Signiﬁcance of Divergent Expression of Prostaglandin EP4 ...and EP4 (15, 16) which are frequently co-expressed in the same cells but use different and/or opposing signaling path-ways

Signal Transduction

Significance of Divergent Expression of Prostaglandin EP4and EP3 Receptors in Human Prostate Cancer

Hosea F. S. Huang1, Ping Shu1, Thomas F.Murphy1, SeenaAisner2, Valerie A. Fitzhugh2, andMark L. Jordan1,3

AbstractPGE2 has been implicated in prostate cancer tumorigenesis. We hypothesized that abnormal prostaglandin

receptor (EPR) expression may contribute to prostate cancer growth. Twenty-six archived radical prostatectomyspecimens were evaluated by immunohistochemistry (IHC) andWestern blotting for the expression of EP1, EP2,EP3, and EP4. As a corollary, EPR expression in one normal (PZ-HPV7) and four prostate cancer cell lines (CA-HPV10, LNCaP, PC3, and Du145) were assessed by Western blotting. Prostate cancer and normal cell growthwere compared in vitro after EPR blockade, siRNA EPR knockdown, or overexpression. EP1, EP2, EP3, and EP4receptors were detected by IHC in all areas of benign tissue within the clinical prostate cancer specimens. In areas ofprostate cancer, EP4 and EP2 were overexpressed in 85% (22 of 26) and 75% (18 of 24) and EP3 expression wasreduced in all (26 of 26, 100%) specimens (P < 0.05 vs. benign tissue). EP1 showed no specific differentialexpression pattern. Increased EP4 and reduced EP3 was confirmed byWestern blotting in fresh clinical specimensand in prostate cancer cell lines (CA-HPV10, LNCaP, PC3, and Du145) compared with the normal prostate cellline (PZ-HPV7). EP2 and EP4 siRNA knockdown resulted in reduced in vitro growth and metastasis-related geneexpression (MMP9 andRunx2) of prostate cancer lines, and in vitromigration was inhibited by EP4 antagonists. Asa corollary, EP3-overexpressing PC3 cells displayed impaired growth in vitro. Human prostate cancer is associatedwith EP4 and EP2 overexpression and reduced EP3 expression. These data suggest that targeting specific EPRmayrepresent a novel therapeutic approach for prostate cancer. Mol Cancer Res; 11(4); 427–39. �2013 AACR.

IntroductionProstate cancer accounts for 30% of all new male cancer

diagnoses, and next to lung cancer, is the second leadingcause of cancer death in men (1). More than 32,000 U.S.men died of prostate cancer in 2010, and the incidence ispredicted to increase as the male population continues to age(1). Epidemiologic studies have shown a direct link betweenthe incidence of prostate cancer and dietary lipid and itsmetabolism (2). For example, there is a higher incidence ofprostate cancer in U.S. Caucasian and African Americanmales than in Asian males, and the incidence of prostatecancer increases in Asianmales who adopt aNorth Americandiet (3). Arachidonic acid (AA) has been shown to promoteprostate cancer cell growth mediated, in part, by the for-mation of prostaglandin E2 (PGE2). PGE2 is known to have

potent effects on immune responses and in oncogenesis of avariety of tumors (4, 5). In prostate cancer, for example,PGE2 acts as a tumor promoter by stimulating cell prolif-eration and inhibiting apoptosis (6, 7). Furthermore, COX-2, the rate-limiting enzyme for PGE2 synthesis from arachi-donic acid, is overexpressed in prostate cancer (8, 9). Theseobservations, taken together with the known suppression ofprostate cancer cell growth by COX-2 inhibitors and non-steroidal anti-inflammatory agents, have suggested thatCOX inhibition may present a logical therapeutic approachfor prostate cancer (10). However, translation of thesefindings into clinical use has been hampered by incompleteknowledge of the mechanisms involved in the effects ofPGE2 on prostate cancer, and COX-2 inhibitors have as yetnot shown clinical efficacy (11).PGE2 exhibits multiple regulatory effects on cell prolif-

eration, apoptosis, motility (4, 6), immune surveillance (12,13) and induces angiogenesis (14). PGE2 exerts both auto-crine and paracrine functions by binding to and activation ofthe specific G-protein–coupled receptors EP1, EP2, EP3,and EP4 (15, 16) which are frequently co-expressed in thesame cells but use different and/or opposing signaling path-ways. Altered EPR expression has been reported in variousmalignancies. For example, EP2 overexpression has beenreported in skin and esophageal cancer (17, 18). EP4 isupregulated in colon and breast cancer (19, 20) and glio-blastoma (21). The functional relevance of EP2 and EP4overexpression in these cancers has been corroborated by

Authors' Affiliations: 1Division of Urology, UMDNJ-New Jersey MedicalSchool; 2Department of Pathology, UMDNJ-New Jersey Medical School,University Hospital, Newark; and 3Department of Veterans Affairs NewJersey Healthcare System Medical Center, East Orange, New Jersey

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

Corresponding Author: Mark L. Jordan, Harris L. Willits Professor andChief, Division of Urology, UMDNJ-NJMS, ACC Suite G1680 140 BergenSt., Newark, NJ 07103. Phone: 973-972-4488; Fax: 973-972-3892; E-mail:[email protected]

doi: 10.1158/1541-7786.MCR-12-0464

�2013 American Association for Cancer Research.

MolecularCancer

Research

www.aacrjournals.org 427

on June 14, 2021. © 2013 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst January 30, 2013; DOI: 10.1158/1541-7786.MCR-12-0464

http://mcr.aacrjournals.org/

enhanced cell proliferation, migration, invasion in vitro andtumorigenesis and metastasis in vivo induced by EP2 andEP4 agonists (18–21). As a corollary, EP2 and EP4 antago-nists suppress cancer cell growth, migration/invasion, andcolony formation in vitro and attenuate tumorigenesis andmetastasis in vivo (22, 23). These data suggest that targetingspecific EP receptors may offer an attractive and morespecific therapeutic approach for the treatment of prostatecancer than has been previously accomplished with broad-based COX inhibition. However, the characteristics of EPreceptor expression in human prostate cancer have not beenwell-established, thus potential rationale for targeting ofspecific receptors is unclear. The current study was under-taken to characterize EPR expression in normal and prostatecancer cells in prostate tissue from patients undergoingradical prostatectomy for clinically localized prostate cancerand to determine whether tissue expression of EP receptorsin human prostate cancer correlates with Gleason grade,stage, and clinical outcome. The patterns of EPR expressionwe detected in resected prostate cancer were recapitulated bythe results observed in human prostate cancer cell lines. Wealso showed that siRNA knockdown of the EP2 and EP4genes and stable overexpression of EP3 attenuated the invitro growth of human prostate cancer cells. EP2 and EP4knockdown was also associated with reduced MMP9 andRunx2 expression. These observations provide a scientificrationale for targeting EPR signaling as a therapeuticapproach in the management of prostate cancer.

Materials and MethodsPatients and tissuesAll human studies described were approved by the Insti-

tutional Review Board (IRB) of the New Jersey MedicalSchool (NJMS). Archived, formalin-fixed, paraffin-embed-ded human prostate tissue from 26 patients who underwentradical prostatectomy for clinically localized prostate cancer inthe Division of Urology, University of Medicine and Den-tistry of New Jersey (UMDNJ)–NJMS were used. None ofthese patients had undergone prior hormonal treatment orradiation. Immunohistochemistry (IHC) using affinity-puri-fied polyclonal antibodies and Alexa Fluor–conjugated sec-ondary antibody assessed expression of EPR in both prostatecancer and normal prostate epithelia within the same patientsample. Tissue sections (8 mm) were first stained withhematoxylin/eosin (HE) for pathologic evaluation by 1 of2 board-certified pathologists (S. Aisner and V. Fitzhugh).Adjacent sections were subsequently used for immunos-taining of EPR in our laboratory. Fresh specimens wereexamined immediately after radical prostatectomy in anadditional 3 cases. The tissue blocks (2 � 2 � 2 mm3)from areas representing a significant cluster (>80%) ofprostate cancer or benign glands were dissected and storedat �80�C for future assay.

Immunohistochemistry of EPRTissue sections for immunostaining were processed by

deparaffinization in xylene, rehydration in graded ethanol,

antigen retrieval in boiling 0.01 mol/L sodium citrate (pH6), and blocking with 10% normal goat serum (NGS). Afterwashing in PBS (5 minutes � 2), sections were incubatedwith affinity purified anti-EPR antibodies (1:100, Caymanbiological) diluted in 4% NGS in PBS overnight at 4�C. Asnegative controls, sections were incubated with normalrabbit serum (1:100) or antigen-absorbed primary antibody,washed in PBS (5 minutes � 2) and incubated in darknesswith anti-rabbit IgG conjugated with Alexa Fluor 488fluorescent dye (1: 100 in PBS) for 60 minutes. Sectionswere examined with a Nikon Eclipse 400 fluorescent micro-scope and photographed. Intensities of immunostaining ofEPR in prostate cancer and adjacent normal epithelia werecompared with a color chart representing increasing degreesof green fluorescence and scored 1 to 4 (1, negative orbackground; 2, weak; 3, moderate; 4, strong). In each area, atotal of 6 to 12 clusters of cancer glands/cells or normalepithelia were scored.

Cell lines and cultureHuman normal prostate epithelial cells line PZ-HPV7

and prostate cancer cell lines from primary tumor (CA-HPV10) or metastasis (LNCaP, PC3, and Du145) werepurchased from American Type Culture Collection. All cellswere maintained in the recommended medium until themonolayers reached 85% to 90% confluence.

RNA isolation and RT-PCRCells were seeded onto 6-well plates in their respective

medium, cultured until 80% confluence, then rinsed withPBS twice and total RNAwas isolated using RNeasyMini kit(Qiagen). Gene expression was examined by reverse tran-scription polymerase chain reaction (RT-PCR) using specificprimer pairs (Supplementary Data). To identify EP3 iso-forms, total RNA from normal and prostate cancer cells weresubjected to RT-PCR using combinations of primer pairs(Supplementary Data) specific for different EP3 isoforms(24).

Real-time PCRQuantitative real-time PCR was carried out using Power

SYBR Green PCR Master Mix (Applied Biosystems) andmonitored using 7500 Real time PCR system (AppliedBiosystems). Negative controls without template wereincluded, and all reactions were conducted in triplicate. Theamplification conditions were 50�C for 2 minutes, 95�C for10 minutes, 40 cycles of 95�C for 15 seconds, and 60�C for1 minute.

Western immunoblottingCultured prostate cancer cells or homogenized freshly

obtained human prostate tissue were lysed in ice-cold lysisbuffer (Cell Signaling Technology) supplemented with pro-tease inhibitor cocktail. Protein concentrations were deter-mined by using BCA protein assay kit (Pierce). Proteinextracts were mixed with 2� Laemmli sample buffer (Bio-Rad Laboratories) and denatured at 95�C for 5 minutes andthen separated by SDS-PAGE (Bio-Rad Laboratories) and

Huang et al.

Mol Cancer Res; 11(4) April 2013 Molecular Cancer Research428




transferred to a polyvinylidene fluoride (PDVF) membrane(Bio-Rad Laboratories). The membrane was subjected toimmunoblot analysis with each primary antibody (EP11:500, EP2 1:500, EP3 1:500, EP4 1:600, Cayman Chem-ical; actin 1:800, Santa Cruz Biotechnology) followed byhorseradish peroxidase–conjugated secondary antibody(1:5,000 dilution, Santa Cruz Biotechnology). The proteinbands were visualized using Amersham ECL Western blot-ting analysis system (GE Healthcare).

Transient EPR knockdownCells (PZ-HPV7, CA-HPV10, and Du145) were seeded

in 6-well plates in 2 mL of appropriate medium to achieve30% to 40% confluence after 24 hours. 2.5 mL of SMART-pool siRNA duplex that specifically targets EP2 or EP4 ornontargetting pool (100 mmol/L, Dharmacon) and 5 mL ofLipofectamine RNAiMAX (Invitrogen) were first diluted in250 mLOpti-MEM I reduced serummedium separately andthen combined and added to each well, mixed gently, andincubated at 37�C in 5%CO2 for 48 to 72 hours. IndividualEPR expression after knockdown was assessed by RT-PCR.To determine the effects of EP2/EP4 knockdown on cellgrowth, PZ-HPV7, CA-HPV10, or Du145 cells (2 � 103per well) were seeded in 96-well plates and allowed to settleovernight. On day 2, EP2/EP4 siRNA knockdown wasconducted using the siRNA-Lipofectamine RNAiMAXmixture described above. The transfection mixture wasremoved 6 hours later and cell growth determined 48 to120 hours later by MTT assay.

Preparation of conditioned mediaTo prepare conditioned medium, 48 hours after siRNA

transfection, cells were rinsed with serum-free medium andincubated with serum-free medium for 24 hours. Mediumwas collected and centrifuged at 1,000 rpm for 5 min at 4�Cto eliminate cellular debris. Cell number was determined foreach sample. The medium was concentrated using AmiconUltra-4 centrifugal filter units (Millipore) with a molecularweight cutoff of 30 KDa.

Gelatin zymographyEqual amounts of protein from conditionedmediumwere

mixed with 2� Laemmli sample buffer (Bio-Rad) andelectrophoresed on 8% PAGE containing 1 mg/mL gelatin.The gel was washed with 2.5% Triton X-100 and incubatedfor 24 hours at 37�C in 1� Developing buffer (Bio-Rad).After incubation, the gel was stained with 0.5% CoomassieBlue R-250 in 50% methanol and 10% acetic acid anddestained in methanol:acetic acid:water (50:10:40). Theintensity of the digested bands on zymograms was quantifieddensitometrically with AlphaImager 3400 (Alpha Innotech).

Construction of EP3II expression plasmidEP3II cDNA was amplified from the pcDNA3.1zeo

(�)/EP3II plasmid (a kind gift from Dr. Barrie Ashby,Temple University, Philadelphia, PA) with the followingprimers: the sense primer included nucleotides 1–23 (bold)and contained aHindIII site (italics): 50-ATTATTAAGCT-

TATGAAGGAGACCCGGGG-CTACGG-30. Antisenseprimer represented nucleotides 1,144 to 1,167 (bold) andcontained an XhoI site (italics): 50-CCGCCGCTCGAGT-CATGCTTCTGTCTGTA-TTATTTC-30. PCR productswere electrophoresed on 1% agarose gel, excised, and puri-fied using QIAQUICK gel purification kit (Qiagen). Toconstruct the EP3II expression vector, the EP3II cDNA andthe pcDNA3.1(þ) expression vector were digested using therestriction enzymesHindIII and XhoI. The EP3-II fragmentwas then ligated into the pcDNA3.1(þ) vector using T4DNA ligase. The plasmid clones were sequenced using T7and BGHrev primers at NJMSMolecular Resource Facility.The plasmid generated (pcDNA3.1/EP3II) was subsequent-ly expanded and used for transfection experiments.

Stable overexpression of EP3 receptor genesPC3 cells seeded onto 6-well plates were grown to 50% to

60% confluence in complete medium and then transfectedwith the EP3 expression plasmid using Lipofectamine 2000for 6 hours. After 48 hours, 500 mg/mL G418 (Invitrogen)was added, and colonies derived from single cells wereselected for cells carrying the plasmid. After 2 weeks,G418-resistant colonies were expanded in the presence ofculture medium with G418 at 500 mg/mL. EP3 overexpres-sion was verified by RT-PCR and immunostaining withspecific antibody.

In vitro migration assayPC3 cells (2.5� 104) suspended in RPMI-1640 medium

supplemented with 10% BCS were seeded into the uppercompartment of a modified Boyden chamber (membranepore size, 8 mm) in 12-well plates in triplicates and culturedovernight. The cells were exposed to reagents in the presenceor absence of 0.1% BCS added to the lower chamber for anadditional 24 hours. The cells that migrated through themembrane into the lower chamber were enumerated after 24hours.

Extracellular signal–regulated kinase phosphorylationPZ-HPV7, CA-HPV10, and Du145 cells were seeded in

6-well plates for 24 hours and then serum-starved (Du145)for an additional 24 hours or protein and growth factor–starved (PZ-HPV7 and CA-HPV10) for 48 hours and thenwere treated with PGE2 (0.5 mmol/L) for 5 to 60 minutes.Cellular proteins were subsequently isolated and subjectedto SDS-PAGE. After transfer to PDVF membranes, phos-phorylated ERK (p-ERK)was detected using anti-phosphor-ylated ERK antibody. The membranes were then strippedand reprobed with antibody against total ERK. To deter-mine effects of EP2 or EP4 blockade, serum or protein/growth factor–starved cells were preincubated with the EP2antagonist AH6809 or EP4 antagonist AH23848 (10 or 50mmol/L, respectively) or vehicle [0.1% or 0.5% dimethylsulfoxide (DMSO), respectively] for 60 minutes beforePGE2 (0.5 mmol/L) treatment. The cells were collected atvarious time points after addition of PGE2, and total and p-ERK were determined and quantified by densitometry. Thep-ERK level of each sample was normalized against the

EP Receptors and Prostate Cancer

www.aacrjournals.org Mol Cancer Res; 11(4) April 2013 429




corresponding total ERK level. Untreated samples were setas unity, and the fold change in p-ERK level of each samplewas calculated as a ratio against the corresponding referencesample.

StatisticsAll experiments were carried out a minimum of 3 times

and results analyzed by ANOVA and Student's t test asappropriate. When significant (P < 0.05), Dunnett's mul-tiple range test was used to compare means among groups.

ResultsExpression of EPR in human prostate cancer tissueThe demographics, Gleason scores, pathologic stage, pre-

and postoperative prostate-specific antigen (PSA), and EPRexpression scores as determined by IHC from the 26archived clinical samples are presented in Table 1. In all26 patients, EP1, EP2, EP3, and EP4 receptors weredetected to varying degrees both in areas of normal prostate(NP) epithelia and prostate cancer (representative patientsample Fig. 1A). However, in prostate cancer, the overallintensity of EP4 immunostaining was consistently stronger(P < 0.001) and EP3 was significantly weaker (P < 0.001)than in the adjacent NP (Fig. 1B). Increased EP4 anddecreased EP3 in prostate cancer cells were observed in22 of 26 (85%) and 25 of 25 (100%) specimens examined.Increased EP2 expression in prostate cancer was alsoobserved in 18 of 24 (75%) patients (P < 0.005 vs. NP).In contrast, the intensity of EP1 immunostaining in areas ofprostate cancer compared with the adjacent NP exhibited nospecific pattern. (Table 1 and summarized in Fig. 1B).Within each patient, the intensities of EP2, EP3, and EP4immunostaining in prostate cancer cells did not correlatewith Gleason grade (P > 0.1).

Western blotting of EPR in fresh clinical samplesTo further validate the IHC results obtained from

archived specimens, Western blot analysis of freshlyobtained prostate cancer specimens using the same antise-rum that had been used for IHC was conducted. Figure 1Cshows representative Western blots of EP1–4 in 3 normaland 3 prostate cancer patient samples assayed in the sameblot, showing greater EP4 and reduced EP3 protein levels inprostate cancer when compared with NP. These observa-tions were confirmed by quantitative comparison of theintensities of protein bands after normalization against thatof b-actin intensity of each sample (Fig. 1D). EP1 and EP2protein levels were not statistically different (not shown).

EPR in human cell linesTo determine whether the differential EPR expression we

observed in the clinical specimens (Table 1) would berecapitulated in human prostate cancer lines, the expressionof EPR was determined in normal (PZ-HPV7, normalprostate) and prostate cancer cell lines (CA-HPV10, primarytumor; LNCaP, PC3, and Du145, prostate cancer linesmetastatic to lymph node, bone, and brain, respectively).

Consistent with our observations in resected human prostatecancer tissue (Table 1 and Fig. 1B), the prostate cancer linesexhibited lower EP3 and higher EP4 expression (comparedwith the benign prostate line PZ-HPV7) after normalizationagainst b-actin in each sample (Fig. 1E). There was noconsistent pattern of EP1 and EP2 in prostate cancer cellscompared with those of normal lines (not shown).

Specificity of siRNA knockdown of EP2 and EP4receptor genesThe increased EP4 and EP2 expression in our clinical

prostate cancer specimens suggested that increased expres-sion might contribute to prostate cancer tumorigenesis. Weconducted transient EP2 andEP4 siRNAknockdown in PZ-HPV7 (NP line) and CA-HPV10 (primary tumor line).These lines were selected as surrogates for normal andnonmetastatic prostate cancer, respectively. The Du145(brain metastatic cell line) was also used as a surrogate formetastatic prostate cancer for comparison with the primarytumor line CA-HPV10. Figure 2A confirms that the recep-tor gene knockdown was specific to the siRNA used andthere was no interference with the expression of the non-relevant EPR.

Physiologic relevance of EP2/EP4 signaling in prostatecancer cellsWe next assessed transient siRNA knockdown of EP2 and

EP4 on the growth of CA-HPV10, PZ-HPV7, and Du145cells. Growth of normal (PZ-HPV7) cells was not affected byEP2 KD (P > 0.05 vs. nontargeting siRNA) but wasmoderately suppressed by EP4 KD (P < 0.05) at 96 hours(Fig. 2B). In contrast, the growth of CA-HPV10 cells wassuppressed by more than 50% (P < 0.001, P < 0.01) by bothEP2 and EP4 KD at 72 and 96 hours. EP2 KD alsosuppressed the growth of Du145 cells (P < 0.001) at 72and 96 hours; however, EP4 KD did not affect Du145 cellgrowth significantly until 96 hours (P < 0.05). This wasconfirmed in 3 subsequent replicate experiments.

EP2 and EP4 KD reduce MMPs and Runx2AsMMP2 andMMP9 have been shown to be involved in

prostate cancer metastasis and invasion (25) and PGE2promotes tumor cell migration, we hypothesized thatMMP2 and 9 may be downstream targets responsive toEPR signaling. Figure 2C shows real-time PCR results ofMMP2, MMP9, and Runx2 expression 72 hours aftertransient EP2 and EP4 KD in PZ-HPV7, CA-HPV10, andDu145 cell lines. EP2 and EP4 KD resulted in reducedexpression of MMP9 mRNA in both PZ-HPV7 and CA-HPV10 cells. MMP2 expression was more variably affected:MMP2 mRNA was reduced in PZ-HPV7 after both EP2and EP4KDbut only by EP2KD inCA-HPV10. InDu145cells, EP2 KD resulted in reduced MMP9 but had no effecton MMP2 expression, whereas EP4 KD appears to haveparadoxically increased MMP2 and MMP9. As Runx2 isabnormally expressed in metastatic prostate cancer cells butnot in benign cells, and Runx2 knockdown has recently beenshown to inhibit prostate cancer cell migration in vitro and

Huang et al.





Tab

le1.

Patient

dem

ographicdataan

dEPRex

pression

Gleas

on

Patho

logic

Preoperative

Postoperative

EP1sc

ore

EP2sc

ore

EP3sc

ore

EP4sc

ore

Ptno

.Age

score

stag

ePSA

PSA

NCaP

NCaP

NCaP

NCaP

147

3þ

4pT2

cNxM

x8.1

0.1

3.3

2.13

2.43

2.34

2.27

1.36

1.69

2.16

274

3þ

5pT2

cN0M

x14

.7

growth in vivo (26), the effects of EP2/4 knockdown onRunx2 expression were also examined. EP2 and EP4 KDresulted in decreased Runx2 expression in all 3 cell lines. Thepotential reasons for differential signaling via EP2 and EP4inDu145 comparedwith the 2 other cell lines are explored inmore detail in the ERK experiment below.

MMP2 andMMP9 enzymatic activity after EP2 and EP4knockdownTo determine whether the altered mRNA expression

observed by real-time PCR in Fig. 2C correlated with func-tional enzymatic activity ofMMP2 andMMP9 after EP2 andEP4KD, conditionedmediumof each siRNA-transfected cellline was collected, and equal amounts of protein were sub-jected to gelatin zymography. The intensity of the digestedbands corresponding to MMP9 and MMP2 (Fig. 3A) wasmeasured by densitometry and expressed as fold level of thenontargeting siRNA-transfected control which was expressed

as unity (Fig. 3B). These results recapitulated the PCR data,with the exception ofMMP2enzymatic activity after EP4KDof Du145 which was not elevated.

EP4 and prostate cancer migration in vitroIn vitromigration as a surrogate for invasive potential was

assessed using a Boyden chamber assay. PC3 was used as arepresentative metastatic cell line in this assay. Controlmigration in Dulbeccos' Modified Eagle's Media (DMEM)was assessed in the presence or absence of 1% BCS as apositive and negative control for chemotaxis, respectively.PGE2 (10–1,000 nmol/L) did not induce chemotaxis in theabsence of BCS. However, PGE2 at physiologic concentra-tions (1,000 nmol/L) significantly (P< 0.05) increased BCS-induced chemotaxis compared with BCS alone (Fig. 4A).PC3 cells were then pretreated with the EP4 antagonistsAH23848 and L161,982 (10 and 50 mmol/L) or vehicle(0.1%–0.5%DMSO) 1 hour before induction of migration

EP1

EP2

EP3

EP4

A

B

C

D

E

0

1

2

3

Rel

ativ

e ab

unda

nce

EP1 EP2 EP3 EP4

4

3

2

1

0

Sco

re

Normal

CaP

P < 0.01

P < 0.05

EP1

EP2

EP3

EP4

EP1 EP2 EP3 EP4N CaPN CaPN CaPN

n = 26 n = 24 n = 25 n = 26

P < 0.001

P < 0.001

P < 0.005P > 0.01

CaP

Normal CaP

β-Actin

β-Actin

Normal CaP linesDu145

EP3

EP4

1.00 1.47 5.03 1.68 1.60

0.450.730.600.691.00

CA-HPV10PZ-HPV7 PC3LNCaP

Figure 1. Expression of EPR in human prostate cancer (CaP) cells. A, micrographs of hematoxylin and eosin (HþE) staining (left), and correspondingimmunostaining for EPR (right) in prostate tissues from a representative patient, showing reduced EP1 and EP3 but enhanced EP2 and EP4 immunostainingin cancer cells (CaP;Gleason 3or 4) comparedwith that in adjacent normal epithelia (N). B, scoring andquantitative comparisonof the intensity of EP3andEP4immunostaining in normal and CaP cells (mean � SD, n ¼ 24–26). The P values indicate significant differences between CaP cells and adjacent areasof normal epithelium within the same section. C, representative Western blots of EPR proteins in freshly obtained human prostate tissues showing 3 normaltissues and 3 CaP tissues showing reduced EP3 and increased EP4 proteins in CaP tissues compared with that in normal tissues. EP1 and EP2 protein levelsvaried among these patients. D, quantitative comparison of the abundance of EP3 and EP4 between normal and CaP tissues after normalized againstthat of b-actin in each sample. Results are presented asmean�SD. n¼ 5. ��,P < 0.01; �,P < 0.05. E,Western blot analyses of EP3 and EP4 proteins in normal(PZ-HPV7) andCaP cell lines (CA-HPV10, LNCaP, PC3, andDu145). The density of each bandwas first normalized against that of b-actin in each sample. Thenumber under each band represent relative abundance of EP3 or EP4 protein in each CaP line relative to that in PZ-HPV7 cells.

Huang et al.





by BCS (1%)þ PGE2 (1,000 nmol/L). As shown in Fig. 4B,EP4 antagonist pretreatment of the cells suppressed migra-tion toward PGE2þBCS by 50% (P < 0.01) compared withrespective vehicle (0.1% or 0.5% DMSO). These resultsindicate that PGE2-induced PC3 migration is mediated, inpart, by EP4.

EP2/4 blockade inhibits PGE2-induced ERKphosphorylationPhosphorylation of ERK is necessary for EP4-mediated

cell migration and cell growth (27, 28) and PGE2-mediatedprostate cancer angiogenesis (14). We therefore examinedwhether the effects of EP2/4 blockade were mediated, inpart, via ERK phosphorylation. Figure 5A shows that max-imal induced ERK1/2 phosphorylation occurred within 5 to10 minutes after exposure to 0.5 mmol/L of PGE2 in the 3cell lines (PZ-HPV7, CA-HPV10, and Du145) evaluated.Cells were then pretreated with the EP2 antagonist AH6809or EP4 antagonist AH23848 (10 or 50 mmol/L, respective-ly), or vehicle (0.1% or 0.5% DMSO) for 1 hour, followedby addition of 0.5 mmol/L PGE2. Resultant p-ERK and totalERK were determined by Western blotting at the corre-sponding time point of peak PGE2-induced ERK phosphor-ylation for each cell line (Fig. 5A: 5minutes for PZ-HPV7 or10 minutes for CA-HPV10 and Du145). Figure 5B showsthat PGE2-elicited ERK1/2 phosphorylation was not affect-ed by vehicle (DMSO, 0.1% or 0.5%) but was markedly

suppressed by the EP2 antagonist (AH6809) at both 10 and50 mmol/L in PZ-HPV7, CA-HPV10, and Du145 cells.The effects of the EP4 antagonist AH23848 were morevariable. AH23848 suppressed PGE2-induced ERK phos-phorylation in CA-HPV10 at 10 and 50 mmol/L, but in PZ-HPV7 cells at only 50 mmol/L, AH23848. AH23848 hadminimal effects on PGE2-induced ERK phosphorylation inDu145 cells. These results are consistent with the effects ofEP2 and EP4 knockdown on PZ-HPV7 and CA-HPV10and Du145 growth shown in Fig. 2B.

EP3II is the major isoform in human prostate cellsThe consistently lower EP3 protein expression observed in

both the prostate cancer specimens and cell lines (Table 1and Fig. 1) suggests that altered EP3 expression may havebiologic significance in prostate cancer. Because of the com-plexity of EP3 isoforms and their diverse cellular functions(29, 30),we sought to identifyEP3 isoforms that are expressedin NP and prostate cancer cells. For this purpose, differentprimers specific for EP3 isoforms (ref. 24; SupplementaryData) were used in RT-PCR. EP3II (primer pairs P1 þ P3)was the only EP3 isoform detected in all prostate cell lines(PZ-HPV7, CA-HPV10, LNCaP, PC3, andDu145) used inour studies (Supplementary Data). This isoform was alsodetected in 2 additional human prostate lines (PREC andPWR-1E). Primer pairs P1þ P4 (Supplementary Data) andP1þ P2 (not shown) did not result in specific EP3 product.

Figure 2. Effects of EP2 and EP4siRNA knockdown (KD) on cellgrowth. A, time course andspecificity of EP2 and EP4 siRNAknockdown. B, effects of EP2 or EP4knockdown on cell growth. Thegrowth of normal (PZ-HPV7) cellswas not affect by EP2 KD (Pp > 0.05)but was slightly but significantlysuppressed by EP4 KD (P < 0.05) at96 hours. In contrast, the growth ofCA-HPV10 cells was suppressed by>50% (P < 0.001, P < 0.01) by eitherEP2 or EP4 KD. While EP2 KD alsosignificantly suppressed the growthofDu145cells (P

These results provide rationale for using the EP3II isoform insubsequent overexpression studies.

Stable EP3 overexpression inhibits prostate cancergrowth in vitroWe cloned the human EP3II isoform and sequence

analysis of the insert revealed 100% match with publishedsequences of human EP3II (24). We then constructed anexpression plasmid of human EP3II (Materials and Meth-ods). Stable transfection of PC3 cells with the EP3IIplasmid resulted in a PC3 subline overexpressing EP3II(EP3II OE) as shown by RT-PCR, Western blotting, andIHC (Fig. 6A–C). Figure 6D shows that EP3 II OE reducesPC3 growth in vitro compared with WT PC3. PC3 WTand EP3II OE cells were cultured in the presence ofindomethacin (10 mmol/L) to prevent endogenous PGE2production with either vehicle (0.01% DMSO) or the EP3agonist, sulprostone (1 mmol/L), for 5 days. The reducedgrowth of EP3II OE cells compared with WT PC3 cells (P< 0.01, Fig. 6D) was further exacerbated by sulprostone (P< 0.01). WT PC3 growth was not affected by sulprostone(Fig. 6E, P > 0.1). These preliminary results stronglysuggest that the overexpressed EP3 receptor is functionally

responsive to exogenous EP3 stimulation by the EP3agonist sulprostone.

DiscussionIt is generally reported that COX-2 is weakly expressed in

normal prostate tissue but overexpressed in CaP (8, 9).Overexpression of COX-2 in prostate cancer tissues (8, 9)is thought to result in higher local endogenous PGE2synthesis which may contribute to disease progression. Theconcurrence of increased COX-2 expression, local inflam-mation, neovascularization, and prostatic intraepithelialneoplasia (PIN) in human prostate tissues (31) is consistentwith COX2/PGE2-mediated carcinogenesis. Multiple invitro studies have also suggested that COX-2/PGE2 signalingplays a role in deregulated prostate cancer cell growth. Forexample, arachidonic acid–induced PC3 cell proliferation ispreceded by increased PGE2 production, and these effectswere associated with an increase in c-fos that in turnstimulates cell proliferation (32). When treated withCOX-2 inhibitors (NS398 or celecoxib), both LNCaP andPC3 cells exhibit increased apoptosis (33), suggesting thatendogenous PGE2 might promote prostate cancer cellgrowth by suppression of apoptosis.

DU145 conditioned media

nTG EP2 EP4

MMP9

MMP2

PZ-HPV-7 conditioned media

CA-HPV-10 conditioned media

nTG EP2 EP4

nTG EP2 EP4

BA

MMP9

MMP9

MMP2

MMP2

0

0.2

0.4

0.6

0.8

1

1.2

MMP2 MMP9

Inte

grat

ed d

ensi

ty

PZ-HPV-7

nTGsi

EP2si

EP4si

0

0.2

0.4

0.6

0.8

1

1.2

1.4

MMP2 MMP9

Inte

grat

ed d

ensi

tyCA-HPV-10

nTGsi

EP2si

EP4si

0

0.5

1

1.5

2

2.5

MMP2 MMP9

Inte

grat

ed d

ensi

ty

DU145

nTGsi

EP2si

EP4si

Figure 3. Effects of transient EP2/EP4 knockdown on MMP2 andMMP9 enzyme activity.Conditioned medium of eachsiRNA-transfected cell line wascollected and equal amounts ofprotein were subjected to gelatinzymography. The intensity of thedigested bands corresponding toMMP9 and MMP2 (A) wasmeasured by densitometry andexpressed as fold level of thenontargeting siRNA-transfectedcontrol which was expressed asunity (B). Data are from arepresentative experiment.

Huang et al.





PGE2 exerts its cellular effects by binding to and activationof the membrane-bound E-prostanoid receptors EP1, EP2,EP3, and EP4 (15, 16). Among these, EP2 and EP4 havebeen shown to be overexpressed in various cancers (17–20)and to promote cell growth and tumorigenesis, as well asmetastasis-related functions including angiogenesis, cellmotility, and invasion (14, 19–21). More recently, theexpression of EPR subtypes has been described in humanprostate cancer (14, 34, 35). Most studies of EP4 in prostatecancer growth had generally shown indirect receptor effectsby the use of pharmacologic antagonists or through inter-mediate effects on angiogenesis factors (14). More recentspecific data from Terada and colleagues (35) reportedupregulation of EP4 in prostate cancer xenografts grownin castrated nude mice compared with xenografts in intactmice, suggesting involvement of EP4 in the development ofthe castration-resistant prostate cancer (CRPC) phenotype.These authors also reported that overexpression of EP4facilitated androgen-independent growth in LNCaP cellsin vitro, which was attenuated by specific EP4 antagonistadministration (35). However, no studies to date haveshown direct involvement of EP2 and EP4 in the growthof human prostate cancer cells or established a direct causalrelationship between EP2/EP4 expression status and thedevelopment of primary tumor.

Our studies in 26 resected prostate cancer specimens fromour clinical urology program revealed consistent overexpres-sion of EP4 and reduced EP3 expression compared withbenign prostate tissue. In vitro experiments using humannormal prostate and prostate cancer cell lines verified theseclinical findings by showing higher EP4 protein levels whencompared with immortalized benign prostate cells. We nextevaluated the effects of siRNA knockdown of EP2 and EP4genes on the growth of PZ-HPV7 (normal) and CA-HPV10(primary tumor). Reduced in vitro growth of CA-HPV10cells after EP4 knockdown strongly suggests a role for EP4 in

0

2,000

1,0001,000101000PGE2 (nM)

BCS (1%) – – – +++

4,000

6,000

8,000

10,000

12,000

Con

trol

1% B

CS

+ 1 μ

M P

GE

2 10 50 10 50 0.1 0.5

AH23848(μmol/L) (μmol/L)

DMSO (%)L161,982

Cel

l num

ber

0

2,000

4,000

6,000

8,000

Cel

l num

ber

** ** ** **

****

*

P < 0.01

A

B

Figure 4. A, PGE2 is chemotactic for PC3 cell migration. Each barrepresents mean � SD of triplicate experiments. �, P < 0.05 versus BCSalone. B, EP4 antagonists, AH23848 and L161982 inhibit PGE2-directedPC3 cell migration. N ¼ 3, ��, P < 0.01 versus BCS þ PGE2.

Time (min)

PGE2 (0.5 μmol/L) 0 5 10 15 30 60

Time (min)

DMSO (%) AH6809 AH23848

DMSO AH6809 AH23848

DMSO AH6809 AH23848

0 5 10 15 30 60

Time (min)

0 5 10 15 30 60

PZ-HPV7

A

B

A

B

A

B

CA-HPV10

Du145

pERK1/2

Total ERK1/2

PGE2 (0.5 μmol/L)pERK1/2

Total ERK1/2

PGE2 (0.5 μmol/L)

PGE2C 50 (μmol/L)1050100.50.1

PGE2C 50 (μmol/L)1050100.50.1

PGE2C

1 1.08 1.11 1.08 0.95 1.01 1.07 1.02

1 1.16 1.13 1.13 0.92 1.95 1.05 1.03

50 (μmol/L)1050100.50.1

pERK1/2

Total ERK1/2

pERK1/2

Total ERK1/2Fold changed

pERK1/2

Total ERK1/2Fold changed

pERK1/2

Total ERK1/2Fold changed 1 1.25 1.13 1.08 0.80 0.70 1.17 1.12

Figure 5. Effects of EP2 and EP4 antagonists on PGE2-induced ERKphosphorylation in humanprostate lines. A, time course of PGE2-inducedERK phosphorylation in PZ-HPV7, CA-HPV10, and Du145 cells. Afterseeding in the specific medium for each line for 24 hours, cell lines wereserum-starved (Du145) for 24 hours or protein and growth factors–starved (PZ-HPV7 and CA-HPV10) for 48 hours, respectively and thenexposed to PGE2 (0.5 mmol/L) for 5 to 60 minutes. Cells weresubsequently collected for protein isolation, and total and p-ERK weredetected at each time point byWestern blotting using antibodies againsttotal ERK or p-ERK as indicated. B, effects of EP2 or EP4 blockade uponthe PGE2-induced ERK phosphorylation. After serum or protein/growthfactor starvation, the cells were preincubated with EP2 and EP4antagonist (AH6809 and AH23848, respectively, 10 or 50 mmol/L) for 60minutes before PGE2 (0.5 mmol/L) treatment. The cells were collected atthe time of maximal ERK phosphorylation (PZ-HPV7: 5 minutes; CA-HPV10 andDu145: 10minutes) specific for each cell line as determined in(A). Total and p-ERK were determined by Western blotting. The foldincrease in p-ERK shown was calculated by normalizing the ratio ofp-ERK:total ERK against that of untreated cells.






primary prostate cancer tumorigenesis. The inhibitoryeffects of EP4 knockdown on human prostate cancer cellgrowth reported herein have been corroborated by our recentfindings showing that stable siRNA knockdown of EP4 inmurine prostate cancer RM1 cells inhibited their growth invitro and tumorigenesis in vivo (36).The effects of EP2 and EP4 knockdown were next

examined in Du145 cells as a surrogate for metastaticprostate cancer. EP4 signaling has been reported to contrib-ute to the development of the androgen-independent phe-notype of prostate cancer cells (35). As Du145 cells arehighly metastatic and exhibit androgen-independentgrowth, the lack of a significant effect of EP4 knockdownon the growth of Du145 cells may suggest less sensitivity ofthese cells to EP4 receptor–mediated cell growth regulationcompared with that of the primary Ca-HPV10 line.Although the growth of PZ-HPV7 cells was not affectedby EP2 knockdown, both CA-HPV10 and Du145 cellgrowth was significantly reduced after EP2 knockdown.Although EP2 and EP4 act through similar signaling path-ways via adenylate cyclase (AC) activation, the amino acid

identity between these EP2 and EP4 is only 31%. EP2actually shares more homology with the prostacyclin recep-tor (IP) than with EP 4 (40% vs. 31%; ref. 15). Thesedifferences may therefore have functional implications: forexample, EP4 regulates murine dendritic cell migration,although both EP2 and EP4 are expressed by these cells(37). Our current study showed that PGE2- induced PC3migration is inhibited by 2 different EP4 receptor antago-nists, suggesting that human prostate cancer migration mayalso be mediated by EP4. Therefore, it is plausible thatalthoughEP2 andEP4may share similar signaling pathways,these receptors may exert different downstream effects inprostate cancer as we observed.The downstream effects of EP2 and EP4 signaling can be

mediated by MMP2 and MMP9 and the Runx familytranscription factors (38). Previous studies have shown thatMMP2 and MMP9 are overexpressed in prostate cancer(25, 38), and the effects of PGE2 on MMP9 expression aremediated through the activation of Runx 2 transcriptionfactors that are abnormally expressed in metastatic prostatecancer cells (38). Recent studies also reported knockdown of

A B

C

PC3, wt

PC3, wt EP3II OE PC3, wt EP3II OE

1.371.00

EP3

Relative density

β-Actin

1.351.00

EP3

Relative density

β-Actin

EP3II OEPC3, wt

N = 3

P < 0.001

P < 0.001

P < 0.01

EP3II OE

PC3, wt

Veh Sulp Veh Sulp

EP3 OE

80

60

40

20

0

800

600

400

200

0

Cel

l num

ber

(×10

4 )/w

ell

Cel

l num

ber

(×10

3 pe

r w

ell)

D

E

Figure 6. Stable overexpression of EP3II inhibits PC3 cell growth. PC3 cells were stably transfected with the EP3II expression plasmid constructed asdescribed in theMaterials andMethods. Stable transfection of PC3cellswith this plasmid and selection of clones derived from single cells resulted in a sublineoverexpressing EP3II as shown by (A) RT-PCR of EP3II mRNA, (B) Western blotting, and (C) immunostaining of EP3 protein. D, EP3II overexpressioninhibits PC3 growth. Wild-type (WT) and EP3II overexpressing PC3 cells (25,000 per well) were seeded in RPMI-1640 supplemented with 10% BCS withoutG418 and cell count enumerated after 6 days. Total cell number per well was significantly reduced (P < 0.001) in EP3II OE cells. Data are mean � SDfrom triplicate samples of a representative experiment of 3. E, EP3 agonist sulprostone exacerbates inhibition of PC3 growth by EP3 overexpression. Wild-type and EP3II stably overexpressing PC3 cells were cultured for 5 days in the presence of indomethacin (10mmol/L) in the presence of vehicle (0.01%DMSO)or the EP3 agonist sulprostone (1 mmol/L). Sulprostone had no effect on PC3 WT growth. However, the reduced growth of EP3II OE PC3 cells was furtherinhibited by sulprostone (P < 0.01 vs. vehicle). Data are mean � SD from triplicate samples of a representative experiment of 3.

Huang et al.





the Runx2 gene results in inhibition of prostate cancer cellmigration and growth in vivo (26). A previous study alsosuggested that Runx 3 transcription factor might be silencedby hypermethylation in prostate cancer (39). We did notobserve consistent changes in the expression of MMP2 andRunx 2 after EP2 or EP4 knockdown. On the other hand,we did observe a significant reduction of MMP9 mRNA inPZ-HPV and CA-HPV10 cells after either EP2 or EP4knockdown. The effects of EP2 and EP4 KD on Du145MMPandRunx2 differed from those in the primary prostatecancerHPV10 line in that both EP2 and EP4KD resulted inreduced expression of Runx2, EP2 KD in reduced MMP9expression, and EP4 KD increased MMP2 and MMP9expression. Whether the differences observed in the Du145line may be attributable to diminished EP4 regulatorycapacity after acquisition of androgen independence (35)remains to be determined.Distinct effects of EP4 siRNA knockdown on the growth

of primary CA-HPV10 cells and metastatic Du145 suggestvariability of EP4-mediated effects occur during diseaseprogression or metastasis. ERK phosphorylation has beenshown to mediate effects of EP4 signaling on various cellularfunctions including cell growth (14, 27, 28). The EP2antagonist AH6809 reduced PGE2-inducedERKphosphor-ylation of PZ-HPV7, CA-HPV10, and Du145 cells; how-ever, the effects of the EP4 antagonist AH23848, althoughinhibitory, were less marked than that of the EP2 antagonist,especially in the normal prostate cell line PZ-HPV7 and theandrogen-independent line Du145. This is consistent withthe idea that primary prostate cancer cells (represented bythe Ca-HPV10 line) overexpressing EP4 may be moresusceptible to EP4-mediated inhibition of deregulatedgrowth than metastatic or androgen-independent prostatecancer (Du145). The marginal effects of the EP4 antagonistAH23848 on the PGE2-induced ERK phosphorylation inDu145 cells also provide a possible explanation for the lackof significant inhibition of the growth of this line afterEP4 KD.Of the 4 EPR subtypes, EP3 is least well-understood and

its role in tumorigenesis unclear with divergent reports ofboth tumor promoting and suppressing properties (29, 30).Several studies have suggested a tumor promoter role forEP3: It was reported that PGE2-induced HCA-7 humancolon cancer cell migration may be mediated by EP3-induced VEGFR-1 expression (40). Others reported thattumor-associated angiogenesis is reduced in EP3 knockout(KO) mice (41). EP3 KO mice exhibit reduced growth ofcarcinogen-induced skin (42) of sarcoma 180 andmetastasisof murine Lewis lung cancer cells compared with wild-typemice (43). EP3 agonists have been shown to stimulate theproliferation of hepatocytes (44) and endometrial stromalcells (45). In contrast, other studies have found EP3 to beantineoplastic role: EP3 expression is reduced in coloncancer in mice and humans compared with normal mucosa,andEP3 agonists caused decreased colon cancer cell viability,and AOK-induced colon cancer was enhanced in EP3 KOmice (46), and overexpression of murine EP3 variantsimpairs human colon cancer (HCT116) and human embry-

onic kidney cell (HEK293) growth (47). EP3 also promotesS-phase arrest of 3T6 fibroblasts (48). EP3 agonists alsoinduced atypical cell death in neutrophils (49). The apparentdiscrepant activities of EP3 signaling may be due, in part, tothe existence of a variety of EP3 isoforms and signalingpathways. For example, it is known that multiple EP3receptor isoforms are generated by alternative splicing ofthe C-terminal tail (50). Functional differences among thesesplice variants have been reported. Macias-Perez and collea-gues (49) showed that overexpression of murine EP3 a, b,and g receptor variants in HCT116 cells resulted indecreased in vitro growth and in vivo tumorigenesis in nudemice and was associated with EP3 coupling to Gi protein.These emerging data suggest a tumor-suppressive role forEP3, at least, in colon cancer.However, to our knowledge, nostudies have examined the role of EP3 in prostate cancer.There are at least 8 human EP3II isoforms which have beenidentified and cloned (24, 30). These EP3 isoforms exert theirfunction throughmembraneGs (EP3I, EP3III, andEPIV) orGi (EP3II) proteins, or Caþþ/phospholipase signaling cas-cades (15, 24, 29, 30). We identified EP3II as the major (ifnot the only) EP3 isoform expressed in the normal andprostate cancer cell lines studied herein. We hypothesizedthat reduced EP3 expression may be consistent with tumor-suppressive properties of EP3 signaling in prostate cancer.Impaired cell growth after EP3overexpression in humanPC3cells and exacerbation of this effect by an EP3 agonistsupports this idea.We have also observed thatmurine EP3IIgOE results in reduced RM1 prostate cancer cell growth invitro and WT RM1 cell colony formation was promoted byan EP3 antagonist (unpublished data). We are undertakingfurther experiments to further characterize EP3II overexpres-sion in human prostate cancer in our laboratory.

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

Authors' ContributionsConception and design: H.F.S. Huang, M.L. JordanDevelopment of methodology: H.F.S. Huang, P. Shu, T.F. Murphy, M.L. JordanAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): H.F.S. Huang, S. Aisner, V. Fitzhugh, M.L. JordanAnalysis and interpretation of data (e.g., statistical analysis, biostatistics, compu-tational analysis): H.F.S. Huang, P. Shu, T.F. Murphy, M.L. JordanWriting, review, and/or revision of the manuscript: H.F.S. Huang, P. Shu, T.F.Murphy, M.L. JordanAdministrative, technical, or material support (i.e., reporting or organizing data,constructingdatabases):H.F.S.Huang,P.Shu,T.F.Murphy,V. Fitzhugh,M.L. JordanStudy supervision: H.F.S. Huang, M.L. Jordan

AcknowledgmentsThe authors thank the kind gift of the pc DNA3.1zeo(�)/EP3II plasmid fromDr.

Barrie Ashby, Temple University.

Grant SupportThe study was supported by the UMDNJ Foundation and VA Merit Review

Program (to M.L. Jordan).The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be herebymarked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Received July 31, 2012; revised December 10, 2012; accepted December 27, 2012;published OnlineFirst January 30, 2013.






References1. American Cancer Society. Cancer facts and figures 2011. Atlanta, GA:

American Cancer Society, 2011.2. Kolonel LN. Fat, meat, and prostate cancer. Epidemiol Rev 2001;

23:72–81.3. DelfinoRJ, Errini RL, Taylor TH, HoweS, Anton-Culver H. Demograph-

ic differences in prostate cancer incidence and stage: an examinationof population diversity in California. Am J Prev Med 1998;14:96–102.

4. Sheng H, Shao J, Washington MK, DuBois RN. Prostaglandin E2increases growth and motility of colorectal carcinoma cells. J BiolChem 2001;276:18075–81.

5. Harris SG, Padilla J, Koumas L, Ray D, Phipps RP. Prostaglandins asmodulators of immunity. Trends Immunol 2002;23:144–50.

6. Cao Y, Prescott SM. Many actions of cyclooxygenase-2 in cellulardynamics and in cancer. J Cell Physiol 2000;190:279–86.

7. Garrison JB, Natasha Kyprianou N. Novel targeting of apoptosispathways for prostate cancer therapy. Curr Cancer Drug Targets2004;4:85–95.

8. Gupta S, Srivastava M, Ahmad N, Bostwick DG, Mukhtar H. Over-expression of cyclooxygenase-2 in human prostate adenocarcinoma.Prostate 2000;42:73–8.

9. Yoshimura R, Sano H,MasudaC, KawamuraM, Tsubouchi Y, CharguiJ, et al. Expression of cyclooxygenase-2 in prostate carcinoma.Cancer 2000;89:589–96.

10. Shigemura K, Shirakawa T, Wada Y, Kamidono S, Fujisawa M, GotohA. Antitumor effects of etodolac, a selective cyclooxygenase-II inhib-itor, against human prostate cancer cell lines in vitro and in vivo.Urology 2005;66:1239–44.

11. Roberts RO, Jacobson DJ, Girman CJ, Rhodes T, Lieber MM, Jacob-sen SJ. A population-based study of daily nonsteroidal anti-inflam-matory druguseandprostate cancer.MayoClin Proc2002;77:219–25.

12. Koga S, Luke PP, Specht SM, Rominski B, Jaquins-Gerstl A, HoffmanRA, et al. Inhibition of the allograft response by donor specific bloodtransfusion: association with reduced local TH1 cytokines and nitricoxide but enhanced prostaglandin E2 production. Transplantation2000;70:1788–96.

13. JordanML, Hoffman RA, Debe EF, Simmons RL. In vitro locomotion ofallosensitized T lymphocyte clones in response to metabolites ofarachidonic acid is subset specific. J Immunol 1986;137:661–8.

14. Jain S, Chakraborty G, Raja R, Kale S, Kundu GC. Prostaglandin E2regulates tumor angiogenesis in prostate cancer. Cancer Res2008;68:7750–9.

15. Breyer RM, Bagdassarian CK, Myers SA, Breyer MD. Prostanoidreceptors: subtypes and signaling. Ann Rev Pharmacol Toxicol2001;41:661–90.

16. Sugimoto Y, Narumiya S. Prostaglandin E receptors. J Biol Chem2007;282:11613–7.

17. Sung YM, He G, Hwang DH, Fischer SM. Overexpression of theprostaglandin E2 receptor EP2 results in enhanced skin tumor devel-opment. Oncogene 2006;25:5507–16.

18. Kuo KT, Wang HW, Chou TY, Hsu WH, Hsu HS, Lin CH, et al.Prognostic role of PGE2 receptor EP2 in esophageal squamous cellcarcinoma. Ann Surg Oncol 2009;16:352–60.

19. Chell SD, Witherden IR, Dobson RR, Moorghen M, Herman AA,Qualtrough D, et al. Increased EP4 receptor expression in colorectalcancer progression promotes cell growth and anchorage indepen-dence. Cancer Res 2006;66:3106–13.

20. Robertson FM, Simeone AM, Lucci A, McMurray JS, Ghosh S, Cris-tofanilli M. Differential regulation of the aggressive phenotype ofinflammatory breast cancer cells by prostanoid receptors EP3 andEP4. Cancer 2010;116:2806–14.

21. KambeA, YoshiokaH, Kamitani H,Watanabe T, BaekSJ, Eling TE. Thecyclooxygenase inhibitor sulindac sulfide Inhibits EP4 expression andsuppresses the growth of glioblastoma cells. Cancer Prev Res2009;2:1088–99.

22. Ma X, Kundu N, Rifat S, Walser T, Fulton AM. Prostaglandin E receptorEP4 antagonism inhibits breast cancer metastasis. Cancer Res2006;66:2923–7.

23. Yang L, Huang Y, Porta R, Yanagisawa K, Gonzalez A, Segi E, et al.Host and direct antitumor effects and profound reduction in tumor

metastasis with selective EP4 receptor antagonism. Cancer Res2006;66:9665–72.

24. KotaniM, Tanaka I, OgawaY,Usui T, TamuraN,Mori K, et al. Structuralorganization of the human prostaglandin EP3 receptor subtype gene(PTGER3). Genomics 1997;40:425–34.

25. DongZ,BonfilRD,ChinniS,DengX,TrindadeFilhoJC,BernardoM,et al.Matrix metalloproteinase activity and osteoclasts in experimental pros-tate cancer bone metastasis tissue. Am J Pathol 2005;166:1173–86.

26. Akech J, Wixted JJ, Bedard K, van der Deen M, Hussain S, Guise TA,et al. Runx2 association with progression of prostate cancer inpatients: mechanisms mediating bone osteolysis and osteoblasticmetastatic lesions. Oncogene 2010;29:811–21.

27. Fujino H, Xu W, Regan JW. Prostaglandin E2 induced functionalexpression of early growth response factor-1 by EP4, but not EP2,prostanoid receptors via the phosphatidylinositol 3-kinase and extra-cellular signal-regulated kinases. J Biol Chem 2003;278:12151–6.

28. Pozzi A, YanX,Macias-Perez I,Wei S,HataAN, Breyer RM, et al. Coloncarcinoma cell growth Is associated with prostaglandin E2/EP4 recep-tor-evoked ERK activation. J Biol Chem 2004;279:29797–804.

29. KotaniM, Tanaka I, OgawaY, Suganami T,Matsumoto T,Muro S, et al.Multiple signal transduction pathways through two prostaglandin Ereceptor EP3 subtype isoforms expressed in human uterus. J ClinEndocrinol Metab 1995;85:4315–22.

30. AnS, Yang J, SoSW,ZengL,Goetzl EJ. Isoforms of theEP3 subtypeofhuman prostaglandin E2 receptor transduce both intracellular calciumand cAMP signals. Biochemistry 1994;33:14496–502.

31. Wang W, Bergh A, Damber JE. Cyclooxygenase-2 expression corre-lates with local chronic inflammation and tumor neovascularization inhuman prostate cancer. Clin Cancer Res 2005;11:3250–6.

32. Chen Y, Hughes-Fulford M. Prostaglandin E2 and the protein kinase Apathway mediate arachidonic acid induction of c-fos in human pros-tate cancer cells. Br J Cancer 2000;82:2000–6.

33. Liu XH, Yao S, Kirschenbaum A, Levine AC. NS398, a selectivecyclooxygenase-2 inhibitor, induces apoptosis and down-regulatesbcl-2 expression in LNCaP cells. Cancer Res 1998;58:4245–9.

34. Miyata Y, Kanda S, Maruta S, Matsuo T, Sakai H, Hayashi T, et al.Relationship between prostaglandin E2 receptors and clinicopatholog-ic features in human prostate cancer tissue. Urology 2006;68:1360–5.

35. Terada N, Shimizu Y, Kamba T, Inoue T, Maeno A, Kobayashi T, et al.Identification of EP4 as a potential target for the treatment of castra-tion-resistant prostate cancer using a novel xenograft model. CancerRes 2010;70:1606–15.

36. Huang HFS, Shu P, Murphy T, Jordan ML. Stable EP4 receptorknockdown inhibits prostate cancer growth in vivo [abstract]. In:Proceedings of the 103rd Annual Meeting of the American Associationfor Cancer Research; 2012 April 1–4; Chicago, Illinois. Philadelphia(PA): AACR; 2012.

37. Kabashima K, Sakata D, Nagamachi M, Miyachi Y, Inaba K, NarumiyaS. Prostaglandin E2-EP4 signaling initiates skin immune responses bypromoting migration and maturation of Langerhans cells. Nat Med2003;9:744–9.

38. Pratap J, Javed A, Languino LR, van Wijnen AJ, Stein JL, Stein GS,et al. The Runx2 osteogenic transcription factor regulates matrixmetalloproteinase 9 in bone metastatic cancer cells and controls cellinvasion. Mol Cell Biol 2005;25:8581–91.

39. Kang GH, Lee S, Lee HJ, Hwang KS. Aberrant CpG island hyper-methylation of multiple genes in prostate cancer and prostatic intrae-pithelial neoplasia. J Pathol 2004;202:233–40.

40. Fujino H, Toyomura K, Chen X-B, Regan JW, Murayama T. Prosta-glandin E2 regulates cellular migration via induction of vascular endo-thelial growth factor receptor-1 in HCA-7 human colon cancer cells.Biochem Pharmacol 2011;81:379–87.

41. Amano H, Hayashi I, Endo H, Kitasato H, Yamashina S, Maruyama T,et al. Host prostaglandin E2-EP3 signaling regulates tumor-associatedangiogenesis and tumor growth. J Exp Med 2003;197:221–32.

42. Shoji Y, Takahashi M, Takasuka N, Niho N, Kitamura T, Sato H, et al.Prostaglandin E receptor EP3 deficiency modifies tumor outcome inmouse two-stage skin carcinogenesis. Carcinogenesis 2005;26:2116–22.

Huang et al.





43. Hatae N, Kita A, Tanaka S, Sugimoto Y, Ichikawa A. Induction ofadherent activity in mastocytoma P-815 cells by the cooperation oftwo prostaglandin E2 receptor subtypes, EP3 and EP4. J Biol Chem2003;278:17977–81.

44. Meisdalen K, Olav F, Dajani O, Thoralf Christoffersen T, DagnySandnes D. Prostaglandins enhance epidermal growth factor-Induced DNA synthesis in hepatocytes by stimulation of E prosta-noid 3 and F prostanoid receptors. J Pharmacol Exp Ther 2007;322:1044–50.

45. OgawaY,Suzuki T,OikawaA,HosonoK,KuboH,AmanoH, et al. Bonemarrow-derived EP3-expressing stromal cells enhance tumor-asso-ciated angiogenesis and tumor growth. Biochem Biophys Res Com-mun 2009;382:720–5.

46. Shoji Y, Takahashi M, Kitamura T, Watanabe K, Kawamori T,Maruyama T, et al. Down regulation of prostaglandin E receptor

subtype EP3 during colon cancer development. Gut 2004;53:1151–8.

47. Macias-Perez IM,ZentR,CarmosinoM,BreyerMD,BreyerRM,Pozzi A.Mouse EP3 alpha, beta, and gamma receptor variants reduce tumor cellproliferationand tumorigenesis in vivo. JBiolChem2008;283:12538–45.

48. Sanchez T, Moreno JJ. GR 63799X, an EP3 receptor agonist, inducedS phase arrest and 3T6 fibroblast growth inhibition. Eur J Pharmacol2006;529:16–23.

49. Liu J, Akahoshi T, JiangS,Namai R, KitasatoH, EndoH, et al. Inductionof neutrophil death resembling neither apoptosis nor necrosis byONO-AE-248, a selective agonist for PGE2 receptor subtype 3. J Leuk Biol2000;68:187–93.

50. Kirschenbaum A, Klausner AP, Lee R, Unger P, Yao S, Liu XH, et al.Expression of cyclooxygenase-1 and cyclooxygenase-2 in the humanprostate. Urology 2000;56:671–6.






2013;11:427-439. Published OnlineFirst January 30, 2013.Mol Cancer Res Hosea F. S. Huang, Ping Shu, Thomas F. Murphy, et al. Receptors in Human Prostate CancerSignificance of Divergent Expression of Prostaglandin EP4 and EP3

Updated version

10.1158/1541-7786.MCR-12-0464doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2013/01/30/1541-7786.MCR-12-0464.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/11/4/427.full#ref-list-1

This article cites 48 articles, 19 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/11/4/427.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/11/4/427To 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-12-0464http://mcr.aacrjournals.org/content/suppl/2013/01/30/1541-7786.MCR-12-0464.DC1http://mcr.aacrjournals.org/content/11/4/427.full#ref-list-1http://mcr.aacrjournals.org/content/11/4/427.full#related-urlshttp://mcr.aacrjournals.org/cgi/alertsmailto:[email protected]://mcr.aacrjournals.org/content/11/4/427http://mcr.aacrjournals.org/

Documents

Signiﬁcance of Divergent Expression of Prostaglandin EP4 ...and EP4 (15, 16) which are frequently co-expressed in the same cells but use different and/or opposing signaling path-ways