Syngeneic Murine Ovarian Cancer Model Reveals That Ascites … · ascites were associated with enhanced sphere-forming ability (Supplementary Fig. S3A and S3B). We chose 50% ascites

Cancer Biology and Signal Transduction

Syngeneic Murine Ovarian Cancer Model RevealsThat Ascites Enriches for Ovarian Cancer Stem-Like Cells Expressing Membrane GRP78Lihong Mo1, Robin E. Bachelder1, Margaret Kennedy1, Po-Han Chen2,3,Jen-Tsan Chi2,3, Andrew Berchuck4, George Cianciolo1, and Salvatore V. Pizzo1

Abstract

Patients with ovarian cancer are generally diagnosed at FIGO(International Federation of Gynecology and Obstetrics) stageIII/IV, when ascites is common. The volume of ascites correlatespositively with the extent of metastasis and negatively withprognosis. Membrane GRP78, a stress-inducible endoplasmicreticulum chaperone that is also expressed on the plasmamembrane (memGRP78) of aggressive cancer cells, plays acrucial role in the embryonic stem cell maintenance. We stud-ied the effects of ascites on ovarian cancer stem-like cells usinga syngeneic mouse model. Our study demonstrates that ascites-derived tumor cells from mice injected intraperitoneally withmurine ovarian cancer cells (ID8) express increased memGRP78levels compared with ID8 cells from normal culture. Wehypothesized that these ascites-associated memGRP78þ cells arecancer stem-like cells (CSC). Supporting this hypothesis, we

show that memGRP78þ cells isolated from murine ascites exhib-it increased sphere forming and tumor initiating abilities com-pared with memGRP78� cells. When the tumor microenviron-ment is recapitulated by adding ascites fluid to cell culture, ID8cells express more memGRP78 and increased self-renewingability compared with those cultured in medium alone. More-over, compared with their counterparts cultured in normalmedium, ID8 cells cultured in ascites, or isolated from ascites,show increased stem cell marker expression. Antibodies direct-ed against the carboxy-terminal domain of GRP78: (i) reduceself-renewing ability of murine and human ovarian cancer cellspreincubated with ascites and (ii) suppress a GSK3a-AKT/SNAI1 signaling axis in these cells. Based on these data, wesuggest that memGRP78 is a logical therapeutic target for late-stage ovarian cancer. Mol Cancer Ther; 14(3); 747–56. �2015 AACR.

IntroductionOvarian cancer is the most lethal gynecologic malignancy (1).

Seventy-five percent of patients with ovarian cancer are diagnosedat stage III/IV, with a long-term survival rate of 10% to 30% (2).Ascites occurs in 17% of patients with FIGO (InternationalFederation ofGynecology andObstetrics) stage I/II ovarian cancerand 89% of patients at stage III/IV (3). Standard management ofovarian cancer consists of surgical staging, operative tumordebulking, and intravenous chemotherapy (2). Although morethan 80% of advanced stage patients benefit from first-line ther-apy (2),most patients suffer from recurrencewithin 18months ofdiagnosis (4). This high relapse rate may result from failure ofconventional therapy to remove cancer stem-like cells (CSC;

ref. 5). CSCs are slow-cycling, therapy-resistant tumor cells capa-ble of self-renewal (6, 7). CSCs play a crucial role in tumorinitiation and metastasis (8).

Glucose-regulated protein 78 (GRP78) is a stress-inducibleendoplasmic reticulum (ER) chaperone that is also expressed onthe plasma membrane (memGRP78) of aggressive cancers (9, 10).GRP78 protects cells from stress by activating the PI3K/AKTpathway and inhibiting proapoptotic cascades. However, if theER is severely stressed, GRP78 can promote cell death (9). Twofunctional domains ofGRP78have been characterized: an amino-terminal domain that drives PI3K/AKT activity (11, 12), and acarboxy (COOH)-terminal domain that promotes apoptotic sig-naling (13, 14). Our previous work demonstrates that targetingGRP78 with mouse IgGs against the GRP78 COOH-terminaldomain causes tumor cell apoptosis by activating the caspasepathway (13, 14), resulting in delayed tumor growth, and pro-longed survival in a mouse melanoma model (15).

GRP78 maintains survival of embryonic (16) and adult mam-mary stem cells (17), and is highly expressed in hematopoieticstem cells (18). Overexpression of GRP78 correlates with malig-nant transformation in epithelial ovarian tumor cells (19), where-as inducible knockout of GRP78 in the hematopoietic systemsuppresses Pten-null leukemogenesis (20). MemGRP78 is associ-atedwith increased cancer stemness in head and neck cancer (21).Based on these findings, in addition to our laboratory's studies ofmemGRP78 functions in cancer (11–15, 22), in the current work,we investigated functions for memGRP78 in in ovarian cancerstemness.

1Department of Pathology, Duke University Medical Center, Durham,North Carolina. 2Department of Molecular Genetics andMicrobiology,Duke University Medical Center, Durham, North Carolina. 3Center forGenomic andComputational Biology, DukeUniversityMedical Center,Durham, North Carolina. 4Department of Obstetrics/Gynecology,Division of Gynecologic Oncology, Duke University Medical Center,Durham, North Carolina.

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

Corresponding Author: Robin E. Bachelder, Duke University Medical Center,DUMC Box 3712, Durham, NC 27710. Phone: 919-684-3072; Fax: 919-684-5077;E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-14-0579

�2015 American Association for Cancer Research.

MolecularCancerTherapeutics

www.aacrjournals.org 747

on August 24, 2020. © 2015 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst January 14, 2015; DOI: 10.1158/1535-7163.MCT-14-0579

http://mct.aacrjournals.org/

To test the hypothesis that memGRP78 is an ovarian CSCmarker, we utilized a syngeneic, immunocompetent ovariancancer model (23). We chose this model because GRP78 auto-antibodies circulating in patients with cancer activate memGRP78(11, 12, 24). We now demonstrate that memGRP78 expression isincreased in ovarian cancer cells isolated from ascites in vivo andovarian cancer cells treated with ascites in vitro. We also show thatmemGRP78þ cells possess higher tumorigenic potential and stem-ness properties than memGRP78� cells. Neutralizing memGRP78with anti-COOH terminal domain antibodies suppresses ovariancancer cell sphere-forming ability. Our work is the first to dem-onstrate that memGRP78þ CSCs are enriched by ascites and thatmemGRP78 is a logical target for eliminating ovarian CSCs.

Materials and MethodsCell culture

ID8 cells (mouse epithelial ovarian cancer cell line obtained in2006 from Dr. Kathy Roby, Kansas University Medical Center)were maintained in DMEM (high glucose, Gibco-Life Technolo-gies) containing 4% FBS, 1� penicillin streptomycin. Compari-son of these cells with the original published ID8 cell line isdescribed (Supplementary Fig. S1). ID8-GFP was a gift from Dr.Brent Berwin (Dartmouth Medical School; 2007). Both ID8 andID8-GFP cellswere shown tobemycoplasma-free (March 20, 2010)and tested to be identical by STR DNA profiling (November 05,2014). Human ovarian cancer cell lines [OvCar3 (January 20,2014) and ES2 (June 06, 2013)] were purchased from Duke CellCulture Facility (CCF) and cultured under ATCC-recommendedconditions. OvCar3 (December 06, 2013) and ES2 (June 20,2009) cells were shown to be mycoplasma-free and authenticatedusing STR DNA profiling before being frozen by Duke CCF. Fiftypercent ascites treatment conditions were adjusted to normalculture conditions with regard to FBS and glucose concentrations.

AntibodiesAntibody sources were as follows: GRP78 N20 and C20 (Santa

Cruz Biotechnology); OCT4, SOX9, and GSK3 (Millipore);CD133 and SCA1 (Biolegend); SNAI1, p-GSK3, AKT, and p-AKT(Cell Signaling Technology). GRP78 C38 and C107 antibodieswere produced in our laboratory (15).

Mouse studiesAnimal experimentation was performed according to the reg-

ulations of the Duke Institutional Animal Care and Use Com-mittee. ID8 cells were injected into the mouse peritoneum offemale 6- to 8-week-oldC57BL/6mice (Charles River). Endpointsfor euthanasia included lethargy, decreased motility, or cross-sectional abdominal diameter increase greater than 1/3.

AscitesDe-identified patient ascites samples (Ov476, Ov480) were

obtained from informed subjects with FIGO stage IIIC grade 2ovarian serous adenocarcinoma (Duke University IRB approvedprotocol Pro00013710). For murine ascites, mice were eutha-nized and peritoneal fluid was collected and centrifuged twice at500 � g for 5 minutes to separate cellular and acellular fractions.Acellular fractions from multiple mice were pooled and filteredthrough 0.22 mm sterile filters.

Microarray analysisTotal RNAwas isolated (Qiagen RNeasyMicro kit) according to

manufacturer's protocol and analyzed byMouseGenome 430 2.0Array (Affymetrix; GEO accession number GSE61285). Array datawere then normalized by Robust Multiarray Averaging (RMA) forfurther analysis. RMA is a normalization procedure for micro-arrays that background corrects, normalizes, and summarizes theprobe level information from the Affymetrix microarray data. Toidentify differentially expressed genes, we applied SAM 4.01 ExcelAdd-In (25) that provides the estimate of FDR formultiple testing.Using the FDR threshold of 2.5%, we identified 1,257 probesetswhose expression values were extracted, mean-centered and clus-tered by Cluster 3.0, and viewed by Treeview v1.6 (26). Theseselected genes (Supplementary Table S1)were thendeposited intoGather (gather.genome.duke; ref. 27) to determine the enrich-ment for the Gene Ontology (GO) and KEGG. For the stemnessscore, the mean expression level of 38 stemness-related genes wascalculated as "stemness score" in amulti-gene signature approach(28, 29) and compared by t test to avoid quantitative bias fromany single gene.

Flow cytometric analysisID8 cells (95% confluence) were harvested using dissociation

buffer (Gibco). C20was used for memGRP78 staining. 7-AAD (BDBiosciences) was used to exclude dead cells. Cells were fixed andpermeabilized using Cytofix/Cytoperm Kit (BD Biosciences)before OCT4 staining. Antibody binding was assessed using aGuava Easycyte Plus flow cytometer (Millipore).

Cell sortingRed blood cells were removed from the cellular fraction of

murine ascites using RBC lysis buffer (BD Biosciences). Remain-ing cells were prepared with Fc block and anti-CD11b antibodycoated magnetic beads (558013; BD Biosciences) and macro-phages were removed with a BD magnet. The remaining samplewas stained with F4/80, 7AAD, and GRP78 N20 and sorted byDuke Cancer Institute Flow Cytometry Shared Resource to selectfor the F4/80�, 7AAD�, and memGRP78þ or F4/80�, 7AAD� andmemGRP78� populations.

Sphere formation assaySingle-cell suspensions were plated in 24-well ultralow attach-

ment plates at 6 � 103 cells per well (ID8 and OvCar3) and 1 �103 cells per well (ES2) in primary assay. Cells were grown inserum-free DMEM supplemented with 1%methyl cellulose (Sig-ma-Aldrich), B27 (Invitrogen), 20 ng/mL EGF, 20 ng/mL bFGF(BD Biosciences), and 4 mg/mL bovine insulin (Invitrogen).Spheres were cultured for 7 to 14 days, after which all spheres(diameter > 50 mm) were counted in the well. For serial passages,spheres were dissociated to single cells with trypsin, and 6 � 103

cells plated and cultured inultralowattachment plates for 14days.

DiD retentionID8 cells were labeled with DiD dye (Life Technologies) using

the manufacturer's protocol. Cells were split into two groupsbeing cultured in either medium or 50% acellular ascites. DiDintensity was measured on day 7. Similar experiments wereperformed with DiO, Dil (Life Technologies), and Cyto-ID Red(Enzo Life Sciences).

Mo et al.

Mol Cancer Ther; 14(3) March 2015 Molecular Cancer Therapeutics748




Statistical analysisPrism (version 5.0; GraphPad Software) was used to perform

statistical analyses. Two-tailed Student t test was used to testdifferences in sample means for data with normally distributedmeans. The c2- test was used to evaluate the survival difference inanimal studies. The likelihood ratio test of single-hit model wasperformed as a standard statistical analysis method for limitingdilution studies. P values less than 0.05 were considered statis-tically significant.

ResultsAcellular ascites increases ovarian cancer cell stemness

A syngeneic mouse model was established by injecting murineepithelial ovarian cancer cells (ID8 cells) intraperitoneally intoC57BL/6mice. We observed a strong linear regression correlationbetween the volume of ascites and tumor burden (R2 ¼ 0.49;Supplementary Fig. S2A). Mice had shortened survival wheninjected with ID8 cells isolated from ascites or with ID8 cells plusascites supernatant from ID8 tumor–bearing mice as comparedwith mice receiving ID8 cells from normal culture (Supplemen-tary Fig. S2B). These results indicate that acellular ascites derivedfrom ID8 tumor–bearingmice is capable of driving ovarian cancerprogression.

We next tested the effect of ascites on ID8 cell self-renewingability. Equal numbers of ID8 cells from normal culture, ID8 cellsisolated from ascites, or ID8 cells pretreated with ascites for 7 dayswere seeded into sphere culture. Increasing concentrations ofascites were associated with enhanced sphere-forming ability(Supplementary Fig. S3A and S3B). We chose 50% ascites forfurther studies because it effectively increased memGRP78 expres-sion and sphere-forming ability of these cells. For all conditions,glucose and FBS concentration was normalized to that of controlmedium.

ID8 cells isolated from ascites and ascites-pretreated ID8 cellsexhibited increased sphere-forming ability compared with ID8cells in regular medium (Fig. 1A). Of note, our sphere assayincludes methylcellulose to avoid nonspecific cell aggregation(30). As shown in Fig. 1A, primary spheres, when dissociated,efficiently established secondary spheres, thus displaying CSCbehaviors. Protein concentration was normalized with albuminas a control for the 50% ascites treatment, and no significantdifference from normal cell culture was detected (SupplementaryFig. S3C).

To confirm that ascites increases sphere-forming ability ofovarian cancer cells, we employed a competition strategy betweenascites-pretreated and -untreated cells. ID8-GFP cells, which sharethe same proliferation rate as ID8 cells (data not shown), werepretreatedwith acellular ascites for 7 days and thenmixed 1:1withuntreated ID8 cells. The cell mixture was seeded into a sphereassay. Serial passage of primary sphere cells into a secondarysphere assay was also performed. Pictures were taken from 5different fields (Fig. 1B, left plot) and the percentages of ID8-GFPand ID8 cells from sphere assays were quantified. As shown in Fig.1B, spheres are composed mostly of ascites-pretreated ID8-GFPcells.

To test whether increased sphere-forming ability was reversibleby removing ascites, we recultured ID8 cells isolated from ascitesin ascites-freemediumor removed ascites fromascites-treated ID8cells. In both situations, sphere-forming ability of ID8 cells wasdecreased significantly (Fig. 1C).

Increased sphere-forming ability of ascites-pretreated ID8 cellscould reflect either ascites stimulation of CSC signaling or ascitesenrichment of a stem cell population. To differentiate betweenthese possibilities, we included ID8 cells exposed to acellularascites for 4hours, a short incubationpromoting signaling but notsufficient for enrichment of a tumor cell subpopulation. Sphere-forming ability of ID8 cells exposed to ascites for 4 hours wassimilar to that of untreated ID8 cells (Fig. 1A), supporting theenrichment hypothesis. After 7 days ascites treatment, 34.5% ID8ovarian cancer cells were Annexin V positive compared with 7.7%ID8 cells in normal medium (Fig. 1D). Collectively, our findingssuggest that ID8 ovarian cancer cells are heterogeneous. Althoughbulk tumor cells do not survive in an ascites microenvironment, asubpopulation of cells with cancer stem-like behavior survivesascites exposure.

To provide further evidence for ascites enrichment of a slow-cycling CSC population, we labeled ID8 cells with a lipophilictracer (DiD) that is diluted inproliferating cells, butmaintained innonproliferating/slow-proliferating cells. We detected 0% DiDþ

cells in ID8 cells cultured for 7 days in normal medium. Incontrast, 66.7% of cells treated with 50% acellular ascites for 7days were DiDþ (Fig. 1E). Similar results were observed using 3other lipophilic dyes (data not shown).

To begin to validate our findings in human ovarian cancer,human ovarian cancer cell lines were preincubated with either oftwopatient ascites samples.Notably, these humanascites samplesincreased sphere-forming ability of both human ovarian cancerlines compared with untreated cells (Fig. 1F and G.).

Microarray analysis of stemness-related genes in ascites-treatedID8 cells

We performed microarray analysis on untreated and ascites-pretreated ID8 cells (Fig. 2A). Gene expression profiles wereinterrogated with Affymetrix mouse 430A2 arrays (GEO acces-sion number GSE61285) and normalized by RMA. To identifydifferentially expressed genes while considering FDR, we usedSAM (25) to identify 1,257 probesets (707 induced and 550repressed; Supplementary Table S1) with 2.5% of FDR. Expres-sion of these genes was extracted, mean-centered, and clusteredto generate the overview heatmap (Fig. 2A). GO and KEGGenrichments of these selected genes were performed (GEOaccession number GSE61285).

We next investigated how ascites treatment of ovarian cancercells affected expression of 38 genes that were previouslyreported to associate with cancer stemness (31). To avoid thebias of any individual stem-related gene, we took a multigene"signature" approach that provides a quantitative measurementof the "stemness" of each sample (32). When the mean expres-sion values of all 38 genes were calculated as a "stemness-score," ascites-treated samples had a significantly higher stem-ness-score than control cells (Fig. 2B; Supplementary Table S2).The expression of 11 stemness genes (Sca-1/Ly6a, Abcb1a/b,Vegfa, Snai1, Sox9, Krt14, Cd44, Kit, Cd24, Kitl, and Ki67l) wasupregulated by ascites (Fig. 2C). Sca-1 (Ly6a), stem cell antigen-1, is a murine CSC marker (32, 33). Snai1 contributes to a stem-like phenotype in ovarian cancer (34–36). Sox9 converts dif-ferentiated breast cancer cells to CSCs (36).

To show that these stemness genes were upregulated at theprotein level in ascites-treated ovarian cancer cells, we nextperformed flow cytometry and Western blotting. By flow cyto-metry, we detected increased SCA1 expression in ascites-treated

Murine Ascites Enriches for Ovarian CSCs Expressing memGRP78

www.aacrjournals.org Mol Cancer Ther; 14(3) March 2015 749




Figure 1.Ascites enriches for CSCs. A, ID8 cells werecultured for the indicated times in medium or50% acellular ascites from ID8-bearing mice.A total of 6� 103 cells perwell were seeded ina primary sphere assay, and then harvested,trypsinized, and passaged to form secondaryspheres. Y axis represents spheres� 50mm�SD. B, competition between ascites-pretreated and -untreated ID8-GFP cells insphere formation. Left, ID8-GFP cells werepretreated with acellular ascites derived fromtumor-bearing mice for 7 days and thenmixed 1:1 with untreated ID8 cells. Themixturewas seeded into a sphere assay (primaryspheres). On day 7, primary spheres weretrypsinized and seeded into a sphere assay(secondary spheres). Pictures were takenfrom 5 random fields. A representative field isshown for primary spheres at 2.5�,secondary spheres at 2.5�, primary spheresat 20�, and secondary spheres at 20�. Right,quantification of percentage of ID8-GFP andID8 cells in 5 fields at 20�magnification fromprimary and secondary sphere assays. Redarrows, GFPþ spheres; white arrows, GFP�

spheres. C, culturing in vivo ascites cellsin vitro for 7 days (recultured; left plot) orreculturing ID8 cells pretreated with ascitesfor 7 days (ascites treated 7 days) in culturefor 9 days (ascites off 9 days; right plot)decreases their sphere-forming ability. Errorbars, SD from 3 trials in triplicate. D, after 7-day ascites treatment, 34.5% of ID8 cellsbecameAnnexinVpositive,whereas 7.7% ID8cells were positive in normal culture. E, ID8cells were labeled with DiD on day 0 and splitinto two groups, receiving either medium or50% ascites for 7 days. The majority ofascites-treated ID8 cells maintained DiD labelon day 7, whereas most ID8 cells in mediumlost thedye. F andG,OvCar3 or ES2 cellswerepretreated with 50% ascites from either oftwo patients with ovarian cancer (Ov476,Ov480) for 7 days, and their ability to formspheres was determined. Error bars, SD from3 different trials in triplicate for this figure.

Mo et al.





Figure 2.Incubatingmurine ovarian cancer cellswith ascites increases stemness geneexpression. A, hierarchy cluster of1,257 probesets identified by SAM tobe differentially expressed betweenID8 cells (n ¼ 3) and ascites-treatedID8 cells (n¼ 2). The heatmap showedthe mean-centered expression valuesof the selected genes (red, induced;green, repressed). B, expression levelsof 38 stemness-associated geneswere extracted, mean-centered, andarranged by hierarchical clustering(Supplementary Table S2).C, stemness-associated genessignificantly upregulated after ascitespretreatment of tumor cells for7 days. D, SCA1 expression on ID8cells (Medium), ID8 cells treated withascites for 7 days (Ascites treated),and ID8 cells from ascites (In vivo).Representative of 3 differentexperiments. E, Western blotting forSNAI1 and SOX9 in ID8 cells and ID8cells treated with ascites for 7 days.Quantification was by densitometry.Error bars, SD from 3 trials in triplicate.






ID8 cells, as well as in ID8 cells harvested from ascites in vivocomparedwith that in control ID8 cells (Fig. 2D). Comparedwithnormal culture, ascites treatment also increased expression ofSNAI1 and SOX9 significantly (Fig. 2E).

MemGRP78 expression in ascites-associated ovarian cancer cellsWe next studied ascites effects on memGRP78 expression.

MemGRP78 levels were significantly higher in ID8 cells isolatedfrom ascites compared with that in ID8 cells cultured in normalmedium (Fig. 3, top plot). When in vivo cells were cultured innormal medium for 7 days, the memGRP78 level shifted back toparental cell levels, leading us to hypothesize that survival of amemGRP78-expressing ovarian cancer cell subpopulation is sup-ported by soluble ascites factors. To test this hypothesis, wecultured ID8 cells with acellular ascites for 7 days and found thatthe%memGRP78þ cells increased significantly from 7.5% (paren-tal) to 43.2% (ascites 7 days), almost reaching the level of in vivotumor cells (Fig. 3, bottom). Removal of ascites from these cellsfor 9 days restored memGRP78þ expression to baseline levels(ascites off 9 days). Notably, memGRP78 levels remained atbaseline following short-term ascites exposure (Fig. 3, bottom).The reversibility in memGRP78 induction by ascites correlates withthe reversibility of ascites-enhanced sphere-forming ability of ID8cells (Fig. 1C), supporting the hypothesis that memGRP78 is anovarian CSC marker.

MemGRP78 is associated with stemnessTo further test whether memGRP78 is a stem cell marker in

murine ovarian cancer cells, we sorted ascites-derived tumor cellsinto memGRP78þ and memGRP78� populations and characterizedtheir self-renewing activity in a sphere assay. 7AADþ deadcells and F4/80þ macrophages were excluded, and the gate for

memGRP78þ andmemGRP78�was set at the extreme ends to insurepurity (Fig. 4A; ref. 37). Although memGRP78þ cells proliferatedslower than memGRP78� cells and unsorted tumor cells (Supple-mentary Fig. S4), memGRP78þ cells formed more spheres thanmemGRP78� cells (Fig. 4B). These studies suggest that memGRP78þ

ovarian cancer cells are similar toCSCs,which are characterized bytheir slow-cycling cells capable of sphere formation (6–8).

We then performed double staining of memGRP78 and twostem cellmarkers [Octamer-binding transcription factor 4 (OCT4;ref. 38) and CD133 (Prom1; ref. 39)]. Acellular ascites treatmentfor 7days increased theOCT4/memGRP78andCD133/memGRP78double-positive populations. ID8 cells isolated from asciteshad 4.9% memGRP78þ/OCT4þ cells (Fig. 4C) and 7.8%CD133þ/memGRP78þ cells (Fig. 4C).

Limiting dilution transplantation assay is a standard methodfor assessing tumor-initiating activity associated with cancerstemness. We investigated the ability of memGRP78þ and mem-

GRP78� cells to initiate tumor growth when injected at cellnumbers ranging from 103 to 106. More mice developed tumorsor tumor-associated ascites in memGRP78þ cells injected groupsthan those injected with memGRP78� cells at 103 to 105 injectionnumbers (Table 1). The likelihood ratio test of single-hit modelwas performed as a standard statistical analysis method forlimiting dilution studies of stem cells (40) and detected P <0.05 between memGRP78þ versus memGRP78� cells. Moreover,mice bearing 106 and 105 memGRP78þ cells died sooner thanthose bearing memGRP78� cells (Fig. 4D).

GRP78 antibodies suppress sphere formationMonoclonal antibodies developed by our laboratory (C38 and

C107), and a commercial antibody (C20) directed against theCOOH-terminal domain of GRP78, suppress tumor growth in

Figure 3.Ascites increases memGRP78expression on ID8 cells. MemGRP78expression was assessed by flowcytometry using C20 antibody. Top,percentage of memGRP78þ cells minusbackground in ID8 cells (Parental), ID8cells freshly isolated from murineascites (In vivo), and ID8 cells isolatedfrom ascites after being reculturedin normal medium for 7 days(Recultured) are shown. Bottom, ID8cells were incubated with 50% ascitesfor 4 hours or 7 days. Ascites wasremoved, and cells were recultured innormal medium for 9 days (Ascites off9 days).

Mo et al.





melanoma and prostate models (13, 15). We studied effects ofthese antibodies on ovarian cancer cells. ID8 cells were treated for7 days with 50% acellular ascites. Antibodies (C38, C107, isotypecontrol at 10 mg/mL; C20 or goat IgG at 5 mg/mL) were added onday 5. On day 7, each group was harvested and sphere-formingability was determined. Ascites treatment significantly increasedsphere number. Notably, C38 and C20 antibodies, but not C107,decreased sphere numbers (Fig. 5A).

To test antibody activity in vivo, 400 mg antibody permousewasinjected intraperitoneally 5 days before implanting 105 ascitestumor cells. Following implantation, 200 mg antibody was deliv-ered intraperitoneally every other week until the endpoint. Micereceiving C38 or C107 antibodies had significantly lengthenedsurvival (C38 median ¼ 64 days, C107 median ¼ 64 days)compared with those receiving IgG2b (median ¼ 55 days;Fig. 5B).

We next investigated signaling pathways associated withmemGRP78-expressing ovarian CSCs. Ascites treatment of ID8cells significantly increased AKT and GSK3a phosphorylation(Fig. 5C), and GSK3a phosphorylation/inactivation led toincreased SNAI1 levels (41, 42), associated with enhanced CSCbehaviors (35, 43). Notably, C20 blocked these signaling eventsby binding to memGRP78 (Fig. 5C and D; refs. 13, 14). These datasuggest that a memGRP78–AKT–GSK3–SNAI1 signaling axis isassociatedwith stemness of ascites-derivedmurine ovarian cancercells.

To begin to validate our findings in human ovarian cancer celllines, we incubated OvCar3 or ES2 cells with acellular humanascites and treated them with either C38, C107, C20, or isotypecontrol IgG. C38, C107, and C20 suppressed sphere-formingability of ascites-pretreated OvCar3 and ES2 cells (Fig. 5E and F).

DiscussionWe demonstrate that treatment of murine ovarian cancer cells

with acellular ascites enriches for memGRP78þ cells with CSCproperties. We also show an ability of human ascites to enrich forhuman ovarian CSCs in vitro. Further studies are needed to showthat ascites enriches for a CSC population relevant to humanovarian cancer disease. Collectively, our findings raise the impor-tant question ofwhether blocking the occurrence of ascites and/or

Figure 4.MemGRP78þ cells exhibit increased sphere-forming ability/tumor-initiatingactivity compared with memGRP78� cells. A, gating for memGRP78þ andmemGRP78� cells from ascites cells generated in ID8 cell–bearing mice. B,sphere-forming ability of memGRP78þ and memGRP78� gated tumor cells. 103

cells perwell were seeded in a sphere assay. Primary sphereswere trypsinizedand passaged to form secondary spheres. Y axis represents spheres� 50 mm

� SD. Similar results were obtained in 3 experiments. C, bar graph showingthe double-positive population for GRP78 and OCT4 or CD133 (n ¼ 2). D,survival ofmice injectedwith 106 or 105 memGRP78þ tumor cell was shortenedcompared to survival of mice injected with the same number of memGRP78�

tumor cells (n¼ 10). Log-rank test was performed between memGRP78þ andmemGRP78� groups at either cell number trial.

Table 1. MemGRP78þ or memGRP78� tumor cells were sorted from ascitesgenerated in ID8 cell–bearing mice

memGRP78 positive memGRP78 negative

Cells injectedTumordetected

Tumortake rate

Tumordetected

Tumortake rate

106 9/9 100% 6/6 100%105 10/10 100% 7/10 70%104 7/10 70% 5/10 50%103 2/10 20% 0/10 0

NOTE: Mice were injected intraperitoneally with 103 to 106 of these sorted cellpopulations, and tumor take rate was determined. The likelihood ratio test ofsingle-hit model was performed and showed P < 0.05.






Figure 5.GRP78 antibodies suppress ovarian cancer sphere-forming ability and prolong mouse survival in a syngeneic model of ovarian cancer. A, ID8 cells werepretreated for 7 days with ascites obtained from ID8-bearing mice plus 10 mg/mL C38, mouse isotype control (IgG2b), 5 mg/mL C20, goat IgG (Goat), or no addedantibody (Blank) for 2 days. Y axis represents spheres� 50 mm. Error bars, SD from 3 trials in triplicate. B, mouse study was performed on 4 groups of mice (n¼ 10/group) receiving C38, C107, IgG2b control, or PBS. We injected 400 mg antibody or PBS per mouse 5 days before implanting 105 ID8 cells isolated from ascites.Following implantation, 200mg antibody or 200mLPBSwas delivered intraperitoneally every otherweek until the endpoint. C, C20or goat IgGwas added to ID8 cellsfrom medium or 7-day ascites pretreatment conditions. After 2 days, cells were lysed and extracts probed for phospho-AKT, phospho-GSK3, GSK-3, andSNAI1 byWestern blotting. A representative blot is shown. D, densitometry of three independent experiments (Western blotting). E and F, OvCar3 or ES2 cells wereincubated with Ov480 ascites for 7 days plus 10 mg/mL C38, C107, mouse isotype control (IgG2b), 5 mg/mL C20, goat IgG, or no antibody (Blank) for 2 days.Y axis represents spheres �50 mm. Error bars represent SD from 3 trials in triplicate.

Mo et al.





clinical removal of ascites may be a useful adjuvant to otherovarian cancer therapies.

We are currently investigating which ascites soluble factor(s)maintain ovarian CSCs. GRP78 ligands are potential candi-dates. We identified in murine ascites several GRP78 ligands[anti-GRP78 antibodies, alpha-2-macroglobulin (a2M) andmurinoglobulin] that activate the amino terminal domain ofGRP78 (Supplementary Fig. S5; ref. 11). Further studies areneeded to determine the importance of these ascites factors inthe maintenance of ascites-enriched ovarian CSC.

GRP78 is generally associated with proproliferative activities(11, 12). However, CSCs are slow-cycling cells (6, 7). Althoughforming tumors faster in vivo, memGRP78þ cells have a slowerproliferation rate than memGRP78� cells (Supplementary Fig. S4).Interestingly, the DiD retention study showed that most of theascites-pretreated ID8 cells were slow-cycling cells (Fig. 1E),correlating with the finding that memGRP78þ ovarian cancer cellsare slowly proliferating.

Our studies show that memGRP78þmurine ovarian cancer cellsexhibit increased tumorigenicity comparedwithmemGRP78� cells(Table 1). These results confirm similar findings in head and neckcancer (21), but differ from a study showing that GRP78þ coloncancer cells exhibit reduced tumorigenicity compared withGRP78� cells (44). We attribute differences between these studiesto the fact that our work, but not the study of Hardy andcolleagues, excluded dead cells during sorting. Dead cells exposetheir endoplasmic reticulum, which is a major source of GRP78.The findings of Hardy and colleagues (44) are likely complicatedby their inclusion of a large percentage of dead GRP78þ cells intheir tumor injection study.

Cancer stem cells are a heterogeneous population consisting ofcells in different differentiation stages, each stage expressingdistinct CSCmarkers (45). The difference in% SCA1þ cells foundin vivo versus in vitro likely reflects the fact that (i) SCA1 isexpressed only during specific CSC differentiation stages and (ii)this SCA1-expressing stem cell subpopulation is representedmorefrequently in our in vitro ascites enrichment model than in the invivo model. This phenomenon may be attributable to microen-vironmental regulation of CSC differentiation state. In contrast,memGRP78 is expressed equally on cancer cells from our in vitromodel and from ascites cells in vivo, suggesting that memGRP78 is auniversal murine CSC marker.

We demonstrated that ascites increased cancer stem cell mar-kers (SOX9 and SNAI1) in ID8 cells (Fig. 2D). Activation of GSK3,a differentiation-related gene that is inactivated by AKT (46),reduces Snai-1 mRNA and SNAI1 protein levels (41, 42). Anti-

bodies against the COOH-terminal GRP78 domain blocked AKTand GSK3a phosphorylation, thus reducing SNAI1 expressionlevel and stem-cell activities. These data demonstrate efficacy ofthese GRP78 antibodies in reducing CSC markers in murineovarian cancer cells.

Our in vivo study, showing that a GRP78 COOH terminaldomain antibody (C38) prolonged survival of ovarian cancer–bearingmice, indicates a potential clinical application of targetingmemGRP78. Although C107 only showed a modest effect onsphere formation in vitro (Fig. 5A), mice receiving this antibodysurvived longer than control mice. This finding may be attribut-able to different conformations of memGRP78 on tumor cells invitro versus in vivo, with the epitope forC107beingmore accessiblein vivo. We believe that efficacy of GRP78 antibodies will beincreased when combined with chemotherapy. According to ourmodel, chemotherapy should target the fast proliferating tumorbulk cells, whereas anti-COOH terminal domain GRP78 anti-bodies will target CSCs. Combination studies are currently underinvestigation.

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

Authors' ContributionsConception and design: L. Mo, R.E. Bachelder, A. Berchuck, S.V. PizzoDevelopment of methodology: L. Mo, R.E. Bachelder, S.V. PizzoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): L. Mo, A. BerchuckAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): L. Mo, R.E. Bachelder, M. Kennedy, P.-H. Chen,J.-T. Chi, S.V. PizzoWriting, review, and/or revision of the manuscript: L. Mo, R.E. Bachelder,M. Kennedy, A. Berchuck, G. Cianciolo, S.V. PizzoAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): L. Mo, R.E. Bachelder, M. Kennedy, P.-H. Chen,J.-T. Chi, A. Berchuck, G. Cianciolo, S.V. PizzoStudy supervision: R.E. Bachelder, S.V. Pizzo

AcknowledgmentsThe authors thank the Duke Microarray Core facility for technical support

and the generation of microarray data. They also thank Dr. Xiao-fan Wang, Dr.Gustaaf de Ridder, Dr. Rupa Ray, and Sturgis Payne for providing comments.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 3, 2014; revised December 23, 2014; accepted December 29,2014; published OnlineFirst January 14, 2015.

References1. Cancer.org [Internet]. Atlanta: American Cancer Society, Inc.; c1913

[updated 2014 August 11;cited 2014 September 19]. Availablefrom: http://www.cancer.org/cancer/ovariancancer/detailedguide/ovari-an-cancer-key-statistics

2. Hennessy BT, Coleman RL,MarkmanM.Ovarian cancer. Lancet 2009;374:1371–82.

3. Burges A,Wimberger P, KumperC,GorbounovaV, SommerH, SchmalfeldtB, et al. Effective relief of malignant ascites in patients with advancedovarian cancer by a trifunctional anti-EpCAM x anti-CD3 antibody: a phaseI/II study. Clin Cancer Res 2007;13:3899–905.

4. Jayson GC, Kohn EC, Kitchener HC, Ledermann JA. Ovarian cancer. Lancet2014;384:1376–88.

5. Kwon MJ, Shin YK. Regulation of ovarian cancer stem cells or tumor-initiating cells. Int J Mol Sci 2013;14:6624–48.

6. Moore N, Lyle S. Quiescent, slow-cycling stem cell populations incancer: a review of the evidence and discussion of significance. J Oncol2011;2011.

7. Li L, Bhatia R. Stem cell quiescence. Clin Cancer Res 2011;17:4936–41.8. Reya T,Morrison SJ, ClarkeMF,Weissman IL. Stem cells, cancer, and cancer

stem cells. Nature 2001;414:105–11.9. Lee AS. GRP78 induction in cancer: therapeutic and prognostic implica-

tions. Cancer Res 2007;67:3496–9.10. Defresne F, Bouzin C, Guilbaud C, Dieu M, Delaive E, Michiels C, et al.

Differential influence of anticancer treatments and angiogenesis on the






seric titer of autoantibody used as tumor and metastasis biomarker.Neoplasia 2010;12:562–70.

11. Quinones QJ, de Ridder GG, Pizzo SV. GRP78: a chaperone with diverseroles beyond the endoplasmic reticulum. Histol Histopathol 2008;23:1409–16.

12. Misra UK, Deedwania R, Pizzo SV. Activation and cross-talk between Akt,NF-kappaB, and unfolded protein response signaling in 1-LN prostatecancer cells consequent to ligation of cell surface-associated GRP78. J BiolChem 2006;281:13694–707.

13. Misra UK, Mowery Y, Kaczowka S, Pizzo SV. Ligation of cancer cell surfaceGRP78 with antibodies directed against its COOH-terminal domain up-regulates p53 activity and promotes apoptosis. Mol Cancer Ther 2009;8:1350–62.

14. Misra UK, Pizzo SV. Ligation of cell surface GRP78 with antibodydirected against the COOH-terminal domain of GRP78 suppressesRas/MAPK and PI 3-kinase/AKT signaling while promoting caspaseactivation in human prostate cancer cells. Cancer Biol Ther 2010;9:142–52.

15. de Ridder GG, Ray R, Pizzo SV. A murine monoclonal antibody directedagainst the carboxyl-terminal domain of GRP78 suppresses melanomagrowth in mice. Melanoma Res 2012;22:225–35.

16. Luo S, Mao C, Lee B, Lee AS. GRP78/BiP is required for cell proliferationand protecting the inner cell mass from apoptosis during early mouseembryonic development. Mol Cell Biol 2006;26:5688–97.

17. Spike BT, Kelber JA, Booker E, KalathurM, Rodewald R, Lipianskaya J, et al.CRIPTO/GRP78 signaling maintains fetal and adult mammary stem cellsex vivo. Stem Cell Rep 2014;2:427–39.

18. Wey S, Luo B, Lee AS. Acute inducible ablation of GRP78 reveals its role inhematopoietic stem cell survival, lymphogenesis and regulation of stresssignaling. PloS One 2012;7:e39047.

19. Huang LW, Lin CY, Lee CC, Liu TZ, Jeng CJ. Overexpression of GRP78 isassociated with malignant transformation in epithelial ovarian tumors.Appl Immunohistochem Mol Morphol 2012;20:381–5.

20. Wey S, Luo B, Tseng CC, Ni M, Zhou H, Fu Y, et al. Inducible knockout ofGRP78/BiP in the hematopoietic system suppresses Pten-null leukemo-genesis and AKT oncogenic signaling. Blood 2012;119:817–25.

21. WuMJ, JanCI, Tsay YG, YuYH,HuangCY, Lin SC, et al. Eliminationof headand neck cancer initiating cells through targeting glucose regulated pro-tein78 signaling. Mol Cancer 2010;9:283.

22. de Ridder GG, Gonzalez-Gronow M, Ray R, Pizzo SV. Autoantibodiesagainst cell surface GRP78 promote tumor growth in a murine model ofmelanoma. Melanoma Res 2011;21:35–43.

23. Roby KF, Taylor CC, Sweetwood JP, Cheng Y, Pace JL, Tawfik O, et al.Development of a syngeneic mouse model for events related to ovariancancer. Carcinogenesis 2000;21:585–91.

24. Arap MA, Lahdenranta J, Mintz PJ, Hajitou A, Sarkis AS, Arap W, et al. Cellsurface expression of the stress response chaperone GRP78 enables tumortargeting by circulating ligands. Cancer Cell 2004;6:275–84.

25. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarraysapplied to the ionizing radiation response. Proc Natl Acad Sci U S A2001;98:5116–21.

26. Tang X, Lucas JE, Chen JL, LaMonte G, Wu J, Wang MC, et al. Functionalinteraction between responses to lactic acidosis and hypoxia regulatesgenomic transcriptional outputs. Cancer Res 2012;72:491–502.

27. Chang JT, Nevins JR. GATHER: a systems approach to interpreting genomicsignatures. Bioinformatics 2006;22:2926–33.

28. Chi JT, Wang Z, Nuyten DS, Rodriguez EH, Schaner ME, Salim A, et al.Gene expression programs in response to hypoxia: cell type specifi-

city and prognostic significance in human cancers. PLoS Med 2006;3:e47.

29. Chen JL, Merl D, Peterson CW, Wu J, Liu PY, Yin H, et al. Lactic acidosistriggers starvation response with paradoxical induction of TXNIP throughMondoA. PLoS Genet 2010;6:e1001093.

30. Jensen JB, Parmar M. Strengths and limitations of the neurosphere culturesystem. Mol Neurobiol 2006;34:153–61.

31. Klonisch T, Wiechec E, Hombach-Klonisch S, Ande SR, Wesselborg S,Schulze-Osthoff K, et al. Cancer stem cell markers in common cancers -therapeutic implications. Trends Mol Med 2008;14:450–60.

32. Xin L, Lawson DA, Witte ON. The Sca-1 cell surface marker enriches for aprostate-regenerating cell subpopulation that can initiate prostate tumor-igenesis. Proc Natl Acad Sci U S A 2005;102:6942–7.

33. Grange C, Lanzardo S, Cavallo F, Camussi G, Bussolati B. Sca-1 identifiesthe tumor-initiating cells in mammary tumors of BALB-neuT transgenicmice. Neoplasia 2008;10:1433–43.

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

35. Kurrey NK, Jalgaonkar SP, Joglekar AV, Ghanate AD, Chaskar PD,Doiphode RY, et al. Snail and slug mediate radioresistance and che-moresistance by antagonizing p53-mediated apoptosis and acquiringa stem-like phenotype in ovarian cancer cells. Stem Cells 2009;27:2059–68.

36. Guo W, Keckesova Z, Donaher JL, Shibue T, Tischler V, Reinhardt F, et al.Slug and Sox9 cooperatively determine the mammary stem cell state. Cell2012;148:1015–28.

37. Raha D, Wilson TR, Peng J, Peterson D, Yue P, Evangelista M, et al. Thecancer stem cell marker aldehyde dehydrogenase is required to main-tain a drug-tolerant tumor cell subpopulation. Cancer Res 2014;74:3579–90.

38. Peng S,Maihle NJ, Huang Y. Pluripotency factors Lin28 andOct4 identify asub-population of stem cell-like cells in ovarian cancer. Oncogene 2010;29:2153–9.

39. Baba T, Convery PA, Matsumura N, Whitaker RS, Kondoh E, Perry T, et al.Epigenetic regulation of CD133 and tumorigenicity of CD133þ ovariancancer cells. Oncogene 2009;28:209–18.

40. Hu Y, Smyth GK. ELDA: extreme limiting dilution analysis for comparingdepleted and enriched populations in stem cell and other assays. J Immu-nol Methods 2009;347:70–8.

41. Bachelder RE, Yoon SO, Franci C, de Herreros AG, Mercurio AM. Glycogensynthase kinase-3 is an endogenous inhibitor of Snail transcription: implica-tions for the epithelial-mesenchymal transition. J Cell Biol 2005;168:29–33.

42. Zhou BP, Deng J, Xia W, Xu J, Li YM, Gunduz M, et al. Dual regulation ofSnail by GSK-3beta-mediated phosphorylation in control of epithelial-mesenchymal transition. Nat Cell Biol 2004;6:931–40.

43. Fan F, Samuel S, Evans KW, Lu J, Xia L, Zhou Y, et al. Overexpression ofsnail induces epithelial-mesenchymal transition and a cancer stem cell-like phenotype in human colorectal cancer cells. Cancer Med 2012;1:5–16.

44. Hardy B, Raiter A, YakimovM,VilkinA,Niv Y. Colon cancer cells expressingcell surface GRP78 as a marker for reduced tumorigenicity. Cell Oncol2012;35:345–54.

45. TangDG. Understanding cancer stem cell heterogeneity and plasticity. CellRes 2012;22:457–72.

46. Force T, Woodgett JR. Unique and overlapping functions of GSK-3 iso-forms in cell differentiation and proliferation and cardiovascular devel-opment. J Biol Chem 2009;284:9643–7.


Mo et al.




2015;14:747-756. Published OnlineFirst January 14, 2015.Mol Cancer Ther Lihong Mo, Robin E. Bachelder, Margaret Kennedy, et al. GRP78

MembraneEnriches for Ovarian Cancer Stem-Like Cells Expressing Syngeneic Murine Ovarian Cancer Model Reveals That Ascites

Updated version

10.1158/1535-7163.MCT-14-0579doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2015/01/14/1535-7163.MCT-14-0579.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/14/3/747.full#ref-list-1

This article cites 44 articles, 13 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/14/3/747.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://mct.aacrjournals.org/content/14/3/747To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-14-0579

http://mct.aacrjournals.org/content/suppl/2015/01/14/1535-7163.MCT-14-0579.DC1

http://mct.aacrjournals.org/content/14/3/747.full#ref-list-1

http://mct.aacrjournals.org/content/14/3/747.full#related-urls

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

mailto:[email protected]

http://mct.aacrjournals.org/content/14/3/747


Documents

Syngeneic Murine Ovarian Cancer Model Reveals That Ascites … · ascites were associated with enhanced sphere-forming ability (Supplementary Fig. S3A and S3B). We chose 50% ascites