Dll4 Inhibition plus Aflibercept Markedly Reduces Ovarian ...€¦ · Delta-like ligand 4 (Dll4), one of the Notch ligands, plays an ... University, 128 Shenyang Road, Shanghai, P.R

Cancer Biology and Signal Transduction

Dll4 Inhibition plus Aflibercept Markedly ReducesOvarian Tumor GrowthJie Huang1,Wei Hu1, Limin Hu2, Rebecca A. Previs1, Heather J. Dalton1, Xiao-Yun Yang1,Yunjie Sun1, Michael McGuire1, Rajesha Rupaimoole1, Archana S. Nagaraja1, Yu Kang1,Tao Liu1, Alpa M. Nick1, Nicholas B. Jennings1, Robert L. Coleman1, Robert B. Jaffe2, andAnil K. Sood1,3,4

Abstract

Delta-like ligand 4 (Dll4), one of the Notch ligands, isoverexpressed in ovarian cancer, especially in tumors resistantto anti-VEGF therapy. Here, we examined the biologic effects ofdual anti-Dll4 and anti-VEGF therapy in ovarian cancer models.Using Dll4-Fc blockade and anti-Dll4 antibodies (murineREGN1035 and human REGN421), we evaluated the biologiceffects of Dll4 inhibition combined with aflibercept or chemo-therapy in orthotopic mouse models of ovarian cancer. We alsoexamined potential mechanisms by which dual Dll4 and VEGFtargeting inhibit tumor growth using immunohistochemicalstaining for apoptosis and proliferation markers. Reverse-phaseprotein arrays were used to identify potential downstreamtargets of Dll4 blockade. Dual targeting of VEGF and Dll4 withmurine REGN1035 showed superior antitumor effects in ovar-ian cancer models compared with either monotherapy. In the

A2780 model, REGN1035 (targets murine Dll4) or REGN421(targets human Dll4) reduced tumor weights by 62% and 82%,respectively; aflibercept alone reduced tumor weights by 90%.Greater therapeutic effects were observed for Dll4 blockade(REGN1035) combined with either aflibercept or docetaxel(P < 0.05 for the combination vs. aflibercept). The superiorantitumor effects of REGN1035 and aflibercept were relatedto increased apoptosis in tumor cells compared with themonotherapy. We also found that GATA3 expression wassignificantly increased in tumor stroma from the mice treatedwith REGN1035 combined with docetaxel or aflibercept,suggesting an indirect effect of these combination treatmentson the tumor stroma. These findings identify that dual target-ing of Dll4 and VEGF is an attractive therapeutic approach.Mol Cancer Ther; 15(6); 1–9. �2016 AACR.

IntroductionOvarian cancer remains the leading cause of death from a

gynecologic malignancy in the United States. Although tumor-reductive surgery and taxane- and platinum-based chemotherapyregimens are effective for most patients with primary ovariancancer, recurrence is common and often deadly. New therapeuticagents are thus needed to improve survival rates and, eventually,to cure this deadly disease.

VEGF plays an important role in tumor angiogenesis (1).We have previously demonstrated that the combination of asoluble VEGF decoy receptor (VEGF Trap, also called afliber-cept) plus a taxane is effective in decreasing tumor burden in amouse model of human ovarian cancer (2). The success ofantiangiogenic therapies, such as bevacizumab and afliber-cept, in solid tumors, including ovarian cancer, has confirmedthe clinical viability of this approach (3). However, not allovarian cancers are responsive to VEGF inhibitors (4). More-over, ovarian cancers develop resistance to VEGF inhibitorsover time (5, 6), leading to progressive tumor growth. Thus,additional angiogenic signaling pathways will likely need tobe targeted to enhance or extend the benefit of antiangiogenictherapy (7).

Delta-like ligand 4 (Dll4), one of the Notch ligands, plays animportant role in tumor angiogenesis (8). VEGF upregulates Dll4expression in endothelial cells (9), and in turn, Dll4-Notchsignaling modulates the response of endothelial cells to VEGF(10, 11). Blockade of Dll4–Notch signaling results in excessivesprouting of markedly abnormal, tortuous tumor vessels, which,seemingly paradoxically, reduces tumor growth because theseabnormal vessels are not perfused (12, 13). Our previous workdemonstrated the biologic roles of Dll4 in tumor and endothelialcells in ovarian cancer, showing that targeting Dll4 can restoreVEGFR2 expression via epigenetic mechanisms (14). Moreover,ovarian cancer patients with increased Dll4 expression in thetumor endothelial cells had poor response to anti-VEGF therapy(14). These findings prompted us to evaluate the biologic effects

1Department of Gynecologic Oncology and Reproductive Medicine,The University of Texas MDAnderson Cancer Center, Houston, Texas.2Center for Reproductive Sciences, University of California, San Fran-cisco, San Francisco, California. 3Center for RNA Interference andNon-Coding RNAs, The University of Texas MD Anderson CancerCenter,Houston,Texas. 4DepartmentofCancerBiology,TheUniversityof Texas MD Anderson Cancer Center, Houston, Texas.

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

W. Hu, L. Hu, and R.A. Previs contributed equally to this article.

Current address for Yu Kang: Obstetrics and Gynecology Hospital, FudanUniversity, 128 Shenyang Road, Shanghai, P.R. China; and current address forTao Liu: Xiehe Hospital, 1277 Jiefang Road, Wuhan, Hubei, P.R. China.

Corresponding Author: Anil K. Sood, The University of Texas MD AndersonCancer Center, 1155 Pressler Street, Unit 1362, Houston, TX 77030. Phone: 713-745-5266; Fax: 713-792-7586; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-15-0144

�2016 American Association for Cancer Research.

MolecularCancerTherapeutics

www.aacrjournals.org OF1

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

Published OnlineFirst March 23, 2016; DOI: 10.1158/1535-7163.MCT-15-0144

http://mct.aacrjournals.org/

of Dll4 blockade combined with aflibercept in ovarian cancermouse models.

In this study, we examined the antitumor activity of bothhuman- and mouse-derived anti-Dll4 antibodies (REGN421and REGN1035) with and without aflibercept in orthotopicmouse models of ovarian cancer. We also examined potentialmechanisms by which dual Dll4 and VEGF targeting inhib-its tumor growth using immunohistochemical staining forapoptosis and proliferation markers, and reverse phase proteinarrays (RPPA).

Materials and MethodsReagents and cell cultures

Docetaxel was obtained from Sanofi-Aventis. Soluble Dll4extracellular domain fused to Fc (mouse Dll4-Fc), aflibercept(inhibits the activity of VEGF-A, VEGF-B, and PIGF; refs. 15,16), REGN1035 (a surrogate that crossreacts with mouse Dll4),and REGN421 (a fully humanmAbdirectly against Dll4), vehicle,and human Fc control were obtained from Regeneron Pharma-ceuticals (12, 15, 17). REGN421 binds to Dll4 and blocks theinteractionofDll4withNotch receptors, thereby inhibitingNotchsignaling. The human OVCAR3 cell line (2) was kindly providedby T. Hamilton (Fox Chase Cancer Center, Philadelphia, PA).HeyA8 and A2780 cancer cell lines were obtained from the MDAnderson Characterized Cell Line Core Facility (Houston, TX),which supplies authenticated cell lines. The cell lines were rou-tinely tested to confirm the absence of mycoplasma, and allexperiments were performed with cell lines at 60% to 80%confluence.

Experimental animalsFemale athymic nude mice (NCr-nu) were purchased from the

Animal Production Program of the NCI, Frederick CancerResearch and Development Center (Frederick, MD). The animalswere kept under specific pathogen-free conditions in facilitiesapproved by the American Association for Accreditation of Lab-oratory Animal Care. Mice used in these experiments were 5 to 7weeks of age. All protocols involving mice were approved by theCommittee on Animal Research at the University of California atSan Francisco and by the MD Anderson Cancer Center Institu-tional Animal Care and Use Committee.

Aflibercept, Dll4-Fc, REGN1035, and REGN421 treatmentsTo test the effects of Dll4-Fc both alone and combined with afli-

bercept in an OVCAR3 model, we prepared OVCAR-3 cells fromascites fluid of athymic mice inoculated with OVCAR-3 cells, asdescribed in our previous study (2). Briefly, ascites fluid was collect-ed and placed in a 4�C refrigerator for 1 to 2 hours. The supernatantwas discarded and the cells were diluted with RPMI1640 supple-mented with 2.0 g/L glucose and 0.3 g/L L-glutamine that hadbeen prewarmed in a 37�C incubator. Athymic mice (5–7 weeks)were inoculated intraperitoneally with OVCAR-3 cells (2 � 106

cells/mouse in 500 mL RPMI1640). Two weeks after inoculation,the mice were randomized into 4 groups (10 mice/group). Onegroup was treated with intraperitoneal aflibercept (10 mg/kgbody weight) twice weekly plus intravenous Dll4-Fc (25 mg/kg)three times weekly for 3 weeks. The second group was treatedwith aflibercept alone (10 mg/kg) three times per week. The thirdgroup was treated with 25 mg/kg Dll4-Fc alone. The fourthgroup received the vehicle alone.

Abdominal circumference and body weight were measuredtwice weekly during treatment. After 5 weeks of treatment, themice were euthanized by CO2 inhalation. The number of tumornodules and distribution of tumors were recorded. The volumesof the ascites were measured, and tumor tissue was excised,weighed, fixed in 4% paraformaldehyde (pH 7.4) at 4�C for 24hours, and embedded in paraffin. Paraffin sections (5 mm) wereused for immunohistochemical analysis.

Additional experiments were carried out to test the effects ofaflibercept, REGN1035, REGN421, and docetaxel in mice inoc-ulated with A2780 and HeyA8 ovarian cancer cells, which we hadpreviously found to be Dll4 positive (14). Following trypsintreatment, these cell culturesweremixedwithmediumcontaining10% FBS, were subjected to centrifugation at 1,200 rpm for 5minutes, and were washed in PBS. Female athymic nude micewere then injected intraperitoneally with Dll4-positive A2780cells (2 � 106 cells/0.2 mL) or Dll4-weak positive HeyA8 cells(2.5 � 105 cells/0.2 mL).

Seven days after inoculation, the mice were randomized intoseven groups (10 mice/group): (i) a control, (ii) docetaxel, (iii)aflibercept, (iv) REGN1035 or REGN421, (v) docetaxel plusaflibercept, (vi) docetaxel plus REGN1035 or REGN421, or (vii)aflibercept plus REGN1035 or REGN421. Docetaxel was givenintraperitoneally at 35 mg per mouse three times per week,aflibercept was given intraperitoneally at 12.5 mg/kg three timesper week, and REGN1035 and REGN421 were given intravenous-ly at 5 mg/kg twice per week (http://www.google.com/patents/WO2010151770A1). Fifteen minutes before sacrifice, the micewere infused with 100 ml of Hypoxyprobe-1 (pimonidazole HCl,100mg/kg, NPI, Inc) through the tail vein.Mice were killed after 4to 6 weeks of therapy, when animals in the control group becamemoribund. At the timeofdeath, themouse and tumorweights andthe number of nodules were recorded.

Immunohistochemical stainingParaffin-embedded tumor tissues were used to detect cell pro-

liferation (Ki67), apoptosis (cleaved caspase-3), GATA3, CX3CL1,hypoxia, andF4/80-positive cells. Sectionsweredeparaffinizedandrehydrated. After antigen retrieval with citrate buffer (pH 6.0), thesections were blocked with 3% hydrogen peroxide in methanoland protein blocker at room temperature. The sections were thenincubated with a monoclonal anti-Ki67 antibody (1:200 dilution;Biocare Medical), anti-cleaved caspase-3 antibody (1:800; BiocareMedical), or anti-mouse GATA3 antibody (HG3-31; 1:100; SantaCruz Biotechnology), Rat anti-mouse F4/80 antibody (1:50, Ser-otec), anti-rabbit polyclonal CX3CL1 (1:100; Abcam), andHypox-yprobe-1-Mab 1 (1:50; Natural Pharmacia International, Inc)overnight at 4�C. After being washed with PBS, all sections wereincubated with 4 plus biotinylated goat anti-rabbit (for Ki67 andcleaved caspase-3) or anti-mouse antibody (for GATA3; BiocareMedical) for 20 minutes. Then, the slides were washed with PBSand incubated with 4 plus streptavidin horseradish peroxidase(HRP; Biocare Medical) for 20 minutes.

To detect microvessels as represented by CD31, frozen ovariantumor sections were fixed in cold acetone for 15minutes, washedwithPBS, blockedwithproteinblocker (4%fish gelatin), and thenincubated with primary antibody against CD31 (1:800; BD Phar-mingen/BDBiosciences) overnight at 4�C. Theywere thenwashedwith PBS and incubated with HRP-conjugated goat anti-ratIgG (1:200; Jackson ImmunoResearch Laboratories) for 1hour. Reactive tissues were visualized using staining with 3,30-

Huang et al.

Mol Cancer Ther; 15(6) June 2016 Molecular Cancer TherapeuticsOF2




diaminobenzidine (Research Genetics), followed by counter-staining with Gill's Hematoxylin (BioGenex Laboratories).

Quantification of microvessel density, cleaved caspase-3, Ki67,GATA3, CX3CL1, hypoxia, and F4/80 cells

For the quantification of microvessel density, cleaved caspase-3, Ki67, and GATA3, five ovarian tumor samples from eachtreatment group were examined. To quantify microvessel densityfor each sample, CD31-positive blood vessels were countedwithin five randomly selected 0.159-mm2

fields at 200� magni-fication. To quantify the expression of cleaved caspase-3, Ki67,GATA3, CX3CL1, and F4/80 cells, we determined the percentageof positive cells in five randomly selected 0.159-mm2

fields at200� magnification. The percentage of hypoxic area was deter-mined by subtracting areas of necrosis from total pimonidazole-positive areas and then normalizing by whole section areas.

RT-PCRThe relative expression of E-cadherin in ovarian tumors

obtained from in vivo studies was determined by RT-PCR. EachRT-PCR used 5 mg total RNA isolated from homogenized tissuestreated with a single agent or the combination from each treat-ment group. GAPDH was used as a control (15). The primersequences of humanE-cadherin primers (991bp)were as follows:H-E-cadherin-F (forward): 50-GTGACTGATGCTGATGCCCC-CAATACC-30; H-E-cadherin-R (reverse): 50-GACGCAGAATCA-GAATTAGGAAAGCAAG-30. The human CX3CL1 primers wereobtained from Sigma.

RPPA analysisTumor tissues collected from in vivo studies were homoge-

nized using a digital homogenizer in the following lysis buffer:1% Triton X-100, 50 nmol/L 4-(2-hydroxyethyl)-1-piperazi-neethanesulfonic acid (pH 7.4), 150 mmol/L MgCl2, 1mmol/L ethylene glycol tetraacetic acid, 100 mmol/L NaF,10 mmol/L sodium pyrophosphate, 1 mmol/L Na3VO4, 10%glycerol, and freshly added protease and phosphatase inhibi-tors. Cellular proteins were denatured using 1% SDS, and five2-fold serial dilutions were performed in lysis buffer containing1% SDS (dilution buffer). These diluted lysates were arrayed onnitrocellulose-coated FAST slides (Whatman) using an Aushon2470 Arrayer (Aushon Biosystems). Slides were probed with176 validated human primary antibodies. The SuperCurveFitting logistic model, developed by the Department of Bioin-formatics and Computational Biology at MD Anderson CancerCenter (Houston, TX; http://bioinformatics.mdanderson.org/OOMPA), was used to generate a fitted curve for each dilution.For both observed and fitted data, the fitted curve was thenplotted with the signal intensities on the y-axis and the log2

values of protein levels on the x-axis. The positive fold changeswere calculated by dividing each linear value greater than 1.0 bythe average control linear value for each antibody tested. Thenegative fold changes (for linear values less than 1.0) werecalculated using the formula �1/linear fold change and thevalues were plotted as a bar graph.

Quantification of plasma markers by ELISAPlasma mouse IFNg (catalog number: KMC4021; Life Tech-

nologies) level was measured by ELISA based on the manufac-turer's instructions. All plasma samples were run in duplicate. AmAb specific for IFNg was coated onto the wells of the microtiter

strips provided. Samples, including standards of known IFNgcontent, control plasma, and treated plasma were pipetted intothese wells.

Statistical analysisFor the in vivo therapy experiments, 10 mice were used in

each group, which provided the power to detect a 50%reduction in tumor size (b error ¼ 0.2). For nonparametricdistributions, the Mann–Whitney U test was used. A value ofP < 0.05 with two-tailed testing was deemed statisticallysignificant.

ResultsEffects of Dll4-Fc and aflibercept on tumor burden and ascitesformation

Inmice inoculatedwithOVCAR3ovarian cancer cells, themeantumor burden in the group treated with aflibercept plus Dll4-Fc(0.51� 0.08 g) was 82% less than that in the control group (2.86� 0.26 g; Fig. 1A, left, and Supplementary Fig. S1A). Mean tumorburden in the aflibercept-alone (1.30� 0.20 g) and Dll4-Fc alone(2.19 � 0.41 g) groups were 55% less and 24% less, respectively,than that in the control group (P < 0.01). Ascites were almostcompletely prevented in the aflibercept plus Dll4-Fc–treated andaflibercept-treated groups, compared with the controls (Fig. 1A,right, and Supplementary Fig. S1B).

At the end of the study period (after 4 weeks of treatment),100% of the control and Dll4-Fc–alone groups and 60% of theaflibercept-alone group had tumors in the hilum of the liver,whereas only 20% of the Dll4-Fc plus aflibercept group had anyperihepatic tumors. In addition, 90%of the control group, 70%ofthe Dll4-Fc–alone group, and 50% of the aflibercept-alone grouphad tumors on the diaphragm. In contrast, nomice in the Dll4-Fcplus aflibercept groupdisplayeddiaphragmatic tumors (P<0.05),suggesting that the combined treatment suppressed direct metas-tasis. Observable behaviors in the mice (e.g., degrees of activityand eating) did not significantly differ between the treatmentgroups.

In vivoDll4 inhibitionwithhumanormouse anti-Dll4 antibodyTo investigate the translational relevance of Dll4-targeted ther-

apy in ovarian cancer, we next studied the effects of anti-Dll4antibodies in combination with anti-VEGF treatment in vivo ontumor growth and angiogenesis. To address the biologic signif-icance of Dll4 expression in tumor and endothelial cells, we usedhuman (tumor) and mouse (endothelial) anti-Dll4 antibodiesand A2780 andHeyA8 cells, which we had previously found to beDll4 positive. In the mice inoculated with A2780 ovarian cancercells, treatment with mouse (REGN1035) or human (REGN421)anti-Dll4 antibodies alone resulted in significant growth inhibi-tion compared with the control (62% and 82%, P < 0.05 forboth; Fig. 1B and C, left, respectively). Aflibercept alone led to areduction in tumor weights by 90% less than those of the controls(P < 0.05). However, the combination of REGN1035 plus afli-bercept resulted in the greatest inhibition of tumor growth com-pared with the controls (96%, P < 0.0001 for the combination vs.control). Notably, superior therapeutic effects were observed forDll4 blockade (REGN1035) combined with either aflibercept ordocetaxel (P<0.05 for the combination vs. aflibercept;P<0.01 forthe combination vs. REGN1035 or docetaxel) compared withmonotherapy.

Dll4 Inhibition in Ovarian Cancer

www.aacrjournals.org Mol Cancer Ther; 15(6) June 2016 OF3




Figure 1.In vivo study of Dll4 blockade combined withaflibercept in OVCAR3, A2780, and HeyA8models. A, mean tumor burden and ascites inOVCAR3-inoculated nude mice treated withaflibercept (Afliber), Dll4-Fc, or both for 3 weeks.B and C, in vivoDll4 inhibition bymouse or humananti-Dll4 antibody in the A2780 (B and C) andHeyA8 (D) models. Afliber, aflibercept; Doce,docetaxel. � , P < 0.05, compared with controlgroup; �� , P < 0.01; �� , P < 0.001.

Huang et al.





Figure 2.Dll4 blockade inhibits angiogenesis and proliferation and induces apoptosis in the A2780 model. A, proliferation, cleaved caspase-3, and microvessel density(CD31) were detected by IHC staining in A2780-inoculated mice treated with human anti-Dll4 antibody (REGN421), docetaxel, or aflibercept, alone or incombination. B, proliferation, cleaved caspase-3, and microvessel density (CD31) were detected by IHC staining in A2780-inoculated mice treated with mouseanti-Dll4 antibody (REGN1035) alone or in combination. All pictures were taken at original magnification (200�). Error bars, SEM. �, P < 0.05, compared withcontrols; �� , P < 0.01; �� , P < 0.001. MVD, microvessel density.






However, tumor reduction did not differ significantlybetween the REGN421 combined treated groups and the sin-gle-drug treated groups. We also examined the effects of treat-ment on the number of tumor nodules and body weight in theA2780 model. The number of nodules was significantly lowerin all treated groups than in the controls (P < 0.05; Fig. 1B andC, right), but nodules did not differ between the combinationand single-drug treatment groups (Fig. 1Band C, right). Meanbody weights did not differ significantly between treatmentgroups (Supplementary Fig. S2A and S2B).

In the mice inoculated with HeyA8 ovarian cancer cells, theeffects of Dll4 inhibition on ovarian tumor weight were, ingeneral, similar to those in the A2780 model (Fig. 1D, left). Inaddition, tumor weights in mice treated with single agents weresignificantly lower than those in the controls: 61% less withaflibercept (P < 0.05), 68% less with REGN1035 (P < 0.05), and57% less with docetaxel (P < 0.01). However, the extent ofreduction in tumor weight and number of nodules did not differsignificantly between the combined- and single-drug treatmentgroups (Fig. 1D, right). Mean mouse body weights did not differsignificantly between any treatment groups in the HeyA8 model(Supplementary Fig. S2C).

Effects of Dll4 and VEGF blockade on microvessel density,apoptosis, and proliferation

To identify potential mechanisms underlying the efficacy ofDll4 and VEGF blockade, we examined the effects of Dll4 andVEGF blockade on angiogenesis using CD31 staining in theA2780 ovarian cancer model. As shown in Fig. 2, microvesselswere less dense in the groups treated with docetaxel alone oraflibercept alone than in the controls (P < 0.01). Furthermore,aflibercept combined with REGN421 or with REGN1035 led tolowermicrovessel density than the single-drug treatments did (P <0.05; Fig. 2A and B, respectively). In contrast, vessel density wasnot significantly changed in the groups treated with REGN421 orREGN1035 with or without docetaxel (Fig. 2A and B, respective-ly), which is consistent with prior reports (12, 14) that Dll4blockade increased nonproductive tumor vascularity.

To determine the effect of Dll4 and VEGF blockade on theinduction of apoptosis, we performed immunohistochemicalstaining for cleaved caspase-3 in the A2780 model. As shownin Fig. 2A andB, the levels of apoptosiswere significantly higher ingroups treated with REGN421, REGN1035, aflibercept, or doc-etaxel alone than in controls (P < 0.05 for each). Furthermore,aflibercept combined with REGN1035 led to significantly greaterlevels of apoptosis in tumor cells than aflibercept alone (P < 0.01for aflibercept plus REGN1035 vs. aflibercept alone).

In addition, when we examined the effect of Dll4 and VEGFblockade on proliferation in the A2780 model, proliferationwas significantly lower with REGN421 than in controls, sug-gesting a direct effect of REGN421 on ovarian cancer cells.Furthermore, the combinations of aflibercept plus REGN421and docetaxel plus REGN421 led to less proliferation than thesingle-drug treatments did (Fig. 2A; P < 0.05). However, pro-liferation did not differ significantly between the groups treatedwith REGN1035 plus aflibercept or docetaxel and the single-drug treatment groups (Fig. 2B).

RPPA analysis of novel Dll4 downstream targetsTo identify potential downstream targets of Dll4 blockade, we

used RPPA to quantify proteins involved in cell cycle, apoptosis,

angiogenesis, and adhesion in response to single-drug or com-bination treatments in the A2780 model (Fig. 3A). Analysis ofRPPA data revealed that 37 proteins were significantly differen-tially modulated by Dll4 blockade and the combination. Surpris-ingly, the expression level of total GATA3, a master transcriptionfactor, was significantly increased inmice treatedwith REGN1035combined with docetaxel or aflibercept (Fig. 3A). Immunohisto-chemical analysis further validated that GATA3 level was signif-icantly higher in mice treated with REGN1035 combined withdocetaxel or aflibercept than in the controls (Fig. 3B). To furtherinvestigate the potential role of GATA3 in enhancing the efficacyof dual targeting of Dll4 and VEGF signaling in ovarian cancer, wetested GATA3 interacting genes (CX3CL1 and E-cadherin) intumors and found that CX3CL1 expression was significantlyincreased in tumors (Fig. 3B) from mice treated with REGN1035combined with docetaxel or aflibercept, but CX3CL1 was slightlyincreased in the tumor stroma of mice treated with REGN1035combined with docetaxel or aflibercept, which could result froman interaction betweenmurineGATA3 andCX3CL1 (Supplemen-tary Fig. S3). However, the exact mechanism by which CX3CL1expression was increased in cancer cells (Fig. 3) is not clear. It ispossible thatmurineGATA3 could transactivate human genes in aparacrinemanner (Supplementary Fig. S4), but the exactmechan-isms of direct interaction betweenGATA3 andCX3CL1 need to befurther investigated. F4/80 cells and plasma IFNg levels wereexamined in the A2780model. F4/80 cells and plasma IFNg weresignificantly higher in mice treated with combination ofREGN1035 and docetaxel or aflibercept (Fig. 3B and Supplemen-tary Fig. S5).

Given the role of GATA3 in metastasis inhibition via inhib-ition of epithelial–mesenchymal transition (EMT; refs.18–20),we next performed RT-PCR analysis for E-cadherin and foundthat E-cadherin levels were significantly higher in tumors fromthe mice treated with REGN1035 combined with docetaxelor aflibercept compared with controls (Supplementary Fig. S6;ref. 20). E-Cadherin levels were also significantly higher in tu-mor treated with REGN1035 alone and in those treated withdocetaxel plus aflibercept than in the controls.

DiscussionThe key findings from this study are that dual targeting of

VEGF and Dll4 with murine REGN1035 results in superiorantitumor effects in ovarian cancer compared with eithermonotherapy. These effects were related to increased apoptosis,and deregulated angiogenesis, accompanied by the inductionof hypoxia. The mechanisms by which dual targeting of VEGFand Dll4 signaling significantly reduces tumor growth are notfully understood. We demonstrated that the combination ofaflibercept and anti-Dll4 antibody (REGN1035) induced exten-sive apoptosis in ovarian cancer; this apoptosis may have beendue to a profound inhibition of functional tumor vessels.Moreover, downregulation of Dll4-Notch results in aberrantnetwork formation in tumor angiogenesis (13), and Dll4-Fcreduces tumor growth in part by increasing vascular tortuositywhile also decreasing functional flow (8). Other mechanismshave also been reported by our group (14) and others in thatthe interaction of VEGF inhibition and Dll4 inhibitionincreases tumor hypoxia and necrosis (11, 21, 22). VEGF notonly is a major hypoxia-responsive gene but also upregulatesDll4 expression (14, 23). Blocking VEGF inhibits neovessel

Huang et al.





Figure 3.RPPA analysis of Dll4-Notch–targeted genes in A2780-inoculated mice treated with REGN1035, aflibercept, or docetaxel, alone or in combinations.A, heatmap of selective target modulation (e.g., GATA 3) in tumor tissues from mice treated with single-agent or combination therapy. Doce,docetaxel. B, immunohistochemical staining of GATA3, CX3CL1, hypoxia, and F4/80 cells in the A2780 model. All pictures were taken at originalmagnification (200�). Error bars, SEM. � , P < 0.05 compared with controls; �� , P < 0.01; and �� , P < 0.001 compared with controls.






sprouting and reduces Dll4 expression in tumors (23). Con-versely, Notch signaling can downregulate VEGFR2 expressionvia epigenetic mechanisms (14, 23).

In addition, another possible explanation for how dual target-ing of VEGF and Dll4 signaling with murine REGN1035 signif-icantly reduces tumor growth involves GATA3 and its interactingtargets (e.g., E-cadherin and CX3CL1). These potent effects ofcombined aflibercept andDll4 inhibitionmaybedue to increasedGATA3 expression under hypoxic conditions, which could resultfrom an indirect effect of these combination treatments on thetumor stroma.

Increased GATA3 expression in triple-negative breast cancercells has been related to reduced tumorigenicity and metasta-ses by reversing EMT (20). GATA3 can bind GATA-like motif ofE-cadherin promoter and could transactivate gene expression,as human and murine GATA3 shows a high degree of aminoacid sequence identity and similar patterns of tissue specificityof expression in chicken, murine, and human (18, 24). More-over, GATA-3 has also been reported to promote natural killer(NK) cell maturation and IFN production (25). TumoralCX3CL expression has been shown to enhance the recruitmentof NK cells and DCs (26), but the exact mechanisms by whichGATA3 expression in the tumor microenvironment occurs andregulates the human gene expression will require furtherinvestigation.

In summary, superior therapeutic effects were observed forDll4 blockade combined with either aflibercept or docetaxel inovarian cancer models. Data from the phase I study of REGN421(R)/SAR153192, a fully human delta-like ligand 4 (Dll4) mAb inpatients with advanced solid tumors, reported clinical responseand prolonged stable disease in patients with ovarian or othersolid tumors (clinical trial information: NCT00871559; ref. 27).Moreover, a phase I/Ib clinical trial of demcizumab (OMP-21M18, anti-Dll4 antibody, NCT 01952249) in combinationwith weekly paclitaxel in patients with ovarian cancer is ongoingat our institution. On the basis of this work, additional oppor-tunities for clinical development could include the combinationof anti-Dll4 antibody with anti-VEGF drugs and cytotoxic agents(e.g., taxanes).

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

Authors' ContributionsConception and design: J. Huang, W. Hu, L. Hu, M. McGuire, R.L. Coleman,R.B. Jaffe, A.K. SoodDevelopment of methodology: J. Huang, W. Hu, L. Hu, A.K. SoodAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J. Huang, W. Hu, L. Hu, R.A. Previs, H.J. Dalton,X.-Y. Yang, Y. Sun, R. Rupaimoole, Y. Kang, N.B. Jennings, R.B. Jaffe, A.K. SoodAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J. Huang, W. Hu, L. Hu, H.J. Dalton, Y. Sun, N.B.Jennings, R.L. Coleman, R.B. Jaffe, A.K. SoodWriting, review, and/or revision of the manuscript: J. Huang, W. Hu, L. Hu,R.A. Previs, H.J. Dalton, X.-Y. Yang, Y. Sun, M. McGuire, R. Rupaimoole,A.S. Nagaraja, Y. Kang, T. Liu, A.M. Nick, N.B. Jennings, R.L. Coleman, A.K. SoodAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): J. Huang, W. Hu, L. Hu, R.A. Previs, H.J. Dalton,Y. Sun, M. McGuire, A.M. Nick, R.L. Coleman, R.B. Jaffe, A.K. SoodStudy supervision: W. Hu, L. Hu, A.K. Sood

AcknowledgmentsThe authors thank Sarah Bronson and Dawn Chalaire in the Department of

Scientific Publications at The University of Texas MD Anderson Cancer Centerfor editing their manuscript. The authors also thank Regeneron for providingDll4 antibodies and aflibercept.

Grant SupportThis work was supported by the NIH (P50 CA083639, P50 CA098258,

CA109298, UH2 TR000943, and CA177909; to A.K. Sood), the OvarianCancer Research Fund, Inc. (Program Project Development Grant), the JudiA. Rees Ovarian Cancer Research Fund, the Marcus Foundation, the Chap-man Foundation, the Betty Anne Asche Murray Distinguished Professorship(to A.K. Sood), and the institutional Core Grant CA16672 to MD AndersonCancer Center from the NIH. H.J. Dalton and R.A. Previs were supported bya T32 Training Grant (T32CA101642) from the NCI, the U.S. Department ofHealth and Human Services, and the NIH.

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 13, 2015; revised February 26, 2016; accepted February 26,2016; published OnlineFirst March 23, 2016.

References1. Folkman J.Tumor angiogenesis: therapeutic implications. N Engl J Med

1971;285:1182–6.2. Hu L, Hofmann J, Holash J, Yancopoulos GD, Sood AK, Jaffe RB. Vascular

endothelial growth factor trap combined with paclitaxel strikingly inhibitstumor and ascites, prolonging survival in a human ovarian cancer model.Clin Cancer Res 2005;11:6966–71.

3. Coleman RL, Duska LR, Ramirez PT, Heymach JV, Kamat AA, Modesitt SC,et al. Phase 1–2 studyof docetaxel plus aflibercept in patientswith recurrentovarian, primary peritoneal, or fallopian tube cancer. Lancet Oncol 2011;12:1109–17.

4. JainRK,DudaDG,Clark JW, Loeffler JS. Lessons fromphase III clinical trialson anti-VEGF therapy for cancer. Nat Clin Pract Oncol 2006;3:24–40.

5. Casanovas O, Hicklin DJ, Bergers G, Hanahan D. Drug resistance byevasion of antiangiogenic targeting of VEGF signaling in late-stage pan-creatic islet tumors. Cancer Cell 2005;8:299–309.

6. Kerbel RS, Yu J, Tran J, Man S, Viloria-Petit A, Klement G, et al. Possiblemechanisms of acquired resistance to anti-angiogenic drugs: implicationsfor the use of combination therapy approaches. Cancer Metastasis Rev2001;20:79–86.

7. Thanapprapasr D, HuW, Sood AK, Coleman RL. Moving beyond VEGF foranti-angiogenesis strategies in gynecologic cancer. Curr Pharm Des 2012;18:2713–9.

8. Thurston G, Noguera-Troise I, Yancopoulos GD. The Delta paradox: DLL4blockade leads to more tumour vessels but less tumour growth. Nat RevCancer 2007;7:327–31.

9. Hayashi H, Kume T. Foxc transcription factors directly regulate Dll4 andHey2 expression by interactingwith the VEGF-Notch signaling pathways inendothelial cells. PLoS One 2008;3:e2401.

10. Jakobsson L, Bentley K, Gerhardt H. VEGFRs and Notch: a dynamiccollaboration in vascular patterning. Biochem Soc Trans 2009;37:1233–6.

11. Williams CK, Li JL, Murga M, Harris AL, Tosato G. Up-regulation of theNotch ligandDelta-like 4 inhibits VEGF-induced endothelial cell function.Blood 2006;107:931–9.

12. Noguera-Troise I,DalyC, PapadopoulosNJ, Coetzee S, BolandP,GaleNW,et al. Blockade of Dll4 inhibits tumour growth by promoting non-pro-ductive angiogenesis. Nature 2006;444:1032–7.

13. Li JL, Harris AL. Crosstalk of VEGF and Notch pathways in tumourangiogenesis: therapeutic implications. Front Biosci 2009;14:3094–110.

14. HuW, LuC,DongHH,Huang J, ShenDY, StoneRL, et al. Biological roles ofthe Delta family Notch ligand Dll4 in tumor and endothelial cells inovarian cancer. Cancer Res 2011;71:6030–9.

15. Kuhnert F, Chen G, Coetzee S, Thambi N, Hickey C, Shan J,et al. Dll4 blockade in stromal cells mediates antitumor effects


Huang et al.




in preclinical models of ovarian cancer. Cancer Res 2015;75:4086–96.

16. Byrne AT, Ross L, Holash J, Nakanishi M, Hu L, Hofmann JI, et al. Vascularendothelial growth factor-trap decreases tumor burden, inhibits ascites,and causes dramatic vascular remodeling in an ovarian cancer model. ClinCancer Res 2003;9:5721–8.

17. Ridgway J, Zhang G, Wu Y, Stawicki S, Liang WC, Chanthery Y, et al.Inhibition of Dll4 signalling inhibits tumour growth by deregulatingangiogenesis. Nature 2006;444:1083–7.

18. YanW, CaoQJ, Arenas RB, Bentley B, Shao R. GATA3 inhibits breast cancermetastasis through the reversal of epithelial-mesenchymal transition. J BiolChem 2010;285:14042–51.

19. Chu IM, Lai WC, Aprelikova O, El Touny LH, Kouros-Mehr H, Green JE.Expression of GATA3 in MDA-MB-231 triple-negative breast cancer cellsinduces a growth inhibitory response to TGFss. PLoS One 2013;8:e61125.

20. Chou J, Lin JH, Brenot A, Kim JW, Provot S, Werb Z. GATA3 suppressesmetastasis and modulates the tumour microenvironment by regulatingmicroRNA-29b expression. Nat Cell Biol 2013;15:201–13.

21. Jubb AM, Turley H, Moeller HC, Steers G, Han C, Li JL, et al. Expression ofdelta-like ligand 4 (Dll4) and markers of hypoxia in colon cancer. Br JCancer 2009;101:1749–57.

22. Patel NS, Li JL, Generali D, Poulsom R, Cranston DW, Harris AL. Up-regulation of delta-like 4 ligand in human tumor vasculature and therole of basal expression in endothelial cell function. Cancer Res 2005;65:8690–7.

23. Thurston G, Kitajewski J. VEGF and Delta-Notch: interacting signallingpathways in tumour angiogenesis. Br J Cancer 2008;99:1204–9.

24. Linda JK, Yamamoto M, Leonard MW, Geogre KM, TING P, Engel JD.Murine and human T-Lymphocyte GATA-3 factors mediate transcriptionthrough a cis-regulatory element within the human T-Cell receptor 8 geneenhancer. Mol Cell Biol 1991;11:2778–2784.

25. SamsonSI, RichardO, TavianM,RansonT,VosshenrichCA,Colucci F, et al.GATA-3 promotes maturation, IFN-gamma production, and liver-specifichoming of NK cells. Immunity 2003;19:701–11.

26. Park MH, Lee JS, Yoon JH. High expression of CX3CL1 by tumor cellscorrelates with a good prognosis and increased tumor-infiltrating CD8þ Tcells, natural killer cells, and dendritic cells in breast carcinoma. J SurgOncol 2012;106:386–92.

27. Jimeno AP, Strother RM, Diamond JR, Plato L, Younger A, MessersmithWA, et al. Phase I study of REGN421 (R)/SAR153192, a fully-human delta-like ligand 4 (Dll4) monoclonal antibody (mAb), in patients withadvanced solid tumors. J Clin Oncol 31, 2013 (suppl; abstr 2502).






Published OnlineFirst March 23, 2016.Mol Cancer Ther Jie Huang, Wei Hu, Limin Hu, et al. Tumor GrowthDll4 Inhibition plus Aflibercept Markedly Reduces Ovarian

Updated version

10.1158/1535-7163.MCT-15-0144doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2016/03/23/1535-7163.MCT-15-0144.DC1

Access the most recent supplemental material at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://mct.aacrjournals.org/content/early/2016/05/17/1535-7163.MCT-15-0144To 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-15-0144

http://mct.aacrjournals.org/content/suppl/2016/03/23/1535-7163.MCT-15-0144.DC1

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

mailto:[email protected]

http://mct.aacrjournals.org/content/early/2016/05/17/1535-7163.MCT-15-0144


Documents

Dll4 Inhibition plus Aflibercept Markedly Reduces Ovarian ...€¦ · Delta-like ligand 4 (Dll4), one of the Notch ligands, plays an ... University, 128 Shenyang Road, Shanghai, P.R