eﬂornithine (dfmo) prevents progression of pancreatic ...€¦ · research article eﬂornithine...

Research Article

Eflornithine (DFMO) Prevents Progression of PancreaticCancer by Modulating Ornithine Decarboxylase Signaling

Altaf Mohammed1, Naveena B. Janakiram1, Venkateshwar Madka1, Rebekah L. Ritchie1, Misty Brewer1,Laura Biddick1, Jagan Mohan R. Patlolla1, Michael Sadeghi1, Stan Lightfoot1, Vernon E. Steele2, andChinthalapally V. Rao1

AbstractOrnithine decarboxylase (ODC) is the key rate-limiting enzyme in the polyamine synthesis pathway and

it is overexpressed in a variety of cancers. We found that polyamine synthesis and modulation of ODC

signaling occurs at early stages of pancreatic precursor lesions and increases as the tumor progresses in Kras-

activated p48Cre/þ-LSL-KrasG12D/þ mice. Interest in use of the ODC inhibitor eflornithine (DFMO) as a

cancer chemopreventive agent has increased in recent years sinceODCwas shown tobe transactivated by the

c-myc oncogene and to cooperate with the ras oncogene in malignant transformation of epithelial tissues.

We tested the effects of DFMO on pancreatic intraepithelial neoplasias (PanIN) and their progression to

pancreatic ductal adenocarcinoma (PDAC) in genetically engineered Kras mice. The KrasG12D/þ mice fed

DFMO at 0.1% and 0.2% in the diet showed a significant inhibition (P < 0.0001) of PDAC incidence

compared with mice fed control diet. Pancreatic tumor weights were decreased by 31% to 43% (P < 0.03–

0.001)with both doses ofDFMO.DFMOat 0.1%and0.2%caused a significant suppression (27%and31%;

P < 0.02–0.004) of PanIN 3 lesions (carcinoma in situ). DFMO-treated pancreas exhibitedmodulated ODC

pathway components along with decreased proliferation and increased expression of p21/p27 as compared

with pancreatic tissues derived from mice fed control diet. In summary, our preclinical data indicate that

DFMO has potential for chemoprevention of pancreatic cancer and should be evaluated in other PDAC

models and in combination with other drugs in anticipation of future clinical trials. Cancer Prev Res; 7(12);

1198–209. �2014 AACR.

IntroductionPancreatic cancer is the fourth leading cause of cancer-

related mortality in both men and women in the UnitedStates (1) and the seventh leading cause of cancer-relateddeaths worldwide. Despite tremendous scientific effort forover three decades, pancreatic cancer still remains an almostuniformly lethal devastating disease with <6% 5-year sur-vival. The incidence of pancreatic cancer has been increasingover the past 10 years. In 2013, about 46,420 people(23,530 men and 22,890 women) were expected to bediagnosed with pancreatic cancer and 39,590 people

(20,170 men and 19,420 women) were expected to die ofpancreatic cancer in the United States (1). Worldwide,338,000 people were diagnosed with pancreatic cancer in2012 (2). The Pancreatic Action Network estimates that by2020, pancreatic cancer will become the second leadingcause of cancer-related deaths in the United States. Lack ofearly diagnosis and effective interventions are the majorfactors in the poor prognosis and dismal survival rates(3–5). First-line chemotherapeutic agents for pancreaticcancer treatment provide a marginal survival benefit. Sofar, a range of targeted therapies has failed to improvesurvival significantly in many clinical trials. Identifyingsuccessful intervention strategies using preclinical modelsthat mimic human pancreatic cancer progression wouldhelp in the development of novel agents for pancreaticcancer prevention and treatment. The pancreatic cancerprecursors, pancreatic intraepithelial neoplasias (PanIN),progress slowly overmany years to development of invasivepancreatic cancer (3–5). Hence, it is important to developnovel strategies to delay or inhibit progression of PanIN topancreatic cancer. The K-rasG12D genetically engineeredmice (GEM) is an excellent pancreatic cancer model thatshows development of lesions closely resembling humanPanINs with progression to pancreatic ductal adenocarci-noma (PDAC) and it has been used successfully for che-moprevention studies.

1Center for Cancer Prevention and Drug Development, Department ofMedicine, Hem-Onc Section, University of Oklahoma Health SciencesCenter, Oklahoma City, Oklahoma. 2Division of Cancer Prevention, Che-mopreventive Agent Development Research Group, National Cancer Insti-tute, Bethesda, Maryland.

Note:Supplementary data for this article are available atCancer PreventionResearch Online (http://cancerprevres.aacrjournals.org/).

Corresponding Authors: Chinthalapally V. Rao, Center for Cancer Pre-vention and Drug Development, 975 NE 10th Street, BRC II, Room 1203,University ofOklahomaHealthSciencesCenter,OklahomaCity,OK73104.Phone: 405-271-3224; Fax: 405-271-3225; E-mail: [email protected];and Altaf Mohammed, [email protected]

doi: 10.1158/1940-6207.CAPR-14-0176

�2014 American Association for Cancer Research.

CancerPreventionResearch

Cancer Prev Res; 7(12) December 20141198

Research. on June 8, 2020. © 2014 American Association for Cancercancerpreventionresearch.aacrjournals.org Downloaded from

Published OnlineFirst September 23, 2014; DOI: 10.1158/1940-6207.CAPR-14-0176

http://cancerpreventionresearch.aacrjournals.org/

Activation of ornithine decarboxylase (ODC), the keyregulatory enzyme in polyamine biosynthesis, and theconsequent increase in concentrations of polyamines(putresine, spermine, and spermidine) are associated withtumor promotion and progression. ODC is modulated byimportant genes of the polyamine biosynthesis pathway,such as increased arginase (Arg1) and ornithine amino-transferase (Oat), and decreased ODC antizyme (Oaz) andspermidine/spermine N(1)-acetyltransferase (Sat1). A linkbetween polyamine levels and tumor grade and stage is wellestablished in colorectal cancers (6, 7). Elevated levels ofODC and polyamines have also been found in pancreaticcancer (8). Pancreatic adenocarcinoma cell lines have beenshown previously to have elevated ODC activities (9).Although it is clear that ODC overexpression is associatedwith advanced pancreatic cancer development, it has notbeen established whether overexpression of ODC isinvolved in the initiation or progression of pancreaticcarcinogenesis.Clinical and preclinical studies have demonstrated that

eflornithine (DFMO), an ODC inhibitor, is a potentialchemopreventive agent for several cancers (10–16). How-ever, expression of ODC during progression of pancreaticcancer and the chemopreventive efficacy of DFMO in apreclinical model that recapitulates human pancreatic can-cer have not yet been evaluated. The rationale for the use ofDFMO as a cancer chemopreventive agent has beenstrengthened in recent years because ODC has been shownto be transactivated by the c-myc oncogene in various celland tissue types and to cooperate with the ras oncogene inmalignant transformation of epithelial tissues. To addressthe role of ODC in pancreatic carcinogenesis, we firstcompared ODC activity in pancreatic tumor tissue fromnormal mice and that from 2-, 6-, and 10-month-oldp48Cre/þ-LSL-KrasG12D/þ mice. We confirmed overexpres-sion ofODC in pancreatic cancer tissues as themice age andthen tested the ODC inhibitor DFMO as a chemopreven-tive. DFMO inhibited tumor progression and modulatedseveral components of ODC signaling and cell prolifera-tion, supporting future testing of DFMO in clinical trials forpancreatic cancer prevention.

Materials and MethodsMouse model, diet, and handlingGeneration of p48Cre/þ; LSL-KrasG12D/þ mice expressing

the activated KrasG12D oncogene has been described previ-ously (17). All animal research was performed under theanimal protocols approved by the University of OklahomaHealth Sciences Center (OUHSC; Oklahoma City, OK)Institutional Animal Care and Use Committee (IACUC).Animals were housed in ventilated cages under standard-ized conditions (21�C, 60% humidity, 12-hour light/12-hour dark cycle, 20 air changes/h) in the OUHSC rodentbarrier facility. Semipurified modified AIN-76A diet ingre-dients were purchased from Bioserv, Inc. DFMO wasprocured from the NCI-DCP chemoprevention drug repos-itory. DFMO (0.1% and 0.2%) was premixed with smallquantities of casein and then blended into the diet using a

Hobart Mixer. The rationale for selecting 0.2% DFMO wasbased on earlier experiments; this dose is equivalent to�40% of the maximum-tolerated dose when it is given insemipurifiedAIN-76Adiet. The lower dose (0.1%) chosen iscomparablewith doses ofDFMOused in clinical trials. Bothcontrol and experimental diets were prepared weekly andstored in the cold room.Mice were allowed ad libitum accessto the respective diets and to automated tap water purifiedby reverse osmosis.

Breeding and genotyping analysisLSL-KrasG12D/þ and p48Cre/þ mice were maintained in a

C57BL/6 heterozygous genetic background and bred. Off-springs of activated p48Cre/þ.LSL-KrasG12D/þ and C5BL/6wild-type mice were generated at required quantities. Thegenotype of each pupwas confirmed by tail DNA extractionand polymerase chain reaction (PCR). Briefly, genomicDNA was isolated from tail tissue samples using the Mini-prep Kit (Invitrogen). PCR was performed for K-ras and Cregenes using the following conditions: denaturation at 95�Cfor 5 minutes, followed by 35 cycles at 95�C for 1 minute,60�C for 1 minute and 72�C for 1 minute. Oligonucleotideprimer sequences used were as follows: K-ras 50-CCTTTA-CAAGCGCACGCAGAG-30 sense, 50-AGCTAGCCACCATG-GCTTGAGTAAGTCTGCA-30 antisense; and p48Cre 50-AC-CGTCAGTACGTGAGATATCTT-30 sense and 5-ACCTGAA-GATGTTCGCGATTATCT-30 antisense. PCR products wereseparated on a 2% agarose gel. Successful recombinationyields were 550- and 350-bp products for Kras and Cregenes, respectively (17).

ODC assayp48Cre/þ-LSL-KrasG12D/þ GEM (n¼ 6/group) at 2, 6, and

10 months of age, along with wild-type mice, were killedand pancreata were snap-frozen in liquid nitrogen forfurther analysis. ODC activity in the pancreata of thesemice was measured as release of CO2 from L-[1-14C]-orni-thine. Briefly, pancreas samples were trypsinized, washedwith phosphate-buffered saline (PBS), and lysed in 100 mLof ODC lysis buffer [25 mmol/L Tris–HCl (pH 7.4), 0.1mmol/L EDTA, 0.1% Triton X-100, 0.1 mmol/L pyridoxyl-5-phosphate, 1 mmol/L dithiothreitol (DTT), and 1� pro-tease inhibitors]. Lysates were frozen at �80�C, thawed onice, and centrifuged at 12,000 � g for 20 minutes at 4�C toremove cell debris. Supernatants were transferred to newEppendorf tubes, and 100 mL (5 mg/mL protein) aliquotswere transferred to stoppered glass vials containing 10 mL(55 nmol) of L-[1-14C]ornithine hydrochloride (ThermoScientific) in triplicate. ODC assay buffer (110 mL) contain-ing [1-14C]ornithine hydrochloride was added and incubat-ed for 1 hour at 37�C. Reactions were terminated with anequal volume of 10% trichloroacetic acid (TCA) injectedwith a syringe through the rubber stopper top and left foranother 20 minutes. Released 14CO2 was captured with afilter paper, whichwas soakedwith 80mL sodiumhydroxideand inserted through a hole in the stopper top (KontesGlass). The filter paper was transferred to a scintillation vial,5 mL of scintillation fluid was added, and liquid

DFMO Prevents Pancreatic Cancer In Vivo

www.aacrjournals.org Cancer Prev Res; 7(12) December 2014 1199




scintillation counting was done for 1 minute. ODC activitywas determined as picomoles of CO2 released (calculatedfrom cpm) per milligram of protein per hour.

Preclinical efficacy assayMale and female p48Cre/þ-LSL-KrasG12D/þmicewereused

in the efficacy study. Briefly, 5-week-old mice were selectedand randomized so that average bodyweights in each groupwere equal (n ¼ 30–34/group for p48Cre/þ-LSL-KrasG12D/þ

mice and n ¼ 12/group for C57BL/6 wild-type mice) andwere fed AIN-76A diet for 1 week. At 6 weeks of age, micewere fed AIN-76A experimental diets containing 0%, 0.1%,or 0.2% DFMO in the diet until termination of the study.Mice were checked routinely for signs of weight loss or anysigns of toxicity or any abnormalities. Body weight of eachanimal was measured once weekly for the first 6 weeks andthen once a month until termination. After 265 days (38weeks) on experimental diets, all mice were euthanized byCO2 asphyxiation at 44 weeks (�300 days) of age andnecropsied; pancreata were collected from all groups,weighed, and snap-frozen in liquid nitrogen for furtheranalysis. Pancreata required for histopathology and immu-nohistochemistry (IHC) to identify PanIN lesions andPDAC and for evaluation of various molecular markerswere fixed (head to tail) in 10% neutral-buffered formalinas previously described (17–19).

Analysis of PanIN lesions and PDACNormal and tumor pancreatic tissues were fixed in 10%

neutral-buffered formalin for 24 to 48 hours and thenprocessed and embedded in paraffin according to standardprotocols. Tissue sections (4 mm) of each pancreas stainedwith hematoxylin and eosin (H&E) were evaluated histo-logically by a pathologist blinded to the experimentalgroups. PanIN lesions and carcinomas were classifiedaccording to histopathology criteria as previously described(17–19). To quantify the progression of PanIN lesions, thetotal number of ductal lesions and their grades were deter-mined. The relative proportion of each PanIN lesion gradeto the overall number of analyzed ducts was recorded foreach animal. Similarly, pancreatic carcinoma and normalappearing pancreatic tissue were evaluated for all the ani-mals. The incidence of PDAC is defined as the percentage ofmice with PDAC as previously described by us (17–19).Briefly, we evaluate pancreas sections derived from top,middle, and bottom portion of the tissue block (from headto tail) from the GEM mice (N ¼ 30–34 mice/group maleand female). The efficacy endpoints used in this study wereinhibition of PanINs and PDAC. The criterion for invasivedisease incidence is groups of adenocarcinoma cells (at least3%) invading with adenomatous stroma streaming aroundthe cells.

IHC and immunofluorescenceFixed sections (5-mm) were incubated with primary anti-

bodies in a hybridization chamber for 1 hour at roomtemperature or overnight at 4�C. The primary antibodiesfor proliferating cell nuclear antigen (PCNA), Cav-1, b-cate-

nin, and p21, were from Santa Cruz Biotechnology; thosefor Ki67 and cyclin D1 were from Cell Signaling Technol-ogy; those for p16 were from LS Bio; and those for Rb werefrom Abcam. Following incubations with primary anti-body, sections were incubated for 1 hour with anti-mouseor anti-rabbit secondary antibodies, as appropriate for eachprimary, then visualized with diaminobenzidine (DAB)and counterstained with hematoxylin for IHC and counter-stained with DAPI (40,6-diamidino-2-phenylindole) forimmunohistofluorescence (IHF). Slides were observedunder an Olympus microscope 1 � 701 and digital com-puter imageswere recordedwith anOlympusDP70 camera.

Quantitative real-time PCR analysisTotal RNA from pancreatic samples (consisting of nor-

mal, PanIN, and PDAC) was extracted using an RNA Kit forisolation of total cellular RNA (Invitrogen) as per themanufacturer’s instructions. Equal quantities of DNA-freeRNAwere used in reverse transcription reactions formakingcDNA using SuperScript reverse transcriptase (Invitrogen).The real-time PCR was carried out in a 25-mL reactionvolume using 3 mL of a 1:10 cDNA dilution containingSYBR Green master mix (Bio-Rad) and primers for PCNA,caveolin-1 (Cav-1),Oaz,Oat, Sat1, Arg1, spermine synthase(Sms), and spermidine synthase (Srm; SupplementaryTable S1). All PCRs were done in a Bio-Rad iCycler iQreal-time PCR detection system and fluorescence thresholdvalues (Ct) were calculated. Relative mRNA levels wereassessed by standardization to actin. Results are expressedas a fold difference in gene expression.

Cell linesThe human MiaPaCa 2 cancer cell line was obtained

from the American Type Culture Collection (ATCC) andmaintained in Dulbecco’s Modified Eagle Medium(DMEM; Life Technologies Inc.) supplemented with10% fetal bovine serum and 1% penicillin–streptomycinin a humidified chamber at 37�C and 5% CO2. The cellline was authenticated by short-tandem repeat analysis bythe ATCC. MiaPaCa 2 cell line was initially expanded andcryopreserved within 1 month of receipt. Cells weretypically used for 3 months, and at that time a fresh vialof cryopreserved cells was used. All experiments werecarried out with cells grown to 70% to 80% confluence.To assess growth inhibition and molecular markers, weapplied subtoxic concentrations of DFMO, ranging from2 to 4 mmol/L.

Protein extraction and Western blot assayPancreata (normalþPanINsþPDAC) harvested from

mice fed diets with or without DFMO, or MiaPaCa 2 cellstreatedwith or withoutDFMO in the presence or absence ofspermidine (polyamine), were homogenized and lysed inice-cold lysis buffer. After a brief vortexing, the lysates werecollected by centrifugation at 12,000 � g for 15 minutesat 4�C, and protein concentrations were measured withthe Bio-Rad Protein Assay reagent. An aliquot (50 mg pro-tein/lane) of the total protein was separated with 10%

Mohammed et al.

Cancer Prev Res; 7(12) December 2014 Cancer Prevention Research1200




sodium dodecyl sulfate-polyacrylamide gel electrophoresis(SDS-PAGE) and transferred to nitrocellulose membranes.After blocking with 5% milk powder, membranes wereprobed for expression of ODC, c-myc, p21, and a-tubulinin hybridizing solution [1:500, in Tris-buffered saline(TBS)-Tween 20 solution] using the respective primaryantibodies and then probed with the appropriate horserad-ish peroxidase (HRP)–conjugated secondary antibodies.Detection was performed using the SuperSignal West PicoChemiluminescence procedure (Pierce). All Western blotexperiments were repeated two times. ImageJ software wasused for quantification of the blots.

Statistical analysisThe data are presented asmean� SE. Differences in body

weights were analyzed by ANOVA. Statistical differencesbetween control and treated groups were evaluated usingthe Fisher exact test for PDAC incidence and unpaired t test

with Welch correction was used for PanINs and PDAClesions. Differences between groups are considered signif-icant at P < 0.05.

ResultsActivation of the ODC pathway during progression ofpancreatic cancer

p48Cre/þ-LSL-KrasG12D/þ and wild-type mice were ana-lyzed for the expression of ODC pathway components(Fig. 1A–F). ODC catalyzes the first step in the polyaminebiosynthetic pathway, the decarboxylation of ornithine toputrescine. ODC is modulated by other important compo-nents of this pathway, such as Arg1, Oat, Oaz, and Sat1;and its activation can lead to tumor promotion and pro-gression (Fig. 1G). As the mice aged from 2 to 6 to 10months, we observed a significant increase in the ODCactivity (Fig. 1A). ODC activity was >2-fold elevated in the

OD

C a

ctiv

ity, n

mol

es o

f CO

2re

leas

ed/h

r/mg

prot

ein

Rel

ativ

e m

RN

A e

xpre

ssio

n (O

DC

)

Rel

ativ

e m

RN

A e

xpre

ssio

n (o

az)

3

2

1

0

3

2

1

0

Rel

ativ

e m

RN

A e

xpre

ssio

n (O

at) 4

3

2

1

0

1.5

1.0

0.5

0.0

Rel

ativ

e m

RN

A e

xpre

ssio

n (S

at) 1.5

1.0

0.5

0.0

40

30

20

10

00 0.1

DFMO (%) in the diet

Aver

age

BW

(g) M

ean

± S

Eaf

ter 3

8 w

eeks

of t

reat

men

t 40

30

20

10

0Aver

age

BW

(g) M

ean

± S

Eaf

ter 3

8 w

eeks

of t

reat

men

t

0.2 0 0.1DFMO (%) in the diet

0.2

Rel

ativ

e m

RN

A e

xpre

ssio

n (A

rg) 15

10

5

0

P < 0.03

WT 2M 6M 10MNormal Tumor

Normal TumorNormal

AIN-76Adiet

Experimental diets

HistopathologyMolecular analysis

Tumor

Normal Tumor

Arginase OAT

OAZODC

38 wks

SAT

SAT

44 wks6 wks5 wks

Normal Tumor

Urea cycle

Arginine

Agmatine

A B C G

D

H I J

E FPutrescine

Acetyl spermidine

Spermidine

Acetyl spermine

Spermidinesynthase

Sperminesynthase

PanIN1 and PanIN2 PanIN3 PDAC

Spermine

SAM

SAM

Ornithine

Antizyme

GlutamateProline

P < 0.01

P < 0.0001 P < 0.0005

P < 0.01

P < 0.0001

P < 0.0001

Figure 1. A, ODC activity in the wild-type (WT) and p48Cre/þ-LSL-KrasG12D/þ GEM mice. ODC activity increased in the GEM mice as they aged from 2 to10 months. mRNA expression of ODC (B), Oaz (C), Arg1 (D), Oat (E), and Sat1 (F) in normal and tumor-bearing pancreas. A significant increase in ODC, Arg1,and Oat and a decrease in Oaz and Sat1 mRNA expression were observed in the pancreatic tumors compared with normal pancreas. G, schematicrepresentation of polyamine biosynthesis. H, experimental design for evaluation of DFMO efficacy in pancreatic cancer prevention in male and femalep48Cre/þ-LSL-KrasG12D/þ mice. At 6 weeks of age, groups of mice (30–34/group for activated p48Cre/þ-LSL-KrasG12D/þ or 12/group for wild-type) werefed AIN76-A diets containing 0%, 0.1%, or 0.2% DFMO continuously for 38 weeks and each pancreas was evaluated histopathologically for markerexpression as described in the text. I and J, effect of DFMO on body weight (BW; mean � SE) at termination of the experiment. No statistically significantdifference was observed between control and DFMO-treated p48Cre/þ-LSL-KrasG12D/þ or wild-type mice.






pancreas of 10-month-old GEM mice compared with con-trol pancreas (Fig. 1A). Similarly, other important ODC-regulating enzymes were modulated in the tumors com-pared with normal pancreas as determined by quantitativereal-time PCR (Fig. 1B–F). The ODC-inhibitory enzymeOaz and the polyamine-degrading enzyme Sat1were down-regulated significantly in the pancreatic tumors comparedwith normal pancreas. The Arg and Oat required forincreased ODC were increased in the tumors.

General health of animals treated with DFMOThe experimental protocol for evaluating DFMO in

pancreatic cancer progression is summarized in Fig. 1H.p48Cre/þ-LSL-KrasG12D/þ mice fed AIN-76A or DFMOdiets had steady body weight gain. As shown in Fig. 1Iand J, there was no significant difference in body weightin the mice fed either AIN-76A or AIN-76A diets supple-mented with either 0.1% or 0.2% DFMO. None of theanimals fed the DFMO diets exhibited any observabletoxicity or any gross morphologic changes to liver, spleen,kidney, or lung.

Effect of DFMO diet on pancreatic tumor weight andPDAC incidence in GEM

Pancreatic weight is a simple marker to assess the pro-gression of tumor. A significant increase in pancreas weight(0.8–1.5 g) was observed in p48Cre/þ-LSL-KrasG12D/þ micecompared with wild-type mice (�200–300 mg). Figure 2Ashows the H&E staining of the pancreatic tumors with andwithout DFMO treatment. As summarized in Fig. 2B and C,DFMO treatment caused a significant decrease in the weightof pancreatic tumors in p48Cre/þ-LSL-KrasG12D/þ mice. InGEM fed 0.1% or 0.2% DFMO, the pancreas weights were0.75 � 0.1 g and 0.86 � 0.01 g, respectively, in males and0.65 � 0.08 g and 0.61 � 0.04 g in females. The pancreatictumor weights were decreased by 31% to 43% (P < 0.03–0.001) with both doses of DFMO inmale and female mice.

Extensive histopathologic analysis of the pancreas usingH&E-stained slides revealed no microscopic pathologicalterations in wild-type mice fed either AIN-76A orDFMO-supplemented diets. In contrast, AIN76A diet-fedmale and female p48Cre/þLSL-KrasG12D/þmice demonstrat-ed >80% and 60% incidence of PDAC (% of mice with

Untreated DFMO treated

1.5

1.0

0.5

0.0

1.2

1.0

0.8

0.6

0.4

0.2

0.0

0%0.1%0.2%

0%0.1%0.2%

0%

0.1%

0.2%

0%

0.1%

0.2%

0.2% (WT)0% (WT)

Pan

crea

s w

eigh

ts (

g) M

ean

± S

Eaf

ter

38 w

eeks

of t

reat

men

t

100

80

60

40

20

0

Pan

crea

tic c

arci

nom

a in

cide

nce

60

40

20

0

Pan

crea

tic c

arci

nom

a in

cide

nce

Pan

crea

s w

eigh

ts (

g) M

ean

± S

Eaf

ter

38 w

eeks

of t

reat

men

t

DFMO (%) in the diet DFMO (%) in the diet


A

B C

D E

P <

0.0

01

P < 0.0001P < 0.0001

P < 0.0001

P < 0.0001

P <

0.0

6

P <

0.0

3

P <

0.0

1

Figure 2. A, histopathologicanalysis of untreated andDFMO-treated pancreas usingH&E staining. Pancreas fromuntreated animals showing poorlydifferentiated adenocarcinomawith some of the cells invadingstroma (left). Pancreas fromanimals treated with DFMOshowing PanIN lesions (right).B–C, effect of DFMO on pancreasweight at the termination of theexperiment in male (B) and female(C) mice. Both doses of DFMOsignificantly reduced thepancreatic tumorweights. D andE,effect of DFMO on the incidenceof PDAC in male (D) and female(E) mice.

Mohammed et al.





PDAC), respectively (Fig. 2D and E). DFMO treatment at0.1% and 0.2% decreased PDAC incidence to 10% and 6%,respectively, in male and to 16% and 9%, respectively, infemale GEM (Fig. 2D and E).

Inhibition of PanIN lesion progression and carcinomapercentage by DFMOHistologic analysis showed 100% penetrance of pan-

creatic precursor PanIN lesions in the GEM fed AIN76A orDFMO-supplemented diets. The number of PanIN 1, 2,and 3 lesions in male GEM fed AIN76-A diet was (mean�SE): 180 � 28, 162 � 20, and 155 � 15, respectively; inthe mice fed 0.1% DFMO, PanIN 1, 2, and 3 numberswere 375 � 59, 225 � 42, and 115 � 33; and in mice fed0.2%DFMO they were 260� 33, 192� 20, and 109� 17,respectively (Fig. 3A). The number of PanIN 1, 2, and 3lesions in female GEM fed AIN76A diet were 183 � 23,178 � 13, and 240 � 19, respectively; in the mice fed0.1% DFMO, the numbers were 255 � 36, 220 � 21, and188 � 28; and in mice fed 0.2% DFMO, they were 280 �49, 195 � 24, and 158 � 16, respectively (Fig. 3B). Thenumber of PanIN 3 lesions or carcinoma in situ was

suppressed by 21% to 33% in the DFMO-treated groups(Fig. 3A and B). An increase in the number of PanIN 1 and2 lesions was observed in pancreas of DFMO-supplemen-ted mice, suggesting a potential blockade of further pro-gression of these lesions to carcinoma in situ and PDAC.Pancreas of GEM fed AIN76 A diet showed 30% � 6%(male; Fig. 3C) and 12% � 4% (female; Fig. 3D) PDACwithin the pancreas. The carcinoma percentage within thepancreas was inhibited significantly (>60% in males and>50% in females; Fig. 3C and D) by both doses of DFMOin GEM. Up to 23% of the pancreas from mice treatedwith DFMO was normal appearing (i.e., free from PanINlesions and carcinoma), whereas only up to 5% wasnormal appearing in the untreated GEM (Fig. 3E andF). Although DFMO significantly inhibited PDAC inci-dence and carcinoma spread, tumor outgrowth wasobserved in the treatment groups.

Modulation of ODC signaling by DFMO in pancreaticcancer

Activation of the ODC pathway is critical in polyaminebiosynthesis. ODC pathway signaling molecules were

450400350300250200150100

500

350

300

250

200

150

100

50

0

0%0.1%0.2%

0%0.1%0.2%

0%0.1%0.2%

0%0.1%0.2%

0%0.1%0.2%

0%0.1%0.2%

Num

ber

of P

anIN

lesi

ons

Mea

n ±

SE

Num

ber

of P

anIN

lesi

ons

Mea

n ±

SE

40

30

20

10

0

353025201510

50

% C

arci

nom

a

17.515.012.510.0

7.55.02.50.0

% C

arci

nom

a

% N

orm

al p

ancr

eas

25

20

15

10

5

0

% N

orm

al p

ancr

eas

DFMO (%) in the dietPanIN 1 PanIN 2 PanIN 3 PanIN 1 PanIN 2 PanIN 3

DFMO (%) in the diet


DFMO (%) in the dietDFMO (%) in the diet

A B

C D

E F

P < 0.006

P < 0.01

P <

0.0

2

P <

0.0

4

P <

0.0

04

Figure 3. A–B, effect of DFMO onPanIN multiplicity in male (A) andfemale (B) GEM mice (mean � SE).C and D, effect of DFMO on thepercentage of pancreas withcarcinoma (C, male; D, female).E and F, effect of DFMO on thepercentage of normal appearingpancreas (E, male; F, female). Thedata in the panelswere analyzed byunpaired "t" test with Welchcorrection; values are consideredstatistically significant at P < 0.05.






measured via IHC and quantitative real-time PCR in pan-creatic tissues. As shown in Fig. 4A, markedly decreasedODC staining was observed in PanIN lesions and carcino-ma in p48Cre/þ-LSL-KrasG12D/þ mice fed DFMO-supple-mented diets compared with the pancreatic tissues frommice fed AIN-76Adiet alone. The expression ofODCand itsrelated signaling molecules in the pancreas (consistingnormal pancreas, PanINs, and PDAC) was analyzed furtherusing real-time quantitative PCR assays. Snap-frozen pan-creatic tissues were analyzed for fold change in mRNAexpression of ODC, Arg, OAZ, and Sat1 (Fig. 4B–D). TotalRNA was isolated, reverse-transcribed into cDNA, andquantified with a real-time PCR assay. As shownin Figure 4B–E, compared with pancreas from AIN-76A-fedp48Cre/þ-LSL-KrasG12D/þ mice, the pancreatic tissues ofDFMO-fed GEM showed significantly decreased (>2-fold;P < 0.05) expression ofODC and ArgmRNA (Fig. 4B andC)and an increase (>2-fold) in Oaz mRNA (Fig. 4D). Expres-sion of Sms and Srm was reduced upon DFMO treatment(Fig. 4E and F). The decrease in the ODC signaling mRNAs

in the DFMO treatment groups is likely to be due to thegreater amount of normal pancreas present in the treatmentgroups.

DFMO inhibits tumor cell proliferation and inducesp21, p27, and p53

To determine the effects of DFMO on tumor cell prolif-eration in pancreatic tissues of the GEM, PCNA, Ki67, p21,p27, cyclin D1, Rb, p16, and p53 were analyzed using IHC/IHF and real-time PCR approaches (Fig. 5A and B andSupplementary Fig. S1). IHC staining showed that PanINlesions and carcinoma were labeled positively for PCNAand Ki67 in the p48Cre/þ-LSL-KrasG12D/þ mice fed the AIN-76A diet alone (Fig. 5A and Supplementary Fig. S1).Markedly decreased numbers of PCNA- and Ki67-positivecells and decreased proliferative index were observed inmice fed a DFMO-supplemented diet (Fig. 5A and B andSupplementary Fig. S1). IHC analysis of p21 revealed asignificant increase in labeling with the higher DFMO dosecompared with untreated pancreatic tumor tissues (Fig. 5A

Untreated

OD

CO

DC

DFMO 0.1% DFMO 0.2%

Rel

ativ

e m

RN

A e

xpre

ssio

n (O

DC

)

Rel

ativ

e m

RN

A e

xpre

ssio

n (O

az) 3

2

1

0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

Rel

ativ

e m

RN

A e

xpre

ssio

n(s

perm

ine

synt

hase

)

Rel

ativ

e m

RN

A e

xpre

ssio

n(s

perm

idin

e sy

ntha

se)

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

DFMO in diet

Rel

ativ

e m

RN

A e

xpre

ssio

n (A

rg1)

A B

C D E F

0 0.1% 0.2%

DFMO in diet0 0.1% 0.2%




Figure 4. A, effect of DFMO on ODC expression in pancreatic tumors. IHC and immunofluorescence analyses were performed with paraffin-embedded andmicrosectioned pancreatic tissues as described in theMaterials andMethods. A significantly decreased expression of ODCwas seen with DFMO treatment.B–F, effect of DFMO onODC (B), Arg1 (C), Oaz (D), Sms (E), and Srm (F) mRNA expression in pancreatic tumors as determined by quantitative real-time PCR.DFMO treatment led to decreased ODC, Arg1, Sms, Srm, and increased Oaz. (�, P < 0.01; ��, P < 0.001).

Mohammed et al.





and B). We also measured the relative mRNA levels ofPCNA, p21, and p27 in pancreatic tissues by quantitativereal-time PCR (qRT-PCR) and found that DFMO signifi-cantly decreased the expression of PCNA and increased p21mRNA in pancreatic tissues of the GEM. The higher dose ofDFMO also led to an increase in p53 and p27 expression inthe pancreatic tissues (Fig. 5B). These results suggest that theinhibition of tumor progression may be associated withdecreased tumor cell proliferation (Fig. 5A and B) and areconsistent with our in vivo tumor inhibition findings. Fur-thermore, we have observed a decrease in cyclin D1 and anincrease in the tumor-suppressor Rb upon DFMO treat-ment. Higher p16 expression was observed in the normalappearing pancreatic tissues from the DFMO-treated pan-creatic tumors compared with the untreated tumors (Sup-plementary Fig. S1).

Effect of DFMO on Cav-1 and b-cateninC-myc activation leads to increased proliferation and

downregulates p21, which can lead to tumor growth. Toexamine the effect of DFMO on the expression of c-myc,we analyzed the pancreatic tissues (containing normalpancreas, PanINs, and PDAC) and MiaPaCa cells by

Western immunoblotting. As shown in Fig. 5C, theDFMO treatment caused a nonsignificant reduction inthe expression of c-myc, with no dose-dependency. Thisreduction most likely is due to higher amounts of normalpancreas in treatment groups. We observed no significantchange in the ODC protein expression in the DFMO-treated and polyamineþDFMO–treated MiaPaCa cells.The c-myc expression was unchanged in the lower dosetreatment, with only a slight increase in high-dose group;and cells treated with polyamines with or without DFMOshowed no change in c-myc protein. Moreover, we foundthat p21 was increased to some extent upon DFMOtreatment (Supplementary Fig. S2). Because Cav-1 cantransport polyamines into the cells (20), we analyzedCav-1 protein expression by IHC/IHF and also measuredmRNA expression by real-time PCR (Fig. 6). A substantialdecrease in Cav-1 was observed upon DFMO treatment ascompared with untreated mouse pancreatic tumor tissues(Fig. 6). Similarly, we observed a decrease in the b-cate-nin protein and mRNA expressions in the treatmentgroups (Fig. 6). These results are consistent with our invivo tumor inhibition findings and suggest that the inhi-bition of tumor progression may be associated with the

Untreated

PC

NA

p21

DFMO 0.1% DFMO 0.2%DFMO

0% 0.1% 0.2%

c-Myc

α-Tubulin

Rel

ativ

e pr

otei

n ex

pres

sion

(c-m

yc)

Rel

ativ

e m

RN

A e

xpre

ssio

n(P

CN

A)

100

80

60

40

20

0

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0 Rel

ativ

e m

RN

A e

xpre

ssio

n(p

21)

Rel

ativ

e m

RN

A e

xpre

ssio

n(p

27)

6

4

2

0 Rel

ativ

e m

RN

A e

xpre

ssio

n(p

53)

4

3

2

1

0

4

3

2

1

0

DFMO (%) in diet

PC

NA

inde

x

A

B

C

0 0.1 0.2

DFMO (%) in dietC 0.1 0.2

DFMO (%) in diet0 0.1 0.2




Figure 5. Effect of DFMO on pancreatic tumor cell proliferation. A, IHC analysis for PCNA and p21 was performed on paraffin-embedded and microsectionedpancreatic tissues as described in the Materials and Methods. A significantly decreased expression of PCNA and increased p21 expression was seenwith DFMO treatment. B, effect of DFMOon the PCNA index and onmRNA expression of PCNA, p21, p53, and p27 as determined with quantitative real-timePCR. A significantly decreased PCNA index and increased p21, p27, and p53 mRNA expression were seen in the DFMO-treated pancreatic tumor tissue.C, effect of DFMO on c-Myc protein expression. A decrease in c-Myc expression was seen upon DFMO treatment.






modulation of molecular markers related to the ODCpathway.

DiscussionPolyamines are essential for cell division and growth of

neoplasia. Overexpression of ODC and increased poly-amines are observed in tumor samples from varioushuman cancers including pancreatic cancer (7, 21). ODCactivity was 3.6-fold elevated in human adenocarcinomasand 3.9-fold elevated in neuroendocrine tumors com-pared with control pancreas (8). We observed increasedODC activity in p48Cre/þ-LSL-KrasG12D/þ mice, a well-established model that simulates the stepwise progressionof human pancreatic cancer and which has been usedextensively by us and others for studies of pancreaticcancer chemoprevention. As the mice aged, the ODCactivity increased significantly (Fig. 1A), suggesting thattumor cell proliferation is occurring along with the tumorprogression. We evaluated mRNA expression for severalconstituents of the ODC pathway in the tumors as com-pared with normal pancreas. Oaz, which regulates poly-amine synthesis by binding to and inhibiting ODC, andSat1, a rate-limiting enzyme in the catabolic pathway ofpolyamine metabolism, were decreased significantly inthe pancreatic tumors compared with normal pancreas(Fig. 1). Arginase and Oat, which increase polyaminelevels by increasing the ODC activity, were also increasedin the tumors (Fig. 1). These results are consistent withprevious reports on polyamine biosynthesis in different

cancers (22) and are consistent with a role for polyaminebiosynthesis in pancreatic cancer progression.

Previous studies have shown cytostatic effects of theODCinhibitor DMFO against several tumors (23–25) but not inpancreatic cancers. Clinical and preclinical studies haveshown that DFMO has significant chemopreventive effectsagainst several cancers (10–16); and DFMO has been usedsuccessfully for chemoprevention of colon, prostate, skin,and cervical cancers (26–31). However, no in vivo chemo-prevention studies had been reported with DFMO usingGEM that mimic human pancreatic cancer.

Inhibition of ODC activity by DFMO caused decreasedcell growth and increased apoptosis in pancreatic tumor-derived cell lines (8). DFMOwas also examined in hamsterH2T cells, an experimental xenograft model of pancreaticcarcinogenesis, both in vitro and in vivo by Marx and col-leagues (32). In vitro they showed that cytotoxicity andinhibition of polyamine biosynthesis increased with pro-longed DFMO treatment, suggesting that continuous use ofDFMO for pancreatic cancer prevention might improveefficacy. In the in vivo studies, H2T cells were injected intothe cheek pouch of Syrian golden hamsters and DFMOtreatment (3% in drinking water) decreased tumor size andinhibited growth of the pancreatic cancer by as much as50% of control (32). However, that xenograft model lacksclassic initiation and promotion steps that characterizehuman pancreatic cancer.

We have extended these findings by showing the poten-tial of dietary DFMO (0.1% and 0.2%) to prevent progres-sion of pancreatic cancer in the p48Cre/þ-LSL-KrasG12D/þ

Untreated

Cav

-1C

av-1

ββ-C

aten

inβ-

Cat

enin

DFMO 0.1% DFMO 0.2%

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

Rel

ativ

e m

RN

A e

xpre

ssio

n (C

av1)

Rel

ativ

e m

RN

A e

xpre

ssio

n (β

-cat

enin

)



Figure 6. Effect of DFMO on Cav-1and b-catenin. IHC (top row ofeach set), immunofluorescence(bottom row of each set) andmRNA analysis (bar graphs to theright) revealed a decrease in theCav-1 expression with both dosesof DFMO; however, only the highdose (0.2%) caused a decrease inb-catenin expression.

Mohammed et al.





mice. We found that DFMO inhibits formation of PanINlesions and their progression to ductal adenocarcinoma.DFMO (0.1% and 0.2%) caused significant inhibition ofPDAC incidence in the male and female GEM (Fig. 2). A21% to 33% suppression of PanIN 3 lesions (carcinoma insitu) was observed with DFMO treatment, and carcinomapercentage was inhibited by >60% inmales and by >50% infemales (Fig. 3). There was no dose-dependency observedwith DFMO treatment; instead, the lower dose was associ-ated with higher amounts of normal appearing pancreas(Fig. 3).We observed DFMO-induced inhibition of ODC and

modulation of several components of polyamine biosyn-thesis in the pancreatic tumors. The DFMO treatment led toa significant decrease in expression ofODC, Arg1, Sms, Srm,and increased Oaz expression (Fig. 4). DFMO caused asignificant decrease in tumor progression, with a largeamount of the pancreas still appearing normal (Fig. 3). Thesignificant reduction in pancreatic ODCmRNAmost likelywas due to the reduction in the overall tumor area (PanINsand PDAC) in the pancreata ofmice exposed toDFMO. Thepancreas tissue of untreated mice consists of greater num-bers of tumor and stromal cells in comparison with theDFMO treatment groups (Figs. 2 and 3). Mechanistically,although DFMO inhibits ODC activity, few studies havereported that ODC mRNA is reduced upon DFMO treat-ment (33–35). ODC is a critical transcription target of theMyc oncogene, which regulates checkpoints that guardagainst tumorigenesis; and is an effective target for cancerchemoprevention (33). Overexpression of Myc at levelsfound in cancer is sufficient to drive normal quiescent cellsinto the cycle and to accelerate their rates of cell-cycletraverse (36). These responses are, at least, in part, depen-dent upon Myc’s ability to downregulate the expression ofthe cyclin-dependent kinase (Cdk) inhibitors p27Kip1 andp21Cip1 (37–39). In other systems, DFMO treatment hasbeen associatedwith a reduction in c-Myc and an increase inp16 and p27 (40–42). Consistent with these reports, weobserved a decrease in expression of c-myc upon DFMOtreatment, although it did not reach statistical significance,and a significant increase in expression of p21, p27, and thetumor-suppressor p53 (Fig. 5).Cav-1 has been predicted to help in the uptake of poly-

amines into cells (20). Clinical studies have shown thatCav-1 expression is associated with pancreatic tumor pro-gression and poor prognosis for patient survival (43–44). Inthis study, overexpression of Cav-1 was correlated with anincrease in proliferation markers (Fig. 6). The high expres-sion of the proliferation markers Ki67, PCNA, Cav-1, andb-catenin observed in KrasG12D/þ mouse pancreatic lesionswas inhibited by dietary DFMO (Figs. 5 and 6). The exactmechanismbywhichDFMOinfluencesCav-1 is not known;however, our previous studies on colon cancer cell linesexposed to lovastatin or pancreatic cancer inhibition byatorvastatin showed significantly suppressed Cav-1 expres-sion and its localization to membrane lipid rafts (19, 20).DFMO treatment results in the depletion of polyamines

leading to a significant decrease in the steady-state levels of

transcription factors such as c-myc. It is suggested thatregulation of c-myc expression may be one of the mechan-isms by which polyamines modulate cell growth. A fewstudies have reported a decrease in c-myc upon DFMOtreatment (40–42, 45, 46). However, a number of signalingmechanisms, especially those involving Wnt, can contrib-ute to regulation of c-myc and polyamines (22, 47). In thepresence of Wnt, b-catenin target genes including c-myc areexpressed.Myc expression, in turn, leads to the expressionofODC, the first enzyme in polyamine metabolism. Dimin-ished b-catenin leads to reduced c-Myc (48). Wnt signalingalso regulates Oaz, a protein that regulates ODC activity bytargeting its degradation. In addition, peroxisome-prolif-erator–activated receptor-g (PPAR-g), which activates sper-midine/spermine N(1)-acetyltransferase (SSAT) transcrip-tion, is repressed by active KRAS. However, in normal cellsand tissues, most of the KRAS protein is inactive as asignaling molecule. Therefore, in normal cells and tissues,Wnt signaling and inactive KRAS lead to reduced prolifer-ation, increased apoptosis, and reduced neoplasia. Undertumor conditions, ODC expression is elevated by alteredWnt signaling, in part, due to inhibition of OAZ. Studiesalso support the existence of oncogenic mutations in KRASthat suppress polyamine catabolism (22). Consequently,alteredWnt signaling and KRAS together can act to promoteneoplasia by increasing polyamine biosynthesis and sup-pressing polyamine catabolism. In addition, Myc overex-pression is able to downregulate the expression of the Cdkinhibitors p27Kip1 and p21Cip1 (37–39). In the presentstudies, DFMO is shown to suppress the pancreatic cancerprogression by modulating ODC signaling, decreasing pro-liferation, altering Wnt signaling, and increasing p21, p27,Rb, and p53 (Supplementary Fig. S3).

Although DFMO inhibited PDAC in Kras mice, there wasstill some tumor outgrowth in animals treated with bothtested doses, indicating potential limitations to the use ofDFMO in human clinical trials. Some possible reasons forthis tumor escape include the following: (i) althoughendogenous polyamines are reduced upon DFMO treat-ment, dietary polyamines are readily available exogenously.Thus, we think that one of the reasons for outgrowth oftumors even after DFMO treatment is the availability ofexogenous polyamines through these diets. (ii) It is possiblethat drug resistance occurs as a result of decreased intracel-lular drug accumulation due to overexpression of certaindrug-resistant efflux pumps. It may be possible to test thispossibilitywith short-term in vitro culturingof theoutgrowntumors. (iii) Another potential mechanism for DFMO resis-tance relates to the alternative mechanisms available for thesynthesis of polyamines within the tumor cells, includingalterations of the agmatinemechanism (Fig. 1G). As evidentfrom Fig. 4E and F, although the ODC inhibition is irre-versible, polyamines are still available from these alternatepathways and may lead to the tumor outgrowth. (iv)Another potential explanation for tumor outgrowth isdependent on ratio of DFMO and ODC concentrations.The half-life (t1/2) of DFMO in the serum is 2 to 4 hours andit is eliminated in 3.2 to 3.6 hours. The ODC enzyme has






turn-over rate or t1/2 of 30 minutes. Thus, any observedpersistent biologic effects of DFMO are questionable. Pre-viously, we and others have shown dose–response effects ofDFMOinother organ site cancers. The failure to detect dose-dependent effects of DFMO on PDAC in this study suggeststhat both doses tested were on the plateau of the dose–response curve, and thus that tumor outgrowth was not dueto insufficient DFMO.However, the ratio of DFMO toODCconcentration may also vary at different time pointsthroughout the period of the experiment. Although miceare fed DFMO in the diet, DFMO levels in the serum mayvary periodically depending on the dietary intake by themice. This possibility can be tested by determining the levelsof DFMO in serum as a function of time and correlatingaverage, maximal and nadir levels with antitumor efficacy.(v) It is also possible that tumor outgrowth is related to thespecies or strain of animal used. Effects of DFMO should betested in other models of pancreatic cancer to see if tumoroutgrowth occurs. (vi) Finally, it should also be noted that,there are several other signalingmechanisms apart from theODC pathway that can drive the tumorigenesis process.Thus, developing combinations ofDFMOwith other agentshaving different tumor-inhibitory mechanisms mayimprove the efficacy.

Because the Kras GEMmodel shows stepwise progressionof precursor lesions to carcinoma as occurs in humans, andis well optimized for studies of pancreatic cancer chemo-prevention (5, 17–19, 49, 50), our findings suggest (i) thatODC plays a key role in progression of PanIN lesions topancreatic cancer; and (ii) that low-dose DFMO may havepotential for chemoprevention in patients at high risk forpancreatic cancer. Patients with chronic pancreatitis, thosewith PanIN lesions, intrapapillary mucinous neoplasms,diabetes, or those of older age might be evaluated for

polyamine levels to evaluate their candidacy for such che-moprevention. Further evaluation is also warranted ofDFMO effects on tumors at an earlier stage and in combi-nation with other agents to increase the efficacy.

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

Authors' ContributionsConception and design: A. Mohammed, N.B. Janakiram, V.E. Steele,C.V. RaoDevelopment of methodology: A. MohammedAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.):A.Mohammed, V.Madka, R.L. Ritchie,M. Brewer,M. Sadeghi, C.V. RaoAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): A. Mohammed, N.B. Janakiram,V. Madka, R.L. Ritchie, J.M.R. Patlolla, S. Lightfoot, C.V. RaoWriting, review, and/or revision of the manuscript: A. Mohammed,N.B. Janakiram, C.V. RaoAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): V. Madka, M. Brewer, L. Biddick,C.V. RaoStudy supervision: A. Mohammed, V.E. Steele, C.V. Rao

AcknowledgmentsThe authors thank the University of Oklahoma Health Sciences Center

(OUHSC) Rodent Barrier Facility (RBF) staff. The authors also want to thankDr. Julie Sando for valuable suggestions and editorial help.

Grant SupportThisworkwas supported by theNational Cancer InstituteN01-CN-53300

to C.V. Rao.The costs of publication of this article were defrayed in part by the

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

Received May 30, 2014; revised September 8, 2014; accepted September16, 2014; published OnlineFirst September 23, 2014.

References1. American Cancer Society. Cancer facts and figures 2014. Atlanta, GA:

American Cancer Society; 2014.2. Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, et al.

GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide:IARCCancerBaseNo. 11 [Internet]. Lyon, France: International Agencyfor Research on Cancer; 2013. Available from: http://globocan.iarc.fr.

3. Mazur PK, Siveke JT. Genetically engineered mouse models of pan-creatic cancer: unravelling tumor biology andprogressing translationaloncology. Gut 2012;61:1488–500.

4. Yachida S, Jones S, Bozic I, Antal T, Leary R, Fu B, et al. Distantmetastasis occurs late during the genetic evolution of pancreaticcancer. Nature 2011;467:1114–17.

5. Mohammed A, Janakiram NB, Lightfoot S, Gali H, Vibhudutta A, RaoCV. Early detection and prevention of pancreatic cancer: use ofgenetically engineered mouse models and advanced imaging tech-nologies. Cur Med Chem 2012;19:3701–13.

6. Weiss TS, Bernhardt G, Buschauer A, Thasler WE, Dolgner D, ZirngiblH, et al. Polyamine levels of human colorectal adenocarcinomas arecorrelated with tumor stage and grade. Int J Colorectal Dis 2002;17:381–87.

7. Hu HY, Liu XX, Jiang CY, Lu Y, Liu SL, Bian JF, et al. Ornithinedecarboxylase gene is overexpressed in colorectal carcinoma. WorldJ Gastroenterol 2005;11:2244–48.

8. Subhi AL, Tang B, Balsara BR, Altomare DA, Testa JR, Cooper HS,et al. Loss of methylthioadenosine phosphorylase and elevated orni-

thine decarboxylase is common in pancreatic cancer. Clin Cancer Res2004;10:7290–96.

9. Black O Jr, Chang BK. Ornithine decarboxylase enzyme activity inhuman and hamster pancreatic tumor cell lines. Cancer Lett 1982;17:87–93.

10. Meyskens FL Jr, Gerner EW. Development of difluoromethylornithineas a chemoprevention agent. Clin Cancer Res 1999;5:945–51.

11. Meyskens FL Jr, McLaren CE, Pelot D, Fujikawa-Brooks S, CarpenterPM, Hawk E, et al. Difluoromethylornithine plus sulindac for theprevention of sporadic colorectal adenomas: a randomized place-bo-controlled, double-blind trial. Cancer Prev Res 2008;1:32–38.

12. Sintoneau AR, Gerner EW, Nagle R, Ziogas A, Fujikawa-Brooks S,YerushalmiH, et al. The effect of difluoromethylornithineondecreasingprostate size and polyamines in men: results of a year-long phase IIbrandomized placebo-controlled chemoprevention trial. Cancer Epide-miol Biomarkers Prev 2008;17:292–99.

13. Witherspoon M, Chen Q, Kopelovich L, Gross SS, Lipkin SM.Unbiased metabolite profiling indicates that a diminished thymi-dine pool is the underlying mechanism of colon cancer chemo-prevention by alpha-difluoromethylornithine. Cancer Discov 2013;3:1072–81.

14. Kreul SM,Havighurst T, KimK,Mendonca EA,WoodGS, SnowS, et al.A phase III skin cancer chemoprevention study of DFMO: long-termfollow-up of skin cancer events and toxicity. Cancer Prev Res 2012;5:1368–74.

Mohammed et al.





15. Rao CV, Tokumo K, Rigotty J, Zang E, Kelloff G, Reddy BS. Chemo-prevention of colon carcinogenesis by dietary administration of pir-oxicam, alpha-difluoromethylornithine, 16 alpha-fluoro-5-androsten-17-one, and ellagic acid individually and in combination. Cancer Res1991;51:4528–34.

16. Meyskens FL Jr, Gerner EW, Emerson S, Pelot D, Durbin T, Doyle K,et al. Effect of alpha-difluoromethylornithine on rectalmucosal levels ofpolyamines in a randomized, double-blinded trial for colon cancerprevention. J Natl Cancer Inst 1998;90:1212–18.

17. Mohammed A, Janakiram NB, Brewer M, Ritchie RL, Marya A, Light-foot S, et al. Antidiabetic Drug Metformin Prevents Progression ofPancreatic Cancer by Targeting in Part Cancer Stem Cells and mTORSignaling. Transl Oncol 2013;6:649–59.

18. Mohammed A, Janakiram NB, Li Q, Madka V, Ely M, Lightfoot S, et al.The epidermal growth factor receptor inhibitor gefitinib preventsthe progression of pancreatic lesions to carcinoma in a conditionalLSL-KrasG12D/þ transgenic mouse model. Cancer Prev Res 2010;3:1417–26.

19. Mohammed A, Qian L, Janakiram NB, Lightfoot S, Steele VE, Rao CV.Atorvastatin delays progression of pancreatic lesions to carcinoma byregulating PI3/AKT signaling in p48Cre/þ-LSL-KrasG12D/þ mice. Int JCancer 2012;131:1951–62.

20. Guruswamy S, Rao CV. Synergistic effects of lovastatin and celecoxiboncaveolin-1 and itsdownstreamsignalingmolecules: implications forcolon cancer prevention. Int J Oncol 2009;35:1037–43.

21. L€oser C, F€olsch UR, Paprotny C, Creutzfeldt W. Polyamine concen-trations in pancreatic tissue, serum, and urine of patients with pan-creatic cancer. Pancreas 1990;5:119–27.

22. Gerner EW, Meyskens FL Jr. Polyamines and cancer: old molecules,new understanding. Nature Reviews Cancer 2004;4:781–92.

23. Grossie VB Jr, Ota DM, Ajani JA, Nishioka K. Effects of intravenous a-difluoromethylornithine on the polyamine levels of normal tissue and atransplantable fibrosarcoma. Cancer Res 1987;47:1836–40.

24. Kingsnorth AN. The chemotherapeutic potential of polyamine anti-metabolites. Ann R Coll Surg Engl 1986;68:76–81.

25. Tutton PJM, Barkla DH. Comparison of the effects of an ornithinedecarboxylase inhibitor on the intestinal epithelium and on intestinaltumors. Cancer Res 1986;46:6091–94.

26. Laukaitis CM, Gerner EW. DFMO: targeted risk reduction therapy forcolorectal neoplasia. BestPractResClinGasteroenterol 2011;25:495–506.

27. Esmat AY, Refaie FM, Shaheen MH, Said MM. Chemoprevention ofprostate carcinogenesis byDFMOand/or finasteride treatment inmaleWistar rats. Tumori 2002;88:513–21.

28. Gupta S, Ahmad N, Marengo SR, MacLennan GT, Greenberg NM,Mukhtar H. Chemoprevention of prostate carcinogenesis by alpha-difluoromethylornithine in TRAMPmice. Cancer Res 2000;60:5125–33.

29. Bailey HH, Kim K, Verma AK, Sielaff K, Larson PO, Snow S, et al. Arandomized, double-blind, placebo-controlled phase 3 skin cancerprevention study of {alpha}-difluoromethylornithine in subjects withprevious history of skin cancer. Cancer Prev Res 2010;3:35–47.

30. Vlastos AT, West LA, Atkinson EN, Boiko I, Malpica A, Hong WK.Results of a phase II double-blinded randomized clinical trial of difluor-omethylornithine for cervical intraepithelial neoplasia grades 2 to 3.Clin Cancer Res 2005;11:390–96.

31. Basuroy UK, Gerner EW. Emerging concepts in targeting the poly-amine metabolic pathway in epithelial cancer chemoprevention andchemotherapy. J Biochem 2006;139:27–33.

32. Marx M, Townsend CM Jr, Barranco SC, Glass EJ, Thompson JC.Treatment of hamster pancreatic cancer with alpha-difluoromethylor-nithine, an inhibitor of polyamine biosynthesis. J Natl Cancer Inst1987;79:543–48.

33. Nilsson JA, Keller UB, Baudino TA, Yang C, Norton S, Old JA, et al.Targeting ornithine decarboxylase in Myc-induced lymphomagenesisprevents tumor formation. Cancer Cell 2005;7:433–44.

34. Rounbehler RJ, Li W, Hall MA, Yang C, Fallahi M, Cleveland JL.Targeting ornithine decarboxylase impairs development of MYCN-amplified neuroblastoma. Cancer Res 2009;69:547–53.

35. Wan X, Jiang B, Liu YS, Liu HW, Wu GY, Chen R. Effect of alpha-difluoromethylornithine on the expression of ODC mRNA in the cortexand hippocampus in rats after cerebral ischemia reperfusion. ZhongNan Da Xue Xue Bao Yi Xue Ban 2005;30:579–82.

36. Bouchard C, Staller P, Eilers M. Control of cell proliferation by Myc.Trends Cell Biol 1998;8:202–06.

37. Baudino TA, Maclean KH, Brennan J, Parganas E, Yang C, AslanianA, et al. Myc-mediated proliferation and lymphomagenesis, but notapoptosis, are compromised by E2f1 loss. Mol Cell 2003;11:905–14.

38. Herold S, Wanzel M, Beuger V, Frohme C, Beul D, Hillukkala T, et al.Negative regulation of the mammalian UV response by Myc throughassociation with Miz-1. Mol Cell 2002;10:509–21

39. Vlach J, Hennecke S, Alevizopoulos K, Conti K, Amati B. Growth arrestby the cyclin-dependent kinase inhibitor p27Kip1 is abrogated by c-Myc. EMBO J 1996;15:6595–04.

40. Tao L, Kramer PM,WangW, Yang S, Lubet RA, Steele VE, Pereira MA.Altered expression of c-myc, p16 and p27 in rat colon tumors and itsreversal by short-term treatment with chemopreventive agents. Car-cinogenesis 2002;23:1447–54.

41. Raul F. Revival of 2-(difluoromethyl)ornithine (DFMO), an inhibitor ofpolyaminebiosynthesis, as acancer chemopreventive agent. BiochemSoc Trans 2007;35:353–5.

42. Liu L, Li L, Rao JN, ZouT, ZhangHM,BonevaD,BernardMS,Wang JY.Polyamine-modulated expression of c-myc plays a critical role instimulation of normal intestinal epithelial cell proliferation. AmJPhysiolCell Physiol 2005;288:C89–99.

43. Suzuoki M, Miyamoto M, Kato K, Hiraoka K, Oshikiri T, Nakakubo Y,et al. Impact of caveolin-1 expression on prognosis of pancreaticductal adenocarcinoma. Br J Cancer 2002;87:1140–44.

44. Hehlgans S, Eke I, Haase M, Baretton G, Cordes N. Caveolin-1mediated radioresistance of 3D grown pancreatic cancer cells. Radio-ther Oncol 2009;92:363–70.

45. KlinkenSP,HolmesKL,MorseHC III, ThorgeirssonSS. Transcriptionaland post-transcriptional regulation of c-myc, c-myb, and p53 duringproliferation and differentiation ofmurine erythroleukemia cells treatedwith DFMO and DMSO. Exp Cell Res 1988;178:185–98.

46. Celano P, Berchtold CM, Giardiello FM, Casero RAJr. Modulation ofgrowthgene expressionby selective alteration of polyamines in humancolon carcinoma cells. Biochem Biophys Res Commun 1989;165:384–90.

47. GernerGW. Impact of dietary amino acids andpolyamineson intestinalcarcinogenesis and chemoprevention inmousemodels. BiochemSocTrans 2007;35:322–5.

48. Juan J, Muraguchi T, Iezza G, Sears RC, McMahon M. DiminishedWNT -> b-catenin -> c-MYC signaling is a barrier for malignantprogression of BRAFV600E-induced lung tumors. Genes Dev 2014;28:561–75.

49. MohammedA, JanakiramNB,BrewerM,Duff A, LightfootS, BrushRS,et al. Endogenous n-3polyunsaturated fatty acids delayprogressionofpancreatic ductal adenocarcinoma in Fat-1.P48Cre/þ-LSL-KrasG12D/þ

mice. Neoplasia 2012;14:1249–59.50. Rao CV, Mohammed A, Janakiram NB, Qian L, Ritchie RL, Lightfoot S,

et al. Inhibition of pancreatic intraepithelial neoplasia progression tocarcinomaby nitric oxide-releasing aspirin in p48Cre/þ-LSL-KrasG12D/þ

mice. Neoplasia 2012;14:778–87.






2014;7:1198-1209. Published OnlineFirst September 23, 2014.Cancer Prev Res Altaf Mohammed, Naveena B. Janakiram, Venkateshwar Madka, et al. Modulating Ornithine Decarboxylase SignalingEflornithine (DFMO) Prevents Progression of Pancreatic Cancer by

Updated version

10.1158/1940-6207.CAPR-14-0176doi:

Access the most recent version of this article at:

Material

Supplementary

1

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2014/09/24/1940-6207.CAPR-14-0176.DCAccess the most recent supplemental material at:

Cited articles

http://cancerpreventionresearch.aacrjournals.org/content/7/12/1198.full#ref-list-1

This article cites 48 articles, 16 of which you can access for free at:

Citing articles

http://cancerpreventionresearch.aacrjournals.org/content/7/12/1198.full#related-urls

This article has been cited by 4 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://cancerpreventionresearch.aacrjournals.org/content/7/12/1198To request permission to re-use all or part of this article, use this link



http://cancerpreventionresearch.aacrjournals.org/lookup/doi/10.1158/1940-6207.CAPR-14-0176

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2014/09/24/1940-6207.CAPR-14-0176.DC1

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2014/09/24/1940-6207.CAPR-14-0176.DC1

http://cancerpreventionresearch.aacrjournals.org/content/7/12/1198.full#ref-list-1

http://cancerpreventionresearch.aacrjournals.org/content/7/12/1198.full#related-urls

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

mailto:[email protected]

http://cancerpreventionresearch.aacrjournals.org/content/7/12/1198


eﬂornithine (dfmo) prevents progression of pancreatic ...€¦ · research article eﬂornithine...

Documents