EphB4 Expression and Biological Significance in Prostate Cancer

Guangbin Xia,1S. Ram Kumar,

2,3Rizwan Masood,

2Sutao Zhu,

1Ramchandra Reddy,

6

Valery Krasnoperov,6David I. Quinn,

1Susan M. Henshall,

7Robert L. Sutherland,

7

Jacek K. Pinski,1Siamak Daneshmand,

4Maurizio Buscarini,

4John P. Stein,

4Chen Zhong,

2

Daniel Broek,5Pradip Roy-Burman,

2and Parkash S. Gill

1,2,6

Departments of 1Medicine, 2Pathology, 3Surgery, 4Urology, and 5Biochemistry and Molecular Biology, University of Southern California KeckSchool of Medicine; 6VasGene Therapeutics, Inc., Los Angeles, California; and 7Cancer Research Program, Garvan Institute of MedicalResearch, St. Vincent’s Hospital, Sydney, New South Wales, Australia

Abstract

Prostate cancer is the most common cancer in men. Advancedprostate cancer spreading beyond the gland is incurable.Identifying factors that regulate the spread of tumor into theregional nodes and distant sites would guide the developmentof novel diagnostic, prognostic, and therapeutic targets. Theaim of our study was to examine the expression and biologicalrole of EphB4 in prostate cancer. EphB4 mRNA is expressed in64 of 72 (89%) prostate tumor tissues assessed. EphB4 proteinexpression is found in the majority (41 of 62, 66%) of tumors,and 3 of 20 (15%) normal prostate tissues. Little or noexpression was observed in benign prostate epithelial cell line,but EphB4 was expressed in all prostate cancer cell lines tovarying degrees. EphB4 protein levels are high in the PC3prostate cancer cell line and several folds higher in ametastatic clone of PC3 (PC3M) where overexpression wasaccompanied by EphB4 gene amplification. EphB4 expressionis induced by loss of PTEN, p53, and induced by epidermalgrowth factor/epidermal growth factor receptor and insulin-like growth factor-I/insulin-like growth factor-IR. Knockdownof the EphB4 protein using EphB4 short interfering RNA orantisense oligodeoxynucleotide significantly inhibits cellgrowth/viability, migration, and invasion, and induces apo-ptosis in prostate cancer cell lines. Antisense oligodeoxynu-cleotide targeting EphB4 in vivo showed antitumor activity inmurine human tumor xenograft model. These data show a rolefor EphB4 in prostate cancer and provide a rationale to studyEphB4 for diagnostic, prognostic, and therapeutic applica-tions. (Cancer Res 2005; 65(11): 4623-32)

Introduction

Prostate cancer remains the most frequently diagnosed solidtumor and is the second leading cause of cancer-related death inmen in Western countries (1). Most advanced prostate cancers aretreated with androgen ablation, which is initially successful andresults in subjective improvement in about 80% of cases. However,the response is usually temporary and many tumors overcomeandrogen blockage and develop a hormone-unresponsive pheno-type after 12 to 18 months of endocrine therapy (2, 3). Theepidermal growth factor (EGF) and insulin-like growth factor (IGF)signaling pathways and mutation of p53 and PTEN have all beenimplicated in prostate cancer progression and metastasis (4–10).

These factors or their downstream effectors responsible for thecontinued proliferation and survival of resistant tumor cells arepotential targets for treatment of hormone-unresponsive prostatecancers (11).EphB4 is a member of the largest known family of receptor

protein tyrosine kinases, which plays an important role in a varietyof processes during embryonic development including patternformation, cell aggregation and migration, segmentation, neuraldevelopment, angiogenesis, and vascular network assembly (12–15).Specifically, EphB4, along with its ligand, Ephrin-B2, is essential forvascular remodeling, maturation and directed growth. BothEphrin-B2 and EphB4 targeted gene knock-outs in mice show lackof remodeling of the primitive vascular plexus to form maturevascular structures (13). EphB4 expression is restricted to venousendothelial cells, whereas Ephrin-B2 is restricted to arterialendothelial cells (16, 17). EphB4 and Ephrin-B2 are also essentialfor defining the boundaries between arterial and venous domains(17), which persist in the adult (18).EphB4 selectively binds Ephrin-B2 and not other Ephrin B

ligands (19, 20). The B class Ephrins are comprised of transmem-brane proteins with an intracellular domain that can elaboratereverse signaling (21, 22). Ligand-receptor binding leads to proteinclustering followed by receptor activation (21). Furthermore,interaction between EphB4 and Ephrin-B2 can provoke bidirec-tional signaling (23, 24). It has been shown in vitro that EphB4-mediated forward signaling restricts intermingling of cells andsupports cellular segregation, whereas reverse signaling fromEphrin-B2 stimulates migration and sprouting angiogenesis(25, 26). Downstream signaling by EphBs regulates cell prolifera-tion, cytoskeletal organization, and migration (27–29).We focused on EphB4 for its potential role in tumor biology

because of its location on chromosome 7q22 , a locus frequentlyamplified in cancers including prostate cancer, and because of itsexpression in various cancers (30–34). However, little is knownabout the regulation and role of EphB4 in tumor cell biology. Inthis report, we show that EphB4 is expressed in prostate tumorcells, and its induction may occur either through gene amplifica-tion or the regulatory effect of prosurvival signaling molecules.Furthermore, inhibition of EphB4 expression inhibits tumor cellsurvival, migration, and invasion. These findings indicate thatEphB4 is a novel target for diagnostic and therapeutic applicationin prostate cancer.

Materials and Methods

Reagents. Neutralizing IGF-IR antibody was obtained from R&DSystems, Inc. (Minneapolis MN), anti-IGF-IR, -EGFR, -EphB4 (C-16),

and CD31 (M20) from Santa Cruz Biotech (Santa Cruz, CA), anti-

phosphotyrosine (4G10) from Upstate, Inc., (Chicago, IL), anti-Ki-67 from

Requests for reprints: Parkash S. Gill, USC/Norris Comprehensive Cancer Center,1441 Eastlake Avenue, NOR 6332, Los Angeles, CA 90033-9172. Phone: 323-865-3909;Fax: 323-865-0092; E-mail: parkashg@usc.edu.

I2005 American Association for Cancer Research.

www.aacrjournals.org 4623 Cancer Res 2005; 65: (11). June 1, 2005

Research Article

DAKO (Carpinteria, CA), anti-human Fc from Jackson Laboratory (BarHarbor, ME), anti–glyceraldehyde-3-phosphate dehydrogenase (GAPDH)

from Research Diagnostics, Inc. (Flanders, NJ), and anti-actin monoclonal

antibody from Sigma Chemical Co. (St. Louis, MO). Media and fetal bovine

serum (FBS) were obtained from Invitrogen (Carlsbad, CA). AG 1478 [4-(3V-Chloroanilino)-6,7-dimethoxy-quinazoline] was from Calbiochem (San

Diego, CA). Ephrin-B2/Fc chimeric protein was from R&D Systems, Inc.,

and human IgG Fc fragment from Jackson Laboratory.

Gene expression microarray studies and analysis. Fresh-frozenprostate tumor tissues were harvested from 72 patients treated with radical

prostatectomy for clinically localized prostate cancer at St. Vincent’s

Hospital, Sydney between 1996 and 2000 as previously described (35, 36).

Total RNA was extracted with Trizol reagent (Life Technologies, Rockville,MD) and was reverse-transcribed using a primer containing oligo (dT) and a

T7 promoter sequence. The resulting cDNAs were then in vitro–transcribed

in the presence of biotinylated nucleotides (Bio-11-CTP and Bio-16-UTP)using the T7 MEGAscript kit (Ambion, Austin, TX). The biotinylated targets

were hybridized to the Eos Hu03, a customized oligonucleotide array

comprising >59,000 oligonucleotide probe-sets for the interrogation of a total

of 46,000 gene clusters that were based on the first draft of the humangenome. Hybridization signals were visualized using phycoerythrin-conju-

gated streptavidin (Molecular Probes, Eugene, OR). EphB1, EphB2, and

EphB4–specific gene expression data was examined in all cases and relative

signal intensity was assessed for each of these three receptors.EphB4 short interfering RNAs and antisense oligodeoxynucleotides.

Phosphorothioate-modified EphB4-specific antisense and sense oligodeox-

ynucleotides used were: sense, TCC TGC AAG GAG ACC TTC AC (5V!3V);antisense-1, GTG CAG GGA TAG CAG GGC CAT (5V!3V); antisense-10, ATGGAG GCC TCG CTC AGA AA (5V!3V), antisense-9, CTT CAG GGT CTT GAT

TGC CAC (5V-3V), which were synthesized and purified using Qiagen

reagents (Valencia, CA). Short interfering RNA (siRNA) used were greenfluorescent protein (GFP)-siRNA, 5V-CGCUGACCCUGAAGUUCATUU-3V;EphB4-siRNA (50), 5V-GAGACCCUGCUGAACACAAUU-3V; EphB4-siRNA

(472), 5V-GGUGAAUGUCAAGACGCUGUU-3V, EphB4-siRNA(2303)

5VCUCUUCCGAUCCCACCUACUU-3Vwhich were synthesized at the USC/Norris Comprehensive Cancer Center Microchemical Core Laboratory.

Cell lines and culture. The prostate cancer cell lines, PC3, PC3M,

DU145, ALVA31, LAPC-4, LNCaP, CWR22R, and BPH-1 were obtained andcultured as described previously (7). Kaposi sarcoma cell line (SLK) was

obtained from American Type Culture Collection (Rockville, MD).

Generation of EphB4 monoclonal antibody. The extracellular domain

of EphB4 was cloned into pGEX-4T-1 to generate glutathione S-transferase–fused extracellular domain. EphB4-extracellular domain expressed as a

glutathione S-transferase fusion protein in BL21 E. coli was purified by

affinity chromatography and the glutathione S-transferase domain was

cleaved by thrombin. A monoclonal antibody was generated and thesensitivity and specificity of the antibody was reconfirmed by Western blot

with whole cell lysate of 293 cells stably transfected with EphB4. This

monoclonal antibody was used for the Western blot and immunoprecip-

itation in this study.One-Step RT-PCR and quantitative PCR. Total RNA was extracted

using RNA STAT-60 (Tel-Test, Inc., Friendswood, TX) from prostate cancer

specimens and adjacent normal specimens. For quantitative RT-PCR first-strand cDNA was synthesized from 5 Ag of total RNA using SuperScript III

(Invitrogen). Quantitative RT-PCR was done on the Stratagene MX3000P

system (Stratagene, La Jolla, CA) using SYBR Green I Brilliant Mastermix

according to the manufacturer’s instructions. Optimized reactions forEphB4 and h-actin (used as the normalizer numerator molecule for

comparative gene expression) were 150 nmol/L each of the forward primers

(h-actin, 5V-GGA-CCT-GAC-TGA-CTA-CCT-A-3V; EphB4, 5V-AAG-GAG-ACC-TTC-ACC-GTC-TT-3V) and reverse primers (h-actin, 5V-TTG-AAG-GTA-GTT-TCG-TGG-AT-3V; EphB4, 5V-TCG-AGT-CAG-GTT-CAC-AGT-CA-3V) with ther-

mal profile of 95jC for 10 minutes followed by 40 cycles of 95jC for

30 seconds, 60jC for 1 minute, and 72jC for 1 minute. The specificity of theamplification was confirmed by the presence of a single dissociation peak.

All reactions were done in triplicate and with no reverse transcriptase or no

template as negative controls.

Gene amplification was analyzed by quantitative PCR. The primerswere EphB4-forward, 5V-TCC TGC AAG GAG ACC TTC AC-3V; EphB4-

reverse, 5V-CAG AGG CCT CGC AAC TAC AT-3V; GAPDH-forward, 5V-GAGGGG TGA TGT GGG GAG TA-3V; and GAPDH-reverse, 5V-GAG CTT CCC

GTT CAG CTC AG-3V. DNA was extracted from peripheral bloodmononuclear cells of normal donors and the prostate cancer cell lines

using the Blood and Cell Culture DNA Midi kit from Qiagen according to

the manufacturer’s instructions. Quantitative PCR was done on 50 ng of

DNA at the annealing temperature of 64jC for both primers. Amplificationsignal for EphB4 was normalized to GAPDH and gene copy number was

normalized to healthy donor peripheral blood mononuclear cells.

Immunohistochemistry. Freezing solution (optimum cutting temper-

ature [OCT]) embedded tissues were sectioned at 5 Am and fixed inphosphate-buffered 4% paraformaldehyde. Sections were washed thrice for

5 minutes in PBS and endogenous peroxidase was blocked by incubation

in 0.3% H2O2 in PBS for 10 minutes at room temperature. Sections wereincubated with Eph4 (C-16) antibody (1:50) for 1 hour at room

temperature, followed by three washes in PBS and incubation with

donkey anti-goat secondary antibody (Santa Cruz Biotechnology) for 1

hour at room temperature. After three washes in PBS, peroxidase activitywas localized by incubation in 3,3V-diaminobenzidine substrate solution

(Vector Laboratories, Inc., Burlingame, CA) for 10 minutes at room

temperature. Sections were counterstained with hematoxylin for 20

seconds, dehydrated and mounted. The negative control for staining wassubstitution of normal IgG isotope for primary antibody. Immunohisto-

chemical staining on the formalin-fixed prostate tissues microarray

(BioMedia, Foster City, CA) and animal tumors was done using avidin-biotin complex method staining system (Santa Cruz Biotechnology)

according to the manufacturer’s instructions. The antibodies used were

EphB4 (C-16, Santa Cruz Biotechnology; 1:100 dilution), CD31 (M20, Santa

Cruz Biotechnology; 1:200 dilution) and Ki-67 (DAKO; 1:100 dilution).Western blot. Cell lysates were prepared using cell lysis buffer

(GeneHunter, Basgvukke, TN) supplemented with protease inhibitor

cocktail (Pierce, Rockford, IL), unless otherwise noted. Total protein was

determined using the DC reagent system (Bio-Rad, Hercules, CA). Typically,20 Ag whole cell lysate was run on 4% to 20% Tris-glycine gradient gel. The

samples were electrotransferred to polyvinylidene difluoride membrane and

nonspecific binding was blocked in TBST buffer [0.5 mmol/L Tris-HCl,45 mmol/L NaCl, 0.05% Tween 20 (pH 7.4)] containing 5% nonfat milk.

Membranes were first probed with primary antibody overnight, washed and

probed with the secondary antibody and developed. The membranes were

then stripped with Restore Western Blot Stripping Buffer (Pierce) andreprobed with h-actin to confirm equivalent loading and transfer of protein.

Signal was detected using SuperSignal West Femto Maximum Sensitivity

Substrate (Pierce).

Phosphorylation analysis. Cells growing in 60 mm dishes were eitherserum-starved (1% FBS supplemented RPMI 1640 for 24 hours) or cultured

in normal conditions (10% FBS) and then treated with or without clustered

1 Ag/mL mouse Ephrin-B2/Fc for 10 minutes to activate EphB4 receptor.

Cleared cell lysates were incubated with EphB4 monoclonal antibodyovernight at 4jC. Antigen-antibody complexes were immunoprecipitated by

the addition of 20 AL of protein G-Sepharose in 20 mmol/L sodium

phosphate (pH 7.0) with incubation overnight at 4jC. Immunoprecipitateswere analyzed by Western blot with phosphotyrosine-specific antibody at

1:1,000 dilution. To monitor immunoprecipitation efficiency, a duplicate

membrane was probed for EphB4.

Transient transfection and sorting of transfected cells. PC3 cellswere cotransfected with pMACS 4.1 coding for a truncated CD4 and wild-

type p53 (pC53-SN3) or PTEN vector or both using LipofectAMINE 2000

(Invitrogen) according to the manufacturer’s instructions. The molar ratio

of CD4 to p53 or PTEN or vector was 1:3 and total plasmid was 24 Ag fora 10 cm2 dish of 90% confluent cells using 60 AL of LipofectAMINE

2000. Twenty-four hours after transfection, a single cell suspension was

made and sorted using truncated CD4 as a surface marker according tothe manufacturer’s protocol (Miltenyi Biotec, Germany). Equal numbers

of sorted cells were lysed in 1� SDS sampling buffer and analyzed by

Western blot.

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4624 www.aacrjournals.org

Study of insulin-like growth factor and epidermal growth factorsignaling pathway on the expression of EphB4. PC3 cells were seeded

onto six-well plates and cultured until 80% confluent and treated with

2 Ag/mL neutralizing IGF-IR monoclonal antibody, MAB391 (37), or with

1 Amol/L AG 1478, a strong EGFR-kinase inhibitor (38) for 24 hours. Crudecell lysates were analyzed by Western blot to detect EphB4 and h-actin as

described above. Band density was quantified with the Bio-Rad QuantityOne

System software.

Cell viability assay. PC3 cells were seeded on 48-well plates at a densityof 1 � 104 cells per well in a total volume of 200 AL. Medium was changed

after the cells were attached and triplicate samples of cells were treated

with various concentrations (1-10 Amol/L) of EphB4 antisense or sense

oligodeoxynucleotides as control. After 3 days, medium was changed andfresh oligodeoxynucleotides were added. Following a further 48 hours

incubation, viable cells were measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide as described previously (39). EphB4 siRNAs(10-100 nmol/L) or control siRNAs were introduced into triplicate samples

of 1 � 104 PC3 cells per well of a 48-well plate using 2 AL of LipofectAMINE

2000 according to the manufacturer’s instructions. Four hours post-

transfection, the cells were returned to growth medium (RPMI1640 supplemented with 10% FBS). Viable cells were measured by 3-(4,5-

dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide 48 hours following

transfection. The SLK cell line, which lacks EphB4 expression, was used as a

negative control.Wound healing migration assay. PC3 cells were seeded onto six-well

plates and cultured until confluent. Antisense-10 or sense oligodeoxynu-

cleotides (10 Amol/L) as control were introduced to the wells as describedfor the viability assay 12 hours before wounding the monolayer by scraping

it with a sterile pipette tip. Medium was changed to RPMI 1640

supplemented with 5% FBS and fresh oligodeoxynucleotides. Confluent

cultures transfected with 50 nmol/L EphB4 siRNA 472 or GFP siRNA 12hours prior to wounding were also examined. The healing process was

examined dynamically and recorded with a Nikon Coolpix 5000 digital

camera with microscope adapter.

Invasion assay. PC3 cells were transfected with EphB4 siRNA 472 orGFP siRNA using LipofectAMINE 2000 and 6 hours later 0.5 � 105 cells were

transferred into 8 Am Matrigel-precoated inserts (BD Bioscience, Palo Alto,

CA). The inserts were placed in companion wells containing RPMI

supplemented with 5% FBS and 5 Ag/mL fibronectin as a chemoattractant.Following 22 hours of incubation, the inserts were removed and the

noninvading cells on the upper surface were removed with a cotton swab.

The cells on the lower surface of the membrane were fixed in 100%

methanol for 15 minutes, air-dried, and stained with Giemsa stain for2 minutes. The cells were counted in five individual high-powered fields for

each membrane under a light microscope. Assays were done in triplicate for

each treatment group.Cell cycle analysis. PC3 cells (80% confluence) in six-well plates were

transfected with EphB4 siRNA 472 (100 nmol/L) using LipofectAMINE 2000.

Twenty-four hours after transfection, cells were trypsinized, washed in PBS

and incubated for 1 hour at 4jC in 1 mL of hypotonic solution containing50 Ag/mL propidium iodide, 0.1% sodium citrate, 0.1 Triton X-100, and

20 Ag/mL DNase-free RNaseA. Cells were analyzed in linear mode at the

USC Flow Cytometry Facility. Results were expressed as percentages of

elements detected in the different phases of the cell cycle, namely sub-G0

peak (apoptosis), G0/G1 (no DNA synthesis), S (active DNA synthesis), G2

(premitosis), and M (mitosis).

Apoptosis ELISA. Apoptosis was studied using the Cell Death DetectionELISAplus Kit (Roche, Piscataway, NJ) according to the manufacturer’s

instructions. Briefly, PC3 cells in 24-well plates cultured to 80% confluence

were transfected using LipofectAMINE 2000 with various concentrations

Figure 1. Expression of EphB4 in prostate cell lines. A , Western blot of total cell lysates of various prostate cancer cell lines, benign prostate hyperplasia (BPH)epithelial cell line, and SLK cells probed for EphB4. The membrane was stripped and reprobed for h-actin to show equivalent loading and transfer of proteins. B, geneamplification in prostate cancer cell lines by quantitative PCR as described in Materials and Methods. Columns, mean; bars, F SD of EphB4 copy number in duplicatesamples relative to peripheral blood mononuclear cells. The scale was set at 2 for the diploid status of normal peripheral blood mononuclear cells. C, phosphorylation ofEphB4 in PC-3 cells determined by Western blot. PC-3 cells cultured under normal conditions (RPMI 1640 supplemented with 10% FBS) or serum-starved for 24 hourswere exposed to 1 Ag/mL Ephrin-B2 for 10 minutes. Cell lysates were immunoprecipitated with EphB4 monoclonal antibody and analyzed by Western blot forphosphotyrosine. A duplicate blot was probed for total EphB4. D, PC3 cells were treated with varying concentrations of Ephrin-B2/Fc or Fc alone for the time periodsshown. Total EphB4 and h-actin was measured by Western blot (left). EphB4 phosphorylation status was assessed by immunoprecipitation of EphB4 from 50 Agof total cell lysate and Western blot for pTyr (top right ) and EphB4 mRNA level was quantitated by real-time RT-PCR (bottom right ).

Expression of EphB4 in Prostate Cancer

www.aacrjournals.org 4625 Cancer Res 2005; 65: (11). June 1, 2005

(0-100 nmol/L) of EphB4 siRNA 472 or GFP siRNA. Sixteen hours later,cells were detached and 1 � 104 cells were incubated in 200 AL lysis

buffer. Nuclei were pelleted by centrifugation and 20 AL of supernatant

containing the mono- or oligonucleosomes was taken for ELISA analysis.

Briefly, the supernatant was incubated with anti-histone-biotin and anti-DNA-POD in streptavidin-coated 96-well plates for 2 hours at room

temperature. The color was developed with ABST and absorbance at 405

nm was read in a microplate reader (Molecular Devices, Sunnyvale, CA).

Apoptosis was detected in deparaffinized sections of tumor tissue frommurine xenograft studies using the in situ cell death detection kit (Roche)

according to manufacturer’s instructions.

In vivo tumor growth studies. Balb/C nu/nu mice (male, 9 weeks old)

were implanted with tumor cells propagated in vitro . Orthotopic xenograftswere done under ketamine anesthesia. A lower abdominal midline incision

was made. The posterior face of prostatic lobe was exposed and PC3 cell

(1 � 105) in 20 AL PBS was injected under the capsule of prostate. Theabdominal wound was closed in two layers with 4-0 surgical sutures. Mice

bearing the xenografts were randomly divided into three groups (six mice

per group). Oligo EphB4 antisense-10 or sense oligodeoxynucleotide

dissolved in sterile physiologic saline (0.9% NaCl) was given i.p. at a doseof 20 mg/kg, starting 10 days after implantation, daily for 2 weeks. The mice

were monitored by general clinical observation as well as by body weight.

The prostate glands were harvested at the completion of the study. Whole

tissue was sectioned and stained by H&E. The section, which has themaximum longitudinal diameter of the tumor, was selected. The volume

was calculated as 0.52 � length � width2. In vivo study of s.c. xenograft

using PC3 cells was carried out as previously described (39) with themodification that mice were treated with EPHB4 antisense-10 or sense

oligodeoxynucleotide dissolved in sterile physiologic saline (0.9% NaCl)given i.p at a dose of 20 mg/kg, daily for 3 weeks (n = 6).

Statistical analysis. The significance of EphB4 expression in normal and

prostate cancer tissues was analyzed by v2 test. Tumor volumes in vivo and

number of CD31-positive vessels and Ki-67 or terminal nucleotidyltransferase–mediated nick end labeling positive cells were compared with

Students’ t test and P < 0.05 was accepted for significance.

Results

Expression of EphB4 in prostate cancer. We first examinedthe expression of EphB4 protein in a variety of prostate cancer celllines by Western blot. Prostate cancer cell lines uniformlyexpressed EphB4 but showed a marked variation in expressionprotein level. The levels were relatively high in PC3 and even higherin PC3M, a metastatic clone of PC3, whereas benign prostateepithelial cells and SLK cells showed no expression (Fig. 1A). Usingquantitative PCR, we found that the gene copy number of EphB4was amplified in PC3M when compared with peripheral bloodmononuclear cells or PC3 cells (Fig. 1B). We next examined ifEphB4 on tumor cells is functional by testing whether EphB4 isphosphorylated in response to ligand (Ephrin-B2). A low-level basalsignal was detected in cells grown in serum-containing conditions,and this signal markedly increased following optimal time exposureto 10 Ag/mL ligand, Ephrin-B2/Fc (Fig. 1C). Serum-deprived PC3cells showed no basal phosphorylation of EphB4 (Fig. 1C). Thus,

Figure 2. Expression of EphB4 in prostate tumor specimens. A, immunohistochemical staining on clinical prostate cancer samples: a, b, e, and g, prostatecancer showing EphB4 in tumor cells; c, isotype control IgG was used in place of primary antibody to probe a serial section of (a); d, lack of EphB4 staining in benignprostate hyperplasia; f, high-power view of (e) showing membrane staining of EphB4; h, EphB4 signal is blocked when a serial section of (g) is probed with primaryantibody premixed with antigen. B, summary of EphB4 staining in formalin-fixed prostate tissue microarray consisting of 62 prostate cancers and 20 normal prostatetissues. C, gene expression for EphB1, EphB2, and EphB4 by DNA microarray in 72 prostate cancer tissues. EphB4 was expressed in 64 (89%) cases.

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4626 www.aacrjournals.org

EphB4 in PC3 cells is functional. We next determined the fate ofEphB4 following treatment with Ephrin-B2/Fc for prolongedperiods. EphB4 levels began to decline following 4 hours ofstimulation with 3 or 10 Ag/mL Ephrin-B2/Fc, whereas 1 Ag/mL ofEphrin-B2/Fc or Fc (10 Ag/mL) had no effect (Fig. 1D , left). Theseresults are expected due to the endocytosis of receptor-ligandcomplex and degradation (50, 51). Concomitant to the decrease inEphB4 level, the extent of receptor phosphorylation also declined.Whereas robust phosphorylation was observed following stimula-tion with 10 Ag/mL Ephrin-B2/Fc for 15 minutes, there was nodetectable phosphorylation after 8 or 24 hours of stimulation(Fig. 1D , top right). EphB4 mRNA levels were quantitated by real-time RT-PCR in PC3 cells treated simultaneously with Ephrin-B2/Fc (10 Ag/mL) for 15 minutes, 8 hours, and 24 hours. No significantchange was observed in mRNA levels (Fig. 1D , bottom right). Inparallel experiments, cell viability was studied in PC3 cells treatedwith Ephrin-B2/Fc at 10 Ag/mL. A 45% decline in cell viability wasobserved after 24 hours, whereas Fc alone at the same dose had no

effect (data not shown). Decline in EphB4 was thus associated withreduction in the measure of viable cells.To determine whether EphB4 is expressed in clinical prostate

samples, tumor tissues and adjacent normal tissue from prostatecancer surgical specimens were examined. The histologic distribu-tion of EphB4 in the prostate specimens was determined byimmunohistochemistry. EphB4 protein expression is confined tothe neoplastic epithelium. Both membrane and cytoplasmicstaining were seen (Fig. 2A , a, b, e, f, and g ), whereas the expres-sion is absent in normal glands (Fig. 2A , d). Antibody specificitywas shown by loss of signal when the antibody was premixed withantigen/peptide (Fig. 2A , g and h). In a prostate tissue microarraygenerated from archival tissue, 41 of 62 (66%) prostate cancersexamined were positive, whereas 3 of 20 (15%) normal prostatetissues showed expression at low intensity (P < 0.01; Fig. 2B).We also examined gene expression in 72 prostate cancer patients

using DNA microarrays. EphB4 was expressed in the majority of thecases (64 of 72 or 89% with normalized expression above 50 units)

Figure 3. Regulation of EphB4 in prostatecancer cells by tumor suppressors andgrowth factors. A, PC3 cells werecotransfected with truncated-CD4 and p53or PTEN or empty vector. Twenty-fourhours later, transfected (CD4-sorted) cellswere analyzed by Western blot for theexpression of EphB4 and h-actin. B,Western blot of PC3 cells treated with orwithout EGFR-specific inhibitor AG1478(1 Amol/L) for 24 hours (left ). Western blotof PC3 cells treated with or withoutIGF-IR-specific neutralizing antibodyMAB391 (2 Ag/mL, 24 hours; right). Therelative EphB4 signal normalized forh-actin is shown as a percentage of control(no treatment) beneath the Western blot.C, PC3 cells were treated with variousoligodeoxynucleotides (10 Amol/L) andsiRNAs (100 nmol/L). Antisense-10 andsiRNA 472 showed the most prominentand specific repression of EphB4, whereasEGFR, GAPDH and h-actin wereunchanged. D, measure of viable cellsafter treatment with siRNA (top ) oroligodeoxynucleotides (bottom ) asdescribed in Materials and Methods. PC3and SLK cells were plated with equaldensity and treated as shown. Viablecells were measured 48 hours aftersiRNA transfection or 5 days ofoligodeoxynucleotide treatment. Valueswere normalized to the LipofectAMINE-only (0 nmol/L of siRNAs) or no treatmentgroups (0 Amol/L oligodeoxynucleotide).Columns, mean; bars, F SEM. In contrastwith PC3 cells (left), siRNAs and antisenseoligodeoxynucleotides have no effect onthe EphB4 negative cell line SLK (right ).

Expression of EphB4 in Prostate Cancer

www.aacrjournals.org 4627 Cancer Res 2005; 65: (11). June 1, 2005

with signal intensity higher than that seen for EphB1 and EphB2(Fig. 2C).Regulation of EphB4 expression by tumor suppressor genes

and growth factors. PC3 cells are known to lack wild-type p53function (8) and PTEN expression (9). We investigated whether therelatively high expression of EphB4 is related to p53 and/or PTENby reintroducing wild-type p53 and/or PTEN into PC3 cells. Cellswere cotransfected with a truncated-CD4 expression plasmid toallow selection of transfected cells. Transfected cells were thensorted for the cotransfected truncated CD4 marker. We found thatthe expression of EphB4 in PC3 cells was reduced >75% by thereintroduction of either wild-type p53 or PTEN (Fig. 3A). Thecotransfection of both p53 and PTEN did not further inhibitthe expression of EphB4. EGFR and IGF-IR have both been shownto promote PC3 cell growth and survival. We postulated thatEphB4 expression might correlate with these prosurvival growthfactors. We tested the relationship by independently blockingEGFR and IGF-IR signaling. In the PC3 cell line, EphB4 was down-regulated by >50% after blocking EGFR signaling using the EGFRkinase inhibitor AG 1478 (Fig. 3B , left) or upon blockade of theIGF-IR signaling pathway using IGF-IR neutralizing antibody(Fig. 3B , right).EphB4 short interfering RNA and antisense oligodeoxynu-

cleotides inhibit PC3 cell viability. To define the significance ofEphB4 overexpression in the prostate cancer model, we concen-

trated our study on PC3 cells, which have a relatively highexpression of EphB4. The two approaches used for decreasingEphB4 expression were siRNAs and antisense oligodeoxynucleo-tides. A number of different phosphorothioate-modified antisenseoligodeoxynucleotide complementary to different segments of theEphB4 coding region were tested for specificity and efficacy ofEphB4 inhibition. Antisense-10 was found to be a specific andpotent inhibitor of EphB4 expression (Fig. 3C). A similar approachwas applied to the selection of specific siRNA. EphB4 siRNA 472effectively knocks down EphB4 protein expression, whereas EphB4siRNA 2,303 was inactive (Fig. 3C). Another EphB4-specific siRNA50 was also highly effective in knocking down protein expression(data not shown). Both EphB4 siRNAs (472 and 50) and antisense-10 oligodeoxynucleotide reduced the number of viable PC3 cells (byas much as 80%) in a dose-dependent manner (Fig. 3D , top andbottom left). Unrelated (GFP-siRNA) or inactive (2,303) siRNAs orsense oligodeoxynucleotide had no effect on the number of viablecells. We found that neither siRNA 472 nor antisense-10 had aninhibitory effect on viable cell number in the EphB4-negative cellline, SLK (Fig. 3D , top right and bottom).EphB4 short interfering RNA and antisense oligodeoxynu-

cleotides inhibit prostate tumor cell migration and invasion.PC3 cells can grow locally and form lymph node metastases wheninjected orthotopically into mice. In an effort to study the role ofEphB4 on migration of PC3 cells in vitro , we did a wound-healing

Figure 4. EphB4 regulates prostate tumor cell migration and invasion. A, PC3 cells were grown to confluence and 50 nmol/L EphB4 siRNA 472 or GFP siRNAtransfected with LipofectAMINE. The cell layer was scraped and photographed immediately or after 20 hours. Photomicrographs of representative 20� fields. B, asimilar assay for cells treated with antisense-10 or sense oligodeoxynucleotide (both 10 Amol/L). C, Matrigel invasion assay. Invasion of PC3 cells into Matrigel wasstudied in a modified Boyden chamber assay. PC3 cells were transfected with 50 nmol/L EphB4 siRNA 472 or GFP siRNA. Cells migrating to the underside of theMatrigel-coated insert in response to 5 mg/mL fibronectin in the lower chamber were fixed and stained with Giemsa. Cell numbers were counted in five individualhigh-powered fields. Columns, mean; bars, F SEM.

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4628 www.aacrjournals.org

assay. When a wound was introduced into a monolayer of PC3cells, over the course of the next 20 hours, cells progressivelymigrated into the cleared area. However, when cells were trans-fected with 50 nmol/L EphB4 siRNA 472 and the wound wasintroduced, this migration was significantly inhibited whencompared with control (GFP-siRNA; Fig. 4A). Pretreatment ofPC3 cells with 10 Amol/L EphB4 antisense-10 for 12 hours alsogenerated inhibition of cell migration when compared with a senseoligodeoxynucleotide (Fig. 4B). In addition, knockdown of EphB4expression in PC3 cells with 50 nmol/L EphB4 siRNA 472 severelyreduced the ability of these cells to invade Matrigel as assessed by adouble-chamber invasion assay (Fig. 4C), whereas control GFP-siRNA showed little or no effect.EphB4 short interfering RNA induces cell cycle arrest and

apoptosis in PC3 cells. Because knockdown of EphB4 resulted in areduced measure of viable cells, we sought to determine whetherthis was due to effects on the cell cycle. In comparison with GFP-siRNA–transfected cells, 50 nmol/L EphB4 siRNA 472 resulted in anaccumulation of cells in the sub-G0 and S phase fractions. The sub-G0 fraction increased from 1% to 7.9% in EphB4 siRNA 472–treatedcells compared with GFP-siRNA–treated cells (Fig. 5A). Apoptosis asa result of EphB4 knockdown was confirmed by measuring levels ofcytoplasmic nucleosomes (a marker of DNA fragmentation) and theeffect was dose-dependent when PC3 cells were transfected withEphB4 siRNA 472, but not GFP siRNA. DNA fragmentation in EphB4siRNA 472 transfected PC3 cells increased 15-fold at a dose of 100nmol/L when compared with GFP-siRNA treated cells (Fig. 5B).In vivo activity of EphB4 antisense oligodeoxynucleotide in a

prostate cancer xenograft model. Following implantation of PC3cells in their prostate glands, male athymic mice were eithertreated with EphB4 antisense-10, sense oligodeoxynucleotide, ordiluent (PBS). Sense oligodeoxynucleotide- or diluent-treated miceshowed large tumors in each of the implanted prostate glands,whereas those treated with antisense-10 oligodeoxynucleotides hadsmall tumors in only two of the six prostate glands (Fig. 6A).Furthermore, antisense-10 reduced cell proliferative index asmeasured by Ki-67 staining, increased apoptosis and reducedmicrovascular density in the tumor (Fig. 6B). The in vivo study wasalso conducted with a s.c. xenograft model using PC3 cells. Tumorvolumes in antisense-10-treated mice were significantly smallerthan those in control groups (sense oligodeoxynucleotide- or PBS-treated; P < 0.01). Western blot of extracts of tumor tissue showedthat the EphB4 expression is inhibited in the antisense-10-treatedgroup (Fig. 6C). There was no evidence of systemic toxicity to themice as evidenced by normal food intake and body weight.Furthermore, histologic evaluation of heart, lung, liver, and kidneydid not reveal any evidence of toxicity (data not shown).

Discussion

EphB4 protein is expressed in a majority of prostate carcinomaswith expression limited predominantly to the neoplastic epithelium.This is in contrast with little or no expression in normal prostategland. To our knowledge, this is the first documentation of EphB4expression in prostate cancer. Previous studies have reportedoverexpression of EphB4 in colon cancer and in a variety of humanlung and breast cancer cell lines (30, 31, 40). However, these studiesprovided limited functional significance of EphB4 expression incancer. Our survey of prostate cell lines first indicates an associationbetween the level of EphB4 expression and aggressive growth. Thisis suggested by the fact that cells derived from benign prostatic

hyperplasia has no expression of EphB4 compared with the PC3prostate cancer metastasis-derived cell line, and a clone of PC3 withhigher metastatic potential (PC3M), which expressed the highestamounts of EphB4. This trend corroborates the observations inendometrial carcinoma and breast cancer where overexpression ofEphB4 has a suggested potential role in cancer promotion (31–33,41), although a contradictory report exists (42). We hypothesize thatthere is a direct correlation of EphB4 expression with higher gradeand metastatic potential that needs to be further explored insubsequent studies.One possible mechanism of increased EphB4 expression may be

gene amplification as we have observed in PC3M cells. Amplifica-tion of chromosome 7q is a common occurrence in androgen-independent prostate carcinomas (43), and the EphB4 locus is at7q22 . We next sought to investigate the functional role and

Figure 5. Effect of EphB4 siRNA 472 on cell cycle and apoptosis. A, PC3 cellstransfected with 50 nmol/L GFP siRNA or EphB4 siRNA 472 as indicated andwere analyzed 24 hours posttransfection for cell cycle status by flow cytometryas described in Materials and Methods. Plots of cell number versus propidiumiodide fluorescence intensity are shown: 7.9% of the cell population is apoptotic(in the sub-G0 peak) when treated with EphB4 siRNA 472 compared with 1% withGFP-siRNA. B, apoptosis of PC3 cells detected by Cell Death DetectionELISAplus kit as described in Materials and Methods. Absorbance at 405 nmincreases in proportion to the amount of histone and DNA-POD in the nuclei-freecell fraction. Columns, mean; bars, F SEM of triplicate samples at the indicatedconcentrations of EphB4 siRNA 472 and GFP-siRNA.

Expression of EphB4 in Prostate Cancer

www.aacrjournals.org 4629 Cancer Res 2005; 65: (11). June 1, 2005

mechanism of EphB4 expression in prostate cancer using the PC3cell line as a model. The EphB4 pathway in PC3 cells is functionalas shown by markedly increased phosphorylation of EphB4following stimulation by Ephrin B2 above the basal level whencells were grown in serum. Loss-of-function mutations in tumorsuppressor genes or induction of growth factor pathwaysparticipate in prostate cancer development or progression (4–10).It is thus not surprising that introduction of the native forms ofPTEN and p53 reduced EphB4 levels in PC3 cells, which lack PTENexpression and wild-type p53 function (8, 9, 44). Notably,concurrent introduction of both PTEN and p53 did not result inadditive down-regulation of EphB4, suggesting overlapping path-ways. In light of this regulation by tumor-suppressor genes, it isinteresting to note that EphB4 promotes oncogenesis andmetastasis initiated by the neuT oncogene. Mammary tumorlatency is reduced by 50% in EphB4/neuT double transgenic micecompared with neuT mice, which develop spontaneous mammarycarcinomas (45). In addition, these tumors produce metastasis,whereas tumors in neuT mice are localized. EphB4 expression wasalso reduced by inhibition of both the EGFR and IGF-IR pathways,which are implicated in prostate cancer and may contribute toescape from androgen dependence (4–6). Together, these resultssuggest that EphB4 can be up-regulated by a variety ofmechanisms. EphB4 induction in prostate cancer may be complex,as we have shown that EphB4 is regulated by oncogenes (EGFR andIGF-IR) and tumor suppressor genes (PTEN and p53). Other

possible mechanisms of regulation include disruption of Notch andactivation of hoxA9 signaling (46–48).Targeted disruption of EphB4 expression in prostate cancer cells

proved to be highly informative. We showed that EphB4 is requiredfor tumor cell viability, migration, invasion, and cell cycleprogression. In endothelial cells, where EphB4 is normallyexpressed, it regulates cell migration, proliferation, and invasion(29). However, in contrast with endothelial cells, EphB4 expressionin prostate cancer cell lines provides a critical survival signal. Here,we showed that EphB4 knockdown increased the fraction of cells insub-G0 fraction by cell cycle analysis. In addition, a dose-dependentincrease in cytoplasmic nucleosomes (a hallmark of apoptosis) wasobserved after ablation of EphB4 expression in PC3 cells. Thus,reduction in EphB4 levels leads to apoptosis. In a recent report,Noren et al. (49) found that activation of EphB4 with clusteredEphrin-B2/Fc inhibited tumor proliferation and concluded thatEphB4 activation inhibits tumor growth. This seemingly contra-dicts our results. However, we found that prolonged incubationwith Ephrin-B2/Fc reduced levels of total and phosphorylatedEphB4 without any change in mRNA levels. This may occur byligand-induced receptor endocytosis and degradation (50, 51).Thus, Noren et al., over a period of 5 days, see a decrease in totalcell number, likely due to the loss of EphB4.In vitro studies were confirmed in vivo using both orthotopic and

s.c. xenograft prostate tumor models. Inhibition of EphB4 geneexpression with antisense-10 oligodeoxynucleotides but not sense

Figure 6. In vivo effect of EphB4 antisense-10 on prostate tumors. A, mice implanted with PC3 cells in the prostate gland were treated with PBS, senseoligodeoxynucleotide, or EphB4 antisense-10 oligodeoxynucleotides. Treatment was begun 10 days after implantation and continued daily throughout the study.Prostate glands were harvested and sectioned for H&E staining at the conclusion of the study. Tumor volume was measured as described in Materials and Methods.Pictures were taken under 10� magnification. Black arrow, tumor; yellow arrow, normal prostate glands; bar, mean volume. B, immunohistochemical staining ontumor-bearing glands for microvascular density (CD31), cell proliferation index (Ki-67), and apoptosis (TUNEL) in tumor (black arrow) and normal mouse prostate(yellow arrow ). Median values of positive signal counted in five random high power fields, and P values comparing control and antisense-10 treated tumors are shown.C, athymic mice bearing s.c. PC3 xenografts were treated with PBS (control), sense or EphB4 antisense-10 oligodeoxynucleotide for 21 days. Tumors were harvestedand lysates analyzed by Western blotting for EphB4, EGFR, and h-actin.

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4630 www.aacrjournals.org

oligodeoxynucleotides showed marked reduction in tumor growth.This reduction corresponds to a down-regulation of EphB4. Wefound a reduction in intratumoral microvessel density andpostulate that down-regulation of EphB4 may also deter reversesignaling of Ephrin-B2, which is involved in angiogenesis. Anantiangiogenic effect in tumor by blocking EphB4/Ephrin-B2signaling has been reported recently (49, 52). To our knowledge,this is the first study that has related the expression of EphB4 tocancer cell growth, survival, migration, and invasion. Inhibition ofEphB4 expression did not cause any evidence of toxicity in adultmice in contrast with embryonic lethality in EphB4 knock-outmice. These results are not unexpected as angiogenesis in adult lifeis limited to sites of remodeling.In summary, EphB4 is a novel marker for prostate cancer that

seems to positively correlate with an invasive and metastaticphenotype. Its expression is actively regulated under the control ofgrowth factors and tumor suppressor genes. EphB4 endows tumorcells with increased migratory and invasive properties. Loss of

EphB4 retards cell cycle progression, reduces cell viability, andinduces apoptosis. EphB4 is highly expressed in a majority ofclinical prostate cancer specimens. Future studies will address therelationship of EphB4 expression with other known genetic lesionsin prostate cancer for diagnostic and prognostic implications.Furthermore, targeting EphB4 has the potential to provide noveltherapeutic opportunities. Blocking EphB4 function could blockangiogenesis as well as directly inhibit survival and invasion ofEphB4-positive prostate cancers.

Acknowledgments

Received 7/26/2004; revised 2/28/2005; accepted 3/24/2005.Grant support: Supported in part by grants from the National Health and Medical

Research Council of Australia, The Cancer Council New South Wales, the RT HallTrust, the Prostate Cancer Foundation of Australia, and the National Cancer Institute,R01 CA79218 (P.S. Gill).

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

References

1. Nomura AM, Kolonel LN. Prostate cancer: a currentperspective. Epidemiol Rev 1991;13:200–27.

2. Brewster SF, Gillatt DA. Advanced prostate cancer:what’s new in hormonal manipulation? Br J Hosp Med1993;49:710–1, 714–5.

3. Griffiths K, Eaton CL, Harper ME, Turkes A, PeelingWB. Hormonal treatment of advanced disease: somenewer aspects. Semin Oncol 1994;21:672–87.

4. Vicentini C, Festuccia C, Gravina GL, Angelucci A,Marronaro A, Bologna M. Prostate cancer cell prolifer-ation is strongly reduced by the epidermal growthfactor receptor tyrosine kinase inhibitor ZD1839 in vitroon human cell lines and primary cultures. J Cancer ResClin Oncol 2003;129:165–74.

5. Nickerson T, Chang F, Lorimer D, Smeekens SP,Sawyers CL, Pollak M. In vivo progression of LAPC-9and LNCaP prostate cancer models to androgenindependence is associated with increased expressionof insulin-like growth factor I (IGF-I) and IGF-I receptor(IGF-IR). Cancer Res 2001;61:6276–80.

6. Hellawell GO, Turner GD, Davies DR, Poulsom R,Brewster SF, Macaulay VM. Expression of the type 1insulin-like growth factor receptor is up-regulated inprimary prostate cancer and commonly persists inmetastatic disease. Cancer Res 2002;62:2942–50.

7. Zhong C, Yang S, Huang J, Cohen MB, Roy-Burman P.Aberration in the expression of the retinoid receptor,RXRa, in prostate cancer. Cancer Biol Ther 2003;2:179–84.

8. Eastham JA, Grafton W, Martin CM, Williams BJ.Suppression of primary tumor growth and the progres-sion to metastasis with p53 adenovirus in humanprostate cancer. J Urol 2000;164:814–9.

9. Li J, Yen C, Liaw D, et al. PTEN, a putative proteintyrosine phosphatase gene mutated in human brain,breast, and prostate cancer. Science 1997;275:1943–7.

10. Quinn DI, Henshall SM, Head DR, et al. Prognosticsignificance of p53 nuclear accumulation in localizedprostate cancer treated with radical prostatectomy.Cancer Res 2000;60:1585–94.

11. Levitzki A. Targeting signal transduction for diseasetherapy. Curr Opin Cell Biol 1996;8:239–44.

12. Pasquale EB. The Eph family of receptors. Curr OpinCell Biol 1997;9:608–19.

13. Gerety SS, Wang HU, Chen ZF, Anderson DJ.Symmetrical mutant phenotypes of the receptor EphB4and its specific transmembrane ligand Ephrin-B2 incardiovascular development. Mol Cell 1999;4:403–13.

14. Tickle C, Altabef M. Epithelial cell movements andinteractions in limb, neural crest and vasculature. CurrOpin Genet Dev 1999;9:455–60.

15. Holder N, Klein R. Eph receptors and ephrins:

effectors of morphogenesis. Development 1999;126:2033–44.

16. Wang HU, Chen ZF, Anderson DJ. Moleculardistinction and angiogenic interaction between embry-onic arteries and veins revealed by ephrin-B2 and itsreceptor Eph-B4. Cell 1998;93:741–753.

17. Adams RH, Wilkinson GA, Weiss C, et al. Roles ofephrin B ligands and EphB receptors in cardiovasculardevelopment: demarcation of arterial/venous domains,vascular morphogenesis, and sprouting angiogenesis.Genes Dev 1999;13:295–306.

18. Shin D, Garcia-Cardena G, Hayashi S, et al. Expres-sion of Ephrin-B2 identifies a stable genetic differencebetween arterial and venous vascular smooth muscle aswell as endothelial cells, and marks subsets of micro-vessels at sites of adult neovascularization. Dev Biol2001;230:139–50.

19. Brambilla R, Schnapp A, Casagranda F, et al.Membrane-bound LERK2 ligand can signal throughthree different Eph-related receptor tyrosine kinases.EMBO J 1995;14:3116–26.

20. Sakano S, Serizawa R, Inada T, et al. Characterizationof a ligand for receptor protein-tyrosine kinase HTKexpressed in immature hematopoietic cells. Oncogene1996;13:813–22.

21. Davis S, Gale NW, Aldrich TH, et al. Ligands for EPH-related receptor tyrosine kinases that require mem-brane attachment or clustering for activity. Science1994;266:816–9.

22. Gale NW, Yancopoulos GD. Ephrins and their recept-ors: a repulsive topic? Cell Tissue Res 1997;290:227–41.

23. Bruckner K, Pasquale EB, Klein R. Tyrosine phos-phorylation of transmembrane ligands for Eph recep-tors. Science 1997;275:1640–3.

24. Holland SJ, Gale NW, Mbamalu G, Yancopoulos GD,Henkemeyer M, Pawson T. Bidirectional signallingthrough the EPH-family receptor Nuk and its trans-membrane ligands. Nature 1996;383:722–5.

25. Fuller T, Korff T, Kilian A, Dandekar G, Augustin HG.Forward EphB4 signaling in endothelial cells controlscellular repulsion and segregation from Ephrin-B2positive cells. J Cell Sci 2003;116:2461–70.

26. Hamada K, Oike Y, Ito Y, et al. Distinct roles of ephrin-B2 forward and EphB4 reverse signaling in endothelialcells. Arterioscler Thromb Vasc Biol 2003;23:190–7.

27. Oates AC, Lackmann M, Power MA, et al. An earlydevelopmental role for eph-ephrin interaction duringvertebrate gastrulation. Mech Dev 1999;83:77–94.

28. O’Leary DD, Wilkinson DG. Eph receptors andephrins in neural development. Curr Opin Neurobiol1999;9:65–73.

29. Steinle JJ, Meininger CJ, Forough R, Wu G, Wu MH,Granger HJ. Eph B4 receptor signaling mediatesendothelial cell migration and proliferation via the

phosphatidylinositol 3-kinase pathway. J Biol Chem2002;277:43830–5.

30. Stephenson SA, Slomka S, Douglas EL, Hewett PJ,Hardingham JE. Receptor protein tyrosine kinase EphB4is up-regulated in colon cancer. BMC Mol Biol 2001;2:15–23.

31. Berclaz G, Andres AC, Albrecht D, et al. Expression ofthe receptor protein tyrosine kinase myk-1/htk innormal and malignant mammary epithelium. BiochemBiophys Res Commun 1996;226:869–75.

32. Andres AC, Zuercher G, Djonov V, Flueck M,Ziemiecki A. Protein tyrosine kinase expression duringthe estrous cycle and carcinogenesis of the mammarygland. Int J Cancer 1995;63:288–96.

33. Berclaz G, Karamitopoulou E, Mazzucchelli L, et al.Activation of the receptor protein tyrosine kinaseEphB4 in endometrial hyperplasia and endometrialcarcinoma. Ann Oncol 2003;14:220–6.

34. Takai N, Miyazaki T, Fujisawa K, Nasu K, Miyakawa I.Expression of receptor tyrosine kinase EphB4 and itsligand ephrin-B2 is associated with malignant potentialin endometrial cancer. Oncol Rep 2001;8:567–73.

35. Henshall SM, Afar DE, Hiller J, et al. Survival analysisof genome-wide gene expression profiles of prostatecancers identifies new prognostic targets of diseaserelapse. Cancer Res 2003;63:4196–203.

36. Quinn DI, Henshall SM, Haynes AM, et al. Prognosticsignificance of pathologic features in localized prostatecancer treated with radical prostatectomy: implicationsfor staging systems and predictive models. J Clin Oncol2001;19:3692–705.

37. Hailey J, Maxwell E, Koukouras K, Bishop WR,Pachter JA, Wang Y. Neutralizing anti-insulin-likegrowth factor receptor 1 antibodies inhibit receptorfunction and induce receptor degradation in tumorcells. Mol Cancer Ther 2002;1:1349–53.

38. Liu W, Akhand AA, Kato M, et al. I.4-hydroxynonenaltriggers an epidermal growth factor receptor-linkedsignal pathway for growth inhibition. J Cell Sci 1999;112:2409–17.

39. Masood R, Kundra A, Zhu S, et al. Malignantmesothelioma growth inhibition by agents that targetthe VEGF and VEGF-C autocrine loops. Int J Cancer2003;104:603–10.

40. Tang XX, Brodeur GM, Campling BG, Ikegaki N.Coexpression of transcripts encoding EphB receptorprotein tyrosine kinases and their ephrin-B ligands inhuman small cell lung carcinoma. Clin Cancer Res 1999;5:455–60.

41. Wu Q, Suo Z, Risberg B, Karlsson MG, Villman K,Nesland JM. Expression of Ephb2 and Ephb4 in breastcarcinoma. Pathol Oncol Res 2004;10:26–33.

42. Berclaz G, Flutsch B, Altermatt HJ, et al. Loss ofEphB4 receptor tyrosine kinase protein expression

Expression of EphB4 in Prostate Cancer

www.aacrjournals.org 4631 Cancer Res 2005; 65: (11). June 1, 2005

during carcinogenesis of the human breast. Oncol Rep2002;9:985–9.

43. Nupponen N, Visakorpi T. Molecular biology ofprogression of prostate cancer. Eur Urol 1999;35:351–4.

44. Vlietstra RJ, van Alewijk DC, Hermans KG, vanSteenbrugge GJ, Trapman J. Frequent inactivation ofPTEN in prostate cancer cell lines and xenografts.Cancer Res 1998;58:2720–3.

45. Munarini N, Jager R, Abderhalden S, et al. Alteredmammary epithelial development, pattern formationand involution in transgenic mice expressing the EphB4receptor tyrosine kinase. J Cell Sci 2002;115:25–37.

46. Liu ZJ, Shirakawa T, Li Y, et al. Regulation of Notch1

and D114 by vascular endothelial growth factor inarterial endothelial cells: implications for modulatingarteriogenesis and angiogenesis. Mol Cell Biol 2003;23:4–25.

47. D’Amore PA, Ng Y-S. Won’t you be my neighbor?Local induction of arteriogenesis. Cell 2002;110:289–92.

48. Bruhl T, Urbich C, Aicher D, Acker-Palmer A, ZeiherAM, Dimmeler S. Homeobox A9 transcriptionallyregulates the EphB4 receptor to modulate endothelialcell migration and tube formation. Circ Res 2004;94:743–51.

49. Noren NK, Lu M, Freeman AL, Koolpe M, Pasquale

EB. Interplay between EphB4 on tumor cells andvascular ephrin-B2 regulates tumor growth. Proc NatlAcad Sci U S A 2004;101:5583–8.

50. Marston DJ, Dickinson S, Nobes CD. Rac-dependenttrans -endocytosis of ephrinBs regulates Eph-ephrincontact repulsion. Nat Cell Biol 2003;5:879–88.

51. Zimmer M, Palmer A, Kohler J, Klein R. EphB-ephrinB bi-directional endocytosis terminates adhesionallowing contact mediated repulsion. Nat Cell Biol 2003;5:869–78.

52. Martiny-Baron G, Korff T, Schaffner F, et al.Inhibition of tumor growth and angiogenesis by solubleEphB4. Neoplasia 2004;6:248–57.

Cancer Research

Cancer Res 2005; 65: (11). June 1, 2005 4632 www.aacrjournals.org