Differential Regulation of ZEB1 and EMT by MAPK ... · Krishan Kumar1,2, Christina R. Chow1,3, Kazumi Ebine1,2, Ahmet D. Arslan1,3, ... exhibit increased activation of the MAPK-interacting

Oncogenes and Tumor Suppressors

Differential Regulation of ZEB1 and EMT byMAPK-Interacting Protein Kinases (MNK) andeIF4E in Pancreatic CancerKrishan Kumar1,2, Christina R. Chow1,3, Kazumi Ebine1,2, Ahmet D. Arslan1,3,Benjamin Kwok1, David J. Bentrem2,3,4, Frank D. Eckerdt3, Leonidas C. Platanias1,2,3, andHidayatullah G. Munshi1,2,3

Abstract

Human pancreatic ductal adenocarcinoma (PDAC) tumorsare associated with dysregulation of mRNA translation. In thisreport, it is demonstrated that PDAC cells grown in collagenexhibit increased activation of the MAPK-interacting proteinkinases (MNK) that mediate eIF4E phosphorylation. Pharma-cologic and genetic targeting of MNKs reverse epithelial–mes-enchymal transition (EMT), decrease cell migration, and reduceprotein expression of the EMT-regulator ZEB1 without affectingZEB1 mRNA levels. Paradoxically, targeting eIF4E, the best-characterized effector of MNKs, increases ZEB1 mRNA expres-sion through repression of ZEB1-targeting miRNAs, miR-200cand miR-141. In contrast, targeting the MNK effector hnRNPA1,which can function as a translational repressor, increases ZEB1

protein without increasing ZEB1 mRNA levels. Importantly,treatment with MNK inhibitors blocks growth of chemoresis-tant PDAC cells in collagen and decreases the number ofaldehyde dehydrogenase activity–positive (Aldefluorþ) cells.Significantly, MNK inhibitors increase E-cadherin mRNA levelsand decrease vimentin mRNA levels in human PDAC organoidswithout affecting ZEB1 mRNA levels. Importantly, MNK inhi-bitors also decrease growth of human PDAC organoids.

Implications: These results demonstrate differential regulation ofZEB1 and EMT by MNKs and eIF4E, and identify MNKs as poten-tial targets in pancreatic cancer.Mol Cancer Res; 14(2); 216–27.�2015AACR.

IntroductionPancreatic ductal adenocarcinoma (PDAC) is currently the

fourth leading cause of cancer-related deaths in the United States,with a median survival of approximately 6 months and 1-yearsurvival of approximately 20% (1, 2). The continuing dismaloutcome is attributed to the fact that PDAC is an aggressive cancer.Contributing to the aggressive nature of this cancer is the intensefibrotic reaction that is associated with the primary tumor and themetastatic PDAC lesions (3, 4). The fibrotic reaction, which canaccount for over 80%of the tumormass (4, 5), contains extensiveamounts of fibrillar collagen. We have previously shown thatPDAC cells respond to type I fibrillar collagen by increasingmotility, decreasing epithelial markers, and upregulating mesen-chymal markers (6–8).

PDAC tumors are also associated with dysregulation ofmRNA translation of proinvasive genes that contributes totumor progression (9, 10). In addition, increased phosphory-lation of the cap-binding protein eukaryotic translation initi-ation factor 4E (eIF4E) in human PDAC tumors is associatedwith high-grade tumors and poor prognosis (9). The functionof eIF4E is critical for malignant transformation and promotestumor development in vivo (11, 12). The MAPK-interactingprotein kinases 1 and 2 (MNK1 and MNK2) mediate eIF4Ephosphorylation on Ser209 and regulate eIF4E-mediatedmRNA translation (13, 14). An additional MNK effector ishnRNPA1 (heterogeneous nuclear ribonucleoprotein A1),which has previously been shown to function as a translationalrepressor of some genes (15, 16). Because of the importance ofmRNA translation in tumorigenesis (17, 18), a better under-standing of the contribution of MNKs to pancreatic cancerprogression may provide a targeted approach for the treatmentof PDAC patients.

In this report, we show that PDAC cells grown in three-dimen-sional (3D) type I collagen demonstrate increased eIF4E phos-phorylation. PDAC cells that have undergone epithelial–mesen-chymal transition (EMT) also demonstrate increased activationof MNKs in collagen. Pharmacologic and genetic targeting ofMNKs reverses EMT, decreases cellmigration, and reduces proteinexpression of the EMT-regulator ZEB1 without affecting ZEB1mRNA levels. Significantly, MNK inhibitors increase E-cadherinmRNA levels and decrease vimentin mRNA levels in humanpancreatic organoids without affecting ZEB1 mRNA levels. Par-adoxically, targeting eIF4E increases ZEB1 mRNA and proteinexpression. In contrast, targeting the MNK effector hnRNPA1

1Department of Medicine, Feinberg School of Medicine, NorthwesternUniversity, Chicago, Illinois. 2Jesse BrownVAMedical Center, Chicago,Illinois. 3The Robert H. Lurie Comprehensive Cancer Center of North-westernUniversity,Chicago, Illinois. 4DepartmentofSurgery, FeinbergSchool of Medicine, Northwestern University, Chicago, Illinois.

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

K. Kumar and C.R. Chow are co-first authors of the article.

Corresponding Author: Hidayatullah G. Munshi, Feinberg School of Medicine,Northwestern University, 303 E. Superior Street, Lurie Building, Room 3-117,Chicago, IL 60611. Phone: 312-503-2301; Fax: 312-503-0386; E-mail:[email protected]

doi: 10.1158/1541-7786.MCR-15-0285

�2015 American Association for Cancer Research.

MolecularCancerResearch

Mol Cancer Res; 14(2) February 2016216

on June 16, 2020. © 2016 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 25, 2015; DOI: 10.1158/1541-7786.MCR-15-0285

http://mcr.aacrjournals.org/

increases ZEB1 protein without increasing ZEB1 mRNA levels.Importantly, treatment with MNK inhibitors blocks growth ofchemoresistant PDAC cells in collagen, inhibits growth of PDACorganoids, and decreases the number of Aldefluor(þ) cells, sug-gesting that MNKs may regulate cancer stem cells and may bepotential targets in pancreatic cancer.

Materials and MethodsReagents

General tissue culture materials were obtained from VWRInternational. Antibodies against eIF4E, tubulin, HSP90, ZEB1,and Dicer were obtained from Santa Cruz Biotechnology, where-as antibodies against p-eIF4E,MNK1, p-MNK1, andDrosha werepurchased from Cell Signaling Technology. Anti-GAPDH anti-body was from Millipore, anti–E-cadherin antibody wasobtained from BD Biosciences, whereas anti-vimentin antibodywas from Abcam. Secondary antibodies were purchased fromSigma. CGP57380 was obtained from Santa Cruz Biotechnolo-gy. siRNAs against MNK1 and MNK2 were purchased fromDharmacon, ZEB1 siRNA was obtained from Life Technologies,whereas eIF4E and hnRNPA1 siRNAs were from Santa CruzBiotechnology. The Aldefluor Assay Kit was purchased fromStemcell Technologies.

Cell cultureAsPC1, CD18/HPAF-II, and Panc1 cells were obtained from the

ATCC. Cells were maintained in DMEM containing 10% FBS andantibiotics (100 U/mL penicillin and 100 mg/mL streptomycin).Chemoresistant CD18 (CD18-CR) cells were generated by treat-ing parental CD18 cells with increasing concentration of 5-fluo-rouracil (5-FU) over a period of 3months (19). The surviving cellswere maintained in 10 mmol/L concentration of 5-FU. The CD18and CD18-CR cells were authenticated by short tandem repeatprofiling at the Johns Hopkins Genetic Resources Core Facility inOctober 2013, whereas AsPC1 and Panc1 cells were authenticatedin June 2010.

Embedding and examination of cells in 3D type I collagen gelsCollagen mixture (2 mg/mL) was made by adding the appro-

priate volumes of sterile water, 10X DMEM, and NaOH and kepton ice until needed (8, 20). Cells were then suspended in thecollagen solution and allowed to gel at 37�C. For protein analysis,the collagen gels were treated with collagenase to extract cells forWestern blotting. For morphologic examination of cells, cellcolonies in 3D collagen were examined using a Zeiss Axiovert40CFLmicroscope and pictures takenwith aNikonCoolpix 4500camera (8). The relative size of individual colonies was measuredusing ImageJ.

TransfectionCells were transfected with siRNA againstMNK1,MNK2, ZEB1,

eIF4E, or control siRNA using RNAimax (Invitrogen) according tothe manufacturer's instructions before plating into collagen (8).

Quantitative real-time PCR analysisQuantitative gene expression was performedwith gene-specific

probes as described previously (8, 20). Similarly, expression ofmiR-200a/b/c, miR-141, and RNU48 was analyzed as previouslypublished (21).

Isolation of polysomal RNAPolysomal fractionation was performed as previously

described (22, 23). Briefly, cell pellets were lysed in hypotonicpolysomal lysis buffer, clarified by centrifugation, and opticaldensity (OD) at 260 nm was measured for each of the super-natant samples. DMSO- and CGP57380-treated supernatantscontaining 300 OD were then layered over 10% and 50%continuous sucrose gradients. Following ultracentrifugation, thefractions were collected while monitoring the absorbance at254/260 nm as a function of gradient depth. The polysomalfractions were pooled, total RNA from polysomal fractions wasisolated, and the levels of ZEB1 and GAPDH mRNA in thepolysomal fractions and in the whole-cell lysates were deter-mined by qRT-PCR. The relative amounts of ZEB1 mRNA in thepolysomal fractions were then compared with the relativeamounts of ZEB1 mRNA in the whole-cell lysates.

Human PDAC tissue analysisPancreatic tissue was obtained from patients with pancreatic

adenocarcinoma on an Institutional review board (IRB)-approved protocol. The tissue microarray specimens were stainedwith p-eIF4E antibody (Abcam), and also trichrome stained toassess for fibrosis (6).

Human PDAC organoidsDeidentified human PDAC tumor specimens were processed

using the recently published Tuveson Laboratory protocol (24).Briefly, the tumors were minced and digested with collagenaseII and TrypLE, embedded in growth factor–reduced Matrigel,and maintained in human complete media (24). To examinethe effects of targeting MNKs on gene expression in theseorganoids, organoids were treated with DMSO or CGP57380for 96 hours, and the effect on gene expression was determinedby qRT-PCR.

ImmunoblottingImmunoblotting for p-eIF4E, eIF4E, p-MNK1, MNK1, E-cad-

herin, vimentin, ZEB1, Dicer, Drosha, hnRNPA1, GAPDH,HSP90, and tubulin was done as previously described (8, 20).

Statistical analysisAll statistical analyses were done using Microsoft Excel or

GraphPad Instat using a two-tailed t test analysis. Error barsrepresent SEs.

ResultsCollagen-dependent phosphorylation of eIF4E in PDAC cells

Increased eIF4E phosphorylation on Ser209 in human PDACtumors correlates with higher-grade tumors and worse progno-sis (9). Because increased fibrosis can be associated with worseprognosis in human PDAC tumors (3, 4), we examined theeffect of the collagen microenvironment on eIF4E phosphory-lation. As shown in Fig. 1A, PDAC cells in 3D collagen dem-onstrate increased eIF4E phosphorylation on Ser209. BecauseMNKs are known to phosphorylate eIF4E on Ser209 in othersystems (25, 26), we examined the effects of the MNK inhibitorCGP57380 on collagen-induced eIF4E phosphorylation. Ini-tially, we examined the effect of the collagen microenvironmenton MNK phosphorylation/activation. PDAC cells in collagendemonstrate increased MNK1 phosphorylation on Thr197/202

MNKs and Pancreatic Cancer Progression

www.aacrjournals.org Mol Cancer Res; 14(2) February 2016 217




(Fig. 1B), whereas treatment with the MNK inhibitor CGP57380blocked collagen-induced eIF4E phosphorylation (Fig. 1C).

In parallel studies, we examined the effect of siRNA-medi-ated knockdown of MNK1/2 on collagen-induced eIF4E phos-phorylation. Transfection with MNK1/2 siRNAs successfullydownregulated MNK1 and MNK2 levels and blocked colla-gen-induced eIF4E phosphorylation (Fig. 1D). Importantly,immunohistochemical staining of a human pancreatic tissuemicroarray showed that there was a trend toward associationbetween eIF4E phosphorylation in human PDAC tumorsand fibrosis (Fig. 1E). These results demonstrate that the col-lagen microenvironment modulates MNK1/2-dependent eIF4Eregulation.

Collagen-dependent phosphorylation of eIF4E inchemoresistant PDAC cells

It is now well recognized that cells that have undergoneEMT are more invasive and metastatic (27, 28). Because cellsdeveloping resistance to chemotherapy can undergo EMT (27),we treated CD18 cells with increasing concentrations of 5-FUchemotherapy to generate chemoresistant (CD18-CR) cells.These cells demonstrated loss of E-cadherin and increased ex-pression of vimentin (Fig. 2A), indicating that the CD18-CR cellshave undergone EMT. CD18-CR cells also showed increasedexpression of ZEB1 (Fig. 2A), a well-known regulator of EMT(27, 29, 30). The CD18-CR cells were also more migratory onthe collagen-coated surfaces (Fig. 2B). Although CD18-CR cellsdemonstrated reduced levels of eIF4E phosphorylation com-pared with CD18 cells (Fig. 2C), the CD18-CR cells also dem-

onstrated increased eIF4E and MNK phosphorylation whengrown in 3D collagen (Fig. 2C and Supplementary Fig. S1).Furthermore, treatment with the MNK inhibitor CGP57380, ortransfection with MNK1/2 siRNAs, blocked collagen-inducedeIF4E phosphorylation in CD18-CR cells (Fig. 2D and E).

Targeting MNKs increases E-cadherin, decreases vimentin, andreduces migration of PDAC cells

We next examined the effect of pharmacologic targetingMNKs on E-cadherin and vimentin protein and mRNAlevels in CD18-CR cells. Treatment with the MNK inhibitorCGP57380 increased E-cadherin protein and mRNA levels anddecreased vimentin protein and mRNA levels (Fig. 3A). Treat-ment with CGP57380 also decreased migration of CD18-CRcells on collagen-coated surfaces (Fig. 3B).

We also evaluated the relative role of MNK1 and MNK2 inregulating EMT by downregulating MNK1, MNK2, or bothMNK1 and MNK2. Downregulating either MNK1 or MNK2decreased eIF4E phosphorylation (Fig. 3C); however, com-bined downregulation of MNK1 and MNK2 completelyblocked eIF4E phosphorylation (Fig. 3C). Combined knock-down of MNK1 and MNK2 particularly increased E-cadherinmRNA and protein levels (Fig. 3C and D). Although MNK1knockdown had minimal effect on vimentin mRNA and pro-tein levels, MNK2 knockdown decreased vimentin protein andmRNA levels (Fig. 3C and D). Moreover, the effect of MNK2siRNA on vimentin levels was not further affected by cotrans-fection with MNK1 siRNA. These results suggest that MNK2 inparticular regulates EMT in PDAC cells.

AsPC1 CD18 Panc1

AsPC1 CD18 Panc1

AsPC1 CD18 Panc1

AsPC1 CD18 Panc1

AsPC1 CD18 Panc1

MNK2 MNK2 MNK2

MNK1 MNK1 MNK1

Plastic

Plastic Plastic Plastic

Plastic PlasticCollagen

Collagen Collagen Collagen

Collagen Collagen

25 kDa

25 kDa

55 kDa

55 kDa

50 kDa

50 kDa25 kDa

25 kDa

25 kDa

25 kDa

36 kDa

50 kDa

36 kDa

p-elF4E

p-MNK1

MNK1

elF4E

Tubulin

p-elF4E

elF4E

GAPDH

GAPDH

DMSO DMSO DMSOCGP CGP CGP

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

Fol

d ch

ange

(gen

e/G

AP

DH

)F

old

chan

ge(g

ene/

GA

PD

H)

siCtrl siCtrl siCtrlsiMNK1/2 siMNK1/2

MNK1

siMNK1/2

p-elF4E

elF4E

p-elF4E

p-el

F4E

Tubulin

siCtrl

siMNK1/

2

siCtrl

siMNK1/

2siC

trl

siMNK1/

2siC

trl

siMNK1/

2

siCtrl

siMNK1/

2siC

trl

siMNK1/

2

Trichrome

Trichrome

PD

AC

Tum

ors

0–10–1

2–3

2–3

3

2

4

18

7

20

P = 0.09

Total

A

B

C

D E

Figure 1.Collagen-dependent phosphorylation of eIF4E in PDAC cells. A, PDAC cells (AsPC1, CD18, and Panc1) were grown on tissue culture plastic or in 3D type Icollagen (2 mg/mL) for 24 hours. The cell lysates were analyzed for eIF4E phosphorylation on Ser209 (p-eIF4E) by Western blotting. B, PDAC cellswere grown for 24 hours on tissue culture plastic or in 3D type I collagen (2 mg/mL). The cell lysates were analyzed for MNK1 phosphorylation on Thr197/202by Western blotting. C, PDAC cells growing in 3D type I collagen were treated with DMSO or CGP57380 (2.5 mmol/L) for 24 hours and the effect oneIF4E phosphorylation was determined by Western blotting. D, PDAC cells were transfected with control siRNA (siCtrl) or a combination of MNK1- andMNK2-specific siRNAs (siMNK1/2) for 48 hours. The cells were then grown in 3D type I collagen (2 mg/mL) for additional 24 hours. The effect on MNK1and MNK2 mRNA expression was determined by qRT-PCR. Effect on MNK1 protein levels and eIF4E phosphorylation was determined by Western blotting.The results are representative of at least three independent experiments. E, human pancreatic TMAs containing 27 pancreatic tumor specimens wereimmunostained with anti–p-eIF4E antibody and assessed for fibrosis using trichrome staining. Samples were obtained on an IRB-approved protocol.The relative immunostaining and the extent of fibrosis were graded as low (0 or 1þ) or high (2þ or 3þ). The relationship between p-eIF4E expressionin the tissue samples and the extent of fibrosis was assessed by the Fisher exact test.

Kumar et al.

Mol Cancer Res; 14(2) February 2016 Molecular Cancer Research218




Targeting MNKs decreases the protein expression of ZEB1without reducing ZEB1 mRNA levels

We next evaluated the effect of targeting MNKs on ZEB1protein and mRNA levels. Initially, we examined the effect ofZEB1 siRNA in reversing EMT and attenuating migration.Downregulating ZEB1 restored E-cadherin expression inCD18-CR cells (Fig. 4A), and reduced migration of CD18-CRcells on collagen-coated surfaces (Fig. 4B). Significantly, treat-ment with the MNK inhibitor CGP57380 decreased ZEB1protein levels without affecting ZEB1 mRNA levels in CD18-CR and AsPC1 cells (Fig. 4C). Moreover, knockdown of MNK1,

MNK2, or both MNK1 and MNK2 did not affect ZEB1 mRNAlevels (Fig. 4D). However, while MNK1 siRNA had minimaleffect on ZEB1 protein levels, MNK2 siRNA decreased ZEB1protein levels (Fig. 4D). These results indicate that MNK2 inparticular regulates EMT by blocking translation of ZEB1mRNAin pancreatic cancer cells.

We also examined the effect of CGP57380 on ZEB1 mRNAtranslation using polysomal analysis. CD18-CR cells were trea-ted with CGP57380 for 48 hours, polysomal mRNA wasisolated, and the effect on ZEB1 mRNA in the polysomalfractions was determined. As shown in Fig. 4E, ZEB1 mRNA

1.5

1.0

0.5

0.0

30

20

10

0

10

5

0

E-cadherin

E-cadherin

Vimentin ZEB1

CD18

CD18

CD18 CD18CD18-CR

CD18-CR

CD18

CD18

CD18CD18-CR

CD18-CR

CD18-CR

CD18 CD18-CR CD18 CD18-CR

CD18-CR CD18-CR

120 kDa

90 kDa HSP90

54 kDa 181 kDa

90 kDa90 kDa HSP90 HSP90

Vimentin ZEB1

2D collagen migration

400

300

200

100

0

Rel

ativ

e m

igra

tion

25 kDa

25 kDa

36 kDa

25 kDa

25 kDa

36 kDa

p-elF4E

elF4E

GAPDH

Plastic PlasticCollagen Collagen

Fol

d ch

ange

(gen

e/G

AP

DH

)

CD

18C

D18

-CR

DMSO CGP

p-elF4E

elF4E

GAPDH

CD18-CR

CD18-CR1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

MNK1 MNK2

Fol

d ch

ange

(gen

e/G

AP

DH

)

siCtrl

siCtrl

siMNK1

/2

siMNK1

/2siCtrl siMNK1/2

MNK1

p-elF4E

elF4E

Tubulin

CD18-CR

50 kDa

25 kDa

25 kDa

55 kDa

A

B C

D E

Figure 2.Collagen-dependent phosphorylation of eIF4E in chemoresistant PDAC cells chemoresistant CD18 (CD18-CR) cells were generated as detailed in Materials andMethods. A, CD18 and CD8-CR cells growing on tissue culture plastic were examined by phase contrast microscopy. Lysates from CD18 and CD18-CR cellswere analyzed for E-cadherin, vimentin, and ZEB1 protein andmRNAexpression byWestern blotting and qRT-PCR. B, CD18 and CD18-CR cells were plated onto thin-layer type I collagen matrix overlaid with colloidal gold, allowed to migrate for 24 hours, and the tracks were photographed and quantified; � , P < 0.05. C,CD18 and CD18-CR cells were grown on tissue culture plastic or in 3D type I collagen (2 mg/mL) for 24 hours. The cell lysates were analyzed for eIF4Ephosphorylation on Ser209 by Western blotting. D, CD18-CR cells growing in 3D type I collagen were treated with DMSO or CGP57380 (2.5 mmol/L) for 24 hoursand the effect on eIF4E phosphorylation was determined by Western blotting. E, CD18-CR cells were transfected with control siRNA (siCtrl) or a combinationof MNK1- and MNK2-specific siRNAs for 48 hours. The cells were then grown in 3D type I collagen (2 mg/mL) for additional 24 hours. The effect on MNK1and MNK2 mRNA expression was determined by qRT-PCR. Effect on MNK1 protein levels and eIF4E phosphorylation was determined by Western blotting. Theresults are representative of at least three independent experiments. See also Supplementary Fig. S1.






levels in the polysomal fractions were decreased followingtreatment with CGP57380, thus further suggesting that MNKsregulate ZEB1 mRNA translation in pancreatic cancer cells.

Targeting eIF4E increases ZEB1 protein and mRNA levels anddecreases ZEB1-targeting miR-200c and miR-141 miRNAs

Because eIF4E is the best-characterized target of MNKs(13, 14), we examined the effect of downregulating eIF4E onEMT in CD18-CR cells and AsPC1 cells. Paradoxically, down-regulation of eIF4E increased ZEB1mRNA and protein levels inCD18-CR cells (Fig. 5A), and in AsPC1 and Panc1 cells (Sup-plementary Fig. S2). The increase in ZEB1 levels was associatedwith a decrease in E-cadherin mRNA levels and an increase invimentin mRNA and protein levels (Fig. 5A). Consistent with

increased ZEB1 levels, eIF4E knockdown cells demonstrated amore "fibroblastic" phenotype (Fig. 5A). In contrast, MNK1/2knockdown cells, which have decreased ZEB1 (Fig. 4), showeda more rounded phenotype (Fig. 5A).

It is now well established that the miR-200 family of miRNAscan regulate ZEB1 expression (30, 31). The miR-200 family iscomposed of two clusters of miRNA located on two differentchromosomes, one on chromosome 1 at 1p36 (miR-200b, miR-200a, and miR-429) and the other on chromosome 12 at 12p13(miR-200c and miR-141; refs. 32, 33). Thus, to understandthe differential effect of eIF4E and MNK1/2 on ZEB1 expression,we evaluated the effect on miR-200b and miR-200a on chromo-some 1, and on miR-200c and miR-141 on chromosome 12.Although eIF4E knockdown in pancreatic cancer cells did not

DMSO CGP

120 kDa

90 kDa

54 kDa

90 kDa

E-cadherin

HSP90

Vimentin

HSP90

4

3

2

1

0

1.5

1.0

0.5

0.0

Fol

d ch

ange

(ge

ne/G

AP

DH

)

Fol

d ch

ange

(ge

ne/G

AP

DH

)

E-cadherin Vimentin

DMSO DMSOCGP CGP

CD18-CR

DMSO CGP

CD

18-C

R

2D collagen migration

300

200

100

0DMSO CGP

Rel

ativ

e m

igra

tion

50 kDa

25 kDa

25 kDa

36 kDa

MNK1

p-elF4E

elF4E

GAPDH

CD18-CR

CD18-CR

CD18-CR

siCtrl siMNK1 siMNK2 siMNK1/2 siCtrl siMNK1 siMNK2 siMNK1/2

120 kDa

54 kDa

90 kDa

E-cadherin

E-cadherin

Vimentin

Vimentin

HSP90

CD18-CR

5.0

4.0

3.0

2.0

1.0

0.0

1.5

1.0

0.5

0.0siCtrl siCtrlsiMNK1 siMNK1siMNK2 siMNK2siMNK1/2 siMNK1/2

A

B

C

D

Figure 3.Targeting MNKs increases E-cadherin,decreases vimentin and reducesmigration of PDAC cells. A, CD18-CRcells were treated with CGP57380 (2.5mmol/L) for 72 hours and the effect onE-cadherin and vimentin protein wasdetermined byWestern blotting (left).The effect on E-cadherin and vimentinmRNA expression was determined byqRT-PCR (right). B, CD18-CRpreviously treated with DMSO orCGP57380 (2.5 mmol/L) for 48 hourswere plated onto thin-layer type Icollagen matrix overlaid with colloidalgold, allowed to migrate for 24 hoursin the presence of DMSOor CGP57380(2.5 mmol/L), and the tracks werephotographed and quantified;� , P < 0.05. C and D, CD18-CR cellswere transfected with control siRNA(siCtrl), MNK1 siRNA (siMNK1), MNK2siRNA (siMNK2), or a combination ofMNK1- and MNK2-specific siRNAs(siMNK1/2) for 48 hours. The cellswere then grown in 3D type I collagen(2mg/mL) for additional 24hours. Theeffect on p-eIF4E, E-cadherin, andvimentin expression was determinedbyWestern blotting (C). The effect onE-cadherin and vimentin mRNAexpression was determined byqRT-PCR (D). The results arerepresentative of at least threeindependent experiments.

Kumar et al.





affect the levels of pri-miRNA transcripts of miR-200b, miR-200a,or miR-200c-141 (Supplementary Fig. S3), there was 70% and80% decrease in miR-200c and miR-141 levels (Fig. 5B). Todemonstrate that these miRNAs regulate ZEB1 in CD18-CR cells,we transfected the cells with pre–miR-200c and pre–miR-141 and

evaluated the effect on ZEB1, E-cadherin, and vimentin. Increas-ing the levels of miR-200c and miR-141 in CD18-CR cellsdecreased ZEB1 mRNA and protein levels that were associatedwith an increase in E-cadherin and a decrease in vimentin mRNAand protein levels (Supplementary Fig. S4). In contrast with the

siCtrl siZEB1

ZEB1

E-cadherin

Vimentin

HSP90

181 kDa

120 kDa

54 kDa

90 kDa

CD18-CR

siZ

EB

1si

Ctr

l

CD

18-C

R

siCtrl siZEB1

siCtrl siZEB12D collagen migration

300

200

100

0

Rel

ativ

e m

igra

tion

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

ZEB1 ZEB1

Fol

d ch

ange

(gen

e/G

AP

DH

)

Fol

d ch

ange

(gen

e/G

AP

DH

)

Fol

d ch

ange

(gen

e/G

AP

DH

)

Fol

d ch

ange

(gen

e/G

AP

DH

)50 kDa

181 kDa

55 kDa

CD18-CR

CD18-CR

MNK1

MNK2

ZEB1

ZEB1

CD18-CR

DMSO

DMSO

DMSO

DMSO

CGP

CGP

CGP

CGP

CD18-CR AsPC1

181 kDa 181 kDa

90 kDa 90 kDa

ZEB1 ZEB1

HSP90 HSP90

CD18-CR AsPC1

siCtrl

siMNK1

siMNK2

siMNK1/

2

siCtrl

siMNK1

siMNK2

siMNK1/

2

siCtrl

siMNK1

siMNK2

siMNK1/

2

CD18-CR

siCtrl siMNK1 siMNK2 siMNK1/2

MNK1

ZEB1

Tubulin

OD

254

nm

Gradient depth Gradient depth

1.5

1.0

0.5

0.0Fol

d ch

ange

mR

NA

(p

olys

omal

/who

le c

ell)

DMSO CGP

CD18-CR

E-cadherin/P

halloidinD

AP

I

A B

C D

E

Figure 4.Targeting MNKs decreases the protein expression of ZEB1 without reducing ZEB1 mRNA levels. A and B, CD18-CR cells were transfected with controlsiRNA (siCtrl) or ZEB1-specific siRNA (siZEB1) for 72 hours, and the effect on E-cadherin and vimentin protein expression was determined by Westernblotting (A, left). CD18-CR cells growing on glass coverslips were transfected with siCtrl or siZEB1 for 72 hours. The cells were then stained for E-cadherin, actinusing phalloidin and DAPI to counterstain the nuclei (A, right). The transfected cells were plated onto thin-layer type I collagen matrix overlaid withcolloidal gold, allowed to migrate for 24 hours, and the tracks were photographed and quantified (B); �� , P < 0.01. C, CD18-CR and AsPC1 cells in 3D collagenwere treated with CGP57380 (2.5 mmol/L) for 72 hours and the effect on ZEB1 mRNA was determined by qRT-PCR. The effect on ZEB1 protein wasdetermined by Western blotting. D, CD18-CR cells were transfected with control siRNA (siCtrl), MNK1 siRNA (siMNK1), MNK2 siRNA (siMNK2), or acombination of MNK1- and MNK2-specific siRNAs (siMNK1/2) for 48 hours. The cells were then grown in 3D type I collagen (2 mg/mL) for additional 24 hours.The cells were then processed for MNK1, MNK2, and ZEB1 mRNA and protein expression by Western blotting and qRT-PCR. E, CD18-CR cells on tissueculture plastic were treated with CGP57380 (5 mmol/L) for 48 hours. The cells were subjected to hypotonic lysis followed by separation of equal amountsof supernatant, as measured by OD at 260 nm, on 10% to 50% sucrose gradients and the OD at 254 nm was recorded. The OD254 is shown as afunction of gradient depth. The levels of ZEB1 and GAPDH mRNA in the whole cell lysates and in the polysomal fractions were determined by qRT-PCR. Therelative amounts of ZEB1 mRNA in the polysomal fractions were compared with the relative amounts of ZEB1 mRNA in the whole-cell lysates. The resultsare representative of three independent experiments.






1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0siCtrl sielF4E

siCtrl siCtrlsielF4E

siCtrl sielF4E

elF4E

siCtrl sielF4E

sielF4E

siCtrl

siCtrl

siCtrl

siCtrl

siCtrl

sihnRNPA1

sihnRNPA1 sihnRNPA1

sihnRNPA1

sielF4E siCtrl sielF4E

siCtrl sielF4E siCtrl siMNK1/2

siCtrl siMNK1/2 siCtrl siMNK1/2

siCtrl siMNK1/2

3

2

1

0

5

4

3

2

1

0

3

2

1

0

CD18-CR

CD18-CR CD18-CR

CD18-CR

CD18-CR CD18-CR

ZEB1F

old

chan

ge(g

ene/

GA

PD

H)

Fol

d ch

ange

(gen

e/G

AP

DH

)F

old

chan

ge(m

iRN

A/R

NU

48)

Fol

d ch

ange

(miR

NA

/RN

U48

)

Fol

d ch

ange

(miR

NA

/RN

U48

)F

old

chan

ge(g

ene/

GA

PD

H)

Fol

d ch

ange

(gen

e/G

AP

DH

)

Fol

d ch

ange

(miR

NA

/RN

U48

)

E-cadherin Vimentin

E-cadherin Vimentin

miR-200b miR-200a miR-200b

miR-200c

miR-200a

miR-141miR-200c miR-141

CD18-CR

hnRNPA1 ZEB1

siCtrl

siCtrl

sihnR

NPA1

sielF

4E

25 kDa

36 kDa

181 kDa

90 kDa

120 kDa

54 kDa

90 kDa

elF4E

GAPDH

ZEB1

HSP90

E-cadherin

Vimentin

HSP90

Phalloidin/DAPI

siC

trl

siel

F4E

siM

NK

1/2

40 kDa34 kDa

181 kDa

120 kDa

54 kDa

90 kDa

hnRNPA1

ZEB1

E-Cadherin

Vimentin

HSP90

A

B

D

C

Kumar et al.





effect of eIF4E knockdown, MNK1/2 knockdown resulted inminimal effect on the levels of miR-200 miRNAs in CD18-CRcells (Fig. 5C).

Targeting the MNK effector hnRNPA1 increases ZEB1 proteinwithout increasing ZEB1 mRNA levels

As we have found that MNK knockdown and eIF4E knock-down have opposite effects on ZEB1 protein levels, it is possiblethat additional MNK effectors may regulate ZEB1 mRNA trans-lation. As hnRNPA1 was previously shown to function as atranslational repressor of TNFa mRNA downstream of MNKs(15), we evaluated the extent to which the MNK effectorhnRNPA1 regulated ZEB1 mRNA translation. We downregu-lated hnRNPA1 in CD18-CR and AsPC1 cells using siRNA andexamined the effect on ZEB1 protein and mRNA levels. Signif-icantly, downregulation of hnRNPA1 increased ZEB1 proteinlevels without affecting ZEB1 mRNA levels in both CD18-CR(Fig. 5D) and AsPC1 cells (Supplementary Fig. S5) that wasassociated with a decrease in E-cadherin and an increase invimentin mRNA and protein levels. These results indicate thatthe MNK effector hnRNPA1, similar to its regulation of TNFamRNA translation (15), functions as a translational repressor ofZEB1 mRNA in PDAC cells.

MNK1/2 inhibitors decrease growth of CD18-CR cells in 3Dcollagen and decrease Aldefluor(þ) cells

We next evaluated the effect of targeting MNKs on growth ofCD18-CR cells in collagen. Treatment with CGP57380 andsiRNA-mediated knockdown of MNK1/2 decreased the growthof CD18-CR cells in 3D collagen (Fig. 6A and B).

As chemoresistant cells can have increased numbers of cancerstem cells (34, 35), we evaluated the presence of Aldefluor(þ)cells in CD18-CR cells. Aldefluor(þ) cells have previously beenshown to be associated with cancer stem cells (36, 37). CD18-CRcells demonstrate increased numbers of Aldefluor(þ) cells com-pared with CD18 cells (Fig. 6C). Significantly, treatment with theMNK inhibitor CGP57380 decreased the number of Aldefluor(þ)cells in CD18-CR cells (Fig. 6D).

Treatment of human PDAC organoids with the MNK inhibitorCGP57380 increases E-cadherin mRNA levels and decreasesvimentin mRNA levels without affecting ZEB1 mRNA levels

Finally, to provide in vivo support for our findings that targetingMNKs can reverse EMT, we evaluated the effect of CGP57380 onhuman PDAC organoids. Because these organoids can accuratelymodel PDAC progression by recapitulating the key features of

human PDAC tumors (24), we established human PDAC orga-noids from deidentified PDAC specimens obtained on an IRB-approved protocol using the recently published Tuveson Labo-ratory protocol (24). Human PDAC specimens were minced,proteolytically digested, and then embedded in Matrigel to gen-erate human pancreatic organoids (Fig. 7A). Treatment of threedifferent PDACorganoidswith theMNK inhibitor CGP57380 didnot significantly affect ZEB1mRNA levels (Fig. 7B). Although wewere not able to analyze the effect on ZEB1 protein levels due toinsufficient amount of protein lysates, treatment of organoidswith CGP57380 increased E-cadherinmRNA levels and decreasedvimentin mRNA levels. In addition, CGP57380 inhibited growthof human PDAC organoids (Fig. 7C). These results suggest thatMNK inhibitors may be able to reverse EMT in human PDACtumors by targeting ZEB1 mRNA translation and may limitgrowth of PDAC tumors.

DiscussionElevated mRNA translation, especially of genes that regulate

key cellular processes, is an important feature of cancer cells(9, 10). Dysregulation of mRNA translation contributes to treat-ment resistance and tumor progression (17, 18). Moreover,increased eIF4E phosphorylation in human PDAC tumors isassociated with high-grade tumors and poor prognosis (9). Thus,there is increasing interest inunderstanding howenhancedmRNAtranslation contributes to PDAC progression. In this report, weshow that the collagenmicroenvironment promotesMNK-depen-dent eIF4E phosphorylation to regulate ZEB1 mRNA translationand EMT in PDAC cells.

We show that MNK inhibitors decrease migration of PDACcells. This is in agreement with a recent report demonstratingthat MNK inhibitors block migration of breast and oral cancercells (38). Although MNK inhibitors decrease vimentin proteinlevels without affecting vimentin mRNA levels in breast cancercells (38), we have found that pharmacologic and genetic target-ing ofMNK inhibitors decrease both vimentin protein andmRNAlevels in PDAC cells. Significantly, it was recently shown thatMNKs could regulate mRNA translation of EMT transcriptionfactors. For example, treatment of mouse NMuMG breast cellswith MNK inhibitors decreased TGFb-induced Snail proteinexpression without affecting Snail mRNA levels (39). We havefound that targeting MNKs decreases ZEB1 protein levels withoutaffectingZEB1mRNA levels. Significantly,we also show treatmentof human pancreatic organoids with MNK inhibitors does notaffect ZEB1 mRNA levels, but increases E-cadherin mRNA levels

Figure 5.Targeting eIF4E increases ZEB1 protein and mRNA levels and decreases ZEB1-targeting miR-200c and miR-141 microRNAs. A, CD18-CR cells weretransfected with control siRNA (siCtrl) or eIF4E-specific siRNA (sieIF4E) for 48 hours. The cells were then grown in 3D type I collagen (2 mg/mL) for additional24 hours. The cells were processed for eIF4E, ZEB1, E-cadherin, and vimentin mRNA expression by qRT-PCR and the lysates were analyzed for eIF4E, ZEB1,E-cadherin, and vimentin protein expression by Western blotting. CD18-CR cells growing on glass coverslips were transfected with siCtrl, sieIF4E, or acombination of MNK1- and MNK2-specific siRNAs (siMNK1/2) for 72 hours. The cells were then stained for actin using phalloidin and DAPI to counterstainthe nuclei. B, CD18-CR cells were transfected with siCtrl or sieIF4E for 48 hours. The cells were then grown in 3D type I collagen (2 mg/mL) foradditional 24 hours. The cells were processed for miR-200b, miR-200a, miR200c, and miR-141 microRNAs and normalized using RNU48 as internal control.C, CD18-CR cells were transfected with siCtrl or siMNK1/2 for 48 hours. The cells were then grown in 3D type I collagen (2 mg/mL) for additional 24 hours.The cells were then processed for miR-200b, miR-200a, miR200c, and miR-141 miRNAs and normalized using RNU48 as internal control. D, CD18-CRcells were transfected with control siRNA (siCtrl) or hnRNPA1-specific siRNA (sihnRNPA1) for 48 hours. The cells were then grown in 3D type I collagen(2 mg/mL) for additional 24 hours. The cells were processed for hnRNPA1, ZEB1, E-cadherin, and vimentinmRNA expression by qRT-PCR and the lysates wereanalyzed for hnRNPA1, ZEB1, E-cadherin, and vimentin protein expression by Western blotting. The results are representative of three independentexperiments. See also Supplementary Figs. S2–S5.






and decreases vimentin mRNA levels. Thus, our findings demon-strate that MNKs regulate EMT through modulation of ZEB1mRNA translation.

ZEB1 plays an important role in PDAC progression. ZEB1expression is increased in poorly differentiated human PDACtumors and in invasive cells arising from differentiated PDACtumors (40, 41). ZEB1 expression is increased in patients withearly recurrence compared with patients with long-term remis-sion following surgery (40). Increased ZEB1 expression is alsoassociated with chemoresistance in pancreatic cancer cells(42, 43). Consistent with our findings that PDAC cells devel-oping resistance to 5-FU have increased ZEB1 levels, others haveshown that PDAC cells developing resistance to gemcitabinechemotherapy demonstrate increased ZEB1 expression (42, 43).These chemoresistant cells have increased numbers of cancerstem cells that can be reduced by targeting ZEB1 (40). We showthat MNKs decrease the number of ALDH(þ) cells in pancreaticcancer, suggesting that targeting MNKs may enable targetingof PDAC cancer stem cells.

Paradoxically, eIF4E knockdown in PDAC cells resulted in amoremesenchymal phenotype, and increased vimentin andZEB1mRNAand protein expression.Our findings contrast with a recentreport demonstrating that targeting eIF4E in breast cancer cellsinhibited TGFb-induced vimentin and Snail expression (44). Weshow that the effect of eIF4E on ZEB1 levels in PDAC cells isthrough its effect on miR-200c and miR-141 miRNAs. We havefound that eIF4E knockdown, but not MNK1/2 knockdown,decreases miR-200c and miR-141 levels in PDAC cells. Signifi-cantly, miR-200c levels are reduced in lung and lymph nodemetastasis compared with primary human PDAC tumors (45).In addition, reduced levels of miR-200c are associated with worsesurvival after pancreatic cancer surgery (46). Consistent with ourfindings on the effect of eIF4E knockdown on ZEB1-miR-200c inPDAC cells, a previous report had demonstrated that chronictreatment of breast cancer cells with mTORC1 inhibitor inducedZEB1 expression through repression of miR-200c (47). As mTORinhibitors block eIF4E function by preventing phosphorylation-dependent release of 4E-BP (eIF4E binding protein) from eIF4E

DMSO CGP

Collagen

Collagen

CD

18-C

RC

D18

-CR

Sid

e sc

atte

rS

ide

scat

ter

siCtrl siMNK1/2

120

80

40

0

120

80

40

0

DMSO CGP

siCtrl siMNK1/2

Ave

rage

col

ony

siz

e (A

U)

Ave

rage

col

ony

siz

e (A

U)

CD18 CD18-CR

CD18-CR

+DEAB +DEAB–DEAB –DEAB

Aldefluor Aldefluor

Aldefluor

A

B

C

D

Figure 6.MNK1/2 inhibitors decrease growth ofCD18-CR cells in 3D collagen anddecrease Aldefluor(þ) cells. A, CD18-CR cells were grown in 3D type Icollagen (2 mg/mL) and fresh serum–

containing medium supplementedwith DMSO or CGP57380 (2.5 mmol/L)was added every other day for 5 days.The effect on colony size in 3D type Icollagen was examined by phase-contrast microscopy and size of theindividual colonies was measured; �� ,P < 0.01. B, CD18-CR cells weretransfectedwith control siRNA (siCtrl)or a combination of MNK1- and MNK2-specific siRNAs (siMNK1/2) for 48hours. The cells were then grown in 3Dtype I collagen (2 mg/mL) foradditional 72 hours. The effect oncolony size in 3D collagen wasexamined by phase-contrastmicroscopy and size of the individualcolonies was measured; � , P < 0.05. C,CD18 and CD18-CR cells wereprocessed for Aldefluor activity in thepresence or absence of aldehydedehydrogenase inhibitor DEAB byFACS analysis. D, CD18-CR cells weretreated with DMSO or CGP57380 (2.5mmol/L) for 48 hours and thenprocessed for Aldefluor activity byFACS analysis. The results arerepresentative of three independentexperiments.

Kumar et al.





(48, 49), it is possible that mTOR regulates ZEB1 through aneIF4E-dependent mechanism.

Although eIF4E knockdown decreased the levels of miR-200cand miR-141 miRNAs, eIF4E knockdown did not affect the

levels of miR-200b and miR-200a miRNAs, suggesting thatthe effect of eIF4E knockdown is specific for miR-200c andmiR-141 miRNAs. It is known that miR-200b and miR-200aare part of a cluster on chromosome 1, whereas miR-200c and

Human PDAC tumor

Human PDAC organoid

Human PDAC organoid

DMSO CGP

DMSO CGP

ZEB11.5

1.0

0.5

0.0

P = 0.63

P = 0.028

P = 0.046

E-cadherin8

6

4

2

0

1.5

1.0

0.5

0.0




T1 T2 T3

T1 T2 T3

T1 T2 T3

Vimentin

8

6

4

2

0

Fo

ld c

han

ge

(g

ene/

GA

PD

H)

Fo

ld c

han

ge

(g

ene/

GA

PD

H)

Fo

ld c

han

ge

(g

ene/

GA

PD

H)

Ave

rag

e O

rgan

oid

siz

e (A

U)

n.d.

A

C

D

B

Figure 7.Treatment of human PDAC organoidswith the MNK inhibitor CGP57380increases E-cadherin and decreasesvimentin levels without affecting ZEB1mRNA levels. A, pancreatic cancerorganoids were established from a de-identified human PDAC tumorspecimen using the published TuvesonLaboratory protocol as detailed inMaterials and Methods. B, threedifferent pancreatic cancer organoidswere treated with DMSO or CGP57380(2.5 mmol/L) for 96 hours and theeffects on ZEB1, E-cadherin, andvimentinmRNA levelswere determinedby qRT-PCR. n.d., not detected. The pvalues were calculated using paired ttest. C, pancreatic cancer organoidswere treated with DMSO or CGP57380(2.5 mmol/L) for 7 days, and the effecton pancreatic organoid size wasexamined by phase-contrastmicroscopy and size of the individualorganoidswasmeasured. Shownhere isbox plot with whiskers representing10th to 90th percentiles; �� , P < 0.001.D, the model depicts differentialregulation of ZEB1 by MNKs and eIF4E:Targeting MNKs attenuates EMT byinhibiting the translation of ZEB1 inpancreatic cancer cells, whiledownregulating theMNKeffector eIF4Eincreases ZEB1 through repression ofZEB1-targeting microRNAs miR-200cand miR-141. Note that although wedemonstrate that hnRNPA1 functions asa translational repressor of ZEB1mRNAin PDAC cells, Buxade and colleagues(15) have previously shown that MNK-mediated phosphorylation of hnRNPA1relieves its repression of mRNAtranslation.






miR-141 belong to a cluster on chromosome 12 (32, 33).Significantly, it has been shown that these two clusters can bedifferentially regulated, resulting in differential expression ofmiR-200b-200a and miR-200c-141 primary miRNA (pri-miRNA) transcripts. For example, in HMLE breast cells, poly-comb group proteins EZH2 and SUZ12 associate with andsilence miR-200b-200a-429 cluster, but do not silence themiR-200c-141 cluster (33). However, we have found that eIF4Eknockdown in pancreatic cancer cells did not affect the levels ofpri-miRNA transcripts of miR-200b, miR-200a, or miR-200c-141, suggesting that the differential effects of eIF4E knockdownon mature miR-200 miRNAs is not at the level of transcriptionof these miRNAs. Moreover, eIF4E knockdown also did notaffect the levels of miRNA-processing proteins, Drosha andDicer (Supplementary Fig. S3). Future experiments will deter-mine how eIF4E knockdown preferentially decreases the levelsof miR-200c-141 miRNAs in pancreatic cancer cells.

Although eIF4E is the best-characterized phosphorylation tar-get of MNKs, an additional MNK effector is hnRNPA1, a nuclearprotein that is involved in regulation of alternative splicing,mRNA export, and mRNA translation (50, 51). Previously, it wasshown that MNK-mediated phosphorylation of hnRNPA1decreases binding of hnRNPA1 to TNFa mRNA and therebyincreases translation of TNFa mRNA (15), demonstrating thathnRNPA1 functions as a translational repressor of TNFa mRNA.The hnRNPA1 protein has affinity for AU-rich sequences, inparticular the AUUUA sequence, which is contained withinthe 30-untranslated region (30-UTR) of eukaryotic genes (52).Importantly, ZEB1 has 9 AUUUA sequences in its 30-UTR(AREsite). Significantly, we have found that knockdown ofhnRNPA1 using siRNA increases ZEB1 protein levels withoutaffecting ZEB1 mRNA levels, indicating that the MNK effectorhnRNPA1 also functions as a translational repressor of ZEB1mRNA in PDAC cells.

Overall, we demonstrate that MNKs play an importantrole in PDAC progression. Targeting MNKs can reverse EMT,

decrease migration, limit growth of PDAC cells, and mayreduce pancreatic cancer stem cells. We also demonstrate thattargeting MNKs, in contrast with targeting eIF4E, does notaffect ZEB1-targeting miRNAs (Fig. 7D, model). Thus, MNKsmay be particular appropriate targets to consider in the treat-ment of this deadly cancer.

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

Authors' ContributionsConception and design: F.D. Eckerdt, L.C. Platanias, H.G. MunshiDevelopment of methodology: K. Kumar, H.G. MunshiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): K. Kumar, C.R. Chow, A.D. Arslan, B. Kwok, D.J.Bentrem, H.G. MunshiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): K. Kumar, C.R. Chow, A.D. Arslan, B. Kwok, F.D.Eckerdt, L.C. Platanias, H.G. MunshiWriting, review, and/or revision of themanuscript: K. Kumar, C.R. Chow, F.D.Eckerdt, L.C. Platanias, H.G. MunshiAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): K. Kumar, K. Ebine, D.J. Bentrem, H.G. MunshiStudy supervision: H.G. Munshi

Grant SupportThis work was supported by grant R01CA186885 (to H.G. Munshi) and

R01CA121192 and R01CA155566 (to L.C. Platanias) from the NCI andMerit awards I01BX001363 (to H.G. Munshi) and I01CX000916 (to L.C.Platanias) from the Department of Veterans Affairs. This work was alsosupported by the training grant T32CA070085 (to C.R. Chow and A.D.Arslan) from the NCI.

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 June 25, 2015; revised November 6, 2015; accepted November 15,2015; published OnlineFirst November 25, 2015.

References1. Ryan DP, Hong TS, Bardeesy N. Pancreatic adenocarcinoma. N Engl J Med

2014;371:1039–49.2. Garrido-Laguna I, Hidalgo M. Pancreatic cancer: from state-of-the-art

treatments to promising novel therapies. Nat Rev Clin Oncol 2015;12:319–34.

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

4. Whatcott CJ, Diep CH, Jiang P, Watanabe A, LoBello J, Sima C, et al.Desmoplasia in primary tumors and metastatic lesions of pancreaticcancer. Clin Cancer Res 2015;21:3561–8

5. Maitra A, Hruban RH. Pancreatic cancer. Annu Rev Pathol 2008;3:157–88.6. Ottaviano AJ, Sun L, Ananthanarayanan V, Munshi HG. Extracellular

matrix-mediated membrane-type 1 matrix metalloproteinase expressionin pancreatic ductal cells is regulated by transforming growth factor-beta1.Cancer Res 2006;66:7032–40.

7. Shields MA, Krantz SB, Bentrem DJ, Dangi-Garimella S, Munshi HG.Interplay between beta1-integrin and Rho signaling regulates differentialscattering andmotility of pancreatic cancer cells by snail and Slug proteins.J Biol Chem 2012;287:6218–29.

8. Shields MA, Dangi-Garimella S, Krantz SB, Bentrem DJ, Munshi HG.Pancreatic cancer cells respond to type I collagen by inducingsnail expression to promote membrane type 1 matrix metallopro-teinase-dependent collagen invasion. J Biol Chem 2011;286:10495–504.

9. Adesso L, Calabretta S, Barbagallo F, Capurso G, Pilozzi E, Geremia R, et al.Gemcitabine triggers a pro-survival response in pancreatic cancer cellsthrough activation of the MNK2/eIF4E pathway. Oncogene 2013;32:2848–57.

10. Martineau Y, Azar R, Muller D, Lasfargues C, El Khawand S, Anesia R, et al.Pancreatic tumours escape from translational control through 4E-BP1 loss.Oncogene 2014;33:1367–74.

11. Wendel HG, Silva RL, Malina A, Mills JR, Zhu H, Ueda T, et al. DissectingeIF4E action in tumorigenesis. Genes Dev 2007;21:3232–7.

12. Furic L, Rong L, Larsson O, Koumakpayi IH, Yoshida K, Brueschke A, et al.eIF4E phosphorylation promotes tumorigenesis and is associated withprostate cancer progression. Proc Natl Acad Sci U S A 2010;107:14134–9.

13. Joshi S, Platanias LC. Mnk kinase pathway: cellular functions and biolog-ical outcomes. World J Biol Chem 2014;5:321–33.

14. Proud CG. Mnks, eIF4E phosphorylation and cancer. Biochim BiophysActa 2015;1849:766–73

15. Buxade M, Parra JL, Rousseau S, Shpiro N, Marquez R, Morrice N, et al. TheMnks are novel components in the control of TNF alpha biosynthesis andphosphorylate and regulate hnRNP A1. Immunity 2005;23:177–89.

16. Buxade M, Parra-Palau JL, Proud CG. The Mnks: MAP kinase-interactingkinases (MAP kinase signal-integrating kinases). Front Biosci 2008;13:5359–73.

17. Pelletier J, Graff J, Ruggero D, Sonenberg N. Targeting the eIF4F translationinitiation complex: a critical nexus for cancer development. Cancer Res2015;75:250–63.

Kumar et al.





18. Bhat M, Robichaud N, Hulea L, Sonenberg N, Pelletier J, Topisirovic I.Targeting the translation machinery in cancer. Nat Rev Drug Discov2015;14:261–78.

19. Sahai V, Kumar K, Knab LM, Chow CR, Raza SS, Bentrem DJ, et al. BETbromodomain inhibitors block growth of pancreatic cancer cells in three-dimensional collagen. Mol Cancer Ther 2014;13:1907–17.

20. Kumar K, Raza SS, Knab LM, Chow CR, Kwok B, Bentrem DJ, et al. GLI2-dependent c-MYC upregulation mediates resistance of pancreatic cancercells to the BET bromodomain inhibitor JQ1. Sci Rep 2015;5:9489.

21. Dangi-Garimella S, Strouch MJ, Grippo PJ, Bentrem DJ, Munshi HG.Collagen regulation of let-7 in pancreatic cancer involves TGF-beta1-mediated membrane type 1-matrix metalloproteinase expression. Onco-gene 2011;30:1002–8.

22. Kaur S, Sassano A, Dolniak B, Joshi S, Majchrzak-Kita B, Baker DP, et al.Role of the Akt pathway in mRNA translation of interferon-stimulatedgenes. Proc Natl Acad Sci U S A 2008;105:4808–13.

23. Joshi S, Kaur S, Redig AJ, Goldsborough K, David K, Ueda T, et al. Type Iinterferon (IFN)-dependent activation of Mnk1 and its role in the gener-ation of growth inhibitory responses. Proc Natl Acad Sci U S A 2009;106:12097–102.

24. Boj SF, Hwang CI, Baker LA, Chio II, Engle DD, Corbo V, et al. Organoidmodels of human and mouse ductal pancreatic cancer. Cell 2015;160:324–38.

25. Grzmil M, Huber RM, Hess D, Frank S, Hynx D, Moncayo G, et al. MNK1pathway activity maintains protein synthesis in rapalog-treated gliomas.J Clin Invest 2014;124:742–54.

26. Altman JK, Szilard A, Konicek BW, Iversen PW, Kroczynska B, Glaser H,et al. Inhibition of Mnk kinase activity by cercosporamide and sup-pressive effects on acute myeloid leukemia precursors. Blood 2013;121:3675–81.

27. Krantz SB, Shields MA, Dangi-Garimella S, Munshi HG, Bentrem DJ.Contribution of epithelial-to-mesenchymal transition and cancer stemcells to pancreatic cancer progression. J Surg Res 2012;173:105–12.

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

29. Thiery JP, Acloque H, Huang RY, Nieto MA. Epithelial–mesenchymaltransitions in development and disease. Cell 2009;139:871–90.

30. Zhang P, Sun Y, Ma L. ZEB1: at the crossroads of epithelial-mesenchymaltransition, metastasis and therapy resistance. Cell Cycle 2015;14:481–7.

31. Brabletz S, Brabletz T. The ZEB/miR-200 feedback loop—a motor ofcellular plasticity in development and cancer? EMBO Rep 2010;11:670–7.

32. Gregory PA, Bert AG, Paterson EL, Barry SC, Tsykin A, Farshid G, et al. ThemiR-200 family and miR-205 regulate epithelial to mesenchymal transi-tion by targeting ZEB1 and SIP1. Nat Cell Biol 2008;10:593–601.

33. LimYY,Wright JA, Attema JL,Gregory PA, Bert AG, Smith E, et al. Epigeneticmodulation of the miR-200 family is associated with transition to a breastcancer stem-cell-like state. J Cell Sci 2013;126:2256–66.

34. Izumiya M, Kabashima A, Higuchi H, Igarashi T, Sakai G, Iizuka H, et al.Chemoresistance is associated with cancer stem cell-like properties andepithelial-to-mesenchymal transition in pancreatic cancer cells. AnticancerRes 2012;32:3847–53.

35. Yamashina T, Baghdadi M, Yoneda A, Kinoshita I, Suzu S, Dosaka-Akita H,et al. Cancer stem-like cells derived from chemoresistant tumors have a

unique capacity to prime tumorigenic myeloid cells. Cancer Res 2014;74:2698–709.

36. Charafe-Jauffret E, Ginestier C, Iovino F, Wicinski J, Cervera N, Finetti P,et al. Breast cancer cell lines contain functional cancer stem cells withmetastatic capacity and a distinct molecular signature. Cancer Res 2009;69:1302–13.

37. Kolev VN,Wright QG, Vidal CM, Ring JE, Shapiro IM, Ricono J, et al. PI3K/mTOR dual inhibitor VS-5584 preferentially targets cancer stem cells.Cancer Res 2015;75:446–55.

38. Beggs JE, Tian S, Jones GG, Xie J, Iadevaia V, Jenei V, et al. The MAP kinase-interacting kinases regulate cellmigration, vimentin expression, and eIF4E/CYFIP1 binding. Biochem J 2015;467:63–76.

39. Robichaud N, Del Rincon SV, Huor B, Alain T, Petruccelli LA, Hearnden J,et al. Phosphorylation of eIF4E promotes EMT and metastasis via trans-lational control of SNAIL and MMP-3. Oncogene 2015;34:2032–42.

40. WellnerU, Schubert J, BurkUC, SchmalhoferO, ZhuF, SonntagA, et al. TheEMT-activator ZEB1 promotes tumorigenicity by repressing stemness-inhibiting microRNAs. Nat Cell Biol 2009;11:1487–95.

41. Bronsert P, Kohler I, Timme S, Kiefer S, Werner M, Schilling O, et al.Prognostic significance of Zinc finger E-box binding homeobox 1 (ZEB1)expression in cancer cells and cancer-associated fibroblasts in pancreatichead cancer. Surgery 2014;156:97–108.

42. ArumugamT,RamachandranV, Fournier KF,WangH,Marquis L, Abbruzz-ese JL, et al. Epithelial-to-mesenchymal transition contributes to drugresistance in pancreatic cancer. Cancer Res 2009;69:5820–8.

43. Wang Z, Li Y, Kong D, Banerjee S, Ahmad A, Azmi AS, et al. Acquisition ofepithelial-mesenchymal transition phenotype of gemcitabine-resistantpancreatic cancer cells is linked with activation of the notch signalingpathway. Cancer Res 2009;69:2400–7.

44. Pettersson F, Del Rincon SV, Emond A, Huor B, Ngan E, Ng J, et al. Geneticand pharmacologic inhibition of eIF4E reduces breast cancer cell migra-tion, invasion, and metastasis. Cancer Res 2015;75:1102–12.

45. Mohr AM, Bailey JM, Lewallen ME, Liu X, Radhakrishnan P, Yu F, et al.MUC1 regulates expression of multiple microRNAs involved in pancreatictumor progression, including the miR-200c/141 cluster. PLoS ONE 2013;8:e73306.

46. Yu J, Ohuchida K, Mizumoto K, Sato N, Kayashima T, Fujita H, et al.MicroRNA, hsa-miR-200c, is an independent prognostic factor in pancre-atic cancer and its upregulation inhibits pancreatic cancer invasion butincreases cell proliferation. Mol Cancer 2010;9:169.

47. Mikaelian I, Malek M, Gadet R, Viallet J, Garcia A, Girard-Gagnepain A,et al. Genetic and pharmacologic inhibition ofmTORC1promotes EMT bya TGF-beta-independent mechanism. Cancer Res 2013;73:6621–31.

48. Laplante M, Sabatini DM. mTOR signaling in growth control and disease.Cell 2012;149:274–93.

49. Beauchamp EM, Platanias LC. The evolution of the TOR pathway and itsrole in cancer. Oncogene 2013;32:3923–32.

50. Mayeda A, Krainer AR. Regulation of alternative pre-mRNA splicing byhnRNP A1 and splicing factor SF2. Cell 1992;68:365–75.

51. Izaurralde E, Jarmolowski A, Beisel C, Mattaj IW, Dreyfuss G, Fischer U.A role for the M9 transport signal of hnRNP A1 in mRNA nuclear export.J Cell Biol 1997;137:27–35.

52. Jean-Philippe J, Paz S, Caputi M. hnRNP A1: the Swiss army knife of geneexpression. Int J Mol Sci 2013;14:18999–9024.






2016;14:216-227. Published OnlineFirst November 25, 2015.Mol Cancer Res Krishan Kumar, Christina R. Chow, Kazumi Ebine, et al. Protein Kinases (MNK) and eIF4E in Pancreatic CancerDifferential Regulation of ZEB1 and EMT by MAPK-Interacting

Updated version

10.1158/1541-7786.MCR-15-0285doi:

Access the most recent version of this article at:

Material

Supplementary

http://mcr.aacrjournals.org/content/suppl/2015/12/15/1541-7786.MCR-15-0285.DC1

Access the most recent supplemental material at:

Cited articles

http://mcr.aacrjournals.org/content/14/2/216.full#ref-list-1

This article cites 52 articles, 24 of which you can access for free at:

Citing articles

http://mcr.aacrjournals.org/content/14/2/216.full#related-urls

This article has been cited by 3 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://mcr.aacrjournals.org/content/14/2/216To request permission to re-use all or part of this article, use this link



http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-15-0285

http://mcr.aacrjournals.org/content/suppl/2015/12/15/1541-7786.MCR-15-0285.DC1

http://mcr.aacrjournals.org/content/14/2/216.full#ref-list-1

http://mcr.aacrjournals.org/content/14/2/216.full#related-urls

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

mailto:[email protected]

http://mcr.aacrjournals.org/content/14/2/216


Documents

Differential Regulation of ZEB1 and EMT by MAPK ... · Krishan Kumar1,2, Christina R. Chow1,3, Kazumi Ebine1,2, Ahmet D. Arslan1,3, ... exhibit increased activation of the MAPK-interacting