kif3a promotes proliferation and invasion via wnt...

Cell Cycle and Senescence

KIF3a Promotes Proliferation and Invasion via Wnt Signalingin Advanced Prostate Cancer

Zun Liu, Ryan E. Rebowe, Zemin Wang, Yingchun Li, Zehua Wang, John S. DePaolo,Jianhui Guo, Chiping Qian, and Wanguo Liu

AbstractAberrant activation of the Wnt/b-catenin signaling pathway is a critical event in advanced prostate cancer,

but the genetic alterations that activate the Wnt signaling pathway in many other cancers are rarely observed inprostate cancer. Other molecular mechanisms that regulate the Wnt signaling pathway in prostate cancerremain to be identified. Here, it is demonstrated that KIF3a, a subunit of kinesin-II motor protein, functions asan agonist of the Wnt signaling pathway in prostate cancer. KIF3a is upregulated in the majority of humanprostate cancer cell lines and primary tumor biopsies. The expression levels of KIF3a correlate with a higherGleason score, tumor–node–metastasis stage, and metastatic status of prostate cancer. Moreover, exogenousexpression of KIF3a promoted cell growth in the benign prostate cells, whereas silencing KIF3a in cancer cellsdecreased cell proliferation, anchorage-independent cell growth, and cell migration/invasion. Mechanistically,KIF3a increases CK1-dependent DVL2 phosphorylation and b-catenin activation in prostate cancer cells,leading to transactivation of the Wnt-signaling target genes such as cyclin D1, HEF1, and MMP9. Thesefindings support the notion that upregulation of KIF3a is causal of aberrant activation of Wnt signaling inadvanced prostate cancer through the KIF3a–DVL2–b-catenin axis.

Implications: Inactivation of KIF3a may improve survival of patients with advanced prostate cancer in whichWntsignaling is activated. Mol Cancer Res; 12(4); 491–503. �2014 AACR.

IntroductionProstate cancer remains themost diagnosed cancer and the

second leading cause of cancer-related death in men in theUnited States (1). Although the incidence and mortality ofthis cancer are decreasing, advanced or aggressive prostatecancer remains the major cause of prostate cancer–relateddeath (1). Aggressive prostate cancer is characterized by earlyage onset, high Gleason score (7–10), high prostate-specificantigen level (more than 50 ng/mL), and/or androgenindependence (2, 3). The mortality of patients with aggres-sive prostate cancer is about four times higher than nonag-gressive prostate cancer (4). Because aggressive and nonag-gressive prostate cancer have dramatically different clinicaloutcomes, markers associated with aggressiveness of prostatecancer would be of great clinical utility. Identifying thesemarkers and understanding their mechanisms will identify

novel therapeutic or chemopreventive targets for patientswith aggressive prostate cancer.Wnt signaling plays a crucial role in the progression of

prostate cancer, and particularly, in aggressive prostatecancer (5–11). Wnt activation in other tumor types, suchas colon cancer, is induced by gene mutations in APC,b-catenin, or AXIN1/2, components of the b-catenindestruction complex of the Wnt signaling pathway (12–14). In cells harboring these mutations, b-catenin accumu-lates, translocates to the nucleus, and interacts with the T-cell factor/lymphoid enhancer factor (TCF/LEF) family oftranscription factors to activate specific Wnt target genes,such as c-Myc, cyclin D1, HEF1, and MMPs (matrix metal-loproteinases) leading to tumorigenesis (15–18). In addi-tion, increased phosphorylation of DVL2 also activatesWnt/b-catenin signaling in certain cancers (19). However,in contrast with colon cancer, mutations in b-catenin andAPC are rare in prostate cancer (20). Only 5% of prostatetumors harbor activating mutations in b-catenin and evenless contain APC mutations; however, the frequency ofnuclear accumulation of b-catenin was reported in 23% to83% of prostate cancer (21). Thus, the mechanisms acti-vating the Wnt/b-catenin signaling pathway in a largeproportion of prostate cancer have yet to be identified.KIF3a is amember of the kinesin family ofmotor proteins.

It has been implicated downstream of the Hedgehog-signal-ing complex and has been shown to regulate early develop-ment, ciliogenesis, and tumorigenesis (22). Interestingly,

Authors' Affiliation: Department of Genetics, Stanley S. Scott CancerCenter, Louisiana State University School of Medicine, New Orleans,Louisiana

Z. Liu and R.E. Rebowe contributed equally to this work.

Corresponding Author: Wanguo Liu, Department of Genetics andStanley S. Scott Cancer Center, Louisiana State University HealthSciences Center, LCRC building Room 904, 1700 Tulane Avenue, NewOrleans, LA 70112. Phone: 504-210-3326; Fax: 504-210-2970; E-mail:[email protected]

doi: 10.1158/1541-7786.MCR-13-0418

�2014 American Association for Cancer Research.

MolecularCancer

Research

www.aacrjournals.org 491

on April 29, 2018. © 2014 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst January 10, 2014; DOI: 10.1158/1541-7786.MCR-13-0418

http://mcr.aacrjournals.org/

P69

p69

BPH1

RWPE-2

PC-3

LNCaP

DU-145

M12

M2182

C4-2B

RW

PE-2

BPH-1

LNCaP

PC-3

PCa

Normal

Normal Tumor

39/74

a b

c d

% o

f P

ositiv

e K

IF3a I

HC

sta

inin

g

0/9

70

60

50

40

30

20

10

0

M12

M21

82

C4-

2B

KIF3a

Sample (pair)

Gleason score

1 2 3 4 5

5

6

6 6 6 6

7

7 78

8 8 7 7 7 7 78 8

9 10 11 12 1314 15 16 17 1899

1

7 8 9 10 11 12

13 14 15 16 17 18

N T N T N T N T N T N T



2 3 4 5 6

KIF3a

**

*

2.2

2.0

1.8

1.6

1.4

1.2

1.0

0.8

2.0

1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

β-Actin

β-Actin

KIF3a

β-Actin

KIF3a

β-Actin

Re

lative

KIF

3a

leve

lsW

este

rn b

ands (

densitom

etr

y)

DU-1

45

A B

DC

FE

Figure 1. KIF3a is upregulated in prostate cancer (PCa). A andB, KIF3a levels in PCa cell lineswere determined byWestern blotting and by densitometry analysisof the Western bands. The intensity of the KIF3a bands was normalized to the intensity of their respective b-actin bands and quantified against each other.�,P<0.05and ��,P<0.01ascomparedwithbenignprostateepithelial cell lines.CandD,KIF3a levelsweredetermined in18pairedprostate tumorsandmatchednormal prostate tissues by Western blotting and by quantification of the protein bands as described in A and B. (Continued on the following page.)

Liu et al.

Mol Cancer Res; 12(4) April 2014 Molecular Cancer Research492




KIF3a interacts with Wnt-signaling component, APC,through an association with the kinesin superfamily-associ-ated protein (KAP3) for regulating cell migration (23). As amicrotubule-directed motor subunit of the KIF3 complex,KIF3a also plays an important role in the subcellular trans-port of b-catenin–cadherin(s) complex (24). In addition, itwas demonstrated that KIF3a can constrain b-catenin–dependentWnt signaling through dual ciliary and nonciliarymechanisms (25). Moreover, selective deletion of Kif3a inosteoblasts of the Kif3a9Oc-cKO mice impairs osteoblast-mediated bone formation through multiple pathways,including Wnt signaling (26). Although these studies pro-vided evidence that KIF3a regulates the Wnt signalingpathway, whether KIF3a plays a role in the activation of theWnt signaling pathway in prostate cancer remains unknown.In this study, wemeasured the expression levels of KIF3a in

prostate cancer cell lines and primary tumor tissues andshowed the correlation of KIF3a levels with prostate cancerprogression and metastasis. We also examined the role ofKIF3a in phosphorylation of DVL2 and in activation of theWnt signaling pathway and identified the KIF3a downstreamtargets cyclin D1, MMP9, and HEF1. Our data provideevidence to support the hypothesis that upregulation ofKIF3aactivates the Wnt signaling pathway to promote prostatecancer cell proliferation and cancer progression. KIF3a is apotential therapeutic target for advanced prostate cancer.

Materials and MethodsCell lines, primary tumor tissues, and tissue microarrayCell lines, including 293T, LNCaP, DU145, PC-3, and

RWPE-2, were purchased from American Type CultureCollection (ATCC); BPH1 (27), P69 (28), M12 (29),M2182 (30), and C4-2B (31) cell lines were kindly providedby Dr. Haojie Huang (Mayo Clinic, Rochester, MN) or Dr.Shahriar Koochekpour (LSUHSC, New Orleans, LA). Allthe cell lines were maintained in an appropriate mediumaccording to the protocols of the ATCC. The humanprostate cancer tumor tissues were obtained from the Louisi-ana Cancer Research Consortium (LCRC; New Orleans,LA) with patient consent and the Institutional Review Boardapproval. The prostate tissuemicroarray was purchased fromUS Biomax, Inc.

Western blotting, immunofluorescence, andimmunohistochemistry analysesWestern blotting was performed as described previously

(32) using anti-KIF3a (Sigma), anti–b-catenin (BDTransduction Laboratories), anti-MMP9 (EMD Milli-pore), anti-HEF1 (ImmuQuest), anti-DVL2 (Cell Signal-ing Technology) and anti-cyclin D1 (BD TransductionLaboratories) antibodies. Protein bands were visualizedusing the Enhanced Chemiluminescence Kit (ThermoScientific). For immunofluorescence analysis, cells weregrown on 12-well chamber slides coated with 100 nmol/L

poly-l-lysine (Invitrogen) for 24 hours. Cells were thenwashed, fixed, and blocked in 1% bovine serum albumin(BSA)–PBS, incubated with primary antibodies and goatanti-rabbit secondary antibody conjugated to Alexa 488(Invitrogen). Imaging was performed on Carl Zeiss fluo-rescence microscope or Confocal Laser Scanning Micro-scope. ImageJ 1.47 (NIH, Bethesda, MD) software wasused for fluorescence intensity measurement and densi-tometry analysis of Western blot. Immunohistochemicalanalyses of the human prostate tissue microarray wereconducted using the anti-KIF3a antibody (Sigma). Tissuesections were deparaffinized and rehydrated. Antigenretrieval was achieved by boiling the sections for 20minutes in 10 mmol/L citric acid buffer (pH6.0). Aftera 10-minute treatment with 3% hydrogen peroxide in 1�PBS to remove endogenous peroxidase and a 30 minutesincubation with 5% BSA for blocking nonspecific bind-ing, tissue sections were incubated with KIF3a antibody(1:100 in 2% BSA) overnight and followed by incubationwith biotinylated goat anti-rabbit antiserum (JacksonImmunoResearch Laboratories) diluted 1:200 in 2% BSAfor 30 minutes at room temperature, and detected withthe ABC Immunodection Kit (Vector Laboratories). Sec-tions were counterstained with Harris hematoxylin solu-tion (Sigma), passed through a dehydration process andcover slipped. The stained sections were scored as 0 (nostaining), 1 (þ, weak positive), 2 (þþ, medium positive),and 3 (þþþ, strong positive), respectively. To assess thespecificity of antibodies used for immunohistochemistry(IHC), omission of the primary antibodies or incubationof samples with normal rabbit IgG as the negative controlwas performed in tumor samples in parallel with regularanti-KIF3a antibody incubation. These negative controlsdemonstrated no immunoreactivity.

Overexpression and shRNA-mediated silencing of KIF3aThe KIF3a-expressing construct was generated by insert-

ing human KIF3a cDNA into pCMV–Tag2b vector (Stra-tagene) with BamHI/HindIII restriction sites, and verified bysequencing. Transient overexpression of KIF3a was per-formed, according to the manufacturer's instructions, bytransfection using X-tremeGene HP DNA transfectionreagent (Roche). Lentiviral short hairpin RNA (shRNA)constructs targeting KIF3A or a scramble control weregenerated by inserting the sequences 50-TCCGCCAGTTT-CAGAAAGAAA-30 (shKIF3a-1), 50-CCAAAGACATTTA-CTTTCGAT-30 (shKIF3a-2), or 50-CCTAAGGTTAAGT-CGCCCTCG-30 (shControl) into the pLKO.1 vector(Addgene) with the AgelI/EcoRI restriction sites. The len-tiviral particles expressing the control or KIF3a shRNA wereproduced in 293T cells using pPACKH1-XL Packaging Mix(System Biosciences) and FuGENE Transfection Reagent(Roche) following the manufacture's instructions. Transfec-tion of LNCaP cells was conducted following a protocol

(Continued.) E, thehistogramshows thepercentageofpositive IHCstainingofKIF3a in9normaland74PCatissuessamples.The IHCstainingofaprostate tissuemicroarray was performed with a KIF3a antibody. F, representative microphotographs of IHC analysis of KIF3a in the prostate tissue microarray.Images a to d represent IHC staining in normal prostate and grade 1–3 PCa tissues (magnification, �200).

KIF3a Activates Wnt Signaling in Prostate Cancer

www.aacrjournals.org Mol Cancer Res; 12(4) April 2014 493




previously described (33). Knockdown efficiency was deter-mined by Western blot analysis 48 hours after transduction.

Cell proliferation and anchorage-independent growthassaysLNCaP cellswere transducedwith lentiviral particles expres-

sing shKIF3a or control shRNAfor 48hours and thenplated ata density of 5 � 103 cells per well in 96-well plates. The p69cells were transfected with pCMV KIF3a or pCMV vectorconstructs at a density of 5 � 103 cells per well in 96-wellplates. The growth and viability effects of KIF3a knockdownon LNCaP cell or overexpression on p69 cell were determinedusingMTTassay (Promega) as described in themanufacturer'sinstruction. Three independent experiments were performedand each experiment was performed in triplicates. Clonogenicgrowth of cells was evaluated by assessing plating efficiency foranchorage-independent growth in soft agar using an approachdescribed previously (34). Colonies were scored at day 16 afterplating. The results were shown as a mean � SD of coloniesfrom three independent experiments.

In vitro cell migration and invasion assaysIn vitro cell migration was determined with a 24-well

Transwell plate (8-mmpore polycarbonate membrane inserts,

Corning Inc.) according to the manufacturer's protocol. Thecells were trypsinized, washed, and suspended in serum-freemedium. A total of 1 � 105 cells were placed in the upperchamber of the insert and the lower chamber of the insert wasfilled with migration-inducing medium that contained 10%FBS and 40 mg/mL Collagen I (Sigma). After incubation for24 hours, the cells on the top chambers were gently removedwith cotton swabs. The migratory cells on the bottom of themembrane were fixed with 100% methanol for 10 minutesand stained with 0.1% crystal violet for 30 minutes. Fivedifferent views were randomly chosen under a microscope,and the cell number on the bottom side of the membrane wascounted.The resultswere shown as amean� SDofmigrationcells from three independent experiments.The cancer cell invasive ability with KIF3a knockdown

and control was examined with a BD BioCoat MatrigelInvasion Chamber (BD Biosciences) according to the man-ufacturer's protocol. For these assays, 2.5 � 104 cells wereplated in the Matrigel-coated chamber inserts. Cells weresuspended in amediumwithout serumor growth factors, andmedium with 10% serum was used as a chemoattractant inthe lower chamber. After incubation at 37�C for 18 hours,the noninvasive cells on the top chambers were gentlyremoved with cotton swabs. The invading cells on the

Table 1. KIF3a expression levels and clinic characteristics of prostate cancer

a. KIF3a expression levels and the Gleason scores of prostate cancer

IHC score

Gleason scores 0–1þ 2þ 3þ Total

Normal 9 (100%) 0 0 93–5 4 (50%) 4 (50%) 0 86–7 16 (47%) 17 (50%) 1 (3%) 348–10 16 (50%) 9 (28%) 7 (22%) 32ND 6Total 89

b. KIF3a expression levels and the TNM stages of prostate cancer

IHC score

TNM stages 0–1þ 2þ 3þ Total

T0 9 0 0 9T1 2 (67%) 1 (33%) 0 3T2 29 (66%) 12 (27%) 3 (7%) 44T3 7 (28%) 16 (64%) 2 (8%) 25T4 3 (38%) 2 (25%) 3 (38%) 8Total 89

c. KIF3a expression levels and metastasis status of prostate cancer

IHC score

Metastatic stages 0–1þ 2þ 3þ Total

Normal 9 (100%) 0 0 9M0 32 (56%) 23 (40%) 2 (4%) 57M1 10 (43%) 8 (35%) 5 (22%) 23Total 89

Liu et al.





underside of themembranewere fixed in 100%methanol for10 minutes, air dried, stained in 0.1% crystal violet, andcounted under amicroscope. The numbers of invaded cells infive random microscopic fields were counted and the resultswere shown as the fold change relative to shRNA control.

Cell-cycle analysisCell-cycle analysis was performed using propidium iodide

staining as described previously (35). In brief, LNCaP cellswere transduced with the lentiviral constructs for 2 days, andwere trypsinized and washed with PBS, and fixed in 70%ethanol for 24 hours. Cells were then washed once in coldPBS and stained with PBS containing 20 mg/mL propidiumiodide, 0.1% triton-100, and 200mg/mLDNase-free RNasefor 30 minutes at room temperature while protected from

light. The cell cycle was analyzed using FACSCalibur flowcytometer and Cell Quest software (BD Biosciences).

TCF reporter assayTo determine the effects of KIF3a on b-catenin/TCF

reporter constructs, cells were transfected in 24-well plates,with either TOPFLASH or FOPFLASH (100 ng) and theinternal control plasmid pRL–TK (5 ng; Promega) and otherplasmids indicated, using LipofectAMINE 2000 Reagent(Invitrogen). The cells were incubated for 36 hours at 37�C,washed with PBS, and then lysed to measure luciferasereporter gene expression by dual-luciferase reporter assaysystem (Promega). Firefly luciferase activity was normalizedto Renilla luciferase activity. All results are expressed as amean � SD for independent triplicate cultures.

Control

0

0

2

21 3 5

4

4

6 8

*

*

Co

ntr

ol

Co

ntr

olshControl

KIF3a

sh

KIF

3a

-1

shKIF3a-1

shKIF3a-2

sh

KIF

3a

-2

Co

ntr

ol

sh

KIF

3a

-1

sh

KIF

3a

-2

Control shKIF3a-1 shKIF3a-2

pC

MV

-KIF

3apCMV-KIF3a

Re

lative

ce

ll p

rolif

era

tio

n (

%)

Re

lative

ce

ll p

rolif

era

tio

n (

%)

Re

lative

nu

mb

er

of

co

lon

ies

Day

Day

P < 0.005

P < 0.005

120

100

80

60

40

20

0

120

100

80

60

40

20

0

1.2

1

0.8

0.6

0.4

0.2

0

KIF3a

β-Actin

β-Actin

A

B

C

*

Figure 2. KIF3a regulates prostatecell proliferation and anchorage-independent cell growth. A,proliferation rate of p69 cells withoverexpression of KIF3a versusvector control. The cells werecultured in 96-well plates at adensity of 5�103perwell in 100mLgrowth medium after transienttransfection of pCMV–KIF3aconstruct or pCMV vector control.The proliferation of KIF3a-overexpressing cells wasmeasured by MTT assays at theindicated number of days. KIF3aoverexpression efficiency wasdetermined using Westernblotting. b-Actin was used as aloading control. Values, mean �SD of three independentexperiments performed intriplicates. �, P < 0.05. B,proliferation rate of LNCaP cellswith knockdown of KIF3a versusvector control. LNCaP cells (2 �105/well) were transducted withtwo different lentiviral shKIF3a(shKIF3a-1 and shKIF3a-2) orcontrol shRNAand then cultured in96-well plates at a density of 5 �103 per well in 100 mL growthmedium. MTT assay wasperformed at the indicated timepoints. Values, mean�SD of threeindependent experimentsperformed in triplicates. �,P < 0.05.KIF3a knockdown efficiency wasdetermined using Westernblotting as shown. C, quantitativedata of soft agar colony formationassay and representative imagesof colonies by LNCaP cells16 days posttransduction with twoshRNA-targeting KIF3a ornegative control. Colonies >50 mmin diameter were counted. Data,mean � SD of three independentexperiments performed intriplicates.






Quantitative real-time PCRTotal RNA was extracted from cultured cells by homog-

enizing in TRizol Reagent according to the manufacturer'sinstruction (Invitrogen). The Superscript First-Strand Kitwas used to synthesize first-strand cDNA (Invitrogen).Quantitative real-time PCR (qRT-PCR) was carried outwith iQ SYBR Green Supermix (Bio-Rad) using primers:HEF1, forward primer 50-ACTGATGCAGCAGACCT-

TTG-30, reverse primer 50-TGGCACTTGGTAGATGG-TGT-30; MMP9, forward primer: 50-CCGGACCAAGGA-TACAGTTT-30, reverse primer: 50–CGGCACTGAGGA-ATGATCTA-30; cyclin D1, forward primer 50-CTGTTT-GGCGTTTCCCAGAGTCATC-30, reverse primer 50-AGCCTCCTCCTCACACCTCCTC-30. Quantitation ofgene expressionwas calculated by the comparativeCt method.GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was

sh

Co

ntr

ol

sh

KIF

3a

-1

sh

KIF

3a

-2

sh

Co

ntr

ol

sh

KIF

3a

-1

sh

KIF

3a

-2

shControl shKIF3a-1 shKIF3a-2 shControl

shControl

shKIF3a-1

shKIF3a-1

shKIF3a-2

shKIF3a-2shControl shKIF3a-1 shKIF3a-2

KIF3a

LNCaP C4-2B

β-Actin

KIF3a

1.4

1.2

1

0.8

0.6

0.4

0.2

0

1.4

1.2

1

0.8

0.6

0.4

0.2

0

β-Actin

P < 0.005

P < 0.005

P < 0.005

P < 0.005

Rela

tive

mig

rate

d c

ells

(fold

change)

Rela

tive

mig

rate

d c

ells

(fold

change)

A

B

C F

E

D

Figure 3. Effects of KIF3a depletion on PCa cell migration and invasion. A to C, cell migration analysis of LNCaP cells with lentiviral shRNA–mediateddepletion of KIF3a or shRNA control. Cells that migrated through Transwell microporous membrane were fixed in methanol and stained with crystal violet.A, KIF3a knockdown efficiencywas determined usingWestern blotting as shown. b-Actin was used as a loading control. B, photos are representative fields ofmigration cells on the membrane. Scale bars, 100 mm. C, bar graphs represent the relative fold difference of migrated cell numbers compared withcontrol cells.P < 0.005 comparedwith shRNA control–transducted cells. D to F, effects of KIF3a depletion on invasion of C4-2B cells. C4-2B cells transducedwith scramble shRNA, KIF3a shRNA-1, or KIF3a shRNA-2 were plated on Matrigel-coated Transwell filters containing 10% FBS in the bottom well.Cells that migrated through Matrigel-coated Transwell membrane were fixed in methanol and stained with crystal violet. D, KIF3a knockdown efficiency wasdetermined using Western blotting as shown. b-Actin was used as a loading control. E, representative images of cells on the Transwell filters. Scalebars, 100 mm. F, the numbers of invading cells were counted and the invasion values are shown as fold change of invading cells relative to control with themeans � SD of three experiments performed in triplicate. P < 0.005 compared with shRNA control cells.

Liu et al.





KIF3a

A B

C D

E F

KIF3a

KIF3a

KIF3a

200

160

120

80

40

0

β-Actin

β-catenin

β-catenin

β-catenin

β-catenin

β-catenin

shControl

shCon

trol

pCM

V-Con

trol

pCM

V-KIF

3a

shKIF3a

shCTR

*

**

*

**

**

**

shKIF3a

shCTR80

70

60

50

40

30

20

10

0

shKIF3a

shCTR

1.4

1.2

1

0.8

0.6

0.4

0.2

0

14

12

10

8

6

4

2

0

C4-2Bp69

shKIF3a

KIF3a

pCMV

pCMV KIF3a

pCMV

pCMV KIF3a

shKIF

3a

Rela

tive

density (

%)

Rela

tive

Flu

ore

scence

(arb

itra

ry u

nits)

Rela

tive

lucife

rase a

ctivity

(fold

change)

Rela

tive

lucife

rase a

ctivity

(fold

change)

**

Figure 4. KIF3a modulates the expression of b-catenin in PCa cells. A, KIF3a modulates the protein expression of b-catenin. p69 cells were transientlytransfected with pCMV–tag2b–KIF3a or pCMV–tag2b vector for 48 hours, and LNCaP cells were transduced with lentiviral shKIF3a or shControlfor 48 hours, cells were then collected for Western blot analysis using antibodies as indicated. b-Actin was used as a loading control. B, densitometricquantification of the protein bands of (A) as described in Fig. 1. �, P < 0.05; ��, P < 0.005. C, KIF3a and b-catenin expression was analyzed byimmunostaining in C4-2B cell following silencing of KIF3a. C4-2B cells were transduced with lentiviral shKIF3a or shControl for 48 hours, and cellswere then immunostained with the indicated antibodies after methanol fixation and permeabilization with 0.5% Triton X-100. Of note, 40,6-diamidino-2-phenylindole was used to counter stain the nuclei. D, the intensities of immunostained fluorescence in individual cell were measured by ImageJ softwareand expressed as relative fluorescence (arbitrary units). Figure shows mean � SD of four independent measurements. �, P < 0.05 and ��, P < 0.005compared with control cells. E and F, KIF3a regulates b-catenin/TCF–driven transcription. KIF3a overexpression p69 cell (E), or KIF3a knockdownC4-2B cell (F) were transfected with TOPflash (100 ng) and FOPflash (100 ng) along with 5 ng of pRL-SV40, and luciferase activities were determined36 hours after transfection. Renilla luciferase vector was used as an internal control. Normalized luciferase activity of control cells was set to 1, and resultsare given in relative luciferase activity. Statistics, ��, P < 0.005 compared with control cells.






Control

ControlshControl

shControl

Perc

enta

ge o

f cells

(%

)

shKIF3a

shKIF3a

MMP-9

MMP-9

700

600

500

400

300

200

100

0

140

120

100

80

60

40

20

0

60

50

40

30

20

10

0

HEF1

LNCaPHEF1

******

***

**

**

**

*

*

*

*

**

**

**

*****

***

p69

Cyclin D1

Cyclin D1

MMP-9KIF3a HEF1

G0–G1SG2–M

Cyclin D1 MMP-9KIF3a HEF1 Cyclin D1

Control

shC

ontrol

shK

IF3a

KIF3a

pC

MV

KIF

3a

pCMV KIF3a

OE KIF3a

2.5

2

1.5

1

0.5

0

Re

lative

mR

NA

fo

ld c

ha

ng

es

Rela

tive

density (

%)

Rela

tive

density (

%)

KD KIF3a

β-Actin

A

C

B

D

E

Figure 5. KIF3a regulates MMP9, HEF1, and cyclin D1 expression in PCa cells. A, qRT-PCR analysis of MMP9, HEF1, and cyclin D1 transcripts from p69 cellsafter overexpression of KIF3a for 48 hours or from LNCaP cells after KIF3a depletion with lentiviral shKIF3a for 48 hours. Data, mean values � SDfrom three independent experiments. Statistics, �, P < 0.05; ��, P < 0.005; ��, P < 0.001. B, Western blot analysis of whole-cell lysates against antibodies asindicated from the same cells described in A. b-Actin was used as a loading control. C and D, quantification of the Western bands of B. All proteinlevels were normalized to b-actin level. Densitometric quantification of protein was relative to the control set in each case (indicated by a value of percentage).Bar graph, mean � SD of values from three independent experiments. �, P < 0.05; ��, P < 0.01, and ��, P < 0.005. (Continued on the following page.)

Liu et al.





chosen as the housekeeping gene for template concentrationcontrol. Data were analyzed according to the comparativecycle threshold (Ct)method andwere normalized byGAPDHexpression in each sample as we previously described (32).

Statistical analysisData are expressed as a mean � SD from at least three

experiments. Statistical analyses were performed by using theStudent t test and c2 tests as appropriate. The c2 test wasapplied to compare KIF3a expression level with the Gleasonscores, TNM (tumor–node–metastasis) stage, and metasta-sis status of prostate cancer. P < 0.05 was consideredstatistically significant.

ResultsUpregulation of KIF3a in prostate cancer cell lines andprimary tumor tissuesTo investigate the role of KIF3a in prostate tumorigenesis,

we first compared KIF3a protein levels in six prostate cancercell lines and three benign prostate epithelial cell lines byWestern blotting.We found that the average expression levelof KIF3a in prostate cancer cell lines was almost double thatof the level in the benign cell lines (Fig. 1A andB).Moreover,the three metastatic prostate cancer cell lines (M12, M182,and C4-2B) had the highest expression levels of KIF3aamong all prostate cancer cell lines analyzed (Fig. 1B). Theaberrant expression of KIF3a observed in prostate cancer celllines was further validated in vivo byWestern blot analysis of18 paired fresh-frozen primary prostate tumors and theirmatched adjacent normal prostate tissues (Fig. 1C). Of the18 tumors, 12 of the 13 tumors with Gleason scores of 7 orabove showed relatively higher levels of KIF3a expressioncompared with matched normal tissues, whereas five (#3, 5,6, 10, and 15) of the six tumors with lowGleason scores (5 or6) showed relatively lower levels of KIF3a expression (Fig.1D). These data indicate that KIF3a is upregulated inprostate cancer and suggest that the expression level ofKIF3a likely increases as the tumor becomesmore aggressive.

KIF3a upregulation and prostate cancer progressionFor clinical relevance, we determined the expression levels

of KIF3a by IHC in a prostate tumor tissue microarray with89 unrelated prostate tissue sample cores. We found a highlevel expression of KIF3a (scored asþ2 andþ3) in 39 of 74(53%) tumor tissues but absent or very low level expressionof KIF3a (scored as 0 and þ1) in all nine normal prostatetissues (Table 1; Fig. 1E). KIF3a staining was observed inboth the cytoplasm and nucleus of cancer cells and stainingscores were obtained by intensity� percentage of cells. Therepresentative images of the KIF3a staining in normal orprostate tumor tissues are presented in Fig. 1F.To examine the potential correlation of KIF3a expres-

sion level and tumor characteristics, we compared the

tumors with strong KIF3a expression (the highest level ofIHC staining scored as þ3) to their respective Gleasonscore, TNM stage, and metastatic status. We observed astepwise correlation between strong KIF3a expression andthe tumor Gleason score (Table 1a). High KIF3a expres-sion was observed in 22% of tumors with Gleason scoresof 8 to 10 and 3% with Gleason scores of 6 to 7 but wasnot observed in normal or tumor tissues with Gleasonscores below 6. The correlation between strong KIF3aexpression and high Gleason scores of 8 to 10 is statis-tically significant (c2 test; P < 0.005). Similarly, elevatedKIF3a expression correlates significantly with the highesttumor TNM grade, T4 (c2 test; P < 0.005; Table 1b). Ofnote, 38% of T4 tumors were scored as high KIF3aexpression compared with 8% of T3, 7% of T2, and0% of T1 or T0 samples. Consistently, increased KIF3aexpression was observed in 22% of metastatic prostatetumors compared with only 4% of the localized prostatetumors (Table 1c). The different expression levels ofKIF3a between metastatic (M1) and nonmetastatic pros-tate tumors (M0) are statistically significant (c2 test; P <0.01). Taken together, these data indicate that high-levelexpression of KIF3a is associated with aggressive prostatecancer, suggesting that increased KIF3a expression maypromote prostate cancer progression and metastasis.

KIF3a regulates prostate cancer cell proliferation in vitroTo investigate the effect of KIF3a on prostate cancer

progression and aggressiveness, we performed cell prolif-eration assay and anchorage-independent colony forma-tion assay. As shown in Fig. 2A, transient overexpressionof KIF3a in benign prostate p69 cells, which had low-levelexpression of KIF3a, increased cell proliferation (P <0.05). In contrast, silencing KIF3a in LNCaP cells, whichhave a high-level expression of KIF3a, using the twodifferent lentiviral KIF3a–shRNA constructs specificallyagainst human KIF3a significantly reduced the cell growthrate compared with the cells expressing scramble shRNA(P < 0.05; Fig. 2B). Silencing KIF3a in LNCaP cellssignificantly reduced (P < 0.005) the number of colonieson soft agar compared with the control cells (Fig. 2C).These data demonstrate that upregulation or depletion ofKIF3a promotes or inhibits cell growth, respectively,suggesting that KIF3a upregulation observed in the major-ity of prostate cancer probably promotes prostate cancercell progression.

KIF3a induces prostate cancer cell migration andinvasion in vitroThe highest levels of KIF3a observed in metastatic pros-

tate cancer cell lines and in high percentage of metastaticprostate tumor tissues suggest that KIF3amay be involved in

(Continued.) E, KIF3amodulatescell-cycleprogression. LNCaPcells after KIF3adepletionwith lentiviral shKIF3a for 48hours asdescribed inBwere subjectedto flow cytometry analysis. Silencing of KIF3a leads to accumulation of LNCaP cells in G0–G1 phase of the cell cycle. On the basis of DNA content,cells in G0–G1 can be distinguished from those in S and G2–M. The percentages of cells in each phase are shown. Values, means � SD; n ¼ 2. Bars withasterisks are significantly different from control values of the corresponding cell-cycle phase, �, P < 0.05; ��, P < 0.005.






prostate cancer metastasis. Therefore, we assessed the effectsof KIF3a manipulation on cell migration and invasion byBoyden Chamber assays. Silencing of KIF3a by lentiviralKIF3a–shRNA in LNCaP cells reduced cell migration ratebymore than 80% comparedwith the control cells (Fig. 3A–C). In addition, the invasive capability of the C4-2B cell (abone metastatic LNCaP-derivative prostate cancer cell line)was reduced by more than 75%when KIF3a was reduced byshRNA compared with the control cells (Fig. 3D–F). Theseresults provide evidence that upregulation of KIF3a may beinvolved in the metastatic process of prostate cancer.

KIF3a activates the Wnt/b-catenin signaling pathway inprostate cancer cellsTo determine the potential downstream effectors of the

KIF3a signaling, which may have contributed to prostatecancer cell proliferation and invasion, we examined theexpression levels of b-catenin byWestern blotting followingexogenous KIF3a expression in p69 cells or silencing ofKIF3a in LNCaP cells. We found that overexpression ofKIF3a in p69 cells induced upregulation of b-catenin,whereas silencing of KIF3a in LNCaP cells reduces theexpression level of b-catenin (Fig. 4A and B). In agreementwith the Western blot data, immunofluorescence analysisindicates that silencing of KIF3a in LNCaP cells reducedb-catenin levels in both the cytoplasm and nuclei (Fig. 4Cand D). Moreover, TCF reporter assays showed that over-expression of KIF3a in p69 cells increases TCF activity,whereas silencing of KIF3a in C4-2B cells dramaticallyreduces TCF reporter activity (Fig. 4E and F). These resultsindicate that KIF3a contributes to prostate cancer cellproliferation and invasion probably through activation ofthe Wnt/b-catenin signaling pathway.

KIF3a regulates expression levels of cyclin D1, MMP9,and HEF1 in prostate cancer cellsCyclin D1 is a Wnt/b-catenin–signaling target, and eleva-

tion of cyclin D1 accelerates cell-cycle progression and cellproliferation (36, 37). We, therefore, investigated whetherKIF3a regulates the expression of cyclinD1 in prostate cancer.The results from both Western blot analysis and quantitativereal-time reverse transcription PCR analysis indicated that thelevel of cyclin D1 was increased in p69 cells ectopically exp-ressing KIF3a but was reduced in LNCaP cells expressingKIF3a-specific shRNA (Fig. 5A–D). In addition, cell flowcytometry analysis of LNCaP cells expressing either KIF3a-specific shRNAor scramble shRNAdemonstrated thatKIF3a-dependent downregulation of cyclin D1 by KIF3a shRNAincreased G0–G1 cell population (Fig. 5E). These resultssuggest that KIF3a-induced cell proliferation is probablythrough transactivation of cyclinD1 due to KIF3a-dependentactivation of the Wnt signaling pathway in prostate cancer.Next, we examined the effect of KIF3a on expression of

MMP2/9 and HEF1, the three Wnt-signaling targets thathave been shown to be involved in cancer cell migration/invasion and are the hallmark of prostate cancer metastasis(38–40). qRT-PCR analysis showed that the RNA levels ofMMP9 and HEF1 were increased in p69 cells following

overexpression of KIF3a while being reduced in C4-2B cellsfollowing KIF3a depletion by KIF3a shRNA (Fig. 5A).However, no obvious change of MMP2 activity or expres-sion was detected in these cells (data not shown). Proteinlevels of MMP9 and HEF1 showed the consistent up-/downregulatory tendency as mRNA level of KIF3a aftermanipulation of KIF3a expression level except the MMP9protein was barely detected in p69 cells (Fig. 5B–D). Takentogether, these data indicate that KIF3a may modulateprostate cancer migration and invasion throughWnt-depen-dent activation of MMP9 and HEF1 contributing to pros-tate cancer metastasis.

KIF3a increases DVL2 phosphorylation via CK1 inprostate cancer cellsTo understand the mechanism of KIF3a-dependent acti-

vation of Wnt signaling, we studied the stability and thephosphorylation of DVL2 in prostate cancer because theincrease of DVL2 phosphorylation activates canonical Wntsignaling in prostate cancer (41). We found that the expres-sion of KIF3a in p69 cells increased DVL2 phosphorylationand its protein level (Fig. 6A and B). In contrast, silencing ofKIF3a in LNCaP cells reduced DVL2 phosphorylation andits protein level (Fig. 6A and B). In addition, silencing ofDVL2 by DVL2-specific shRNA following overexpressionof KIF3a in p69 cells reduced phosphorylation of DVL2 andexpression level of b-catenin (Fig. 6C and D). Furthermore,we also showed that KIF3a-induced DVL2 phosphoryla-tion is CK1-dependent because treatment of p69 cellsoverexpressing KIF3a with the CK1 inhibitor D4476abolished KIF3a-induced DVL2 phosphorylation (Fig.6E). These results indicate that upregulation of KIF3ain advanced prostate cancer activates the Wnt signalingpathway probably through induction of CK1-dependentDVL2 phosphorylation.

DiscussionThe Wnt signaling pathway is frequently activated in

advanced prostate cancer. But the underlying molecularmechanisms are poorly understood. In this study, we showedfor the first time that the motor protein KIF3a is a potentactivator of the Wnt signaling pathway in human prostatecancer.We showed that KIF3a is upregulated in themajorityof prostate cancer cell lines and primary tumor tissues. Theupregulation of KIF3a activates the Wnt signaling pathwayas demonstrated by increased CK1-dependent DVL2 phos-phorylation, b-catenin activation, and increased expressionof Wnt-signaling target genes such as cyclin D1. Thesefindings suggest the presence of a novel KIF3a–DVL2–b-catenin axis in human prostate cancer (Fig. 6F). CyclinD1 is a well-known Wnt target gene. It is often associatedwith advanced prostate cancer for increasing the proliferativepotential of human prostate cancer cells (42). We show thatKIF3a modulates cyclin D1 expression and KIF3a-depen-dent downregulation of cyclin D1 increases cell populationin G0–G1 phase. These data provide evidence that upregula-tion of KIF3a is one of the molecular mechanisms activating

Liu et al.





shC

ontr

ol

shC

ontr

ol

shC

ontr

ol

shControl

shK

IF3a

shKIF3aC

ontr

ol

Contr

ol

pC

MV

KIF

3a

pC

MV

KIF

3a

pC

MV

KIF

3a

pCMV KIF3a

* *

*

***

Control

250

200

150

100

50

0

400

350

300

250

200

150

100

50

0

120

100

80

60

40

20

0

Control

LNCaPp69 Phospho

DVL2

DVL2

shD

VL2

DVL2

DVL2

Phospho

DVL2

KIF3a + shDVL2

Phospho DVL2

KIF3a

Re

lative

de

nsity (

%)

Rela

tive

density (

%)

KIF3a + shControl

KIF3a

KIF3a

KIF3a

KIF3a

KIF3a

Dvl2 P

KIF

3a

KIF

3a

+ D

4476

KIF3a

**

**

β-Actin

β-Actin

β-Cat

β-Cat β-Cat

β-Catβ-Cat

β-Cat

β-Catβ-Cat

β-Cat β-Cat β-Cat

β-Catβ-Cat

MMP9 MMP9

Cyclin D1 Cyclin D1 Cell cycle

MetastasisHEF1 HEF1

β-Cat

β-catenin

β-catenin

β-Actin

A B

C

E F

D

Figure 6. KIF3a modulates CK1-dependent DVL2 phosphorylation in PCa cells. A, KIF3a induces DVL2 expression and phosphorylation. Westernblot analysis of DVL2 in whole-cell lysates of the same samples as described in Fig. 5A for DVL2 expression and phosphorylation levels. B, densitometryanalysis of protein bands in figure A for DVL2 expression and phosphorylation levels are shown as means � SD of three independent experiments.�, P < 0.05; ��, P < 0.001. C and D, b-catenin levels were assessed by Western blot analysis following overexpression of KIF3a and silencing of DVL2in p69 cells. Levels of phosphorylated DVL2 and b-catenin were normalized to b-actin. Densitometry analyses of protein bands in C were displayedas the mean � SD of values from three independent experiments. ��, P < 0.005. E, CK1 inhibitor abolishes KIF3a-induced DVL2 phosphorylation.Western blot analysis of DVL2 in p69 cells overexpressing KIF3a following starvation and treatment with 100 mmol/L D4476 or vehicle (dimethylsulfoxide) for 2 hours. F, ligand-independent activation of KIF3a–DVL2–b-catenin signaling in PCa. In PCa, KIF3a probably functions as an agonist of theb-catenin/TCF signaling by promoting DVL2 phosphorylation and stabilizing b-catenin in the absence of Wnt stimulation. Subsequently, stabilizedb-catenin activates the transcriptional targets, including MMP9, HEF1, and cyclin D1. Elevation of cyclin D1 increases PCa cell proliferation andtumorigenesis, whereas elevation of MMP9 and HEF1 contributes to the regulation of PCa cell migration and invasion for potential metastasis of PCa.






the Wnt signaling pathway in a subgroup of patients withprostate cancer, particularly in advanced prostate cancer.Metastatic potential is the hallmark of the advanced pros-

tate cancer. Many studies on dysregulation of Wnt signalingin prostate cancer suggest that aberrant activation ofb-cateninis correlated with tumor progression and metastasis(11, 43, 44). In this report, we showed that KIF3a regulatesprostate cancer cell migration and invasion through KIF3a-dependent activation of MMP-9 and HEF1, which are twoWnt-signaling target genes (45, 32). It has been reportedthat excessive production of MMP-9, the major MMP,correlates with aggressive prostate cancer (46). HEF1, a keyprotein in cell migration, has recently been shown to promoteprostate cancer invasion (47). We showed that KIF3a mod-ulates the expression of MMP9 and HEF1 in metastaticprostate cancer cell lines. Importantly, depletion of KIF3a byshRNA reduced the expression of MMP9, HEF1, and theinvasive properties of metastatic C4-2B cells. These datasupport the notion that KIF3a activates the Wnt signalingpathway, leading to transactivation of MMP-9 and HEF1,and increasing the metastatic potential of prostate cancer.The association of KIF3a upregulation with metastatic

potential in prostate cancer is also clinically evident in prostatecancer specimens. We observed that among all the prostatecancer cell lines analyzed, the three metastatic prostate cancercell lines expressed the highest levels of KIF3a (Fig. 1A).Consistently, the highest level expression of KIF3a, shown bystrong IHC staining (scored as þ3), was detected in a highfrequency of tumors with a higher Gleason score (8–10), highTNM grade (T4), and metastasis (Table 1). These datafurther suggest that upregulation of KIF3a may associatewith prostate cancer aggressiveness and metastasis. However,although the association is statistically significant, the num-bers of tissue specimens analyzed in this study are rather small.Further analysis of large sample sets of prostate cancer will benecessary to determine the significance of upregulation ofKIF3a in clinical practice and possible therapeutic interven-tions for advanced prostate cancer.

In summary, we demonstrated for the first time thatKIF3a is a novel agonist of the Wnt signaling pathway inhuman prostate cancer and an oncoprotein associated withprostate cancer progression and metastasis. KIF3a controlsprostate cancer cell proliferation and invasion, at least in part,through the induction of DVL2 phosphorylation and tran-scriptional activation of three Wnt target genes: cyclin D1,HEF1, and MMP-9 (Fig. 6F). Furthermore, this studyreveals a new molecular mechanism for activation of theWnt signaling pathway in prostate cancer through theKIF3a–DVL2–b-catenin axis. KIF3a may be a potentialtherapeutic target for advanced prostate cancer inwhichWntsignaling is activated.

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

Authors' ContributionsConception and design: Z. Liu, Y. Li, W. LiuDevelopment of methodology: Z. Liu, Z. Wang, Y. Li, Z. Wang, W. LiuAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): Z. Liu, R.E. Rebowe, Z. Wang, J.S. DePaolo, W. LiuAnalysis and interpretation of data (e.g., statistical analysis, biostatistics, compu-tational analysis): Z. Liu, R.E. Rebowe, J.S. DePaolo, W. LiuWriting, review, and/or revision of the manuscript: Z. Liu, R.E. Rebowe, W. LiuAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): Z. Liu, J. Guo, C. Qian, W. LiuStudy supervision: Z. Liu, W. Liu

AcknowledgmentsThe authors thankDr.HaojieHuang andDr. Shahriar Koochekpour for providing

some of the prostate cell lines and Euraque Samantha for critical editing of this article.

Grant SupportThis work was supported by NIH grants R01 CA115555, R01 CA115555-03S1,

and P20 MD004817, and Research Enhancement Fund from Louisiana StateUniversity School of Medicine (New Orleans, LA; to W. Liu).

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be herebymarked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Received August 5, 2013; revisedNovember 1, 2013; acceptedDecember 23, 2013;published OnlineFirst January 10, 2014.

References1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer

J Clin 2013;63:11–30.2. Stark JR, Perner S, Stampfer MJ, Sinnott JA, Finn S, Eisenstein AS,

et al. Gleason score and lethal prostate cancer: does 3þ4 ¼ 4þ3?J Clin Oncol 2009;27:3459–64.

3. Greene KL, Albertsen PC, Babaian RJ, Carter HB, Gann PH, Han M,et al. Prostate-specific antigen best practice statement: 2009 update.J Urol 2013;189:S2–11.

4. Albertsen PC, Hanley JA, Fine J. 20-year outcomes following conser-vativemanagement of clinically localized prostate cancer. JAMA2005;293:2095–101.

5. Uysal-Onganer P, Kawano Y, Caro M, Walker MM, Diez S, Darring-ton RS, et al. Wnt-11 promotes neuroendocrine-like differentiation,survival, and migration of prostate cancer cells. Mol Cancer 2010;9:55.

6. Hall CL, Bafico A, Dai J, Aaronson SA, Keller ET. Prostate cancer cellspromote osteoblastic bone metastases through Wnts. Cancer Res2005;65:7554–60.

7. Verras M, Brown J, Li X, Nusse R, Sun Z. Wnt3a growth factorinduces androgen receptor-mediated transcription and enhances

cell growth in human prostate cancer cells. Cancer Res 2004;64:8860–6.

8. Zhu H, Mazor M, Kawano Y, Walker MM, Leung HY, Armstrong K,et al. Analysis of Wnt gene expression in prostate cancer: mutualinhibition by WNT11 and the androgen receptor. Cancer Res 2004;64:7918–26.

9. de la Taille A, Rubin MA, Chen MW, Vacherot F, de Medina SG,BurchardtM, et al. b-Catenin–related anomalies in apoptosis-resistantand hormone-refractory prostate cancer cells. Clin Cancer Res2003;9:1801–7.

10. Chesire DR, Ewing CM, Gage WR, Isaacs WB. In vitro evidence forcomplex modes of nuclear b-catenin signaling during prostate growthand tumorigenesis. Oncogene 2002;21:2679–94.

11. Chen G, Shukeir N, Potti A, Sircar K, Aprikian A, Goltzman D, et al.Upregulation of Wnt-1 and b-catenin production in patients withadvanced metastatic prostate carcinoma: potential pathogenetic andprognostic implications. Cancer 2004;101:1345–56.

12. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B,et al. Activation of beta-catenin-Tcf signaling in colon cancer bymutations in beta-catenin or APC. Science 1997;275:1787–90.

Liu et al.





13. Liu W, Dong X, Mai M, Seelan RS, Taniguchi K, Krishnadath KK, et al.Mutations in AXIN2 cause colorectal cancer with defective mismatchrepair by activating beta-catenin/TCF signalling. Nat Genet 2000;26:146–7.

14. Jin LH, Shao QJ, Luo W, Ye ZY, Li Q, Lin SC. Detection of pointmutations of the Axin1 gene in colorectal cancers. Int J Cancer2003;107:696–9.

15. He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT, et al.Identification of c-MYC as a target of the APC pathway. Science 1998;281:1509–12.

16. Tetsu O,McCormick F. Beta-catenin regulates expression of cyclin D1in colon carcinoma cells. Nature 1999;398:422–6.

17. Xia D, Holla VR, Wang D, Menter DG, DuBois RN. HEF1 is a crucialmediator of the proliferative effects of prostaglandin E(2) on coloncancer cells. Cancer Res 2010;70:824–31.

18. Zucker S, Vacirca J. Role of matrix metalloproteinases (MMPs) incolorectal cancer. Cancer Metastasis Rev 2004;23:101–17.

19. Smalley MJ, Sara E, Paterson H, Naylor S, Cook D, Jayatilake H, et al.Interaction of axin and Dvl-2 proteins regulates Dvl-2-stimulated TCF-dependent transcription. EMBO J 1999;18:2823–35.

20. Chesire DR, Ewing CM, Sauvageot J, Bova GS, Isaacs WB. Detectionand analysis of beta-catenin mutations in prostate cancer. Prostate2000;45:323–34.

21. Voeller HJ, Truica CI, Gelmann EP. Beta-catenin mutations in humanprostate cancer. Cancer Res 1998;58:2520–3.

22. Huangfu D, Liu A, Rakeman AS, Murcia NS, Niswander L, AndersonKV. Hedgehog signalling in the mouse requires intraflagellar transportproteins. Nature 2003;426:83–7.

23. Jimbo T, Kawasaki Y, KoyamaR, Sato R, Takada S, Haraguchi K, et al.Identification of a link between the tumour suppressor APC and thekinesin superfamily. Nat Cell Biol 2002;4:323–7.

24. Teng J, Rai T, Tanaka Y, Takei Y, Nakata T, HirosawaM, et al. The KIF3motor transports N-cadherin and organizes the developing neuroe-pithelium. Nat Cell Biol 2005;7:474–82.

25. Corbit KC, Shyer AE, DowdleWE, Gaulden J, Singla V, ChenMH, et al.Kif3a constrains beta-catenin-dependent Wnt signalling through dualciliary and non-ciliary mechanisms. Nat Cell Biol 2008;10:70–6.

26. Qiu N, Xiao Z, Cao L, Buechel MM, David V, Roan E, et al. Disruption ofKif3a in osteoblasts results in defective bone formation and osteope-nia. J Cell Sci 2012;125:1945–57.

27. Hayward SW, Dahiya R, Cunha GR, Bartek J, Deshpande N, et al.Establishment and characterization of an immortalized but non-trans-formed human prostate epithelial cell line: BPH-1. In VitroCell Dev BiolAnim 1995;31:14–24.

28. Bae VL, Jackson-Cook CK, Brothman AR, Maygarden SJ, Ware JL.Tumorigenicity of SV40 T antigen immortalized human prostate epi-thelial cells: association with decreased epidermal growth factorreceptor (EGFR) expression. Int J Cancer 1994;58:721–9.

29. Bae VL, Jackson-Cook CK, Maygarden SJ, Plymate SR, Chen J,WareJL.Metastatic sublines of anSV40 large T antigen immortalized humanprostate epithelial cell line. Prostate 1998;34:275–82.

30. Jackson CC, Bae VL, Edelman W, Brothman A, Ware J. Cytogeneticcharacterization of the human prostate cancer cell line P69SV40T andits novel tumorigenic sublines M2182 and M15. Cancer Genet Cyto-genet 1996;87:14–23.

31. Thalmann GN, Anezinis PE, Chang SM, Zhau HE, Kim EE, HopwoodVL, et al. Androgen-independent cancer progression and bonemetas-tasis in the LNCaPmodel of human prostate cancer. Cancer Res 1994;54:2577–81.

32. Li Y, Bavaria JH, Wang Z, Guo J, Qian C, Thibodeau SN, et al. HEF1, anovel target of Wnt signalling, promotes colonic cell migration andcancer progression. Oncogene 2011;30:2633–43.

33. Guo J, Zheng L, Liu WY, Wang X, Wang Z, French AJ, et al. Frequenttruncating mutation of TFAM induces mitochondrial DNA depletionand apoptotic resistance in microsatellite-unstable colorectal cancer.Cancer Res 2011;71:2978–87.

34. Leone A, Flatow U, VanHoutte K, Steeg PS. Transfection of humannm23-H1 into the human MDA-MB-435 breast carcinoma cell line:effects on tumor metastatic potential, colonization and enzymaticactivity. Oncogene 1993;8:2325–33.

35. Darzynkiewicz Z, Huang X. Analysis of Cellular DNA Content by FlowCytometry. Curr Protoc Immunol, Chapter 5:unit 5.7, 2004.

36. ShtutmanM, Zhurinsky J, Simcha I, Albanese C, D'AmicoM, Pestell R,et al. The cyclin D1 gene is a target of the beta-catenin/LEF-1 pathway.Proc Natl Acad Sci U S A 1999;96:5522–7.

37. Resnitzky D, Gossen M, Bujard H, Reed SI. Acceleration of the G1/Sphase transition by expression of cyclins D1 and E with an induciblesystem. Mol Cell Biol 1994;14:1669–79.

38. Trudel D, Fradet Y, Meyer F, Harel F, Tetu B. Significance of MMP-2expression in prostate cancer: an immunohistochemical study.CancerRes 2003;63:8511–5.

39. Saleem M, Kweon MH, Johnson JJ, Adhami VM, Elcheva I, Khan N,et al. S100A4 accelerates tumorigenesis and invasion of humanprostate cancer through the transcriptional regulation of matrix metal-loproteinase 9. Proc Natl Acad Sci U S A 2006;103:14825–30.

40. O'Neill GM, Seo S, Serebriiskii IG, Lessin SR, Golemis EA. A newcentral scaffold for metastasis: parsing HEF1/Cas-L/NEDD9. CancerRes 2007;67:8975–9.

41. Mimeault M, Batra SK. Recent advances on multiple tumorigeniccascades involved in prostatic cancer progression and targetingtherapies. Carcinogenesis 2006;27:1–22.

42. Drobnjak M, Osman I, Scher HI, Fazzari M, Cordon-Cardo C. Over-expression of cyclin D1 is associated with metastatic prostate cancerto bone. Clin Cancer Res 2000;6:1891–5.

43. Culig Z, Hobisch A, Hittmair A, Peterziel H, Cato AC, Bartsch G, et al.Expression, structure, and function of androgen receptor in advancedprostatic carcinoma. Prostate 1998;35:63–70.

44. Hall CL, Keller ET. The role of Wnts in bone metastases. CancerMetastasis Rev 2006;25:551–8.

45. Wu B, Crampton SP, Hughes CC. Wnt signaling induces matrixmetalloproteinase expression and regulates T-cell transmigration.Immunity 2007;26:227–39.

46. Zhong WD, Han ZD, He HC, Bi XC, Dai QS, Zhu G, et al. CD147,MMP-1, MMP-2 and MMP-9 protein expression as significantprognostic factors in human prostate cancer. Oncology 2008;75:230–6.

47. GuoW, Ren D, Chen X, Tu X, Huang S, Wang M, et al. HEF1 promotesepithelial mesenchymal transition and bone invasion in prostate can-cer under the regulation of microRNA-145. J Cell Biochem 2013;114:1606–15.






2014;12:491-503. Published OnlineFirst January 10, 2014.Mol Cancer Res Zun Liu, Ryan E. Rebowe, Zemin Wang, et al. Advanced Prostate CancerKIF3a Promotes Proliferation and Invasion via Wnt Signaling in

Updated version

10.1158/1541-7786.MCR-13-0418doi:

Access the most recent version of this article at:

Cited articles

http://mcr.aacrjournals.org/content/12/4/491.full#ref-list-1

This article cites 46 articles, 19 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/12/4/491.full#related-urls

This article has been cited by 1 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/12/4/491To 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-13-0418

http://mcr.aacrjournals.org/content/12/4/491.full#ref-list-1

http://mcr.aacrjournals.org/content/12/4/491.full#related-urls

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

mailto:[email protected]

http://mcr.aacrjournals.org/content/12/4/491


kif3a promotes proliferation and invasion via wnt...

Documents