Tumor and Stem Cell Biology

EGFL6 Regulates the Asymmetric Division,Maintenance, and Metastasis of ALDHþ OvarianCancer CellsShoumei Bai1, Patrick Ingram2, Yu-Chih Chen2, Ning Deng1, Alex Pearson1,Yashar S. Niknafs1, Patrick O'Hayer1, Yun Wang1, Zhong-Yin Zhang3, Elisa Boscolo4,Joyce Bischoff5, Euisik Yoon2, and Ronald J. Buckanovich1,6

Abstract

Little is known about the factors that regulate the asymmetricdivision of cancer stem–like cells (CSC).Here, we demonstrate thatEGFL6, a stem cell regulatory factor expressed in ovarian tumorcells and vasculature, regulates ALDHþ ovarian CSC. EGFL6 sig-naled at least in part via the oncoprotein SHP2 with concomitantactivation of ERK. EGFL6 signaling promoted the migration andasymmetric division of ALDHþ ovarian CSC. As such, EGFL6increased not only tumor growth but also metastasis. Silencing of

EGFL6 or SHP2 limited numbers of ALDHþ cells and reducedtumor growth, supporting a critical role for EGFL6/SHP2 inALDHþ cell maintenance. Notably, systemic administration of anEGFL6-neutralizing antibodywe generated restricted tumor growthand metastasis, specifically blocking ovarian cancer cell recruit-ment to the ovary. Together, our results offer a preclinical proof ofconcept for EGFL6 as a novel therapeutic target for the treatmentof ovarian cancer. Cancer Res; 76(21); 6396–409. �2016 AACR.

IntroductionAlthough controversy persists regarding cancer stem–like cells

(CSC), ALDH-expressing CSC have been linked with ovariancancer chemotherapy resistance, disease recurrence (1–4), andmetastasis (5). We reported an epithelial ovarian cancer (EOC)cell differentiation hierarchy consistent with a CSC model (1).Within this hierarchy, ALDHþCD133þ cells and ALDHþCD133�

cells have the greatest cancer-initiating capacity (2), indicatingthat ALDHþ cells are important for ovarian cancer biology.

Normal stem cells are closely associated with vascular cells in a"stem cell niche" (6, 7). Like normal stem cells, CSC reside in aperivascular location (8, 9). Tumor vascular cells secrete "angio-crine" factors (10), which promote stem cell proliferation. Little isknown about the vascular factors that regulate CSC. Two studiesidentifiedEGFL6 in tumor vascular cells of EOC(11, 12). EGFL6 isa secreted protein (13) that regulates stem cell proliferation and

differentiation in different biological systems. EGLF6 regulatesstem cells in hair follicle morphogenesis (14, 15), stimulatesendothelial cell migration/proliferation in a p-ERK-dependentmanner during osteoblast differentiation (16), and promotes theadhesion and proliferation of stromal vascular cells during adi-pocyte differentiation (17).

Herein, we evaluate the role of EGFL6 in ovarian cancer. Wefind EGFL6 is expressed in tumor vascular cells and in somecancer cells. We demonstrate in vitro that EGFL6 inducesALDHþ ovarian CSC to undergo asymmetric division. EGFL6signaling is mediated in part via integrin-dependent activationof the phosphatase SHP2 and ERK. EGFL6 or SHP2 knock-down/inhibition is associated with a significant reduction inALDHþ cells and a reduction in tumor growth. EGFL6 expres-sion in vascular cells increases tumor growth and metastasis.EGFL6 blockade reduces cancer growth and reduces metastasis.Interestingly, EGFL6 blockade completely eliminated metasta-ses to the ovary, suggesting that EGFL6 might play a critical rolein the recruitment of cancer cells to the ovary. Together, ourresults indicate that EGFL6 is a novel tumor and angiocrinefactor that regulates ALDHþ cell asymmetric division, migra-tion, and metastasis. EGFL6 thus represents a potential thera-peutic target in ovarian cancer.

Materials and MethodsPrimary tumor processing

All studies were approved by the Institutional Review Boardof the University of Michigan, and tumors were obtainedwith informed patient consent. All tumors were stage III or IVhigh-grade serous ovarian (HGSC) or primary peritoneal can-cer. Single-cell isolation from tumor tissues and ascites was asdescribed (2, 18).

Cell culture, tumor sphere culture, and treatmentCulture methods are detailed in Supplementary Methods.

1Division of Hematology-Oncology, Department of Internal Medicine, Universityof Michigan, Ann Arbor, Michigan. 2Department of Electrical Engineering,University of Michigan, Ann Arbor, Michigan. 3Department of Biochemistry andMolecular Biology, Indiana University School of Medicine, Indianapolis, Indiana.4Department of Pediatrics, Cincinnati Children's Hospital Medical Center, Cin-cinnati, Ohio. 5Department of Surgery, Boston Children's Hospital and HarvardMedical School, Boston, Massachusetts. 6Division of Gynecologic-Oncology,Department of Obstetrics and Gynecology, University of Michigan, Ann Arbor,Michigan.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Current address of N. Deng: Lifescience and Technology School, Jinan Univer-sity, Guangzhou, China.

Corresponding Author: Ronald J. Buckanovich, University of Michigan, 5219Cancer Center, 1500 E Medical Center Drive, Ann Arbor, MI 48109-5940. Phone:734-764-2395; Fax: 734-936-7376; E-mail: ronaldbu@med.umich.edu

doi: 10.1158/0008-5472.CAN-16-0225

�2016 American Association for Cancer Research.

CancerResearch

Cancer Res; 76(21) November 1, 20166396

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Quantitative real-time PCRcDNA synthesis, PCR, and primer information are described

in Supplementary Methods.

Tissue microarray stainingA tissue microarray (TMA) contained primary debulking

tissues from 154 chemotherapy-na€�ve ovarian cancer patients.Note that 12.5%, 10.7%, 66.1%, and 10.7% patients hadstage I–IV disease, respectively. Median age was 58 years(minimum, 30; maximum, 84). TMA sections were processedas described (2) with two anti-EGL6 antibodies (Sigma;1:200; and a mouse anti-EGFL6 we generated, 1:400). Tumorswere scored by two reviewers. Tumors were scored as EGFL6þ

if vascular EGFL6 expression was detected in either primarytumor or metastatic sites. The method of Kaplan and Meierwas used to estimate overall and recurrence-free survival.Follow-up time was calculated from the date of diagnosis/staging surgery until the date of first documented relapse ordeath. Data were censored at 5 years. The log-rank test wasconducted to test for a significant difference (P < 0.05)between groups. We used the Cox proportional hazardsmodel to assess individual variable effect on time-to-eventoutcome. Statistical programming was performed using Rversion 3.0.1.

BioinformaticsFor EGFL6-expression analysis in normal ovary and different

ovarian cancer histologies, gene expression data were obtainedfrom ONCOMINE (gene accession #NM_015507, Probe ID219454_at, Hendrix dataset, Affymetrix HG_U133A array;ref. 19). Detailed methodologies are provided in SupplementaryMethods.

EGFL6-expressing cell linesEGFL6 was cloned into p3xFlag and pRSV-GFP vectors.

SKOV3 cells were transfected with EGFL6-p3xFLAG usingFuGene 6 reagent (Promega) per protocol. EGFL6-expressingclones were selected by G418 treatment and confirmed byWestern blotting with Flag antibody. Transduced cell linesexpressing EGFL6 or control were obtained by lentiviral infec-tion followed by FACS sorting of GFP-positive cells.

EGFL6 productionHEK293 cells were transiently transfected with EGFL6 or

empty vector plasmid using FuGENE 6 reagent as above inRPMI-5% FBS. Supernatant was collected at 36 hours, EGFL6secretion was confirmed via Western blotting analysis, andsupernatant was used for cell treatment. For purification, celllysates of transiently transfected HEK293 cells were loadedonto the FLAG M2 Affinity Gel (Sigma) column under gravityflow 4�C, washed with TBS, and FLAG-EGFL6 protein elutedwith 0.1 mol/L glycine HCl, pH 3.5, and neutralized with1 mol/L Tris, pH 8.0. Unless otherwise indicated, EGFL6treatment in in vitro studies was daily for 72 hours.

Cell-cycle analysisSKOV3 cells were synchronized by serum starvation for 24

hours, treated with EGFL6 or control for 24 hours, fixed with ice-cold ethanol, washed with PBS, and then stained with propidiumiodide (10mg/mL) andRNase-A (100mg/mL) inPBS, followedbyFACS analysis and analysis by FlowJo.

Flow cytometry and FACSNIHOVCAR3, SKOV3 cells, or primary ovarian tumor/ascites

cells were stained with DAPI and ALDEFLUOR (Stem Cell Tech-nologies) as previously described (2). For FACS isolation, equalnumbers of ALDHþ and ALDH(�) cells were collected for subse-quent experiments.

Microfluidics cultureCells were FACS-isolated and loaded into the microfluidic

device as previously described (1) and photographed to con-firm ALDH expression. Twelve hours after loading, cells weretreated with every 12-hour EGFL6 or vehicle. After 48 or96 hours of treatment (cell lines vs. primary cells, respectively),cells were restained with ALDEFLUOR in situ and photo-graphed. All samples were evaluated in at least two replicatedevices from at least two experiments. A Student t test wasused to compare results for each sample. ANOVA analysis wasused to determine statistically significant findings of aggregateanalyses.

Microfluidic migration assays were performed in RPMI-5%FBS as previously described (20). Loading was photographed toconfirm cell location. EGFL6-containing media or controlmedia were provided in right inlet. Chips were reimaged after24 hours and distance migrated measured. Results presentedrepresent mean� SDs. A two-tailed Student t test (unpaired)was used to measure significance.

Immunohistochemistry and immunofluorescenceExperiments were performed as previously described

(2, 21). The primary antibodies are provided in SupplementaryMethods.

Western blottingCells were lysed in NP40 cell lysis buffer. Antibodies used are

provided in Supplementary Methods.

EGFL6 antibody purification and treatmentHybridoma cells were maintained in RPMI-5% FBS. Super-

natant was collected after approximately 7 days, cells/debriswere removed by centrifugation, and supernatant loaded onProtein G Agarose (Millipore), washed, and antibody elutedwith 50 mmol/L glycine, pH 2.7, neutralized to PH 7.2–7.4with 1 mol/L Tris (Ph 9.0). EGFL6 antibody was applied at10 mg/mL for in vitro cell treatment and 10 mg/kg twice weeklyfor in vivo treatment.

EGFL6 and SHP2 shRNA knockdownEGFL6 or SHP2 shRNA (Sigma) and control scrambled

shRNA were cotransfected with Pspax2 and pMD2.G lentiviralpackaging plasmids into HEK293T to produce lentivirus.NIHOVCAR3 or SKOV3 cells were transduced with shRNAstargeting EGFL6 or SHP2 or control, and selected with puro-mycin. RNA and cell lysates were prepared to confirm EGFL6 orSHP2 knockdown via quantitative real-time PCR (qRT-PCR)and Western blot.

Murine tumor modelsAnimal experiments were approved by the University Com-

mittee for Use and Care of Animals. For transfected EGFL6-SKOV3 tumor model, three G418-selected EGFL6-SKOV3clones or three empty-vector clones, at 1 � 106 each, were

EGFL6 Regulates Asymmetric Division

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6397

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

injected into mouse axilla. Tumor volume was monitoredweekly, and mice were euthanized when tumors reachedapproximately 1,000 mm3 (using L � W � W/2 calculation).EGFL6shRNA or scrambledshRNA control cells (1 � 106 SKOV3 or1 � 106 NIHOVCAR3) were injected and tumors monitoredas above. For intraperitoneal (IP) tumor models, 5 � 105

SKOV3 cells were injected IP into NSG mice. Mice were treatedwith mIgG or anti-EGFL6 biweekly, as above. Mice were sacri-ficed when IP tumor nodules reached approximately 400 mm3

or mice had significant weight gain/loss due to tumor/ascitesburden. At the time of euthanasia, mice were inspected forgross metastasis. Lung, liver, ovary, and omentum/intestineswere resected for histologic analysis of metastatic disease. Forintravenous tumor cell injection, mice were mock-treated ortreated with anti–EGFL6-blocking antibody for 3 days prior totumor cell injection. SKOV3 cells (1 � 106) with mIgG or anti-EGFL6 were injected via tail vein into NSG mice. Therapy wasmaintained for 2 weeks after tumor cell injection. Mice weremonitored, euthanized, and metastases assessed as describedabove.

Establishment of HemSCEGFL6 tumor modelHemangioma stem cells (HemSC) were transduced with

EGFL6-pRSVGFP or control lentivirus and FACS-sorted basedon GFP expression. EGFL6 expression was confirmed by West-ern blot. Lucif-SKOV3 cells (2 � 105) were mixed with EGFL6-expressing HemSC (HemSCEGFL6, 1 � 106) or control HemSC(HemSCControl). For primary cell–derived tumors, 5 � 105

primary cells (from 2 patients, injected bilaterally in 4 mice)were mixed with either 5 � 105 HemSCEGFL6 or 5 � 105

HemSCControl cells; then injected into axilla of NOD-SCIDmice. Three days after tumor injection, mice were treated withmIgG or EGFL6 antibody (5 mg/kg) bi-weekly. At the time ofeuthanasia, SKOV3 flank tumors were resected and luciferaseimaging was performed to identify microscopic metastases.From all animals, lung, liver, ovary, and omentum/intestineswere resected for histologic analysis of metastatic disease.

ResultsEGFL6 is upregulated in ovarian cancer and predicts a poorpatient prognosis

We analyzed EGFL6 mRNA expression across ovarian can-cer histologies in two different datasets (19). Compared withnormal ovary, EGFL6 mRNA levels are significantly elevatedin all ovarian tumor histologies, but are greatest in HGSC(Fig. 1A). We next stained a TMA of 154 ovarian tumorswith EGFL6 antibody. EGFL6 protein was detected in tumorendothelial cells in approximately 80% of tumors (Fig. 1Band C). EGFL6 was also detected in tumor-associated adipose(Fig. 1B) and, surprisingly, in cancer cells in 25% of tumorsevaluated (Fig. 1B). EGFL6 expression in cancer cells wasprimarily in HGSC tumors, whereas EGFL6 expression inother histologies was primarily in the stroma (SupplementaryFig. S1A).

Based on prior studies, tumor cell EGFL6 expression wasunexpected. However, analysis of EGFL6 expression in humanovarian cancer cell lines using the Cancer Cell Line Encyclope-dia (CCLE) supported EGFL6 expression in many HGSCcell lines (Supplementary Fig. S1B). qRT-PCR analysis ofHGSC cell lines confirmed clear expression of EGFL6 in

NIHOVCAR3 and OVKATE HGSC cells (Fig. 1D). Little or noexpression was noted in other ovarian cancer lines, includingSKOV3 and CAOV3 cells.

Using the TMA, we scored vascular EGFL6 expression aspresent/absent in each tumor and performed univariate pro-portional hazards regression analysis to determine if vascularEGFL6 protein expression was a prognostic factor for ovariancancer. EGFL6 expression was related both to death (HR¼ 3.91,P ¼ 0.01) and recurrence (HR ¼ 2.67, P ¼ 0.01; Fig. 1E).Multiple proportional hazards regression demonstrated nodependence between EGFL6 and age, stage, tumor grade, ordebulking status.

We next screened RNAseq data from 261 serous ovarian car-cinomas (TCGA) and investigated the correlation of expression ofEGFL6 to all other genes in the genome. A P value cutoff (1e�8)was applied to generate a list of the 538 most correlated genes.Numerous tumor vascular–specific genes (including MXRA,TDO2, TNFAIP6, SEMA3D, and Coll11A1) and EGF-like genes(including EGFLAM and MEGF10) were highly correlatedwith EGFL6 expression. GSEA demonstrated that EGFL6-corre-lated genes were strongly correlated with invasive ovarian cancerand embryonic stem cell core genes (Fig. 1F and SupplementaryFig. S1C).

EGFL6 induces ovarian cancer cell proliferationWe expressed EGFL6 in non–EGFL6-expressing cell lines,

including HEK293 cells, NIH3T3 cells, and SKOV3 ovariancancer cells. Western blot confirmed secretion of EGLF6 in bothtransient and stable transfectants (Fig. 2A,i). We purified EGFL6protein to >95% purity (Fig. 2A,ii) and treated ovarian cancercells with either purified EGFL6, supernatant from EGFL6-expressing HEK293 cells, or supernatant from control-trans-fected HEK293 cells. Purified fusion protein and supernatantfrom EGFL6-transfected cells had similar effects. EGFL6 treat-ment of SKOV3, OVCAR3, OVCAR8, and primary ovarian tumorcells was associated with a 30% to 40% increase in total cellnumber (Fig. 2B). Cell-cycle analysis demonstrated that EGFL6treatment resulted in a 1.8-fold decrease in the number of cells inG1 phase, and a concomitant increase in the number of cells in Sand G2–M phases (Fig. 2C).

EGFL6 promotes asymmetric division of ALDHþ ovarian CSCGiven GSEA correlation of EGFL6 with a core stem cell

signature, we assessed the impact of EGFL6 on ovarian CSC.Aldehyde dehydrogenase enzymatic activity (ALDH) is anestablished marker of ovarian CSC (2–4, 22, 23). Treatmentof ovarian cancer cells with increasing concentrations ofEGFL6 was associated with increasing total cell numbers, butdecreasing percentages of ALDHþ CSC, with a resultant stableabsolute ALDHþ CSC number (Fig. 3A and B).

A dividing ALDHþ CSC can theoretically undergo at least threedistinct types of cell division related to the expression of ALDH:(i) division yielding 2 ALDHþ cells, (ii) differentiation yielding2 ALDH(�) cells, or (iii) asymmetric division yielding an ALDHþ

cell (self-maintenance) and an ALDH(�) cell. The finding thatEGFL6 increases total cancer cell number and decreases thepercentage of ALDHþ cells without affecting the absolute numberof ALDHþ cells is consistent with EGFL6-stimulating ALDHþCSCasymmetric division. Alternatively, EGFL6 could preferentiallypromote proliferation of ALDH(�) cells. We used single-cellmicrofluidic culture (24) to evaluate EGFL6 impact on

Bai et al.

Cancer Res; 76(21) November 1, 2016 Cancer Research6398

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Figure 1.

Expression of EGFL6 in ovarian tumors and normal tissue. A, expression of EGFL6 across ovarian cancers histologies in the Hendrix dataset (i), TCGA, GTEx(ii) datasets. B, IHC of EGFL6 in the indicated tissues. C, immunofluorescence of EGFL6 (red) and CD31 (green) in human tumor vasculature (blue,DAPI staining of nucleus). D, qRT-PCR analysis of EGFL6 expression in ovarian cancer cell lines, MCF7 (breast cancer), HEK293 (kidney), and HemSC controls.E, Kaplan–Meier curves for recurrence-free and overall survival for ovarian cancer patients with or without vascular EGFL6 expression in primary debulkingspecimens. F, GSEA demonstrating EGFL6-correlated genes are associated with invasive ovarian cancer phenotype and an embryonic stem cell coresignature. See Supplementary Fig. S1C for ES scores, P values, and false discovery rate q-values.

EGFL6 Regulates Asymmetric Division

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6399

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

asymmetric division. We FACS-sorted ALDHþ and ALDH(�)

SKOV3 cells into separate microfluidic devices and confirmedALDHexpression (Fig. 3C, i, left). Cells were thenmock-treated ortreated with EGFL6. After 48 hours, live cells were restained/imaged with ALDEFLUOR (Fig. 3C, i, right). Cell divisions, typeof daughter cells (ALDH(�) or ALDHþ), and total cell numberwere scored. Thirty-fivepercent of untreatedALDH(�) SKOV3 cellsdemonstrated no cell division, whereas 65% of cells underwentdivision to produce additional ALDH(�) cells, to yield an average2.2 daughter cells/well after 72 hours (Fig. 3C, ii). No ALDHþ

daughter cells were observed. EGFL6 treatment of ALDH(�)

SKOV3 cells had no significant impact on the number or type ofcell divisions (Fig. 3C, ii).

Compared with ALDH(�) cells, ALDHþ SKOV3 cells weremoreproliferative, with only 10% of untreated cells not dividing.Consistent with prior studies, ALDHþ cells demonstrated theability to produce both ALDHþ and ALDH(�) cells. ALDHþ cellscould divide to yield two ALDHþ cells (symmetric divisionrelative toALDH)or undergo an asymmetric division yieldingoneALDHþ cell and an ALDH(�) cell. EGFL6 treatment of ALDHþ

SKOV3 cells resulted in a 2-fold increase in the percentage ofALDHþ cells undergoing asymmetric division (Fig. 3C, ii). Thiswas associated with a statistically significant increase in total cellnumbers: 6.5 versus 3.7 average cells/well and a total of 325 versus185 daughter cells generated for every 50 captured cells in EGFL6-treated ALDHþ versus control ALDHþ cells, respectively. EGFL6

treatment of ALDH(�) SKOV3 cells had no significant impact onproliferation rates.

We confirmed these results with cells from three separateprimary ovarian cancer debulking specimens. Primary cellsdivided more slowly, and thus division was assessed after 96hours of EGFL6 treatment. For primary cells, we observed thatover 50% of untreated ALDH(�) cells underwent no division(Fig. 3C, iii). ALDH(�) cells that underwent cell division gen-erated only ALDH(�) progeny. EGFL6 treatment of ALDH(�)

cells was associated with a nonstatistically significant (P¼ 0.15)increase in average number of progeny/well and total cellnumbers (Fig. 3C, iii). As in cell lines, primary ALDHþ cellswere more proliferative than ALDH(�) cells with only 20% ofALDHþ cells not dividing (Fig. 3C, iii). EGFL6 treatment result-ed in an increase in number of progeny cells per well and totalcell numbers (Fig. 3C, iii). Likely due to the slower growth ofprimary cells, this did not reach statistical significance (P ¼0.09). EGFL6 treatment of primary ALDHþ cells was associatedwith a statistically significant (P ¼ 0.02) 1.9-fold increase inthe number of ALDHþ cells undergoing asymmetric division(Fig. 3C, iii).

EGFL6 signaling involves integrin binding and SHP2activation

EGFL6 activity is reported to be dependent on an intactRGD domain (16), suggesting signaling via integrins. To

Figure 2.

Effect of EGFL6 on ovarian cancer cell proliferation. A, i,Western blot of EGFL6 following control or EGFL6 transfectionof HEK293 cells; ii, Coomassie stain of the steps of EGFL6purification. B, total cell number for EGFL6 and vehiclecontrol–treated SKOV3 cells and primary tumor cells (PT112and PT122). C, cell-cycle analysis of EGFL6-treated SKOV3cells showing summary of three independent analyses (i) andrepresentative cell-cycle profile (ii). Experiments wereperformed in duplicate. Error bars, SDs.

Bai et al.

Cancer Res; 76(21) November 1, 2016 Cancer Research6400

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Figure 3.

EGFL6 promotes ALDHþ cell asymmetric division. A, summary of three replicate experiments demonstrating EGFL6 treatment is associated with increasingtotal cell numbers (i), decreasing percentages of ALDHþ cells (ii), but no change in absolute ALDHþ cell number (iii). B, percentages of ALDHþ primaryovarian cancer cells following treatment with EGFL6 or vehicle. C, single-cell microfluidic culture showing representative immunofluorescence images (i)demonstrating initial ALDEFLUOR stain (ALDHþ, green; ALDH(�), gray) in captured single cells and the observed types of cell division outcomes for ALDH(�)

vs. ALDHþ cells after capture, summary of percentages of division events and average number of progeny/microfluidic well with EGFL6 (ii and iii), or vehicletreatment of SKOV3 cells (ii) and three primary patient samples (iii). SKOV3 cells were analyzed in three independent experiments. Primary samples wereanalyzed in two independent experiments.

EGFL6 Regulates Asymmetric Division

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6401

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Bai et al.

Cancer Res; 76(21) November 1, 2016 Cancer Research6402

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

determine if EGFL6 signals via integrins in cancer cells, wegenerated an EGFL6 protein with an RGD-to-RGE mutation(EGFL6RGE). Mutation of the RGD domain eliminated theproliferative effects of EGFL6 (Fig. 4A).

To identify integrin family members involved in EGFL6 sig-naling, we performed qRT-PCR analysis of integrin family mRNAexpression in ALDHþ and ALDH(�) ovarian cancer cells. Wefound that integrin b3 (ITGB3), but not integrin b1 or b5, wasspecifically enriched in ALDHþ SKOV3 and A2008 cells (Fig. 4B).We tested the impact of the integrin b1/b3 inhibitor echistatin onovarian cancer cell response to EGFL6. Echistatin blocked bothEGFL6-mediated cancer cell proliferation and the decrease inALDHþ percentage (Fig. 4C). Interestingly, integrin b3–blockingantibody independently restricted ovarian cancer cell growth butonly partially prevented EGFL6-induced proliferation (Supple-mentary Fig. S2A). Finally, we performed coimmunoprecipitationstudies of EGFL6, EGFL6RGE, and integrin-b3. Wild-type EGFL6coimmunoprecipitated with integrin-b3 4.6-fold more effectivelythan EGFL6RGE. Interestingly, mixing wild-type EGFL6 withEGFL6RGE also compromised EGFL6 binding to integrin b3 (Sup-plementary Fig. S2B).

We next examined cellular signaling changes associated withEGFL6 treatment. Western blot analysis of ALDHþ andALDH(�) SKOV3 cells demonstrated 2.4-fold increased levelsof pSHP2 in ALDHþ cells versus ALDH(�) cells (Fig. 4D, i).EGFL6 treatment resulted in an additional 2-fold increase inpSHP2 levels specifically in ALDHþ cells (Fig. 4D, i). EGFL6treatment also resulted in a 4-fold increase in p-ERK levels inALDHþ cells, and a 1.8-fold increase in p-ERK in ALDH(�) cells(Fig. 4D, i). EGFL6RGE did not increase pSHP2 (Fig. 4D, ii) or p-ERK, although effects on p-ERK were more variable. EGFL6-mediated phosphorylation of SHP2 and ERK could be blockedby echistatin- or EGFL6-blocking antibodies (Fig. 4D, iii, seebelow for EGFL6-blocking antibody validation). Integrin b3–blocking antibodies blocked EGFL6-mediated increases inpSHP2, but only partly abrogated the increase in p-ERK (Sup-plementary Fig. S2C).

Direct interactions of SHP2 with Integrin protein complexeshave been reported (25). Immunoprecipitation of either integrinb3 or SHP2 confirmed interactions of the two proteins in ovariancancer cells (Supplementary Fig. S2D). Confirming a critical rolefor SHP2 in EGFL6 signaling on ovarian CSC, shRNA knockdownof SHP2 with three independent SHP2 shRNA (Fig. 4E, i) wasassociated with a significant decrease in ALDHþ cells in all cases(Fig. 4E, ii; Supplementary Fig. S3A). This is analogous to that seenin breast cancer (26). SHP2 knockdown was associated with asignificant decrease in total cell numbers (Fig. 4E, iii), and a 5- to8.5-fold decrease in the absolute number of ALDHþ cells (Fig. 4E).SHP2 knockdown eliminated EGFL6-mediated tumor cell prolif-eration (Fig. 4E, iii). Similarly, treatment of ovarian cancer cell

lines with the SHP2 inhibitor 11a-1 (27) resulted in a dose-dependent reduction in the total cell number, and percentageand absolute number of ALDHþ cells (Fig. 4F, i–ii; SupplementaryFig. S3B). Increasing doses of 11a-1 were associated with anincrease in cell death, though the quantity of cell deaths mightnot completely explain the significant reduction in total cellnumbers (Supplementary Fig. S3B, ii).

EGFL6 expression by tumor cells increases tumor growthin vivo

We next assessed in vivo tumor-growth effect of EGFL6 expres-sion by tumor cells. As SKOV3 cells do not express EGFL6, weevaluated the growth of two stably transfected EGFL6-expressingSKOV3 clones. Both clones demonstrated increased growth ratesrelative to vector-only transfected control cell clones (Fig. 5A, i;Supplementary Fig. S4A, i–S4A, ii). EGFL6-expressing tumorsdemonstrated an increase in percentage of Ki67-expressing cellsand a decrease in the concentration of ALDHþ cells (Fig. 5A, ii;Supplementary Fig. S4A, iii). Given the increased tumor volume,there was no estimated change in absolute number of ALDHþ

cells. Similarly, SKOV3 cells transduced with lentivirus EGFL6-GFP, compared with GFP-only controls, demonstrating increasedtumor growth based on both tumor weight and GFP intensity(Supplementary Fig. S4B).

We next evaluated the impact of EGFL6 knockdown on thegrowth of NIHOVCAR3 HGSC cells, which express high levelsof EGFL6. We validated EGFL6 knockdown with three EGFL6shRNAs (Fig. 5B, i). EGFL6 knockdown was associated with asignificant reduction in pSHP2 for all shRNAs (Fig. 5B, iii), and,for two of three shRNAs, EGFL6 knockdown was associatedwith a reduction in total SHP2 levels (Fig. 5B, iii). Analogous toSHP2 knockdown, EGFL6 knockdown resulted in a significant2.4-fold reduction in ALDHþ cells (Fig. 5B, ii; SupplementaryFig. S5). EGFL6 knockdown was associated with significantlyreduced tumor growth in vivo, an almost 2-fold increase inanimal survival and a 1.9-fold reduction in ALDHþ cells (Fig.5Biv–vi).

To further test the role of EGFL6 in vivo, we developedan EGFL6-blocking antibody (anti-EGFL6), which blockedEGFL6-triggered (i) increases in cell proliferation, (ii) changesin ALDHþ cell percentages, and (iii) increases in p-SHP2 andp-ERK (Fig. 4Diii; Supplementary Fig. S6). Although this anti-body was nonreactive on Western blot, the antibody detectedEGFL6 via immunofluorescence and was able to immunopre-cipitate EGFL6 in transduced cells (Supplementary Fig. S6).Anti-EGFL6 treatment of mice bearing NIHOVCAR3 flanktumors resulted in a significant reduction of tumor growth(P < 0.0001, Fig. 5C, i–ii), a 35% reduction in the percentageof ALDHþ cells, and a 7-fold reduction in absolute ALDHþ cellnumber (Fig. 5C, iii–iv).

Figure 4.EGFL6 signaling requires Integrin-mediated phosphorylation of SHP2. A, SKOV3 cell numbers after 72 hours of treatment with EGFL6 or EGFL6RGE. B,qRT-PCR demonstrating increased expression of integrin b3, but not b1 or b5, mRNA levels in ALDHþ vs. ALDH(�) ovarian cancer cells. C, FACS plotdemonstrating the integrin b1/b3–competitive inhibitor echistatin inhibits EGFL6-mediated reduction in ALDHþ cell percentages. D, Western blotanalysis of the indicated proteins with and without EGFL6 treatment demonstrating SHP2 is preferentially phosphorylated in ALDHþ cells and EGFL6 furtherincreases SHP2 activation in ALDHþ cells (i). EGFL6 treatment is associated with increased p-ERK in both ALDHþ and ALDH(�) cells, and EGFL6RGE

mutant does not significantly increase p-SHP2 or p-ERK (ii); iii, EGFL6-mediated SHP2 and ERK phosphorylation is suppressed by echistatin andanti-EGFL6 treatment. Bar graphs below graphs indicate densitometric quantification of p-SHP2. E, SHP2 Western blot of three independent SHP2 shRNA(Sh-SHP2; i), ALDH percentage (ii), and proliferation in Sh-SHP2 cells treated with EGFL6 or EGFL6RGE (iii). F, ALDHþ cell percentage (i) and cellproliferation (ii) in SHP2 inhibitor treated cells. All experiments were performed at least twice. Error bars, SDs.

EGFL6 Regulates Asymmetric Division

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6403

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Figure 5.

EGFL6 expression in tumor cells promotes ovarian tumor growth. A, i, tumor growth curves and tumor weights of EGFL6 and control vector–transfectedovarian cancer cells (n ¼ 10/group in two independent experiments); ii, IHC analysis and quantification of ALDH1A1 expression in EGFL6 vs. controltumors. B, qRT-PCR analysis of EGFL6 expression in control and EGFL6 shRNA knockdown (Sh-EGFL6) NIHOVCAR3 cells (i), ALDH FACS (ii), andWestern blot analysis in control and Sh-EGFL6 cells (iii); iv–vi tumor growth curves and overall survival, and ALDH percentage for OVCAR3 control (n ¼ 10)and Sh-EGFL6 cells (n ¼ 6/group). C, tumor growth curves and weights (i and ii) and ALDHþ cell percentages and absolute number for NIHOVCAR3tumors mock-treated or treated with anti-EGFL6 (EGFL6Ab, n ¼ 10/group; iii–iv). Error bars, SD.

Bai et al.

Cancer Res; 76(21) November 1, 2016 Cancer Research6404

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Vascular EGFL6 promotes tumor growth and metastasisin vivo

We next investigated the impact of vascular EGFL6 expressionon ovarian tumor growth. To do this, we used human HemSC tocreate human tumor blood vessels expressing EGFL6. HemSChave been shown to proliferate and generate blood-filledhuman vessels, in vivo in mice (28, 29). HemSC have low/noexpression of EGFL6 at baseline (Fig. 1D). We transducedHemSC with EGFL6 lentivirus and confirmed strong EGFL6expression (Fig. 6A). When SKOV3 cells and HemSCEGFL6 cellswere coinjected in vivo, we observed robust human CD31þ

tumor vessels, which expressed EGFL6 (Fig. 6B). SKOV3 cellsgrown with HemSCEGFL6 cells demonstrated increased growthcompared with SKOV3 cells grown with HemSCControl (Fig. 6C).As observed with EGFL6 expression in tumor cells, expression ofEGFL6 by HemSC cells was associated with increased numbersof Ki67þ cells (Fig. 6D). Confirming growth was related toEGFL6, treatment of SKOV3:HemSCEGFL6 tumors with anti-EGFL6 abrogated EGFL6-stimulated tumor growth (Supplemen-tary Fig. S7).

This experimentwas then repeatedwith primary cells from fourdifferent patients (n ¼ 2 tumors each). We observed tumorinitiation in eight of eight primary tumors coinjected withHemSCEGFL6 compared with three of eight tumor samples coin-jected with HemSCControl. As in cell lines, EGFL6 expression byHemSC cells was associated with more rapid tumor growth andincreased Ki67 stain (Fig. 6E and F).

Mice treated with anti-EGFL6 lost hair. Given EGFL6 isknown to be expressed in the hair follicle (14), this is pre-sumably an on-target side effect, suggesting anti-EGFL6 couldtarget murine EGFL6. We thus tested whether anti-EGFL6treatment could inhibit the growth of non–EGFL6-expressingSKOV3 tumors. Although anti-EGFL6 did not affect thegrowth of SKOV3 cells grown in vitro, twice-weekly intraper-itoneal treatment with anti-EGFL6 treatment of SKOV3 flanktumors resulted in an approximately 2-fold reduction ofin vivo SKOV3 tumor growth (Fig. 6G, i–ii). Analysis of tumorsindicated a 3.2-fold reduction in the percentage of ALDHþ

cells and a 9-fold reduction in absolute ALDHþ cell number(Fig. 6G, iii–iv).

Integrin and SHP2 signaling have been linked with meta-static capacity in several tumor types. Suggesting EGFL6 couldplay a role in metastasis, GSEA analysis revealed that EGFL6expression correlated with the expression of multiple metas-tasis-related gene sets (Fig. 7A and Supplementary Fig. S1C).We therefore evaluated the ability of EGFL6 to act as achemotropic factor. We used a microfluidic channel (20) tocreate a gradient of EGFL6 and evaluated the migration ofALDHþ and ALDH(�) ovarian cancer cells toward EGFL6-containing media. Although EGFL6 had no impact onALDH(�) cells, ALDHþ cells demonstrated statistically signif-icant migration toward EGFL6-containing media (Fig. 7B).Similarly, EGFL6 treatment was associated with increased"wound healing" in a standard scratch assay (SupplementaryFig. S8).

To evaluate the impact of EGFL6 on metastasis in vivo, we firstevaluated mice bearing EGFL6-expressing flank tumors. Whentumor cells expressed EGFL6 (from experiment in Fig. 5A;Supplementary Fig. S4), we observed no metastases from flanktumors either grossly or via luciferase imaging. Similarly, whenSKOV3 cells were grown with HemSCControl cells in vivo, we

observed no metastasis. In contrast, when SKOV3 cells weregrown with HemSCEGFL6, we observed metastases in 4 of 10mice (Fig. 7C, i—metastases were observed in the liver, lung,and ovary with associated ascites). Lymphovascular space inva-sion of ALDHþ cells was detected in HemSCEGFL6

flank tumors(Fig. 7C, ii). We similarly evaluated metastases from primaryovarian cells grown with HemSCControl or HemSCEGFL6. In thepresence of HemSCControl, we observed pulmonary metastasesin 1 of 3 mice that generated tumors, whereas withHemSCEGFL6, we observed metastases in 2 of 4 mice withmetastases in the lungs, liver, and the ovary (Fig. 7D).

We next assessed the ability of anti-EGFL6 to inhibit ovariancancer metastasis. We performed IP injections of ovarian cancercells and evaluated the impact of anti-EGFL6. Anti-EGFL6 led toa 3-fold reduction in the average number of tumor nodules(Fig. 7D). Furthermore, although control animals had tumornodules in the ovary and liver, anti–EGFL6-treated animals haddisease limited to the bowel/omentum with no detectableovarian tumors on/in the liver or ovaries.

We have recently shown that ovarian cancer cells injectedintravenously preferentially metastasize to the ovary (30).EGFL6 is expressed in the ovary. We assessed if EGFL6 couldrecruit ovarian cancer cells to the ovary. Mice were treated withanti-EGFL6 for 3 days and then injected with SKOV3 cellsintravenously via the tail vein. Anti-EGFL6 therapy was con-tinued for an additional 2 weeks. As expected, control micedeveloped metastases in not only the liver and lung but alsoin the ovaries, and in the peritoneal cavity with the develop-ment of ascites. Although treatment with anti-EGFL6 had noimpact on lung metastases, it reduced liver nodules 6-foldand completely eliminated ovarian metastases and ascites(Fig. 7E). Together, these data suggest that EGFL6 plays a rolein ovarian cancer metastasis and that anti-EGFL6 therapy mightrestrict the spread of ovarian cancer.

DiscussionVascular endothelial cells are regulators of the stem cell

niche (6, 31). Given the proximity of CSC and endothelialcells, a parallel role for endothelial cells and CSC has beenproposed. Endothelial cells have been suggested to directlyprovide "angiocrine" factors that direct stem cell fate (32, 33).EGFL6 represents one such angiocrine factor. Consistent with arole for EGFL6 as an angiocrine factor in ovarian cancer, it isreported to be present in the serum of patients with ovariancancer at levels 200 to 300X higher than that of healthycontrols (34).

We found EGFL6 promotes tumorigenesis by inducingALDHþ cells to undergo asymmetric division. This results inALDHþ cell maintenance, the production of ALDH(�) cells, andan increase in tumor growth. Supporting a role of EGFL6in asymmetric division, acute EGFL6 exposure is associatedwith a decrease in ALDHþ cell percentage, but no change in theabsolute ALDHþ cell number. Similarly, short-term EGFL6blockade increases ALDHþ cell percentage without changingabsolute cell number. In contrast, EGFL6 knockdown orlong-term blockade was associated with a significant reductionin both the percentage and absolute number of ALDHþ cells.This suggests that EGFL6, either related to or independentof the ability to promote asymmetric division, might play arole in ALDHþ cell maintenance. This would be consistent with

EGFL6 Regulates Asymmetric Division

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6405

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

EGFL6 signaling via SHP2, which plays a role in breast cancerstem cell maintenance (26). Interestingly, SHP2 signals inconjunction with the adaptor protein GAB2 (35, 36). Consis-tent with a critical role for SHP2 in ovarian cancer, GAB2 wasrecently reported to be an oncogene in ovarian cancer (37).

Parallel to our studies with EGFL6, GAB2 expression wasassociated with increased activation of ERK. Additional studieswill be necessary to determine whether EGFL6 signals viaGAB2. However, our work supports EGFL6 as a therapeutictarget in ovarian cancer.

Figure 6.

Vascular EGFL6 promotes tumor growth. A, EGFL6 Western of control and EGFL6-lentivirally transduced infantile hemangioma stem cells (HemSCEGFL6).B, coimmunofluorescence of human CD31 (hCD31) and EGFL6 in tumor vessels in SKOV3:HemSCEGFL6 tumor xenografts. C, tumor growth curves of SKOV3:HemSCEGFL6 tumors vs. SKOV3:HemSCControl tumors (n ¼ 10/group in two independent experiments). D, IHC analysis of Ki67 expression in EGFL6-expressingvs. control tumors. E, tumor growth curves of freshly isolated primary patient cells coinjected with HemSCEGFL6 or HemSCControl (n ¼ 4 patients withtwo tumors each). F, hematoxylin and eosin staining and Ki67 IHC of tumors generated with primary patient cells coinjected with HemSCEGFL6 orHemSCControl (n ¼ 6/group). G, tumor growth curve (i), weights (ii), ALDHþ cell percentage (iii), and absolute cell number (iv) for control andanti–EGFL6-treated SKOV3 (non–EGFL6-expressing cells) flank tumors (n ¼ 10/group in two separate experiments). Error bars, SDs.

Bai et al.

Cancer Res; 76(21) November 1, 2016 Cancer Research6406

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

EGFL6 is expressed in both cancer cells and vascular cells. Tostudy the role of human EGFL6 in a human vascular niche, wedeveloped a new model of human tumor vasculature usingHemSC. When combined with human ovarian tumor cells, theHemSC contributed to the tumor vasculature by creatinghuman tumor vessels, which were viable for as long as tumorswere maintained. Using the HemSC-derived tumor model, weobserved that vascular EGFL6 expression, similar to tumor cellEGFL6 expression, promoted the growth of tumor xenografts.In addition, vascular EGFL6 expression was associated with anincrease in metastases for both cell lines and primary cancercells. This is consistent with recent studies that not onlydemonstrated a role of hematogenous metastasis for ovariancancer, but also identified upregulation of EGFL6 with hema-

togenous ovarian cancer metastases (38). Interestingly, a pre-dominant site of metastasis was the ovary, which also expressesEGFL6. Anti-EGFL6 demonstrated an ability to completelyeliminate hematogenous spread of ovarian cancer cells to theovary, suggesting that EGFL6 might play a critical role in theovarian microenvironment. Consistent with this, EGFL6 exon-1knockout mice are prenatal lethal specifically in female mice(MMRRC). Inducible EGFL6 knockout mice will need to bedeveloped to further explore the role of EGFL6 in ovariancancer.

Our data strongly support EGFL6 as a therapeutic target inovarian cancer. EGFL6 blockade/knockdown strongly suppressedtumor growth and reduced metastasis. mRNA microarray expres-sion data suggest that EGFL6 is also expressed at very high levels in

Figure 7.

The role of EGFL6 in ovarian cancermetastasis. A, GSEA demonstratingEGFL6 expression correlated withmetastatic gene signatures inendometrial ovarian cancer andmelanoma. (See Supplementary Fig.S1C for ES scores, P values, and falsediscovery rate q-values.) B, i,immunofluorescent GFP-labeledALDHþ SKOV3 cells after capture(top) and migration (bottom) inmicrofluidic migration device. Controlcells (bottom left) have no gradient vs.EGFL6 gradient (bottom right),summary of distance (ii) migrated forthe indicated cells from replicateexperiments. C, percentage of micewith identifiable metastases whenSKOV3 cells or primary human ovariancancer cells were grownsubcutaneously (SQ) combined withHemSCControl or HemSCEGFL6 (i) andIHC demonstrating ALDHþ cells inSKOV3-HemSCEGFL6 tumor vessels(ii).D, percentage of metastasis to theindicated body sites in mice injectedintraperitoneally with SKOV3 mock-treated (control) or treated withanti-EGFL6. E, percentage of micewith metastasis to the indicated bodysites in mice injected intravenouslywith SKOV3 cells and mock-treated(control) or anti–EGFL6-treated cells(n ¼ 10/group in two separateexperiments).

EGFL6 Regulates Asymmetric Division

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6407

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

uterine tumors and aerodigestive tumors (11, 34). EGFL6 iselevated in the serum of patients with glioblastoma, bladder,breast, gastric, liver, lung, pancreatic, and prostate cancers (34).Thus, EGFL6 could be an important target for other tumor types.

In conclusion, EGFL6 is an angiocrine and tumor cell factorthat promotes tumorigenesis via SHP2-mediated signaling.EGFL6 stimulates ALDHþ ovarian CSC to migrate and undergoasymmetric division. As such, EGFL6 specifically promotes ovar-ian cancer cell metastasis. Finally, a novel EGFL6-blocking anti-body abrogated the impact of EGFL6 on ovarian tumor growthand metastasis, suggesting this antibody represents a noveltherapeutic.

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

Authors' ContributionsConception and design: S. Bai, Z.-Y. Zhang, E. Yoon, R.J. BuckanovichDevelopment ofmethodology: S. Bai, P. Ingram, Y.-C. Chen,N.Deng, Y.Wang,E. Boscolo, J. Bischoff, E. Yoon, R.J. BuckanovichAcquisition of data (provided animals, acquired and managed pati-ents, provided facilities, etc.): S. Bai, P. Ingram, N. Deng, Z.-Y. Zhang,J. Bischoff

Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S. Bai, P. Ingram, A. Pearson, Y. Niknafs, P. O'Hayer,E. Yoon, R.J. BuckanovichWriting, review, and/or revision of the manuscript: S. Bai, P. O'Hayer,Z.-Y. Zhang, E. Boscolo, J. Bischoff, E. Yoon, R.J. BuckanovichAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): Y. Wang, E. YoonStudy supervision: E. Yoon, R.J. Buckanovich

AcknowledgmentsWe thank David Cheresh for providing the Integrin b3–blocking

antibody.

Grant SupportThis work was supported by the OCRFA and the DOD-OCRP Idea Award

W81XWH-14-1-0187. UMCC core facilities were supported by theNIH SupportGrant (P30 CA046592).

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received February 1, 2016; revised July 1, 2016; accepted July 25, 2016;published OnlineFirst November 1, 2016.

References1. Choi Y-J, Ingram PN, Yang K, Coffman L, Iyengar M, Bai S, et al. Identifying

an ovarian cancer cell hierarchy regulated by bone morphogenetic protein2. Proc Natl Acad Sci 2015;112:E6882–8.

2. Silva IA, Bai S, McLean K, Yang K, Griffith K, Thomas D, et al. Aldehydedehydrogenase in combination with CD133 defines angiogenic ovariancancer stem cells that portend poor patient survival. Cancer Res 2011;71:3991–4001.

3. Steg AD, Bevis KS, Katre AA, Ziebarth A, Dobbin ZC, Alvarez RD, et al. Stemcell pathways contribute to clinical chemoresistance in ovarian cancer.Clin Cancer Res 2012;18:869–81.

4. Landen CN Jr, Goodman B, Katre AA, Steg AD, Nick AM, Stone RL, et al.Targeting aldehyde dehydrogenase cancer stem cells in ovarian cancer.Mol Cancer Ther 2010;9:3186–99.

5. Charafe-Jauffret E, Ginestier C, Iovino F, TarpinC,DiebelM, Esterni B, et al.Aldehyde dehydrogenase 1-positive cancer stem cells mediate metastasisand poor clinical outcome in inflammatory breast cancer. Clin Cancer Res2010;16:45–55.

6. ShenQ,Goderie SK, Jin L, KaranthN, SunY, AbramovaN, et al. Endothelialcells stimulate self-renewal and expand neurogenesis of neural stem cells.[see comment]. Science 2004;304:1338–40.

7. Holmberg J, Genander M, Halford MM, Anneren C, Sondell M, ChumleyMJ, et al. EphB receptors coordinate migration and proliferation in theintestinal stem cell niche. Cell 2006;125:1151–63.

8. Calabrese C, Poppleton H, Kocak M, Hogg TL, Fuller C, Hamner B, et al.A perivascular niche for brain tumor stem cells. Cancer Cell 2007;11:69–82.

9. Gilbertson RJ, Rich JN. Making a tumour's bed: Glioblastoma stem cellsand the vascular niche. Nat Rev Cancer 2007;7:733–6.

10. Butler JM, Kobayashi H, Rafii S. Instructive role of the vascular niche inpromoting tumour growth and tissue repair by angiocrine factors. Nat RevCancer 2010;10:138–46.

11. Buckanovich RJ, Sasaroli D, O'Brien-Jenkins A, Botbyl J, Hammond R,Katsaros D, et al. Tumor vascular proteins as biomarkers in ovarian cancer.J Clin Oncol 2007;25:852–61.

12. Lu C, Bonome T, Li Y, Kamat AA, Han LY, Schmandt R, et al. Genealterations identified by expression profiling in tumor-associated endo-thelial cells from invasive ovarian carcinoma. Cancer Res 2007;67:1757–68.

13. Yeung G,Mulero JJ, Berntsen RP, Loeb DB, Drmanac R, Ford JE. Cloning ofa novel epidermal growth factor repeat containing gene EGFL6: expressedin tumor and fetal tissues. Genomics 1999;62:304–7.

14. Osada A, Kiyozumi D, Tsutsui K, Ono Y, Weber CN, Sugimoto N, et al.Expression of MAEG, a novel basement membrane protein, in mouse hairfollicle morphogenesis. Exp Cell Res 2005;303:148–59.

15. Fujiwara H, Ferreira M, Donati G, Marciano DK, Linton JM, Sato Y, et al.The basement membrane of hair follicle stem cells is a muscle cellniche. Cell 2011;144:577–89.

16. Chim SM, Qin A, Tickner J, Pavlos N, Davey T, Wang H, et al. EGFL6promotes endothelial cell migration and angiogenesis through theactivation of extracellular signal-regulated kinase. J Biol Chem 2011;286:22035–46.

17. Oberauer R, Rist W, Lenter MC, Hamilton BS, Neubauer H. EGFL6 isincreasingly expressed in human obesity and promotes proliferation ofadipose tissue-derived stromal vascular cells. Mol Cell Biochem 2010;343:257–69.

18. Pulaski HL, Spahlinger G, Silva IA, McLean K, Kueck AS, ReynoldsRK, et al. Identifying alemtuzumab as an anti-myeloid cell anti-angiogenic therapy for the treatment of ovarian cancer. J Transl Med2009;7:49.

19. Rhodes DR, Kalyana-Sundaram S, Mahavisno V, Varambally R, Yu J,Briggs BB, et al. Oncomine 3.0: Genes, pathways, and networks in acollection of 18,000 cancer gene expression profiles. Neoplasia 2007;9:166–80.

20. Chen YC, Allen SG, Ingram PN, Buckanovich R, Merajver SD, Yoon E.Single-cell migration chip for chemotaxis-based microfluidic selectionof heterogeneous cell populations. Sci Rep 2015;5:9980.

21. McLean K, Gong Y, Choi Y, Deng N, Yang K, Bai S, et al. Human ovariancarcinoma-associated mesenchymal stem cells regulate cancer stem cellsand tumorigenesis via altered BMP production. J Clin Invest 2011;121:3206–19.

22. Kryczek I, Liu S, Roh M, Vatan L, Szeliga W, Wei S, et al. Expression ofaldehydedehydrogenase andCD133defines ovarian cancer stem cells. Int JCancer 2012;130:29–39.

23. Deng S, Yang X, Lassus H, Liang S, Kaur S, Ye Q, et al. Distinct expressionlevels and patterns of stem cell marker, aldehyde dehydrogenase iso-form 1 (ALDH1), in human epithelial cancers. PLoS One 2010;5:e10277.

24. Chung J, Kim YJ, Yoon E. Highly-efficient single-cell capture in micro-fluidic array chips using differential hydrodynamic guiding structures.Appl Phys Lett 2011;98:123701.

25. Lacalle RA, Mira E, Gomez-Mouton C, Jimenez-Baranda S, MartinezAC, Manes S. Specific SHP-2 partitioning in raft domains triggers

Cancer Res; 76(21) November 1, 2016 Cancer Research6408

Bai et al.

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

integrin-mediated signaling via Rho activation. J Cell Biol 2002;157:277–89.

26. Aceto N, Sausgruber N, Brinkhaus H, Gaidatzis D, Martiny-Baron G,Mazzarol G, et al. Tyrosine phosphatase SHP2 promotes breast cancerprogression and maintains tumor-initiating cells via activation of keytranscription factors and a positive feedback signaling loop. Nat Med2012;18:529–37.

27. Zeng LF, Zhang RY, Yu ZH, Li S, Wu L, Gunawan AM, et al. Therapeuticpotential of targeting the oncogenic SHP2 phosphatase. J Med Chem2014;57:6594–609.

28. Boscolo E, Bischoff J. Vasculogenesis in infantile hemangioma. Angiogen-esis 2009;12:197–207.

29. Khan ZA, Melero-Martin JM, Wu X, Paruchuri S, Boscolo E, Mulliken JB,et al. Endothelial progenitor cells from infantile hemangioma and umbil-ical cord blood display unique cellular responses to endostatin. Blood2006;108:915–21.

30. Coffman LG, Burgos-Ojeda D, Bai S, Buckanovich RJ. Murine ovariancancer metastasis models demonstrate preferential spread of ovariancancer cells to the ovary. Transl Res 2016;175:92–102.e2.

31. Cleaver O, Dor Y. Vascular instruction of pancreas development. Devel-opment 2012;139:2833–43.

32. Bergers G, Hanahan D. Modes of resistance to anti-angiogenic therapy.Nat Rev Cancer 2008;8:592–603.

33. Hanahan D, Weinberg RA. Hallmarks of cancer: The next generation.Cell 2011;144:646–74.

34. Wang X, Gong Y, Wang D, Xie Q, Zheng M, Zhou Y, et al. Analysis of geneexpression profiling in meningioma: Deregulated signaling pathwaysassociated with meningioma and EGFL6 overexpression in benignmenin-gioma tissue and serum. PLoS One 2012;7:e52707.

35. Futami M, Zhu QS, Whichard ZL, Xia L, Ke YH, Neel BG, et al. G-CSFreceptor activation of the Src kinase Lyn is mediated by Gab2 recruitmentof the Shp2 phosphatase. Blood 2011;118:1077–86.

36. Gu H, Pratt JC, Burakoff SJ, Neel BG. Cloning of p97/Gab2, the majorSHP2-binding protein in hematopoietic cells, reveals a novel pathwayfor cytokine-induced gene activation. Mol Cell 1998;2:729–40.

37. Dunn GP, Cheung HW, Agarwalla PK, Thomas S, Zektser Y, Karst AM,et al. In vivo multiplexed interrogation of amplified genes identifiesGAB2 as an ovarian cancer oncogene. Proc Natl Acad Sci U S A 2014;111:1102–7.

38. Pradeep S, Kim SW, Wu SY, Nishimura M, Chaluvally-Raghavan P,Miyake T, et al. Hematogenous metastasis of ovarian cancer: Rethinkingmode of spread. Cancer Cell 2014;26:77–91.

www.aacrjournals.org Cancer Res; 76(21) November 1, 2016 6409

EGFL6 Regulates Asymmetric Division

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Correction

Correction: EGFL6Regulates the AsymmetricDivision, Maintenance, and Metastasis ofALDHþ Ovarian Cancer Cells

In this article (Cancer Res 2016;76:6396–409), which appeared in the November 1,2016, issue of Cancer Research (1), the correct name of the sixth author isYashar S. Niknafs.The online version of the article has been corrected and no longer matches the print.

Reference1. Bai S, Ingram P, Chen Y-C, Deng N, Pearson A, Niknafs YS, et al. EGFL6 regulates the asymmetric

division, maintenance, and metastasis of ALDHþ ovarian cancer cells. Cancer Res 2016;76:6396–409.

Published online April 14, 2017.doi: 10.1158/0008-5472.CAN-17-0421�2017 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 2175

2016;76:6396-6409. Cancer Res   Shoumei Bai, Patrick Ingram, Yu-Chih Chen, et al.  

Ovarian Cancer Cells+Metastasis of ALDHEGFL6 Regulates the Asymmetric Division, Maintenance, and

  Updated version

  http://cancerres.aacrjournals.org/content/76/21/6396

Access the most recent version of this article at:

  Material

Supplementary

  http://cancerres.aacrjournals.org/content/suppl/2017/02/01/76.21.6396.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/76/21/6396.full#ref-list-1

This article cites 38 articles, 14 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/76/21/6396.full#related-urls

This article has been cited by 5 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

  .pubs@aacr.org

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://cancerres.aacrjournals.org/content/76/21/6396To request permission to re-use all or part of this article, use this link

on March 30, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from