Inhibition of Endoglin GIPC Interaction Inhibits Pancreatic Cancer …€¦ · 22/09/2014 · Inhibition of Endoglin–GIPC Interaction Inhibits Pancreatic Cancer Cell Growth Krishnendu

Small Molecule Therapeutics

Inhibition of Endoglin–GIPC Interaction Inhibits PancreaticCancer Cell Growth

Krishnendu Pal1, Alexandre A. Pletnev2, Shamit K. Dutta1, Enfeng Wang1, Ruizhi Zhao2, Aradhita Baral3,Vinod Kumar Yadav4, Suruchi Aggarwal4, Soundararajan Krishnaswamy5, Khalid M. Alkharfy1,6,Shantanu Chowdhury3,4, Mark R. Spaller2, and Debabrata Mukhopadhyay1

AbstractEndoglin, a 180-kDa disulfide-linked homodimeric transmembrane receptor protein mostly expressed in

tumor-associated endothelial cells, is an endogenous binding partner of GAIP-interacting protein, C terminus

(GIPC). Endoglin functions as a coreceptor of TbRII that binds TGFb and is important for vascular develop-

ment, and consequently has become a compelling target for antiangiogenic therapies. A few recent studies in

gastrointestinal stromal tumor (GIST), breast cancer, and ovarian cancer, however, suggest that endoglin is

upregulated in tumor cells and is associated with poor prognosis. These findings indicate a broader role of

endoglin in tumor biology, beyond angiogenic effects. The goal of our current study is to evaluate the effects of

targeting endoglin in pancreatic cancer both in vitro and in vivo.Weanalyzed the antiproliferative effect of both

RNAi-based andpeptide ligand-based inhibition of endoglin in pancreatic cancer cell lines, the latter yielding a

GIPC PDZdomain-targeting lipopeptidewith notable antiproliferative activity.We further demonstrated that

endoglin inhibition induced a differentiation phenotype in the pancreatic cancer cells and sensitized them

against conventional chemotherapeutic drug gemcitabine. Most importantly, we have demonstrated the

antitumor effect of both RNAi-based and competitive inhibitor–based blocking of endoglin in pancreatic

cancer xenograft models in vivo. To our knowledge, this is the first report exploring the effect of targeting

endoglin in pancreatic cancer cells. Mol Cancer Ther; 13(10); 1–12. �2014 AACR.

IntroductionPancreatic cancer is among the most lethal and unfor-

giving of human cancers, and continues to be a majorunsolved health problem (1). It has a 5-year survival rateof only 6% and is estimated to have 46,420 new cases andcause 39,590 deaths in 2014 in the United States, a numberthat has been steadily increasing for more than a decade(2, 3). Conventional treatment approaches, including sur-gery, radiotherapy, chemotherapy, or combinations

thereof, have had little impact on the course of thisaggressive cancer. To address an unmet clinical need fornew agents to treat pancreatic cancer, there is a corre-sponding, urgent requirement to identify novel targetsthat could lead the way toward therapeutic strategies tocombat this disease.

Endoglin (or CD105) is a transmembrane glycoproteinand a component of the TGFb receptor system (4). It ismainly expressed in endothelial cells and promotes theirproliferation by modulating TGFb signaling through itsPDZ domain-binding association with GAIP-interactingprotein, C terminus (GIPC; refs. 5–7). Mice lacking endo-glin die at the mid-gestation stage as a result of defectivevascular development, suggesting that endoglin has rolesin endothelial proliferation and vascular development, i.e., angiogenesis and vasculogenesis (8). Several groupshave shown that endoglin is also expressed inperitumoraland intratumoral blood vessels of brain, prostate, breast,colorectal, pancreatic, and hepatic cancers (9–15). Thespecific expression pattern of endoglin in tumoral vesselsindicates that it may be a worthy target molecule forantiangiogenic therapy in cancer. Indeed, anti-endoglintherapy is currently being explored as an antiangiogenictherapy in several cancers (16–19).

However, endoglin may serve in a capacity beyondangiogenesis alone. Studies in GIST, breast cancer, andovarian cancer suggest that endoglin is upregulated not

1Department of Biochemistry and Molecular Biology, Mayo Clinic, Roche-ster, Minnesota. 2Department of Pharmacology and Toxicology, GeiselSchool of Medicine at Dartmouth and Norris Cotton Cancer Center,Lebanon, New Hampshire. 3Proteomics and Structural Biology Unit, Insti-tute of Genomics and Integrative Biology, Council for Scientific andIndustrial Research, New Delhi, India. 4G.N.R. Knowledge Center forGenome Informatics, Institute of Genomics and Integrative Biology, Coun-cil for Scientific and Industrial Research, New Delhi, India. 5Department ofBiochemistry, King Saud University, Riyadh, Saudi Arabia. 6Department ofPharmacy, King Saud University, Riyadh, Saudi Arabia.

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

Corresponding Author: Debabrata Mukhopadhyay, Department of Bio-chemistry and Molecular Biology, Gugg 1321C, Mayo Clinic, 200 FirstStreet SW, Rochester MN 55905. Phone: 507-538-3581; Fax: 507-293-1058; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-14-0291

�2014 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org OF1

Research. on January 24, 2021. © 2014 American Association for Cancermct.aacrjournals.org Downloaded from

Published OnlineFirst August 14, 2014; DOI: 10.1158/1535-7163.MCT-14-0291

http://mct.aacrjournals.org/

only in tumor endothelial cells, but also in tumor cells, andis associated with a poor prognosis (20–22). In addition, ithas been recently shown that while endoglin is rarelyexpressed in primary ovarian cancer cells, it is frequentlyexpressed in recurrent platinum-resistant tumor cells, ascompared with the primary untreated tumor (23). Thesefindings point to a broader role of endoglin in tumor cellbiology beyond that of endothelial expression alone.

A recent study reported a wide range of endoglinexpression levels in pancreatic cancer cell lines but nodetectable levels of endoglin in normal pancreatic epithe-lial cells (24). Endoglin expression could also be seen in asubset of pancreatic cancer cells obtained from patientswith pancreatic ductal adenocarcinoma (PDAC). Interest-ingly, downregulation of epithelial marker E-cadherinand overexpression of mesenchymal marker vimentinwere observed in endoglin-positivepancreatic cancer cellscompared with those in endoglin-negative cells. Thesefindings prompted us to evaluate the effects of endoglininhibition in pancreatic cancer progression.

GIPC was originally identified as a binding partnerof the GTPase-activating protein RGS-GAIP (25). ThePDZ domain of GIPC binds to several endogenouspartner proteins such as RGS-GAIP, APPL1, Glut-1 trans-porter, Semaphorin-F, neuropilin-1, IGF1R, and endoglinthrough a direct associationwith the C-terminal tail of thepartner proteins (7, 25–30) and this PDZ interaction hasbeen implicatedas a critical player in thebiologyofnormaland malignant cells (30–34). Therefore, it is quite under-standable that disruption of this GIPC PDZ-specific inter-action with selective inhibitors should inhibit protein–protein interactions within cellular signaling pathwaysthat are required for cancer growth. Previous studies fromour group had utilized peptide-based inhibitors for thispurpose (30, 34, 35).

In the current study, we have analyzed the antiproli-ferative effect of both RNAi-based and competitive, pep-tide-based inhibition of endoglin via the PDZ domain ofone of its endogenous binding partners, GIPC, in pancre-atic cancer cell lines.Wehave further shown that endoglininhibition induced a differentiation phenotype in thepancreatic cancer cells and sensitized them against theconventional chemotherapeutic drug gemcitabine. Mostimportantly, we have shown the antiproliferative effect ofboth RNAi-based and molecular competitive peptide-based inhibition of endoglin in pancreatic cancer xeno-graft models in vivo. To our knowledge, this is the firstreport exploring the effect of targeting endoglin in pan-creatic cancer cells.

Materials and MethodsReagents

The antibodies against phospho-Smad 1/5, Smad 1,phospho-Smad 2, and Smad 2/3 were purchased fromCell Signaling Technology. Antibodies against Id-1, Ki67,and CD31 were purchased from Santa Cruz Biotechnol-ogy. Anti-b-actin and anti-E-cadherin were from BD Bios-ciences. Anti-endoglin antibody was purchased from

Sigma-Aldrich. Anti-GIPC antibody was purchased fromPierce. Immunohistochemistry was performed using theIHC Select HRP/DAB Kit from Millipore.

Cell culturePancreatic cancer cell lines ASPC-1, MiaPaca-2, and

cells isolated from pancreatic cancer patient-derivedxenografts (MCPAN002, 5160-1, MCPAN014, 4866-1,5647-1, and 4482-1) were used in the present study.ASPC-1 (CRL-1682; purchased in January, 2006) andMiaPaca-2 (CRL-1420; purchased in November, 2005)cells were from ATCC. No authentication of the celllines was done by the authors. The procedure for iso-lation of cells from pancreatic cancer patient-derivedxenografts is described below. ASPC-1 cell line wasmaintained in RPMI1640 medium supplemented with10% FBS and 1% penicillin-streptomycin. MiaPaca-2 cellline was maintained in DMEM supplemented with 10%FBS and 1% penicillin-streptomycin. Cells isolated frompatients were cultured in DMEM-F12 medium supple-mented with 10% FBS and 1% antimycotic/antibiotic(Invitrogen). Cells from a culture of 70% to 80% con-fluence were used in experiments.

Isolation of cells from pancreatic cancer patientxenografts

The human primary pancreatic cell lines were devel-oped from successful patient-derived xenografts. To gen-erate our xenografts, we obtained human pancreatictumor tissue from the surgical pathology department ofthe Mayo Clinic (Rochester, MN). All tissues wereobtained and used in accordance with the Mayo ClinicInstitutional Review Board and the Institutional AnimalCare and Use Committee. The tumor tissue was main-tained under sterile conditions and was washed twicewith PBS. It was then minced with a razor blade andmixed with high growth factor Matrigel (BD Biosciences)to create a slurry. Approximately 0.1 mL slurry wasinjected subcutaneously into the right flank of SCIDmice.Tumors were grown to a spherical diameter of approxi-mately 1.7 cm, if permitted by the health of the animal.Animals were then euthanized and xenografts were har-vested. The harvested xenograft was washed twice withHank’s balanced salt solution (HBSS) supplemented withglucose (0.9 g/L) and 1% antibiotic-antimycotic. The tis-sue was then cut into 1 to 2 mm pieces and digested withCollagenase V (Sigma). After a 20-minute incubation in a37�C water bath, the tissue was washed three times withthe HBSS glucose solution. The tissue was further disso-ciated with TrypLE (Life Technologies) for 5 minutes atroom temperature and then washed with complete cul-ture medium. The sample was finally resuspended in theculture medium and plated on rat tail collagen-coated 60mm plates (BD Biosciences) and incubated at 37�C with5% CO2/95% relative humidity. The plates were thentrypsinized over a series of weeks to remove nontumorcells. Tumor cells were replated on rat tail collagen-coatedplates for expansionwhen needed. Cells weremaintained

Pal et al.

Mol Cancer Ther; 13(10) October 2014 Molecular Cancer TherapeuticsOF2




in DMEM-F12 medium (Invitrogen) supplemented with10% FBS and 1% antimycotic/antibiotic (Invitrogen).

Quantitative RT-PCRRNA was isolated from pancreatic cancer cell lines

using RNAEasy Plus Mini Kit (Qiagen), and was used toprepare cDNA using iScript cDNA Synthesis Kit (Bio-Rad); both procedures followed the manufacturers’ pro-tocols. They were subjected to a quantitative PCR toevaluate the expression levels of Endoglin using RT2

qPCR Primer Assay (SABioscience). The values werenormalized against b-actin.

Western blot analysis and immunoprecipitationWhole-cell lysates were prepared in NP-40 lysis buffer

supplemented with protease inhibitor cocktail (Sigma)and Halt phosphatase inhibitor (Thermo Scientific).Supernatantwas collected by centrifugation at 12,000 rpmfor 10 minutes at 4�C. Thereafter, samples were subjectedto SDS-PAGE and then transferred to polyvinylidenedifluoride membranes. Themembranes were then immu-noblotted and antibody-reactive bands detected by usingthe Supersignal West Pico substrate (Thermo Scientific).Immunoprecipitation experiments were performed aspreviously described (30).

siRNA transfectionTransient knockdown of endoglin was performed in

ASPC-1 and MiaPaca-2 cell lines with two different anti-endoglin siRNAs (Qiagen, Target sequence si 1: 50-CGCCATGACCCTGGTACTAAA-30, Target sequence si2: 50-CAGCAATGAGGCGGTGGTCAA-30) using Dhar-mafect-4 (Dharmacon). A control siRNA (Qiagen) lackingknown human or mouse targets was used as a nontargetcontrol. RNAorproteinwas isolated from the cells 72 to 96hours after the siRNA transfection and endoglin down-regulation was determined by quantitative reverse tran-scription-PCR (RT-PCR) and Western blot analysis.

Lentiviral shRNA transductionStable knockdownof endoglinwasperformed inASPC-

1 cells using lentivirus shRNA transductionmethod. Twodifferent Endoglin shRNAs (Sigma TRC1.5, Targetsequence for sh1: GTCTTGCAGAAACAGTCCATT, Tar-get sequence for sh2: CCACTTCTACACAGTACCCAT)were used. The lentivirus particles were prepared using293T cells following standardmethods.ASPC-1 cellsweretransduced with the prepared lentivirus particles andstable colonies were isolated after selection with puromy-cin (1 mg/mL).

Design and synthesis of peptidesPeptides were designed on the basis of the native eight-

residue C-terminal sequence of endoglin (PCSTSSMA),with varying degrees of N-terminal lipidation to enhancecellular membrane permeability (Table 1). In addition,two positions were independently mutated to Lys resi-dues, in which the side chain was benzoylated [AP1035,

myristoyl-PCSTSS[K(eN-benzoyl)]A and AP1036, myris-toyl-PCST[K(eN-benzoyl)]SMA]. The details of the syn-thesis procedures are given in Supplementary Materialsand Methods.

Surface plasmon resonanceThe affinity of N-myristoyl-PCSTSSMA (AP1013) for

recombinant GIPC PDZ was measured by surface plas-mon resonance, using a Biacore X100 (GE Healthcare LifeSciences). After pH scoutingwas performed, determiningan optimal pH of 5.5, GIPC PDZ was immobilized on aCM5 sensor chip via amine coupling reaction usingN-hydroxysuccinimide (NHS) and N-ethyl-N0-(dimethy-laminopropyl) carbodiimide (EDC). Solutions of AP1013were prepared using HBS-EP buffer, in dilutions rangingfrom10�Kd to 0.1�Kd. SPRexperimentswere conductedusing the kinetic/affinity measurement workflow of theinstrument. Collected data from the sensorgrams wereanalyzed using the Biacore evaluation software includedwith the instrument.

Thymidine incorporation assayApproximately, 2 � 104 cells were seeded in 24-well

plates and cultured for indicated time in presence ofsiRNA or peptides. Then, 1 mCi of [3H] thymidine wasadded to each well and incubated for 4 hours. The cellswere then washed with ice-cold PBS, fixed with chilledmethanol, and lysed with 0.1 N NaOH. The lysates werecollected for measurement of trichloroacetic acid-precip-itable radioactivity.

Drug sensitivity assayApproximately, 5 � 103 cells were seeded in 96-well

plates and treatedwith Endoglin siRNAor peptides for 24hours followed by gemcitabine for 48 hours. At the end ofthe treatment period, cell viability was measured usingCellTiter 96 Aqueous One Solution Cell ProliferationAssay (Promega) as per themanufacturer’s protocol. Halfmaximal inhibitory concentration (IC50) values were cal-culated as concentrations corresponding to a 50% reduc-tion of cell viability.

In vivo tumor progression analysisSix- to 8-week-old male SCID mice were obtained from

NIH (Bethesda, MD) and housed in the institutionalanimal facilities. All animal work was performed underprotocols approved by Mayo Clinic Institutional AnimalCare and Use Committee. Either control or endoglinshRNA–treatedASPC-1 cells (1� 106) suspended in 50 mLPBS were injected orthotopically into the pancreas of 6- to8-week-old male SCID mice (5 mice in each group).Tumors were allowed to grow for 3 weeks. After 3 weeks,mice were sacrificed and tumor growth was analyzed. Inanother set of experiments, 5 � 106 ASPC-1 cells sus-pended in 100 mL PBS were injected subcutaneously intothe right flanks of 6- to 8-week-oldmale SCIDmice (7micein each group). After 9 days, mice were randomized andeither AP1063 or AP1032 dissolved in PBS containing 80%

Targeting Endoglin in Pancreatic Cancer

www.aacrjournals.org Mol Cancer Ther; 13(10) October 2014 OF3




DMSOwere injected intratumorally everyday for 3weeks(500 mg/mouse/d). After 3 weeks of treatment, miceweresacrificed and tumor growth was analyzed. Tumorvolumes were calculated using the formula: V ¼ 0.5 � a� b2, where "a" is the longest tumor axis and "b" is theshortest tumor axis.

Histologic studyTumors were removed and fixed in 10% neutral buff-

ered formalin at room temperature for 24 hours beforeembedding in paraffin and sectioning. Sections weredeparaffinized and then subjected to different immuno-histochemical staining according to manufacturer’sinstructions (DAB 150, Millipore). Stable diaminobenzi-dine was used as a chromogen substrate and the sectionswere counterstained with a hematoxylin solution. Imageswere acquired using Zeiss Axioplan 2 Microscope.

Statistical analysisThe independent samples t test was used to test the

probability of significant differences between groups.Statistical significance was defined as P < 0.05 (�) andstatistical high significance was defined as P < 0.01 (��).Error bars are given on the basis of calculated SD values.

ResultsEndoglin downregulation inhibits cell proliferation

Endoglin expression could be seen in both the pancre-atic cancer cell lines tested (e.g., ASPC-1, MiaPaca-2). Itwas also expressed in several cell lines isolated frompancreatic cancer patient-derived xenografts such as5160-1, MCPAN014, 5647-1, and 4482-1 (Fig. 1A and B).However, the expression levels were varied among thecell lines. To check whether the expression of endoglin isimportant for pancreatic cancer growth, downregulationof endoglin was performed in two different cell lines withdifferent amount of endoglin expression (ASPC-1 withlower expression andMiaPaca-2 with higher expression).Two different siRNAs (ENG si 1 and ENG si 2) andshRNAs (ENG sh1 and ENG sh2) effectively reduced the

endoglin expression at the mRNA and protein levels(Fig. 1C–E). Both siRNAs inhibited cell proliferation inASPC-1 and MiaPaca-2 cell lines (Fig. 1F). Similarly, bothshRNAs showed significant inhibition of proliferation inASPC-1 (Fig. 1G). Overall, these observations suggest thatendoglin plays a significant role in proliferation.

Endoglin downregulation inhibits tumor growthWhen endoglin-downregulated ASPC-1 cells were

injected orthotopically into the pancreas of 6- to 8-week-old SCID mice (5 mice in each group), and theresulting tumors were allowed to grow for 3 weeks, theywere significantly smaller compared with the tumorsarising from control shRNA-treated cells (Fig. 2A andB). The tumor volumes were 416.94 � 125.24 mm3 incontrol shRNA group versus 232.97 � 102.4 mm3 and215.34 � 63.4 mm3 in ENG sh1 and ENG sh2 groups,respectively. The tumorweightswere 436.72�76.21mg incontrol shRNAgroupversus 232.97� 102.4mgand215.34� 63.4 mg in ENG sh1 and ENG sh2 groups, respectively.The proliferation of tumor cells was significantly lower intumors from ENG sh1 and ENG sh2 groups comparedwith control shRNA group, as evident from Ki67 stainingof the tumor tissue sections (Fig. 2C). The abundance ofKi67-positive nuclei was 50.86 � 4.58% in control shRNAgroup versus 25.3 � 4.39% and 21.24 � 3.7% in ENGsh1 and ENG sh2 groups, respectively (Fig. 2D). This isin accord with our hypothesis that endoglin plays animportant role in cancer progression, which is affectedby endoglin downregulation, thus causing a slower tumorgrowth. However, no antiangiogenic effects such asreducedmicrovessel density (MVD) or increased necrosiswere observed in tumor tissue sections obtained fromendoglin-downregulated ASPC-1 cells (SupplementaryFig. S1). This was not surprising, considering these ortho-topic tumors were mostly nonvascular. This also con-firmed that the observed effect was due to the inhibitionof tumor cell–specific endoglin rather than an antiangio-genic effect.

Endoglin–GIPC interaction is necessary forproliferation

Endoglin has been shown to act in endothelialcells through its association with GIPC through a PDZdomain-bindingmotif (7).However, little has beenknownabout the importance of endoglin–GIPC association incancer cells. To this end, we designed an octapeptide(PCSTSSMA) mimicking the C-terminus of endoglin, aPDZ domain-binding motif, with the intent of selectivelyblocking endoglin–GIPC association in a competitivemanner. A biochemical binding assay, using SPR underequilibrium analysis mode, demonstrated that N-myris-toyl-PCSTSSMA (AP1013) binds to recombinant GIPCPDZ with a binding constant of Kd ¼ 7.1 � 2.8 mmol/L.

Derivatives in which the amino terminus of the peptidewas acylated with fatty acids other than myristic (i.e.,lauric, palmitic, and stearic), as well as simple acetic acid,were prepared to serve as a formof control for the effect of

Table 1. List of peptides designed on the basisof the native eight-residue C-terminal sequenceof endoglin (PCSTSSMA), with varying degreesof N-terminal lipidation

Compoundcode Lipidated peptide N-acyl C length

AP1030 Acetyl-PCSTSSMA 2AP1031 Lauroyl-PCSTSSMA 12AP1013 Myristoyl-PCSTSSMA 14AP1032 Palmitoyl-PCSTSSMA 16AP1033 Stearoyl-PCSTSSMA 18AP1063 Palmitoyl-SMTSCAPS 16AP1035 Myristoyl-PCSTSS[K(Bn)]A 14AP1036 Myristoyl-PCST[K(Bn)]SMA 14

Pal et al.





lipidation (Table 1). Two additionalN-myristoylated pep-tides were synthesized in which one of two native resi-dues was sequentially replaced with Lys, in which theside chain had beenmodified with benzoyl moieties. Wescreened all these lipidated peptides in ASPC-1 cellsand monitored their antiproliferative activities, with the

intent of selecting the analog with maximal activity.Only myristoyl, palmitoyl, and stearoyl derivativesshowed significant antiproliferative activity, witha noticeable dose–response relationship (Fig. 3A). TheN-acetyl analog (AP1030), lacking a sufficiently longaliphatic tail that might enhance cell membrane

0.09

0.06

0.03

0

6

4

2

0

Con Si

ENG Si 1

ENG Si 2

Rel

ativ

e en

do

glin

exp

ress

ion

Rel

ativ

e en

do

glin

exp

ress

ion

Rel

ativ

e en

do

glin

exp

ress

ion

ASPC-1 MiaPaca-2ASPC-1

ASPC-1

Mia Pac

a-2

Mia

Pac

a-2

MCP AN00

2

MCP A

N002

MCP AN01

4

MCP A

N014

5160

-1

5160

-1

Con S

i

Con S

i

Con s

h

ENG Si 1

ENG Si 1

ENG Si 2

ENG Si 2

ENG sh

1ENG s

h 2

4866

-1

4866

-1

5647

-1

5647

-1

4482

-1

4482

-1Endoglin

b-ActinASPC-1

ASPC-1

ASPC-1 ASPC-1

ASPC-1

Con sh

ENG sh1

ENG sh2

1.2

0.8

0.4

0

MiaPaca-2

MiaPaca-2

Endoglin

b-Actin

Con SiCon sh

ENG Si 1ENG sh1

ENG sh2ENG Si 2

120

80

40

0

120

80

40

0

Pro

lifer

atio

n (

%)

Pro

lifer

atio

n (

%)

A C

B

D

E

F G

Figure 1. Endoglin downregulation inhibits cell proliferation. A and B, endoglin mRNA and protein expression in pancreatic cancer cell lines as well as cellsisolated from patients. C–E, estimation of endoglin downregulation in ASPC-1 and MiaPaca-2 cell lines by endoglin siRNAs and shRNAs. F, inhibition of cellproliferation in ASPC-1 and MiaPaca-2 after siRNA-mediated endoglin downregulation. G, inhibition of cell proliferation in ASPC-1 after shRNA-mediatedendoglin downregulation. ��, P < 0.01 compared with respective controls.






penetration and permeability, was, unsurprisingly, bothineffective and lacked a dose–response effect.

We selected palmitoyl-PCSTSSMA (AP1032) for furtherstudies, as it exhibited the highest antiproliferative activ-ity. We also synthesized a scrambled peptide with thesame residues (SMTSCAPS), withN-terminal palmitoyla-tion, to use as a negative control (AP1063). TreatmentwithAP1032 effectively inhibited endoglin–GIPC interactionin both ASPC-1 and MiaPaca-2 cell lines, compared withcontrolAP1063 (Fig. 3B).AP1032 inhibitedproliferation inASPC-1 (Fig. 3C) and MiaPaca-2 (Fig. 3D) cell lines.Interestingly, the peptide AP1032 is more effective inASPC-1, which has lower endoglin expression, than inMiaPaca-2, with higher endoglin expression. To checkwhether this observation can be generalized, two sets ofcell lines isolated from patient-derived xenografts werechosen. 5647-1 and 4482-1 express comparatively loweramount of endoglin, whereas 5160-1 and MCPAN014express higher amount of endoglin (Fig. 1A). As wepresumed, the peptide AP1032 is more effective in5647-1 and 4482-1, thus corroborating our model of com-petitive inhibition (Fig. 3E).

Inhibition of endoglin–GIPC interaction inhibitstumor growth

Intratumoral injections of AP1032 in a subcutaneousASPC-1 tumor model developed in 6- to 8-week-old

SCID mice (7 mice in each group) showed significantinhibition of tumor growth compared with control pep-tide AP1063 (Fig. 4A and B). The tumor volumes were385.96 � 94.95 mm3 in the AP1063-treated group versus153.95 � 27.13 mm3 in the AP1032-treated group. Thetumor weights were 333.57 � 47.53 mg in the AP1063-treated group versus 166.7 � 29.49 mg in the AP1032-treated group. The proliferation of tumor cells wassignificantly lower in the AP1032-treated group as evi-dent fromKi67 staining of the tumor tissue sections (Fig.4C). The abundance of Ki67-positive nuclei was 44.44 �3.94% in the AP1063-treated group versus 11.22 � 2.25%in the AP1032-treated group (Fig. 4D). This suggeststhat endoglin–GIPC interaction plays an important rolein cancer progression. Interestingly, treatment with thepeptide AP1032 resulted in a significant decrease inMVD and increased necrotic area compared withAP1063 in the subcutaneous model suggesting that thepeptide AP1032 was also affecting the tumor vascula-ture (Supplementary Fig. S2).

Endoglin downregulation does not affect Smadphophorylation but inhibition of endoglin–GIPCinteraction does

As endoglin is a coreceptor of TGFb signaling pathway,we thought to seewhether the canonical Smad signaling isaffected due to endoglin downregulation. Surprisingly,

600

400

200

0

600

400

200

0

Con sh ENG sh1 ENG sh2

Con sh ENG sh1 ENG sh2

H&E Ki67

Con sh

ENG sh1

ENG sh2

60

40

20

0Con sh ENG sh1 ENG sh2K

i67-

po

siti

ve c

ells

(%

)

Tum

or

wei

gh

t (m

g)

Tum

or

volu

me

(mm

3 )

A

B

C

D

Figure 2. Endoglin downregulationinhibits tumor growth. A and B,1 � 106 of control or endoglinshRNA-treated ASPC-1 cells wereinjected orthotopically into thepancreas of 6- to 8-week-old maleSCID mice and tumors wereallowed to grow for 3 weeks.Tumors arising from endoglin-downregulated ASPC-1 cellswere significantly smaller.�, P < 0.05 compared withcontrol. C, H&E and Ki67immunohistochemical staining inthe tumor tissues. Scale, 200 mm.D, quantification of Ki67-stainednuclei in control and endoglinshRNA-treated tumor tissuesections, respectively. ��, P < 0.01compared with control.

Pal et al.





endoglin downregulation did not have significant effecton Smad phosphorylation (Fig. 5A). We then checkedwhether blocking endoglin–GIPC interaction acts in thesame manner. However, treatment with AP1032 clearlyinhibits Smad-1/5 phosphorylation, which is involved insignaling pathways responsible for cell proliferation (Fig.5B). Expressions of Smad1 and Smad 2were also inhibiteddue to the treatment with AP1032. From these results, wemay conclude that downregulation of endoglin andblock-ing endoglin–GIPC interaction acts differently in pancre-atic cancer cells.

Both endoglin downregulation and inhibition ofendoglin–GIPC interaction induce differentiationAs both endoglin downregulation and inhibition of

endoglin–GIPC interaction inhibited tumor growth, wethought to check whether they induce differentiation inthe pancreatic cancer cells. Endoglin downregulationincreased E-cadherin expression in ASPC-1 and inhibitedId-1 expression in both ASPC-1 and MiaPaca-2 (Fig. 6A).Treatment with AP1032 did not significantly increase E-cadherin but inhibited Id-1 expression in both ASPC-1and MiaPaca-2 (Fig. 6B). Therefore, both endoglin down-regulation and inhibition of endoglin–GIPC interaction

induce a differentiation phenotype to pancreatic cancercells.

Both endoglin downregulation and inhibition ofendoglin–GIPC interaction sensitize pancreaticcancer cells toward gemcitabine

As both endoglin downregulation and inhibition ofendoglin–GIPC interaction induced differentiation inpancreatic cancer cells, we explored whether they couldsensitize pancreatic cancer cells toward standard chemo-therapeutic drug gemcitabine. Our results show thatendoglin downregulation decreases the IC50 value ofgemcitabine from 44.92 to 18.14 nmol/L (ENG si 1) and24.54 nmol/L (ENG si 2) in ASPC-1 cells (Fig. 6C). Sim-ilarly, the IC50 values for MiaPaca-2 cell line are 52.08nmol/L (con si), 15.73 nmol/L (ENG si 1), and 19.99nmol/L (ENG si 2; Fig. 6D). Inhibition of endoglin–GIPCinteraction also decreases the IC50 value from 47.37 nmol/L (AP1063) to 9.59 nmol/L (AP1032) in ASPC-1 (Fig. 6E)and from 42.73 nmol/L (AP1063) to 15.83 nmol/L(AP1032) in MiaPaca-2 (Fig. 6F). So, in both ASPC-1 andMiaPaca-2 cell lines, either genetic inhibition of endoglinor inhibition of endoglin–GIPC interaction elicited a

150

100

50

0Pro

lifer

atio

n (

%)

Pro

lifer

atio

n (

%)

Pro

lifer

atio

n (

%)

Pro

lifer

atio

n (

%)

AP1013 AP1030 AP1031 AP1032

AP1063 AP1032

AP1033 AP1035 AP1036

UT 50 50 50 50 50 50 50100

200

100

200

100

200

100

200

100

200

100

200

100

200

DMSO

UT 50 100 50 100

DMSO

UT 50 50100

100

DMSO

120

80

40

0

IP: GIPC IP: GIPC

ASPC-1 MiaPaca-2

IgG

AP106

3AP1

032

AP106

3AP1

032

IgG

IB: Endoglin

IB: GIPC

DMSO

AP1063

AP1032

150

100

50

0

120

80

40

0

5160-1 MCPAN014 5647-1 4482-1

Higherendoglin endoglin

LowerAP1063 AP1032

A

B C

D E

Figure 3. Endoglin–GIPCinteraction is necessary forproliferation. A, screening ofendoglin PDZ domain-bindingsequence peptides N-acylatedwith acetic and fatty acids inASPC-1 cells. Palmitoyl-PCSTSSMA (AP1032) wasselected for further experiments.B, AP1032 effectively blockedendoglin–GIPC interactionbut palmitoyl-SMTSCAPS (controlpeptide, AP1063) couldnot, as seen from theimmunoprecipitation experiment.C and D, AP1032 showed greaterinhibition of proliferation inASPC-1 cells compared withMiaPaca-2 cells. AP1063 wasused as a control. E, AP1032 alsoinhibited proliferation in cellsisolated from patient-derivedxenografts. Remarkably, AP1032was more effective in cells with alower level of endoglin expression.






significantly greater drug response compared with con-trol, when treated with gemcitabine.

DiscussionThe primary canonical role of endoglin is as a TGFb

coreceptor (4). It is primarily expressed in endothelialcells (5, 6). It is also found in solid tumor vasculatureand is a reliable marker of angiogenesis (9, 11–15).However, to date, only a few studies address the expres-sion of endoglin on tumor cells and its potential role incancer progression (20–22). A recent study reported theexpression of endoglin in pancreatic cancer cells andcells isolated from patients with PDAC (24). Here, we

have shown that endoglin is expressed in several pan-creatic cancer cells and RNAi-based downregulation ofendoglin inhibits in vitro pancreatic cancer cell prolif-eration. In vivo tumorigenic property of pancreatic can-cer cell line ASPC-1 was also significantly reduced uponendoglin downregulation. In addition, endoglin down-regulation sensitized the cells toward conventional che-motherapeutic drugs possibly through the induction ofdifferentiation as indicated from increased E-cadherinexpression and reduced Id-1 expression. Surprisingly,endoglin downregulation had no significant effecton Smad phosphorylation which suggests that otherTGFb coreceptors are possibly compensating for the

Tum

or

volu

me

(mm

3 )

Tum

or

wei

gh

t (m

g)600

400

200

0

400

200

0AP1063 AP1032 AP1063 AP1032

H&E Ki67

AP1063

AP1063

AP1032

AP1032K

i67-

po

siti

ve c

ells

(%

)

60

40

20

0

A B

C D

Figure4. Inhibition of endoglin–GIPC interaction inhibits tumor growth. A andB, 5�106ASPC-1 cellswere injected subcutaneously into the right flanks of 6- to8-week-old male SCID mice and tumors were allowed to grow. After 9 days, mice were randomized into two groups (n ¼ 7) and intratumoral injections ofAP1032 or AP1063 (dissolved in PBS containing 80%DMSO) at a dosage of 500 mg/day/mousewere started. The daily treatment continued for 3weeks, afterwhich the mice were sacrificed, and the tumors were collected and analyzed. Tumors obtained from AP1032-treated group were significantly smallercompared with AP1063-treated group. C, H&E and Ki67 immunohistochemical staining in the tumor tissues. Scale, 200 mm. D, quantification of Ki67-stainednuclei in AP1063-treated and AP1032-treated tumor tissue sections, respectively. ��, P < 0.01 compared with AP1063-treated group.

ASPC-1 MiaPaca-2 ASPC-1 MiaPaca-2

Con

Si

Con

Si

AP10

63AP

1032

AP10

63AP

1032

ENG

Si 1

ENG

Si 1

ENG

Si 2

ENG

Si 2

phospho Smad1/5 phospho Smad1/5

phospho Smad2 phospho Smad2

Smad1Smad1

Smad2/3 Smad2/3

β-Actin β-Actin

A BFigure 5. Endoglin downregulationdoes not affect Smadphophorylation but inhibition ofEndoglin–GIPC interaction does.A, Smad phosphorylation was notsignificantly affected by endoglindownregulation. B, inhibition ofendoglin–GIPC interaction withtreatment of AP1032 showedinhibition of Smad1/5phosphorylation.

Pal et al.





downregulation of endoglin. Nonetheless, from theseresults it appears that targeting endoglin in pancreaticcancer may prove to be useful.In regards to therapeutic development in patients with

cancer, delivery of siRNA constructs has the potential tooffer long duration of target inhibition as well as reducedtoxicity compared with other approaches. However,development of a delivery modality for siRNA constructs

remains the rate-limiting step in translational research.This is the reason we chose a peptide-based competitiveinhibition model. We designed an octapeptide based onthe C-terminal PDZ domain-binding sequence of endo-glin (PCSTSSMA), to competitively inhibit the associationof endoglin with GIPC. Our SPR studies with AP1013demonstrated binding to recombinantly produced PDZdomain derived from the full-length GIPC gene, with a

Con S

iEN

G S

i 1

ENG

Si 1

ENG

Si 2

ENG

Si 2

AP106

3AP1

032

AP106

3

AP103

2

Con S

i

ASPC-1ASPC-1MiaPaca-2 MiaPaca-2

E-Cadherin E-Cadherin

Id-1 Id-1

b-Actin b-Actin

Con Si+GemENG Si 1+GemENG Si 2+Gem

Con Si+GemENG Si 1+GemENG Si 2+Gem

AP1063+GemAP1032+Gem

AP1063+GemAP1032+Gem

Via

bili

ty (

%)

Via

bili

ty (

%)

Via

bili

ty (

%)

Via

bili

ty (

%)

Gemcitabine concentration (nmol/L) Gemcitabine concentration (nmol/L)

Gemcitabine concentration (nmol/L) Gemcitabine concentration (nmol/L)

100

75

50

25

0

100

75

50

25

0

100

75

50

25

0

100

75

50

25

010-1 100 101 102 103 104 105 10-1 100 101 102 103 104 105

10-1 100 101 102 103 104 10510-1 100 101 102 103 104 105

A B

C D

E F

Figure 6. Both endoglin downregulation and inhibition of endoglin–GIPC interaction induce differentiation and sensitize pancreatic cancer cells towardgemcitabine. A, endoglin downregulation increased E-cadherin expression in ASPC-1 and inhibited Id-1 in both ASPC-1 and MiaPaca-2 cell lines.B, treatment with 100 mmol/L AP1032 inhibited Id-1 in both ASPC-1 and MiaPaca-2 cell lines. C and D, endoglin downregulation sensitized bothASPC-1 and MiaPaca-2 cell lines toward gemcitabine. E and F, treatment with 50 mmol/L AP1032 sensitized both ASPC-1 and MiaPaca-2 cell linestoward gemcitabine.






dissociation constant of Kd¼ 7.1� 2.8 mmol/L. This valueis in line with binding affinities of peptides for otherPDZ domains, which, biologically, are believed to beprincipally engaged in transient interactions notintended to bind their endogenous partner proteinswith high affinity. It is possible, however, throughjudicious selection of standard (36) or nonstandardresidues (37) or through chemical modification (38), toenhance the affinity of a peptide for a PDZ domain, andeven selectivity, plasma stability, and cellular perme-ability properties (39). These and other approacheshighlight the significant amount of activity that hasarisen around the design and development of molecularagents to target of PDZ domains (40).

To promote cell permeability, we adopted andexpanded upon two strategies we had employed pre-viously (35). For the first, we synthesized derivatives inwhich the amino terminus of the peptide was acylatedwith natural fatty acids of increasing carbon length(lauric, myristic, palmitic, and stearic), as well as simpleacetate, to serve as a form of control for the effect oflipidation (Table 1). Usually, the mode of cell membraneuptake and transmigration is influenced by both thelength of the fatty acid (i.e., its lipophilicity) and thepeptide sequence, that is, the overall length and com-bined contribution of polar and nonpolar residues.Thus, with different peptide sequences, it is prudentto explore lipid appendages of varying length, as it maynot be possible to predict in advance which fatty acidwould prove to be the best lipid conjugate. Addition ofmoieties at the N-terminal region of peptide ligands forPDZ domains is generally considered unobtrusive, asthe primary recognition and binding energetics for PDZdomain-mediated interactions often seems to reside in4 to 6 residues at the C-terminus.

The second influence from our earlier work (35)inspired us to chemically modify the side chain in eachof two different positions within the endoglin peptidesequence. Sequentially substituting Lys for the endog-enous Met and Ser residues at the "�1" and "�3"positions, respectively, we acylated the side chainamine with benzoate. This lysine-(eN-benzoyl) unitserves as a "probe moiety", to gauge the extent to whichthat binding position will not only tolerate such asubstitution, but also potentially enhance binding affin-ity and, consequently, biologic efficacy. In the absenceof structural or other binding data, the chances that asingle substitution would improve potency are limited,but the primary motivation would be, as just stated, totest for tolerance. If efficacy is either maintained, or atleast not irredeemably compromised, this would pro-vide a degree of justification for pursuing a larger scalechemical library approach, in which hundreds of dif-ferent organic acids, many of which would be aromaticacids with different functionality, would be placed atone or both position, and each tested for affinity forGIPC PDZ. We have employed this strategy for adifferent PDZ domain, and were able to generate chem-

ically modified peptide ligands with enhanced bindingaffinity (38). Although the modified analogs AP1035and AP1036 were not superior to the correspondingnonmodified AP1013, the parent from which theyderived, they did exhibit a degree of efficacy and doseresponse that could justify a chemical library approach,as described above. This could prove to be the desiredapproach for devising next-generation compoundswith improved potency over that of the native endoglinsequence.

Such peptide-based inhibitors will be important fortargeting the PDZ domain-containing protein GIPC, acentral player in various signaling pathways. Previous-ly, it has been shown that endoglin modulates TGFbsignaling through its association with GIPC in endothe-lial cells (7). To the best of our knowledge, however,there have been no reports exploring the role of endo-glin–GIPC association in cancer cells. In this study, wehave shown that blocking the interaction of endoglinand GIPC, by means of targeting the PDZ domain of thelatter with a lipidated peptide inhibitor, reduced pan-creatic cancer cell proliferation. The in vivo tumor pro-gression was also significantly reduced by this peptide.In addition, this peptide sensitized the cells towardgemcitabine possibly through the induction of differ-entiation as indicated from reduced Id-1 expression.Interestingly, Smad 1/5 phopsphorylation was inhib-ited here, which could not be observed in RNAi-basedinhibition. We presume the reason behind this differ-ence was due to the similarity between the PDZ-domainbinding sequences of endoglin and other TGFb core-ceptors; allowing them to bind to GIPC possiblythrough the same PDZ domain. As a result, blockingthat domain by treatment with AP1032 may block otherinteractions, thus inhibiting the downstream signalingpathway. However, downregulation of endoglin shouldhave no effect on the binding of other coreceptors withGIPC. So, in the absence of endoglin, other coreceptorsmay still bind to GIPC and activate the downstreamsignaling pathway. Taken together, our data suggestthat PDZ domain-targeting peptide-based inhibitorsmay have a better therapeutic potential than that ofRNAi-based inhibition.

Currently, anti-endoglin therapy is being explored as atherapeutic in several cancers as an antiangiogenic agent(16–19). The present study shows that endoglin-targetedtherapies in pancreatic cancer may offer an additionaladvantage by targeting not only the vasculature but alsothe tumor cells expressing endoglin. Their effects ondifferentiation will also contribute to drug sensitivity, soa combination therapy may prove more efficacious inchemotherapy-resistant tumors. Overall, our data strong-ly suggest endoglin-targeted therapy as a double-spearedapproach to improve drug sensitivity through inductionof differentiation in tumor cells in addition to antiangio-genic therapy. This approach should be actively exploredas a potential therapy for the treatment of pancreaticcancer.

Pal et al.





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

Authors' ContributionsConception and design: K. Pal, M.R. Spaller, D. MukhopadhyayDevelopmentofmethodology:K.Pal, S.Krishnaswamy,D.MukhopadhyayAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): K. Pal, S.K. Dutta, E. Wang, R. Zhao, A. Baral,V.K. Yadav, S. Aggarwal, S. ChowdhuryAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): K. Pal, A. Baral, V.K. Yadav, S. Aggarwal,S. Chowdhury, D. MukhopadhyayWriting, review, and/or revisionof themanuscript:K.Pal, K.M.Alkharfy,M.R. Spaller, D. MukhopadhyayAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): D. MukhopadhyayStudy supervision: M.R. Spaller, D. MukhopadhyayOther (peptide chemical synthesis and purification): A.A. Pletnev

AcknowledgmentsThe authors thank the Mayo Clinic Histology Core Facility and Optical

Morphology Core Facility for their assistance with this work. The authorsalso thank Dr. Thomas C. Smyrk for providing kind and helpful sugges-tions with histologic analysis.

Grant SupportThis work is supported by NIH grants CA78383 and CA150190 (to D.

Mukhopadhyay).The costs of publication of this article were defrayed in part by the

payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received April 4, 2014; revised July 16, 2014; accepted August 4, 2014;published OnlineFirst August 14, 2014.

References1. Li D, Xie K, Wolff R, Abbruzzese JL. Pancreatic cancer. Lancet

2004;363:1049–57.2. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer

J Clin 2014;64:9–29.3. Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. Future of

cancer incidence in the United States: burdens upon an aging, chang-ing nation. J Clin Oncol 2009;27:2758–65.

4. BarbaraNP,Wrana JL, LetarteM.Endoglin is an accessory protein thatinteracts with the signaling receptor complex of multiple members ofthe transforming growth factor-beta superfamily. J Biol Chem1999;274:584–94.

5. Gougos A, Letarte M. Primary structure of endoglin, an RGD-contain-ing glycoprotein of human endothelial cells. J Biol Chem 1990;265:8361–4.

6. Bernabeu C, Conley BA, Vary CP. Novel biochemical pathways ofendoglin in vascular cell physiology. J Cell Biochem 2007;102:1375–88.

7. Lee NY, Ray B, How T, Blobe GC. Endoglin promotes transforminggrowth factor beta-mediated Smad 1/5/8 signaling and inhibits endo-thelial cell migration through its association with GIPC. J Biol Chem2008;283:32527–33.

8. Li DY, Sorensen LK, Brooke BS, Urness LD, Davis EC, Taylor DG, et al.Defective angiogenesis in mice lacking endoglin. Science 1999;284:1534–7.

9. Behrem S, Zarkovic K, Eskinja N, Jonjic N. Endoglin is a better markerthan CD31 in evaluation of angiogenesis in glioblastoma. Croat Med J2005;46:417–22.

10. Munson J, Bonner M, Fried L, Hofmekler J, Arbiser J, Bellamkonda R.Identifying new small molecule anti-invasive compounds for gliomatreatment. Cell Cycle 2013;12:2200–9.

11. El-Gohary YM, Silverman JF, Olson PR, Liu YL, Cohen JK, Miller R,et al. Endoglin (CD105) and vascular endothelial growth factor asprognostic markers in prostatic adenocarcinoma. Am J Clin Pathol2007;127:572–9.

12. Gomez-Esquer F, Agudo D, Martinez-Arribas F, Nunez-Villar MJ,Schneider J. mRNA expression of the angiogenesis markers VEGFand CD105 (endoglin) in human breast cancer. Anticancer Res2004;24:1581–5.

13. Saad RS, Liu YL, Nathan G, Celebrezze J, Medich D, SilvermanJF. Endoglin (CD105) and vascular endothelial growth factor asprognostic markers in colorectal cancer. Mod Pathol 2004;17:197–203.

14. Yoshitomi H, Kobayashi S, Ohtsuka M, Kimura F, Shimizu H, Yoshi-dome H, et al. Specific expression of endoglin (CD105) in endothelialcells of intratumoral blood and lymphatic vessels in pancreatic cancer.Pancreas 2008;37:275–81.

15. Yang LY, Lu WQ, Huang GW, Wang W. Correlation between CD105expression and postoperative recurrence and metastasis of hepato-cellular carcinoma. BMC Cancer 2006;6:110.

16. Takahashi N, Haba A, Matsuno F, Seon BK. Antiangiogenic therapy ofestablished tumors in human skin/severe combined immunodeficien-cy mouse chimeras by anti-endoglin (CD105) monoclonal antibodies,and synergy between anti-endoglin antibody and cyclophosphamide.Cancer Res 2001;61:7846–54.

17. UnedaS,ToiH, Tsujie T, TsujieM,HaradaN, TsaiH, et al. Anti-endoglinmonoclonal antibodies are effective for suppressing metastasis andthe primary tumors by targeting tumor vasculature. Int J Cancer2009;125:1446–53.

18. Rosen LS,Hurwitz HI,WongMK,Goldman J,MendelsonDS, FiggWD,et al. A phase I first-in-human study of TRC105 (Anti-Endoglin Anti-body) in patients with advanced cancer. Clin Cancer Res 2012;18:4820–9.

19. Liu Y, Starr MD, Brady JC, Dellinger A, Pang H, Adams B, et al.Modulation of circulating protein biomarkers following TRC105(anti-endoglin antibody) treatment in patients with advanced cancer.Cancer Med 2014;3:580–91.

20. Gromova P, Rubin BP, Thys A, Cullus P, Erneux C, Vanderwinden JM.ENDOGLIN/CD105 is expressed in KIT positive cells in the gut and ingastrointestinal stromal tumours. J Cell Mol Med 2012;16:306–17.

21. Davidson B, Stavnes HT, Forsund M, Berner A, Staff AC. CD105(Endoglin) expression in breast carcinoma effusions is a marker ofpoor survival. Breast 2010;19:493–8.

22. Ziebarth AJ, Nowsheen S, Steg AD, Shah MM, Katre AA, Dobbin ZC,et al. Endoglin (CD105) contributes to platinum resistance and is atarget for tumor-specific therapy in epithelial ovarian cancer. ClinCancer Res 2013;19:170–82.

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

24. Fujiwara K, Ohuchida K, Ohtsuka T,Mizumoto K, Shindo K, Ikenaga N,et al. Migratory activity of CD105þ pancreatic cancer cells is stronglyenhanced by pancreatic stellate cells. Pancreas 2013;42:1283–90.

25. DeVries L, LouX, ZhaoG, ZhengB, FarquharMG.GIPC, aPDZdomaincontaining protein, interacts specifically with the C terminus of RGS-GAIP. Proc Natl Acad Sci U S A 1998;95:12340–5.

26. Varsano T, Dong MQ, Niesman I, Gacula H, Lou X, Ma T, et al. GIPC isrecruited by APPL to peripheral TrkA endosomes and regulates TrkAtrafficking and signaling. Mol Cell Biol 2006;26:8942–52.

27. Wieman HL, Horn SR, Jacobs SR, Altman BJ, Kornbluth S, RathmellJC. An essential role for the Glut1 PDZ-binding motif in growth factorregulation of Glut1 degradation and trafficking. Biochem J 2009;418:345–67.

28. Wang LH, Kalb RG, Strittmatter SM. A PDZ protein regulates thedistribution of the transmembrane semaphorin, M-SemF. J Biol Chem1999;274:14137–46.

29. Wang L, Mukhopadhyay D, Xu X. C terminus of RGS-GAIP-interactingprotein conveys neuropilin-1-mediated signaling during angiogenesis.FASEB J 2006;20:1513–5.






30. Muders MH, Vohra PK, Dutta SK, Wang E, Ikeda Y, Wang L, et al.Targeting GIPC/synectin in pancreatic cancer inhibits tumor growth.Clin Cancer Res 2009;15:4095–103.

31. Muders MH, Dutta SK, Wang L, Lau JS, Bhattacharya R, Smyrk TC,et al. Expression and regulatory role of GAIP-interacting protein GIPCin pancreatic adenocarcinoma. Cancer Res 2006;66:10264–8.

32. Kedlaya RH, Bhat KM, Mitchell J, Darnell SJ, Setaluri V. TRP1 inter-acting PDZ-domain protein GIPC forms oligomers and is localized tointracellular vesicles in human melanocytes. Arch Biochem Biophys2006;454:160–9.

33. Kirikoshi H, Katoh M. Expression of human GIPC1 in normal tissues,cancer cell lines, and primary tumors. Int J Mol Med 2002;9:509–13.

34. Wang L, Lau JS, Patra CR, Cao Y, Bhattacharya S, Dutta S, et al. RGS-GAIP-interacting protein controls breast cancer progression. MolCancer Res 2010;8:1591–600.

35. Patra CR, Rupasinghe CN, Dutta SK, Bhattacharya S, Wang E, SpallerMR, et al. Chemically modified peptides targeting the PDZ domain ofGIPC as a therapeutic approach for cancer. ACS Chem Biol2012;7:770–9.

36. Saro D, Li T, Rupasinghe C, Paredes A, Caspers N, Spaller MR. Athermodynamic ligand binding study of the third PDZ domain (PDZ3)from the mammalian neuronal protein PSD-95. Biochemistry2007;46:6340–52.

37. Memic A, Spaller MR. How do halogen substituents contributeto protein-binding interactions? A thermodynamic study of pep-tide ligands with diverse aryl halides. Chembiochem 2008;9:2793–5.

38. Udugamasooriya DG, Sharma SC, Spaller MR. A chemical libraryapproach to organic-modified peptide ligands for PDZ domain pro-teins: a synthetic, thermodynamic and structural investigation. Chem-biochem 2008;9:1587–9.

39. Bach A, Eildal JN, Stuhr-Hansen N, Deeskamp R, Gottschalk M,Pedersen SW, et al. Cell-permeable and plasma-stable peptidomi-metic inhibitors of the postsynaptic density-95/N-methyl-D-aspartatereceptor interaction. J Med Chem 2011;54:1333–46.

40. Grillo-Bosch D, Choquet D, Sainlos M. Inhibition of PDZ domain-mediated interactions. Drug Discov Today Technol 2013;10:e531–40.


Pal et al.




Published OnlineFirst August 14, 2014.Mol Cancer Ther Krishnendu Pal, Alexandre A. Pletnev, Shamit K. Dutta, et al. Cancer Cell Growth

GIPC Interaction Inhibits Pancreatic−Inhibition of Endoglin

Updated version

10.1158/1535-7163.MCT-14-0291doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2014/08/19/1535-7163.MCT-14-0291.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/2014/09/22/1535-7163.MCT-14-0291To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-14-0291

http://mct.aacrjournals.org/content/suppl/2014/08/19/1535-7163.MCT-14-0291.DC1

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

mailto:[email protected]

http://mct.aacrjournals.org/content/early/2014/09/22/1535-7163.MCT-14-0291


Documents

Inhibition of Endoglin GIPC Interaction Inhibits Pancreatic Cancer …€¦ · 22/09/2014 · Inhibition of Endoglin–GIPC Interaction Inhibits Pancreatic Cancer Cell Growth Krishnendu