694–710. dysfunction in Drosophila neural stem cells causes tumors that are Bailly E, Pines J, Hunter T, Bornens M. (1992). Cytoplasmic not due to genome instability. Curr Biol 18: 1209–1214. accumulation of cyclin B1 in human cells: association with Chen Z, Indjeian VB, McManus M, Wang L, Dynlacht BD. (2002). a detergent-resistant compartment and with the centrosome. J Cell CP110, a cell cycle-dependent CDK substrate, regulates centrosome Sci 101(Part 3): 529–545. duplication in human cells. Dev Cell 3: 339–350. Basto R, Brunk K, Vinadogrova T, Peel N, Franz A, Khodjakov A Chrzan P, Skokowski J, Karmolinski A, Pawelczyk T. (2001). et al. (2008). Centrosome amplification can initiate tumorigenesis in Amplification of c-myc gene and overexpression of c-Myc protein flies. Cell 133: 1032–1042. in breast cancer and adjacent non-neoplastic tissue. Clin Biochem Bearss DJ, Lee RJ, Troyer DA, Pestell RG, Windle JJ. (2002). 34: 557–562. Differential effects of (WAF1/CIP1) deficiency on MMTV-ras Clair T, Miller WR, Cho-Chung YS. (1987). Prognostic significance of and MMTV-myc mammary tumor properties. Cancer Res 62: the expression of a ras protein with a molecular weight of 21,000 by 2077–2084. human breast cancer. Cancer Res 47: 5290–5293. Berman H, Zhang J, Crawford YG, Gauthier ML, Fordyce CA, D’Cruz CM, Gunther EJ, Boxer RB, Hartman JL, Sintasath L, McDermott KM et al. (2005). Genetic and epigenetic changes in Moody SE et al. (2001). c-MYC induces mammary tumorigenesis by mammary epithelial cells identify a subpopulation of cells involved means of a preferred pathway involving spontaneous Kras2 in early . Cold Spring Harb Symp Quant Biol 70: mutations. Nat Med 7: 235–239. 317–327. Doxsey S. (2002). Duplicating dangerously: linking centrosome Berns EM, Foekens JA, van Putten WL, van Staveren IL, Portengen duplication and aneuploidy. Mol Cell 10: 439–440. H, de Koning WC et al. (1992). Prognostic factors in human Duensing A, Ghanem L, Steinman RA, Liu Y, Duensing S. (2006). primary breast cancer: comparison of c-myc and HER2/neu p21(Waf1/Cip1) deficiency stimulates centriole overduplication. amplification. J Steroid Biochem Mol Biol 43: 13–19. Cell Cycle 5: 2899–2902. Blakely CM, Sintasath L, D’Cruz CM, Hahn KT, Dugan KD, Duesberg P, Rasnick D. (2000). Aneuploidy, the somatic mutation Belka GK et al. (2005). Developmental stage determines the that makes cancer a species of its own. Cell Motil Cytoskeleton 47: effects of MYC in the mammary epithelium. Development 132: 81–107. 1147–1160. Elenbaas B, Spirio L, Koerner F, Fleming MD, Zimonjic DB, Blancato J, Singh B, Liu A, Liao DJ, Dickson RB. (2004). Correlation Donaher JL et al. (2001). Human breast cancer cells generated by of amplification and overexpression of the c-myc oncogene in oncogenic transformation of primary mammary epithelial cells. high-grade breast cancer: FISH, in situ hybridisation and immuno- Genes Dev 15: 50–65. histochemical analyses. Br J Cancer 90: 1612–1619. Finn RS, Dering J, Conklin D, Kalous O, Cohen DJ, Desai AJ et al. Carroll PE, Okuda M, Horn HF, Biddinger P, Stambrook PJ, (2009). PD 0332991, a selective kinase 4/6 inhibitor, Gleich LL et al. (1999). Centrosome hyperamplification in human preferentially inhibits proliferation of luminal estrogen receptor-

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5111 positive human breast cancer cell lines in vitro. Breast Cancer Res Mangues R, Seidman I, Pellicer A, Gordon JW. (1990). Tumorigenesis 11: R77. and male sterility in transgenic mice expressing a MMTV/N-ras Fisk HA, Mattison CP, Winey M. (2003). Human Mps1 protein oncogene. Oncogene 5: 1491–1497. kinase is required for centrosome duplication and normal mitotic McDermott KM, Zhang J, Holst CR, Kozakiewicz BK, Singla V, progression. Proc Natl Acad Sci USA 100: 14875–14880. Tlsty TD. (2006). (INK4a) prevents centrosome dysfunction and Fisk HA, Winey M. (2001). The mouse Mps1p-like kinase regulates genomic instability in primary cells. PLoS Biol 4: e51. centrosome duplication. Cell 106: 95–104. Meraldi P, Lukas J, Fry AM, Bartek J, Nigg EA. (1999). Centrosome Fleisch MC, Maxwell CA, Kuper CK, Brown ET, Barcellos-Hoff MH, duplication in mammalian somatic cells requires E2F and Cdk2- Costes SV. (2006). Intensity-based signal separation algorithm for . Nat Cell Biol 1: 88–93. accurate quantification of clustered centrosomes in tissue sections. Miyakis S, Sourvinos G, Spandidos DA. (1998). Differential expres- Microsc Res Tech 69: 964–972. sion and mutation of the ras family genes in human breast cancer. Fry AM, Meraldi P, Nigg EA. (1998). A centrosomal function for the Biochem Biophys Res Commun 251: 609–612. human Nek2 protein kinase, a member of the NIMA family of cell Nelsen CJ, Kuriyama R, Hirsch B, Negron VC, Lingle WL, Goggin cycle regulators. EMBO J 17: 470–481. MM et al. (2005). Short term cyclin D1 overexpression induces Fukasawa K. (2005). Centrosome amplification, chromosome centrosome amplification, mitotic spindle abnormalities, and instability and cancer development. Cancer Lett 230: 6–19. aneuploidy. J Biol Chem 280: 768–776. Fukasawa K, Choi T, Kuriyama R, Rulong S, Vande Woude GF. Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T et al. (1996). Abnormal centrosome amplification in the absence of p53. (2006). A collection of breast cancer cell lines for the study of Science 271: 1744–1747. functionally distinct cancer subtypes. Cancer Cell 10: 515–527. Ganem NJ, Godinho SA, Pellman D. (2009). A mechanism linking Oikawa T, Okuda M, Ma Z, Goorha R, Tsujimoto H, Inokuma H extra centrosomes to chromosomal instability. Nature 460: 278–282. et al. (2005). Transcriptional control of BubR1 by p53 and Ghadimi BM, Sackett DL, Difilippantonio MJ, Schrock E, Neumann suppression of centrosome amplification by BubR1. Mol Cell Biol T, Jauho A et al. (2000). Centrosome amplification and instability 25: 4046–4061. occurs exclusively in aneuploid, but not in diploid colorectal cancer Omer CA, Chen Z, Diehl RE, Conner MW, Chen HY, Trumbauer cell lines, and correlates with numerical chromosomal aberrations. ME et al. (2000). Mouse mammary tumor virus-Ki-rasB transgenic Genes Chromosomes Cancer 27: 183–190. mice develop mammary carcinomas that can be growth-inhibited Goepfert TM, McCarthy M, Kittrell FS, Stephens C, Ullrich RL, by a farnesyl:protein transferase inhibitor. Cancer Res 60: Brinkley BR et al. (2000). Progesterone facilitates chromosome 2680–2688. instability (aneuploidy) in p53 null normal mammary epithelial cells. Pihan GA, Purohit A, Wallace J, Knecht H, Woda B, Quesenberry P FASEB J 14: 2221–2229. et al. (1998). Centrosome defects and genetic instability in malignant Grisendi S, Bernardi R, Rossi M, Cheng K, Khandker L, Manova K tumors. Cancer Res 58: 3974–3985. et al. (2005). Role of nucleophosmin in embryonic development and Podsypanina K, Politi K, Beverly LJ, Varmus HE. (2008). Oncogene tumorigenesis. Nature 437: 147–153. cooperation in tumor maintenance and tumor recurrence in mouse Guo HQ, Gao M, Ma J, Xiao T, Zhao LL, Gao Y et al. (2007). mammary tumors induced by Myc and mutant Kras. Proc Natl Analysis of the cellular centrosome in fine-needle aspirations of the Acad Sci USA 105: 5242–5247. breast. Breast Cancer Res 9: R48. Quintyne NJ, Reing JE, Hoffelder DR, Gollin SM, Saunders WS. Habedanck R, Stierhof YD, Wilkinson CJ, Nigg EA. (2005). The Polo (2005). Spindle multipolarity is prevented by centrosomal clustering. kinase Plk4 functions in centriole duplication. Nat Cell Biol 7: Science 307: 127–129. 1140–1146. Reddy HK, Mettus RV, Rane SG, Grana X, Litvin J, Reddy EP. Hanashiro K, Kanai M, Geng Y, Sicinski P, Fukasawa K. (2008). (2005). Cyclin-dependent kinase 4 expression is essential for neu- Roles of cyclins A and E in induction of centrosome amplification in induced breast tumorigenesis. Cancer Res 65: 10174–10178. p53-compromised cells. Oncogene 27: 5288–5302. Rummukainen JK, Salminen T, Lundin J, Kytola S, Joensuu H, Isola Hayward DG, Clarke RB, Faragher AJ, Pillai MR, Hagan IM, Fry JJ. (2001). Amplification of c-myc by fluorescence in situ hybridiza- AM. (2004). The centrosomal kinase Nek2 displays elevated tion in a population-based breast cancer tissue array. Mod Pathol levels of protein expression in human breast cancer. Cancer Res 14: 1030–1035. 64: 7370–7376. Saavedra HI, Fukasawa K, Conn CW, Stambrook PJ. (1999). MAPK Hoadley KA, Weigman VJ, Fan C, Sawyer LR, He X, Troester MA mediates RAS-induced chromosome instability. J Biol Chem 274: et al. (2007). EGFR associated expression profiles vary with breast 38083–38090. tumor subtype. BMC Genomics 8: 258. Saavedra HI, Maiti B, Timmers C, Altura R, Tokuyama Y, Fukasawa Hundley JE, Koester SK, Troyer DA, Hilsenbeck SG, Barrington RE, K et al. (2003). Inactivation of E2F3 results in centrosome Windle JJ. (1997a). Differential regulation of cell cycle character- amplification. Cancer Cell 3: 333–346. istics and apoptosis in MMTV-myc and MMTV-ras mouse Sarkisian CJ, Keister BA, Stairs DB, Boxer RB, Moody SE, Chodosh mammary tumors. Cancer Res 57: 600–603. LA. (2007). Dose-dependent oncogene-induced senescence in vivo Hundley JE, Koester SK, Troyer DA, Hilsenbeck SG, Subler MA, and its evasion during mammary tumorigenesis. Nat Cell Biol 9: Windle JJ. (1997b). Increased tumor proliferation and genomic 493–505. instability without decreased apoptosis in MMTV-ras mice deficient Schliekelman M, Cowley DO, O’Quinn R, Oliver TG, Lu L, Salmon in p53. Mol Cell Biol 17: 723–731. ED et al. (2009). Impaired Bub1 function in vivo compromises Iovino F, Lentini L, Amato A, Di Leonardo A. (2006). RB acute loss tension-dependent checkpoint function leading to aneuploidy and induces centrosome amplification and aneuploidy in murine primary tumorigenesis. Cancer Res 69: 45–54. fibroblasts. Mol Cancer 5: 38. Schneeweiss A, Sinn HP, Ehemann V, Khbeis T, Neben K, Krause U Janes PW, Daly RJ, deFazio A, Sutherland RL. (1994). Activation et al. (2003). Centrosomal aberrations in primary invasive breast of the Ras signalling pathway in human breast cancer cells cancer are associated with nodal status and hormone receptor overexpressing erbB-2. Oncogene 9: 3601–3608. expression. Int J Cancer 107: 346–352. Kleylein-Sohn J, Westendorf J, Le Clech M, Habedanck R, Stierhof Shackney SE, Smith CA, Pollice A, Brown K, Day R, Julian T et al. YD, Nigg EA. (2007). Plk4-induced centriole biogenesis in human (2004). Intracellular patterns of Her-2/neu, ras, and ploidy cells. Dev Cell 13: 190–202. abnormalities in primary human breast cancers predict post- Lingle WL, Barrett SL, Negron VC, D’Assoro AB, Boeneman K, operative clinical disease-free survival. Clin Cancer Res 10: Liu W et al. (2002). Centrosome amplification drives chromosomal 3042–3052. instability in breast tumor development. Proc Natl Acad Sci USA Sinn E, Muller W, Pattengale P, Tepler I, Wallace R, Leder P. (1987). 99: 1978–1983. Coexpression of MMTV/v-Ha-ras and MMTV/c-myc genes in

Oncogene Centrosome amplification in mammary precursor lesions X Zeng et al 5112 transgenic mice: synergistic action of oncogenes in vivo. Cell 49: epithelium induces genetic instability preceding mammary tumor 465–475. formation. Oncogene 25: 7148–7158. Suizu F, Ryo A, Wulf G, Lim J, Lu KP. (2006). Pin1 regulates Weaver BA, Cleveland DW. (2007). Aneuploidy: instigator and centrosome duplication, and its overexpression induces inhibitor of tumorigenesis. Cancer Res 67: 10103–10105. centrosome amplification, chromosome instability, and oncogenesis. Weaver ZA, McCormack SJ, Liyanage M, du Manoir S, Coleman A, Mol Cell Biol 26: 1463–1479. Schrock E et al. (1999). A recurring pattern of chromosomal Tarapore P, Horn HF, Tokuyama Y, Fukasawa K. (2001). Direct aberrations in mammary gland tumors of MMTV-cmyc transgenic regulation of the centrosome duplication cycle by the p53-p21Waf1/ mice. Genes Chromosomes Cancer 25: 251–260. Cip1 pathway. Oncogene 20: 3173–3184. Wu G, Qiu XL, Zhou L, Zhu J, Chamberlin R, Lau J et al. (2008). Tsunoda N, Kokuryo T, Oda K, Senga T, Yokoyama Y, Nagino M Small molecule targeting the Hec1/Nek2 mitotic pathway suppresses et al. (2009). Nek2 as a novel molecular target for the treatment tumor cell growth in culture and in animal. Cancer Res 68: of breast carcinoma. Cancer Sci 100: 111–116. 8393–8399. Tutt A, Gabriel A, Bertwistle D, Connor F, Paterson H, Peacock J Xu X, Weaver Z, Linke SP, Li C, Gotay J, Wang XW et al. (1999). et al. (1999). Absence of Brca2 causes genome instability Centrosome amplification and a defective G2-M cell cycle by chromosome breakage and loss associated with centrosome checkpoint induce genetic instability in BRCA1 exon 11 isoform- amplification. Curr Biol 9: 1107–1110. deficient cells. Mol Cell 3: 389–395. Wang X, Zhou YX, Qiao W, Tominaga Y, Ouchi M, Ouchi T et al. Yu Q, Geng Y, Sicinski P. (2001). Specific protection against breast (2006). Overexpression of aurora kinase A in mouse mammary cancers by cyclin D1 ablation. Nature 411: 1017–1021.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene