Cell Cycle Dependent Tumor Engraftment and Migration Are … · ra-A activity at centrosomes, as...

Cell Cycle and Senescence

Cell Cycle–Dependent Tumor Engraftment andMigration Are Enabled by Aurora-ATony L.H. Chu1, Marisa Connell1, Lixin Zhou2, Zhengcheng He1, Jennifer Won3,Helen Chen1, Seyed M.R. Rahavi1, Pooja Mohan1, Oksana Nemirovsky1,Abbas Fotovati1, Miguel Angel Pujana4, Gregor S.D. Reid1,5,Torsten O. Nielsen3, Nelly Pante2, and Christopher A. Maxwell1,5

Abstract

Cell-cycle progression and the acquisition of a migratoryphenotype are hallmarks of human carcinoma cells that areperceived as independent processes but may be intercon-nected by molecular pathways that control microtubule nucle-ation at centrosomes. Here, cell-cycle progression dramaticallyimpacts the engraftment kinetics of 4T1-luciferase2 breastcancer cells in immunocompetent BALB/c or immunocom-promised NOD-SCID gamma (NSG) mice. Multiparameterimaging of wound closure assays was used to track cell-cycleprogression, cell migration, and associated phenotypes inepithelial cells or carcinoma cells expressing a fluorescenceubiquitin cell-cycle indicator. Cell migration occurred with anelevated velocity and directionality during the S–G2-phase ofthe cell cycle, and cells in this phase possess front-polarizedcentrosomes with augmented microtubule nucleation capac-ity. Inhibition of Aurora kinase-A (AURKA/Aurora-A) dam-pens these phenotypes without altering cell-cycle progression.During G2-phase, the level of phosphorylated Aurora-A atcentrosomes is reduced in hyaluronan-mediated motility

receptor (HMMR)-silenced cells as is the nuclear transport ofTPX2, an Aurora-A–activating protein. TPX2 nuclear transportdepends upon HMMR-T703, which releases TPX2 from acomplex with importin-a (KPNA2) at the nuclear envelope.Finally, the abundance of phosphorylated HMMR-T703, asubstrate for Aurora-A, predicts breast cancer–specific survivaland relapse-free survival in patients with estrogen receptor(ER)–negative (n ¼ 941), triple-negative (TNBC) phenotype(n ¼ 538), or basal-like subtype (n ¼ 293) breast cancers, butnot in those patients with ER-positive breast cancer (n ¼2,218). Together, these data demonstrate an Aurora-A/TPX2/HMMR molecular axis that intersects cell-cycle progres-sion and cell migration.

Implications: Tumor cell engraftment, migration, and cell-cycleprogression share common regulation of the microtubule cyto-skeleton through the Aurora-A/TPX2/HMMR axis, which hasthe potential to influence the survival of patients with ER-negative breast tumors. Mol Cancer Res; 16(1); 16–31. �2017 AACR.

IntroductionProliferation and migration are hallmarks of carcinoma cells

(1). These critical processes are often perceived as independent,but molecular pathways that control microtubule nucleation at

centrosomes may interconnect them. The nucleation of micro-tubules at centrosomes (2), and not at cell–cell contacts (3), isneeded to prepare polarized epithelial cells for cell division.Similarly, the scattering of epithelial cells relies upon the nucle-ation of microtubules at centrosomes (4). Although epithelial-to-mesenchymal transition (EMT)maybe dispensable formetastasisin animal models (5, 6), motile cancer cell populations that haveundergone EMT gain gene expression profiles that are shared withproliferative, tumor-initiating populations (7, 8). The examina-tion of single metastatic cells derived from patient-derived xeno-grafts of triple-negative breast tumors also indicates convergencebetween the expression of EMT-related gene products and pro-liferation. Metastatic cells isolated from low metastatic burdentissues express gene profiles similar to normal basal cells (9),analogous to rare invasive "leader" cells at the periphery ofprimary tumors (10). In these xenograft models, metastatic pro-gression requires a switch from dormancy to cell cycle, and highmetastatic burden associates with increased proliferation (9).Thus, carcinoma cell division, EMT, and metastasis may beinterconnected by a requirement for the acquisition of microtu-bule nucleation capacity at the centrosome. Indeed, increasedmicrotubule nucleation at supernumerary centrosomes is suffi-cient to trigger cell invasion in nontransformed humanmammaryepithelial cells (11).

1Department of Pediatrics, University of British Columbia, Vancouver, BritishColumbia, Canada. 2Department of Zoology, University of British Columbia,Vancouver, British Columbia, Canada. 3Department of Pathology and Labora-tory Medicine, University of British Columbia, Vancouver, British Columbia,Canada. 4Breast Cancer and Systems Biology Unit, Program Against CancerTherapeutic Resistance (ProCure), Catalan Institute of Oncology, IDIBELL,L'Hospitalet del Llobregat, Barcelona, Spain. 5Michael Cuccione ChildhoodCancer Research Program, BC Children's Hospital, Vancouver, British Columbia,Canada.

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

T.L.H. Chu and M. Connell contributed equally to this article.

Corresponding Author: Christopher A. Maxwell, University of British Columbia,BC Children's Hospital, Room 3086, 950 West 28th Avenue, Vancouver, BritishColumbia V5Z4H4, Canada. Phone: 604-875-2000; Fax: 604-875-3120; E-mail:[email protected]

doi: 10.1158/1541-7786.MCR-17-0417

�2017 American Association for Cancer Research.

MolecularCancerResearch

Mol Cancer Res; 16(1) January 201816

on February 15, 2021. © 2018 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst October 9, 2017; DOI: 10.1158/1541-7786.MCR-17-0417

http://crossmark.crossref.org/dialog/?doi=10.1158/1541-7786.MCR-17-0417&domain=pdf&date_stamp=2017-12-19

http://mcr.aacrjournals.org/

Aurora kinase A (Aurora-A) increases the microtubule nucle-ation capacity at centrosomes prior to and during mitosis (12).During mitosis, Aurora-A is optimally active when in a com-plex with targeting protein for Xklp2 (TPX2; ref. 13). A gradientof Ran-GTP near mitotic chromosomes releases TPX2 fromimportin-a, a nuclear import receptor, and enables the for-mation of an Aurora-A–TPX2 heterodimer (14, 15). The mitot-ic localization of TPX2, and in turn its access to the kinase, ispromoted by hyaluronan-mediated motility receptor (HMMR;refs. 16, 17), a nonmotor adaptor protein (18). However, lessis known about Aurora-A activation and function withinnonmitotic cells.

During interphase, Aurora-A and HMMR localize to centro-somes and microtubules, while TPX2 is largely nuclear (19);despite their occupation of distinct subcellular locations, Auro-ra-A activity at centrosomes, as determined by a fluorescenceresonance energy transfer (FRET)–based reporter, is dampenedin TPX2-silenced cells (20). However, it remains largely unex-plored whether molecular pathways that control microtubuleorganization during mitosis also regulate these processes innonmitotic cells and, if so, whether the augmented microtubulenucleation at centrosomes that occurs as a cell progresses throughthe cell cycle enhances that cell's migratory capacity.

Here, we find that cell-cycle progression dramatically impactsthe engraftment kinetics of breast cancer cells. We use multipa-rameter imaging of wound closure assays to track cell-cycleprogression, cell migration, and associated phenotypes in epithe-lial cells or carcinoma cells expressing a fluorescence ubiquitincell-cycle indicator (FUCCI). We find cell migration occurs withan elevated velocity and directionality during S–G2-phase, andcells in this phase possess front-polarized centrosomes withaugmented microtubule nucleation capacity. Aurora-A andHMMR promote wound closure in G2-phase cells and, in thesecells, the silencing of HMMR dampens Aurora-A activity andimpedes the nuclear transport of TPX2. Mechanistically, TPX2nuclear transport relies upon HMMR-T703, which releases TPX2from a complex with importin-a at the nuclear envelope. Finally,our analysis of 3,922 clinically annotated mammary carcinomatissues finds phosphorylatedHMMR-T703 (pHMMR) abundanceto be a significant predictor of breast cancer–specific survival(BCSS) and relapse-free survival (RFS) in patients with estrogenreceptor (ER)–negative breast cancer. We propose that epithelialcell migration and cell-cycle progression share common regula-tion of the microtubule cytoskeleton through the Aurora-A–TPX2–HMMR axis, which could represent an effective therapeutictarget in ER-negative breast cancer.

Materials and MethodsAnimals

All mice were maintained in the specific pathogen-free animalfacility at BC Children's Hospital on a 12-hour light cycle, 20�C�2�C, with 50% � 5% relative humidity, and with food and waterad libitum. All procedures involving animals were in accordancewith theCanadianCouncil onAnimal Care andUBCAnimal CareCommittee (protocol no. A15-0187).

4T1-luciferase 2 (luc2) engraftment assayA total of 10,000 4T1-luc2 Bioware Ultra cells (PerkinElmer,

124087) single cells suspended in 0.2 mL PBS were injectedthrough the tail veins of BALB/c or nonobese diabetic (NOD)-

severe combined immunodeficient (scid) IL2 receptor gammachain gene null (gamma) mice (NSG mice). Successful injectionwas confirmed by bioluminescent detection of cells in lungs 2 to 4hours after injection. Recipientmice were subsequently imaged atindicated time points after engraftment. All imaging was per-formed on the Spectral Instruments Imaging Ami-X platform 5minutes after intraperitoneal injection of 1 mg D-luciferin potas-sium salt (dissolved in 0.1 mL PBS). Images were analyzed usingthe Amiview software program version 1.7.06 (Spectral Instru-ments Imaging).

Cell culture4T1-luc2BiowareUltra cells (PerkinElmer, 124087)were grown

in RPMI media supplemented with 10% FBS and penicillin/streptomycin. For synchronization in G1-phase, cells were grownin media without serum supplementation for 48 hours. Normalmurinemammary gland cells expressing FUCCI (nMuMG-FUCCI;Riken Institute) were grown in DMEM (high glucose) supple-mented with 10% FBS, penicillin/streptomycin, and 10 mg/mLinsulin. MCF-10A cells were grown as described previously(19). HeLa (ATCC) and HeLa-FUCCI cells (Riken Institute) weregrown in DMEM (high glucose) supplemented with 10% FBSand penicillin/streptomycin. Synchronization of cells in G2-phase(2 mmol/L thymidine block and release for 6 hours) or G2–M-phases (100 ng/mL nocodazole block) and microtubule nucle-ation assays were performed as described previously (16).

Flow cytometry analysisCells were collected, suspended in PBS, and fixed in ice-cold

ethanol at �20�C. Ethanol-fixed samples were washed with cold1% BSA in PBS and then suspended in 1% BSA in PBS containing30 mg/mL propidium iodide and 40 mg/mL RNase A, in the dark atroom temperature for 30 minutes. Cells were analyzed using theAccuri C6 flow cytometer (BD Biosciences).

Real-time PCRRNAwas extracted using TRIzol (Invitrogen, 15596026), quan-

tified with NanoDrop, and converted to cDNA using the High-Capacity cDNA Reverse Transcriptase Kit (Applied Biosystems,4368813) as per the manufacturer's protocols. Primers arelisted in Supplementary Table S1 [generously provided by BradHoffman (University of British Columbia, British Columbia,Canada)]. Real-time PCR reactions were run in triplicate with anApplied Biosystems 7000 series machine (Invitrogen). Resultswere analyzed using the DDCt method. Expression of each tran-script was normalized to TATA box binding protein (TBP) levels;expression levels of each transcript/gene measured in G1-phaseenriched 4T1-luc2 cells were normalized to levels measured inasynchronously growing cells.

Scratch wound closure assayCells were grown in standard growth medium until 100%

confluence and then serum starved for 72 hours. Prior to imaging,a wound mark was made and the media were replaced withstandard growth media. Cells were imaged using an ImageXpressMicro High Content Screening System (Molecular Devices).Image analysis was performed using MetaXpress and ImageJ.

Virus packaging and transductionshRNA sequences targeting HMMR (50-CGTCTCCTCT-

ATGAAGAACTA-30 and 50-GCCAACTCAAATCGGAAGTAT-30;

Engraftment and Migration Require Aurora-A

www.aacrjournals.org Mol Cancer Res; 16(1) January 2018 17




Sigma-Aldrich) were packaged in lentivirus using psPAX2 andpMD2.G (psPAX2 and pMD2.G were gifts from Didier Trono;Addgene plasmids #12260 and 12259). Production and collec-tion of lentiviral particles was performed as described previously(19). Virus was added to media for 24 hours, the media werereplaced, and cells were analyzed 72 to 96 hours later.

Constructs and transfectionOn-target plus siRNA (Dharmacon) and scrambled siRNAwere

used as described previously (16). Plasmids used were: GFP-HMMRWT and GFP (16); GFP-HMMRT703A was created fromGFP-HMMRWT cDNA using a QuikChange Site-Directed Muta-genesis Kit (Agilent, 200515). Transfection of DNA and siRNAused JetPrime (Polyplus Transfection, 114-07) following themanufacturer's protocols. Cells were harvested or analyzed 96hours posttransfection with siRNA (24 hours after rescue con-struct transfection).

ImmunofluorescenceCells were fixed with ice-cold methanol at �20�C for 15

minutes and blocked with 3% BSA in PBST. Primary antibodieswere incubated for 2 hours at room temperature, or overnight at4�C, and secondary antibodies were incubated for 1 hour atroom temperature. The following primary antibodies were used:Alexa-647-b-tubulin (TUBB; Thermo Fisher Scientific, MA5-16308-A647, 1:100), phosphorylated Aurora-A (pAurora-A; CellSignaling Technology, 3079, 1:1,000), DDK (Origene, TA50011-100, 1:1,000), phosphorylated HMMR (pHMMR)(1:7,500;ref. 19), Ran-binding protein 2 (RANBP2;ms, Abcam, ab111811,1:200), RANBP2 [rb, 1:2,000; generously provided by Dr. MaryDasso (National Institute for Child Health and Human Devel-opment, Rockville, MD) and Dr. Frauke Melchior (Zentrum f€urMolekulare Biologie der Universit€at Heidelberg, Heidelberg,Germany)], TPX2 (Novus, mb500179, 1:1,000), and g-tubulin(TUBG1; Sigma, T6557, 1:1,000). Antibodies conjugated withAlexa-488, Alexa-549, and Alexa-647 (Life Technologies) wereused as secondary antibodies. Coverslips were mounted withProlong Gold antifade reagent with DAPI (Life Technologies,P36935), and images were acquired with a confocal microscope(FluoView Fv10i, Olympus or Leica SP8, Leica). For the nuclearimport assay, cells were visualized using a Fluoview FV1000confocal laser-scanning microscope (Olympus). The quantita-tion of the ratio of intensity in the nucleus and cytoplasm wasmeasured as described previously (21). Image analysis wasperformed using ImageJ.

Immunoprecipitation and Western blot analysisHeLa cells were synchronized at G2-phase or at G2–M-phase

and homogenized with lysis buffer with a phosphatase inhib-itor (PhosSTOP; Roche, 04906845001) and cOmplete proteaseinhibitor cocktail (Roche, 11697498001). Immunoprecipita-tion with IgG control antibodies or antibodies targetingpHMMR, or importin-a, was carried out overnight at 4�C,followed by incubation with protein A/G beads at 4�C (SantaCruz Biotechnology, sc-2003). Protein A/G beads were washedwith lysis buffer three times. Bound proteins were separated bySDS-PAGE and analyzed by Western blot analysis with thefollowing antibodies: actin (Sigma, a5060), Aurora-A (Abcam,ab13824), pAurora-A (Cell Signaling Technology), GAPDH(Proteintech, 60004-1-IG), histone H3 (Cell Signaling Tech-nology, 9715), p-histone H3 (Cell Signaling Technology,

9701s), HMMR (Abcam, ab124729), importin (Novus,NB100-79807), and TPX2 (Novus). Secondary antibodies wereHRP conjugated (GE Healthcare).

Preparation of HMMR-depleted cytosolsTo remove HMMR from rabbit reticulocyte lysate (RRL,

Promega, L4960), 50 mL of Dynabeads Protein G suspension(Life Technologies 1003D) was washed three times with 200 mLof PBS, followed by incubation with 2 mg of anti-HMMR/CD168 antibody (Abcam ab124729) in 200 mL PBS with0.02% Tween-20 (PBST) for 2 hours at 4�C. Dynabead com-plexes were washed twice with 200 mL of PBST and thencrosslinked with 400 mL of RRL for 2 hours at 4�C. After theincubation, the RRL was removed from the beads, analyzed byWestern blot for successful depletion of HMMR, and used toperform nuclear import assay.

Nuclear import assay in digitonin-permeabilized HeLa cellsBSA covalently attached to the NLS of SV40T antigen

(CGGGPKKKRKVED; NLS-BSA) was custom made (GenScript).NLS-BSA was labeled with Cy3 (Amersham Biosciences,PA33000) according to the manufacturer's protocol.

Adherent HeLa cells grown as monolayers on coverslips werepermeabilized with 40 mg/mL digitonin (Sigma-Aldrich, D141)in transport buffer (TB: 20 mmol/L HEPES, pH 7.4, 110 mmol/Lpotassium acetate, 1 mmol/L EGTA, 5 mmol/L sodium acetate,2 mmol/L magnesium acetate, and 2 mmol/L dithiothreitol)for 4 minutes. Permeabilized cells were washed with TB andincubated with TB containing 70-kDa Texas Red-labeled dextran(Invitrogen, D1864), Cy3-labeled NLS-BSA, or DDK-taggedTPX2 (TPX2-DDK; OriGene, MR221968) for 30 minutes at37�C in the presence or absence of 20% RRL (Promega)or HMMR-depleted RRL, an energy-regenerating system(0.4 mmol/L ATP, 0.45 mmol/L GTP, 4.5 mmol/L phospho-creatine, and 18 U/mL phosphocreatine kinase; Sigma-Aldrich),cOmplete-Mini/EDTA-free Protease Inhibitor Cocktail (Roche)at 10 mg/mL, and 1.6 mg/mL of BSA. After incubation, cells werewashed with TB three times and fixed with 4% paraformalde-hyde for 10 minutes. Cells incubated with Texas Red dextran orCy3-labeled NLS-BSA were washed with TB three times andmounted with Prolong Gold antifade reagent with DAPI. Cellsincubated with TPX2-DDK were permeabilized with 0.2%Triton X-100 for 5 minutes and prepared for immunofluores-cent microscopy as indicated above.

Patient information and tissue microarrayThe cohort comprises breast cancer patients newly diagnosed

between 1986 and 1992 referred to the British Columbia CancerAgency and has been described previously (22–24). Detaileddescription of the demographic, pathologic, and treatment char-acteristics are described in Supplementary Table S2. Clinicopath-ologic information including staging, treatment, and follow-upinformation was available with amedian follow-up of 12.3 years.Additional information includes histology, grade, tumor size,lymphovascular invasion (LVI), ER status, axillary lymph nodeinvolvement, type of systemic therapy, dates of diagnosis, recur-rences (local, regional, or distant), or death. The Clinical ResearchEthics Board of the BritishColumbiaCancer Agency approved thisstudy. Two pathologists reviewed hematoxylin and eosin slidepreparations from these blocks to identify areas of invasivecarcinoma for inclusion into 0.6-mm core tissue microarrays.

Chu et al.

Mol Cancer Res; 16(1) January 2018 Molecular Cancer Research18




IHC stainingAbundance of pHMMR was analyzed using a rabbit antihu-

man antibody (described in ref. 19) on 4-mm sections of the 17tissue microarray blocks using previously described IHC meth-ods (25). Tissue microarrays were scored visually by threeobservers, with the third observer being the arbitrator whenthe first two scores were discordant, using a scoring system thatcaptured intensity of staining ranging from 0 (no/low stain-ing), 1 (moderate), and 2 (strong). IHC methods for ER,progesterone receptor (PR), HER2, and Ki67 were as describedpreviously (26).

Statistical analysisStatistical analysis was performed using GraphPad Prism v5.01

for Windows (Graphpad Software). Pairwise comparisons weremade using unpaired Student t test. Comparisons of multiplegroups were made using one-way ANOVA with a Bonferroniposttest. Colocalization was measured using the M1 colocaliza-tion coefficient (TPX2/RANBP2).

For the tissue microarray, data distribution in groups andsignificance between different conditions was analyzed byusing an unpaired t test (95% confidence interval, two tailed)or other comparison tests as indicated in GraphPad Prismsoftware. P < 0.05 was considered statistically significant. Theentire cohort was randomly divided into a training set and avalidation set. Complete data were available for 1,583 and1,592 cases from the training and validation sets, respectively(N ¼ 3,175). Prespecified analyses for pHMMR associationwith clinicopathologic variables and outcome were initiallyconducted on the training set, confirmed on the validation set,and then performed on the entire cohort as presented in thisstudy. Exploratory analyses in the different treatment subsetswere done to further assess the association of pHMMR abun-dance with survival.

All statistical analyses were performed using SPSS 18.0 and R2.15.0. c2 analysis was used to test associations of pHMMRabundance (as a categorical variable) with age, menstrual status,nodal status, local treatment and systemic treatment, tumor size,grade, histologic type, LVI, and expression of ER, PR, HER2, Ki67(<14% vs.�14%), andmolecular subtypes by IHC (27). LuminalA tumors were defined as ER or PR positive, negative for HER2,and low Ki67. Luminal B subtypes were all tumors positive eitherfor ER or PR as well as high Ki67 but negative for HER2. Luminal/HER2 positive subtype was HER2 positive as well as positive foreither ER or PR. HER2 subtypes were all those exclusively positivefor HER2 with hormone receptor negativity. Tumors negative forall three receptors, ER, PR, and HER2, but positive for either ofCK5/6 or EGFR were defined as basal-like subtype. Systemictherapy was categorized as no adjuvant systemic therapy (noAST), tamoxifen only, chemotherapy only, and tamoxifen withchemotherapy.

Univariate survival analyses were performed using the Kaplan–Meier method, and survival differences were estimated using thelog-rank test. For multivariate analysis, a Cox proportionalhazards ratio model was used to estimate the adjusted HRsignificance. The primary endpoint for survival analyses wasBCSS, defined as the time from date of diagnosis of primarybreast cancer to date of death due tobreast cancer as the primary orunderlying cause. RFS, defined as time from date of diagnosis ofprimary breast cancer to date of first local, regional, or distantrecurrence, was used as a secondary endpoint.

ResultsCells in G1-phase engraft less efficiently than asynchronouslygrowing cells

Triple-negative breast cancer cells engaged in the cell cyclemetastasize better than quiescent cells in patient-derived xenograftmodels (9). We tested whether a particular phase of the cell cycleallows for better engraftment of murine breast cancer cells insyngeneic animals. A total of 10,000 4T1-luc2 cells that were eithergrowing asynchronously or G1-phase enriched (SupplementaryFig. S1A) were injected into the tail vein of BALB/c mice. Usingluciferase activity, we monitored engraftment and tumor growthover 3 weeks. 4T1-luc2 cell populations enriched for cells in G1-phase were significantly impaired in their ability to engraft BALB/cmice (Fig. 1A and B). We performed similar experiments byinjecting asynchronously or G1-phase–enriched 4T1-luc2 cells intothe tail vein of NSG mice (Supplementary Fig. S1B). Again, G1-phase enriched cells injected in NSG mice were significantlyimpaired in the kinetics of engraftment and exhibited an approx-imately 1-week delay in tumor growth (Fig. 1C). We examinedwhether the reduction in engraftment potential observed in G1-phase enriched 4T1-luc2 cells correlated with changes to theexpression of defined EMT markers, including Cdh1, Snai1, Snai2,vimentin, orCdh2. The expression level forHmmr,which is elevatedduring G2–M-phase (28), was included as a positive control.Relative to the expression observed in asynchronously growingcells, G1-phase enriched cells express elevated levels of the epithe-lial cell marker Cdh1 and 4- to 10-fold lower levels of Snai1, Snai2,or vimentin, markers for mesenchymal cells, and Hmmr, a prolif-eration marker (Fig. 1D). These data indicate that the enrichmentof 4T1-luc2 cells in G1-phase dampens their expression of EMTmarkers and severely diminishes their engraftment potential.

Cells in S–G2-phase close a wound more efficiently thancells in G1-phase

To determine whether G1-phase cells are impaired in migra-tory capacity measured in vitro, we utilized live cell imagingto follow the kinetics of wound closure for nonmalignant,murine mammary epithelial cells expressing a cell-cycle indi-cator (nMuMG-FUCCI; Fig. 2A). nMuMG-FUCCI cells wereseeded in 96-well plates, grown to confluence, and serumstarved to synchronize at G1-phase. Wounds were introducedand complete closure was observed 48 hours following thescratch (Fig. 2B). We noted at 24 hours postwounding that thecells occupying the leading edge and wound area were largelyin S–G2-phase, while cells more distal from the scratch werelargely in G1-phase (Fig. 2C and D), suggesting that either G1-phase cells at the edge of the wound transition into S–G2-phasemore quickly than more centrally located cells or that S–G2-phase cells have a migratory advantage. To distinguish betweenthese possibilities, individual cells at the wound edge weretracked for the first 24 hours postwounding using a wide-field,high-content cell imaging system with multiparameter assess-ment of cell-cycle phase, wound closure, and migration veloc-ity (Fig. 2E). The region of the wound was defined, and onlycells (indicated by masked nuclei) within this region of interestwere measured. Based upon the FUCCI signal, cells wereclassified as those that remained within one phase (cells 1and 4, Fig. 2E) and those that transitioned through cell-cyclephases during the imaging window (cells 2 and 3, Fig. 2E). Wefound that cells in S–G2-phase migrated with approximately






1.7-fold greater velocity than neighboring cells that remainedin G1-phase (14.1 � 4.8 mm/hour vs. 8.8 � 4.1 mm/hour, P <0.0001; Fig. 2F). When we restricted the analysis to individualcells that transitioned from G1-phase to S–G2-phase duringimaging, we confirmed a significant increase in migrationvelocity during S–G2-phase (13.2 � 3.7 mm/hour vs. 9.9 �3.9 mm/hr, P < 0.001; Fig. 2G). We repeated these studies usingmalignant, cervical carcinoma HeLa-FUCCI cells and found acongruent correlation between progression into S–G2-phaseand elevated migration velocity (Supplementary Fig. S2).

Cell cycle–specific microtubule organization promotes cellmigration and polarity

For mechanistic insight into why S–G2-phase cells maymigrate more efficiently than cells in G1-phase, we incubated

nMuMG-FUCCI cells with thymidine, or equivalent DMSOcontrol, to prevent their entry into G2-phase during woundclosure. Relative to control-treated cells, wound closure wasimpaired by nearly 45% for thymidine-synchronized cells(Fig. 3A and B). Next, we augmented the frequency ofnMuMG-FUCCI cells in G2-phase by incubation with noco-dazole, which depolymerizes microtubules and prevents entryinto mitosis. Although we observed a robust enrichment ofG2-phase cells following nocodazole treatment, wound clo-sure was abolished (Fig. 3A and B), indicating a requirementfor the microtubule cytoskeleton. Thus, cell-cycle progressioninto G2-phase is strongly correlated with an increased velocityduring wound closure assays with nMuMG-FUCCI cells, andthe microtubule cytoskeleton functions as a critical mediatorfor migration.

0 6 9 14 18

A B

C D

ASG1

Days postengraftmentCdh1

Cdh2Snai1

Snai2

Vimen

tinHmmr

Expr

essi

on(n

orm

aliz

ed to

TB

P an

d A

S) ASG1

0.000.250.500.751.001.251.501.752.00

Day 0

ASG1

BALB/c

BALB/cA

sync

hron

ous

G1-

phas

e en

riche

dPh

oton

/sec

NOD scid gamma1011

1010

1009

1008

1007

1006

1005

0 9 13 17 21Days postengraftment

Phot

on/s

ec

1010

1009

1008

1007

1006

1005

P < 0.0001

P < 0.0001

Day 9 Day 21

Day 0 Day 9 Day 21

Figure 1.

Cells in G1-phase engraft less efficiently than asynchronously growing cells. A, 4T1-luc2 cells were G1-phase enriched or grown asynchronously, and 10,000cells were injected into BALB/c mice. Representative images are shown for monitoring over 3 weeks of tumor engraftment. B, Light emitted byasynchronously growing (AS) or G1-phase–enriched 4T1-luciferase tumors grown in BALB/c mice was measured at the indicated times postengraftment.Data are represented as mean � SD for 3 mice injected with AS cells and 4 mice injected with G1-enriched cells. P value derived by two-way ANOVA. C, Lightemitted by asynchronously growing (AS) or G1-phase enriched 4T1-luciferase tumors grown in NOD scid gamma mice was measured at the indicatedtimes postengraftment. Data are represented as mean � SD for 5 mice injected with AS cells and 5 mice injected with G1-enriched cells. P valuederived by two-way ANOVA. D, Expression levels of EMT markers, including Cdh1, Cdh2, Snai1, Snai2, and vimentin, and the proliferation markerHmmr were assessed in G1-phase enriched and asynchronously growing (AS) 4T1-luc2 cells. Expression levels were measured by real-time PCR andnormalized to the level of expression of TATA box binding protein (TBP) within each sample. Expression levels in G1-enriched samples werenormalized to levels in appropriate AS sample and plotted as mean � SD, n ¼ 2 experiments.

Chu et al.





By imaging wounds during the first 24 hours of closure, wewere able to measure the velocity and directionality of migra-tion for individual cells (Fig. 3C). Consistent with the kinetics

of wound closure (Fig. 3B), thymidine-synchronized, G1-phasecells and nocodazole-treated, G2-phase cells migrated with asignificantly reduced mean velocity relative to DMSO-treated,

Edge Central

24 h

0 h

24 h

48 h

CentralEdge

P < 0.02P < 0.002

A

C D

E F G

G1S–G2

S–G

2-ph

ase

G1-

phas

e

0 h

0–24 h

Wound24 h

Wound

S–G2–MG1

Wound

0 h

11’

22’

3

3’4

4’

0–24 h

Masked Time-lapse

BulkG1 S–G2

PairedG1 S–G2

Hours0 24

Hours0 24

Cel

l-cyc

le s

tage

(%)

B

Wou

nd c

losu

re (%

)

Hours

P < 0.0001 P < 0.001

100

80

60

40

20

00 16 24 36 48

100

80

60

40

20

0

Cel

l-cyc

le s

tage

(%)

100

80

60

40

20

0Ve

loci

ty (m

m/h

)

Velo

city

(mm

/h)

50

40

30

20

10

0

50

40

30

20

10

0

Figure 2.

Epithelial cells close a wound more efficiently in S–G2-phase. A, Representative images of a wound closure assay for nMuMG-FUCCI cells imaged every30 minutes for 48 hours. Red box indicates region depicted in panels on the right hand side. White triangles indicate the wound position and thewhite lines in the 24-hour image indicate the leading edges of the closing wound. Scale bar, 200 mm. B, Kinetics of wound closure for nMuMG-FUCCIcells. Data are represented as mean � SD, n ¼ 7 experiments. C, Representative image of a wound closure assay 24 hours postwound for nMuMG-FUCCIcells. Red boxes indicate edge and central regions used for quantification in D. Scale bar, 200 mm. D, Quantification of the percentage of nMuMG-FUCCIcells in G1-phase (red) or S–G2-phase (green) at the leading edge and a central region during wound closure. Data are represented as mean � SD,n ¼ 2 experiments (central region) and 3 experiments (leading edge). P values from Student t test. E, nMuMG-FUCCI cells were imaged every30 minutes for 24 hours following wounding, and cell velocity was tracked and quantitated by MetaXpress software. Individual time points are overlaidin the 0- to 24-hour time-lapse image and in the masked image. White triangles indicate the wound position. Cells 1 and 4 remain in one phase, whilecells 2 and 3 transition from G1-phase (red) into and through S–G2-phase (green). Scale bar, 100 mm. F, Velocity of G1-phase (red) and S–G2-phase(green) nMuMG-FUCCI cells during wound closure. Data are presented as box and whisker graph displaying the median, 25–75 percentiles, and 1 and99 percentiles, n ¼ 45 cells (G1), 42 cells (S–G2) from two experiments. P value from Student t test. G, Velocity of nMuMG-FUCCI cells as they passfrom G1-phase (red) into and then through S–G2-phase (green). Data are presented as box and whisker graph displaying the median, 25–75 percentiles,and 1 and 99 percentiles, n ¼ 34 cells from two experiments. P value from Student t test.






E DMSO (S–G2-phase) Nocodazole (G2-phase)Thymidine (S-phase)

52.6% ±6.6 6.9% ±2.1

19.6% ±5.6

20.8% ±4.2

32.9% ±1.4

31.5% ±3.8

22.2% ±3.4

13.4% ±3.889.4% ±6.0

3.9% ±2.1

6.8% ±4.9

0.0%-250

-100

-50

50

100

-250

-100

-50

50

100

-250

-100

-50

50

100

G1TUBG1DAPI

G1

G2

G2

F G H I

TUBB

Scra

tch

18ho

urs

G1S–G2TUBG1

f

f s

s

ss

cr

r

f

f

s s

sf

sf

f

ff

f

ff

ff

c

ff

s

s

r

DM

SO

Noc

odaz

ole

A

C D

DMSO Thymidine Nocodazole

G1S–G2TUBG1

G1S–G2

G1 S–G2

Mic

rotu

bule

leng

th (m

m)

0 h

24 h

Wound Wound

0 h

24 h

Wound Wound

0 h

24 h

Wound Wound

Wound0–24 h

Time-lapse0–24 h

Time-lapse

(vs. DMSO, P < 0.001) (vs. DMSO, P < 0.05)

B

Wou

nd c

losu

re (%

)

0

20

40

60

80

DMSO Thy. Noc. DMSO Thy. Noc.0

5

10

15

20

Velo

city

(mm

/h)

0

20

40

60

80

100

Front Side Rear CenterCentrosome position

Cel

l pro

port

ion

(%)

0

2

4

6

P < 0.0001P < 0.001

P < 0.001 P < 0.0001

P = 0.01 P < 0.01

Thym

idin

e

0–24 h

Time-lapse

Figure 3.

Cell cycle–specific microtubule organization promotes cell migration and polarity. A, Representative images of scratch wound closure assay fornMuMG-FUCCI cells treated with DMSO (control), thymidine, or nocodazole and imaged every 30 minutes for 24 hours. White triangles indicate thewound position and the white lines in the 24-hour images indicate the leading edges of the closing wound. Scale bars, 200 mm. B, Quantification ofwound closure at 24 hours postwounding for nMuMG-FUCCI cells treated with DMSO (control), thymidine, or nocodazole. Data are represented asmean � SD, n ¼ 8 experiments. P values from Student t test. C, Representative images for scratch wound closure assays in nMuMG-FUCCI cells treatedwith DMSO (control), thymidine, or nocodazole and imaged every 30 minutes for 24 hours. Individual time points are overlayed in the 0- to 24-hourtime-lapse images. White triangles indicate the original position of the wound. Scale bar, 50 mm. D, Velocities for nMuMG-FUCCI cells treated withDMSO (control), thymidine, or nocodazole during wound closure. Data are presented as box and whisker graph displaying the median, 25–75 percentiles,and 1 and 99 percentiles, n ¼ 7 experiments (DMSO, nocodazole) and 8 experiments (thymidine). (Continued on the following page.)

Chu et al.





S–G2-phase cells (Fig. 3D). Moreover, the migration of DMSO-treated cells was highly biased to be directed toward the woundas indicated by plotting the displacement of individual cellsalong the x-axis and y-axis (Fig. 3E). At the end of 24 hours,89.4% � 6.0% of DMSO-treated cells were displaced within aquadrant defined by �45-degree angles from their origin.However, the displacement of thymidine-synchronized, G1-phase cells and nocodazole-treated, G2-phase cells was far lessdirectional (Fig. 3E).

Movement of the centrosome, or microtubule-organizing cen-ter, toward the leading edge is a critical step in the front-rearpolarization, scattering, and migration of certain cell types, includ-ing epithelial cells (4, 29). Moreover, augmented microtubulenucleation at supernumerary centrosomes is known to triggerinvasive behavior in mammary epithelial cells (11). Therefore,we examined whether G1-phase and S–G2-phase nMuMG-FUCCIcells differ in centrosome polarity and microtubule nucleationcapacity during wound closure. At 18 hours postwounding,nMuMG-FUCCI cells were fixed and immunostained for TUBG1.The frequency of front-polarized TUBG1-positive centrosomes(i.e., those positioned between the nucleus and wound, Fig. 3F)was augmented in S–G2-phase cells (Fig. 3G). Next, we treatedscratch wound assays with 15micromol/L nocodazole for 60min-utes at 18 hours postwounding to depolymerize microtubulesin nMuMG-FUCCI cells. Cells were washed and microtubuleswere allowed to polymerize for 2 minutes prior to fixation(Fig. 3H). The microtubule fibers regrown in nMuMG-FUCCI cellslocated at the leading edge were significantly longer in S–G2-phasecells than in G1-phase cells (Fig. 3I), indicating an augmentedmicrotubule nucleation capacity in S–G2-phase. Together, thesedata indicate fundamental differences in centrosome orientationand microtubule nucleation capacity between S–G2-phase cellsand G1-phase cells during wound closure assays, which may altertheir relative abilities to break epithelial cohesion and migrateinto a wound.

Aurora-A activity enables cell polarity and migrationAurora-Apromotesmicrotubule nucleationwithinmitotic cells

(30). To assess whether active Aurora-A is required for the migra-tion of nonmitotic cells duringwound closure, wefirst establishedthat 2.5 mmol/L MLN8237 (a specific small-molecule Aurora-Ainhibitor) was a minimal dose needed to reduce either the levelsof phosphorylated histoneH3 (Supplementary Fig. S3A), a kinasesubstrate, or the levels of the autophosphorylated kinase (Sup-plementary Fig. S3B) in lysates prepared from nocodazole-syn-chronized nMuMG-FUCCI or MCF-10A cells, respectively. Thetreatment of nMuMG-FUCCI cells with 2.5 mmol/L MLN8237 forthe first 24 hours following wounding was sufficient to signifi-

cantly reduce the area of wound closure (Fig. 4A and B). Thetreatment of nMuMG-FUCCI cells with 2.5 mmol/L MLN8237,however, did not significantly change the proportion of S–G2-phase cells at the leading edge (Fig. 4C, P¼ 0.13), indicating thatinhibition of Aurora-A dampens migration without altering cell-cycle progression into G2-phase. We next analyzed the kineticsand directionality of cell migration for the first 24 hours in thepresence of DMSO or MLN8327. Aurora-A inhibition not onlydiminished the velocity (Fig. 4D) but also dampened the direc-tionality of nMuMG-FUCCI cellmigration (Fig. 4E). These alteredmigration phenotypes induced through Aurora kinase inhibitionwere accompanied by a significantly reduced frequency of front-polarized centrosomes (Fig. 4F and G) and diminished microtu-bule nucleation capacity at the centrosome (Fig. 4H and I) whenthe analyses were restricted to S–G2-phase cells at the leadingedge. Thus, Aurora-A activity enables microtubule nucleation atcentrosomes, centrosome polarization, and directed migration inscratch wound closure assays.

HMMR promotes phosphorylated Aurora-A levels and cellmigration

HMMR localizes TPX2 during mitosis and acts as anupstream regulator of Aurora-A activity that enables microtu-bule nucleation at duplicated centrosomes and near chromo-somes (16, 17). As TPX2 enhances the accumulation of Aurora-A during G2-phase (31) and Aurora kinase activity at centro-somes (20), we postulated that HMMR may also promoteAurora-A activity and cell migration during G2-phase. To testthis postulate, we silenced HMMR by separately treating HeLacells with scrambled siRNA control or siRNA targeting the 50- or30- untranslated regions (UTR) in HMMR (Fig. 5A), whichenabled rescue with cDNA lacking those UTRs. Alternatively,we treated HeLa cells with lentivirus encoding a nonhairpinshRNA control or shRNA targeting HMMR (Fig. 5B). Followingthese treatments, we determined that HMMR was dispensablefor the centrosome localization of pAurora-A (Fig. 5C), but thelevels of the active kinase were dampened in HMMR-silencedcells relative to control-treated cells (Fig. 5D). To determine theeffect of this dampened activity on the cell's capacity to nucleatemicrotubules, microtubule regrowth assays were performed 3days following transduction with lentivirus. We restricted ouranalysis of microtubule regrowth capacity to cyclin B1–posi-tive, G2-phase cells (Fig. 5E) and found that silencing HMMRresulted in a significant reduction in the length of microtubulefibers at the centrosome (Fig. 5F). Moreover, the fraction ofcyclin B1–positive, G2-phase cells with front-polarized centro-somes (Fig. 5G) were also reduced in HMMR-silenced cellpopulations (Fig. 5H), which correlated with significantly

(Continued.) P values from Student t test. E, Direction traveled by nMuMG-FUCCI cells treated with DMSO (control), thymidine, or nocodazoleduring wound closure assays. The final position of migrating cells is plotted relative to their origin (0,0) along the x-axis and y-axis. Quadrants weredefined by � 45-degree angles from their origin. The mean proportion in the front polarizing quadrant was compared between DMSO-treated,thymidine-treated, and nocodazole-treated groups, n ¼ 3 experiments. P values from Student t test. F, Representative images at the time of thewounding and 18 hours postwounding that indicate the position of TUBG1-positive centrosomes relative to the nucleus and wound in nMuMG-FUCCI cellsat the leading edge. Images are labeled with the following designations of centrosome position: Front, f; side, s; center, c; rear, r. Scale bar, 20 mm.G, Quantification of centrosome position in G1-phase (red) and S–G2-phase (green) nMuMG-FUCCI cells at 18 hours postwounding. Data arerepresented as mean � SD, n ¼ 4 experiments. P value from Student t test. H, nMuMG-FUCCI cells were treated with nocodazole at 18 hourspostwounding, washed, and microtubules were allowed to regrow for 2 minutes. Representative images are shown for microtubule fibers (TUBB) atTUBG1-positive centrosomes in G1-phase and S–G2-phase nMuMG-FUCCI cells after microtubule regrowth assay. Scale bar, 10 mm. I, Quantificationof microtubule fiber length in G1-phase and S–G2-phase nMuMG-FUCCI cells. Data are presented as box and whisker graph displaying the median,25–75 percentiles, and 1 and 99 percentiles, n ¼ 4 experiments. P value from Student t test.






DMSO

2.5 mmol/L

5 mmol/L

MLN

8237

G1

S–G2

A B C

D E

F G H I

MLN8237 (S–G2-phase)DMSO (S–G2-phase)

-250

-100

-50

50

100

49.3% ± 2.3

(vs. DMSO, P < 0.001)

28.7% ±5.5

5.3% ±2.9

16.7% ±3.9

G1S–G2TUBG1

F

F

F

FC

C

SS

R

DM

SO

FS

SF S

SC

RF

F

MLN

8237

TUBG1DAPI

TUBB

G2

MLN8237DMSO

TUBB

DMSO MLN8237

DMSOMLN8237

Cel

l pro

port

ion

(%)

TUBG1DAPI

Mic

rotu

bule

leng

th (m

m)

DMSO 2.5 5.00

20

40

60

80

100

0

5

10

15

Velo

city

(mm

/h)

-250

-100

-50

50

100

96.7% ±5.8

2.4% ±4.1

0.95% ±1.7

0.00%

DMSO MLN8237

0

4

6

8

2Cel

l pro

port

ion

(%)

Front Side Rear CenterCentrosome position

DMSO 2.5 5.0MLN8237 (mmol/L) MLN8237 (mmol/L)

Wou

nd c

losu

re (%

)

P < 0.001

P < 0.05

P < 0.001

P < 0.005

P < 0.001

0

20

40

60

0

20

40

60

80

G1S–G2TUBG1 G2

G2

G2

Figure 4.

Aurora-A is required for cell migration and microtubule nucleation. A, Representative images for wound closure assay at 24 hours postwounding during whichnMuMG-FUCCI cells were treated with DMSO or MLN8237 (Aurora-A inhibitor) at the indicated doses. White line indicates leading edge. Scale bar, 200 mm.B, Quantification of wound closure at 24 hours postwounding for nMuMG-FUCCI cells treated with DMSO or MLN8237. Data are represented as mean � SD,n ¼ 8 (DMSO), 6 (2.5 mmol/L), and 4 (5 mmol/L) experiments. P values from Student t test. C, Quantification of the proportion of nMuMG-FUCCIcells at the leading edge in G1-phase or S–G2-phase. Cells were treated with either DMSO or MLN8237 at the indicated doses. Measurements were made at24 hours postwounding. Data are represented as mean � SD, n ¼ 2 experiments. P value from one-way ANOVA. D, Velocity of nMuMG-FUCCI cellsduring wound closure and treated with either DMSO or MLN8237 (2.5 mmol/L). Data are presented as box and whisker graph displaying the median,25–75 percentiles, and 1 and 99 percentiles, n ¼ 8 (DMSO) and 7 (MLN8237) experiments. P value from Student t test. E, Direction traveled by nMuMG-FUCCIcells treated with DMSO or MLN8237 (2.5 mmol/L) during wound closure. The final position of migrating cells is plotted relative to their origin (0,0) along thex-axis and y-axis. Quadrants were defined by � 45-degree angles from their origin. The mean proportion of cells in the front polarizing quadrant wascompared between treatment groups (n ¼ 3 experiments). P value from Student t test. F, The position of TUBG1-positive centrosomes relative to thenucleus and wound for nMuMG-FUCCI cells at the leading edge. Cells were treated with DMSO or MLN8237 (2.5 mmol/L) and fixed at 24 hours postwounding.Images are labeled with the following designations of centrosome position: Front, f; side, s; center, c; rear, r. Scale bar, 10 mm. G, Quantification ofcentrosome position in S–G2-phase (green) nMuMG-FUCCI cells at the leading edge 24 hours postwounding. Cells were treated with DMSO or MLN8237(2.5 mmol/L). Data are represented as mean � SD, n ¼ 3 experiments. P value from Student t test. H, Representative images for microtubule fibers (TUBB)at TUBG1-positive centrosomes in S–G2-phase nMuMG-FUCCI cells after microtubule regrowth assay. nMuMG-FUCCI cells were treated with DMSO orMLN8237 (2.5 mmol/L). At 24 hours postwounding, cells were treated with nocodazole, washed, and microtubules were allowed to regrow for 2 minutes.Scale bar, 10 mm. P value from Student t test. I, Quantification of microtubule length in S–G2-phase nMuMG-FUCCI cells following the microtubuleregrowth assay. Cells were treated with DMSO or MLN8237 (2.5 mmol/L). Data are presented as box and whisker graph displaying the median,25–75 percentiles, and 1 and 99 percentiles, of microtubule length measurements, n ¼ 5 experiments. P value from Student t test.

Chu et al.





impaired wound closure measured at 24 hours (Fig. 5I and J)and dampened migration velocity (Fig. 5K). Together, thesedata indicate that the migration of G2-phase epithelial cells

requires Aurora-A activity, which can be reduced by treatmentwith a small-molecule inhibitor (MLN8237) or through thesilencing of HMMR, an upstream molecular regulator.

Scra

tch

24 h

0

2

4

6

8

10

A B C D

E F G H

I J K

CyclinB1DAPI

rs

r

TUBBTUBB

ff

sr

CyclinB1DAPI

TUBB

CyclinB1DAPI

CyclinB1DAPI

TUBB

Scr HMMR Scrambled HMMRNHP HMMR

Nonhairpin HMMR

NHP HMMR

Nonhairpin HMMRM

icro

tubu

le le

ngth

(mm

)

pAurora-A pAurora-ATUBB TUBB

Scr HMMR

pAur

ora-

A le

vels

(rel

ativ

e to

TU

BB

)

0

20

40

60

80

100

Front Side Rear Center

NHPshHMMR

Cel

l pro

port

ion

(%)

Wou

nd c

losu

re (%

)

0

20

40

60

80

0.0

0.2

0.4

0.6

0.8

0

5

10

15

20

Velo

city

(mm

/h)

P < 0.01

P < 0.001P < 0.0001

P < 0.0001

HMMR

Actin

HMMR

Actin

Nonhairpin HMMR

NHP HMMR NHP HMMR

G2 G2

P = 0.08

Figure 5.

HMMR augments phosphorylated Aurora-A, microtubule nucleation, and migration. A, Western blot analysis of HMMR abundance in HeLa cell populationstransfected with scrambled (Scr) siRNA or siRNA targeting HMMR. Lysates were prepared 72 hours after transfection. Actin served as a loadingcontrol. B, Western blot analysis of HMMR abundance in HeLa cell populations transduced with lentivirus encoding nonhairpin (NHP) shRNA or shRNAtargeting HMMR. Lysates were prepared 72 hours after transduction. Actin served as a loading control. C, Representative images of phosphorylated Aurora-A(pAurora-A) location and levels in HeLa cells treated with scrambled siRNA or siRNA targeting HMMR. Centrosomes, indicated by b-tubulin (TUBB), areboxed and enlarged beneath the panel. Scale bar, 10 mm. D, Quantification of the levels of pAurora-A at centrosomes, normalized to the levels of b-tubulin,in HeLa cells treated with scrambled siRNA or siRNA targeting HMMR. Data represented as mean � SE, n ¼ 3 experiments. P value from Student t test.E, Representative images of microtubule (TUBB) growth in cyclin B1–positive, G2-phase, HeLa cells transduced with lentivirus encoding nonhairpinshRNA or shRNA targeting HMMR. Scale bars, 10 mm. F, Quantification of microtubule fiber length in cyclin B1–positive, G2-phase, HeLa cells transducedwith lentivirus encoding nonhairpin (NHP) shRNA or shRNA targeting HMMR. Data are presented as box and whisker graph displaying the median,25–75 percentiles, and 1 and 99 percentiles from 47 cells (NHP) or 52 cells (HMMR) across two experiments. P value from Student t test. G, Representativeimages of centrosome position indicated by TUBB immunofluorescence in cyclin B1–positive, G2-phase, HeLa cells transduced with lentivirus encodingnonhairpin shRNA or shRNA targeting HMMR. Wound closure assays were fixed at 24 hours postwounding. Scale bar, 10 mm. H, Quantification ofcentrosome position in cyclin B1–positive, G2-phase HeLa cells transduced with lentivirus encoding nonhairpin (NHP) shRNA or shRNA targeting HMMR.Wound closure assays were fixed at 24 hours postwounding, and centrosome position relative to the wound was measured in cells at the leading edge.Data represented as mean � SD, n ¼ 2 experiments. I, Wound closure for HeLa cells transduced with lentivirus encoding nonhairpin shRNA or shRNAtargeting HMMR. Yellow lines indicate position of leading edge at time of wounding and 24 hours postwounding. Scale bar, 200 mm. J, Quantificationof wound closure measured at 24 hours postwounding for HeLa cells transduced with lentivirus encoding nonhairpin (NHP) shRNA or shRNA targetingHMMR. Data are presented as box and whisker graph displaying the median, 25–75 percentiles, and 1 and 99 percentiles, n ¼ 8 (NHP) and 11 (HMMR)experiments. P value from Student t test. K, Velocity of HeLa cells at the leading edge measured over the first 24 hours postwounding. Cells weretransduced with lentivirus encoding nonhairpin (NHP) shRNA or shRNA targeting HMMR. Data are presented as box and whisker graph displaying the median,25–75 percentiles, and 1 and 99 percentiles, n ¼ 5 experiments. P value from Student t test.






HMMR-T703 is required for the nuclear localization of TPX2In nonmitotic cells, little is known about the extent or mech-

anism by which TPX2 in the nucleus can activate Aurora-A otherthan the observation that the subcellular localization of TPX2 isaltered in HMMR-silenced cells (19). Indeed, TPX2 colocalizeswith the nucleoporin Nup153 in HMMR-silenced cells (Supple-mentary Fig. S4A), suggesting a role for HMMR in the nucleartransport of TPX2. HMMR is a microtubule-associated proteinand is usually segregated fromTPX2 in the nucleus (28); however,Aurora-A phosphorylates HMMR at threonine-703 (pHMMR)and pHMMR localizes to the nucleus (Supplementary Fig. S4B;ref. 19). To test whether this phosphorylation event in HMMRaffects the nuclear localization of TPX2, we created a GFP-taggedphospho-dead mutant (GFP-HMMRT703A) as well as wild-typeHMMR (GFP-HMMRWT; Supplementary Fig. S4C). We treatedcells with siRNA targeting the 50- or 30-UTRs in HMMR, orscrambled siRNA control, and transiently transfected cells withcDNAs encoding GFP-HMMR variants, which are resistant to theHMMR-targeting siRNA. In scrambled siRNA-treated control cells,the expression of GFP, GFP-HMMRWT, or GFP-HMMRT703A didnot alter the nuclear localization of TPX2 (Fig. 6A). In HMMR-silenced cells, only the expression of GFP-HMMRWT, and not GFPalone or GFP-HMMRT703A, rescued the nuclear localization ofTPX2 (Fig. 6A and B). These data indicate that the nuclearlocalization of TPX2 is promoted by HMMR threonine-703.

To investigate a requirement for HMMR in the nuclear tran-sport of TPX2, we used an in situ nuclear import assay whereinHeLa cells were transfected with siRNA targeting HMMR, orscrambled siRNA, and subsequently permeabilized with digi-tonin to release soluble cytosolic components (32). Recombi-nant flag-tagged TPX2 (TPX-DDK), an energy regenerationmixture, and rabbit reticulocyte lysate, which was immunode-pleted of HMMR (Supplementary Fig. S5A), were then added tothese cells to enable nuclear import. The addition of fluores-cently labeled BSA conjugated with a nuclear localization signal(NLS-BSA) served as a positive control for nuclear import.To monitor nuclear integrity, we confirmed that fluorescentlylabeled dextran remained cytosolic and did not enter thenucleus during the experimental procedures (SupplementaryFig. S5B). Moreover, in the absence of energy and cytosol(demarked No Transport), nuclear transport did not occur foreither TPX2-DDK (Fig. 6C) or NLS-BSA (Supplementary Fig.S5C). When scrambled siRNA control-treated cells were incu-bated cytosol and energy (demarked Transport), both NLS-BSA(Supplementary Fig. S5C) and TPX2-DDK (Fig. 6C) were effi-ciently transported into the nucleus. However, the levels ofTPX2-DDK in the nucleus, measured as a ratio with the levelsin the cytoplasm, were quantitatively reduced in HMMR-silenced cells (Fig. 6C and D). HMMR-silenced cells did nothave alterations in the organization of the microtubule cyto-skeleton or the shape and area of the nucleus (SupplementaryFig. S5D and S5E), which may explain the deficient nucleartransport of TPX2-DDK. Similar to the localization of endog-enous TPX2 in HMMR-silenced cells, we also observed anaugmented colocalization of TPX2-DDK with the nucleoporinRANBP2 (Fig. 6E and F) indicative of TPX2-DDK accumulationat the nuclear envelope. We next tested whether complexes ofTPX2 and importin-a were augmented in HMMR-silenced cells.Cell lysates from control-treated or HMMR-silenced cells wereimmunoprecipitated with antibodies targeting importin-a. InHMMR-silenced cell lysates, we observed an overall reduction

in the total levels of TPX2 (input, Fig. 6G); in spite of this, wealso observed an increase in the levels of TPX2 coprecipitatedwith antibodies targeting importin-a from HMMR-silenced celllysates (Fig. 6G). Taken together, our data uncover HMMR-T703as a critical regulator of TPX2 nuclear transport that is needed toreduce TPX2–importin-a complexes at the nuclear envelope.These data support an axis of Aurora-A–TPX2–HMMR in thedetermination of microtubule nucleation and migration innonmitotic cells.

Phosphorylated HMMR is a prognostic marker inER-negative breast cancer

To study the putative significance of the HMMR–TPX2–Aurora kinase A axis in primary breast tumor tissues, weevaluated pHMMR abundance in a large tumor tissue array.Tumor was present for evaluation in 3,175 of the 3,992 cases;817 cases were excluded for no or insufficient tissue present.Phosphorylated HMMR abundance (intensity score �1) wasseen in 58.7% (1863 of 3175) tumors. The intensity of stainingin the tumors ranged from no or weak staining (41.3%), tomoderate staining (44.1%), or strong staining (14.6%; Fig. 7A).A significant positive correlation between pHMMR and Ki67was observed (P < 0.005).

In patients with ER-positive breast cancer, the intensity ofpHMMR staining did not predict BCSS or RFS (Fig. 7B and F).However, pHMMR was a significant predictor of BCSS and RFSin patients with ER-negative breast cancer (Fig. 7C and G),triple-negative phenotype breast cancer (Fig. 7D and H), andbasal-like subtype breast cancer (Fig. 7E and I). In the multi-variate model, including age at diagnosis, grade, nodal status,tumor size, LVI, ER, and HER2 status, pHMMR was an inde-pendent predictor of RFS in patients with basal-like subtype(P ¼ 0.02), but was not significant in multivariate analysesof BCSS (P ¼ 0.07). Similarly, in multivariate analyses,pHMMR was not a significant independent predictor of BCSSor RFS in all ER-negative breast cancer (P ¼ 0.09 and P ¼ 0.15,respectively).

Within the entire cohort, patients were treated with either acombination of chemotherapy and tamoxifen (n¼ 237), chemo-therapy alone (n ¼ 588), tamoxifen alone (n ¼ 968), or no AST(n¼ 1,362).Wewere specifically interested in the predictive valueof pHMMR in the cohort of patients that received no AST.Phosphorylated HMMR abundance did not predict BCSS or RFSacross all tumors that received no AST (Fig. 7J and L). However,when the analysis was restricted to the patients with ER-negativetumors that received no AST (n ¼ 453), pHMMR was an inde-pendent predictor of both BCSS and RFS (Fig. 7K and M), andthis association was retained in multivariate analyses (P ¼ 0.01and P¼ 0.002, respectively); in patients with ER-negative tumorsthat received no AST, there was also an association betweenpHMMR and distant nodal metastasis (P < 0.05).

DiscussionApicobasal-polarized epithelial cells organize microtubules at

cell–cell contacts (3) but, prior to cell division, Aurora-A enablescilia dissolution (2) and promotes the microtubule nucleationcapacity, or maturation, of the centrosome (33, 34). Similarly,microtubule nucleation at the centrosome, whichmust be reposi-tioned toward the leading edge to establish front-rear polarity(29), is critical to motile or scattering epithelial cells (4). Thus,

Chu et al.





TPX2

Merge/DAPI

TPX2

Merge/DAPI

TPX2

Merge/DAPI

TPX2

Merge/DAPI

TPX2

GFP GFPGFP-RHAMMWT

Merge/ DAPI

GFP-RHAMMWTGFP-RHAMM703A

Scrambled Scrambled Scrambled HMMR HMMR HMMR

TPX2

Merge/DAPI

GFP-RHAMM703A

0.0

0.5

1.0

1.5P < 0.0001

Col

ocal

izat

ion

coef

ficie

nt

HMMR

Input

Actin

IP: Importin-a IgG

Importin-a

TPX2

RANBP2 RANBP2

TPX2-DDK TPX2-DDK

Transport(+ cytosol, + energy)

TPX2-DDKTPX2-DDK

No transport

TPX2-DDKTPX2-DDK

TPX2RANBP2DAPI

Scrambled HMMR Scrambled HMMR

Scrambled HMMR

Scr HMMR

TPX2RANBP2DAPI

TPX2RANBP2DAPI

TPX2RANBP2DAPI

(- cytosol, - energy)

A B

C D

E F G

0

2

4

6

TPX2-DDK

Nuc

lear

/cyt

opla

smic

inte

nsity

Scr HMMR Scr HMMR

NLS-BSA

0.00

0.25

0.50

0.75

1.00

Nuc

lear

/cyt

opla

smic

inte

nsity

(nor

mal

ized

)

+GFP

+GFP

-WT

+GFP

-703

A

+GFP

+GFP

-WT

+GFP

-703

A

Scrambled HMMR

TPX2

P < 0.01

P < 0.01P = 0.49

P < 0.05

Scr HMMR

Scr HMMR Scr HMMR

TPX2-DDKRANBP2DAPI

TPX2-DDKRANBP2DAPI

Figure 6.

HMMR-T703 is required for the nuclear localization of TPX2. A, Representative images for TPX2 localization following rescue of HMMR knockdown in HeLa cells usingGFP alone, wild-type HMMR (GFP-HMMRWT), or phospho-dead HMMR (GFP-HMMR703A). Transfected cells were indicated by GFP immunofluorescence. Scale bar,10 mm. B,Quantification of nuclear to cytoplasmic fluorescence intensity ratio for TPX2 following HMMR knockdown and rescue with either GFP, wild-type GFP-HMMR(GFP-FL), or phospho-dead GFP-HMMR703A (GFP-703A) from two experiments (730 cells/group). TPX2 nuclear/cytoplasmic ratio in cells expressing GFPconstructs are normalized to the corresponding control-treated cells. Data represented as mean � SDs, n ¼ 2 experiments. P values from one-way ANOVA.C, Representative images for nuclear import assays measuring the localization of TPX2-DDK in digitonin-permeabilized HeLa cells treated with scrambled siRNAor siRNA targeting HMMR. During the assay, cells were provided with (Transport), or without (No Transport), an energy regeneration mixture and rabbitreticulocyte lysate (cytosol). Scale bars, 20 mm. D, Quantification of nuclear to cytoplasmic fluorescence intensity ratio for nuclear import assays with NLS-BSA orTPX2-DDK in digitonin-permeabilized HeLa cells treated with scrambled (Scr) siRNA or siRNA targeting HMMR and provided an energy regeneration mixtureand rabbit reticulocyte lysate. Data represented as mean � SDs, n ¼ 2 experiments. P values from Student t test. E, Representative images for TPX2-DDKcolocalization with RANBP2 in digitonin-permeabilized HeLa cells treated with scrambled siRNA or siRNA targeting HMMR and provided an energy regenerationmixture and rabbit reticulocyte lysate. Inverted images are presented for RANBP2 and TPX2 immunofluorescence to more clearly demarcate the nuclearenvelope localization. Scale bar, 20 mm. F, Quantification of TPX2-DDK-RANBP2 co-localization in digitonin-permeabilized HeLa cells treated with scrambled (Scr)siRNA or siRNA targeting HMMR and provided an energy regeneration mixture and rabbit reticulocyte lysate. Data are presented as box and whisker graphdisplaying the median, 25–75 percentiles, and 1 and 99 percentiles for 23 cells (Scr) or 16 cells (HMMR) across two experiments. P value from Student t test.G, Coimmunoprecipitation of TPX2 with antibodies targeting importin-a, but not with control IgG, is augmented from cell lysates prepared from HeLa cells previouslytransfected with siRNA targeting HMMR relative to scrambled (Scr) siRNA. The levels of HMMR and TPX2 in the inputted lysates are shown with actin serving as aloading control. Images are representative of three experiments.






P = 0.89

Overall relapse-free survival time (years)0 5 10 15

Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0

Any ER+ tumors

n = 2,218


Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0

Any ER– tumors

P = 0.05n = 941


Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0

TNP tumors

n = 538 P = 0.03


Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0 P < 0.01n = 293

Overall relapse-free survival time (years)

Cum

ulat

ive

surv

ival

All tumors (no AST)

Overall relapse-free survival time (years)

Cum

ulat

ive

surv

ival

P = 0.550 5 10 15

1.0

0.8

0.6

0.4

0.2

0.0

0 5 10 15

1.0

0.8

0.6

0.4

0.2

0.0n = 1,362 P = 0.01

ER-tumors (no AST)

n = 453

P = 0.58 P = 0.02

pHMMR High

pHMMR Low

Breast cancer–specific survival time (years)0 5 10 15

Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0

Any ER+ tumors

n = 2,218


Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0

Any ER– tumors

n = 941


Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0

TNP tumors

P = 0.04n = 538


Cum

ulat

ive

surv

ival

1.0

0.8

0.6

0.4

0.2

0.0 P = 0.03n = 293

Basal-like tumors

Breast cancer–specific survival time (years)

Cum

ulat

ive

surv

ival

All tumors (no AST)

Breast cancer–specific survival time (years)

Cum

ulat

ive

surv

ival

ER-tumors (no AST)

P = 0.310 5 10 15

1.0

0.8

0.6

0.4

0.2

0.0

0 5 10 15

1.0

0.8

0.6

0.4

0.2

0.0n = 1,362 P = 0.003n = 453

A

B C D E

F G H I

J K L M

Intensity = 0 Intensity = 1 Intensity = 2

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

pHMMR High

pHMMR Low

Basal-like tumors

Figure 7.

Phosphorylated HMMR is a prognostic marker in ER-negative breast cancer. A, pHMMR expression in mammary ductal carcinoma. Low/no staining(intensity ¼ 0), moderate staining (intensity ¼ 1), or strong staining (intensity ¼ 2). B, pHMMR correlation with breast cancer–specific survival inER-positive tumors (n ¼ 2,218). Survival differences were estimated using the log-rank test. C, pHMMR correlation with breast cancer–specific survival inER-negative tumors (n ¼ 941). Survival differences were estimated using the log-rank test. D, pHMMR correlation with breast cancer–specificsurvival in triple-negative phenotype (TNP) tumors (n ¼ 538). Survival differences were estimated using the log-rank test. E, pHMMR correlation withbreast cancer–specific survival in basal-like subtype tumors (n ¼ 293). Survival differences were estimated using the log-rank test. F, pHMMRcorrelation with overall relapse-free survival in ER-positive tumors (n ¼ 2,218). Survival differences were estimated using the log-rank test.G, pHMMR correlation with overall relapse-free survival in ER-negative tumors (n ¼ 941). Survival differences were estimated using the log-rank test.H, pHMMR correlation with overall relapse-free survival in triple-negative phenotype (TNP) tumors (n ¼ 538). Survival differences were estimatedusing the log-rank test. I, pHMMR correlation with overall relapse-free survival in basal-like subtype tumors (n ¼ 293). Survival differences wereestimated using the log-rank test. J, pHMMR correlation with breast cancer–specific survival in all tumors from patients not receiving adjuvantsystemic therapy (n ¼ 1,362). Survival differences were estimated using the log-rank test. K, pHMMR correlation with breast cancer–specific survival inER-negative tumors from patients not receiving adjuvant systemic therapy (n ¼ 453). Survival differences were estimated using the log-rank test. L,pHMMR correlation with overall relapse-free survival in all tumors from patients not receiving adjuvant systemic therapy (n ¼ 1,362). Survivaldifferences were estimated using the log-rank test. M, pHMMR correlation with overall relapse-free survival in ER-negative tumors from patients notreceiving adjuvant systemic therapy (n ¼ 453). Survival differences were estimated using the log-rank test.

Chu et al.





cell-cycle progression and cell migration necessitate comparablechanges with the organization of microtubules in epithelial orcarcinoma cells. Here, we find that cell-cycle progression aug-ments both engraftment kinetics in vivo and the expression ofmesenchymal markers in vitro in asynchronously growing 4T1-luc2 cells. Cells in S–G2-phase possess an increased velocity anddirectionality of migration as well as front-polarized centrosomeswith augmented microtubule nucleation capacity. Thus, we findthat epithelial cell migration and cell-cycle progression sharecommon regulation of the microtubule cytoskeleton through theAurora-A–TPX2–HMMR axis.

The connection between cell-cycle progression and migrationin epithelial cells likely explains the strong relationshipbetween cell populations that express an EMT signature andthose with an augmented tumor-initiating potential (7, 8).When EMT is induced in renal fibrosis, for instance, the cellsbecome delayed in G2–M, and this delay is only reversed withthe removal of the EMT induction (35). This relationship hasimportant implications. Conventional theories hypothesizethat rare malignant cells in a primary tumor possess uniqueproperties or transcriptional programs that enable them to seedmetastasis (9); analysis of transcriptional profiles in single cellsisolated from low burden metastatic sites suggest that thesepioneer cells may express basal-like or stem-like gene expres-sion patterns, including dormancy-related genes, and that met-astatic progression requires a switch to the cell cycle (9). Here,we find that the simple process of enriching for G1-phase cellshas dramatic negative consequences on the expression ofdefined basal-like cell markers and the efficiency of tumorengraftment. Moreover, we identified significant changes inmigration velocity, centrosome polarization, and microtubulenucleation capacity as individual G1-phase cells progressedthrough the cell cycle. Thus, a carcinoma cell transitioningthrough the cell cycle demonstrates plasticity that may beinterpreted by gene expression profiling, or through xenotrans-plantation assays, as an EMT or increasing cancer "stemness."But, a connection between cell-cycle progression and migratorycapacity is likely cell-type specific given that hematopoieticcells, for example, do not engraft well during S–G2–M butoptimally engraft during G0–G1-phase (36, 37). Our findingsare also highly relevant to the study of tumor initiation andcancer stem cells (CSC). The tumor xenotransplantation assayis the current gold standard for the functional detection ofcancer stem cells. Here, we find that simple adjustment in cell-cycle progression within a standard breast cancer cell linesignificantly impacts the efficiency of engraftment analogousto the influence of the immune system in the recipient animal(38) or the use of tumor fragments as opposed to disassociatedcells (39). Thus, for carcinoma samples, the CSC readoutobtained from a tumor xenotransplantation assay may bestrongly impacted by the cell-cycle profile of the inputtedtumor sample.

A FRET-based reporter for Aurora-A activity at centrosomesindicates the need for nuclear-localized TPX2 to activate thekinase and stabilize microtubules during G1 phase (20), but themechanism for this regulation is not well studied. We havepreviously reported that silencing HMMR reduces the nuclearlocalization of TPX2 and alters Aurora-A activity (19). Here, weshow that HMMR, and specifically T703, is needed for the nucleartransport of TPX2. In HMMR-silenced, nonmitotic cells, TPX2–importin-a complexes are augmented and TPX2 accumulates at

the nuclear envelope while the activity of Aurora-A at the centro-some, as measured by the levels of pAurora-A, is dampened,similar to the effect observed in TPX2-silenced, G1-phase cells(20). This observation contrasts with our prior interpretation thatpHMMR negatively regulates Aurora-A activity (19), which wasbased upon measurements of kinase activity in cell lysates asopposed to individual cells. Rather, our analysis of pAurora-Aindicates that HMMRpromotes kinase activity in nonmitotic cellsthrough the nuclear transport of TPX2 and facilitating TPX2release from a complex with importin-a, similar to the roleHMMR plays in mitotic cells and extracts (16, 17). Aurora-Aactivity at the centrosome is known to control microtubuledynamics in migrating endothelial cells (40) and neurons (41).Thus, in these cells and tissues, it will be important to determinewhether Aurora-A activity is also reliant upon the control of TPX2location by pHMMR.

A limitation of our study of pHMMR in the large breastcancer tissue microarrays relates to IHC approaches, which havebeen outlined previously (27). Briefly, the study has limitedtechnical reproducibility and the readout for pHMMR wassubjective and qualitative, although we relied upon a consensusscore from three blinded observers to offset the latter limita-tion. Moreover, the study was trained and validated on tissuemicroarrays, which may miss focally higher areas of immunos-taining that are more easily appreciated in whole sections.Survival data were derived from a median follow-up periodof 12.3 years and, as a consequence, the treatment recommen-dations at the time of tissue collection (1986–1992) differedfrom those in contemporary practice, which tend to be moreaggressive. Putting these limitations aside, however, the abun-dance of pHMMR appears to be a significant negative prog-nostic factor specifically in ER-negative breast tumors.

Proliferation is one of the most important prognostic factorsfor invasive breast cancer (42). In ER-positive, HER2-negativesubtypes, a proliferative molecular module, defined in part byAurora-A expression, was the common denominator of mostprognostic signatures and the strongest parameter predictingclinical outcome (43). Indeed, Aurora-A expression, used incombination with the expression of HER2 and ER, is sufficientto discriminate between low and high proliferative luminal Aand luminal B tumors (44). Moreover, stable DNA amplifica-tion of TPX2 was identified in ER-positive breast cancer and wasfound to be enriched in prognostic gene sets for ER-positivebreast tumors (45). However, we found that elevated pHMMRlevels did not discriminate ER-positive tumors with a poorprognosis (Fig. 7B and F). When considering this discrepancy, itis important to note that elevated transcript expression does notabsolutely correlate with protein expression nor with kinaseactivity (phosphoproteome) at steady state or following drugtreatment (46), so ER-positive tumors with elevated Aurora-Aor TPX2 gene expression may not be detected with highpHMMR by IHC. More importantly, in the TMAs analyzedhere, the Ki67 index associated with poor BCSS and RFS inluminal breast tumors (27). Thus, it is likely that luminal breasttumors detected to express high pHMMR are not simply Ki67-positive tumors with elevated levels of proliferation or withconsistent elevated expression of TPX2 and Aurora-A.

The selectivity of the prognostic power for elevated pHMMRin TNBC and basal-like breast cancer may relate to a specificnonproliferative role of Aurora-A signaling in promoting theprogression of these subtypes. Genomic and transcriptomic






analyses of 2,000 breast tumors implicated elevated Aurorakinase signaling in a basal-like cancer–enriched subgroup(47). In this large cohort of breast tumors, basal-like breasttumors harbor chromosome 5q deletions associated in transwith elevated Aurora kinase signaling and cell division genes(47). In other tumor cells, genomic imbalance in chromosome5q32-qter correlated with markers for Aurora kinase activityand elevated sensitivity to aurora kinase inhibitors (48).Chromosome 5q33-34 and specifically polymorphisms ofHMMR and CCNG1 in this region (19, 49), modifies breastcancer risk in BRCA1 mutation carriers, which tend to developER-negative or basal-like breast tumors. Thus, heightenedAurora-A activity, as indicated by elevated pHMMR levels,may promote the progression of ER-negative breast tumorsand consequently indicate a poor prognosis. If so, thesetumors may possess heightened sensitivity to small-moleculeAurora kinase inhibitors.

Taken together, our data indicate that cell-cycle progression andthe acquisition of migratory phenotypes are interconnected incarcinoma cells by a requirement for the acquisition of microtu-bule nucleation capacity at the centrosome that is enabled by theAurora-A/TPX2/HMMR axis. Specifically, this molecular axis pro-motes the establishment of directional migration. The precisereason why this pathway is more strongly responsible for theprogression of ER-negative breast tumors, and whether this path-way enables rare invasive "leader" cells or epithelial cell scattering,which is known to rely upon centrosome repositioning (4),remains unclear and warrants further study. Finally, our findingsmay also have implications for therapeutic targeting strategies incurrently difficult-to-treat ER-negative breast cancer.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: T.L.H. ChuDevelopmentofmethodology: T.L.H.Chu,M.Connell, L. Zhou, S.M.R. Rahavi,A. Fotovati, N. Pante, C.A. MaxwellAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): T.L.H. Chu, M. Connell, L. Zhou, Z. He, H. Chen,S.M.R. Rahavi, P. Mohan, O. Nemirovsky, G.S.D. Reid, T.O. Nielsen,C.A. MaxwellAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): T.L.H. Chu, M. Connell, L. Zhou, Z. He, J. Won,H. Chen, S.M.R. Rahavi, P. Mohan, O. Nemirovsky, M.A. Pujana, G.S.D. Reid,T.O. Nielsen, N. Pante, C.A. MaxwellWriting, review, and/or revision of the manuscript: T.L.H. Chu, M. Connell,S.M.R. Rahavi, P. Mohan, G.S.D. Reid, N. Pante, C.A. MaxwellAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): T.L.H. Chu, L. Zhou, Z. He, S.M.R. Rahavi,C.A. MaxwellStudy supervision: T.L.H. Chu, N. Pante, C.A. MaxwellOther (animal study and bioluminescent imaging): S.M. Rahavi

AcknowledgmentsThis study was funded by the Canadian Institutes of Health Research

(OBC-134038, awarded to C.A. Maxwell and N. Pante; MSH-136647 salaryaward to C.A. Maxwell) and the Canadian Breast Cancer Foundation (PhDfellowship, awarded to T.L.H.Chu). C.A.Maxwell andG.S.D.Reid are supportedby salary awards from the Michael Cuccione Childhood Cancer ResearchProgram, BCCH.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received August 1, 2017; revised September 27, 2017; accepted October 4,2017; published OnlineFirst October 9, 2017.

References1. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell

2011;144:646–74.2. Pugacheva EN, Jablonski SA, Hartman TR, Henske EP, Golemis EA. HEF1-

dependent Aurora A activation induces disassembly of the primary cilium.Cell 2007;129:1351–63.

3. Meng W, Mushika Y, Ichii T, Takeichi M. Anchorage of microtubule minusends to adherens junctions regulates epithelial cell-cell contacts. Cell 2008;135:948–59.

4. Burute M, Prioux M, Blin G, Truchet S, Letort G, Tseng Q, et al. Polarityreversal by centrosome repositioning primes cell scattering during epithe-lial-to-mesenchymal transition. Dev Cell 2017;40:168–84.

5. Fischer KR, Durrans A, Lee S, Sheng J, Li F, Wong ST, et al. Epithelial-to-mesenchymal transition is not required for lungmetastasis but contributesto chemoresistance. Nature 2015;527:472–6.

6. Zheng X, Carstens JL, Kim J, Scheible M, Kaye J, Sugimoto H, et al.Epithelial-to-mesenchymal transition is dispensable for metastasis butinduces chemoresistance in pancreatic cancer. Nature 2015;527:525–30.

7. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, et al. Theepithelial-mesenchymal transition generates cells with properties of stemcells. Cell 2008;133:704–15.

8. Morel AP, LievreM, ThomasC,HinkalG, Ansieau S, Puisieux A.Generationof breast cancer stem cells through epithelial-mesenchymal transition.PLoS One 2008;3:e2888.

9. Lawson DA, Bhakta NR, Kessenbrock K, Prummel KD, Yu Y, Takai K, et al.Single-cell analysis reveals a stem-cell program in humanmetastatic breastcancer cells. Nature 2015;526:131–5.

10. Cheung KJ, Gabrielson E, Werb Z, Ewald AJ. Collective invasion inbreast cancer requires a conserved basal epithelial program. Cell 2013;155:1639–51.

11. Godinho SA, Picone R, Burute M, Dagher R, Su Y, Leung CT, et al.Oncogene-like induction of cellular invasion from centrosome amplifica-tion. Nature 2014;510:167–71.

12. Wang G, Jiang Q, Zhang C. The role of mitotic kinases in coupling thecentrosome cycle with the assembly of the mitotic spindle. J Cell Sci 2014;127:4111–22.

13. Bayliss R, Sardon T, Vernos I, Conti E. Structural basis of Aurora-Aactivation by TPX2 at the mitotic spindle. Mol Cell 2003;12:851–62.

14. GrussOJ, Carazo-Salas RE, SchatzCA,GuarguagliniG, Kast J,WilmM, et al.Ran induces spindle assembly by reversing the inhibitory effect of importinalpha on TPX2 activity. Cell 2001;104:83–93.

15. Kufer TA, SilljeHH,Korner R,GrussOJ,Meraldi P,Nigg EA.Human TPX2 isrequired for targeting Aurora-A kinase to the spindle. J Cell Biol 2002;158:617–23.

16. Chen H, Mohan P, Jiang J, Nemirovsky O, He D, Fleisch MC, et al.Spatial regulation of Aurora A activity during mitotic spindle assem-bly requires RHAMM to correctly localize TPX2. Cell Cycle 2014;13:2248–61.

17. Scrofani J, SardonT,Meunier S, Vernos I.Microtubule nucleation inmitosisby a RanGTP-dependent protein complex. Curr Biol 2015;25:131–40.

18. Manning AL, Compton DA. SnapShot: Nonmotor proteins in spindleassembly. Cell 2008;134:694.

19. Maxwell CA, Benitez J, Gomez-Baldo L, Osorio A, Bonifaci N, Fernandez-Ramires R, et al. Interplay between BRCA1 and RHAMM regulates epithe-lial apicobasal polarization and may influence risk of breast cancer.PLoS Biol 2011;9:e1001199.

20. Bertolin G, Sizaire F, Herbomel G, Reboutier D, Prigent C, Tramier M. AFRET biosensor reveals spatiotemporal activation and functions of aurorakinase A in living cells. Nat Commun 2016;7:12674.

Chu et al.





21. Wu WW, Sun YH, Pante N. Nuclear import of influenza A viral ribonu-cleoprotein complexes is mediated by two nuclear localization sequenceson viral nucleoprotein. Virol J 2007;4:49.

22. Liu S, Foulkes WD, Leung S, Gao D, Lau S, Kos Z, et al. Prognosticsignificance of FOXP3þ tumor-infiltrating lymphocytes in breast cancerdepends on estrogen receptor and human epidermal growth factor recep-tor-2 expression status and concurrent cytotoxic T-cell infiltration. BreastCancer Res 2014;16:432.

23. Liu S, Chia SK, Mehl E, Leung S, Rajput A, Cheang MC, et al. Progesteronereceptor is a significant factor associated with clinical outcomes and effectof adjuvant tamoxifen therapy in breast cancer patients. Breast Cancer ResTreat 2010;119:53–61.

24. Lohrisch C, Jackson J, Jones A, Mates D, Olivotto IA. Relationship betweentumor location and relapse in 6,781 women with early invasive breastcancer. J Clin Oncol 2000;18:2828–35.

25. Fotovati A, Abu-Ali S, Wang PS, Deleyrolle LP, Lee C, Triscott J, et al. YB-1bridges neural stem cells and brain tumor-initiating cells via its roles indifferentiation and cell growth. Cancer Res 2011;71:5569–78.

26. Mehta RJ, Jain RK, Leung S, Choo J, Nielsen T, HuntsmanD, et al. FOXA1 isan independent prognostic marker for ER-positive breast cancer. BreastCancer Res Treat 2012;131:881–90.

27. Cheang MC, Chia SK, Voduc D, Gao D, Leung S, Snider J, et al. Ki67 index,HER2 status, and prognosis of patients with luminal B breast cancer. J NatlCancer Inst 2009;101:736–50.

28. Maxwell CA, Keats JJ, Belch AR, Pilarski LM, Reiman T. Receptor forhyaluronan-mediated motility correlates with centrosome abnormalitiesin multiple myeloma and maintains mitotic integrity. Cancer Res 2005;65:850–60.

29. YvonAM,Walker JW,Danowski B, FagerstromC,KhodjakovA,WadsworthP. Centrosome reorientation in wound-edge cells is cell type specific. MolBiol Cell 2002;13:1871–80.

30. Carmena M, Ruchaud S, Earnshaw WC. Making the Auroras glow: regu-lation of Aurora A and B kinase function by interacting proteins. Curr OpinCell Biol 2009;21:796–805.

31. Giubettini M, Asteriti IA, Scrofani J, De Luca M, Lindon C, Lavia P, et al.Control of Aurora-A stability through interaction with TPX2. J Cell Sci2011;124:113–22.

32. Cassany A, Gerace L. Reconstitution of nuclear import in permeabilizedcells. Methods Mol Biol 2009;464:181–205.

33. BarrettaML, SpanoD,D'Ambrosio C, Cervigni RI, Scaloni A, CordaD, et al.Aurora-A recruitment and centrosomal maturation are regulated by aGolgi-activated pool of Src during G2. Nat Commun 2016;7:11727.

34. Joukov V, Walter JC, De Nicolo A. The Cep192-organized Aurora A-Plk1cascade is essential for centrosome cycle and bipolar spindle assembly.Mol Cell 2014;55:578–91.

35. Lovisa S, LeBleu VS, Tampe B, Sugimoto H, Vadnagara K, CarstensJL, et al. Epithelial-to-mesenchymal transition induces cell cycle

arrest and parenchymal damage in renal fibrosis. Nat Med 2015;21:998–1009.

36. GlimmH, Oh IH, Eaves CJ. Human hematopoietic stem cells stimulatedto proliferate in vitro lose engraftment potential during their S/G(2)/M transit and do not reenter G(0). Blood 2000;96:4185–93.

37. Bowie MB, McKnight KD, Kent DG, McCaffrey L, Hoodless PA, Eaves CJ.Hematopoietic stem cells proliferate until after birth and show areversible phase-specific engraftment defect. J Clin Invest 2006;116:2808–16.

38. Quintana E, ShackletonM, SabelMS, FullenDR, Johnson TM,Morrison SJ.Efficient tumour formation by single human melanoma cells. Nature2008;456:593–8.

39. Landis MD, Lehmann BD, Pietenpol JA, Chang JC. Patient-derived breasttumor xenografts facilitating personalized cancer therapy. Breast CancerRes 2013;15:201.

40. Braun A, Dang K, Buslig F, Baird MA, Davidson MW, Waterman CM, et al.Rac1 and Aurora A regulate MCAK to polarize microtubule growth inmigrating endothelial cells. J Cell Biol 2014;206:97–112.

41. Takitoh T, Kumamoto K,WangCC, SatoM, Toba S,Wynshaw-Boris A, et al.Activation of Aurora-A is essential for neuronal migration via modulationof microtubule organization. J Neurosci 2012;32:11050–66.

42. van Diest PJ, van der Wall E, Baak JP. Prognostic value of proliferation ininvasive breast cancer: a review. J Clin Pathol 2004;57:675–81.

43. Desmedt C, Haibe-Kains B, Wirapati P, Buyse M, Larsimont D,Bontempi G, et al. Biological processes associated with breast cancerclinical outcome depend on the molecular subtypes. Clin Cancer Res2008;14:5158–65.

44. Haibe-Kains B, Desmedt C, Loi S, Culhane AC, Bontempi G, QuackenbushJ, et al. A three-gene model to robustly identify breast cancer molecularsubtypes. J Natl Cancer Inst 2012;104:311–25.

45. Endesfelder D, Burrell R, Kanu N, McGranahan N, Howell M, Parker PJ,et al. Chromosomal instability selects gene copy-number variants encod-ing core regulators of proliferation in ERþ breast cancer. Cancer Res2014;74:4853–63.

46. Sacco F, Silvestri A, PoscaD, Pirro S,Gherardini PF, Castagnoli L, et al. Deepproteomics of breast cancer cells reveals that metformin rewires signalingnetworks away from a pro-growth state. Cell Syst 2016;2:159–71.

47. Curtis C, Shah SP, Chin SF, Turashvili G, RuedaOM,DunningMJ, et al. Thegenomic and transcriptomic architecture of 2,000 breast tumours revealsnovel subgroups. Nature 2012;486:346–52.

48. Mohan P, Castellsague J, Jiang J, Allen K, Chen H, Nemirovsky O, et al.Genomic imbalance of HMMR/RHAMM regulates the sensitivity andresponse of malignant peripheral nerve sheath tumour cells to aurorakinase inhibition. Oncotarget 2013;4:80–93.

49. Pujana MA, Han JD, Starita LM, Stevens KN, Tewari M, Ahn JS, et al.Network modeling links breast cancer susceptibility and centrosomedysfunction. Nat Genet 2007;39:1338–49.






2018;16:16-31. Published OnlineFirst October 9, 2017.Mol Cancer Res Tony L.H. Chu, Marisa Connell, Lixin Zhou, et al. Enabled by Aurora-A

Dependent Tumor Engraftment and Migration Are−Cell Cycle

Updated version

10.1158/1541-7786.MCR-17-0417doi:

Access the most recent version of this article at:

Cited articles

http://mcr.aacrjournals.org/content/16/1/16.full#ref-list-1

This article cites 49 articles, 13 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/16/1/16.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mcr.aacrjournals.org/content/16/1/16To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-17-0417

http://mcr.aacrjournals.org/content/16/1/16.full#ref-list-1

http://mcr.aacrjournals.org/content/16/1/16.full#related-urls

http://mcr.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://mcr.aacrjournals.org/content/16/1/16


Cell Cycle Dependent Tumor Engraftment and Migration Are … · ra-A activity at centrosomes, as...

Documents

Transcript of Cell Cycle Dependent Tumor Engraftment and Migration Are … · ra-A activity at centrosomes, as...