cd44 expression level and isoform contributes to pancreatic … · that cd44 may be a molecular...

Biology of Human Tumors

CD44 Expression Level and Isoform Contributesto Pancreatic Cancer Cell Plasticity, Invasiveness,and Response to TherapyShujie Zhao1, Chen Chen1, Katherine Chang1,2, Anand Karnad1,2, Jaishree Jagirdar3,Addanki P. Kumar4,5, and James W. Freeman1,2,5

Abstract

Purpose: A subpopulation of pancreatic ductal adenocarci-noma (PDAC) cells is thought to be inherently resistant tochemotherapy or to give rise to tumor cells that becomeresistant during treatment. Here we determined the role ofCD44 expression and its isoforms as a marker and potentialtarget for tumor cells that give rise to invasive and gemcitabine-resistant tumors.

Experimental Design: RT-PCR, Western blotting, and DNAsequencing was used to determine CD44 isoform and expressionlevels. Flow cytometry was used to sort cells on the basis of theirCD44 expression level. CD44 expression was knocked downusing shRNA. Tumorigenic properties were determined by clono-genic and Matrigel assays, IHC, tumor growth in vivo usingluciferase imaging and by tumor weight.

Results:We identified an invasive cell population that gives riseto gemcitabine-resistant tumors. These cancer cells express a highlevel of CD44 standard isoform and have an EMT phenotype(CD44s/EMT). In vivo, CD44s/EMT engraft and expand rapidlyand give rise to tumors that express high levels of CD44 isoformsthat contain multiple exon variants. CD44low-expressing cellsshow continued sensitivity to gemcitabine in vivo and knockdownof CD44 in CD44s/EMT cells increases sensitivity to gemcitabineand decreases invasiveness.

Conclusions: PDAC cells expressing high levels of CD44s witha mesenchymal-like phenotype were highly invasive and devel-oped gemcitabine resistance in vivo. Thus, initial targeting CD44or reversing the CD44high phenotype may improve therapeuticresponse. Clin Cancer Res; 22(22); 5592–604. �2016 AACR.

IntroductionHeterogeneity of cancer cells within tumors and established

cancer cell lines is well known. The impact that this heterogeneityon pathobiology and therapy has not been fully investigated.Epithelial-to-mesenchymal transition (EMT) and acquired cancerstem cell properties are a major cause of cancer cell heterogeneity(1, 2). EMT represents an adaptive plasticity defined as loss ofmembrane polarity and acquired migratory properties that ischaracteristic of mesenchymal cells (3). EMT is thought to playan important role in early cancer cell dissemination and metas-tasis (3, 4). The development of metastasis is believed to requireEMT, which enables cancer cells to migrate from their primarytumor sites. However, recolonizing at distant sites may requireredifferentiation or a mesenchymal-to-epithelial transition

(MET), which is believed to favor tumor engraftment and growthof cancer cells at metastatic sites (3, 5, 6). The phrase adaptiveplasticity explains how cancer cells gain selective growth andsurvival capabilities through phenotypic changes in response totheir environment. During the last few years, the role of epithelialcell plasticity in cancer development has gained more attention;however, the mechanisms that initiate and enable phenotypicchanges are not fully defined.

CD44, a nonkinase transmembrane receptor that binds hya-luronan (HA), has been found to play a role in the EMT process(7, 8). CD44 is a widely expressed adhesion molecule, con-tributing to cell–cell and cell–matrix adhesion, cell growth,differentiation, and trafficking, and is highly expressed on stemcells (2, 9). CD44 is encoded by 20 exons and undergoesextensive alternative splicing to generate CD44 standard(CD44s) and CD44 variant (CD44v) forms. CD44s consiststhe first five exons (exons 1–5) and last five exons (exons 16–20). The CD44 variable exons are typically numbered v1 to v10(v1 is not encoded in human) corresponding to the genomicexons 6–15 and are alternatively spliced and incorporatedbetween the exon 5 and exon 16, either as a single exon variantor as a combination of several exons variant (7, 10). Functionalsignificance of CD44s and CD44v are not fully understood.Recent studies show aberrant levels and variant forms of CD44are expressed in a variety of tumor types including PDAC. Highlevel of CD44 is thought to be associated with poor prognosisof cancer and CD44 v6 and v3 are reported to be relatedto cancer metastasis (9, 11–17). Studies show that CD44v6expression is largely restricted to the advanced stages of tumorprogression and is more prevalent in metastatic than in

1Department of Medicine, Division of Medical Oncology, University ofTexas Health Science Center at San Antonio, San Antonio, Texas.2Experimental and Developmental Therapeutics Program, CancerTherapy and Research Center, San Antonio, Texas. 3Department ofPathology, University of Texas Health Science Center at San Antonio,San Antonio, Texas. 4Department of Urology, University of TexasHealth Science Center at San Antonio, San Antonio, Texas. 5Researchand Development, Audie Murphy Veterans Administration Hospital,San Antonio, Texas.

CorrespondingAuthor: JamesW. Freeman, Department ofMedicine, Division ofMedical Oncology, University of Texas Health Science Center at San Antonio,7703 Floyd Curl Drive, San Antonio, TX 78229-3900; Phone: 210-567-5298; Fax:210-567-6687; E-mail: [email protected].

doi: 10.1158/1078-0432.CCR-15-3115

�2016 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 22(22) November 15, 20165592

on September 3, 2021. © 2016 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst June 7, 2016; DOI: 10.1158/1078-0432.CCR-15-3115

http://clincancerres.aacrjournals.org/

nonmetastatic colorectal adenocarcinomas (18). Elevatedexpression of CD44v6 in pancreatic cancer appears to play afunctional role in metastasis (13, 19). CD44 isoform switchingfrom a variant form to a standard form during the EMT processhas been reported and CD44s is correlated to an EMT pheno-type (1, 20, 21). We previously showed that treating a PDACcell line BxPC-3 with gemcitabine induces cells to undergo anEMT, to become more invasive and to show a phenotypicswitch from expressing mainly low levels of CD44v to expres-sing high levels of CD44s (22). This EMT/CD44s phenotypereverses after multiple passages following removal of gemcita-bine, suggesting that PDAC cells respond to chemotherapy byadaptive plasticity and that CD44 may play a role in thisprocess. Isoform switching of CD44 may be a part of tumorprogression and appear to play a role in EMT (23).

The expressionof CD44s and its isoforms in cancer cells impliesthat CD44 plays a role in promoting tumorigenesis, suggestingthat CD44 may be a molecular target for cancer therapy. Cancerstem cells that express CD44 are reported to be resistant tochemotherapy and to give rise to disease progression duringtherapy or recurrent disease following therapy (24–28). The roleof CD44 in maintenance of the stem cell phenotype is not fullyestablished. In the current study, we demonstrated that CD44high

-expressing cells show predominantly a CD44s isoform, displayan EMT phenotype, and are highly invasive. CD44high CFPAC-1cells showed stem cell properties by forming tumor spheres invitro. Mice implanted with CD44high cells developed rapid tumorgrowth and metastasis. Tumors from CD44high cells initiallyresponded to gemcitabine but rapidly developed resistance,whereas tumors from CD44low cells continued to be responsiveto gemcitabine for the 22-week treatment schedule. This suggeststhat CD44high cells with predominant expression of CD44s havean initial implantation and growth advantage and give rise togemcitabine-resistant tumors. However, there is an uncoupling ofCD44s expression and EMT phenotype during tumor expansionas these tumors show expression of high levels of CD44 variantsand E-cadherin. Thus, this population of CD44high-expressing

cells may show further plasticity by undergoing a CD44 isoformchange and a mesenchymal-to-epithelial (MET) switch. Chemo-therapy or other environmental stresses may induce a CD44high

-expressing cell population that promotes cell survival andincreases invasiveness. Thus, targeting CD44high-expressing cellsmay be an important strategy for increasing the response tochemotherapy.

Materials and MethodsCell culture and materials

Human pancreatic ductal adenocarcinomas (PDAC) cell linesBxPC-3, Capan-2, MIA PaCa-2, CFPAC-1, Panc-1, AsPC-1 werefromATCC,UK Pan-1was established in our laboratory (29), andHPNE (hTERT-immortalized human pancreas nestin expressingcell line) was obtained from Michel M. Ouellette (University ofNebraska Medical Center, Omaha, NE; ref. 30). Cells were main-tained inmedium as recommended and supplemented with 10%FBS in a 37�C, 5% CO2 incubator. The plasmid shCD44-2 pRRLwas a gift from BobWeinberg (Addgene plasmid # 19123), whichtargets all isoforms of human CD44 (31). PMMP-luciferase plas-mid was from Dr. Ricardo Aguiar (University of Texas HealthScience Center, San Antonio, TX). The plasmids were transfectedintohuman embryonic kidney293Tpackaging cells (ATCC)usingFuGENE 6 (Roche Applied Science) according to the manufac-turer's protocol. Cells were then infected with the viral mediumfrom the packaging cells 48 hours after transfection. The CD44knockdown cells were selected by GFP cell sorting (Flow Cyto-metry Core of University of Texas Health Science Center at SanAntonio, TX). The luciferase-expressing clones were selected with300 mg/mL of G418. CD44 levels were determined by Westernblotting analysis and luciferase expression was determined usingluciferase assay kit (Promega). Gemcitabine-HCl for Injectionwasmanufactured by Eli Lily and was dissolved in sterile saline (0.9%sodium chloride) at a concentration of 20 mg/mL.

FACS and ALDEFLUOR assaySubconfluent cells were dissociated with trypsin and washed

twice in serum-free medium. Cells were labeled with FITC (fluo-rescein isothiocyanate) conjugated monoclonal anti-CD44 anti-body (eBioscience) at 4�C for 30 minutes. Brightly and weaklystained cells were separated as gated (Fig. 1B) by FACS and weredesignated as CD44high or CD44low cells respectively. The alde-hyde dehydrogenase (ALDH) activity was determined using anALDEFLUOR assay kit (StemCell Technologies) according to themanufacturer's protocol. Briefly, the cells were stainedwith ALDE-FLUOR reagent (bodipy-aminoacetaldehyde) and/or PE (phyco-erythrin)-conjugated monoclonal CD24 antibody (eBioscience)and incubated for 30 minutes at 37�C. A specific inhibitor ofALDH1, DEBA (diemethylamino-benzaldehyde)was used to con-trol background fluorescence. The ALDH activities and positivepopulation and CD24-positive population of the cells were ana-lyzed using the BD LSRII analyzer (BD Biosciences).

Reverse transcription PCRTotal RNA from cultured cells or from tumors was extracted by

TRIzol reagent (Invitrogen) and 1.0 mg of total RNA was reversetranscribed using the SuperScript III First-Strand Synthesis System(Life Technologies) and 50 ng cDNA was used for the followingPCR reaction. PCR primers used to amplify CD44 isoforms andCD44s only have been described previously (13, 32, 33). The

Translational Relevance

Progression of PDAC after initial response to gemcitabinetreatment is attributed to the expansion of a subpopulation ofinherently resistant tumor cells. It is important to understandthe biology of this resistant subpopulation of PDAC cells andto develop strategies for targeting these cells to prevent tumorregrowth and progression. We found that gemcitabine treat-ment led to expansion of cells that express CD44. Here weshow the separation of a subpopulation of PDAC cells basedon high expression level of CD44. This cell populationexpressed a high level of CD44 standard isoform, was mes-enchymal-like, highly invasive, and gave rise to gemcitabine-resistant tumors that exhibited a CD44 isoform switch withexpression of multi-exon isoform variants of CD44. Knock-downofCD44 inCD44high cell resulted in cells thatweremoresensitive to gemcitabine and less invasive. These studies sug-gest that high CD44 expression is a marker for gemcitabineresistance in PDAC and that targeting CD44high-expressingcellsmay provide a therapeutic strategy for improving survival.

CD44 and Gemcitabine Resistance

www.aacrjournals.org Clin Cancer Res; 22(22) November 15, 2016 5593




A

B

D

E F

C

160 kDa CD44v1.5k

1.0k

0.5k

1.5k1.0k0.5k

1.5k1.0k

0.5k

1.1k1.0k0.7ks10

s8s8

v6v6

v3

F R

Primer CD44s

Primer 4Primer 3Primer 2Primer 1

CD44v

MK Capan

-1

Capan

-2

BxPC-3

CFPAC-1

MIA PaCa-2

Panc-1

UK Pan-1

AsPC-1

HPNECap

an-1

Capan

-2

BxPC-3

CFPAC-1

MIA PaCa-2

Panc-1

UK Pan-1

AsPC-1

CD44 v8-10 (seq)

v3-v6 (seq)

v6-10 (seq)

v6 (seq)

CD44v3isoforms

v3 (seq)

CD44s

ESRP1

b-Actin

v6 (seq)

CD44s

Exon 1 2 3 4 5 6 7 8 9 1011 12 13 14 1516 17 1819 20

s9s7v10v8v6v4v2s5s1 s3

s3s4

s1 s2 s3 s4 s5 s6 s7 s8 s9 s10

0.5k0.3k

CD44s

ZEB1

E-Cad

Vimentin

b-Actin

75 kDa

160140120100

8060402010

86

Rel

ativ

e C

D44

s m

RN

A ex

pres

sion

Rel

ativ

e m

RN

A ex

pres

sion

Rel

ativ

e m

RN

A ex

pres

sion

420

cap-1

Capan

-2

BxPC-3

CFPAC-1

Mia PaC

a-2

Panc-1

UK Pan-1

AsPC-3

60

50

40

30

20

10

0Capan-1 Capan-2 CFPAC-1

CD44s

1.2

1.0

0.8

0.6

0.4

0.2

0.0Mia PaCa-2 Panc-1 UK Pan-1 AsPC-1

CD44v3CD44v6

CD44sCD44v3CD44v6

BxPC-3

Figure 1.

Association of CD44 isoforms with PDAC cell phenotypes. A, Western blot analysis of a panel of PDAC cell lines for the expression of CD44 and EMT markers,ZEB1, E-cadherin, and vimentin. b-actin was used for the protein loading control. B, the diagrams of PCR primer design for CD44s and CD44v mRNAexpression. C, reverse transcription PCR (RT-PCR) analysis of CD44 isoforms and ESRP1 expressed in a panel of PDAC cell lines. CD44v isoforms labeled as CD44v(seq) were confirmed by DNA sequencing. D–F, quantitative RT-PCR with SYBR Green supermix for the expression of CD44s, CD44v3, and CD44v6containing isoforms. Data represent fold change (mean � SD) calculated with 2�DDCt method of triplicate experiments.

Zhao et al.

Clin Cancer Res; 22(22) November 15, 2016 Clinical Cancer Research5594




quantitative PCR reaction was done with iQ SYBR Green Super-mix (Bio-Rad) according to the manufacturer's instruction. TheCD44s, CD44v3, and CD44v6 primers used for qRT-PCR havebeen described elsewhere (13, 33). The 2�DDCt method was usedfor calculating the relative mRNA expression.

DNA sequencingPCR products from PDAC cell lines or tumors using CD44

isoform primers were separated by 1.2% agarose gel. DNA frag-ments of CD44 isoforms were purified by Zymoclean Gel DNARecovery kit (ZYMO Research) according to the manufacturer'sinstruction. In brief, DNA fragments of our interests were excisedby clean scalpel under UV light, agarose gel slices were incubatedat 55�C for 5 to 10 minutes until they are completely dissolved.Melted agarose was applied to a Zymo-spin column andDNAwaseluted from the column. At least 150 ng of each interested DNAwere sent to the DNA Core Facility in the University of TexasHealth Science Center for DNA sequencing and were then con-firmed using the NCBI blast alignment tool.

Colony and sphere formation assaysThe tumorigenic properties were determined by colony forma-

tion assay and tumor sphere formation assay. For the colonyformation, single-cell suspensionwas plated at density of 50, 200,and 500 cells per 60-mmdishes and cultured for 2weeks. The cellswere washed with PBS, then fixed and stained with 0.1% crystalviolet/40%methanol and colonies� 1mmwere counted. For thesphere formation, 2,000 cells were seeded per well in a 24-welllow attachment plate (Corning) and cultured in the serum-freemedia containing 2% B27, 1% N2, 20 ng/mL human bFGF(Invitrogen, Life Technologies) and 50 ng/mL rhEGF (R&D Sys-tems) for 7 to 10 days. Spheres �30 mm were counted under thelight microscopy. For serial passages, spheres were disassociatedto single cell by trypsin and 2,000 cells were replaced per well.

Western blottingWhole-cell lysateswere obtained in Laemmli sample buffer and

Western blotting analysis was performed using standard techni-ques as described previously (34). Antibodies against CD44 andZEB1were purchased fromCell Signaling Technology. E-cadherinand vimentin were purchased from Santa Cruz Biotechnology.

Orthotopic pancreatic cancer mouse modelFour to 5-week-old female athymic nude mice were purchased

fromHarlan Laboratories, Inc. Mice were housed andmaintainedin accordance with the standards of Animal Care and Use Com-mittee of Audie Murphy Veterans Administration Hospital (SanAntonio, TX). Cells were grown to 80% confluence, trypsinized,and resuspended in saline (0.9% sodium chloride). A total of 2.5� 105 cells/50 mL were injected into the tail of the pancreas ofanesthetized mice. The orthotopic xenograft experiment wasrepeated to determine gemcitabine response in vivo. Two weeksafter implantation, mice received either gemcitabine (80 mg/kg)or saline by intraperitoneal injection, once a week until the end ofexperiment. Tumor growth andmetastasis were monitored week-ly by bioluminescent imaging using the IVIS Imaging system(Caliper Life Sciences). Six weeks after implantation, controlgroup mice showing stress signs were euthanized and remainingmice were continually received gemcitabine once a week untilstress signs shown. During necropsy, mice were evaluated grosslyfor evidence of tumor invading and metastasizing into the peri-

toneum and liver. Liver metastases were also examined by thebioluminescent image during the necropsy.

Histopathology and IHCParaffin-embedded primary tumor and liver sections were

stained with hematoxylin and eosin (H&E) and IHC was per-formed with mAbs to CD44 (Cell Signaling Technology) or toKi67 (Abcam) in the Histology and Pathology Laboratory (Uni-versity of Texas Health Science Center, San Antonio, TX). Thetumor desmoplasia was evaluated using Sirius Red staining pro-tocol from IHC WORLD, LLC and quantitated using ImageJsoftware (NIH, Bethesda, MD). The primary tumor and livermetastasis were examined under the light microscopy.

Statistical analysisStatistical analysis was performed using GraphPad InStat soft-

ware (GraphPad Software, Inc.). The significance of differencesamong groups was determined by one-way ANOVA followed byBonferroni multiple comparison test, t test, or Fisher exact testaccordingly. Statistically significance was considered as P < 0.05.

ResultsExpression of CD44 and EMT makers in PDAC cell lines

A panel of eight PDAC cell lines and HPNE (30), an immor-talized, undifferentiated, nontumorigenic, and nestin-positivecell line was examined by Western blotting for the expression ofCD44 and markers of epithelial and mesenchymal phenotypes.The cell lines that express a higher level of CD44 also expressedgreater levels of themesenchymalmarker vimentin, and the EMT-associated transcription factor ZEB-1 and lower levels of E-cad-herin, an epithelial marker (Fig. 1A). Interestingly, the epithelial-like and mesenchymal-like PDAC cell lines express not onlydifferent levels, but also distinct patterns of the CD44 standardisoform (CD44s) and variant isoforms of CD44 (CD44v). PDACcell lines and the undifferentiated HPNE cell line that possess amesenchymal phenotype express a high level of CD44s and lowlevels of single exon or low molecular size variants of CD44 asdetermined by RT-PCR (Fig. 1C), by DNA sequencing of RT-PCRproducts and by quantitative RT-PCR (Fig. 1D and F). In com-parison, PDAC cell lines showing an epithelial-like phenotypeexpressed CD44v containing combinations of multiple variantexons (Fig. 1A and C). Expression of single exon variant CD44v3and CD44v6 or low molecular size CD44 isoforms possessinglimited numbers of variant exons were more common to men-senchymal-like PDAC cell lines (Fig. 1B, and C, primers 2–4). TherelativemRNA levels of CD44swere evaluated by quantitative RT-PCR for the eight PDAC cell lines (Fig. 1D). In agreement with theWestern blot analysis (Fig. 1A) the mRNA levels of CD44s werehigher in the four cell lines that exhibit a clear mesenchymalphenotype (MIAPaCa-2, Panc-1, UKPan-1, and AsPC-1) com-pared with those cells that showed a mixed (CFPAC-1) or anepithelial (Capan-1, Capan-2, BxPC-3) phenotype (Fig. 1D).PDAC cells with an epithelial phenotype showed a relativelylower level of CD44s protein and mRNA expression (Fig. 1A,D, and E). Interestingly, the ratio of CD44v (containing v3 and/orv6) to CD44s mRNA expression was higher in cells with anepithelial phenotype and the reverse ratio (higher ratio of CD44sto CD44v) was seen in cells with a mesenchymal phenotype(Fig. 1E and F). Correspondingly, ESRP1 was mainly expressedin the epithelial-like phenotype PDAC cell lines (Fig. 1C). Thus,






PDAC cells with a mesenchymal phenotype express greater totallevels of CD44 with a predominance of CD44s isoform andCD44v expression was restricted to single variant exon or lesseramounts of combination variant isoforms (Fig. 1).

Isolation of PDAC cells based on expression level of CD44The cell lines CFPAC-1 and AsPC-1 showed low levels of

expression of both E-cadherin and vimentin and intermediatelevels of CD44s with expression of higher molecular weightCD44v isoforms (Fig. 1A), suggesting that these cell lines mayrepresent an intermediate phenotype or contain a mixture ofepithelial and mesenchymal phenotypes. To further investigatethese possibilities, CD44high andCD44low expressing populationsof CFPAC-1 and AsPC-1 were separated by flow cytometry(Fig. 2A). The CD44high and CD44low populations displayeddifferent morphologies and expressed differential EMT markers;CD44high cells were more fibroblast like in appearance andexpressed high levels of vimentin and low levels of E-cadherin,whereas CD44low cells exhibit a more cobblestone-like appear-

ance and expressed low levels of vimentin and high levels ofE-cadherin (Fig. 2B and C). CD44 expression levels were con-firmed by Western blotting, which showed that CD44high cellsexpressed higher levels of CD44 with CD44s as the predominantisoform. In comparison with CD44high cells, CD44low cells exp-ressed lower level of CD44s (Fig. 2C).

Oncogenic properties of CD44high and CD44low cellsThe oncogenic properties of cells from CD44high and CD44low

populations were compared by determining their clonogenic,invasive, and tumor sphere–forming capacities. CD44high cellsisolated from CFPAC-1 and AsPC3 parental cells were moreclonogenic and more invasive than the CD44low cells as deter-mined by colony-forming assay and Matrigel invasion assays(Fig. 3A and B). CD44 is recognized as a cancer stem cell markerand is widely used to enrich cancer stem cells in vitro in a variety ofcancers (2, 9). A tumor sphere assay was used to test whetherCD44 expression and mesenchymal phenotype were correlatedwith cancer cell stemness, CD44high-expressing CFPAC-1 cells

250

150

100

50

5010

015

020

025

030

0

50

102 103 104 105 102 103 104 105

100

150

200

250

0

102 103 104 105 102 103 104 105

025

050

075

01,

000

200

CD44Low

CFPAC-1 CFPAC-1S

SC

-A

(

x 1,

000)

SS

C-A

(x

1,00

0)

CD44High

CD44Low

CD44 FITC-A

CD44-Hi

CF

PAC

-1A

sPC

-1

CD44-Lo

CD44v

CD44s

E-cadherin

Vimentin

b-Actin

CFPAC-1

p Hi Lo p Hi Lo

AsPC-1

CD44 FITC-A

CD44High

CD44Low

CD44High

CD44Low

Cou

ntC

ount

CD44 FITC-A CD44 FITC-A

AsPC-1AsPC-1

CD44High

B C

A

Figure 2.

Isolation of CD44 high and low cells.A, cells were labeled with FITC-conjugated monoclonal anti-CD44antibody and CD44high and CD44low

cells were separated as gated byFACS. B, light microscopic imagesrepresent the morphology ofCD44high and CD44low cells aftercell sorting. C, expression of CD44s,CD44v, and EMT marker proteinsfrom cell lines after cell sorting wasanalyzed by Western blotting. p,parental cells; Hi, CD44high cells;Lo, CD44low cells.

Zhao et al.





continuously formed spheres through four subsequent serialpassages (Fig. 3C) and cells from these spheres express CD44and vimentin as determined by immunofluorescent staining (Fig.

3D), whereas, CD44low CFPAC-1–expressing cells show adecreased capacity to form spheres after serial passages (Fig.3C). The current understanding of markers for pancreatic cancer

Figure 3.

In vitro tumorigenic and invasive properties of CD44high and CD44low cells. A, clonogenic assays were performed as described previously (34). Briefly, cells wereseeded at 50, 250, and 500 cells per 6-cm2 dishes and cultured for 2 weeks. The cells were stained with crystal violet and colonies �1 mm were counted.The datawere presented asmean� SD of experiments performed in triplicates. B, 3� 104 cells/well were seeded in 24-well Matrigel invasion chambers for 24 hours.Invaded cells were counted under microscope and the data represent mean � SD from triplicate experiments. Top, photos of invasion assays; bottom,quantification data of invasion assays. C, tumor sphere–forming assay of CFPAC-1 cells. Cells (2,000) were plated in a low-attachment 24-well plate and cultured inthe medium described in Materials and Methods for 7–10 days and spheres �30 mm were counted. P1–P4 are the passage numbers; every passage lastedfor 7—10 days. D, confocal microscopy images of immunofluorescence staining of CD44 and vimentin in tumor spheres of CD44high population cells. CD44-Hi,CD44High; CD44-Lo, CD44Low.






stem cells include CD44, CD24, and ALDH1 as reviewed by Vazand colleagues. and Medema (35, 36). To further assess stem cellproperties, CFPAC1/CD44high and CD44low cells were comparedusing FACS analyses to detect CD24 and ALDH-positive cells.CD24was detected using an antibody and ALHDbright cells weredetected using an ALDEFLUOR assay. CD24þALHD bright ordouble positive cells were greater in CFPAC-1-CD44high cells(4.3%) compared with CFPAC1-CD44low cells (1.2%; data notshown).

The CD44high population in AsPC-1 cells showed a clearmesenchymal phenotype and lack of E-cadherin; however, theCD44low population remained rather intermediate with contin-ued higher expression of CD44s than CD44v and intermediatelevels of E-cadherin and vimentin (Fig. 2C). The lack of sphere-forming activity by serial passages in the AsPC-1 high cells (datanot shown) suggests that expression of CD44smay play a role butis not in itself sufficient for tumor sphere formation and stemness.

CD44high/mesenchymal cells aremore tumorigenic and are lessresponsive to gemcitabine in vivo

To compare tumor progression in vivo, we orthotopicallyimplanted either CD44high or CD44low populations of CFPAC-1 cells that express firefly luciferase into the pancreas of athymicnude mice. Bioluminescent imaging was performed weekly tomonitor tumor progression. Six weeks after tumor cell implan-tation, mice were euthanized and tumor growth and metastasiswere evaluated by tissue examination, tumor weight, and histo-pathology. Thequantificationdata of thebioluminescent imagingfor the 12 mice in each of the CD44high and CD44low groups isshown in Fig. 4A (right) and the representative bioluminescentimages up to 6 weeks for a representative mouse in the CD44high

and CD44low groups is shown in Fig. 4A (left). Tumor growth andmetastatic spread were more rapid in mice implanted withCD44high-expressing cells. By week 6, mice implanted withCD44high cells developed larger tumors and showed greater localinvasion (Fig. 4A and B) and liver metastasis (Fig. 4C). Biolumi-nescent imaging of resected livers showed that 9 of 12 mice hadmetastatic foci inCD44high groupwhile only 2of 12mice from theCD44low group showed positive bioluminescent imaging(Fig. 4C). Histologically, the metastatic liver foci were within thelymphatic tissue surrounding the liver or showed bile ductinvolvement with surrounding hepatic invasion (Fig. 4D). Mac-roscopic observation also showed evidence of peritoneal metas-tases and tumor attachment to the intestines and by histology ofseveral mesenteric lymph nodes in some mice.

We next compared the response to gemcitabine for orthotopi-cally implanted CD44high and CD44low cells in vivo. Mice weretreated with vehicle (saline) alone or gemcitabine once a weekbeginning two weeks after implantation. Tumors in the placebogroup grew and progressed more rapidly (Fig. 4E and F). Thesecontrol groups of mice were euthanized at 6 weeks after implan-tation of tumor cells due to tumor burden and signs of physicalstress in the group implanted with CD44high cells. The gemcita-bine-treated group ofmice continued to receiveweekly treatment.In the first 6 weeks, both CD44high and CD44low groups of miceresponded to gemcitabine and tumor growth was inhibited orregressed (Fig. 4E). However, in mice treated weekly with gemci-tabine, the CD44high groupmice displayed signs of tumor growthbyweek twelve and tumors continued to grow exponentially untilthe end of the study at week 22 at which time these mice formedascites. Interestingly, tumors remained small to nondetectable in

the gemcitabine-treatedCD44low group ofmice through 22weeks(Fig. 4E and F).

To determine whether CD44 expression had an effect on tumorcell proliferation, immunohistochemical staining for Ki67 wasperformed. The results showed that tumor cell proliferation wassignificantly lower in CD44low tumors when treated with gemci-tabine, whereas, tumors that grew from the CD44high groupfollowing gemcitabine treatment had proliferation rate similarto the saline-treated group (Fig. 5A). To further investigate the roleof CD44 in tumor recurrence after gemcitabine treatment, tumortissues were analyzed by IHC. CD44low tumors showed lessintensity of staining in saline-treated controls compared withCD44high tumors (Fig. 5B). However, in tumors that grew aftergemcitabine treatment, almost all cells showed intense CD44staining (Fig. 5B). Western blot analysis of tissue from tumorsderived from CD44high cells showed a switch from expressingpredominantly CD44s to expressing high levels of CD44 variantisoforms and these tumors also expressed the epithelialmarker, E-cadherin (Fig. 5C). The expression of epithelial variant form ofCD44 v8-10 and ESRP1 inCD44high tumorswas confirmedbyRT-PCR analysis (Fig. 5D). This suggests that these rapidly growingtumors undergo a CD44 isoform switch towards expressing highlevels of CD44 variant isoforms and a phenotypic change to amore epithelial-like tumor cell. An epithelial phenotype may berequired for the rapid growth of tumor cells. Thus, the greater levelof CD44 expression is associated with an increase in tumorgrowth; however, the specific roles of the CD44 isoforms inengraftment and growth need to be further determined.

Desmoplasia commonly develops in pancreatic cancer and isthought to contribute to chemoresistance. It is possible that CD44may play a role in this process. To investigate this possibility,tumors derived from CD44high and CD44low cells were stainedwith Sirius Red to detect collagen. No differences of the collagenstaining were found between saline-treated CD44high andCD44low tumors; however, gemcitabine-treated tumors displayedlarge area of fibrosis-like pathology and collagen staining(Fig. 5E). This suggests that CD44 expression level does not playa role in the development of desmoplasia although gemcitabinetreatment induces fibrosis perhaps as a result of cell death intumor areas.

Knockdown CD44 inhibits tumorigenicity and increasessensitivity to gemcitabine

To further look at the role of CD44 in mediating tumorigenicproperties, CD44 was knocked down in CD44high cells withshRNA (Fig. 6F). These studies show that knocking down CD44significantly reduced the colony-forming property in CFPAC-1CD44high cells (Fig. 6A) and enhanced their response to gemci-tabine similar to the level of CD44low cells (Fig. 6B). To test ourhypothesis that CD44high cells may contain a subpopulation ofcells that possess stem-like properties and are originally resistantto gemcitabine, cells were exposed to high dose of gemcitabine(50 ng/mL) for 2 weeks. We found that colonies only grewappreciably from the CD44high cells and that knocking downCD44 from CD44high cells abolished the colony-forming activityin the presence of high dose of gemcitabine (Fig. 6C). The invasiveand sphere-forming properties of these cells were tested by in vitroMatrigel invasion and sphere-forming assays. Knocking downCD44 significantly inhibited the invasiveness in both CFPAC-1and AsPC-1 CD44high cells (Fig. 6D) and prevented the sphere-forming activity of CFPAC-1 (Fig. 6E). However, knocking down

Zhao et al.





Figure 4.

CD44mediate tumor growth, invasion, and chemoresistance in vivo. A total of 2.5� 105CF/CD44high (CD44-Hi) andCF/CD44low (CD44-Lo) cellswere implanted intothe tail of pancreas of nude mice. Primary tumor and metastases were examined grossly and pathologically at week 6 after implantation. A, bioluminescentimages of tumors from CD44high and CD44low cells monitored weekly by bioluminescent imaging (left). Tumor growth curve plotted from bioluminescentdata of CF/CD44high and CF/CD44low groups. The data represent means (n ¼ 12) of each group, Bars, SE. � , P < 0.05; �� , P < 0.01, CF/CD44low versus CF/CD44high

group. B, tumors collected at the end of experiment from CF/CD44high and CF/CD44low groups (left) and a quantification data of tumor weight of 12 miceeach group (right).C,bioluminescent images of livermetastases taken at the end of experiment fromCF/CD44high andCF/CD44lowmice.D, representative photos ofhematoxylin and eosin (H&E)-stained liver sections. The arrows indicate the metastatic foci. E, tumor growth curve plotted from bioluminescent data ofCF/CD44high and CF/CD44low groups treated with saline for 6 weeks or gemcitabine for 20 weeks (n ¼ 6 each group). �� , P < 0.01; �� , P < 0.001, comparedwith CF/CD44high group. F, a quantification data of tumor weight (mean �SE, n ¼ 6 each group). CD44-Hi, CD44High; CD44-Lo, CD44Low; Gem, gemcitabine.






CD44 itself was not sufficient to completely reverse the EMTphenotype of these cells, as the mesenchymal marker, vimentinexpression was only slightly reduced and the epithelial marker, E-cadherin was not reexpressed in the CD44 knockdown clones(Fig. 6F).

DiscussionIn this study, we show that the phenotypic and tumorigenic

properties of PDAC cells are related to the level of expression andtype of CD44 isoform. An initial observation from analyses of apanel of PDAC cell lineswas that cells showing anEMTphenotypeexpress higher levels of CD44 with a predominance of the stan-dard isoform. To further look at heterogeneity of CD44 expres-sion, we sorted out CD44high and CD44low-expressing cells fromtwo separate PDAC cell lines, CFPAC-1 and AsPC-1. CD44high

cells express predominantly a CD44s isoform, display an EMTphenotype, and are more invasive in vitro. In comparison,CD44low cells express relative low level of CD44s and low levelsof larger size CD44 variants, are more epithelial like, and are lessinvasive. Our finding that PDAC cells with an EMT phenotypeexpress mainly CD44s isoform is in agreement with a study byBrown and colleagues (20) showing that induction of EMT inHMLE cells was associated with a switch from expressing mainlyCD44v isoforms to expressing predominantly CD44s. Our studyis also in agreement with that ofMani and colleagues (1) showingthat EMT was associated with higher level in expression of CD44.Interestingly, the study by Mani and colleagues (1) shows thatinduction of EMTgenerates cells that have properties of stem cells.Here we also show using the CFPAC-1 model that CD44high butnot CD44low-expressing cells were able to grow in serial passagesas tumor spheres. However, only a low percentage of cells (4.3%)from the CD44high group show coexpression of other stem cellmarkers including CD24 and ALDH. To this point, knockdown ofCD44 in CD44high cells was sufficient to prevent growth of thesecells as tumor spheres. However, CD44high cells fromASPC-1 cellswere not able to form tumor spheres. These results suggests thatCD44 may play a role but is not sufficient by itself for a stem cellphenotype.

In addition to expressing high levels of CD44s, PDAC cells withan EMT phenotype express small molecular size exon variantisoforms of CD44v3 or CD44v6. This phenomenon appearsunique for cells with a mesenchymal phenotype as cells with anepithelial phenotype express CD44v isoforms that contain pre-dominantly multiple variant exons. A number of studies suggestthat expression of CD44 variants including CD44v3 and v6 areassociated withmetastatic lesions (16, 17, 19, 32). The functionaldifferences of CD44 molecules that possess single or multiplevariant exons is not clear andneeds further investigation. The highlevel of expression in CD44high cells suggests that CD44s isoformplays an important role in tumor cell plasticity contributing toincreased invasiveness. However, this study does not rule out thepossibility that minor variant isoforms contribute significantly tothese properties.

The functional significance of the CD44 in CFPAC-1 cells wasfurther assessed using an orthotopic mouse model. Tumors devel-oped from CD44s/EMT cells grew and metastasized more rapidlythan CD44low expressing cells. However, once tumors grew in vivo,the tumor cells underwent a phenotypic change. Western blot andRT-PCR analyses show that tumor specimens from the CD44high

group express higher level of CD44 variant isoforms (v8–10) than

the standard form and showed an increase in expression of theepithelial marker, E-cadherin, which is expressed at a low level ornot present in the original CD44high cells. This phenomenonsuggests that cells from CD44high-induced tumors undergo METand that a CD44 isoform switch may be required for this to occur.As these tumors appear to be resistant to gemcitabine, this suggestsan uncoupling of CD44s expression and gemcitabine resistance.However, these tumors continue to express high levels ofCD44butas variant isoforms. Thus, the level of CD44 may be important inmaintaining resistance. Desmoplasia an important component ofPDAC (37, 38) did not appear to differ appreciably in tumorsgrown from CD44high or CD44low cells but was dramaticallyincreased by gemcitabine treatment.

EMT andMET are dynamic processes in tumor progression, andCD44 may play a role in regulating these processes. EMT isinvolved in invasion, tumor metastasis, and stem cell properties,whereas, conversion back to an epithelial phenotype may favortumor growth (39). Tumor cells once engrafted, may need CD44vto convert EMT-like cells to epithelial-like cells (MET); otherwise,CD44s may be required to facilitate cells to undergo EMT, whichincreases local invasion and formetastasizing to the distal regions.Thus, the total level of CD44 as well as the isoform may beimportant in tumor progression. EMT appears in this study to beclosely associated with loss of ESRP1 and a decrease in expressionof CD44 multi-exon variants. ESRP1 has been reported to play arole in the isoform switch of CD44 (6, 20, 21, 40). ExpressingESRP1 in PDAC cells was shown to decrease proliferation, inva-sion, andmetastases (41)We showed here that ESRP1 expressionwas associated with cells that are epithelial-like and express lowerlevels of total CD44, of whichmost express CD44vmolecules thatpossess multiple variant exons. ESRP1 expression was decreasedor lost in the mesenchymal-like cells that mainly express CD44s.We also demonstrated that the tumors derived from CD44high

cells underwent CD44 isoform switch and gained the ESRP1expression when grown as orthotopic implants. This observationis consistent with the premise that the molecular basis of CD44isoform switch is regulated by ESRP1. Therefore, regulation ofESRP1 expression likely controls, at least in part, the type of CD44isoform expressed and downregulation of ESRP1 likely facilitatesexpression of mainly the form of CD44s.

CD44 is highly expressed in variety of tumors, tumor-initiatingcells, and recurrent tumors, suggesting that it could be a thera-peutic target for cancer. Continued progression and recurrence ofcancers following chemotherapy is thought to be due to inherentor acquired resistance to the therapeutic agent. In this study, wetreated the CD44high and CD44low tumor-bearing mice withgemcitabine. Although both CD44high and CD44low tumorsresponded to gemcitabine at the beginning, CD44high tumorsgradually develop resistance after 12weeks of treatment. CD44low

tumors continued to be sensitive to gemcitabine through 22weeks suggesting that CD44 expression level is related to che-moresistance and disease progression. We have also demonstrat-ed that knocking down CD44 in CD44high cells attenuated col-ony-forming activity and invasiveness and increased the sensitiv-ity to gemcitabine. A subpopulationofCD44high cellsmaypossessstem cell properties that are inherently resistant to gemcitabine.Preclinical studies show promising results by targeting CD44using CD44 antibodies (24, 27). These studies demonstrated thatrelapsing tumors are sensitive to anti-CD44 treatment; therefore,targeting CD44 could have a therapeutic benefit. Our studysuggests the possibility that targeting the CD44high-expressing

Zhao et al.





Figure 5.

A, representative pictures of immunostaining of tumor sections with mAb to Ki67 (left) and quantification data of Ki67-stained tumor sections analyzed byImmunoRatio online software (right). The data were presented as the mean � SD. B, representative pictures of IHC of tumor sections with CD44 mAb. C,tumor tissue lysates from saline-treated group and collected 6 weeks after implantation. Fifty micrograms of the tissue lysates were analyzed by Western blotanalysis with indicated antibodies. b-actin was used for the protein loading control. Last two lanes labeled as "Cell" are lysates from CF/CD44high and CD44low

cells used for the implantation. D, RT-PCR analysis of tumor tissues treated with saline using the primer 1 as indicated in Fig. 1B and CD44 isoforms were separatedin 1.2 % agarose gel. CD44v isoforms labeled as CD44 v8–10 (seq) were confirmed by DNA sequencing. E, representative photos of Sirius Red staining of tumorsections (left) and a quantification data (right) analyzed using ImageJ program (NIH, Bethesda, MD). CD44-Hi, CD44High; CD44-Lo, CD44Low; Gem, gemcitabine.






Figure 6.

A and B, 500 cells were plated per 6-cm2 dishes. The next day, cells were treated with one dose of 0, 2.0, and 5.0 ng/mL gemcitabine and cultured for 2 weeks. Thecells were stained with crystal violet and colonies � 1 mm were counted. A, the data were presented as mean � SD of experiments performed in triplicates.B, the datawere calculated as the percent of colony number relative to control (gemcitabine 0 dose). C, 5,000 cells were plated per 6-cm2 dishes. The next day, cellswere treated with gemcitabine (50 ng/mL) and gemcitabine was replenished every week for 2 weeks. Colonies � 1 mm were counted. The data representthe mean � SD for triplicate experiments. D, 3 � 104/well cells were seeded in 24-well Matrigel invasion chambers for 24 hours. Invaded cells were counted undermicroscope and the data represent mean � SD from triplicate experiments. � , P < 0.05; �� , P < 0.01; and �� , P < 0.001 compared with CD44high cells. E,tumor sphere–forming assayof CF/CD44high, CF/CD44low, and shCD44 cells. Cells (2,000)were plated in a low-attachment 24-well plate and cultured in themediumdescribed in Materials and Methods and passaged every 7–10 days and then the 2,000 cells were replated. The passages–4 spheres (� 30 mm) were counted.F, CD44 levels and EMT marker proteins in CD44 knocking down CF/CD44high and As/CD44high cells analyzed by Western blotting. cl.1 and cl.2 representsshCD44 clone 1 and clone 2, respectively.

Zhao et al.





tumor cells may increase and prolong response to therapy andperhaps prevent the outgrowth of a gemcitabine-resistant popu-lation. Inour studies, CD44high cells formed tumors that graduallydeveloped resistance to gemcitabine. It is of interest that thegemcitabine-resistant tumors continued to express high totallevels of CD44 composed of CD44s but also showed a dramaticincrease in the levels ofCD44v that possessmultiple variantCD44exons.

In summary, PDAC cell lines show heterogeneity in regards toCD44 isoform and level of expression. In vitro, CD44high-expres-sing cells show predominantly a CD44s isoform, display an EMTphenotype, and in vivo are highly invasive and form tumors andmetastases more rapidly. Mice implanted with CD44high cellsshow initial response to gemcitabine but gradually show rapidexpansion and loss of response to gemcitabine. Mice implantedwith CD44low cells showed a continued response to gemcitabineover a 22-week period. This suggests that CD44low-expressingphenotype will have a prolonged response to chemotherapy.Molecular analyses show that CD44high tumors continue toexpress high levels of CD44 but with a switch to expressingincreased levels of CD44v isoforms. Knockdown of CD44 inCD44high cells causes these cells to be less invasive and moresensitive to gemcitabine. This study suggests that targeting CD44in PDAC that show a CD44high expression level may benefittherapy and improve overall response.

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

Authors' ContributionsConception and design: A.P. Kumar, J.W. FreemanDevelopment of methodology: S. ZhaoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): S. Zhao, C. Chen, K.W. Chang, J.W. FreemanAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S. Zhao, C. Chen, K.W. Chang, J.W. FreemanWriting, review, and/or revision of the manuscript: S. Zhao, C. Chen,K.W. Chang, A. Karnad, A.P. KumarAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): S. ZhaoStudy supervision: A.P. Kumar, J.W. FreemanOther (review of pathology slides): J. Jagirdar

AcknowledgmentsWe thank the Flow Cytometry Core Facility and the Histology and

Pathology Laboratory (University of Texas Health Science Center, SanAntonio, TX).

Grant SupportThisworkwas supported by grants from theWilliam and EllaOwensMedical

Research Foundation, VA Merit Award, Cancer Center Grant to CTRC and NIH-P30CA054174 (to J. W. Freeman), and CPRIT training grant (to K. Chang).

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 December 31, 2015; revised April 6, 2016; accepted May 10, 2016;published OnlineFirst June 7, 2016.

References1. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, et al. The

epithelial-mesenchymal transition generates cells with properties of stemcells. Cell 2008;133:704–15.

2. Keysar SB, Jimeno A. More than markers: biological significance of cancerstem cell-defining molecules. Mol Cancer Ther 2010;9:2450–7.

3. Nieto MA. Epithelial plasticity: a common theme in embryonic and cancercells. Science 2013;342:1234850.

4. RhimAD,Mirek ET, Aiello NM,Maitra A, Bailey JM,McAllister F, et al. EMTand dissemination precede pancreatic tumor formation. Cell 2012;148:349–61.

5. Tam WL, Weinberg RA. The epigenetics of epithelial-mesenchymal plas-ticity in cancer. Nat Med 2013;19:1438–49.

6. Yae T, Tsuchihashi K, Ishimoto T, Motohara T, Yoshikawa M, Yoshida GJ,et al. Alternative splicing of CD44 mRNA by ESRP1 enhances lung colo-nization of metastatic cancer cell. Nat Commun 2012;3:883.

7. Goodison S,Urquidi V, TarinD.CD44 cell adhesionmolecules.Mol Pathol1999,52:189–196.

8. Ponta H, Sherman L, Herrlich PA. CD44: from adhesion molecules tosignalling regulators. Nat Rev Mol Cell Biol 2003;4:33–45.

9. Zoller M. CD44: can a cancer-initiating cell profit from an abundantlyexpressed molecule? Nat Rev Cancer 2011;11:254–67.

10. Sneath RJ, ManghamDC. The normal structure and function of CD44 andits role in neoplasia. Mol Pathol 1998;51:191–200.

11. Gansauge F, Gansauge S, Zobywalski A, Scharnweber C, Link KH, NusslerAK, et al. Differential expression of CD44 splice variants in humanpancreatic adenocarcinoma and in normal pancreas. Cancer Res 1995;55:5499–503.

12. KaufmannM, Heider KH, Sinn HP, vonMinckwitz G, Ponta H, Herrlich P.CD44 variant exon epitopes in primary breast cancer and length of survival.Lancet 1995;345:615–9.

13. Li Z, Chen K, Jiang P, Zhang X, Li X, Li ZCD44v/CD44s expression patternsare associated with the survival of pancreatic carcinoma patients. DiagnPathol 2014;9:79.

14. MuramakiM,Miyake H, Kamidono S, Hara I. Over expression of CD44V8-10 in human bladder cancer cells decreases their interaction with hya-

luronic acid and potentiates their malignant progression. J Urol 2004;171:426–30.

15. TodaroM, Gaggianesi M, Catalano V, Benfante A, Iovino F, BiffoniM, et al.CD44v6 is a marker of constitutive and reprogrammed cancer stem cellsdriving colon cancer metastasis. Cell Stem Cell 2014;14:342–56.

16. Wang SJ, Wreesmann VB, Bourguignon LY. Association of CD44 V3-containing isoforms with tumor cell growth, migration, matrix metallo-proteinase expression, and lymphnodemetastasis in head andneck cancer.Head Neck 2007;29:550–58.

17. KuniyasuH, Chihara Y, Kubozoe T, Takahashi T. Co-expression of CD44v3and heparanase is correlated with metastasis of human colon cancer. Int JMol Med 2002;10:333–7.

18. Wielenga VJ, Heider KH, Offerhaus GJ, Adolf GR, van den Berg FM, PontaH, et al. Expression of CD44 variant proteins in human colorectal cancer isrelated to tumor progression. Cancer Res 1993;53:4754–6.

19. Rall CJ, Rustgi AK. CD44 isoform expression in primary and metastaticpancreatic adenocarcinoma. Cancer Res 1995;55:1831–5.

20. Brown RL, Reinke LM, Damerow MS, Perez D, Chodosh LA, Yang J, et al.CD44 splice isoform switching in human and mouse epithelium isessential for epithelial-mesenchymal transition and breast cancer progres-sion. J Clin Invest 2011;121:1064–74.

21. Preca BT, Bajdak K, Mock K, Sundararajan V, Pfannstiel J, Maurer J, et al. Aself-enforcing CD44s/ZEB1 feedback loop maintains EMT and stemnessproperties in cancer cells. Int J Cancer 2015;137:2566–77

22. Bera A, VenkataSubbaRaoK,ManoharanMS,Hill P, Freeman JW. AmiRNAsignature of chemoresistant mesenchymal phenotype identifies novelmolecular targets associated with advanced pancreatic cancer. PLoS One2014;9:e106343.

23. Biddle A, Gammon L, Fazil B, Mackenzie ICCD44 staining of cancer stem-like cells is influenced by down-regulation of CD44 variant isoforms andup-regulation of the standard CD44 isoform in the population of cells thathave undergone epithelial-to-mesenchymal transition. PLoS One 2013;8:e57314.

24. MolejonMI, Tellechea JI, Loncle C, Gayet O, Gilabert M, Duconseil P, et al.Deciphering the cellular source of tumor relapse identifies CD44 as amajor






therapeutic target in pancreatic adenocarcinoma. Oncotarget 2015;6:7408–23.

25. Boulbes DR, Chauhan GB, Jin Q, Bartholomeusz C, Esteva FJ. CD44expression contributes to trastuzumab resistance in HER2-positive breastcancer cells. Breast Cancer Res Treat 2015;151:501–13.

26. Hong SP, Wen J, Bang S, Park S, Song SY. CD44-positive cells are respon-sible for gemcitabine resistance in pancreatic cancer cells. Int J Cancer2009;125:2323–31.

27. Li L, Hao X, Qin J, Tang W, He F, Smith A, et al. Antibody against CD44sinhibits pancreatic tumor initiation and postradiation recurrence in mice.Gastroenterology 2014;146:1108–18.

28. Du YR, Chen Y, Gao Y, Niu XL, Li YJ, DengWM. Effects andmechanisms ofanti-CD44 monoclonal antibody A3D8 on proliferation and apoptosis ofsphere-forming cells with stemness from human ovarian cancer. Int JGynecol Cancer 2013;23:1367–75.

29. Fralix KD, Ahmed MM, Mattingly C, Swiderski C, McGrath PC, Venkata-subbarao K, et al. Characterization of a newly established human pancre-atic carcinoma cell line, UK Pan-1. Cancer 2000;88:2010–21.

30. Lee KM, Nguyen C, Ulrich AB, Pour PM, Ouellette MM. Immortalizationwith telomerase of the Nestin-positive cells of the human pancreas.Biochem Biophys Res Commun 2003;301:1038–44.

31. Godar S, Ince TA, Bell GW, Feldser D, Donaher JL, Bergh J, et al. Growth-inhibitory and tumor- suppressive functions of p53 depend on its repres-sion of CD44 expression. Cell 2008;134:62–73.

32. Kuncova J, Kostrouch Z, Viale M, Revoltella R, Mandys V. Expression ofCD44v6 correlates with cell proliferation and cellular atypia in urothelialcarcinoma cell lines 5637 and HT1197. Folia Biol 2005;51:3–11.

33. Raso-Barnett L, Banky B, Barbai T, Becsagh P, Timar J, Raso E. Demon-stration of a melanoma-specific CD44 alternative splicing pattern thatremains qualitatively stable, but shows quantitative changes duringtumour progression. PLoS One 2013;8:e53883.

34. Zhao S, VenkatasubbaraoK, Lazor JW, Sperry J, JinC,Cao L, et al. Inhibitionof STAT3 Tyr705 phosphorylation by Smad4 suppresses transforminggrowth factor beta-mediated invasion and metastasis in pancreatic cancercells. Cancer Res 2008;68:4221–8.

35. Vaz AP, PonnusamyMP, Seshacharyulu P, Batra SK. A concise reviewon thecurrent understanding of pancreatic cancer stem cells. J Cancer Stem CellRes 2014;2:1–21

36. Medema JP. Cancer stem cells: the challenges ahead. Nat Cell Biol 2013;15:338–44.

37. Shields MA, Dangi-Garimella S, Redig AJ, Munshi HG. Biochemical role ofthe collagen-rich tumour microenvironment in pancreatic cancer progres-sion. Biochem J 2012;441:541–52.

38. Schober M, Jesenofsky R, Faissner R, Weidenauer C, HagmannW, Michl P,et al. Desmoplasia and chemoresistance in pancreatic cancer. Cancers2014;6:2137–54.

39. Brabletz T. EMT and MET in metastasis: where are the cancer stem cells?Cancer Cell 2012;22:699–701.

40. Warzecha CC, Sato TK, Nabet B, Hogenesch JB, Carstens RP. ESRP1 andESRP2 are epithelial cell-type-specific regulators of FGFR2 splicing. MolCell 2009;33:591–601.

41. Ueda J,MatsudaY, YamahatsuK,UchidaE,Naito Z,KorcM, et al. Epithelialsplicing regulatory protein 1 is a favorable prognostic factor in pancreaticcancer that attenuates pancreatic metastases. Oncogene 2014;33:4485–95.


Zhao et al.




2016;22:5592-5604. Published OnlineFirst June 7, 2016.Clin Cancer Res Shujie Zhao, Chen Chen, Katherine Chang, et al. Cancer Cell Plasticity, Invasiveness, and Response to TherapyCD44 Expression Level and Isoform Contributes to Pancreatic

Updated version

10.1158/1078-0432.CCR-15-3115doi:

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/22/22/5592.full#ref-list-1

This article cites 41 articles, 10 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/22/22/5592.full#related-urls

This article has been cited by 5 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/22/22/5592To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-15-3115

http://clincancerres.aacrjournals.org/content/22/22/5592.full#ref-list-1

http://clincancerres.aacrjournals.org/content/22/22/5592.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/22/22/5592


cd44 expression level and isoform contributes to pancreatic … · that cd44 may be a molecular...

Documents