targeted delivery of an antibody mutant human endostatin … · therapeutic discovery targeted...

Therapeutic Discovery

Targeted Delivery of an Antibody–Mutant Human EndostatinFusion Protein Results in Enhanced Antitumor Efficacy

Seung-Uon Shin1,2, Hyun-Mi Cho1,2, Jaime Merchan1,2, Jin Zhang1,2, Krisztina Kovacs1,2,Yawu Jing4, Sundaram Ramakrishnan4, and Joseph D. Rosenblatt1,2,3

AbstractThe antiangiogenic protein endostatin showed considerable preclinical antitumor activity, but

limited efficacy in phase I/II trials. Prior studies using an anti-HER2 antibody–murine endostatin fusion

showed enhanced antitumor activity compared to anti-HER2 antibody or endostatin given alone, or in

combination. We have generated two anti-HER2 human endostatin fusion proteins by fusing either wild-

type or a mutant human endostatin (huEndo-P125A) to the 3’ end of a humanized anti-HER2 IgG3

antibody. Antitumor efficacy was examined in murine and human breast tumor models. HuEndo-P125A

antibody fusion protein (aHER2-huEndo-P125A) inhibited VEGF and bFGF induced endothelial cell

proliferation, and tube formation in vitro, more efficiently than endostatin alone, wild-type endostatin

fusion protein (aHER2-huEndo), or parental anti-HER2 antibody (aHER2 IgG3). Wild-type and mutant

human endostatin was rapidly cleared from serum in mice (T½2 ¼ 2.0–2.1 hours), whereas aHER2-huEndo

fusion proteins had a significantly prolonged half-life (T½2 ¼ 40.7–57.5 hours). Treatment of SK-BR-3 breast

cancer xenografts with anti-HER2 IgG3-huEndo-P125A fusion resulted in greater inhibition of tumor

growth and improved survival, compared to treatment with either aHER2 IgG3 (P ¼ 0.025), human

endostatin (P ¼ 0.034), or anti-HER2 IgG3-huEndo (P ¼ 0.016). aHER2-huEndo-P125A specifically

inhibited tumors expressing HER2 in mice simultaneously implanted with murine mammary tumor

EMT6 cells and with EMT6 engineered to express HER2 antigen (EMT6-HER2). Targeting of endostatin

using antibody fusion proteins could improve antitumor activity of either anti-HER2 antibody and/or

endostatin and provides a versatile approach that could be applied to other tumor targets with alternative

antibody specificities. Mol Cancer Ther; 10(4); 603–14. �2011 AACR.

Introduction

Endostatin, an antiangiogenic fragment of collagenXVIII, is a global inhibitor of endothelial cell proliferationin vitro and angiogenesis in vivo (1). The mechanism ofaction of endostatin is not fully understood. Systemictherapy with murine endostatin (mEndo) inhibited thegrowth of Lewis lung carcinoma, fibrosarcomas (T241),melanomas (B16F10), hemangioepithelioma (EOMA; refs.1 and 2), and human renal cell cancer xenografts (3).Human and mouse collagen XVIII chains show a high

degree of homology (4). Human endostatin (huEndo)inhibited the growth of several different tumors in vivo,such as human glioblastoma (U-87MG), C6 rat glioma, orrat gliosarcoma (BT4Cn; refs. 5–7). Repeated treatmentwith endostatin led to permanent eradication of tumor inseveral rodent models (1, 2, 8).

Although initial trials proved that endostatin is safewhen delivered in a variety of dose schedules, they didnot show comparable antitumor activity to that seen inmurine models. In human phase I trials, huEndo admin-istration at variable dose schedules was feasible and safebut, little or no consistent antitumor activity was shown(9–13). In a phase II study in 42 patients with advancedpancreatic neuroendocrine tumors or carcinoid tumorstreated with huEndo administered as a subcutaneousinjection, huEndo showed minimal toxicity (13), but nopatient achieved a partial response.

Several explanations have been advanced for the fail-ure to see antitumor activity in the human setting. It hasbeen suggested that initial preparations of endostatin hadsuboptimal biologic activity and a modified formulations(e.g., EndoStar) may be more potent and have beenapproved for use in China (14–16). Endostatin may be

Authors' Affiliations: Departments of 1Medicine, 2Hematology-Oncol-ogy, and 3Microbiology and Immmunology, University of Miami, MillerSchool of Medicine and Sylvester Comprehensive Cancer Center, Miami,Florida; and 4Department of Pharmacology, University of Minnesota,Minneapolis, Minnesota

Corresponding Author: Seung-Uon Shin, Department of Medicine,Hematology-Oncology, University of Miami School of Medicine and Syl-vester Comprehensive Cancer Center, 1475 N.W. 12th Avenue (D8-4)Miami, FL 33136. Phone: 305-243-3668; Fax: 305-243-4787; E-mail:[email protected]

doi: 10.1158/1535-7163.MCT-10-0804

�2011 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org 603

Research. on December 1, 2020. © 2011 American Association for Cancermct.aacrjournals.org Downloaded from

Published OnlineFirst March 10, 2011; DOI: 10.1158/1535-7163.MCT-10-0804

http://mct.aacrjournals.org/

more potent when delivered as a dimer or trimer thanwhen delivered in monomeric form (17, 18). Finally, amutant form of endostatin (P125A) has been reported tohave enhanced antiangiogenic activity (19–21). Endosta-tin has also been reported to show a bimodal responsecurve such that optimal concentrations may not havebeen achieved in phase I/II studies (22).

To improve efficacy of trastuzumab and endostatin, weconstructed several endostatin fusion proteins by joininghuman endostatin to the 3’ end of humanized anti-HER2IgG3. We reasoned that delivery of endostatin using anantibody fusion protein would increase endostatin half-life, enhance antiangiogenic activity due to endostatindelivery as a dimer, and selectively deliver endostatin tothe site of tumor, thereby improving efficacy. We hadpreviously validated the efficacy of this approach usingan anti-HER2 antibody fused to murine endostatin. Inthis study, we report on the biological activity of anaHER2 IgG3 fused to either a wild type or a mutant form(P125A) of human endostatin with markedly enhancedantiangiogenic activity. We show that this fusion proteinis associated with significantly enhanced antiangiogenicand antitumor efficacy in several cancer models.

Materials and Methods

Cell lines, materials, and animalsMurine mammary tumor EMT6 cells were transduced

with a retroviral vector encoding human HER2 asdescribed (23, 24). EMT6, SK-BR-3, Sp2/0, andP3�63Ag8.653 cells were purchased from and character-ized by the American Type Culture Collection. EMT6,EMT6-HER2, the human breast cancer cell line SK-BR-3,and transfected Sp2/0 or P3�63Ag8.653 cells were cul-tured in Iscove’s Modified Dulbecco’s Medium with 5%calf serum. Expression of HER2was confirmed on EMT6-HER2 and SK-BR-3 cell lines using flow cytometry. Sp2/0and P3�63Ag8.653 were tested by ELISA for antibodyexpression. Human recombinant endostatin was pur-chased from commercial sources (E8154, Sigma). FemaleBALB/c mice (4–6 weeks) and severe combined immu-nodeficient (SCID) mice (4–6 weeks, Jackson Laboratory)were used for in vivo tumor growth and xenograft experi-ments (SK-BR-3) as indicated. All experiments were con-ducted in compliance with the NIH Guides for the Careand Use of Laboratory Animals and approved by theUniversity of Miami Institutional Animal Care and UseCommittee.

Construction, expression, and characterization ofaHER2-huEndo fusion protein

The huEndo genewas cloned from the human collagen,type XVIII, a1 gene by PCR using primers 50-CCCCTCGCGATATCACAGCCACCGCGACTTCCAG-CCG-30 and 50-CCCCGAATTCGTTAACCCTTGGAGG-CAGTCATGAAGC-30. PCR products were subclonedinto pCR-Blunt II-TOPO vector. A single point mutationin human endostatin at position 125 (proline to alanine;

huEndo-P125A) has been reported to enhance antiangio-genic activity (19–21). Amutant clone inwild-type humanendostatin encoding an alanine residue substituted forproline at position 125 was derived by site-directedmuta-genesis using PCR with phosphorylated primer, 50p-GGCTCGGACGCCAACGGGCGC-30. The subclonedhuEndo and huEndo-P125A genes were ligated in frameto the carboxyl end of the heavy chain constant domain ofhuman IgG3 in the vector pAT135 (24, 25). The endostatinheavy chain constant region was joined to the C-terminusof the anti-HER2 heavy chain of a recombinant huma-nized monoclonal antibody 4D5–8 (Genentech) in theexpression vector (pSV2-his) containing a HisD gene foreukaryotic selection (23, 26).

The aHER2-huEndo fusion constructs were stablytransfected into SP2/0 or P3�63Ag8.653 myeloma cellsexpressing the anti-HER2 k light chain by electroporationas described previously (27). The endostatin fusion pro-teins were purified from culture supernatants using pro-tein A immobilized on Sepharose 4B fast flow (yield: 1–5mg/L of supernatant; Sigma) as described (27). aHER2-huEndo fusion proteins were biosynthetically labeledwith [35S]methionine (Amersham Biosciences), immuno-precipitated using IgGSorb suspension, and analyzed bySDS-PAGE as described (27).

P125A-endostatin was cloned and expressed in Pichiapastoris (20). Following methanol based induction of thetransgene, P125A-endostatin was purified from culturesupernatants using heparin linked ceramic particles (20).

Flow cytometryTo detect the binding of aHER2-huEndo fusion pro-

teins to HER2 antigen and human umbilical veinendothelial cells (HUVEC; Clontech Lab) human SK-BR-3 breast cancer cells, murine mammary tumor cells,EMT6 and EMT6-HER2, and HUVECs were incubated at4�C with 1 mg/mL of endostatin fusion proteins, aHER2IgG3, or isotype control. Fifteen minutes later, the cellswere washed with PBS containing 0.1% BSA and 0.05%NaN3, and the bound fusion proteins were identifiedwith either FITC conjugated antihuman IgG, or the endo-statin domain was recognized with biotinylated antihu-man endostatin antibody (R&D Systems, Inc.) andsecondarily stained with a streptavidin-PE conjugate(Sigma) at 4�C, washed twice and resuspended in PBSand analyzed by flow cytometry. Background stainingwas estimated following incubation with the secondaryFITC or PE-labeled antibody alone.

Binding to HUVECs was used to determine endostatindomain binding, using a biotinylated antihuman endo-statin antibody and streptavidin conjugated PE as indi-cated earlier.

Matrigel tube formation assayHUVECs were used between passages 3 and 5, and

maintained in EGM2-MV medium (Clontech) that con-tained endothelial basal medium 2 (EBM-2), supplemen-tedwith 2% fetal bovine serum, gentamicin, amphotericin

Shin et al.

Mol Cancer Ther; 10(4) April 2011 Molecular Cancer Therapeutics604




B, hydrocortisone, ascorbic acid, vascular endothelialgrowth factor (VEGF), basic fibroblast growth factor(bFGF), human epidermal growth factor, and insulin-likegrowth factor I as described (28, 29).The Matrigel tube formation assay was done in 48-well

plates, as previously reported (28, 29). Each well ofprechilled 48-well cell culture plates was coated with100 mL of unpolymerized Matrigel (7 mg/mL) and incu-bated at 37�C for 30–45minutes. HUVECswere harvestedwith trypsin, and 4 � 104 cells were resuspended in 300mL of full EC growth medium and treated with endo-statin, control antibody or the various aHER2-huEndofusion proteins at indicated concentrations before platingonto the Matrigel-coated plates. Following 16 hours ofincubation at 37�C, EC tube formation was assessed withan inverted photomicroscope, and microphotographs ofthe center of each well were taken at low power (40�).Tube formation by untreated HUVECs in full EC growthmedium was used as a control.

Pharmacokinetics of aHER2-huEndo fusionproteinsTo monitor the levels of human endostatin fusion

proteins in serum, BALB/c mice (n ¼ 3, 4–6 weeks) wereinjected i.v. with either aHER2-huEndo (50 mg, equimolarto 9.52 mg human endostatin), aHER2-huEndo-P125A (50mg, equimolar to 9.52 mg human endostatin), huEndo(10 mg), or huEndo-P125A (10 mg). Blood samples wereserially obtained at indicated intervals aHER2-huEndo oraHER2-huEndo-P125A. Mice injected with eitherhuEndo or huEndo-P125A was bled within 1 minute to2 hours after the i.v. injection. A quantitative ELISA forserum endostatin (QuantikineHuman ES ELISA kit, R&DSystems) was used to detect both endostatin and endo-statin fusion proteins according to the manufacturer’sinstructions.The pharmacokinetic variables were calculated by fit-

ting human endostatin concentration data in serum to abiexponential model with derivative free nonlinearregression analysis (PK Solution, version 2.0.6 Programdeveloped at Summit Research Services. The pharmaco-kinetic variables, such as serum distribution and elim-ination, steady-state area under the serum concentrationcurve (AUC0-¥), and mean residence time, werecalculated.

In vivo tumor growth assaysTo evaluate antitumor activity of aHER2-huEndo

fusion proteins in vivo, human SK-BR-3 breast cancercells (2 � 106) were implanted s.c. in the flanks of SCIDmice (24). Starting on day 5–6, mice were injected i.v. withequimolar amounts of purified aHER2-huEndo fusionproteins (42 mg), aHER2 IgG3 (34 mg), trastuzumab (30mg), human endostatin (8 mg), and human mutant endo-statin-P125A (8 mg) as indicated. Control mice wereinjected with PBS. This treatment was repeated everyother day for 11–13 doses as indicated. Tumor size wasmeasured with calipers and growth rates were recorded

and calculated using the following equation: tumorvolume (mm3) ¼ 4/3 � 3.14 � [(long axis þ shortaxis)/4]3.

The in vivo antitumor efficacy and specificity ofaHER2-huEndo fusion protein was also examined usingthe EMT6 and EMT6-HER2 cell lines simultaneouslyimplanted contralaterally in syngeneic BALB/c mice.The EMT6-HER2 cells that were used in these studiesproliferate at about the same rate in vitro as parentalEMT6 cells (24). To determine efficacy of aHER2-huEndofusion proteins, BALB/c mice (3–8 per group, 4–6 weeks)were injected s.c. with 1 � 106 EMT6-HER2 cells in theright flank and/or control EMT6 cells in the left flank. Onday 6, mice were injected i.v. with the aHER2-huEndofusion proteins (42 mg/injection, 2� 10�10mol, equimolarto 8 mg human endostatin), aHER2 IgG3 alone (34 mg/injection, 2� 10�10 mol), endostatin alone (8 mg/injection,4 � 10�10 mol), or PBS as a control. All mice received atotal of 11 injections at 2-day intervals and growth ofEMT6 and contralateral EMT6-HER2wasmonitoredwithcalipers.

Immunofluorescent staining of tumor vasculatureEMT6 and EMT6-HER2 cell lines (1 � 106) were

implanted contralaterally in syngeneic BALB/c mice(n ¼ 4) as described (24). Starting on day 4, mice wereinjected i.v. with aHER2-huEndo-P125A (42 mg/injec-tion), or PBS as a control every 2 days. On day 12, micewere sacrificed for analysis of vascularity following 4 ofthe aforementioned treatments. Tumors were excisedfrom the sacrificed mice, frozen in liquid nitrogen and8 mm frozen sections were prepared. The tumor sectionswere fixed with methanol for 10 minutes, washed withPBS 3 times, and incubated with blocking solution for 1hour in a humidified chamber and again washed withPBS 3 times. For immunofluorescent staining, dilutedprimary antibodies [rat anti-mouse CD31 antibody con-jugated with biotin, 1:200 in PBS (BD Biosciences)] wereadded to each slide. Following incubation at room tem-perature overnight, sections were incubated with dilutedsecondary antibodies conjugated with Alexa Fluor 488(1:500) with PBS, and thenwith diluted DAPI (1:5000) in ahumidified chamber and mounted with Gel mountingmedia (Biomeda Corp.). Alexa Fluor 488 was obtainedfrom Invitrogen, and DAPI from Molecular Probes. Thestained images from 5 low power fields were analyzedwith a Zeiss microscope. Digital tumor images from eachtreatment were measured for blood vessel area (pixel2),averaged to measure blood vessel density per tumor, andvessel number and density analyzed using NIH ImageJsoftware.

Statistical analysisStatistical analysis was carried out with Graphpad

Prism 4 (GraphPad Software, Inc.). HUVEC proliferationand antitumor efficacy in human breast cancer SK-BR-3xenografts were analyzed by 2-way repeated measures(RM) analyses of variance (ANOVA), followed by

Antibody–Human Endostatin Fusion Protein

www.aacrjournals.org Mol Cancer Ther; 10(4) April 2011 605




Bonferroni posttest. Tumor growth in the murine syn-geneic tumor model was compared using Student’s t test(unpaired, 2-tailed) for each treatment group, and wasanalyzed among 3 different treatments by 1-wayANOVA, followed by Bonferroni’s multiple comparisontest. Graphs were expressed as the mean values with 95%confidence interval (CI). Differences were consideredstatistically significant at P < 0.05.

Results

Construction and purification of aHER2-huEndofusion proteins

The cloned huEndo and huEndo-P125A genes weregenetically fused to the 3’ end of a humanized anti-HER2IgG3 antibody heavy chain gene. The anti-HER2 IgG3-huEndo fusion protein heavy chain constructs were sta-bly transfected into myeloma cells expressing the anti-HER2 kappa light chain in order to assemble the entireanti-HER2 IgG3-huEndo fusion proteins, anti-HER2IgG3-huEndo (aHER2-huEndo) and anti-HER2 IgG3-huEndo-P125A (aHER2-huEndo-P125A; Fig. 1A). Theresulting aHER2-huEndo and (aHER2-huEndo-P125Afusion proteins were biosynthetically labeled with[35S]methionine and analyzed by SDS-PAGE. aHER2-huEndo fusion proteins of the expectedmolecular weightwere secreted as the fully assembled H2L2 form (Fig. 1B).The secreted [35S]methionine-labeled proteins had amolecular weight of 220 kDa under nonreducing condi-tions, the size expected for a complete antibody (170 kDa)with 2 molecules of endostatin (25 kDa each) attached(Fig. 1B). Following reduction, heavy and light chains ofthe expected molecular weight were observed (85 kDaand 25 kDa, respectively; Fig. 1B). The secreted endosta-tin fusion proteins were then purified from culture super-natants using a protein A column.

Binding to HER2 target antigen and to HUVECsHuman HER2þSK-BR-3 cells, and murine EMT6 and

EMT6-HER2 were used to test whether the endostatinfusion proteins could recognize the HER2 antigen(Fig. 1C). Using antihuman IgG antibody as a detectionantibody, aHER2 IgG3, aHER2-huEndo, and aHER2-huEndo-P125A, bound to HER2þSKBR3 and EMT6-HER2 cells (Fig. 1C-I and II, respectively), whereas theisotype control antibody (anti-dansyl IgG3) did not bind.

aHER2 IgG3, aHER2-huEndo, or aHER2-huEndo-P125A did not bind to parental EMT6 cells (Fig. 1C–III).

To investigate structural integrity of the endostatindomain, the human endostatin domain of fusionproteins bound to SK-BR-3 and EMT6-HER2 wasdetected with biotinylated antihuman endostatin andstained with streptavidin-PE conjugate. aHER2-huEndoand aHER2-huEndo-P125A were both recognized fol-lowing binding to SK-BR-3 and EMT6-HER2 by theantihuman endostatin detection antibody (Fig. 1C-Vand VI), whereas aHER2 IgG3 and the isotype controlantibodies were not detected.

To determine whether aHER2-huEndo fusion proteinscould also bind to endothelial cells, HUVECs were trea-ted with aHER2 IgG3, huEndo, huEndo-P125A, oraHER2-huEndo fusion proteins, and binding detectedeither with antihuman IgG-FITC, or with biotinylatedantihuman endostatin antibody and stained with strep-tavidin-PE conjugate. Binding of huEndo, huEndo-P125A, and aHER2-huEndo fusion proteins to HUVECswas readily detected by antihuman endostatin (Fig. 1C–VII), whereas bound aHER2-huEndo fusion proteinswere also readily detected by antihuman IgG-FITC(Fig. 1C–IV). Binding of the isotype control, or of aHER2IgG3 was not detected using either reagent. Therefore,aHER2-huEndo and aHER2-huEndo-P125A bound toHER2 and the fused endostatin domain(s) bound toendothelial cells.

Inhibition of endothelial tube formation andendothelial cell proliferation

We tested the effects of the aHER2-huEndo fusionproteins in an in vitro angiogenesis assay in which humanendothelial cells are plated on Matrigel, and sponta-neously aggregate and assemble into multicellular capil-lary-like tubular structures in response to vascularstimuli (e.g., bFGF, VEGF, fetal bovine serum; refs. 28and 29). Neither parental antibodies (aHER2 IgG3 ortrastuzumab) nor human endostatin (wild type orP125A mutant type) alone showed appreciable inhibitionof tube formation at concentrations tested. In contrast,aHER2-huEndo fusion proteins, aHER2-huEndo (Fig. 2Eand F), and aHER2-huEndo-P125A (Fig. 2H and I)strongly inhibited assembly into tubular structures, com-pared to aHER2 IgG3 (Fig. 2B), trastuzumab (Fig. 2C),huEndo (Fig. 2D), or huEndo-P125A (Fig. 2G). HUVECstreated with aHER2-huEndo fusion proteins remaineddispersed and exhibited a scattered morphology in dose-dependent fashion (Fig. 2). Inhibition of tube assemblyseen with aHER2-huEndo-P125A (Fig. 2H and I) wassignificantly greater than that seen for aHER2-huEndo(Fig. 2E and F) at comparable concentrations, and treat-ment of HUVEC at 45 nmol/L resulted in completedisruption of tube formation (Fig. 2H and I).

Inhibition of tube assembly has been previouslyreported with the NC1 domain of collagen XVIII andoligomeric forms of endostatin (17, 18). The increased invitro antiangiogenic effect of aHER2-huEndo fusion pro-teins relative to native endostatin may be due to presen-tation of endostatin as a dimer as previously reported (17,18). As morphologic changes were observed within 16hours, this dispersed phenotype was due to scattering ofendothelial cells rather than proliferation (division timefor endothelial cells � 24 hours).

We also assessed the effects of aHER2-huEndo fusionproteins on EC proliferation. HUVECs were exposed toincreasing concentrations of the fusion proteins for 72hours in the absence or presence of either VEGF or bFGF.Both wild-type and mutant antibody–endostatin fusionproteins markedly inhibited EC proliferation induced by

Shin et al.





either VEGF or bFGF. HUVEC proliferation was moreeffectively inhibited at comparable concentrations byaHER2-huEndo-P125A than by aHER2-huEndo (P ¼0.0085 in the presence of VEGF, P¼ 0.0034 in the presenceof bFGF) or by endostatin alone (P ¼ 0.0003 in thepresence of VEGF or bFGF; data not shown).

Serum clearance and stability of HER2-endoTo characterize the pharmacokinetics of aHER2-

huEndo, mice were injected i.v. with aHER2-huEndo,

aHER2-huEndo-P125A, huEndo, or huEndo-P125A, andclearance of injected proteins wasmeasured by ELISA forhuman endostatin. Representative results from mice areshown graphically in Fig. 3, and the pharmacokineticdata for all mice are summarized in Table 1. HuEndo andhuEndo-P125A were rapidly removed from the serumcompartment in mice (T½

2 elimination, 2.0–2.1 hours),whereas the T½

2 of aHER2-huEndo and aHER2-huEndo-P125A fusion proteins (T½

2, 40.7 and 57.5 hours, respec-tively) were significantly increased compared to that of

Anti-HER2 lgG3

A

C

B

ααHE

R2

lgG

3

αHE

R2

lgG

3

VH

CH1

VL

Cκ

Endostatin

SK-BR-3

I II III IV

VIIVIV

010

2030

40C

ount

s

FITC100 101 102 103 104

PE100 101 102 103 104

PE100 101 102 103 104

PE100 101 102 103 104

FITC100 101 102 103 104

FITC100 101 102 103 104

hIgG-FITC100 101 102 103 104

5060

010

2030

40C

ount

s

010

20

020

4060

8010

0

3040

Cou

nts

5060

700

1020

3040

Cou

nts 50

010

2030

40C

ount

s

020

4060

8010

012

0C

ount

s

50

6070

80

EMT6-HER2 EMT6 HUVEC

SK-BR-3 EMT6-HER2 HUVEC

Endostatin(P125A)

H2L2

M.W.kDa

-220H

L-

-97

-66

-46

-30

-20Non-reduced

Reduced

M.W.kDa

-220

-97

-66

-46

-30

-20

CH2

CH3

H

αHER2-huEndo

αhu

man

IgG

-FIT

Cαh

uE

nd

o A

b/P

E

αHE

R2-

hE

nd

o

αHE

R2-

hE

nd

o

αHE

R2-

mE

nd

o

αHE

R2-

mE

nd

o

αHER2-huEndo-P125A

αHE

R2-

hE

nd

o-P

125A

αHE

R2-

hE

nd

o-P

125A

Figure 1. A, schematic diagram of anti-HER2 IgG3-human endostatin fusion proteins. Details in text. B, expression of anti-HER2 IgG3-CH3-endostatin fusionproteins. Secreted human endostatin fusion proteins were labeled with [35S]methionine and immunoprecipitated and analyzed under reducing andnonreducing conditions. Anti-HER2 IgG3-CH3-murine endostatin fusion (aHER2-mEndo) was used as a control (24). C, binding of anti-HER2 humanendostatin fusion proteins to HER2 antigen and HUVECs, and recognition by antihuman endostatin antibody. Human breast cancer cells, SK-BR-3 (I, V),murine mammary tumor cells, engineered to express HER2, EMT6-HER2 (II, VI) and EMT6 (III), or HUVECs (IV, VII) were incubated with aHER2-huEndo (filledwith gray), aHER2-huEndo-P125A (red line), aHER2 IgG3 (green line), human endostatin (blue line), human endostatin-P125A (red line filled with pink),or isotype control (anti-dansyl IgG3, black line). The bound fusion proteins were identified with either antihuman IgG-FITC conjugated (I–IV) or with biotinylatedantihuman endostatin antibody and secondarily stained with a streptavidin-PE conjugate (V–VII).






human endostatin. There is no significant difference inserum elimination between huEndo (T½

2

elimination, 2.0 � 0.4 hours) and huEndo-P125A (T½2

elimination, 2.1� 0.2 hours), or between aHER2-huEndo(T½

2 elimination, 57.5 � 3.9 hours) and aHER2-huEndo-P125A (T½

2 elimination, 40.7 � 1.8 hours). Therefore,human endostatin fused with antibody has a serumhalf-life of at least 20-fold greater than human endostatinalone.

AUC of the human endostatin fusion proteins wasincreased by a factor of 56 compared with huEndo orhuEndo-P125A (1.01 � 0.11 mg�h/mL or 0.74 � 0.05mg�h/mL, respectively) as a consequence of both a longerhalf-life of elimination and an increased mean residencetime (Table 1). The AUC of aHER2-Endo-P125A (293.15� 12.78 mg�h/mL) was slightly increased compared withaHER2-huEndo (201.52 � 14.83 mg�h/mL; Table 1).

Antitumor efficacy in human breast cancer SK-BR-3xenografts

SK-BR-3 is a HER2-amplified human breast cancer cellline which grows slowly as a xenograft in SCID mice.Trastuzumab, an anti-HER2 IgG1, is able to inhibit thegrowth of human breast cancer SK-BR-3 alone or incombination with chemotherapy (30). We assayed forantitumor activity of aHER2-huEndo fusion proteinsagainst human breast cancer SK-BR-3 xenografts in SCIDmice. Representative experiments are shown in Figure 4.Equimolar doses of protein were injected every other dayfor 4 weeks (Fig. 4). Endostatin and aHER2 IgG3 did notsignificantly inhibit tumor growth relative to the non-treated group (PBS, P ¼ 0.1504) by day 29, aHER2 IgG3inhibited tumor growth relative to the nontreated group(PBS, P ¼ 0.0045), whereas treatment with aHER2-huEndo or aHER2-huEndo-P125A resulted in signifi-cantly greater inhibition of growth (P < 0.0001, respec-tively; Fig. 4A). Mice treated with aHER2-huEndo-P125Ashowed significantly reduced tumor growth compared to

Figure 3. Serum clearance of human endostatin fusion proteins followingintravenous administration to mice. aHER2-huEndo (open circle),aHER2-huEndo-P125 (filled circle), huEndo (open triangle), and huEndo-P125A (filled triangle) were injected via tail vein of BALB/c mice (n ¼ 3). Atthe indicated time points, serum samples were assayed by an ELISA,which detects human endostatin. The data are presented as themean � 95% CI. A representative experiment is shown.

Figure 2. Effects of anti-HER2IgG3-huEndo fusion proteins onendothelial cell tube formation.HUVECs were resuspended inendothelial cell growth mediumand treated as indicated beforeplating onto the Matrigel-coatedplates. Following 16 hours ofincubation, tube formation wasobserved through an invertedphotomicroscope. Tube formationwith media alone (A), 45.46 nmol/LaHER2 IgG3 (B), 45.46 nmol/Ltrastuzumab (C), 45.46 nmol/LhuEndo (D), and 45.46 nmol/LhuEndo-P125A (G) werecompared to those with 22.73 and45.46 nmol/L aHER2-huEndo (Eand F), and 22.73 and 45.46 nmol/L aHER2-huEndo-P125A (H andI). Experiments were repeatedtwice. A representativeexperiment is shown.

Shin et al.





those treated with aHER2-huEndo (P ¼ 0.0161), humanendostatin (P ¼ 0.0343), or aHER2 IgG3 (P ¼ 0.0253;Fig. 4A). Similar results were seen in 3 separate experi-ments (Fig. 4A). Treatment with aHER2-huEndo-P125Acompletely eradicated tumors after 30 days (Fig. 4A) andno tumor recurrence was observed until the experimentwas terminated at day 40 (data not shown).

In separate experiments, mice treated with aHER2-huEndo-P125A showed improved survival relative tothose treated with huEndo (P ¼ 0.0005), huEndo-P125A (P < 0.0001), trastuzumab (P < 0.0001), or aHER2IgG3 (P < 0.0001; Fig. 4B). Mice treated with huEndo,huEndo-P125A, trastuzumab, or aHER2 IgG3 all showedsignificantly improved survival to the untreated group(PBS: P < 0.0001), but there was no significant differencebetween huEndo and huEndo-P125A (P ¼ 0.2233) orbetween trastuzumab and aHER2 IgG3 (P ¼ 0.5995;Fig. 4B).

Antitumor efficacy requires presence of the HER2target

To investigate whether the ability of aHER2-huEndo-P125A fusion protein to specifically target HER2enhanced efficacy, BALB/c mice were simultaneouslyimplanted with EMT6 and EMT6-HER2 tumors onopposite flanks. Mice were then treated with eitheraHER2-huEndo-P125A, aHER2 IgG3, or human endo-statin (Fig. 5A). aHER2-huEndo-P125A inhibitedEMT6-HER2 tumor growth more effectively than PBS(P < 0.001), endostatin (P < 0.001), or aHER2 IgG3 (P ¼0.0001), but comparable antitumor activity was notseen against EMT6 (P > 0.05; Fig. 5B). aHER2 IgG3inhibited EMT6-HER2 tumor growth more effectivelythan PBS (P ¼ 0.001), but not than human endostatin(P ¼ 0.196; Fig. 5B). Equimolar administration ofaHER2-huEndo-P125A and aHER2 IgG3 showed pre-ferential growth inhibition of EMT6-HER2, when com-pared to parental EMT6 implanted on the contralateralflank (P < 0.0001, P ¼ 0.0004, respectively; Fig. 5C),

Table 1. Pharmacokinetic variables for human endostatin fusion proteins in mice

Variablea aHER2-huEndo aHER2-huEndo-P125A huEndob huEndo-P125Ab

T½1 (h): distribution 0.463 � 0.023 0.772 � 0.049 0.060 � 0.000 0.066 � 0.000

T½2 (h): elimination 57.5 � 3.9 40.7 � 1.8 2.0 � 0.4 2.1 � 0.2

AUC0-96 (mg�h/mL) 126.22 � 4.58 227.40 � 4.97 0.62 � 0.01 0.45 � 0.00AUC0-¥ (mg�h/mL) 201.52 � 14.83 293.15 � 12.78 1.01 � 0.12 0.74 � 0.05MRT (h) 89.4 � 6.0 58.9 � 3.1 2.3 � 0.5 2.3 � 0.3

aFor the pharmacokinetic variables, the superscript 1 represents the distribution phase and the superscript 2 represents theelimination phase. AUC0-96 and AUC0-¥ are the first 96 hours and steady-state AUC, respectively. To calculate pharmacokineticvariables, the concentrations of endostatin in serum were fit to a biexponential model (aHER2-huEndo, aHER2-huEndo-P125A,huEndo, and huEndo-P125A). Data are mean � SEM.bMeasurements of endostatin were made 2 hours after i.v. injection in the mice.

Figure 4. A, antitumor efficacy of anti-HER2 IgG3-huEndo fusionproteins. SCID mice (n ¼ 5) were implanted s.c. with 2 � 106 SK-BR-3on day 0, then i.v. injected with aHER2-huEndo fusion proteins (42 mg),aHER2 IgG3 (34 mg), huEndo (8 mg), or PBS every other day (arrows)starting on day 5. Tumor growth was measured as described earlier. Thevalues represent mean � 95% CI of tumor volume (mm3) of 5 mice.Experiments were repeated 3 times with similar results. A representativeexperiment is shown. B, survival of mice per treatment group. aHER2-huEndo-P125A (42 mg), huEndo (8 mg), huEndo-P125A (8 mg), aHER2IgG3 (34 mg), trastuzumab (30 mg), or PBS every other day (arrows)starting on day 6. Mice with greater than 2500 mm3 tumor volume wereeuthanized. The proportion surviving in each mouse group (%) isindicated. N. Sig., not significant. (This is a separate experiment fromthat shown in A.)






whereas PBS (P ¼ 0.4014) and endostatin treatment(P ¼ 0.1665) showed no significant difference betweenEMT6-HER2 and EMT6 tumor growth (Fig. 5C). There-fore, selective targeting of HER2 by the fusion proteinwas required for maximum efficacy.

Immunofluorescent staining of vasculature oftreated tumors

To investigate the effects of aHER2-huEndo-P125Afusion protein on tumor angiogenesis, histologic sec-tions of tumors were derived from treated anduntreated mice after 4 treatments, and tumor vascula-ture was visualized and total vessel density was quan-tified using anti-PECAM fluorescence immunostaining(Fig. 6). Immunofluorescent staining of EMT6-HER2tumors demonstrated that the antibody-endostatinfusion treated group showed thin, short, and fragmen-ted blood vessels on day 12 (Fig. 6B), compared to moreextensively arborizing vasculature in the PBS treated

group (Fig. 6A). Treatment with aHER2-huEndo-P125Afusion protein caused a statistically significant decreasein total blood vessel density in tumors (Fig. 6C, P ¼0.0038). To investigate average blood vessel area, totalVd was divided by vessel number to yield averagevessel density. Tumors treated with endostatin fusionshowed a significant reduction in average vessel den-sity (Fig. 6D, P ¼ 0.0009). Treatment with aHER2-huEndo-P125A resulted in a marked reduction in tumorvasculature, as well as profound changes in vasculararchitecture.

Discussion

Antiangiogenic therapywith endostatin has been shownto block tumor growth with little or no emergence ofresistance despite multiple cycles of therapy, in a varietyof murine models. In several studies, repeated treatmentwith endostatin resulted in permanent eradication of

Figure 5. Antitumor activity of aHER2-huEndo fusion-P125A protein depends on HER2 targeting. BALB/c mice (n ¼ 3–8 per group) were implanted s.c.contralaterally with EMT6 (E) and EMT6-HER2 (EH; 1 � 106 cells per mouse), followed on day 6 by equimolar injections every other day (11 times) ofaHER2-huEndo-P125A (42 mg), aHER2 IgG3 (34 mg), human endostatin (huEndo, 8 mg), or PBS. A, tumor volume was measured as described earlier. The dataare presented as the mean � 95% CI. B, individual tumor size was compared for EMT6 and EMT6-HER2 on day 16. C, EMT6 and EMT6-HER2 tumormeasurements of individual mice on day 16 were compared for each treatment group as indicated. Similar results were seen in 2 duplicate experimentsof which a representative experiment is shown.

Shin et al.





tumors (1, 2, 8). However, phase I/II studies of humanendostatin did not show the levels of antitumor activityseen inmurinemodels (9–13).We hypothesized that endo-statin performance could be improved if the half-life ofendostatin could be extended and if endostatin could bespecifically targeted to the tumor, to achieve higher localconcentrations. In addition, we hypothesized that endo-statin might be more effective if delivered as a dimer as inthe context of an antibody fusion protein (17, 18). A pro-totypic anti-HER2-murine endostatin fusion proteinretained antiangiogenic activity, prolonged serum half-lifecompared toendostatin, targetedHER2expressing tumors,and inhibited in vivo tumor growth which provided initialvalidation for this concept (24).To reduce the possible antigenicity of the murine

endostatin fusion domain in preparation for humanapplication, we have now constructed 2 new fusionsbased on human endostatin and on a mutated form ofendostatin with increased antiangiogenic properties.Similar to results obtained with a murine endostatinfusion protein (24), human endostatin fused with anti-body had a significantly longer serum half-life thanhuman endostatin alone. Yokoyama and colleaguesreported that a mutant version of endostatin in which

the proline at 125 is substituted with an alanine,showed greater antiangiogenic activity than nativeendostatin in vitro (19–21). Consistent with this, theaHER2-huEndo-P125A fusion protein showed greaterinhibition of tube formation in vitro than either nativeendostatin, mutant endostatin-P125A, or wild typeaHER2-huEndo fusion.

It has been reported that human or murine endostatintreatment inhibits HUVEC assembly into tubular struc-tures in vitro, with cells remaining dispersed and exhi-biting a morphology resembling adherent cells onplastic rather than aggregating into characteristic capil-lary-like tubes (17, 18). Dimers or trimers of endostatinstimulated the motility of endothelial cells, more effi-ciently than monomers and endostatin oligomerizationwas required for the efficient inhibition of tube forma-tion (17, 18). Because of the presence of 2 fused heavychain-endostatin domains in the assembled H2L2 form(Fig. 1B), aHER2-huEndo fusion proteins may effec-tively present endostatin as a dimer. Dimerization ofthe endostatin domains could in theory augment bind-ing to integrins, perlecan, and glypicans, furtherincreasing fusion protein activity in vitro and in vivo(17, 18).

Figure 6. Effect of treatment on tumor vascularity. BALB/c mice (n ¼ 4 per group) were implanted s.c. contralaterally with EMT6 and EMT6-HER2 (1�106

cells per mouse), followed on day 4 by equimolar injections every other day of aHER2-huEndo-P125A (P125A, 42 mg) or PBS. On day 12, mice weresacrificed after 4 treatments. Histologic sections of tumors from the sacrificedmice were analyzed using immunofluorescent staining for PECAM (green). DAPI(blue) was used for counter-staining of the nuclei. Cryosections (n ¼ 5) per each treatment were stained with rat anti-mouse CD31 and anti-ratIgG-Alexa 488 (green fluorescence). Representative immunofluorescent staining of EMT6-HER2 tumor cryosections from mice treated with PBS (A) andaHER2-huEndo-P125A (B) are presented. Blood vessel area (pixel2, n ¼ 5) in EMT6-HER2 tumors was quantified using NIH ImageJ by color image to form abinary image to allowmeasurement of blood vessel density, which is presented as total density of vessels (C). Blood vessel area was then divided by number ofblood vessels on the field, which is presented as average vessel density (D). The data are presented as themean of 5 low power field determinations� 95%CI.






In vivo treatment of established SK-BR-3 xenografts withthe aHER2-huEndo-P125A fusion resulted in greaterinhibition of growth than aHER2 IgG3, trastuzumab,human endostatin, human mutant endostatin (huEndo-P125A), or aHER2-huEndo fusion protein treatment. TheaHER2-huEndo fusion protein specifically targetedtumors expressing HER2 in syngeneic mice simulta-neously implanted with EMT6 and EMT6-HER2. Immu-nofluorescent staining of EMT6-HER2 tumors treatedwiththe aHER2-huEndo-P125A fusion showed thin, short, andfragmented blood vessels. The data presented earlierstrongly suggest that the enhanced antitumor effects ofthe fusion proteinmay be due to the targeting of antiangio-genic activity by the anti-HER2 antibody domain.

The mutant aHER2-huEndo-P125A fusion inhibitedtube formation of HUVEC in vitro, and tumor growth invivomore effectively thanaHER2-huEndo. Four syntheticpeptides corresponding to the sequences 6–49 (I), 50–92(II), 93–133 (III), and 134–178 (IV) of human endostatinhave been examined for their ability to inhibit endothelialcell proliferation, migration, and both in vitro and in vivoangiogenesis (31, 32). Fragment I was found to be anti-angiogenic whereas unexpectedly, fragment III exhibiteda proangiogenic activity, increased endothelial cell migra-tion andneovascularization, and enhanced the angiogenicresponse to vascular endothelial growth factor in a cornealpocket assay. The P125A point mutation may lead aconformational change that reduces the proangiogenicproperties of fragment III. Human endostatin has aninternal Asn-Gly-Arg (NGR) motif at position 126–128following the proline at position 125 (20). Mutation ofthe 125-proline may therefore affect vascular targeting bythe NGR motif. P125A endostatin bound to endothelialcellsmore efficiently thanwild-type endostatin andexhib-ited greater inhibition of both proliferation andmigrationof endotheliel cells (20). Vascular endothelial growthfactor and angiopoietin 1 were down-regulated more byP125A endostatin than by native endostatin (20). Theseresults suggested that an antibody fusion based on themutant P125A endostatin might inhibit tumor growth invivo more effectively than aHER2-huEndo as indeedproved to be the case in our experiments.

Antigenicity of the fusion proteins is a theoreticalproblem which may affect activity in vivo in either immu-nocompetent mice or potentially in man, and reduceefficacy. There are several examples of fusions inapproved clinical use in which antigenicity has not beena major problem including anti-TNFR-fusions (33, 34),DAB389IL-2 (35), and VEGFR-fusion (VEGF-Trap inpatients with solid tumors and non-Hodgkin’s lympho-

mas; ref. 36). Therefore, the effects of antigenicity onefficacy cannot be predicted a priori from mouse models.

Linking endostatin to an antibody may significantlyenhance the antitumor activity of trastuzumab (24).Because the overall response rates of HER2þ breastcancers to trastuzumab remain relatively low (15%–34%; refs. 36–39), this approach holds promise forincreasing both response rate and duration relative totrastuzumab, andmay expand the spectrum of antitumoractivity of trastuzumab given alone or in combinationwith other antitumor strategies such as other cytotoxicagents (carboplatin, docetaxel; refs. 40–43), and/or anti-angiogenic drugs (e.g., bevacizumab; anti-VEGF anti-body, thrombospondin-1; refs. 44–47). Experiments areongoing in our laboratory to determine whether anti-HER2–endostatin fusions may also be active againsttrastuzumab resistant cell lines in vivo.

Because endostatin is known tobeapowerful andglobalregulator of angiogenic gene expression, we concentratedinitial experiments on endostatin as a candidate fusion. Inaddition to endostatin, other antiangiogenic domainscould also theoretically be incorporated into fusions (e.g., angiostatin, tumstatin, etc). Finally, targeting antian-giogenic proteins using antibody is a versatile approachthat could be applied to other tumor targets throughsubstitution with other antibody specificities/variabledomains. This approach could in theory therefore beapplied to enhance efficacy of antibodies directed to othertumor antigens in settingswhere parental antibody showsmodest efficacy (e.g., cetuximab; refs. 48–50).

Disclosure of Potential Conflicts of Interest

S-U. Shin and J.D. Rosenblatt have filed a US patent application.

Acknowledgments

We thank Morgan and Friends (Dr. Herb Krickstein and MorganPressel) and the Kassamali Family and Luminaire for the generous giftsand efforts.

Grant Support

The U.S. Army Medical Research and Material Command under W81XWH-05-1-0351 Grant (DOD Idea Award, BC044744: S.U. Shin), Bankhead Coley Pre-Spore from the State of Florida (S.U. Shin, J.D. Rosenblatt), the National CancerInstitute CA114340 (S. Ramakrishnan).

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 August 25, 2010; revised November 18, 2010; acceptedFebruary 8, 2011; published OnlineFirst March 10, 2011.

References1. O'Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS, et al.

Endostatin: an endogenous inhibitor of angiogenesis and tumorgrowth. Cell 1997;88:277–85.

2. Boehm T, Folkman J, Browder T, O’Reilly MS. Antiangiogenic therapyof experimental cancer does not induce acquired drug resistance.Nature 1997;390:404–7.

3. Dhanabal M, Ramchandran R, Volk R, Stillman IE, Lombardo M,Iruela-Arispe ML, et al. Endostatin: yeast production, mutants, andantitumor effect in renal cell carcinoma. Cancer Res 1999;59:189–97.

4. Standker L, Schrader M, Kanse SM, Jurgens M, Forssmann WG,Preissner KT. Isolation and characterization of the circulating form ofhuman endostatin. FEBS Lett 1997;420:129–33.

Shin et al.





5. Joki T, Machluf M, Atala A, Zhu J, Seyfried NT, Dunn IF, et al.Continuous release of endostatin frommicroencapsulated engineeredcells for tumor therapy. Nat Biotechnol 2001;19:35–9.

6. Read TA, Farhadi M, Bjerkvig R, Olsen BR, Rokstad AM, Huszthy PC,et al. Intravital microscopy reveals novel antivascular and antitumoreffects of endostatin delivered locally by alginate-encapsulated cells.Cancer Res 2001;61:6830–7.

7. Sorensen DR, Read TA, Porwol T, Olsen BR, Timpl R, Sasaki T, et al.Endostatin reduces vascularization, blood flow, and growth in a ratgliosarcoma. Neuro-oncol 2002;4:1–8.

8. Kisker O, Becker CM, Prox D, Fannon M, D'Amato R, Flynn E, et al.Continuous administration of endostatin by intraperitoneallyimplanted osmotic pump improves the efficacy and potency of ther-apy in amouse xenograft tumormodel. Cancer Res 2001;61:7669–74.

9. Eder JP Jr, Supko JG, Clark JW, Puchalski TA, Garcia-Carbonero R,Ryan DP, et al. Phase I clinical trial of recombinant human endostatinadministered as a short intravenous infusion repeated daily. J ClinOncol 2002;20:3772–84.

10. Herbst RS, Hess KR, Tran HT, Tseng JE, Mullani NA, CharnsangavejC, et al. Phase I study of recombinant human endostatin in patientswith advanced solid tumors. J Clin Oncol 2002;20:3792–803.

11. Thomas JP, Arzoomanian RZ, Alberti D, Marnocha R, Lee F, Friedl A,et al. Phase I pharmacokinetic and pharmacodynamic study ofrecombinant human endostatin in patients with advanced solidtumors. J Clin Oncol 2003;21:223–31.

12. Hansma AH, Broxterman HJ, Van Der Horst I, Yuana Y, Boven E,Giaccone G, et al. Recombinant human endostatin administered as a28-day continuous intravenous infusion, followed by daily subcuta-neous injections: a phase I and pharmacokinetic study in patients withadvanced cancer. Ann Oncol 2005;16:1695–701.

13. Kulke MH, Bergsland EK, Ryan DP, Enzinger PC, Lynch TJ, Zhu AX,et al. Phase II study of recombinant human endostatin in patients withadvanced neuroendocrine tumors. J Clin Oncol 2006;24:3555–61.

14. Sun L, Ye HY, Zhang YH, Guan YS, Wu H. Epidermal growth factorreceptor antibody plus recombinant human endostatin in treatment ofhepatic metastases after remnant gastric cancer resection. World JGastroenterol 2007;13:6115–8.

15. Ling Y, Yang Y, Lu N, You QD, Wang S, Gao Y, et al. Endostar, a novelrecombinant human endostatin, exerts antiangiogenic effect viablocking VEGF-induced tyrosine phosphorylation of KDR/Flk-1 ofendothelial cells. Biochem Biophys Res Commun 2007;361:79–84.

16. Song HF, Liu XW, Zhang HN, Zhu BZ, Yuan SJ, Liu SY, et al.Pharmacokinetics of His-tag recombinant human endostatin in Rhe-sus monkeys.Acta Pharmacol Sin 2005;26:124–8.

17. Javaherian K, Park SY, Pickl WF, LaMontagne KR, Sjin RT, Gillies S,et al. Laminin modulates morphogenic properties of the collagen XVIIIendostatin domain. J Biol Chem 2002;277:45211–8.

18. Kuo CJ, LaMontagne KR Jr, Garcia-Cardeña G, Ackley BD, Kalman D,Park S, et al. Oligomerization-dependent regulation of motility andmorphogenesis by the collagen XVIII NC1/endostatin domain. J CellBiol 2001;152:1233–46.

19. Calvo A, Yokoyama Y, Smith LE, Ali I, Shih SC, Feldman AL, et al.Inhibition of themammary carcinoma angiogenic switch in C3(1)/SV40transgenic mice by a mutated form of human endostatin. Int J Cancer2002;101:224–34.

20. Yokoyama Y, Ramakrishnan S. Improved biological activity of amutant endostatin containing a single amino-acid substitution. Br JCancer 2004;90:1627–35.

21. Yokoyama Y, Ramakrishnan S. Addition of integrin binding sequenceto a mutant human endostatin improves inhibition of tumor growth. IntJ Cancer 2004;111:839–48.

22. Celik I, S€ur€uc€u O, Dietz C, Heymach JV, Force J, H€oschele I, et al.Therapeutic efficacy of endostatin exhibits a biphasic dose-responsecurve. Cancer Res 2005;65:11044–50.

23. Challita-Eid PM, Penichet ML, Shin SU, Poles T, Mosammaparast N,Mahmood K, et al. A B7.1-antibody fusion protein retains antibodyspecificity and ability to activate via the T cell costimulatory pathway. JImmunol 1998;160:3419–26.

24. Cho HM, Rosenblatt JD, Kang YS, Iruela-Arispe ML, Morrison SL,Penichet ML, et al. Enhanced inhibition of murine tumor and human

breast tumor xenografts using targeted delivery of an antibody-endo-statin fusion protein. Mol Cancer Ther 2005;4:956–67.

25. Shin SU, Wu D, Ramanathan R, Pardridge WM, Morrison SL. Func-tional and pharmacokinetic properties of antibody-avidin fusion pro-teins. J Immunol 1997;158:4797–804.

26. Coloma MJ, Hastings A, Wims LA, Morrison SL. Novel vectors for theexpression of antibody molecules using variable regions generated bypolymerase chain reaction. J Immunol Methods 1992;152:89–104.

27. Shin SU, Morrison SL. Production and properties of chimeric antibodymolecules. Methods Enzymol 1989;178:459–76.

28. Merchan JR, Chan B, Kale S, Schnipper LE, Sukhatme VP. In vitro andin vivo induction of antiangiogenic activity by plasminogen activatorsand captopril. J Natl Cancer Inst 2003;95:388–99.

29. Merchan JR, Jayaram DR, Supko JG, He X, Bubley GJ, Sukhatme VP.Increased endothelial uptake of paclitaxel as a potential mechanismfor its antiangiogenic effects: potentiation by Cox-2 inhibition. Int JCancer 2005;113:490–8.

30. PegramMD, Konecny GE, O’Callaghan C, Beryt M, Pietras R, SlamonDJ. Rational combinations of trastuzumab with chemotherapeuticdrugs used in the treatment of breast cancer. J Natl Cancer Inst2004;96:739–49.

31. CattaneoMG, Pola S, Francescato P, Chillemi F, Vicentini LM. Humanendostatin-derived synthetic peptides possess potent antiangiogenicproperties in vitro and in vivo. Exp Cell Res 2003;283:230–6.

32. Chillemi F, Francescato P, Ragg E, Cattaneo MG, Pola S, Vicentini L.Studies on the structure-activity relationship of endostatin: synthesisof human endostatin peptides exhibiting potent antiangiogenic activ-ities. J Med Chem 2003;46:4165–72.

33. Leonardi CL, Powers JL, Matheson RT, Goffe BS, Zitnik R, Wang A,et al. Etanercept as monotherapy in patients with psoriasis. N Engl JMed 2003;349:2014–22.

34. Weinberg JM. An overview of infliximab, etanercept, efalizumab, andalefacept as biologic therapy for psoriasis. Clin Ther 2003;25:2487–505.

35. Foss FM, Bacha P, Osann KE, Demierre MF, Bell T, Kuzel T. Biologicalcorrelates of acute hypersensitivity events with DAB(389)IL-2 (deni-leukin diftitox, ONTAK) in cutaneous T-cell lymphoma: decreasedfrequency and severity with steroid premedication. Clin Lymphoma2001;1:298–302.

36. Fricke I, Mirza N, Dupont J, Lockhart C, Jackson A, Lee JH, et al.Vascular endothelial growth factor-trap overcomes defects in den-dritic cell differentiation but does not improve antigen-specificimmune responses. Clin Cancer Res 2007;13:4840–8.

37. Baselga J, Tripathy D, Mendelsohn J, Baughman S, Benz CC, DantisL, et al. Phase II study of weekly intravenous recombinant humanizedanti-p185HER2monoclonal antibody in patients with HER2/neu-over-expressing metastatic breast cancer. J Clin Oncol 1996;14:737–44.

38. Vogel CL, Cobleigh MA, Tripathy D, Gutheil JC, Harris LN, Fehren-bacher L, et al. Efficacy and safety of trastuzumab as a single agent infirst-line treatment of HER2-overexpressing metastatic breast cancer.J Clin Oncol 2002;20:719–26.

39. Burstein HJ, Harris LN, Marcom PK, Lambert-Falls R, Havlin K,Overmoyer B, et al. Trastuzumab and vinorelbine as first-line therapyfor HER2-overexpressingmetastatic breast cancer: multicenter phaseII trial with clinical outcomes, analysis of serum tumor markers aspredictive factors, and cardiac surveillance algorithm. J Clin Oncol2003;21:2889–95.

40. Pegram MD, Lipton A, Hayes DF, Weber BL, Baselga JM, Tripathy D,et al. Phase II study of receptor-enhanced chemosensitivity usingrecombinant humanized anti-p185HER2/neu monoclonal antibodyplus cisplatin in patients with HER2/neu-overexpressing metastaticbreast cancer refractory to chemotherapy treatment. J Clin Oncol1998;16:2659–71.

41. Baselga J, Norton L, Albanell J, Kim YM, Mendelsohn J. Recombinanthumanized anti-HER2 antibody (Herceptin) enhances the antitumoractivity of paclitaxel and doxorubicin against HER2/neu overexpres-sing human breast cancer xenografts. Cancer Res 1998;58:2825–31.

42. Burstein HJ, Harris LN, Gelman R, Lester SC, Nunes RA, Kaelin CM,et al. Preoperative therapy with trastuzumab and paclitaxel followedby sequential adjuvant doxorubicin/cyclophosphamide for HER2






overexpressing stage II or III breast cancer: a pilot study. J Clin Oncol2003;21:46–53.

43. Pegram MD, Pienkowski T, Northfelt DW, Eiermann W, Patel R,Fumoleau P, et al. Results of two open-label, multicenter phase IIstudies of docetaxel, platinum salts, and trastuzumab in HER2-posi-tive advanced breast cancer. J Natl Cancer Inst 2004;96:759–69.

44. Willett CG, Boucher Y, di Tomaso E, Duda DG, Munn LL, Tong RT,et al. Direct evidence that the VEGF-specific antibody bevacizumabhas antivascular effects in human rectal cancer. Nat Med 2004;10:145–7.

45. Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, Hainsworth J,Heim W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucov-orin for metastatic colorectal cancer. N Engl J Med 2004;350:2335–42.

46. Kabbinavar F, Hurwitz HI, Fehrenbacher L, Meropol NJ, Novotny WF,Lieberman G, et al. Phase II, randomized trial comparing bevacizumabplus fluorouracil (FU)/leucovorin (LV) with FU/LV alone in patients withmetastatic colorectal cancer. J Clin Oncol 2003;21:60–5.

47. Rini BI, Halabi S, Taylor J, Small EJ, Schilsky RL, Schilsky RL, et al.Cancer and Leukemia Group B 90206: a randomized phase III trial ofinterferon-alpha or interferon-alpha plus anti-vascular endothelialgrowth factor antibody (bevacizumab) in metastatic renal cell carci-noma. Clin Cancer Res 2004;10:2584–6.

48. Xiong HQ, Rosenberg A, LoBuglio A, Schmidt W,Wolff RA, Deutsch J,et al. Cetuximab, a monoclonal antibody targeting the epidermalgrowth factor receptor, in combination with gemcitabine for advancedpancreatic cancer: a multicenter phase II Trial. J Clin Oncol 2004;22:2610–6.

49. Saltz LB, Meropol NJ, Loehrer PJ Sr, Needle MN, Kopit J, Mayer RJ.Phase II trial of cetuximab in patients with refractory colorectal cancerthat expresses the epidermal growth factor receptor. J Clin Oncol2004;22:1201–8.

50. Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, Santoro A,et al. Cetuximab monotherapy and cetuximab plus irinotecan inirinotecan-refractory metastatic colorectal cancer. N Engl J Med2004;351:337–45.

Shin et al.





2011;10:603-614. Published OnlineFirst March 10, 2011.Mol Cancer Ther Seung-Uon Shin, Hyun-Mi Cho, Jaime Merchan, et al. Fusion Protein Results in Enhanced Antitumor Efficacy

Mutant Human Endostatin−Targeted Delivery of an Antibody

Updated version

10.1158/1535-7163.MCT-10-0804doi:

Access the most recent version of this article at:

Cited articles

http://mct.aacrjournals.org/content/10/4/603.full#ref-list-1

This article cites 50 articles, 24 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/10/4/603.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

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

SubscriptionsReprints 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/10/4/603To 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-10-0804

http://mct.aacrjournals.org/content/10/4/603.full#ref-list-1

http://mct.aacrjournals.org/content/10/4/603.full#related-urls

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

mailto:[email protected]

http://mct.aacrjournals.org/content/10/4/603


targeted delivery of an antibody mutant human endostatin … · therapeutic discovery targeted...

Documents