JTE-522, A Selective Cyclooxygenase-2 Inhibitor, Inhibits Induction but not Growthand Invasion of 1,2-Dimethylhydrazine-Induced Tubular Adenocarcinomas of Colonin RatsMin Wei, Keiichirou Morimura, Hideki Wanibuchi, Jun Shen, Elsayed I. Salim, Masaharu Moku, Katsuo Hakoi andShoji Fukushima*

Department of Pathology, Osaka City University Medical School, Osaka, Japan

We have previously demonstrated that JTE-522, a selective cyclo-oxygenase-2 (COX-2) inhibitor, inhibited development of aberrantcrypt foci (ACF) in rats, a putative preneoplastic lesion in colon,and suggested its inhibitory potential in rat colon carcinogenesis.To evaluate the chemopreventive properties of JTE-522, thepresent study was design to evaluate the inhibitory effects ofJTE-522 on rat colon tumorigenesis induced by 1,2-dimethylhy-drazine (DMH). Rats at 6 weeks of age were divided into 4 groups.One week after the start of the experiment, all rats received DMHby s.c. injection at a dose of 40 mg/kg body weight once a week for4 successive weeks. As the initiation and postinitiation treatmentgroups, groups 1–3 were fed diets containing 0, 50, or 150 ppmJTE-522, respectively, from the start of the study to the end. As thepostinitiation treatment group, group 4 was given 150 ppm JTE-522 from 1 week after the last DMH injection to the end of thestudy. Forty weeks after the start of the experiment, administra-tion of 150 ppm JTE-522 during both initiation and postinitiationstages significantly inhibited the incidences of tubular adenocar-cinomas and total carcinomas, as well as total tumors in the colon.The inhibitory effect of JTE-522 was most prominent for tubularadenocarcinomas, but was not observed in the nontubular carci-nomas (signet-ring cell and mucinous carcinomas). Almost equalinhibitory effects on tubular adenocarcinomas were also observedin the rats given 150 ppm JTE-522 during the postinitiation stage,suggesting that its major anticancer action is at the postinitiationphase. However, JTE-522 had no effect on the size or invasiveextent of tubular adenocarcinomas. Furthermore, microarrayanalyses revealed that JTE-522 had no effect on gene expressionlevels in DMH-induced tubular adenocarcinomas. These findingssuggest that JTE-522 possesses chemopreventive activity againstinduction but not progression of tubular adenocarcinomas in ratcolon. In view of the significant inhibitory effects of JTE-522 onACF, its major anticancer action may occur in the postinitiationstage but before the malignant conversion stage of DMH-inducedcolon carcinogenesis.

Key words: JTE-522; selective cyclooxynenase-2 inhibitor; coloncarcinogenesis; tubular adenocarcinoma; chemoprevention

Colorectal cancer leads to approximately 550,000 annual deathsworldwide1 and is therefore a major public health problem andunderlines the need for effective chemopreventive strategies. Theprotective effect of traditional nonsteroidal antiinflammatory drugs(NSAIDs), such as aspirin and sulindac, for colon cancer has beenwell documented in epidemiologic and animal studies.2 However,their use for chemoprevention of colon cancer has been hinderedby their potential gastrointestinal and renal toxicity. It is nowknown that nonselective NSAIDs inhibit activities of cyclooxygen-ases (COX), and inhibition of COX-2 may well explain theirchemopreventive effect, whereas inhibition of COX-1 is believedto be responsible for many of their adverse side effects.3,4 COX-2has been demonstrated to be overexpressed in human colorectalcancers5,6 and in chemically induced colon cancers in rodents,7 andhas been suggested to be an important target for both cancerprevention and/or therapy. Furthermore, recently developed selec-tive COX-2 inhibitors such as celecoxib have been shown tosuppress colon carcinogenesis in experimental animals8–10 andhumans with less toxicity than nonselective NSAIDs.11,12 Selec-tive COX-2 inhibitors have emerged as a promising chemopreven-tion strategy for colorectal cancers.

JTE-522, a novel selective inhibitor of rat and human COX-2,13

has been shown to reduce the number and growth of adenomas inAPC�-474 knockout mice.14,15 This compound also inhibited liverand lung metastases of colon cancer expressing COX-2 but lackedeffect on those lacking COX-2 expression in the nude mousexenograft model.16,17 In chemically induced colon carcinogenesis,we have previously shown that JTE-522 inhibited the developmentof aberrant crypt foci (ACF), the putative precursors of bothhuman and experimental colon cancer,18,19 when administeredthroughout the study.20

In the present study, to evaluate the chemopreventive propertiesof JTE-522, we have investigated the efficacy of JTE-522 duringdifferent stages of rat colon carcinogenesis using colon tumorformation as an endpoint, including its effects on histopathologicpattern and growth, as well as invasion of colon tumors. We havealso compared the gene expression profile of colon cancers inJTE-522-treated rats to those in rats without JTE-522 treatmentusing cDNA microarray analysis.

Material and methodsChemicals

JTE-522 (purity, 100%) was kindly provided by Japan Tobacco(Osaka, Japan). DMH (purity, 99%) was purchased from TokyoKasei (Tokyo, Japan).

AnimalsA total of 108 male F344 rats, 5 weeks old, were purchased from

Charles River Japan (Hino, Japan). The rats were housed in poly-carbonate cages (3/cage) on paper chip bedding in an animal roomtargeted under controlled conditions of a 12-hr light/dark cycle,55% � 5% humidity and 23 � 2°C room temperature. Rats werequarantined for 1 week before the start of study. The rats weregiven free access to the diets and tap water throughout the study.

DietsAll control powdered CE-2 diet and the experimental diets

containing JTE-522 were prepared by Clea Japan (Osaka, Japan)every 8 weeks and were stored at room temperature in the dark.The levels of JTE-522 in the diet were based on the results of ourprevious experiment in which gavage of JTE-522 at a dose of 10mg/kg body weight, but not 3 mg/kg, inhibited DMH-induced ratcolonic ACF formation.20 Doses of 50 and 150 ppm used in thepresent study corresponded to the doses of 3 and 10 mg/kg bodyweight when administered by oral gavage, respectively.

Experimental procedureThe experiments were designed to evaluate the effect of JTE-

522 administered after carcinogen treatment (postinitiation) and

Grant sponsor: the Ministry of Health, Labor and Welfare, Japan.*Correspondence to: Department of Pathology, Osaka City University

Medical School, 1-4-3 Asahi-machi, Abeno-ku, Osaka 545-8585, Japan.Fax: �81-6-6646-3093. E-mail: fukuchan@med.osaka-cu.ac.jp

Received 18 May 2004; Accepted after revision 6 July 2004DOI 10.1002/ijc.20583Published online 28 September 2004 in Wiley InterScience (www.

interscience.wiley.com).

Int. J. Cancer: 113, 354–358 (2005)© 2004 Wiley-Liss, Inc.

Publication of the International Union Against Cancer

throughout the study (initiation and postinitiation periods). Rats at6 weeks of age were randomized into 4 groups of 27 rats each. Oneweek after the start of the experiment, all rats received DMH bys.c. injection at a dose of 40 mg/kg body weight once a week for4 successive weeks. As the initiation and postinitiation treatmentgroups, groups 1–3 were fed diets containing 0, 50, or 150 ppmJTE-522, respectively, from the start of the study to the end of thestudy. As the postinitiation treatment group, group 4 was given150 ppm JTE-522 from 1 week after the last DMH injection to theend of the experiment. Body weight, water and food consumptionwere measured weekly during the experiment.

Forty weeks after the beginning of the study, or earlier in thecase of animals that became moribund, rats were sacrificed underether anesthesia and subjected to macroscopic and histopathologicexaminations of colon, liver and kidney. The colons were quicklyexcised, distended in 0.9% saline, then opened longitudinally.Estimate of tumor volume was determined using the formula V �L � W � D � �/6, where L is length, W is width and D is depthof colon tumor.10 After recording all suspected neoplastic lesionsfor their location, number and size, a portion of large tumors werecut and snap-frozen in liquid nitrogen, then stored at �80°C untilprocessed for RNA extraction. The remaining tumor and colontissues were fixed in 10% phosphate-buffered formalin. The liverand kidney were removed from all animals, weighed and fixed in10% phosphate-buffered formalin. After fixation, all specimenswere processed for paraffin embedding, sectioned and stained withhematoxylin and eosin for histopathologic examination.

Classification of colon tumorsDMH-induced colon tumors were histologically classified as

adenomas and carcinomas. Carcinomas were further classified into2 types: tubular adenocarcinomas and nontubular carcinomas.21,22

The former are well-differentiated gland-forming adenocarcino-mas. The latter show abundant mucus production and are furtherclassified into mucinous and signet-ring cell carcinomas. Muci-nous carcinomas exhibit in mucin pools and low cellularity themajority of the tumor, whereas the signet-ring cell carcinomasdisplay a number of signet-ring cells. In addition, all carcinomaswere classified as invasive or noninvasive. The noninvasive car-cinomas were protruded into the intestinal lumen. Invasive carci-nomas traversed the muscularis mucosae into submucosa layer orbeyond.

COX-2 immunohistochemistrySections (4 �m) were cut from paraffin-embedded colon tissues

and mounted on poly-L-lysine-coated slides. Established proce-dures for the COX-2 immunohistochemical staining by the avidin-biotin-peroxidase complex (ABC) method was used.23 Briefly,endogenous peroxidase activity was blocked with 0.3% H2O2 indistilled water for 5 min, then sections were trypsinized at 37°C in0.4% trypsin, 0.01% CaCl2 in Tris-buffered saline for 30 min.Antigen retrieval was performed by microwaving at 98°C for20 min in 0.01 M citrate buffer (pH 6.0). After blocking nonspe-cific binding with serum at 37°C for 30 min, sections were incu-bated with mouse monoclonal COX-2 antibody (TransductionLaboratories, Lexington, KY) at 1:500 dilution overnight at 4°C.Immunoreactivity was detected using a Vectastain Elite ABC Kit

(PK-6102; Vector Laboratories, Burlingame, CA) and 3,3-diami-nobenzidine hydrochloride (Sigma Chemical, St. Louis, MO) fol-lowed by counterstaining with Mayer’s hematoxylin. A negativecontrol was included with each staining procedure by omitting theprimary antibody. Normal epithelium consistently demonstratedno COX-2 immunoreactivity, providing a negative internal controlfor each specimen. Overexpression of COX-2 in colon tumors wasdefined as positive when cytoplasmic staining was evident in� 50% of the cells.23

cDNA microarray analysisInhibitory effects of JTE-522 were prominently observed in

tubular adenocarcinomas. Therefore, only this type of carcinomawas selected for microarray analysis. Total RNA was extractedfrom 10 tubular adenocarcinomas from 10 and 6 rats of groups 1(DMH treatment alone) and group 3 (treatment with DMH and 150ppm JTE-522 during initiation and postinitiation stages), respec-tively, using Isogen RNA Isolation Kit (Nippon Gene, Toyama,Japan). The integrity of total RNA was confirmed by agarose gelanalysis. The Atlas Rat 1.2 array containing 1,176 genes wasperformed according to the manufacturer’s instructions (ClontechLaboratories, Palo Alto, CA). The complete listing of the genes isavailable on the Web (http://www.bdbiosciences.com/clontech/atlas/genelists). All of the following reagents used were purchased fromClontech Laboratories. Briefly, 30 �g of total RNA of each groupprepared by pooling equal amounts of RNA from each cancer wereconverted to 32P-labeled cDNA probes using the Atlas Pure RNALabeling Kit. Labeled probes were further purified by NucleoSpinExtraction Column, then were hybridized to the Atlas Rat 1.2 arraymembranes separately. After a high-stringency wash, the arrayswere exposed to a phosphoimage plate for 1 day. The phospho-image plate was scanned with a BAS-2500 Imager (Fujifilm Med-ical Systems, Stamford, CT); expression changes between the 2arrays were analyzed densitometrically using AtlasImage version2.0, according to the manufacturer’s instructions. To minimize theeffects of measurement variation introduced by artificial sourcesduring the experiments, only genes that showed expression levelsmore than 2-fold difference between the 2 arrays were consideredas possible JTE-522-responsive genes.

To validate the gene expression data, RT-PCR using RNA ofindividual cancers was performed. Briefly, cDNA synthesis wasperformed with 1 �g RNA using Advantage RT-for-PCR kit(Clontech Laboratories). Amplification of the cDNAs was carriedout using Titanium Taq PCR Kits (Clontech Laboratories). Allsequences of the primers were purchased from Clontech Labora-tories. The PCR reaction conditions and cycle numbers wererigorously adjusted for each product to ensure that all reactionswere within the linear range of product amplification. PCR prod-ucts were electrophoresed through 2% agarose gels. Gene expres-sion levels were semiquantitated by densitometry of the DNAbands using a FluorImager (Molecular Dynamics, CA) with ImageQuant 5.1 software and normalized to -actin.

Statistical analysisAll mean values were expressed as mean � SD. All statistical

analyses were performed using StatView J-5.0 software (AbacusConcepts, Berkeley, CA) for the Macintosh computer. Group

TABLE I – FINAL BODY, LIVER AND KIDNEY WEIGHTS

Group Treatment Final body weight(g)

Liver weight Kidney weight

Absolute(g)

Relative(%)

Absolute(g)

Relative(%)

1 DMH alone 365 � 19 8.2 � 0.7 2.3 � 0.1 2.2 � 0.1 0.6 � 0.02 DMH � 50 ppm JTE-5221 360 � 24 8.4 � 0.7 2.3 � 0.2 2.2 � 0.2 0.6 � 0.13 DMH � 150 ppm JTE-5221 346 � 283 8.1 � 1.2 2.4 � 0.5 2.2 � 0.2 0.6 � 0.04 DMH 3 150 ppm JTE-5222 352 � 263 8.1 � 0.8 2.3 � 0.3 2.2 � 0.2 0.6 � 0.11JTE-522 was administered throughout the study.–2JTE-522 was administered from 1 week after the last DMH treatment.–3Significantly

different from group 1.

355JTE-522

means for body and organ weights, tumor multiplicity (number oftumors/animal) and tumor volumes were compared by Student’st-test or Welch’s t-test. Tumor incidence was analyzed by chi-square probability analysis. Differences were considered as statis-tically significant at p � 0.05.

Results

There were 4 animals, 1 from each group, that died due toocclusive ileus secondary to the development of mucinous orsignet-ring cell carcinomas in the proximal colon before the end ofthe study. As shown in Table I, final average body weights ingroups treated with 150 ppm JTE-522 (groups 3 and 4) weresignificantly lower than in the control group. There were nosignificant differences in absolute and relative weights of liver orkidney among groups. JTE-522 did not produce any macroscopicor microscopic changes in liver and kidney. JTE-522 treatment hadno effects on food consumption, but caused a significant increasein water consumption (data not shown). The increased waterconsumption and decreased body weight in 150 ppm JTE-522groups are probably due to its diuretic effect through inhibition ofcarbonic anhydrase (data not shown).

Incidences and multiplicities of colon tumorsTable II summarizes the incidences and multiplicities of DMH-

induced colon tumors. Histopathologically, the majority of DMH-induced colon cancers in rats were tubular adenocarcinomas. Ascompared to the control (group 1), administration of 150 ppmJTE-522 during initiation and postinitiation stages (group 3) sig-nificantly inhibited the incidences of tubular adenocarcinomas,total carcinomas, as well as total tumors. It is noteworthy that theinhibitory effects of JTE-522 were confined to tubular adenocar-cinomas. Significant inhibition of tubular adenocarcinomas wasalso observed in the rats given 150 ppm JTE-522 during postini-tiation (group 4). JTE-522 has no effects on nontubular carcinomas(signet-ring cell and mucinous carcinomas). With regard to mul-tiplicities of colon tumors, there was a significant inhibition in themultiplicities of tubular adenocarcinomas, total carcinomas andtotal tumors of the colon in group 4. Similar suppressive effectswere also observed in group 3, but did not reach statistical differ-ence. Administration of 150 ppm JTE-522 during postinitiationstage is equally as effective on inhibiting tubular adenocarcinomaswhen it was administrated during both initiation and postinitiationstages. Administration of 50 ppm JTE-522 (group2) had no effecton the incidences or multiplicities of any types of colon tumor.

Size and invasion of colon cancersTable III shows the data on size and invasion of colon carcino-

mas. All nontubular carcinomas were invasive and observed in thearea of solitary lymphoid aggregates. These carcinomas were notincluded for size comparison because it was difficult to examineaccurately the size. Although 150 ppm JTE-522 showed inhibitoryeffects on the incidence of tubular adenocarcinomas, it did notaffect size or invasive extent.

COX-2 immunohistochemistryColonic epithelial cells of normal epithelium and all adenomas

did not show COX-2 immunoreactivity in all groups. As shown inTable IV, positive COX-2 staining was detected in the cytoplasmof tumor cells in tubular adenocarcinomas. In contrast, none of themucinous or signet-ring cell adenocarcinomas showed COX-2overexpression. These findings suggest that tubular adenocarcino-mas showed a significantly higher COX-2 expression level thannontubular carcinomas. However, JTE-522 treatment did not affectthe COX-2 staining patterns among groups.

cDNA microarray analysisGene expression profiles of tubular adenocarcinomas in rats treated

with and without JTE-522 were compared using Rat 1.2 cDNAmicroarray, which includes 1,176 known genes involved in prolifer-ation, cell cycle regulation, apoptosis, angiogenesis and metastasis.

TA

BL

EII

–E

FFE

CT

OF

JTE

-522

ON

DM

H-I

ND

UC

ED

CO

LO

NC

AR

CIN

OG

EN

ESI

SIN

MA

LE

F344

RA

TS

Gro

upT

reat

men

tN

umbe

rof

rats

Inci

denc

e(%

)M

ultip

licity

(num

ber

ofco

lon

tum

ors/

rats

)

Ade

nom

aC

arci

nom

aT

otal

tum

ors

Ade

nom

aC

arci

nom

aT

otal

tum

ors

Tub

ular

1N

ontu

bula

r2T

otal

Tub

ular

1N

ontu

bula

r2T

otal

1D

MH

alon

e27

4(1

4.8)

21(7

7.8)

4(1

4.8)

25(9

2.6)

25(9

2.6)

0.15

�0.

361.

15�

0.86

0.15

�0.

361.

30�

0.72

1.44

�0.

852

DM

H�

50pp

mJT

E-5

223

271

(3.7

)22

(81.

5)5

(18.

5)24

(88.

9)26

(96.

2)0.

04�

0.19

1.19

�0.

880.

22�

0.51

1.41

�1.

051.

44�

1.01

3D

MH

�15

0pp

mJT

E-5

223

273

(11.

1)11

(40.

7)5

6(2

2.2)

15(5

5.6)

517

(63.

0)5

0.11

�0.

320.

74�

1.13

0.22

�0.

420.

96�

1.16

1.07

�1.

14

4D

MH3

150

ppm

JTE

-522

427

3(1

1.1)

14(5

1.9)

56

(22.

2)20

(74.

1)20

(74.

1)0.

11�

0.32

0.63

�0.

695

0.22

�0.

420.

85�

0.60

50.

96�

0.76

5

1T

ubul

arad

enoc

arci

nom

a.–2

Non

tubu

lar

carc

inom

ain

clud

ing

sign

et-r

ing

cell

and

muc

inou

sca

rcin

omas

.–3JT

E-5

22w

asad

min

iste

red

thro

ugho

utth

est

udy.

–4JT

E-5

22w

asad

min

iste

red

from

1w

eek

afte

rth

ela

stD

MH

trea

tmen

t.–5Si

gnifi

cant

lydi

ffer

ent

from

grou

p1.

356 WEI ET AL.

RNA used for microarray analysis was pooled from 10 individualtubular adenocarcinomas of each group; the expression profiles, there-fore, represent an average of gene expression level of these adeno-carcinomas. Although gene expression data showed that some genesexhibited more than 2-fold differences in mean level between the 2groups, the differences were not found to be statistically significantwhen validated by RT-PCR using RNA from individual cancers (datano shown). The changed average levels appear to be due to the greatchanges in a couple of samples. These results indicate that JTE-522lacks the capability of modifying gene expression level in DMH-induced colon adenocarcinomas.

Discussion

The results of the present study demonstrate that the adminis-tration of the selective COX-2 inhibitor JTE-522 at the level of150 ppm during both initiation and postinitiation stages signifi-cantly inhibited the induction of DMH-induced rat colon carcino-mas, predominantly in tubular adenocarcinomas. These resultsconfirm our previous findings that JTE-522 significantly inhibitedthe development of ACF, the putative precursors of both humanand experimental colon cancer,18,19 when administered duringinitiation and postinitiation stages of rat colon carcinogenesis.20

Furthermore, administration of JTE-522 during the postinitiationstage is almost equally as effective on inhibiting tubular adeno-carcinomas as when it was administrated during both initiation andpostinitiation stages, suggesting that JTE-522 acts to suppressadenocarcinoma formation mainly during the postinitiation phase.These findings indicate that JTE-522 possesses chemopreventiveactivity against colon carcinogenesis of rats.

With regard to the chemopreventive effects of NASIDs duringdifferent stages of colon carcinogenesis, the efficacy of celecoxiband piroxicam has been found to be primarily in the later stages ofcarcinogenesis, as evidenced by the findings that administration ofthese agents after formation of adenomas and adenocarcinoma(beginning 13 or 14 weeks after the last colon carcinogenazoxymethane treatment, respectively) were equally effective onthe incidence and size of colon tumors when they were adminis-trated during initiation and postinitiation stages.9,24 In the presentstudy, JTE-522 inhibited the incidence but not the size or invasiveextent of tubular adenocarcinomas, indicating that it lacks thecapability of retarding the growth of adenocarcinomas. Further-more, it was anticipated that the gene expression profile would be

different between colon tumors from rats treated with or withoutJTE-522 due to a selective environment. However, we did notdetect any significant differences in gene expression levels be-tween the tubular adenocarcinomas with or without JTE-522 treat-ment using microarray analysis. This result indicated that JTE-522has no effect on gene expression levels in DMH-induced colonadenocarcinomas. Therefore, these findings suggest that JTE-522lacks inhibitory effects on the progression of colon adenocarcino-mas in rats in the present model. Considering the significantinhibitory effects of JTE-522 on ACF, the putative precursors ofboth human and experimental colon cancer, its major anticanceraction may occur at an early stage of DMH-induced colon carci-nogenesis. In terms of the administration stage that shows inhib-itory effects on colon carcinogenesis, the effects of JTE-522 arequite different from celecoxib and piroxicam. Further studies eval-uating the efficacy of JTE-522 on the ACF-adenoma and theadenoma-carcinoma process are warranted to evaluate this com-pound for its chemopreventive properties.

As indicated above, the observed inhibitory effects of JTE-522in the present model were predominantly confined on tubularadenocarcinomas. Although it is difficult to reach the conclusionthat JTE-522 may not be an effective chemopreventive agent onsignet-ring cell and mucinous carcinomas in rats due to limitednumbers of these cancers, results on COX-2 immunohistochemis-try provide an indication of selective inhibitory effects. Immuno-histochemical analysis revealed that COX-2 was associated withtubular adenocarcinomas in contrast to nontubular colon carcino-mas, in which COX-2 protein expression was almost undetectable.In the mouse model, JTE-522 has been found to inhibit liver andlung metastases of colon cancer that show COX-2 expression butlack an effect on those lacking COX-2 expression.16,17 Consider-ing that COX-2 is the best-known target for NSAIDs in coloncarcinogenesis, lack of inhibitory effects of JTE-522 on signet-ringcell and mucinous carcinomas could be explained at least in partby a lack of COX-2 expression in those cancers. Furthermore, Parket al.25 have suggested that tubular adenocarcinoma follows theACF-adenoma-carcinoma sequence, whereas poorly differentiatedcarcinoma may arise de novo without passing through the stage ofACF in DMH-induced colon carcinogenesis in rats. The differentpathways followed by the 2 kinds of colon carcinomas may alsoprovide a possible explanation for selective effects of JTE-522 inthe present model.

TABLE III – EFFECT OF JTE-522 ON SIZE AND INVASIVE EXTENT OF DMH-INDUCED COLON CARCINOMAS IN MALE F344 RATS

Group Treatment Number of tubularadenocarcinomas1 Size (mm3)

Number of invasive carcinoma/number of carcinomaobserved (%)

Tubular2 Nontubular3 Total

1 DMH � control diet 28 83.4 � 140.1 12/31 (38.7) 4/4 (100) 16/35 (45.7)2 DMH � 50 ppm JTE-5224 27 64.1 � 120.0 10/32 (31.3) 6/6 (100) 16/38 (42.1)3 DMH � 150 ppm JTE-5224 20 65.0 � 86.7 6/20 (30.0) 6/6 (100) 12/26 (46.2)4 DMH 3 150 ppm JTE-5225 15 65.9 � 117.9 5/17 (29.4) 6/6 (100) 11/23 (47.8)1Number of well-differentiated adenocarcinoma that size was measured at sacrifice.–2Tubular adenocarcinoma.–3Nontubular carcinoma

including signet-ring cell and mucinous carcinomas.–4JTE-522 was administered throughout the study.–5JTE-522 was administered from 1 weekafter the last DMH treatment.

TABLE IV – EFFECTS OF JTE-522 ON COX-2 EXPRESSION OF DMH-INDUCED COLON CARCINOMASIN MALE F344 RATS

Group TreatmentNumber of COX-2 overexpression carcinoma/number of

carcinoma observed (%)

Tubular1 Nontubular2 Total

1 DMH alone 17/31 (54.8) 0/4 (0) 17/35 (48.6)2 DMH � 50 ppm JTE-5223 21/32 (65.6) 0/6 (0) 21/38 (55.3)3 DMH � 150 ppm JTE-5223 13/20 (65.0) 0/6 (0) 13/26 (50.0)4 DMH 3 150 ppm JTE-5224 11/17 (64.7) 0/6 (0) 11/23 (47.8)1Tubular adenocarcinoma.–2Nontubular carcinoma including signet-ring cell and mucinous carcino-

mas.–3JTE-522 was administered throughout the study.–4JTE-522 was administered from 1 week after thelast DMH treatment.

357JTE-522

In conclusion, the present study demonstrates that the selectiveCOX-2 inhibitor JTE-522 significantly inhibited induction ofDMH-induced tubular adenocarcinomas in rats. Lack of effects ofJTE-522 on growth and invasion of adenocarcinomas suggests thatits major anticancer action may occur before the malignant con-version stage of DMH-induced colon carcinogenesis.

Acknowledgements

The authors thank Japan Tobacco for providing the JTE-522,Kaori Touma for technical assistance and Mari Dokoh and YukoOnishi for assistance with the preparation of the manuscript (De-partment of Pathology, Osaka City University Medical School).

References

1. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statistics, 2000.CA Cancer J Clin 2000;50:7–33.

2. Janne PA, Mayer RJ. Chemoprevention of colorectal cancer. N EnglJ Med 2000;342:1960–8.

3. Mitchell JA, Akarasereenont P, Thiemermann C, Flower RJ, Vane JR.Selectivity of nonsteroidal antiinflammatory drugs as inhibitors ofconstitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA1993;90:11693–7.

4. Needleman P, Isakson PC. The discovery and function of COX-2.J Rheumatol 1997;24(Suppl 49):6–8.

5. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S,DuBois RN. Up-regulation of cyclooxygenase 2 gene expression inhuman colorectal adenomas and adenocarcinomas. Gastroenterology1994;107:1183–8.

6. Kutchera W, Jones DA, Matsunami N, Groden J, McIntyre TM,Zimmerman GA, White RL, Prescott SM. Prostaglandin H synthase 2is expressed abnormally in human colon cancer: evidence for atranscriptional effect. Proc Natl Acad Sci USA 1996;93:4816–20.

7. DuBois RN, Radhika A, Reddy BS, Entingh AJ. Increased cycloox-ygenase-2 levels in carcinogen-induced rat colonic tumors. Gastroen-terology 1996;110:1259–62.

8. Kawamori T, Rao CV, Seibert K, Reddy BS. Chemopreventive ac-tivity of celecoxib, a specific cyclooxygenase-2 inhibitor, againstcolon carcinogenesis. Cancer Res 1998;58:409–12.

9. Reddy BS, Hirose Y, Lubet R, Steele V, Kelloff G, Paulson S, SeibertK, Rao CV. Chemoprevention of colon cancer by specific cyclooxy-genase-2 inhibitor, celecoxib, administered during different stages ofcarcinogenesis. Cancer Res 2000;60:293–7.

10. Rao CV, Indranie C, Simi B, Manning PT, Connor JR, Reddy BS.Chemopreventive properties of a selective inducible nitric oxide syn-thase inhibitor in colon carcinogenesis, administered alone or incombination with celecoxib, a selective cyclooxygenase-2 inhibitor.Cancer Res 2002;62:165–70.

11. Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, WheltonA, Makuch R, Eisen G, Agrawal NM, Stenson WF, Burr AM, ZhaoWW, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidalanti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: theCLASS study, a randomized controlled trial. JAMA 2000;284:1247–55.

12. Steinbach G, Lynch PM, Phillips RK, Wallace MH, Hawk E, GordonGB, Wakabayashi N, Saunders B, Shen Y, Fujimura T, Su LK, LevinB. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familialadenomatous polyposis. N Engl J Med 2000;342:1946–52.

13. Matsushita M, Masaki M, Yagi Y, Tanaka T, Wakitani K. Pharma-

cological profile of JTE-522, a novel prostaglandin H synthase-2inhibitor, in rats. Inflamm Res 1997;46:461–6.

14. Sasai H, Masaki M, Wakitani K. Suppression of polypogenesis in anew mouse strain with a truncated Apc (Delta474) by a novel COX-2inhibitor, JTE-522. Carcinogenesis 2000;21:95–8.

15. Sunayama K, Konno H, Nakamura T, Kashiwabara H, Shoji T,Tsuneyoshi T, Nakamura S. The role of cyclooxygenase-2 (COX-2) intwo different morphological stages of intestinal polyps in APC (Del-ta474) knockout mice. Carcinogenesis 2002;23:1351–9.

16. Tomozawa S, Nagawa H, Tsuno N, Hatano K, Osada T, Kitayama J,Sunami E, Nita ME, Ishihara S, Yano H, Tsuruo T, Shibata Y, et al.Inhibition of haematogenous metastasis of colon cancer in mice by aselective COX-2 inhibitor, JTE-522. Br J Cancer 1999;81:1274–9.

17. Nagatsuka I, Yamada N, Shimizu S, Ohira M, Nishino H, Seki S,Hirakawa K. Inhibitory effect of a selective cyclooxygenase-2 inhib-itor on liver metastasis of colon cancer. Int J Cancer 2002;100:515–9.

18. Bird RP. Role of aberrant crypt foci in understanding the pathogenesisof colon cancer. Cancer Lett 1995;93:55–71.

19. Takayama T, Katsuki S, Takahashi Y, Ohi M, Nojiri S, Sakamaki S,Kato J, Kogawa K, Miyake H, Niitsu Y. Aberrant crypt foci of thecolon as precursors of adenoma and cancer. N Engl J Med 1998;339:1277–84.

20. Wei M, Morimura K, Wanibuchi H, Shen J, Doi K, Mitsuhashi M,Moku M, Salim EI, Fukushima S. Chemopreventive effect of JTE-522, a selective cyclooxygenase-2 inhibitor, on 1, 2-dimethylhydra-zine-induced rat colon carcinogenesis. Cancer Lett 2003;202:11–6.

21. Sunter JP, Appleton DR, Wright NA, Watson AJ. Pathological fea-tures of the colonic tumours induced in rats by the administration of1,2-dimethylhydrazine. Virch Arch B Cell Pathol 1978;29:211–23.

22. Sunter JP, Hull DL, Appleton DR, Watson AJ. Cell proliferation ofcolonic neoplasms in dimethylhydrazine-treated rats. Br J Cancer1980;42:95–102.

23. Wei M, Wanibuchi H, Morimura K, Iwai S, Yoshida K, Endo G,Nakae D, Fukushima S. Carcinogenicity of dimethylarsinic acid inmale F344 rats and genetic alterations in induced urinary bladdertumors. Carcinogenesis 2002;23:1387–97.

24. Reddy BS, Maruyama H, Kelloff G. Dose-related inhibition of coloncarcinogenesis by dietary piroxicam, a nonsteroidal antiinflammatorydrug, during different stages of rat colon tumor development. CancerRes 1987;47:5340–6.

25. Park HS, Goodlad RA, Wright NA. The incidence of aberrant cryptfoci and colonic carcinoma in dimethylhydrazine-treated rats varies ina site-specific manner and depends on tumor histology. Cancer Res1997;57:4507–10.

358 WEI ET AL.