fractalkine receptor cx cr1 is expressed in epithelial ovarian … · bd biosciences. transwell...

Angiogenesis, Metastasis, and the Cellular Microenvironment

Fractalkine Receptor CX3CR1 Is Expressed in EpithelialOvarian Carcinoma Cells and Required for Motility andAdhesion to Peritoneal Mesothelial Cells

Mijung Kim1, Lisa Rooper2, Jia Xie1, Andre A. Kajdacsy-Balla2, and Maria V. Barbolina1

AbstractEpithelial ovarian carcinoma (EOC) is a deadly disease, and little is known about the mechanisms underlying its

metastatic progression. Using human specimens and established cell lines, we determined that the G-protein–coupled seven-transmembrane fractalkine receptor (CX3CR1) is expressed in primary and metastatic ovariancarcinoma cells. Ovarian carcinoma cells robustly migrated toward CX3CL1, a specific ligand of CX3CR1, in aCX3CR1-dependent manner. Silencing of CX3CR1 reduced migration toward human ovarian carcinoma ascitesfluid by approximately 70%. Importantly, adhesion of ovarian carcinoma cells to human peritonealmesothelial cellswas dependent on CX3CL1/CX3CR1 signaling. In addition, CX3CL1 was able to induce cellular proliferation.Together, our data suggest that the fractalkine network may function as a major contributor to the progression ofEOC, and further attention to its role in the metastasis of this deadly malignancy is warranted. Mol Cancer Res;10(1); 11–24. �2011 AACR.

Introduction

Epithelial ovarian carcinoma (EOC) is a relatively rare buthighly lethal gynecologic malignancy that claims the lives ofmore than 14,000 women in the United States and morethan 140,000 worldwide (1). EOC is seldomly diagnosed atearly stages, when the rate of survival is nearly 90%. Instead,most patients are diagnosed with EOC at late metastaticstages (FIGO stages III–IV; ref. 2). Treatment of advancedand metastatic EOC cases is clinically difficult. The currentstandard of care includes surgery and chemotherapy withplatinum-based drugs (3). However, many EOC casesfrequently recur and become insensitive to these therapies(4, 5). Improvement of the existing approaches and devel-opment of new therapies to be used, either alone or incombination, with the existing treatments are necessary toprolong and improve patient survival.Metastatic EOC spreads throughout the peritoneum large-

ly by sloughing off of the ovary and subsequently attaching toand invading the organs and tissues of the peritoneal cavity.Dissemination throughblood and lymphatic vessels, however,is minimal (2, 6). Malignant EOC cells are shed from the

ovarian surface and later attach to the mesothelial layer out-lining the peritoneum (2, 7–10). This spreading patternsuggests that the local microenvironment of the peritonealcavity may contain signals that support homing of the EOCmetastasis. Patients with EOC frequently accumulate largevolumes (0.5–4 L) of peritoneal ascitic fluid. This fluid is arich source of growth factors, extracellular matrix proteins,chemokines, and other factors that may support and promoteperitoneal metastasis (11–13).The development of new approaches to treat metastatic

ovarian carcinoma and the identification of novel pathwaysfor drug targeting could be bolstered by the identification ofhighly efficacious treatments against molecules involved inmetastasis. The goal of this study was to identify methods toprevent or retard metastatic dissemination. Given thatidentifying new drug target pathways is our priority, wetook a focused approach to ascertain the molecules that aremost critical for promoting metastasis and can be targetedtherapeutically. Chemokine receptors encompass a subclassof G protein–coupled receptors (GPCR), which in turnrepresent a preferred class of drug targets (14). Chemokinereceptors show promise for development as therapeutictargets because they have been successfully targeted in theclinic (15) and play a role in homing of metastatic cells totheir niches (16–19); however, beyond CXCR4, their role inthe development and progression of EOC is largelyunknown (11, 20, 21). It is possible that ovarian cancerantichemokine receptor antagonists could be useful in pre-ventingmetastasis in patients with early-stage disease, as wellas in restricting metastatic spread in patients presenting withadvanced stages of malignancy.We were interested in determining whether chemokine

networks that have not yet been implicated in EOC could be

Authors' Affiliations: Departments of 1Biopharmaceutical Sciences and2Pathology, University of Illinois at Chicago, Chicago, Illinois

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

Corresponding Author: Maria V. Barbolina, Department of Biopharma-ceutical Sciences, University of Illinois at Chicago, 833 S. Wood St.,PHARM 355, Chicago, IL 60612. Phone: 7-312-355-0670; Fax: 312-996-0098; E-mail: [email protected]

doi: 10.1158/1541-7786.MCR-11-0256

�2011 American Association for Cancer Research.

MolecularCancer

Research

www.aacrjournals.org 11

on December 4, 2020. © 2012 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 7, 2011; DOI: 10.1158/1541-7786.MCR-11-0256

http://mcr.aacrjournals.org/

important for progression of this malignancy. Preliminaryreverse transcriptase PCR (RT-PCR) array data in ourlaboratory suggested that CX3CR1, a member of the che-mokine family of GPCRs, is expressed in cell lines propa-gated from acute EOC samples. CX3CR1 is a member of the7-transmembrane CX3C chemokine receptor group family(22). CX3CL1, or fractalkine, is the major, highly specificligand that activates CX3CR1 (23, 24). This robust ligandspecificity suggests that CX3CR1 could be an attractivetarget for therapeutic interventions. Interestingly, it has beensuggested that CX3CL1 can also activate another surfacereceptor, a receptor tyrosine kinase termed epidermalgrowth factor receptor (EGFR), to stimulate cellular prolif-eration (25). It has also been reported that poor prognosesand EOC progression correlate with the accumulation ofCX3CR1-positive immunosuppressive leukocytes (26).Because CX3CR1-positive immunosuppressive leukocytesare attracted to the peritoneal milieu to suppress the nativeimmune system, these findings may suggest that theCX3CR1-positive EOC cells that are also present in theperitoneum may escape clearance by the native immunesystem. Escape from immune surveillance and use ofimmune cell secretions for tumor progression are commonfeatures of metastatic tumors (27–31). However, in coloncancer expression of CX3CL1 correlated with a betterprognosis and associated with high densities of T lympho-cytes (32).Here, we show that CX3CR1 expression is nearly absent in

the normal ovarian surface epithelium and that its expressionaccumulates during the course of tumorigenesis. We alsoreport that CX3CL1 is present in the ascites of women withEOC and benign gynecologic conditions. Furthermore, weshow that a functional interaction between CX3CL1 andCX3CR1 facilitates cell migration and cell–cell adhesionbetween EOC cells and peritoneal mesothelial cells.CX3CL1 is also required for EOC cell proliferation. Becausecell migration, proliferation, and peritoneal seeding arecrucial for the formation and development of EOC metas-tasis, our data suggest that fractalkine signaling in EOCmaycontribute to metastatic progression.

Materials and Methods

Reagents and antibodiesTissue microarray (TMA) slides containing specimens of

normal ovary and primary and distant metastatic ovariancarcinoma (catalog no. OV208), specimens of normal ovary(catalog no. OV806), specimens of serous cystadenoma(catalog no. OV603), and specimens of multiple types ofovarian carcinoma (catalog no. OV1002) were obtainedfrom US Biomax. cDNA arrays prepared from normal ovaryand primary ovarian carcinoma specimens were obtainedfrom Origene Technologies. CX3CL1 was obtained fromRayBiotech. Human collagen type I, type III, bovine serumalbumin (BSA), fura-2 AM, and ATP were purchased fromSigma-Aldrich. WST-1 proliferation kits were obtainedfrom Takara Bio. CX3CR1 siRNA, control siRNA (#A),a goat anti-human CX3CR1 antibody (K-13), a rabbit anti-

human EGFR antibody, a mouse anti-human glyceralde-hyde-3-phosphate dehydrogenase (GAPDH) antibody, arabbit anti-human proliferating cell nuclear antigen (PCNA)antibody, and total cell lysates from THP-1 and A431 wereobtained from Santa Cruz Biotechnology. DharmaFECTwas obtained fromDharmacon. Matrigel was obtained fromBD Biosciences. Transwell cell migration chambers with 8-micron pores were purchased from Corning. AG1478 wasobtained from Cayman Chemicals. Rabbit polyclonalCX3CR1 (catalog no. ab8020 and ab8021) antibodies andnormal rabbit IgGwere obtained fromAbcam. A fluoresceinisothiocyanate (FITC)-conjugated secondary anti-rabbitIgG was purchased from Invitrogen. Mouse monoclonalanti-human CX3CL1 was procured from R&D Systems,and rabbit anti-human CX3CL1 was purchased from BioVi-sion. Mouse monoclonal anti-human CD44 and mousemonoclonal anti-b1–integrin (MAB2253) were obtainedfrom Millipore. Vectastain ABC and 3,30—diaminobenzi-dine (DAB) kits were obtained from Vector Laboratories.

Cell linesTwo human ovarian carcinoma cell lines that originated

from malignant cells in the ascites fluid and presented aserous histotype, SKOV-1 and A2780, were obtained fromthe National Cancer Institute (NCI) Tumor Cell Reposi-tory (Detrick, MD). These cell lines were cultured asdirected by the manufacturer for no longer than 20 conse-cutive passages. Another human ovarian carcinoma cell lineoriginating from malignant cells in the ascites fluid andpresenting a serous histotype, Caov-3, was obtained fromDr. M.S. Stack (University of Missouri-Columbia, MO)and propagated in minimal essential media supplementedby 10%FBS for no longer than 15 consecutive passages. Thehuman ovarian carcinoma cell line with a serous histotypeand originating from a primary tumor, IGROV-1, wasobtained from the NCI Tumor Cell Repository (Detrick,MD) and cultured as directed by the manufacturer for nolonger than 20 consecutive passages. Two borderline EOCcell lines (HuIOSBT-1.5 andHuIOSBT-3.3) and 2 immor-talized human ovarian surface epithelial cell lines (IOSE29and IOSE80)were obtained fromDr.N. Auersperg throughthe Canadian Ovarian Tissue Bank. The immortalizedhuman ovarian surface epithelial cell line HOSE11-12 wasobtained from Dr. G.S.W. Tsao (The University of HongKong, PRC) and propagated in minimal essential mediasupplemented with 10% FBS for 5 consecutive passages.The immortalized human ovarian surface epithelial cell lineT1074 was obtained from Applied Biological Materials andpropagated as directed by the manufacturer. The normal(not immortalized) human ovarian surface epithelial cell lineHOSEpiC was purchased from ScienCell Research Labo-ratories and maintained in the conditions suggested by themanufacturer between passages 1 and 2. The humanimmortalized peritoneal mesothelial cell line LP-9 wasobtained from the Coriell Aging Cell Repository and cul-tured as indicated by the manufacturer for 5 to 8 passages.All cell lines were routinely assessed for cellular morphologyand average doubling time. All cell lines were propagated

Kim et al.

Mol Cancer Res; 10(1) January 2012 Molecular Cancer Research12




from stocks originally obtained from cell banks and indi-vidual investigators and have been stored in aliquots forfuture use. Each aliquot was further propagated for nolonger than 20 consecutive passages or 4 months, which-ever came first.

Real-time PCRReal-time PCR was carried out using the ABI Prizm

system (Applied Biosystems) according to the manufac-turer's instructions. The following primers were used fordetection of CX3CR1 mRNA in normal ovary and primaryEOC samples (obtained from Origene Technologies):CX3CR1FORW 50-CACAAAGGAGCAGGCATGGAA-G-30; CX3CR1REV 50-CAGGTTCTCTGTAGACACA-AGGC-30. Primers for beta-actin detection (ACTNB) weresupplied with the cDNA sample panel (Origene). The iTaqSYBR Green Supermix (Bio-Rad) was used for quantitativePCR as a double-stranded DNA-specific fluorophore. ThePCR reaction conditionswere as follows: initial denaturationfor 10 minutes at 95�C followed by 40 cycles of 94�C for 15seconds and 60�C for 30 seconds. To determine the spec-ificity of the PCR primers, melting curves were collected byheating the products to 95�C, cooling them to 65�C, andslowly melting them by increasing the temperature 0.5�/secup to 95�C. Primers for mRNA detection of our genes ofinterest were designed using the Primer3 software accordingto the requirements for real-time RT-PCR oligonucleotideprimers.

ImmunohistochemistryTMA slides were deparaffinized by baking at 60�C for 2

hours and rehydrated by incubation in xylene, 100% eth-anol, 95% ethanol, 70% ethanol, and PBS, pH 7.4, for 5minutes each. Peroxidase activity was inhibited with H2O2.Antigen retrieval was carried out by a 15-minute incubationin 1 mmol/L EDTA (pH 8.0) at 95�C. Prior to primaryantibody staining [1:50 dilution (Abcam catalog no.ab8021), 1 hour at room temperature] nonspecific bindingwas blocked by incubation with 10% goat serum for 1 hour.The biotin-conjugated goat anti-rabbit secondary antibodywas used at a 1:200 dilution for 30 minutes at roomtemperature. The Vectashield ABC kit was used as directedby the manufacturer, and tissues were incubated for 45minutes at room temperature. The DAB reagent was pre-pared as instructed by the manufacturer and applied totissues on TMA slides for 2 to 10 minutes until browncolor developed. TMAs were stained with hematoxylin,dehydrated in 50%, 70%, 95%, and 100% ethanol, andmounted with Permount. Pancreatic cancer tissue was usedas a positive control. Staining was evaluated by A.A.K.-B.and L.R., who were both blinded to the experimental out-comes of the study. Staining was scored on the basis of theintensity and percentage of positive cells. Intensity of stain-ing was "0" for negative samples, "1" for weakly positivesamples, "2" for moderately positive samples, and "3" forhighly positive samples. Overall scores were derived as theintensity score multiplied by the percentage. Staining wasassessed separately in the cytoplasm and the membrane.

Flow cytometryCells (1 � 106 per tube) were harvested with trypsin/

EDTA, resuspended in 100 mL of ice-cold PBS supplemen-ted with 10% fetal calf serum and 1% sodium azide, andfixed in a 1% paraformaldehyde solution in PBS for 15minutes on ice. Two micrograms of rabbit anti-humanCX3CR1 antibody (Abcam, ab8021) was added to the cells.For negative controls, the cells were incubated with either 2mg of anti-rabbit IgG antibody or no primary antibody. Cellswere incubated for 1 hour on ice in the dark with agitationfollowing washing and resuspension in 400 mL ice-cold PBS.Two micrograms of goat anti-rabbit FITC-conjugated IgG(Invitrogen) was added to the cells and incubated for 1 houron ice in the dark with agitation. The cells were washed andresuspended in 400 mL of ice-cold PBS supplemented with2% BSA and 1% sodium azide. Labeled cells were analyzedusing a BectonDickinson LSR I flow cytometer or an AccuriC6 flow cytometer on the same day.

Western blottingWestern blotting analysis was used to detect the expres-

sion of CX3CR1 and b-tubulin in normal (not immor-talized) ovarian surface epithelial cells, immortalized ovar-ian surface epithelial cells, and EOC cell lines. Thisprocedure was carried out as previously described (33–35). Antibodies were used at the following dilutions:1:500 rabbit anti-human–CX3CR1 (Abcam, ab2080) ina 1� non-animal blocker solution (Vector Laboratories)or 1:200 goat anti-human CX3CR1 (Santa Cruz Biotech-nology, K-13) in 3% BSA in a solution of 50 mmol/LTBS, pH 7.4, 150 mmol/L NaCl, and 0.05% Tween-20(TBST; Sigma), 1:200 mouse anti-human GAPDH(Santa Cruz Biotechnology) and 1:200 mouse anti-human–b-tubulin (Developmental Studies HybridomaBank, Iowa) in 3% BSA in TBST. Immunoreactive bandswere visualized with an anti-(rabbit-IgG)-peroxidase, anti-(goat-IgG)-peroxidase, or anti-(mouse-IgG)-peroxidase(Sigma; 1:1,000 in 3% BSA in TBST), and enhancedchemiluminescence was read using Chemidoc (Bio-Rad)and Bio-Rad Chemidoc ImageReader software.

Caþþ imagingIntracellular free calcium concentrations were measured

using digital video microfluorimetry as previously described(11). Briefly, SKOV-3 cells were plated on collagen I–coatedglass coverslips, rinsed with HEPES buffer, exposed to 10mmol/L fura-2 AM (Molecular Probes), and then incubatedfor 30 minutes at room temperature. Complete dye deester-ification was achieved by removing the fura-2 AM, rinsingand maintaining the cells in the dark for 30 minutes. Glasscoverslips were mounted on the stage of a Zeiss AxioOb-server inverted epifluorescence microscope equipped fordigital fluorescence microscopy. Fluorescence was digitallymonitored at 520 nm after excitation at 340 nm (boundCa2þ) and 380 nm (free Ca2þ). Ratios of F340/F380 werecollected before and during treatment with 20 nmol/LCX3CL1 using Zeiss software, exported in Excel softwareand plotted.

CX3CR1 in Ovarian Carcinoma

www.aacrjournals.org Mol Cancer Res; 10(1) January 2012 13




Immunofluorescence stainingCells were cultured on glass coverslips to 50% to 70%

confluence, fixed, permeabilized with 0.01% Triton X-100and blocked in goat serum. Rabbit anti-human-CX3CR1(Abcam, ab8021), rabbit anti-human-CX3CL1 (BioVision),and rabbit anti-human PCNA (Santa Cruz Biotechnology)antibodies were used at a dilution of 1:50. Cells wereincubated with the primary antibodies for 1 hour at roomtemperature (22�C). The cells were then incubated withsecondary Alexa 488–conjugated or Alexa 555–conjugatedanti-rabbit antibodies (1:500) for 1 hour at room temper-ature in the dark. 40,6-Diamidino-2-phenylindole (DAPI)was added to the secondary antibody solution to a finalconcentration of 10 mg/mL 10 minutes prior to the end ofthe incubation period. Cells were washed, air dried, andmounted on glass slides using ProlongGold (Invitrogen).Fluorescent imaging was done using a Zeiss AxioObser-verD.1 fluorescent microscope.

Transient transfectionsEOC cells were cultured to 80% confluence and trans-

fected with siRNAs using DharmaFECT (ThermoScienti-fic) according to the manufacturer's instructions.

Transwell cell migrationInserts with 0.8-micron porous membranes were bottom-

coated with Matrigel (diluted 1:100) for 1 hour at 37�C,rinsed, and dried. Next, 5,000 EOC cells in a final volume of300 mL were seeded onto the inserts. The inserts were placedin 24-well plates filled with serum-free minimal essentialmedia in the presence or absence of CX3CL1. Cells wereallowed to migrate for 5 hours at 37�C and 5% CO2.Migration was terminated by removing the nonmigratedcells from the inside of the inserts. Cells that did migratethrough the membranes were fixed, stained, and counted.

ELISAAscites samples from 6 patients with benign (n ¼ 3) or

malignant (n ¼ 3) ovarian disease were obtained fromProteoGenex and analyzed for CX3CL1 using HumanCX3CL1 Quantikine ELISA (R&D Systems) according tothe manufacturer's instructions.

Cell–cell adhesionThe human-derived peritoneal mesothelial cell line LP-9

was cultured in 96-well plates to near confluence. SKOV-3cells were cultured in monolayers and labeled with fluores-cent DiO (Invitrogen) according to the manufacturer'sinstructions prior to the adhesion assays. They were subse-quently released from monolayers with trypsin and resus-pended in serum-free cell culture media. When needed,SKOV-3 cells were transiently transfected with either con-trol or CX3CR1-specific siRNAs and used in adhesion assaysbetween 48 and 72 hours from the start of transfection. Forthe assay, 10,000 DiO-labeled SKOV-3 cells were seededonto LP-9 monolayers (in sextuplicate for each condition)and allowed to adhere for 5 hours at 37�C and 5% CO2.Subsequently, the monolayers were washed 2 times with

PBS and fixed in a methanol-containing cell fixative. Whenrequired, LP-9 cells were pretreated with CX3CL1 blockingantibodies for 30minutes, and SKOV-3 cells were pretreatedwith either normal anti-rabbit IgG, anti-CD44, or anti-b1–integrin blocking antibodies for 30minutes at 37�C and 5%CO2. Adherent cells showing a round cell morphology werevisualized by green fluorescent signals using a Zeiss fluores-cent microscope. The DiO-labeled cells were then counted,averaged, and characterized as a percentage of adherent cells.

Cell–ECM adhesionTissue culture–treated 48-well plates were precoated with

10mg/mLhuman collagen type I, 10mg/mLhuman collagentype III, Matrigel (diluted 1:100), or 1% BSA for 1 hour at37 �C. The plates were subsequently rinsed with PBS and airdried. Next, 10,000 SKOV-3 cells were seeded (in triplicatefor each condition) in coated wells and allowed to adhere for1 hour at 37�C and 5%CO2. This seeding was followed by 2washes with PBS, fixation in a methanol-containing cellfixative, and staining. Cells were counted, averaged, andplotted.

Cell proliferationCells were seeded in tissue culture–treated 96-well

plates and allowed to attach for 6 hours following over-night serum starvation (0% serum) at 37�C in 5% CO2.Cells were plated at a density that yielded confluencebelow 25% at the beginning of the treatment. Cells wereseeded in triplicate for each condition. To assess theproliferative effect of CX3CL1 on EOC cells, cells weretreated with 5, 15, 25 nmol/L CX3CL1, or 25 nmol/LCX3CL1 with 5 mmol/L AG1478 for 24 hours. Cellstreated with complete media containing 10% FBS wereused as a positive control. Cells treated with serum-freemedia served as a negative control. WST-1 was usedaccording to the manufacturer's instructions and incubat-ed with the cells for 1 hour following measurement ofOD440. A baseline OD440 reading of the mixture of mediaand WST-1 containing no cells was subtracted from theexperimental values.

Results

Expression of CX3CR1 in normal and pathologic ovariansurface epitheliaThe mechanisms of malignant transformation and met-

astatic progression of EOC are poorly understood. Currenthypotheses have proposed the ovarian surface epitheliumand the fimbriated edge of the fallopian tube as putativecellular origins of the carcinoma (36, 37). It has recently beenhypothesized that these 2 epithelia are part of a transitionalepithelium that shares a common origin and is prone toneoplastic transformation due to incomplete differentiation(38). Another unique feature of EOC that is particularlypertinent to this study is the pattern of metastatic dissem-ination, which is largely intraperitoneal. The chemokine axishas been associated with homing ofmalignant cells into theirmetastatic niches. We hypothesized that the fractalkine axis

Kim et al.





could play a role in the peritoneal dissemination of EOC.Therefore, we evaluated CX3CR1 expression in normal andpathologic ovarian surface epithelia.CX3CR1 expression in normal ovarian surface epithe-

lium was assessed by 4 methods. Immunohistochemistrywas used to detect CX3CR1 protein expression in normalovary specimens. CX3CR1 expression was also measuredin normal ovarian surface epithelial cells (HOSEpiC)using immunofluorescence staining, flow cytometry, andWestern blot. The majority of the normal epithelial cells(21 cases of 22) were negative for membranous CX3CR1expression (Fig. 1A, Supplementary Table S1). Further-more, 77% of the tested samples (17 of 22) lackedCX3CR1 cytoplasmic staining (Supplementary TableS1). Samples were considered negative if the total scorewas below 100, weakly to moderately positive if the scorewas between 100 and 200, and strongly positive if thescore was above 200. CX3CR1 expression was not detect-able in HOSEpiC cells by immunofluorescence staining,flow cytometry, or Western blot (Fig. 1B and C, Fig. 3A).Next, we evaluated CX3CR1 expression in immortalized

ovarian surface epithelium. Interestingly, all of the immor-talized ovarian surface epithelial cell lines tested (T1074,HOSE11-12, IOSE29, and IOSE80) were CX3CR1 posi-tive, as shown by flow cytometry and Western blot analysis(Fig. 2A, Fig. 3A). These findings suggest that induction ofCX3CR1 expression can occur very early in the tumorigenictransformation process.Subsequently, we evaluated CX3CR1 expression in

specimens of a benign gynecologic condition, ovarianserous cystadenoma, and established cell lines originatingfrom cases with borderline EOC. Flow cytometry alsodetected positive membranous expression of CX3CR1 inthe borderline EOC cell lines HuIOSBT-1.5 andHuIOSBT-3.3 (Fig. 2B). Immunohistochemical analysisof 30 cases of ovarian serous cystadenoma revealed that83% of the specimens were CX3CR1 positive in theepithelium (Fig. 2C, Supplementary Table S1). A small

number of serous cystadenoma cases also showed extensiveCX3CR1 immunoreactivity. Because the clinical progno-sis for these cases is excellent, the significance of thisfinding is unknown.Furthermore, we utilized RT-PCR and immunohis-

tochemistry to test the RNA and protein expression levelsof CX3CR1 in primary ovarian carcinoma specimens ofvarious histotypes. Our findings revealed that 66% and 50%of the primary EOC cases with a serous histotype wereCX3CR1 positive by RT-PCR (19 of 29) and immunohis-tochemistry (12 of 24), respectively (Supplementary TablesS1 and S2, Fig. 3). Among mucinous EOC cases, 3 of 3 and13 of 24 cases were CX3CR1 positive as shown by RT-PCRand immunohistochemistry, respectively (SupplementaryTables S1 and S2, Fig. 3D). RT-PCR analysis identified5 CX3CR1-positive cases of endometrioid EOC of 8 tested(Supplementary Table S2). Notably, nearly all of the casesthat received a negative classification by immunohistochem-istry did contain some CX3CR1-positive cells (Supplemen-tary Table S1).Ovarian carcinoma cell lines derived from primary EOC

(IGROV-1) or ascites (Caov-3, SKOV-3, A2780) all showedCX3CR1 expression, as shown by Western blot, immuno-fluorescence staining, and flow cytometry (Fig. 3). Quan-tification by ImageJ software revealed that the intensity ofthe CX3CR1 signal in these samples was significantly greaterthan the background expression observed in pseudopodialprotrusions of the SKOV-3 EOC cell line (Fig. 3A, graph).SKOV-3 cells responded to the addition of CX3CL1 by arobust release of intracellular calcium, which is indicative ofthe ability of CX3CR1 to support downstream signaling inthese cells (Fig. 3B).We tested specimens of metastatic EOC by immunohis-

tochemistry and found that 64% of the specimens (7 of 11)were CX3CR1 positive (Fig. 3D, Supplementary Table S1).However, nearly all of the samples that were considerednegative by immunohistochemistry did contain someCX3CR1-positive cells (Supplementary Table S2).

HOSEpiC

DAPICX3CR1 300

0

100

A B C

104FITC

Cou

nts

Figure 1. CX3CR1 is not expressed in normal ovarian surface epithelium. A, samples from normal ovaries were examined for CX3CR1 expression byimmunohistochemistry. The black arrow indicates theovarian surfaceepithelium. Brown–CX3CR1; blue–hematoxylin. Imagesweregeneratedusing anAperioScanScope digital slide scanner. Magnification–20�. B, expression of CX3CR1 in normal ovarian surface epithelial cells (HOSEpiC) was evaluated byimmunofluorescent staining. CX3CR1–green; DAPI–blue. Images were collected on the green and blue filters independently and then superimposed. Bar, 50mm.C, cell surface expression of CX3CR1 in HOSEpiCwas analyzed by flow cytometry. CX3CR1 expression in HOSEpiC and negative controls (normal rabbitIgG plus FITC-conjugated goat anti-rabbit IgG and FITC-conjugated goat anti-rabbit IgG only) is shown. The results are representative of 3 independentexperiments.






Together, our data suggested that CX3CR1 expressionmay be induced very early in the course of the tumorigenictransformation of the ovarian surface epithelium.

CX3CL1 is present in the ascites of patients withborderline gynecologic diseases and EOCCX3CL1 is expressed in many cells and tissues, including

the small and large intestine, ovary, and kidney (23, 24, 39,40). CX3CL1 exists in either a membrane-inserted form(23, 24) or a soluble form that has been cleaved to facilitaterelease (41). Peritoneal ascites is a rich source of manycompounds, including growth factors, components of theextracellular matrix, and chemokines. Soluble CX3CL1 ispresent in the low nanomolar range in the ascites of womenwith endometriosis (42). We therefore measured CX3CL1levels in the ascites of women with EOC and women withborderline gynecologic conditions. By ELISA, we foundthat the average concentration of CX3CL1 in the ascitesfrom stages III–IV EOC patients is 2.8 nmol/L. The levelis approximately 10-fold lower, 0.18 nmol/L, in the ascites

collected from patients with borderline gynecologic dis-eases (Table 1).

EOC cells migrate in a CX3CL1/CX3CR1-dependentmannerCell migration is essential for metastatic dissemination.

Chemokine networks and CX3CL1/CX3CR1, in partic-ular, have been implicated in the chemotactic migration oftumor cells to distant sites. The presence of solubleCX3CL1 in the ascites fluid may indicate that this chemo-kine could serve as an attractive force that guides EOCcells to the peritoneum. Therefore, we tested the ability ofCX3CL1 to attract EOC cells in a Transwell cell migrationassay. The addition of 5 nmol/L CX3CL1 significantly androbustly stimulated the migration of both Caov-3 andSKOV-3 cells by 1.7- and 2.2-fold, respectively (Fig. 4A).Silencing of CX3CR1 with specific siRNAs reduced theability of SKOV-3 cells to migrate in response toCX3CL1, to a level that was nearly identical to the control(no CX3CL1; Fig. 4A). We previously reported a 1.5-fold

HOSE11-12

48%T1074

59%IOSE29

68%

IOSE80

43%

HuIOSBT1.5

73%

HuIOSBT3.3

48%

A

B

Weakly positiveNegative

Strongly positiveModerately positive

C

Figure 2. CX3CR1 is expressed in immortalized andborderline EOC cell lines as well as human serouscystadenoma specimens. Cell surface expression ofCX3CR1 in immortalizedOSE (A) and borderline EOCcells lines (B) was analyzed by flow cytometry. Thepercentage of positive cells is indicated. Solid thickline–CX3CR1 expression; solid thin line–negativecontrol (normal rabbit IgG and FITC-conjugated goatanti-rabbit IgG); dotted line–negative control (FITC-conjugated goat anti-rabbit IgG only). These data arerepresentative of at least 3 independentexperiments. C, expression of CX3CR1 incases of benign serous cystadenoma wasdetermined by immunohistochemistry. Brown–CX3CR1; blue–hematoxylin. Images were generatedusing an Aperio ScanScope digital slide scanner.Magnification–10�. Examples of CX3CR1-negative(core C7, Supplementary Table S3), weakly positive(core B4, Supplementary Table S3), moderatelypositive (core D2, Supplementary Table S3), andstrongly positive (core E3, Supplementary Table S3)specimens are presented.

Kim et al.





increase in CXCL12/CXCR4-mediated invasion of EOCcells in the presence of CXCL12 (11). The CXCL12/CXCR4 axis is a well-established driver of metastaticdissemination in EOC (11, 20, 21); in comparison, theCX3CL1/CX3CR1 pathway makes an even greater con-tribution to cell migration.We next tested the ability of SKOV-3 cells with silenced

CX3CR1 to migrate toward human ascitic fluid obtained

from a patient with stage IV EOC (specimen #5, Table 1).Notably, cells with silenced CX3CR1 displayed a 70%reduction in their ability to migrate compared with controlcells expressing CX3CR1 (Fig. 4B). CX3CR1 siRNA treat-ment of SKOV-3 cells reduced the total CX3CR1 expressionto 70% (Fig. 4B).CX3CR1-negative HOSEpiC cells did not display altered

migratory ability in the presence of CX3CL1 (Fig. 4C).

A

CX3CR1 DAPI

SKOV-3 0

50

Inte

nsit

y

0

1,000

[Ca

++

]I n

mo

l/L

CX3CL1

ATP

C

D

B

SKOV-3

20%

Caov-3

23%IGROV-1

59%

A2780

71%

Mucinous

positiveSerous

positive

Mesentery metastasis

positive

Mucinous

negativeSerous

negativeRectum metastasis

negative

50 CX3CR1

GAPDH

Figure 3. CX3CR1 is expressed in primary and metastatic EOC. A, CX3CR1 expression in normal ovarian surface epithelial cells (HOSEpiC), immortalizedovarian surface epithelial cells (T1074), and serous EOC cell line (IGROV-1) was evaluated by Western blot. GAPDH served as a loading control. THP-1 celllysate was used as a positive control for CX3CR1 expression. Position of the marker lane of apparent molecular weight of 50 kDa is indicated. Expression ofCX3CR1 in the serous EOC cell line (SKOV-3) was evaluated using immunofluorescence staining. CX3CR1–green; DAPI–blue. Images were collectedindependently on the green and blue filters and subsequently superimposed. Bar, 10 mm. Histogram shows the intensity of CX3CR1 staining across thepseudopodial protrusions of SKOV-3 cells (white dotted line and red arrow). B, intracellular [Caþþ] changes in response to 20 nmol/L CX3CL1 indicate thepresenceof activeCX3CR1 in EOCcells. AnATP-inducedsignal indicates live cells. C, cell surfaceexpression ofCX3CR1 in the indicated serousEOCcell lineswas analyzed by flow cytometry. The percentage of positive cells is indicated. Solid thick line–CX3CR1 expression; solid thin line–negative control (normalrabbit IgG and FITC-conjugated goat anti-rabbit IgG); dotted line–negative control (FITC-conjugated goat anti-rabbit IgG only). These data are representativeof at least 3 independent experiments. D, expression of CX3CR1 in cases of primary andmetastatic EOCwas determined by immunohistochemistry. Brown–CX3CR1; blue–hematoxylin. Imageswere generated using anAperio ScanScope digital slide scanner. Magnification–10�. Examples of CX3CR1-positive andCX3CR1-negative cases of mucinous adenocarcinoma (cores C10 and C5, respectively, Supplementary Table S1), serous carcinoma (cores I15 and G9,respectively, Supplementary Table S1), and metastatic serous carcinoma (cores J8 and L3, respectively, Supplementary Table S1) are displayed.






Interestingly, CX3CR1-positive T1074 cells did not migratesignificantly faster toward CX3CL1 than controls (Fig. 4C).Together, these results show that CX3CR1-positive EOCcells show an increased migration ability when stimulatedwith CX3CL1 in vitro, whereas immortalized ovarian surfaceepithelium does not.

EOC cells adhere to human peritoneal mesothelial cellsvia the CX3CL1/CX3CR1 axisEOC metastatic dissemination occurs largely in the

peritoneal cavity. Several molecular interactions, includingintegrin–ECM, CD44–hyaluronan, CA125–mesothelin,and L1–neuropilin-1 have been shown by both in vitroand in vivo models to facilitate peritoneal seeding of EOCcells. The CX3CL1/CX3CR1 axis has also been shownto promote cell–cell adhesion via the membrane-boundform of CX3CL1 (22). Thus, we hypothesized that theCX3CL1/CX3CR1 interaction could foster EOC–meso-thelial adhesion and, thereby, facilitate metastatic seed-ing. Prior to confirming this interaction, we determined

Table 1. ConcentrationofCX3CL1 in samplesofovarian carcinoma and borderline gynecologicdiseases determined by ELISA

Sample CX3CL1

# Diagnosis ng/mL nmol/L

1 Borderline endometrioidcystadenoma

2.50 0.29

2 Borderline mucinouscystadenoma

0.86 0.10

3 Serous cyst 1.25 0.154 Serous adenocarcinoma,

Stage IV, Grade 320.00 2.30

5 Serous adenocarcinoma, Stage IV 25.00 3.006 Serous adenocarcinoma,

Stage IIIc, Grade 325.00 3.00

0

1

2

3

Mig

rati

on

Caov-3

Caov-3 + CX3CL1

SKOV-3

SKOV-3 + CX3CL1

SKOV-3 CTRLsi

SKOV-3 CTRLsi + CX3CL1

SKOV-3 CX3CR1si

SKOV-3 CX3CR1si + CX3CL1

*

**

0 0

1

2

Mig

rati

on

SKOV-3

SKOV-3 CTRLsi

SKOV-3 CX3CR1si

SKOV-3SKOV-3 CTRLsiSKOV-3 CX3CR1si

*

A

B CHOSEpiC

HOSEpiC + CX3CL1

T1074

T1074 + CX3CL1

CX

3C

R1 s

i C

TR

Lsi S

KO

V-3

NT Ctrlsi CX3CR1si

CX3CR1

GAPDH

*

0

1

2

Mig

rati

on

CX

3C

R1

ex

pre

ss

ion

1.5

1

0.5

Figure4. CX3CL1/CX3CR1-dependent cellmigration. A, twoEOCcell lines,Caov-3 andSKOV-3,were either not transfected, transfectedwith control siRNA, ortransfected with CX3CR1 siRNA and then subjected to a cell migration assay in a Transwell chamber in the presence or absence of 5 nmol/L CX3CL1 for 5hours. B, SKOV-3 cells were either not transfected or transiently transfected with control or CX3CR1 siRNAs and then subjected to a Transwell cell migrationassay using human peritoneal ascites fluid (specimen #5, respectively, Table 1). Images on the right illustrate the numbers of migrated cells in each condition.Downregulation of CX3CR1 by siRNA in SKOV-3 cells was analyzed by Western blot and quantified by digital densitometry. GAPDH served as a loadingcontrol. �,P <0.05. C, normal ovarian surface epithelial cells (HOSEpiC) and immortalized ovarian surface epithelial cells (T1074) were allowed tomigrate for 5hours in thepresenceor absenceof 5nmol/LCX3CL1 inaTranswell chamber. Thenumberof cells thatmigrated in thecontrol conditionwithout addedCX3CL1was set as "1", and other values were calculated relative to this level. These data represent the average of 3 experiments carried out in triplicate. �, P < 0.05.

Kim et al.





CX3CL1 expression in human-derived peritoneal meso-thelial LP-9 cells. CX3CL1 was expressed by the LP-9 cellsand present on the cell surface (Fig. 5A). Because mem-brane-anchored CX3CL1 is present on mesothelial cells,and CX3CR1 is expressed on EOC cells, this interaction isfeasible. To test the role of CX3CL1/CX3CR1 in EOC–mesothelial cell adhesion, we carried out in vitro adhesionassays by coculturing EOC and mesothelial cells. LP-9cells were cultured to near complete monolayers. DiO-labeled EOC cells SKOV-3 were seeded on top of the LP-9 monolayers. We found that 90% to 93% of SKOV-3cells adhered to the LP-9 monolayer (Fig. 5B, images 1–2,Supplementary Table S3). Interestingly, siRNA-mediateddownregulation of CX3CR1 resulted in an approximately50% loss of EOC cell adhesion to the LP-9 cells (Fig. 5B,image 3, Supplementary Table S3). Moreover, progressiveinhibition of CX3CL1 by pretreating LP-9 monolayerswith increasing doses of anti-CX3CL1 antibodies (0.1–0mg/mL) prior to the addition of EOC cells resulted in a50% to 83% decrease in the ability of EOCs to adhere(Fig. 5B, images 5–7, Supplementary Table S3). Additionof a nonspecific IgG did not affect adhesion (Fig. 5B,image 4, Supplementary Table S3). In control experi-ments that tested the consequences of combined inhibi-tion by anti-CD44 and anti-b1–integrin antibodies, theability of SKOV-3 cells to adhere to LP-9 cells wasreduced by 40% to 50% (Fig. 5B, images 8–9, Supple-mentary Table S3). Downregulation of CX3CR1 or theaddition of CX3CL1 blocking antibodies to CD44- andb1–integrin blocking antibody treatment further exacer-bated the combined effect and resulted in a nearly 70%reduction in SKOV-3–LP-9 adhesion (Fig. 5B, panels10–11, Supplementary Table S3).These data suggested that the CX3CL1/CX3CR1 axis

plays a significant role in the adhesion of EOC cells tomesothelial cells.

The CX3CL1/CX3CR1 pathway plays a small role infacilitating EOC cell adhesion to MatrigelAfter adhesion to mesothelial cells, EOC cells disrupt the

cellular monolayer, expose extracellular proteins of thebasement membrane, and further invade the submesothelialmatrix, which is primarily composed of interstitial collagenstype I and III (43). Chemokine networks and the CX3CL1/CX3CR1 axis, in particular, have been shown to promotecell–ECM adhesion (44, 45). Therefore, we hypothesizedthat this axis could play a role in EOC adhesion to proteins ofthe submesothelial matrix. To address this question, wetested the ability of SKOV-3 cells to adhere to Matrigel,collagen I, and collagen III. Surprisingly, addition ofCX3CL1 resulted in a slight, but significant, concentra-tion-independent 1.5-fold increase in cell adhesion toMatri-gel but not collagens I or III (results not shown). Thesefindings indicate that CX3CL1/CX3CR1 plays a small role inpromoting ECM adhesion in the initial steps of EOC cellinteraction with the proteins of the basement membrane,but it is unlikely to affect adhesion to the deeper layerscomposed of the collagenous matrix.

CX3CL1 is a potent inducer of EOC cell proliferationChemokines are effective inducers of cell proliferation

(46). Proliferation plays a prominent role in the virulence ofovarian carcinoma. Metastatic progression of EOC is char-acterized by a widespread metastasis that far exceeds the sizeof the primary tumor. Furthermore, the size of the residualtumor after debulking surgery plays a significant role indetermining patient survival (47). Our data and previouslypublished results show that high levels of CX3CL1 arepresent in the peritoneal milieu in the ascites of patientswith various malignant and nonmalignant gynecologic dis-eases (Table 1; ref. 42). Therefore, we hypothesized thatCX3CL1 could induce the proliferation of EOC cells. Wetested CX3CL1 in a range of concentrations between 5 and25 nmol/L and found that the majority of immortalizedOSE, borderline, and serous EOC cell lines showed asignificant proliferative response to CX3CL1 stimulation.In several cell lines, including T1074, HuIOSBT-1.5,SKOV-3, and IGROV-1, the addition of CX3CL1completely rescued cellular proliferation to a level equal toor higher than the rate of cells stimulated with completemedia containing 10% FBS (Fig. 6A and B). We alsoobserved a strong induction of nuclear PCNA 24 hoursafter the addition of 25 nmol/L CX3CL1 to serum-starvedSKOV-3 cells (Fig. 6B). However, some cell lines, includingthe immortalized cell line IOSE80 and the borderline EOCcell line HuIOSBT-3.3, showed only slight changes inproliferation in response to CX3CL1 (Fig. 6A). CX3CL1can partner with both EGFR and CX3CR1 to induce cellproliferation (25, 40). Therefore, to assess the contributionof EGFR toCX3CL1-dependent proliferation, we incubatedeach of our cell lines with the EGFR inhibitor, AG1478.Interestingly, CX3CL1-dependent proliferation of all EOCcell lines tested was either inhibited completely or signifi-cantly reduced in the presence of AG1478 (Fig. 6A). For amore detailed understanding of the role of both CX3CR1and EGFR in CX3CL1-induced proliferation of EOC cells,the receptors were transiently downregulated using specificsiRNAs. A proliferation assay revealed that a 70%decrease inCX3CR1 led to a nearly 50% reduction in CX3CL1-depen-dent proliferation. Moreover, a 95% decrease in EGFRresulted in a nearly 90% reduction in CX3CL1-dependentproliferation (Fig. 6C).

Expression of CX3CR1 in multiple types of ovariancarcinomaEOC is the most predominant type of ovarian carcinoma.

However, ovarian tumors can also arise in stromal cells.Thus, to comprehensively evaluate CX3CR1 expression inovarian carcinoma, we evaluated specimens of Krukenbergtumors, yolk sac tumors, dysgerminomas, mature teratomas,Sertoli-stromal cell tumors, granular cell tumors, thecomas,and gynandroblastomas. Our data indicate that the majorityof specimens were CX3CR1 positive (Supplementary Fig.S1, Supplementary Table S4). Specifically, 7 of 10 Kruken-berg, 8 of 8 yolk sac, 8 of 8 dysgerminoma, 13 of 13 matureteratomas, 1 of 1 Sertoli-stromal cell, 13 of 15 granular cell,14 of 14 thecomas, and 1 of 1 gynandroblastoma tumor






CX3CL1

DAPI

Merge

A

LP-9

B 1 2 3 4

5 6 7 8

9 10 11

0

10

20

30

Inte

nsit

y

1009080706050403020100

1 2 3 4 5 6 7 8 9 10 11

Adh

esio

n, %

Figure 5. EOC cells adhere to peritoneal mesothelial cells and Matrigel in a CX3CL1/CX3CR1-dependent manner. A, immunofluorescence analysis ofCX3CL1 expression in LP-9 cells. The rabbit anti-human CX3CL1 antibody (BioVision) was used at a dilution of 1:50, and the goat anti-rabbit Alexa Fluor546 antibody (Molecular Probes) was used at a dilution of 1:500. Nuclei were counterstained with DAPI. CX3CL1–red; DAPI–blue. Images werecollected independently with red and blue filters and subsequently superimposed. Bar, 5 mm. The insert shows a 1.5-fold enlarged rectangular areaoutlined by dotted lines. White arrows indicate membranous CX3CL1 staining. The histogram shows the intensity of CX3CL1 staining across themembrane of an LP-9 cell (along the green dotted line). B, SKOV-3 cell adhesion to LP-9 monolayers is shown. LP-9 cells were cultured in monolayers.SKOV-3 cells were labeled with DiO, a fluorescent tracking dye, prior to the adhesion assays. Adhesion assays were carried out over 5 hours. Adheredcells were fixed with 4% paraformaldehyde, and images were taken using green fluorescence and bright field microscopy with a 5� magnificationobjective and a Zeiss fluorescent microscope. The images were then superimposed. Green–SKOV-3; gray–LP-9. The following conditions wereevaluated (1) SKOV3 adhesion to LP-9 cells; (2) control siRNA-transfected SKOV-3 cell adhesion to LP-9 cells; (3) CX3CR1 siRNA-transfected celladhesion to LP-9 cells; (4) SKOV-3 cell adhesion to LP-9 cells pretreated with IgG; (5) SKOV-3 cell adhesion to LP-9 cells pretreated with 0.1 mg/mLCX3CL1 blocking antibodies; (6) SKOV-3 cell adhesion to LP-9 cells pretreated with 1 mg/mL CX3CL1 blocking antibodies; (7) SKOV-3 cell adhesion toLP-9 cells pretreated with 10 mg/mL CX3CL1 blocking antibodies; (8) SKOV-3 cell adhesion to LP-9 cells pretreated with 1 mg/mL CD44 blockingantibodies and 1 mg/mL b1-integrin blocking antibodies; (9) control siRNA-transfected SKOV-3 cell adhesion to LP-9 cells pretreated with1 mg/mL CD44 blocking antibodies and 1 mg/mL b1-integrin blocking antibodies; (10) CX3CR1 siRNA-transfected SKOV-3 cell adhesion to LP-9 cellspretreated with 1 mg/mL CD44 blocking antibodies and 1 mg/ml b1-integrin blocking antibodies; (9) SKOV-3 cell adhesion to LP-9 cells pretreated with 1mg/mL CX3CL1 blocking antibodies, 1 mg/mL CD44 blocking antibodies, and 1 mg/mL b1-integrin blocking antibodies. These data represent the averageof 3 experiments conducted in sextuplicate. A quantitative analysis of the adhesion data is presented in the histogram.

Kim et al.





specimens were CX3CR1 positive, suggesting that fractalk-ine signalingmay affect the progression of all of these tumors.Importantly, these data show that cancers beyond those ofepithelial origin may depend on fractalkine signaling.

Discussion

The mechanisms underlying peritoneal dissemination ofovarian carcinoma are poorly understood. Ovarian

0

0.5

1

1.5

2

2.5P

rolife

rati

on

G

S

S + 25 nmol/L CX3CL1

S + 25 nmol/L CX3CL1 +5 µmol/L AG1478

A

BSerum starved 10% FBS Serum starved +

CX3CL1

PCNA DAPI PCNA DAPI

Merge

Serum starved + CX3CL1

C

ImmortalizedOSE

Borderline EOC

SerousEOC

*

*

* *

**

*

**

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

CX3CR1

EGFR

EGFR

G

S

Pro

lifer

atio

n

Pro

tein

exp

ress

ionS + 25 nmol/L

CX3CL1

SKOV-3

SKOV-3 C

trl si

SKOV-3 C

X3CR1 s

i

SKOV-3 E

GFR si

SKOV-3 N

T

SKOV-3 C

trl si

SKOV-3 C

X3CR1 s

i

SKOV-3 E

GFR si

1.51

0.50

2.5

2

1.5

1

0.5

0

CX3CR1

GAPDH

Figure 6. CX3CL1 is instrumental in promoting cell proliferation. A, the proliferation rates of serum-starved IOSE, borderline EOC, and serous EOC celllines were evaluated in the presence of 25 nmol/L CX3CL1 ("S þ 25 nmol/L CX3CL1", green bars) or 25 nmol/L CX3CL1 plus 5 mmol/L AG1478 ("S þ 25nmol/L CX3CL1 þ 5 mmol/L AG1478", purple bars). Growth phase cells cultured in complete media containing 10% FBS were used as a positivecontrol ("G", blue bars). Serum-starved cells were used as a negative control ("S", red bars). These data represent the average of 3 experimentsconducted in triplicate. �, P < 0.05. B, SKOV-3 cells were serum-starved overnight, then one group was exposed to complete medium containing10% FBS, second group was serum-starved, and the third group was treated with 25 nmol/L CX3CL1 in serum-free medium for 24 hours, andphotographed using a Zeiss microscope with a 5� objective. Bar, 100 mm. Nuclear PCNA staining (green) was evaluated by immunofluorescence.Nuclei were stained with DAPI (blue). Fluorescent images were generated using a Zeiss AxioObserver fluorescent microscope with a 40� objective.Images were collected independently on the green and blue filters and subsequently superimposed. Bar, 20 mm. C, SKOV-3 cells that were nottransfected, transfected with control, CX3CR1 specific, or EGFR-specific siRNAs were subjected to the cell proliferation assay in the presence of 25nmol/L CX3CL1 ("S þ 25 nmol/L CX3CL1 growth phase cells cultured in complete media containing 10% FBS were used as a positive control ("G", bluebars). Serum-starved cells were used as a negative control ("S", red bars). Data represent the average of 3 experiments conducted in triplicate.�, P < 0.05. Expression of CX3CR1 and EGFR in transfected and nontransfected SKOV-3 cells was determined by Western blot. GAPDH served as aloading control. THP-1 and A431 cell lysates were used as positive controls for expression of CX3CR1 and EGFR, respectively. Histogramsshow the levels of CX3CR1 and EGFR expression, as determined by digital densitometry. �, P < 0.05.






carcinoma is a malignancy with an exceptionally high mor-tality rate (1), largely due to the lack of effective antimeta-static treatment approaches. A more detailed understandingof the mechanisms underlying the formation and develop-ment of EOC metastases could offer insights into how latestages of this disease might be effectively targeted. Proteinsthat are presented on the cell surface have consistently beenconsidered attractive molecular targets for disease treatment.Our data suggest that a member of GPCR family, thechemokine fractalkine receptor, functions to support theprometastatic properties of EOC cells, which include migra-tion, peritoneal adhesion, and proliferation. In addition, thefractalkine chemokine can support cell proliferation throughanother receptor, EGFR. Our data support the hypothesisthat multiple types of ovarian carcinoma rely on fractalkinesignaling for disease progression.Peritoneal metastasis is a terminal stage of EOC pro-

gression. Nevertheless, the mechanistic details of meso-thelial adhesion are poorly understood. Several molecularinteractions between EOC cells and CD44, b1-integrin,L1, and CA125 as well as between mesothelial cells andhyaluronan, fibronectin, neuropilin-1, and mesothelinhave been shown to contribute to peritoneal attachmentboth in vitro and in vivo (8, 48–51). Despite thesefindings, which show the importance of mesothelial adhe-sion, a complete understanding of all factors that playa role in EOC cell adhesion to the peritoneum has yetto be presented. Moreover, it is expected that differencesin patient-specific expression levels and isoform hetero-geneity of peritoneal adhesion-stimulating molecules(7, 52, 53) will diminish the extensive inhibition observedin both cell culture and xenograft models. Thus, it isimperative to elucidate other factors that contribute tomesothelial adhesion, which is a key initial step in themetastatic colonization of the peritoneal cavity. To thisend, our data strongly suggest that the fractalkine networkoriginating between CX3CR1-positive EOC cells andCX3CL1-positive mesothelial cells could play a role incell–cell adhesion. There may be several potential mechan-isms whereby this interaction bridges cell–cell attachment.A traditional mechanism utilized by rolling leukocytesinvolves interaction of surface expressed chemokine re-ceptors with chemokines presented on the glycosamino-glycans chains of endothelial cells following integrinactivation (22, 54). CX3CL1, however, is a transmem-brane protein, and it has been previously shown thatCX3CL1-mediated cell–cell adhesion can take place inthe absence of integrin activation (22, 54). Our datasupport the latter mechanism as we have observed anadditive negative effect on cell adhesion from simulta-neous disruption of CX3CL1/CX3CR1-, b1-integrin-,and CD44-dependent modes of adhesion.Our data indicate that 2 additional properties of the

metastatic cell, migration and proliferation, are dependenton fractalkine signaling as well. We show that the CX3CL1/CX3CR1 pair plays a pivotal role in facilitating cell motility.This property may have implications for all stages of EOCmetastatic progression.CX3CL1 is present in the ascites fluid

of EOC patients at levels that allow for productive CX3CL1/CX3CR1 signaling, which might induce shedding of themalignant cells and locoregional migration of the cellscomprising growing metastatic lesions. Interestingly, ourdata, albeit limited by the number of specimens, indicatethat the level of CX3CL1 in the ascites fluid of patients withbenign gynecologic diseases is an order of magnitude lowerthan that of the EOC ascites. These data suggest thatalthough CX3CR1 is expressed in transformed cells,CX3CL1-mediated migration off of the ovarian surface isunlikely to occur due to the lower than optimal levels of theligand, as the IC50 andKd of CX3CL1 and CX3CR1 bindingwere estimated at 2.4 and 1.7 nmol/L, respectively (55, 56).Thus, the benign tumors might harbor the potentialfor future activation of CX3CL1/CX3CR1-dependentmigration.This work also reveals a role for fractalkine signaling in

EOC cell proliferation. Our data suggest that both EGFRand CX3CR1 could be the primary receptors supportingCX3CL1-dependent EOC cell proliferation. Proliferationplays a significant role in the progression of peritonealmetastasis from EOC. EOC is characterized by a vastuncontrolled spread of metastasis through the peritone-um. Moreover, because an estimated 48% of EOC casesare EGFR positive (57), and our studies show that themajority of EOC cases are CX3CR1 positive, CX3CL1could be considered an important contributor to EOC cellproliferation.Taken together, our data indicate that both fractalkine and

its receptor may facilitate the metastatic progression of EOCby participating in migration, proliferation, and peritonealadhesion. Hence, both fractalkine and its receptor maybecome useful therapeutic targets. Targeted inhibition ofboth membrane-bound and soluble CX3CL1, as well as itsreceptor CX3CR1, could result in impaired migration,adhesion, and proliferation. However, additional studies arerequired to validate our in vitro data and disclose the impactof these pathways (CX3CL1/EGFR andCX3CL1/CX3CR1)on the development of metastasis in vivo. Interestingly, apossibility of the "cross-talk" between CXCL12/CXCR4and EGFR, which may result in linking signals of cellproliferation in ovarian carcinoma, has been suggestedbefore (58). Thus, it is interesting to speculate that activatedCX3CR1, which belongs to the family of chemokine recep-tors along with CXCR4, may activate EGFR-dependentproliferation using a similar "cross-talk" mechanism.It has been proposed that the CX3CL1/CX3CR1 pathway

could become a major therapeutic target due to its involve-ment in various pathologic conditions, such as inflamma-tion, pain, and cancer (59). Although no anti-CX3CL1 oranti-CX3CR1 therapies are currently available, a few studieshave shown the efficacy of several compounds in preclinicalstudies. For example, an analog of CX3CL1 that has aninhibitory effect on CX3CR1 during inflammation wasrecently discovered (60). A small molecule inhibitor ofCX3CR1, AZ12201182, was effective in abrogating mitosisand apoptosis (25). In another report, nicotinic acid wasshown to effectively suppress fractalkine and other

Kim et al.





chemokines in adipocytes (61). Although there are cases inwhich various diseases were successfully alleviated by target-ing chemokine receptors, challenges in drug design andefficacy, the existence of multiple receptors driving thedisease, species differences, and relevance of the target tothe human disease, among others, have also been reported(62). Thus, the clinical efficacy of future CX3CR1-targetingagents remains to be confirmed.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Dr. Penzes for the critical reading of the manuscript.

Grant Support

This work was supported by the NIH grant 1R21CA160917 (to M.V. Barbolina).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 June 3, 2011; revised October 31, 2011; accepted November 3, 2011;published OnlineFirst November 7, 2011.

References

1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer JClin 2010;60:277–300.

2. Cannistra SA. Cancer of the ovary. N Engl J Med 2004;351:2519–29.3. ArmstrongDK, BundyB,Wenzel L, HuangHQ,BaergenR, Lele S, et al.

Intraperitoneal cisplatin and paclitaxel in ovarian cancer. N Engl J Med2006;354:34–43.

4. Martin LP, Schilder RJ. Management of recurrent ovarian carcinoma:current status and future directions. Semin Oncol 2009;36:112–25.

5. Vasey PA.Management of recurrent epithelial ovarian carcinoma. AustN Z J Obstet Gynaecol 2005;45:269–77.

6. Lengyel E. Ovarian cancer development and metastasis. Am J Pathol2010;177:1053–64.

7. Cannistra SA, DeFranzo B, Niloff J, Ottensmeir C. Functional hetero-geneity of CD44molecules in ovarian cancer cell lines. Clin Cancer Res1995;1:333–42.

8. Cannistra SA, Kansas GS, Niloff J, DeFranzo B, Kim Y, Ottensmeier C.Binding of ovarian cancer cells to peritoneal mesothelium in vitro ispartly mediated by CD44H. Cancer Res 1993;53:3830–8.

9. Casey RC, Skubitz AP. CD44 and beta1 integrins mediate ovariancarcinoma cell migration toward extracellular matrix proteins. Clin ExpMetastasis 2000;18:67–75.

10. Khan SM, Funk HM, Thiolloy S, Lotan TL, Hickson J, Prins GS, et al. Invitro metastatic colonization of human ovarian cancer cells to theomentum. Clin Exp Metastasis 2010;27:185–96.

11. Barbolina MV, Kim M, Liu Y, Shepard J, Belmadani A, Miller RJ, et al.Microenvironmental regulation of chemokine (C-X-C-motif) receptor 4in ovarian carcinoma. Mol Cancer Res 2010;8:653–64.

12. Kohn EC, Travers LA, Kassis J, Broome U, Klominek J. Malignanteffusions are sources of fibronectin and other promigratory and proin-vasive components. Diagn Cytopathol 2005;33:300–8.

13. Sood A,Moudgil A, SoodN, Kharay AS, Kaushal S, Narang AP. Role offibronectin in diagnosis of malignant ascites. J Assoc Physicians India1997;45:283–5.

14. Williams C, Hill SJ. GPCR signaling: understanding the pathway tosuccessful drug discovery. Methods Mol Biol 2009;552:39–50.

15. Viola A, Luster AD. Chemokines and their receptors: drug targets inimmunity and inflammation. Annu Rev Pharmacol Toxicol 2008;48:171–97.

16. Balkwill F. Cancer and the chemokine network. Nat Rev Cancer2004;4:540–50.

17. Jamieson WL, Shimizu S, D'Ambrosio JA, Meucci O, Fatatis A.CX3CR1 is expressed by prostate epithelial cells and androgensregulate the levels of CX3CL1/fractalkine in the bonemarrow: potentialrole in prostate cancer bone tropism. Cancer Res 2008;68:1715–22.

18. Marchesi F, Piemonti L, Fedele G, Destro A, Roncalli M, Albarello L,et al. The chemokine receptorCX3CR1 is involved in the neural tropismandmalignant behavior of pancreatic ductal adenocarcinoma. CancerRes 2008;68:9060–9.

19. Shulby SA, Dolloff NG, Stearns ME, Meucci O, Fatatis A. CX3CR1-fractalkine expression regulates cellularmechanisms involved in adhe-sion, migration, and survival of human prostate cancer cells. CancerRes 2004;64:4693–8.

20. Kajiyama H, Shibata K, Terauchi M, Ino K, Nawa A, Kikkawa F.Involvement of SDF-1alpha/CXCR4 axis in the enhanced peritonealmetastasis of epithelial ovarian carcinoma. Int J Cancer 2008;122:91–9.

21. Scotton CJ, Wilson JL, Scott K, Stamp G, Wilbanks GD, Fricker S,et al. Multiple actions of the chemokine CXCL12 on epithelialtumor cells in human ovarian cancer. Cancer Res 2002;62:5930–8.

22. Imai T, Hieshima K, Haskell C, Baba M, Nagira M, Nishimura M, et al.Identification and molecular characterization of fractalkine receptorCX3CR1, whichmediates both leukocytemigration and adhesion. Cell1997;91:521–30.

23. Bazan JF, Bacon KB, Hardiman G, Wang W, Soo K, Rossi D, et al. Anew class of membrane-bound chemokine with a CX3C motif. Nature1997;385:640–4.

24. Pan Y, Lloyd C, Zhou H, Dolich S, Deeds J, Gonzalo JA, et al.Neurotactin, a membrane-anchored chemokine upregulated in braininflammation. Nature 1997;387:611–7.

25. White GE, Tan TC, John AE, Whatling C, McPheat WL, Greaves DR.Fractalkine has anti-apoptotic and proliferative effects on humanvascular smooth muscle cells via epidermal growth factor receptorsignalling. Cardiovasc Res 2010;85:825–35.

26. Hart KM, Bak SP, Alonso A, Berwin B. Phenotypic and functionaldelineation of murine CX(3)CR1 monocyte-derived cells in ovariancancer. Neoplasia 2009;11:564–73, 1 p following 73.

27. Lin A, Schildknecht A, Nguyen LT, Ohashi PS. Dendritic cells integratesignals from the tumor microenvironment to modulate immunity andtumor growth. Immunol Lett 2010;127:77–84.

28. Nelson DS, Nelson M. Evasion of host defences by tumours. ImmunolCell Biol 1987;65:287–304.

29. Ohm JE, Carbone DP. Immune dysfunction in cancer patients. Oncol-ogy (Williston Park) 2002;16:11–8.

30. Takaishi K, Komohara Y, Tashiro H, Ohtake H, Nakagawa T, KatabuchiH, et al. Involvement of M2-polarizedmacrophages in the ascites fromadvanced epithelial ovarian carcinoma in tumor progression via Stat3activation. Cancer Sci 2010;101:2128–36.

31. Labidi-Galy SI, Sisirak V,MeeusP, GobertM, Treilleux I, Bajard A, et al.Quantitative and functional alterations of plasmacytoid dendritic cellscontribute to immune tolerance in ovarian cancer. Cancer Res2011;71:5423–34.

32. Mlecnik B, Tosolini M, Charoentong P, Kirilovsky A, Bindea G, BergerA, et al. Biomolecular network reconstruction identifies T-cell homingfactors associated with survival in colorectal cancer 2010;138:1429–40.

33. Barbolina M, Adley BP, Kelly DL, Fought AJ, Scholtens DM, SheaLD, Stack MS. Motility-related actinin alpha-4 is associated withadvanced and metastatic ovarian carcinoma. Lab Invest 2008;88:602–14.

34. Barbolina MV, Adley BP, Ariztia EV, Liu Y, Stack MS. Microenviron-mental regulation of membrane type 1 matrix metalloproteinase activ-ity in ovarian carcinoma cells via collagen-inducedEGR1 expression. JBiol Chem 2007;282:4924–31.






35. Barbolina MV, Adley BP, Shea LD, Stack MS. Wilms tumor geneprotein 1 is associated with ovarian cancer metastasis and modulatescell invasion. Cancer 2008;112:1632–41.

36. Levanon K, Crum C, Drapkin R. New insights into the pathogenesis ofserous ovarian cancer and its clinical impact. J Clin Oncol 2008;26:5284–93.

37. Pothuri B, Leitao MM, Levine DA, Viale A, Olshen AB, Arroyo C, et al.Genetic analysis of the early natural history of epithelial ovariancarcinoma. PLoS One 2010;5:e10358.

38. Auersperg N. The origin of ovarian carcinomas: a unifying hypothesis.Int J Gynecol Pathol 2011;30:12–21.

39. Lucas AD, Chadwick N, Warren BF, Jewell DP, Gordon S, Powrie F,et al. The transmembrane form of theCX3CL1 chemokine fractalkine isexpressed predominantly by epithelial cells in vivo. Am J Pathol2001;158:855–66.

40. Gaudin F, Nasreddine S, Donnadieu AC, Emilie D, Combadi�ere C,Pr�evot S, et al. Identification of the chemokine CX3CL1 as a newregulator of malignant cell proliferation in epithelial ovarian cancer.PLoS One 2011;6:e21546.

41. Garton KJ, Gough PJ, Blobel CP, Murphy G, Greaves DR, DempseyPJ, et al. Tumor necrosis factor-alpha-converting enzyme (ADAM17)mediates the cleavage and shedding of fractalkine (CX3CL1). J BiolChem 2001;276:37993–8001.

42. ShimoyaK, ZhangQ, Temma-AsanoK,Hayashi S,Kimura T,MurataY.Fractalkine in the peritoneal fluid of women with endometriosis. Int JGynaecol Obstet 2005;91:36–41.

43. Harvey W, Amlot PL. Collagen production by human mesothelial cellsin vitro. J Pathol 1983;139:337–47.

44. Goda S, Imai T, Yoshie O, Yoneda O, Inoue H, Nagano Y, et al. CX3C-chemokine, fractalkine-enhanced adhesion of THP-1 cells to endo-thelial cells through integrin-dependent and -independent mechan-isms. J Immunol 2000;164:4313–20.

45. Lloyd AR, Oppenheim JJ, Kelvin DJ, Taub DD. Chemokines regulate Tcell adherence to recombinant adhesion molecules and extracellularmatrix proteins. J Immunol 1996;156:932–8.

46. Wang JM, Deng X, Gong W, Su S. Chemokines and their role in tumorgrowth and metastasis. J Immunol Methods 1998;220:1–17.

47. Chi DS, Eisenhauer EL, Zivanovic O, Sonoda Y, Abu-Rustum NR,Levine DA, et al. Improved progression-free and overall survival inadvanced ovarian cancer as a result of a change in surgical paradigm.Gynecol Oncol 2009;114:26–31.

48. Arlt MJ, Novak-Hofer I, Gast D, Gschwend V, Moldenhauer G,Gr€unberg J, et al. Efficient inhibition of intra-peritoneal tumor growthand dissemination of human ovarian carcinoma cells in nude mice byanti-L1-cell adhesion molecule monoclonal antibody treatment. Can-cer Res 2006;66:936–43.

49. Sawada K, Mitra AK, Radjabi AR, Bhaskar V, Kistner EO, TretiakovaM, et al. Loss of E-cadherin promotes ovarian cancer metastasis viaalpha 5-integrin, which is a therapeutic target. Cancer Res 2008;68:2329–39.

50. Scholler N, Garvik B, Hayden-Ledbetter M, Kline T, Urban N. Devel-opment of aCA125-mesothelin cell adhesion assay as a screening toolfor biologics discovery. Cancer Lett 2007;247:130–6.

51. Strobel T, Cannistra SA. Beta1-integrins partly mediate binding ofovarian cancer cells to peritoneal mesothelium in vitro. Gynecol Oncol1999;73:362–7.

52. Daponte A, Kostopoulou E, Kollia P, Papamichali R, Vanakara P,Hadjichristodoulou C, et al. L1 (CAM) (CD171) in ovarian serousneoplasms. Eur J Gynaecol Oncol 2008;29:26–30.

53. Bast RC Jr, Klug TL, St John E, Jenison E, Niloff JM, Lazarus H, et al. Aradioimmunoassay using amonoclonal antibody tomonitor the courseof epithelial ovarian cancer. N Engl J Med 1983;309:883–7.

54. Fong AM, Robinson LA, Steeber DA, Tedder TF, Yoshie O, Imai T, et al.Fractalkine and CX3CR1 mediate a novel mechanism of leukocytecapture, firm adhesion, and activation under physiologic flow. J ExpMed 1998;188:1413–9.

55. Davis CN, Zujovic V, Harrison JK. Viral macrophage inflammatoryprotein-II and fractalkine (CX3CL1) chimeras identify molecular deter-minants of affinity, efficacy, and selectivity at CX3CR1.Mol Pharmacol2004;66:1431–9.

56. Nakayama T, Watanabe Y, Oiso N, Higuchi T, Shigeta A, Mizuguchi N,et al. Eotaxin-3/CC chemokine ligand 26 is a functional ligand forCX3CR1. J Immunol 2010;185:6472–9.

57. Lafky JM, Wilken JA, Baron AT, Maihle NJ. Clinical implications ofthe ErbB/epidermal growth factor (EGF) receptor family and its ligandsin ovarian cancer. Biochim Biophys Acta 2008;1785:232–65.

58. Porcile C, Bajetto A, Barbieri F, Barbero S, Bonavia R, Biglieri M, et al.Stromal cell-derived factor-1alpha (SDF-1alpha/CXCL12) stimulatesovarian cancer cell growth through the EGF receptor transactivation.Exp Cell Res 2005;308:241–53.

59. D'Haese JG,Demir IE, FriessH,CeyhanGO. Fractalkine/CX3CR1:whya single chemokine-receptor duobears amajor andunique therapeuticpotential. Expert Opin Ther Targets 2010;14:207–19.

60. DorghamK,Ghadiri A, HermandP,RoderoM,Poupel L, IgaM, et al. Anengineered CX3CR1 antagonist endowed with anti-inflammatoryactivity. J Leukoc Biol 2009;86:903–11.

61. Digby JE,McNeill E, DyarOJ, LamV,GreavesDR,ChoudhuryRP.Anti-inflammatory effects of nicotinic acid in adipocytes demonstrated bysuppression of fractalkine, RANTES, and MCP-1 and upregulation ofadiponectin. Atherosclerosis 2010;209:89–95.

62. Horuk R. Chemokine receptor antagonists: overcoming developmen-tal hurdles. Nat Rev Drug Discov 2009;8:23–33.

Kim et al.





2012;10:11-24. Published OnlineFirst November 7, 2011.Mol Cancer Res Mijung Kim, Lisa Rooper, Jia Xie, et al. Peritoneal Mesothelial CellsCarcinoma Cells and Required for Motility and Adhesion to

CR1 Is Expressed in Epithelial Ovarian3Fractalkine Receptor CX

Updated version

10.1158/1541-7786.MCR-11-0256doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2011/11/07/1541-7786.MCR-11-0256.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/10/1/11.full#ref-list-1

This article cites 61 articles, 18 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/10/1/11.full#related-urls

This article has been cited by 4 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/10/1/11To 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-11-0256

http://mcr.aacrjournals.org/content/suppl/2011/11/07/1541-7786.MCR-11-0256.DC1

http://mcr.aacrjournals.org/content/10/1/11.full#ref-list-1

http://mcr.aacrjournals.org/content/10/1/11.full#related-urls

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

mailto:[email protected]

http://mcr.aacrjournals.org/content/10/1/11


fractalkine receptor cx cr1 is expressed in epithelial ovarian … · bd biosciences. transwell...

Documents