effect of chemopreventive agents on intermediate biomarkers … · vol. 1, 405-4 1 1, july/august i...

Vol. 1, 405-4 1 1, July/August I 992 Cancer Epidemiology, Biomarkers & Prevention 405

Effect of Chemopreventive Agents on Intermediate Biomarkersduring Different Stages of Azoxymethane-induced Colon

Carcinogenesis1

Jagveer Singh, Gary Kelloff, and Bandaru S. Reddy2Aniie’ricani tlealth Foundation, t)mvmsenni mmtNutrmti(mnal C,mrcmnogenesis,Valhalla, New York 10595 [1. S., B. S. R.], and National Cancer Institute’,

Cheniopre’ventmon Branch, Division (mt Cancer Control ,mnd Pre’ve’ntion,Bethe’sdla, M,mrvlaiid 20892 [G. K.]

Abstract

Chemoprevention of colon cancer is emerging as analternative to therapy with a broad potential forreducing cancer incidence in defined high-risk groupsand the general population. Besides severalchemopreventive agents in use and under investigation,D,L-a-difluorOmethylOrnithine (DFMO) and piroxicamhave been shown to effectively inhibit coloncarcinogenesis in rodents. A variety of proliferation-related parameters have been suggested as potentialintermediate markers of cancer risk that could be usedto monitor the progress of chemoprevention in clinicaltrials. We have investigated the effect ofchemopreventive agents, DFMO, and piroxicam onmucosal ornithine decarboxylase (ODC) and tyrosine-specific protein kinase (TPK) activities during differentstages of azoxymethane (AOM)-induced coloniccarcinogenesis in male F344 rats in order to examinethe plausibility of using these enzymes as intermediatebiochemical markers of colon cancer. Groups of maleF344 rats were fed modified AIN-76A diets containing0 or 150 ppm piroxicam or 4000 ppm DFMO and givens.c. injections of AOM dissolved in normal saline at adose of 1 5 mg/kg body weight/week, once weekly, for4 weeks. Vehicle control groups received s.c. equalvolumes of normal saline. Groups of animals were thensacrificed at 0, 4, 16, 24, and 32 weeks after AOM orsaline treatment, and their colonic mucosa wasanalyzed for ODC and TPK activities. AOM treatmentsignificantly increased mucosal ODC as well as TPKactivities. AOM-induced ODC and TPK activities weresignificantly suppressed by dietary DFMO progressivelyat all stages of colon carcinogenesis. Dietary piroxicamincreased AOM-induced colonic mucosal ODC andTPK activities but significantly reduced tumor incidenceas well as tumor multiplicity. DFMO exerted a morepronounced inhibitory effect on AOM-induced colontumor development. These results emphasize the

Received 9/25/91.‘ Supporte’dI by USPHS Contract NO1-CN-85095-01 and Grant CA17613froni the National Caner Institute.2 To whom requests for re’prinits should he addressed, at American Health

Foundation, Division of Nutritional C,ircinogene-’sis, One’ Dana Road,Valhalla, NY 10595.

importance of development of agent-specificintermediate biomarkers to be used as effectivepredictors of colon carcinogenesis.

Introduction

Colonectal cancer is a major cause of morbidity andmortality among men and women in the United States,afflicting 167,000 people and causing about 65,000deaths in 1991 (1). Despite several advances made in thetreatment of colon cancer, the cure rate has remainedlargely unchanged in the past two decades (2). Chemo-prevention, which is emerging as an alternative to then-apy, relies on the concept of inhibition or reversal ofcancer formation through chemical intervention (3, 4).Several chemopreventive agents in use and under inves-

tigation include retinoids, purified dietary constituents,vitam ins, m icron utnients, anti m utagens, d ithiolthiones,metabolic products, protease inhibitors, and polyaminesand prostaglandin synthesis inhibitors. DFMO,3 a spe-cific, enzyme-activated, irreversible inhibitor of ODCactivity, and piroxicam, a nonsteroidal, antiinflammatorydrug have been shown to effectively inhibit carcinogen-induced colon tumor development in rodents (5, 6).Recent pneclinical studies in laboratory animals and din-ical investigations in high-risk patients of colon cancerhave identified DFMO and pinoxicam as ideal candidatesfor chemopnevention trials. However, the success of suchclinical trials has been hampered by the lack of appro-pniate intermediate biomankers or end points whichcould indicate the progress of chemoprevention andpredict the subsequent reduction of cancer occurrence.

Intermediate end points may be defined as meas-

unable markers of cellular or molecular events associatedwith specific stages of the multistep evolution and pro-gression of carcinogenesis. This indicates that the risk of

carcinogenic transformation must correlate with quanti-tative degree and pattern of biomarker expression. ODC,which catalyzes the conversion of omnithine to putrescinein a rate-limiting step of crucial polyamine biosynthesis,plays an important role in normal and neoplastic cellproliferation (7) and is associated with the tumor-pro-moting ability of a variety of agents (7-8). Increased ODCactivity has been demonstrated in benign colonic ade-nomas (9-12) as well as normal-appearing mucosa adja-

cent to adenomas (13-15), reflecting the underlying hy-

perproliferative state of colonic mucosa. Significantly

3 The abbreviations used are: DFMO, o,c-ce-difluoromethylornithine;

ODC, ornithine decarboxylase; TPK, tyrosine-specific protein kinase;AOM, azoxyniethane; DTT, dithiothreitol.

on May 3, 2020. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

http://cebp.aacrjournals.org/

406 Intermediate Biomarkers of Colon Cancer

high levels of ODC activity have been reported for 5ev-eral rodent carcinogenesis models (16-17).

TPKs which catalyse phosphorylation of tyrosine res-idues of protein substrates are important growth-pro-moting signal-transducing enzymes involved in cellulargrowth and differentiation (18-19). Tyrosine phosphor-ylation is now recognized as an important regulatorymechanism in response to a number of processes, in-cluding the action of several growth factors and onco-genes. Enhanced TPK activity has been demonstrated inneunoblastoma (20), rhabdomyosarcoma (21), and mam-mary (22) and colon (23) carcinomas. A significant num-ber of TPKs that are products of protooncogenes havebeen implicated in the etiology of a number of humancancers (24).

Do these intermediate biomarkers, ODC and TPK,predict the known chemopreventive influence of DFMOand pinoxicam on experimental colon tumor develop-ment? This study was designed to investigate the effectof chemopreventive agents, DFMO and piroxicam, onmucosal ODC and TPK activities during different stagesof AOM-induced colon carcinogenesis in male F344 ratsin order to examine the plausibility of using these en-zymes as intermediate biochemical markers for coloncancer. The major goal of this study was to determinewhich intermediate biomarkers, ifaltered by these chem-opreventive agents, could be used as effective predictorsof colon cancer.

Materials and Methods

Animals, Diets, and Carcinogen. Weanling male F344rats were obtained from Charles River Breeding Labora-tories (Kingston, NY). Piroxicam and DFMO were gen-erous gifts from Pfizer Central Research (Groton, CT) andMerrill-Dow Research Institute (Cincinnati, OH), respec-tively. AOM (COS: 25843-45-2) was purchased from Ash-

Stevens (Detroit, MI). All ingredients of semipunified dietwere obtained from Dyets, Inc. (Bethlehem, PA) andstored at 4#{176}Cprior to preparation of the diets. Piroxicamand DFMO were incorporated into the control diet atdose levels of 150 ppm and 4000 ppm, respectively. Allcontrol and experimental diets were prepared weekly inour laboratory and stored in a cold room.

Male F344 rats received at weaning were quaran-tined for 10 days. All animals were housed in plasticcages with filter tops under controlled environmentalconditions of 21 #{176}Ctemperature, 50% humidity, and a1 2-h light/dark cycle. At 5 weeks of age, all animals were

assigned to AOM-treated and vehicle-treated groups andfed the control diet (Ref. 25; Table 1).

Experimental Procedure. Beginning at 6 weeks of age,animals from AOM- and vehicle-treated groups werereallocated into different dietary subgroups and fed one

of the experimental diets containing 1 50 ppm piroxicamor 4000 ppm DFMO or control diet (Table 1). The ration-ale for selecting 150 ppm piroxicam and 4000 ppmDFMO has been based on our previous results, whichindicated a colon tumor-inhibitory effect of these agentsat these levels (5). One week later, animals intended forcarcinogen treatment were administered s.c. AOM dis-solved in normal saline at a dose rate of 15 mg/kg bodyweight/week, once weekly, for 4 weeks. Animals in-tended for vehicle treatment received s.c. an equal vol-ume of normal saline. Animals were maintained on their

Table 1 Pe’rce’ntage’ ( �mnip (msition ol experiniiental se’niipurmtue’d clue’ts

Diet mnigreclie’nts

Coniposit (on

Expe’rime’nit,il die’ts”Cumnitrul diet”

Pirumxicani 1)FMO

Casein 20.() 20.0 20.0

u,L-Methionin(-’ 0. 1 0.3 0.3Cornstarch 52.0 5 1 .985 5 1 .6

Dextrose 1.1.0 1 1.0 13.0

Corn oil 5.0 5.0 5.0

Alphacel 5.0 5.0 5.0

Mineral mix, AIN-76A 3.5 3.5 .1.5Vilaniin niix, AIN�76A 1 .0 1 .0 1.0

Choline hitartrate 0.2 0.2 0.2

Piroxic:ani 0 0.01 5 0

DFMO 0 (�) 0.4

a Adlopted from Anie’rican Institute’ ut Nutrition ReIe’re’nice’ Duet )AIN-76A),

xvith the niocliticatiun 01 sunirc e (If arl)umhydlrate.

S Piroxicam arid DFMO were added to the’ dlie’ts at the’ expense ofcornstarch.

respective dietary regimens until termination of the ex-peniment. Body weights were recorded once weeklyduring carcinogen treatment and then every 4 weeks.Ten animals treated with AOM or vehicle from eachdietary subgroup were sacrificed by decapitation at 0, 4,16, 24, and 32 weeks after the last AOM or salineinjection. Their colons were resected and opened longi-tudinally, and the contents were flushed with saline.Colon tumors, if any, were removed before scraping thecolon mucosa with a microscope slide. Mucosal scrapingsand tumors were quick-frozen in liquid nitrogen andstored at -80#{176}C until use. No colon tumors were ob-served until 16 weeks after the AOM or vehicletreatment.

Determination of ODC Activity. Mucosal scrapings andtumor tissue were homogenized separately with polytronhomogenizer in 10 volumes of 50 m�i sodiuni phosphatebuffer, pH 7.2 (containing 5 m�i DTT, 0.2 mrvt pynidoxalphosphate, 0.1 mM EDTA), and centrifuged at 40,000 xg for 30 mm at 4#{176}C.Clear supemnatant fraction was usedfor ODC assay. Enzyme activity was determined by meas-uring -tmCO liberated from L�[1�mC]ornithine as de-scnibed elsewhere (26). The standard assay niixture con-tamed, in a total volume of 0.25 ml, an aliquot of super-natant, 50 mM sodium phosphate buffer (pH 7.2), 0.2mM pynidoxal phosphate, 5 m�i DTT, 0.1 mtvt EDTA, 0.4mM L-ornithine, and 0.50 zCi D,L-[1-4C]ornithine hydro-chloride (56.6 mCi/mmol; Amersham Co., ArlingtonHeights, IL). The reaction mixture was incubated at 37#{176}Cfor 1 h in a 16 x 100 mm glass tube sealed with rubberstopper supporting a center well (Kontes, Morton Grove,IL). The released 4C02 was trapped on a microglass-fiber filter disc (934-AH, GFA; Whatman) soaked in sat-

urated solution of barium hydroxide. The reaction wasstopped by injecting 0.1 ml of 2 N sulfuric acid througha rubber septum directly into the reaction mixture. Theincubation was continued for an additional 1 h to com-pletely trap the released 4C02. Filter paper along withthe center well was then transferred to a scintillation vial,and radioactivity was counted in 10 nil of scintillationcocktail (Scintisol; ISOLAB, Inc.) ODC activity is ex-pressed as pmol iCo) released/mg protein/h.

Determination of Protein Tyrosine Kinase Activity. TPKactivity was measured in both cytosolic and membrane



Cancer Epidemiology, Biomarkers & Prevention 407

Table 2 Body weights of male F344 rats treated with AOM or vehicle and fed control or experimental diets

Experimental groupBody weight Ig) at weeks after carcinogen/vehicle treatmenta

0 4 16 24 32

AOM-treatedControl diet 216 ± 4.2 265 ± 1.8 357 ± 3.7 379 ± 4.2 398 ± 10.1Piroxicam, 150 I)Pfli 216 ± 2.2 263 ± 2.1 358 ± 3.5 382 ± 4.8 402 ± 7.7DFMO, 4000 PP� 210 ± 1.8 258 ± 2.2 352 ± 3.7 372 ± 5.3 399 ± 6.8

Vehicle-treate’d

Control die’t 239 ± 1.6 280 ± 4.3 379 ± 3.8 414 ± 4.5 436 ± 7.1

Piroxicani, 150 ppm 238 ± 2.2 285 ± 2.2 385 ± 3.9 413 ± 5.2 433 ± 7.6DFMO, 4000 Ppm 235 ± 3.2 279 ± 2.6 377 ± 3.2 410 ± 5.9 428 ± 7.4

a Values are mean ± SEM.

fractions. Colonic mucosal or tumor samples were ho-mogenized in 5 volumes of 10 mrvs Tris-HCI, pH 7.5,containing 0.25 M sucrose, 1 mM MgCI2, 1 m�i EDTA, 1mM DII, 1 mM phenylmethylsulfonyl fluoride, and 0.5mg/mi aprotinin. All debris and nuclei were removed bycentnifugingat 800 X gfor 10 mm at 4#{176}C.The supernatantwas further centrifuged at 40,000 X g for 30 mm at 4#{176}C,

and the resulting supernatant fraction was used as cytosolfor assaying TPK activity. The 40,000 X g pellet (mem-

brane fraction) was resuspended in 50 mr�.i Tris-HCI, pH7.5, containing 20 mM magnesium acetate, 5 m�t NaF, 1

mM EDTA/ethyleneglycol-bis-(f3-aminoethyl ether)-N,N-tetraacetic acid, 1 mt’�i DTT, 30 �tM sodium-o-vanadate,and 0.5% Nonidet P-40; sonicated three times for 10 seach time; and centrifuged at 40,000 x g for 30 mm at4#{176}C.The supernatant (solubilized membrane) was usedto assay TPK activity. TPK activity was measured using atyrosine kinase-specific synthetic polymer as substrate(27) according to the method described by Dangot et a!.(28) with modifications. The reaction mixture in a finalvolume of 50 �zI contained 50 m�i Tnis-HCI, pH 7.5, 20mM magnesium acetate, 5 mM NaF, 1 m�i EDTA, 1 m�’tDTT, 30 MM sodium-o-vanadate, 50 �zM AlP ({‘y-m2PJATP)with a specific activity of 0.6 Ci/mmol (Amersham Co.),and 20 �ig poly(L-Glu-L-Tyr, 4:1) (Sigma Chemical Co., St.Louis, MO). The reaction was initiated by adding a mem-brane/cytosol fraction (corresponding to 10-20 �tg pro-tein) and carried at 20#{176}Cfor 20 mm. The reaction wasterminated by spotting 20-�iI aliquots of reaction mixtureon Whatman no. 3 filter paper. Filters were extensivelywashed in ice-cold 10% trichloroacetic acid containing100 mM sodium pyrophosphate, rinsed in ethanol:ether

(1:1) and 100% ether. Radioactivity was counted in 10-ml scintillation cocktails (Scintiverse; Fisher Scientific, FairLawn, NJ). TPK activity was connected for endogenousphosphorylation by assaying in the absence of polymersubstrate. The enzyme activity is expressed as pmol [32P]ATP incorporated/mg protein/mm.

Protein Determination. Protein content in cytosol andmembrane fractions was determined by the method ofBradford (29), using bovine serum albumin as thestandard.

Statistical Analysis. Wherever applicable the data wereanalyzed using an unpaired t test and one-way analysisof variance.

Results

General Observations

The body weights were comparable in all the animals fedcontrol and experimental diets (Table 2). However, the

body weights ofAOM-treated animals were slightly lowerthan those of vehicle-treated animals due to AOM car-cinogenicity and consequent tumor burden.

Tumor Incidence

Table 3 presents gross observations of the effect of pi-roxicam and DFMO on incidence and multiplicity ofcolon tumors observed at different stages. AOM inducedmultiple tumors toward the end of experiment in about100% of animals fed the control diet. Animals fed thepiroxicam diet developed relatively small colon tumors.As reported earlier using different protocol (5), the colon

T,ihle 3 Effect of DEMO and piroxicam on tumor incidence’ and niultiplicity during different stages of AOM-induced colon carcinogenesis in male

F344 rats

Chemopre-ventive agents

Week s after carcinoge n treatment

0

No. of

animalsTotal

no.

4

No. of

animalsTotal

no.No. of

animals

16

Total no.

24 32

No. ofanimals Total no.

No. of

animals Total no.with of tu- with of tu- with of tumors with of tumors with of tumors

tumors mors tumors mors tumors tumors tumors

Control diet 0 0 0 0 2 )20)� 2 [0.2 ± 0.1]b 9 190) 24 [2.4 ± 0.5] 8 (100) 28 [3.5 ± 0.7]Piroxicam, 150 ppm 0 0 0 0 1 110) 1 [0.1 ± 0.1] 2 (20) 3 [0.3 ± 0.2]” 5 (50) 7 [0.7 ± 0.3](2DEMO, 4000 ppni 0 0 0 0 0 0 0 0 2 118) 3 [0.3 ± 0.2](2

aNunibers in parentheses indicate percentage of animals developing grossly visible tumors.

b Numbers in brackets are mean ± SEM.

( Significantly less than that in control group: 1 , P < 0.01 ; 2, P < 0.001.



408 Intermediate Biomarkers of Colon Cancer

Mucosal Protein Tyrosine Kinase Activity. As presentedin Table 5, the modulation of colonic mucosal TPK activ-

Table 4 Effect of DEMO and piroxicam on colonic mucosal ODC activity during different stages of AOM-iniduce’d olon .arcmnioge’nie’sms riniale’ E344 rats”

Experimental groupWeeks alter car ( iniogen tre’,mtmeiit

0 4 16 24 .32 Iurummrs

AOM-treated

Controldiet 360± 33m,m 130± ii�’ 217 ± � 129 ± � 174 ± ii 4192 ± 11511Piroxicam, 150 PP�

DFMO, 4000 ppm

950 ± 85�””

70 ± 451.( I

316 ± 49h2.,2

26 ± 3bI.�2

808 ± 70.43 � 451. � 1

1627 ± 124 �

29 ± 41 . 1

525 ± 27.18 ± .1” �

176 ± 41’

Vehicle-treatedControl diet 32 ± 2 30 ± 4 31 ± 4 12 ± 3 29 ± 2

Piroxicam,lSOppm 170±11” 150±17” 151 ±12’’ .158±37’ 224±41’’

DFMO,4000ppm 24±4 14±V’ 7±1” 6±1’’ 3±1’’

a ODC activity is expressed as pmol ‘�CO2 released/hr/mg protein; values are mean ± SEM ri = 10).b Significantly different from corresponding vehicle-treated groups; 1, P < 0.001 ; 2, P < 0.01.

‘ Significantly different from their respective control groups: 1, P < 0.001 ; 2, P < 0.01.

tumor multiplicity was significantly reduced in animalsfed the piroxicam diet (P < 0.01). The incidence andmultiplicity of AOM-induced colon tumors were signifi-cantly decreased in animals fed the experimental dietcontaining 4000 ppm DFMO (P < 0.001). The latencyperiod was greatly extended by dietary DFMO andpinoxicam.

ODC Activity

Induction of Colonic Mucosal ODC Activity. Table 4summarizes the colonic mucosal ODC activity at differ-ent time periods after AOM or saline treatment. AOMadministration induced persistently elevated levels ofODC activity throughout the duration ofthis experiment.Although AOM-stimulated ODC activity exhibited a dis-tinctly biphasic elevation pattern as demonstrated earlier(16), the differences in levels of ODC activity betweenAOM-treated and vehicle-treated animals remainedhighly significant irrespective of dietary intake of chem-opreventive agent (P < 0.00 1).

Dietary intake of DFMO resulted in a significantsuppression of colonic mucosal ODC activity at all timepoints (Table 4). AOM-treated animals fed the DFMOdiet demonstrated up to 90% inhibition of ODC activity.In saline-treated animals, DFMO suppressed the ODCactivity by more than 90% at the end of the experiment.Interestingly, animals from both AOM-treated as well assaline-treated groups fed the piroxicam diet exhibitedsignificantly increased levels of ODC activity as com-pared to corresponding groups fed the control diet (P <

0.01 -0.00 1).

Colonic Tumor ODC Activity. AOM-induced colon tu-moms obtained at the termination of the experiment fromanimals fed the control and piroxicam diets were ana-lyzed for ODC activity (Table 4). None ofthe animals fedthe DFMO diet showed gross colon tumors. The tumor

ODC activity of animals fed the control diet was dramat-ically increased by 24 times as compared to their mucosalODC activity. In contrast to the results of mucosal ODCactivity, dietary piroxicam significantly suppressed theODC activity of colon tumors as compared to those fedthe control diet.

TPK Activity

ity by piroxicam and DFMO followed somewhat similarpatterns of ODC activity, except that the differences inlevels of TPK activity were not significant in the beginningof the exr)eniment. AOM treatment significantly in-creased membrane-bound as well as cytosolic TPK activ-ity in animals on the control diet as compared to theirvehicle-treated counterparts throughout the duration ofthe experiment (P < 0.05-0.001). In addition, AOM ad-ministration induced progressively increasing levels ofmembrane-bound as well as cytosolic TPK activity. Die-tary piroxicam significantly raised the TPK activity inAOM-treated as well as saline-treated groups as com-pared to their corresponding AOM- and saline-treatedanimals fed the control diet (P < 0.05-0.001 ). In contrast,animals fed the DFMO diet exhibited progressively de-creasing levels of TPK activity in AOM-treated as well as

saline-treated animals. In these animals, the levels of TPKactivity were significantly reduced as compared to theircorresponding control dietary groups (P < 0.05-0.001).

Colonic Tumor TPK Activity. As summarized in Table 5,The TPK activity of colon tumors of animals fed thecontrol diet was about 6 times higher than that of theiruninvolved colon mucosae. Interestingly, both cytosolicand membrane-bound TPK activities were significantlysuppressed in colon tumors of animals fed piroxicam ascompared to those fed the control diet (P < 0.01).

Discussion

The results of this experiment demonstrate that dietaryDFMO and piroxicam significantly reduce colon tumoroccurrence and multiplicity and greatly increase tumorlatency in AOM-treated animals, corroborating earlierfindings (5-6, 16-17, 30-32). Our data clearly demon-strate distinct peaks of ODC activity manifesting themultistep characteristics of AOM-induced colonic carci-nogenesis. The phasic increase in the colonic mucosalODC activity of AOM-treated rats was 10-11 timeshigher than that of saline-treated rats. The levels of ODCactivity in colonic tumors of rats fed the control diet were140-fold higher than the colonic mucosal ODC activityof vehicle-treated rats on the control diet. The results ofpresent investigation and earlier studies (7-17) suggestthat mucosal ODC activity can be used as a reliable andsensitive marker for colon carcinogenesis.

We expected that DFMO would inhibit the inter-mediate biomarkers, since it had inhibited colon tumor



Cancer Epidemiology, Biomarkers & Prevention 409

Table 5 Effect of DFMO and piroxicam on colonic mucosal TPK activity during different stages of AOM-induced colon carcinogenesis inmale E344 ratsa

Experimental group

Membrane-bound

or cytosolicactivity

Weeks after car cinogen treatment

0 4 16 24 32 Tumors

AOM-treatedControldiet Membrane

Cytosolic

1O9±5�

46 ± 3b2

99±3b

48 ± 2�’

169±4b

76 ± 2bi

239±9bm

61 ± 2bi

255±6�

61 ± 2i�

1513 ±200

498 ± 64Piroxicam, 150 ppm Membrane

Cytosolic123 ± 5b1

48 ± 21 b3

112 ± 4bn.,3

49 � 2b3

306 ± 1O�

77 � 2b�

415 ± 16��1�81 ± �

386 ± flbi..�

59 ± �

494 � 58”196 ± 57(

DFMO,4000ppm Membrane

dytosolic98±4�’�

44 ± 3b)

8o±3I�.�42 ± 2b3, 1

113±Sbr.Cn

38 ± 2b�.0

73±Sb.�24 ± 1 b1.�i

66±2�.c

24 ± 1 �

Vehicle-treatedControl diet Membrane

dytosolic85 ± 4

37 ± 2

77 ± 4

31 ± 2

91 ± 4

41 ± 2

86 ± 3

39 ± 2

80 ± 2

42 ± 2Piroxicam, 150 ppm Membrane

Cytosolic89 ± 340 ± 2

83 ± 441 ± 3�2

129 ± 5”50 ± 2(2

163 ± 7”

61 ± 3(198 ± 6’

40 ± 3DEMO, 4000 ppm Membrane

Cytosolic

80 ± 5

36 ± 2

66 ± 2’�36 ± 2

60 ± 2”28 ± 2�2

50 ± 2’18 ± 1”

53 ± 3”12 ± 1”

a TPK activity is expressed as [32P]ATP phosphorylated/mg protein/mm; values are mean ± SEM In = 10).b Significantly different from corresponding vehicle-treated groups; 1 , P < 0.001 ; 2, P < 0.01 ; 3, P < 0.05.

‘ Significantly different from their respective control groups: 1 , P < 0.001 ; 2, P < 0.01 ; 3, P < 0.05.

development; this was the case. DFMO has been shownto strongly inhibit cell proliferation and tumor develop-ment in rodents presumably by blocking ODC-catalyzeddecarboxylation of ornithine into putrescine, therebydepleting intracellular polyamines (5-6). These polyam-ines are known to play an important regulatory role inthe control of normal growth and neoplastic transforma-tion (33-34). In most of the animal studies on chemopre-vention reported to date, ODC inhibitors are shown tobe very effective antitumor agents when treatment wasbegun prior to, concomitant with, or shortly after tumorinitiation (31-32). In this study, where DFMO treatmentbegan 1 week before AOM administration, we observedmore than 90% inhibition of AOM-induced increase inODC activity in a time-dependent manner, as well assuppression of colon tumor development. Tatsula et a!.(35) found that prolonged administration of putrescineinhibited AOM-induced colon tumor incidence, the mu-cosal labeling index, and ODC activity and suggestedthat putrescmne-augmented down-regulation of ODC isthe basis of chemoprevention of AOM-induced canci-nogenesis. It thus appears that inhibition of ODC activityinterferes with the sequence of events in tumor devel-opment. Significant differences in the levels of colonicmucosal ODC activity between saline-treated and AOM-treated rats on the control diet as well as on the DFMOdiet, as observed in this study, further attest to the factthat ODC may be used as a marker of DFMO-mediatedsuppression of colon tumor progression.

We expected that piroxicam would also exert itsinhibitory effect on the biomarkers just as DFMO had.However, this was not the case. The results of the presentstudy demonstrate that, in contrast to suppressive effectof DFMO on ODC activity, piroxicam significantly in-creased colonic mucosal ODC activity. Direct evidenceestablishing any mechanism by which piroxicam mayexert its antitumor activity is lacking. However, the chem-opreventive influence of piroxicam and other nonsteroi-dal antimnflammatory drugs like indomethacin and aspirinmight be mediated through the inhibition of prostaglan-din synthesis at the cyclooxygenase level or possibly byinterfering with cellular mitosis at the G1 stage (31, 36-

37). To our knowledge, there are no data pertaining tothe modulation of ODC activity by piroxicam. Indometh-acm is the only compound of this category so widelystudied in the context of its chemopreventive action.DeRubertis et a!. (38), in a related study, demonstratedthat treatment of rats with indomethacin rapidly in-creased colonic mucosal ODC activity, [3H]thymidineincorporation into mucosal DNA, and enhancement ofproliferative activity of colonic epithelium. These authorsalso showed that concurrent administration of pnosta-glandmn analogue in rats suppressed an indomethacin-induced increase in colonic mucosal ODC activity, DNAsynthesis, and epithelial proliferative activity. Recently,Noguchi et a!. (39) demonstrated that indomethacin de-creased mammary tumor incidence and tumor multiplic-ity and increased tumor latency but significantly pro-moted proliferation of DMBA-induced tumors in bothhigh- and low-fat diet groups as indicated by increasedtumor size, increased bromodeoxyunidine labeling index,and decreased tumor doubling time. Furthermore, sinceprostaglandins have a major role in the maintenance ofnormal gastrointestinal physiology, it is not surprising thatdrugs which inhibit the formation of prostaglandins in-terfere with normal gastrointestinal function. The nonste-roidal antiinflammatony drugs have been shown to causegastric mucosal cell loss due to gastric erosion and pepticulcer formation and perforation (40), which may initiateseveral rounds of compensatory hyperproliferation ingastrointestinal mucosa. In several other studies, aspirinand indomethacin have been shown to increase gastricand duodenal epithelial proliferation (41-42).

TPK activity has only recently been implicated incellular proliferation, differentiation, and neoplastictransformation. The biochemical actions of most of theoncogenes and certain growth factors are thought to bemediated through the effects of specific protein kinasesthat are involved in selective phosphorylation of cellularproteins (18-20). Hennipman et a!. (22) observed pro-gressively increasing levels of TPK activity in benign tomalignant breast tumors as compared to normal breasttissue. Elevated levels of TPK activity have been dem-onstnated in human colon cancer and colon cancer cell



4 10 Intermediate Biomarkers of Colon dancer

lines. Anlow et a!. (43) recently reported that an AOM-induced increase in colonic ODC activity in rats wasassociated with increased TPK activity. In our study,colonic mucosal TPK activity has exhibited a patternsimilar to that of ODC activity. There was a consistent

and significant increase in AOM-induced TPK activity inboth cytosolic and membrane fractions of rats fed thecontrol diet as compared with their saline-treated coun-terparts. A similar pattern of difference in AOM-inducedTPK activity was observed in animals fed the piroxicamor DFMO diet. Although piroxicam appears to moder-ately enhance TPK activity both in AOM-treated as wellas saline-treated animals, the dietary DFMO significantlysuppressed the TPK activity in a systematic time-depend-ent manner (Table 5). Furthermore, we observed drasti-cally reduced levels of membrane-bound TPK activity inrelation to cytosolic TPK activity in AOM-treated animalsfed the piroxicam diet both in colonic mucosa and tu-mors as compared to those in AOM-treated animals fedthe control diet. Sakanoue et a!. (44) have reportedreduced levels of cytosolic tyrosine kinase activity andincreased levels of membrane TPK activity in humancolon carcinomas as compared with their adjacent nor-mal mucosa. It is possible that the progressively increas-ing translocation of TPK from cytosol to membrane maybe occurring with the advancing sequence of events inAOM-induced tumor development as in the case ofphorbol ester-induced translocation of protein kinase C(45). It may also be possible that the cytosolic (nonrecep-ton) tyrosine kinase could be an antioncogene productitself. Recently, multiple genetic alterations have beendetected that appear to contribute to the malignant trans-formation of colonic mucosa. Allelic deletions of chro-mosomes 5, 17, and 18 sequences have been identifiedin human colonic carcinoma and in adenomas, suggestingthat tumor suppressor genes may be important regulatorsof cell growth (46).

In conclusion, the results of this study demonstratethe usefulness of mucosal ODC and TPK in analyzing thestate of proliferating mucosa and events leading to tumordevelopment. But these results also point to the fact thatsome chemopreventive agents, such as piroxicam, caninhibit cancer development without reducing prolifera-tion. This of course raises interesting questions about thediverse mechanisms by which different agents exert theirchemopreventive effects and the general usefulness ofproliferating markers such as ODC as an intermediateend point. We suggest that it may be necessary to use“agent-specific intermediate biochemical markers” for

evaluating the effectiveness of specific agent(s) in chem-

opreventive studies.

AcknowledgmentsWe thank Donna Virgil for preparatic)n of the nianuscript and Ronnie

Bland for preparing the diets, Thanks are also due to the staff c)f theResearch Animal Facility for their technical assistance.

References

1. Cancer Statistics, 1991, CA Cancer Clin., 41: 19-36, 1991.

2. World Health Statistics Annual 1989. Geneva: WHO, 1989.

3. Sporn, M. B., and Roberts, A. B. Role of retinoids in differentiation

and carcinogenesis. I. NatI. Cancer Inst., 7.3: 1381-1387, 1984.

4. Greenwald, P., Nixon, D. W., Malone, W. E., Kelloff, G. I.. Slern, H.R., and Wilkins, K. M. Concepts in cancer chemopreventive research.

Cancer )Phila.) 65: 1483-1490, 1990.

5. Reddy, B. S., Nayini, I.. Tokumu, K., Rigotty, I., Zang, E., arid Kelluff,

G. Chemupreventiun of colon carcinugenesis by concurrent administra-

lion of piruxicani, a non-steroidal anti-inflammatory drug with D,L-a-

difluorumethylurnithinie, an ornithine decarhuxylase’ inhibitor, in diet.

Cancer Res., .50: 2562-2568, 1990.

6. Reddy, B. S., Maruyama, H., arid Kelluff, G. Dose related inhibition of

colon carcinogenesis by dietary piruxicam, a nunsteroidal anti-inflam-

matory drug, during different stages of rat cc)Iun tumor development.

Cancer Res., 47: 5340-5346, 1987.

7. Russel, D. H. Ornithine decarbuxylase: a key regulatory enzyme in

normal and neoplastic growth. Drug Metab. Rev., 16: 1-88, 1985.

8. O’Brian, T. G. The induction of ornithine decarboxylase as an early,possibly obligatory event in mouse skin carcinogenesis. Cancer Res., .36:

2644-2653, 1976.

9. Verma, A. K., Shapas, B. G., Rice, H. M., et a). Correlation of inhibition

by retinoids 01 tumor promc)ter induced mouse epidermal ornithine

decarbuxylase activity and of skin tumor promotion. Cancer Res., 39:

419-425, 1979.

10. LaMuraglia, G. M., Lacaine, F., and Malt, R. A. High urnithine decar-buxylase activity and pulyaniine levels in human culurectal neuplasia.

Anti. Surg., 204: 89-93, 1986.

1 1 . Porter, C. W., Herrera-Ornelas, L., Clark, I.. Pera, P., PeIrelli, N. I..

and Mittleman, A. Polyamine biosynthetic activity in normal and neu-

plastic human culorectal tissue. Cancer (Phila.) 60: 1275-1281, 1987.

1 2. Herrera-Ornielas, L., Porter, C., Pera, P., Gregu, W., PeIrelli, N. I..

and Mettelman, A. A comparison of ornithine decarhoxylase and 5-adenosylmethiune decarbuxylase’ activity in human large bowel mucusa,

pulyps and colureclal adenucarcinunia. I Surg. Res., 42: 56-60, 1987.

13. Mourehead, R. I., Hoper, M., arid McKelvey, S. T. D. Assessment ofornithine decarboxylase in rectal mucosa as a marker fur culorectal

adenornas and carcinomas. Br. I. Surg., 74: 364-365, 1987.

14. Luk, G. D., and Baylin, S. B. Ornithine decarboxylase as biologicalniarker in familiar culoniic polyps. N. EngI. I. Med., .11 1: 80-83, 1984.

15. Colarian, I.. Arluw F, Calzada, R., Luk, G. D., and Majumdar, A. P.

N. Differential activation of urnithine decarbuxylase and tyrusine kiniase

in rectal mucusa of patients with adenomatous and hyperplastic polyps.

Gastruenterology, 96: A93, 1989.

16. Luk, G. D., Hamilton, S. R., Yung, P., Smith, J. A., O’Cealaigh, D.,

McAvinchey, D., and Hyland, 1. Kinetic changes in mucusal ornithinie

decarboxylase activity during azoxyniethane induced iuluni carcinogen-

esis in the rat. Cancer Res., 46: 4449-4452, 1986.

17. Kingsnorth, A. N., King, W. W. K., Diekema, K. A., McCann, P. P.,Ross, I. S., arid Malt, R. A. Inhibition uI urnithine decarbuxylase with 2-

difluuruniiethylurnilhine’: reduced incidence of dimethylhydrazine in-duced colon tumor in mice. Cancer Res., 43: 2545-2549, 1983.

18. Ullrich, A., and Yarden, Y. Molecular analysis of signal transdu�iun

by growth factors. Biochemistry, 27:3113-3119, 1988.

19. Yarden, Y. Growth factor receptor tyrusine kinases. Annu. Rev.

Biocheni., 57: 443-478, 1988.

20. Bulen, I. B., Rosen, N., and Israel, M. A. Increased pp6O’” tyrusinie

kinase activity in huniani neurublastuma is associated with amino terminaltyrosine PhusphurYlatiuni (if the src gene product. Proc. NaIl. Acacl. Sci.

USA, 82: 7275-7279, 1985.

2 1 . [in, M. F., Lee, P. L., and Clinton, C. M. Charaterizatiun of tyrusyl

kinase activity in human serum. I. Biol. Chem., 260: 1582-1587, 1985.

22. Hennipniani, A., van Dirschut, B. A., Rijksen, I. S., arid Staal, G. E. I.

Tyrosune kinase activity in breast cancer, benign breast disease and normalbreast tissue. Cancer Res., 49: 516-521, 1989.

23. Bulen, I. B., Veillette, A., Schwartz, A. M., PeseaL, V., and Rosen, N.

Activation of 01)60’ “ protein kinase activity in human carcinoma. Proc.NatI. Acad. Sci. USA, 84: 2251-2255, 1987.

24. Weinberg, R. A. Oncugenes, anitiuncugenes, and the molecular bases

of niullistep carcinugenesis. Cancer Res., 49: 3713-3721, 1989.

25. Recldy, B. S., and Sugie, S. Effect of different levels of umega.3 andumega-6 fatty acids on azuxynie’th,inie.iniduced colon carcinoge?nesis in

F344 rats. Cancer Res., 48: 6642-6647, 1988.

26. Lapunte, D. S., and Cohen, R. I. A rapid and efficient niicruassay ofurnithine decarboxylase. Anal. Biuchem., 190: 291-294, 1980.

27. Zmek, Y., Grunberger, G., and Rees-lunes, R. Use of tyrusine contain-

ing polymers to characterize the’ substrate specificity of insulin and ditherhormone stimulated tyrusine kinases. Eur. I. Biocheni., 148: 1 77-182,

1985.

28. Dangut, L. I.. Puett, D., and Melner, M. H. Characterization of proteintyrusine kinase activity in niurine levdig tuniior cells. Biuchini. Biuphys.

Ada, 886: 187-194, 1986.



Cancer Epidemiology, Biomarkers & Prevention 4 1 1

29. Bradford, M. A rapid and sensitive method for the determination ofmicrogram quantities of protein utilizing the principles of protein dye

binding. Anal. Biochem., 72: 248-254, 1976.

30. Rosenberg, L., Palmer, I. R., zauber, A. G., Warshauer, M. E., Stolley,

P. D., and Shapiro, S. A hypothesis: nonsteroidal anti-inflammatory drugsreduce the incidence of large bowel cancer. I. NatI. Cancer Inst., 83:

355-358, 1991.

31. Ip, C., and Thompson, H. I. New approaches to cancer chemopre-

vention with difluoromethylornithine and selenite. I. NatI. Cancer Inst.,

81: 839-843, 1989.

32. Luk, G. D., zang, S. z., and Hamilton, S. R. Effects of timing of

administration and dose of difluoromethylornithine on rat colonic card-nogenesis. I. NatI. Cancer Inst., 81: 421-427, 1989.

33. Janne, I., and Alhonen-Hongisto, L. Inhibitors of polyamine biosyn-

thesis as therapeutic agents. In: U. Bachrach and N. M. Heimer (eds.(,The Physiology of Polyamines It. Boca Raton, FL: CRd Press, 1989.

34. Pegg, A. E. Polyamine metabolism and its importance in neoplasticgrowth and as a target for chemotherapy. Cancer Res., 48: 759-773,

1988.

35. Tatsula, M., lishi, H., Baba, M., Ichil, M., Lakaizumi, A., Uehera, H.,and Taniguchi, H. Inhibition by putrescine of experimental carcinogenesisin rat colon induced by azoxymethane. Int. I. Cancer, 47: 738-741, 1991.

36. Abramson, S. B., and Weissmann, G. The mechanisms of action of

non-steroidal anti-inflammatory drugs. Arthritis Rheum., 32: 1 -9, 1989.

37. Bayer, B. M., and Beaven, M. A. Evidence that indomethacin revers-ibly inhibits cell growth in the G, phase of the cell cycle. Biochem.

Pharmacol., 28: 441-443, 1979.

38. DeRubertis, F. R., draven, P. A., and Saito, R. 16,16-Dimethylpros-

taglandin E2 suppresses the increase in the proliferative activity of rat

colonic epithelium induced by indomethacin and aspirin. Gastroenter-ology, 89: 1054-1063, 1985.

39. Noguchi, M., Taniya, T., Koyasaki, N., Kumaki, T., Miyazaki, I., and

Mizukami, Y. Effects of prostaglandin synthetase inhibitor indomethacinon tumorigenesis, tissue proliferation, cell kinetics, and receptor contentsof 7, 1 2-dimethylbenzla)anthracene induced mammary carcinoma inSprague-Dawley rats fed a high or low fat diet. Cancer Res., 51: 2683-

2689, 1991.

40. Henry, D. A. Side effects of non-steroidal antiinflammatory drugs.

Baillieres dIm. Rheumatol., 2: 425-454, 1988.

41 . Eastwood, G. L., and Quimby, G. F. Effect of chronic aspirin ingestion

on epithelial proliferation in rat fundus, antrum, and duodenum. Gastro-

enterology, 82: 852-856, 1982.

42. Kuwayama, H., and Eastwood, G. L. Effects of water immersion,

restraint, stress, and chronic indomethacin ingestion on antral and fundicepithelial proliferation. Gastroenterology, 88: 362-365, 1985.

43. Arlow, F. L., Walczak, S. M., Luk, G. D., and Majumdar, A. P. N.Attenuation of azoxymethane induced colonic mucosal ornithine decar-

boxylase and tyrosine kinase activity by calcium in rats. Cancer Res., 49:5884-5888, 1989.

44. Sakanoue, Y., Kusunoki, M., Hatada, T., Sakiyama, T., Yamamura, T.,and Utsunomiya, J. Altered protein tyrosine kinase levels in human coloncarcinoma. Cancer (Phila.(, 67: 590-596, 1991.

45. Kraft, A. S., and Anderson, W. B. Phorbol esters increase the amount

of Ca2” phospholipid dependent protein kinase associated with plasma

membrane. Nature (Lond.), 307: 621-623, 1983.

46. Vogelstein, B., Fearon, E. R., Hamilton, S. R., et a!. Genetic alterationsduring colorectal tumor development. N. EngI. I. Med., 319: 525-532,

1988.



1992;1:405-411. Cancer Epidemiol Biomarkers Prev J Singh, G Kelloff and B S Reddy carcinogenesis.during different stages of azoxymethane-induced colon Effect of chemopreventive agents on intermediate biomarkers

Updated version

http://cebp.aacrjournals.org/content/1/5/405

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cebp.aacrjournals.org/content/1/5/405To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



effect of chemopreventive agents on intermediate biomarkers … · vol. 1, 405-4 1 1, july/august i...

Documents