~~ ~-------------~-

Cell Culture Model for Colon CancerPrevention and Therapy':An ~I!ernativ~ Approach to AnimalExperlmentatlonNitin Telang ' and Meena Katdare?'Strang Cancer Prevention Center, New York, USA; -wem Medical College of Cornell University, New York, USA

SummaryMouse models for colon cancer that harbor a germ linemutation in the tumor suppressor gene Adenomatous polyposiscoli (Apc) exhibit a primary genetic defect that predisposes toa high incidence of adenomatous polyps in the small intestinerather than in the colon. Colon cell culture models expressingquantifiable markers for carcinogenic risk may represent analternative approach to reduce, refine or replace long-termanimal experimentation. The newly developed colon epithelialcell lines 1638N COL-CIJ (clonal derivative of the parentalApc mutant cell line 1638N COL) and 1638N COL-Pn (tumorderivative of the clone), established from an Apc1638N {+/-]mutant mouse, exhibit aberrant cell cycle progression, down-regulated apoptosis, enhanced carcinogenic risk and tumorformation, indicating that aberrantly proliferative preneo-plastic1638N COL-Cli cells exhibit a quantifiable riskfor car-cinogenesis. Treatment of these preneoplastic Apc mutant cellswith a combination of celecoxib and 5-fluorouracil at clinicallyachievable low concentrations produced a 2.1 fold to 5.5 foldhigher efficacy for cytostatic growth arrest and a 40.2% to52.4% higher efficacy for inhibition of carcinogenic risk,relative to that obtained by these agents used individually. Thus,a low dose combination of mechanistically distinct agentsresulted in enhanced efficacy. These data validate a novel cellculture model and a rapid mechanism-based approach to prior-iti:e efficacious drug combinations for animal studies andclinical trials on cancer prevention and, thereby, support the3R concept by refining and/or reducing the use of animalsin biomedical research relevant to prevention/therapy of coloncancer.

Keywords: APC mutation, colon cancer risk, prevention

1 Introduction

Zusammenfassung: Ein Zellkulturmodell für die Dickdarm-krebs- Vorsorge und Therapie: Eine Alternative zum tierex-perimentellen AnsatzMausmodelle für Dickdarmkrebs, welche die Keimzell-Mutationfür das Tumorsuppressorgen Apc (adenomatöse Polyposis coli)tragen, zeigen ein durch diesen primären Gendefekt bedingteshäufigeres Auftreten adenomatöser Polypen im Dünndarm eherals im Dickdarm. Dickdarm-Zellkulturmodelle, die messbareMarker für das Krebsrisiko exprimieren, könnten geeignet seinLangzeittierversuche zu reduzieren. Die neu entwickeltenDickdarm-Zelllinien 1638N COL-CLJ (klonale Ableitung derparentalen Linie 1638N COL) und 1638N COL-Pn (ausTumoren entnommen), gewonnen aus der Mausmutante Apc1638N {+/-], wiesen anormale Zellryklen, verminderteApoptose, erhöhtes Krebsrisiko und Tumorbildung auf DieBehandlung von präneo-plastischen Apc Mutantenzellen miteiner Kombination aus Celecoxib und 5-Fluorouracilführte zueiner 2,1- bis 5,5- fach höheren Wirksamkeit der zytostatischenWachstumshemmung und zu einer 40,2 bis 52,4% höherenWirksamkeit bei der Unterdrückung des Krebsrisikos imVergleich zum Einsatz der einzelnen Substanzen. Diese Datenbelegen ein neues Zellkulturmodell und einen schnellenmechanistischen Ansatz zur Auswahl effektiver Therapie-kombinationen für Tierversuche und klinische Versuche zurKrebsprävention. Das 3R Konzept wird damit durch eine Ver-minderung der Belastung und eine Reduktion der Tierzahlenin der biomedizinischen Forschung zur Vorsorge und Therapiedes Dickdarmkrebses gefördert.

Colon cancer is a multi-step, multi-facto-rial disease where genetic, environmentaland dietary factors exert a profound ef-

feet on etiology, pathogenesis and pre-ventive/therapeutic intervention (Potter,1996). Traditionally, animal models suchas azoxymethane (AOM)-induced ratcolon cancer or in adenomatous polypo-

sis coli (Apc) gene mutant mice havebeen utilized as preclinical approachesfor the evaluation of toxicity and efficacyof potential preventive/therapeutic agents(Corpet and Pierre, 2003). Rational se-lection of relevant alternative approachesmay reduce, refine or replace the need forReceived 11 September 2006; received in final form and accepted for publication 7 February 2007

16 ALTEX 24, 1/07

_____ ~--------------------------------------------------~L=IN~Z~2~O~O~6

extensive long-term animal experimentson colon carcinogenesis and its preven-tion.The germ line mutation in the tumor

suppressor Apc gene represents a pri-mary genetic defect responsible forclinical familial adenomatous polyposis(FAP), a high-risk syndrome for earlyonset colon cancer (Fodde et al., 2001;Fearon and Vogelstein, 1990). Apc mu-tant mice, however, exhibit adenomasand carcinomas predominantly in thesmall intestine, rather than in the colon(Moser et al., 1990, Fodde et al., 1994).The predominance of carcinogenesis inthe small intestine of Apc mutant micerepresents a target site distinct from thatfor clinical colon cancer, and hence datagenerated from these models require ex-trapolation to determine their clinical rel-evance.Reliable cell culture models estab-

lished from the appropriate target organsite, that express a clinically relevant ge-netic defect, and exhibit quantifiablemarkers of risk for carcinogenesis are ex-pected to offer alternative approachesthat may reduce or refine animal experi-ments, minimize necessary extrapolationand, thereby, complement in vivo animalstudies. Such cell culture studies may al-so identify a rapid mechanism-based ap-proach to evaluate the potential efficacyof agents that inhibit, delay or reduce theincidence of colon cancer.The Apc Min/+mouse, which harbors a

germ line mutation in the Apc tumorsuppressor gene, exhibits a high inci-dence of intestinal adenomas (Moser etal., 1990) that are susceptible to growthinhibition by several mechanistically dis-tinct chemopreventive agents used inde-pendently or in combination (Jacoby etal., 2000; Torrance et al., 2000; Swamyet al., 2006). It is also noteworthy thatpharmacological (Saez et al. , 1998; Paul-son et al. , 2001) or genetic (Rao et al.,2005; Sodir et al., 2006) manipulationleads to colon cancer in the Apc Min/+mice, indicating that multi-step coloncarcinogenesis is inducible in this model.Recent cell culture studies using

murine cell lines have demonstrated thathistopathologically normal colon epithe-lial cells expressing mono-allelic muta-tions in the Apc gene (Katdare et al. ,2002), or in Apc and Mlh, genes (Telang

ALTEX 24. 1/07

et al., 2006), exhibit aberrant prolifera-tion and enhanced risk for carcinogene-sis. Furthermore, these mutant cells arealso susceptible to growth inhibition byseveral chemopreventive agents usedeither individually or in combination(Katdare et al., 2002; Telang et al. , 2006;Telang and Katdare, 2007).The present report summarizes recent

experiments conducted on a newly devel-oped colon epithelial cell culture modelfor FAP to: i) characterize the model withregard to the extent of aberrant prolifera-tion and risk for spontaneous carcino-genic transformation and ii) evaluate thesusceptibility of the model to inhibitionof growth and reduction of carcinogenicrisk by a combination of mechanisticallydistinct preventive/therapeutic agents.The outcome of the present studydemonstrates that the clonally selected,aberrantly proliferative Apc mutant1638N COL-Clt cells displaya definedrisk for carcinogenesis and that combina-tion of mechanistically distinct preven-tive/therapeutic agents given at low con-centrations leads to enhanced efficacy ofgrowth inhibition and to a greater reduc-tion of the risk of carcinogenesis, relativeto that produced by these agents used in-dependently.

2 Materials and methods

2.1 Cell linesApc [+/+] C57 COL (Source: normal de-scending colon of C57BLl6J mouse),1638N COL- Cl, (Source: Apc mutantclonal derivative from parental Apc [+/-]1638N COL cellline) and 1638N COL-Pr, (Source: tumor-derived cells fromtransplantation of 1638N COL-Clt cells)were grown in DMEIF12 medium sup-plemented with 10% heat-inactivatedfetal calf serum, 0.24 lU/mi (10 ug/ml)insulin and 1 ~ dexamethasone. Theculture medium also contained the an-tibiotic mixture (100 IU/ 100 ug/ml peni-cillin! streptomycin +50 ug/ml fungizone+50 ug/ml gentamycin). Routinely, thestock cultures of all the cell lines weremaintained at 37°C in an humidified at-mosphere of 95% air: 5% C02, fed withfresh medium every 48 hr and sub-cul-tured at 1:10 split at about 70-80% con-f1uency (Katdare et al., 2002).

2.2 Population doubling time,anchorage-independent colonyformation and tumorigenecityThese end points were measured usingpreviously optimized assays (Telang etal., 1990; Jinno et al., 1999; Katdare etal., 2002 and Telang et al., 2006). Popu-lation doubling time (PDT) was calculat-ed from the exponential growth phase.Viable cell counts were obtained at 24,48, 72 and 96 h post-seeding of 1xl05

cells. These time points cover the expo-nential growth phase of the celllines.For the anchorage independent colony

formation (AICF) assay, cells were sus-pended at the initial seeding density of100 cells/well in 0.33% agar made up inDMEIF12 medium. This cell suspensionin agar was overlaid on a basement layerof 0.6% agar in six weil cluster plates.The cultures were maintained at 37°C,and the numbers of anchorage indepen-dent colonies per weil were determinedat day 14 post-seeding (Telang et al.,1990; 1991).For the tumorigenicity assay l xl O"

cells suspended in 0.1 ml of the culturemedium were injected sub-cutaneouslyin the flank region of C57BLl6J mice.The transplant site was palpated at week-ly intervals, and palpable tumors wereexcised at 1 cm diameter. The primarytumor cells were expanded in culture toestablish the tumor-derived 1638N COL-Pr, cell line using previously optimizedculture conditions (Telang et al., 1990).

2.3 Cell cycle analysisThe status of cell cycle progression wasdetermined by monitoring percent distri-bution of cells in sub Go, GO/GI, S, andG21M phases of the cell cycle (Jinno etal., 1999; Katdare et al., 2002). The cellcycle data was expressed as S+G21M(proliferative, P): sub Go (apoptotic, A)ratio.

2.4 Preventive/therapeutic testcompoundsThe selective COX-2 inhibitor celecoxib(CLX) and selective thymidylate syn-thase inhibitor 5-fluorouracil (5-FU)were used as prototypical test com-pounds because of their documented pre-clinical and clinical efficacy in preven-tion!therapy of colon cancer (Jacoby etal., 2000; Steinbach et al., 2000 and Dia-

17

LINZ 2006 ~-------------r.;;:J-

sie and Harris, 1989). The stock solu-tions (100 mM) were made up in 100%ethanol (EtOH), and were serially dilutedin the culture medium to obtain 0.01, 0.1,1, 10 and 100 f1M concentrations. Thegrowth inhibitory effects were deter-mined by obtaining viable cell counts atday 5 post-seeding of 1x105 cells (Kat-dare et al., 2002; Telang et al., 2006). Theprimary cell count data were expressedas inhibitory concentration (IC 50) valuesand were compared with the reportedclinically or experimentally achievablemean plasma levels of the two com-pounds.

2.5 Combinatorial eHicacyThe efficacy of low concentration combi-nations of the mechanistically distinctCLX and 5-FU was evaluated using eachcompound independently and in combi-nation at less than its respective individ-ual IC20 concentration. Viable cell num-ber and cell cycle progression at day 5post-seeding and AICF at day 14 post-seeding represented the quantitative endpoints.

3 Results

3.1 Aberrantly proliferativeApc mutant cells exhibit anenhanced risk for carcinogenesisRelative to the wild type Apc [+/+] C57COL cells, the Apc [+/-] mutant clonalderivative 1638N COL-Cl1 and the tumorderived 1638N COL-Prl cells exhibited aprogressive 55.9% and 64.7% decreasein population doubling time (PDT). Fur-thermore, the Apc mutant cell lines alsoexhibited a 100% incidence of anchor-age-independent colony formation(AICF), and a 100% incidence of tumorformation (Tab. 1). It is noteworthy thatthe 1638N COL-Cl1 cellline, represent-ing a clonal derivative obtained from asingle anchorage-independent colony ofthe parental Apc [+/-] 1638N COL cellline, exhibited persistent anchorage-in-dependent colony formation in vitro andtumorigenicity upon in vivo transplanta-tion. Thus, 1638N COL-Cl1 cells exhibitpre-requisites for a preneoplastic pheno-type.

18

3.2 Apc mutant cellsexhibitaltered cell cycle progressionThe aberrant proliferation in the Apc mu-tant cells was associated with an increaseof the cell population in the S+G2/Mphase of the cell cycle and a decrease in

the sub Go (apoptotic) population. Thus,relative to that observed in the C57 COLcells, the 1638N COL-Ch and 1638NCOL-Prl exhibited an 89.4% and 152.8%increase in the cell population in theS+G2/M phase of the cell cycle. In addi-

Tab. 1: Aberrant proliferation and carcinogenic transformationin Apc mutant colon epithelial cell lines

AICFb

(%)Tumour Incidence c(%)

Cell Line PDTa

(hr)

C57 COL 34 0/12(0%)

1638N COL-Ch 15 12/12(100%)

1638N COL-Pr1 12 12/12(100%)

0/10(0%)

10/10(100%)

5/5(100%)

a Population doubling time (POT) determined from the exponential growth phase.bAnchorage independent colony formation (AICF) determined from the number ofcolonies at day 14 post-seeding. Number of wells with colonieslTotal wells.

c Tumour incidence determined after sub-cutaneous transplantation of 1x1 06 cellsin C57BU6J mice.

Tab. 2: Aberrant cell cycle progression in Apc mutant colon epithelial cell lines

Cell Line BiomarkerS+G2/M a Sub Goa P:A Ratio(%P) (%A)

24.6±3.3b 4.3±1.1e 5.7

46.6±1.9c 0.6±0.4 f 77.7

62.2±3.0d 0.3±0.2g 207.3

C57 COL

1638N COL-CI1

1638N COL-Pr1

a Mean ± SO, n=6 per cell line.b-c.e-f p=0.02,

c-d,e-g p=0.01.

Tab. 3: In vive pharmacokinetics and in vitro dose response of preventive/therapeuticcompounds

Agent Plasma Levels 1638N COL-Ch(range, l-lM) (lC50, l-lM) a

Clinical Experimental

CLXb 2.5-10.0 3.0-5.7 5.2±0.55-FU c 3.0-10.0 5.0-10.0 0.3±0.1

a Determined from the concentration range of 0.01-100 l-lM. Mean ± SO, n=6.b Jacoby, R.. F et al., Cancer Res. 60, 5040-5044, 2000.c Takimoto, C. H. et al., Clin. Cancer Res. 5, 1347-1352, 1999.

ALTEX 24, 1/07

~ ~~-~--------------

tion, the Ape mutant cells exhibited an86.0% and 93.0% inhibition in conftuen-cy-dependent spontaneous apoptosis.The cell cycle data, expressed as P:Aratio, revealed a progressive increase infavor of proliferation (Tab. 2).

3.3 Apc mutant cells aresusceptibleto growthinhibition bya combinationof preventive/therapeuticcompoundsThe experiments presented in Table 3compared the efficacy of growth inhibi-tion by CLX and 5-FU in the Apc mutantclonally derived preneoplastie 1638NCOL-CI] eells, determined by the ICsovalues, to the published in vivo clinicaland experimental pharmacokinetic pa-rameters that were determined by mea-suring the plasma levels of the twoagents. These eomparative data demon-strate that the ICso values are within thepharmacologically achievable range.The experiments presented in Table 4,

examined the effects of CLX and 5-FU

independently and in combination on theclonally derived Apc mutant preneoplas-tie 1638N COL-ClJ cells. The status ofthe viable cell population and AICF rep-resented the quantitative end points.A low dose combination of CLX+5-FUat JC < 20 exhibited 2.1 to 5.5 fold in-creased efficacy for growth arrest, rela-tive to that obtained by these compoundsused independently. In addition, the dataobtained from the AICF assay demon-strated that the CLX+5-FU eombinationwas 40.2% to 52.4% more efficaciousthan the two compounds used indepen-dently.

3.4 Combination ofpreventive/therapeuticcompounds alterscell cycle progressionThe experiments presented in Table 5 ex-amined the status of cell cycle progres-sion in clonally derived Apc mutant pre-neoplastic 1638N COL-Cb cells inresponse to treatment with low doseCLX and 5-FU independently and in

Tab. 4: Inhibition of aberrant proliferation in Apc mutantcolon epithelial 1638N COL-Ch ceilline

Agent Concentration(~M)

Inhibition(% of Contral)

Viable Cell Number a AICF b

CLX

5-FU

CLX+5-FU

1.0

0.1

1.0+0.1

7.9

16.551.4

52.9

57.5

80.6

a Number of viable cells in solvent controls on day 5 post-seeding: 49.5±3.1 x1 05.Mean ± SO, n=6 per treatment group.

b Anchorage independent colony formation (AICF). Number of anchorage independentcolonies in solvent controls on day 14 post-seeding: 18.1 ± 1.1.Mean ± SO, n=12 per treatment group.

Tab. 5: Regulation of cell cycle progression in Apc mutantcolon epithelial 1638N COL-Ch cell line

Agent Concentration Phase of Cell Cycle a(~M) (%)

G1 S+G2/M Ratio

1.0 32.1 67.9 0.4

0.1 28.2 71.8 0.4

1.0+0.1 58.0 42.0 1.4

CLX

5-FU

CLX+5-FU

a Oetermined fram at least 104 fluorescent events at day 5 post-seeding.Mean n=6 per treatment group.

ALTEX 24, 1/07

combination. The CLX+5-FU combina-tion induced an at least 80.7 to 105.7%increase in the cell population at GIphase and 38.1 to 41.4% decrease in thecell population at the S+GzIM phase ofthe cell cycle relative to that observedwith CLX and 5-FU used independently.

4 Discussion

In an effort to minimize indiscriminateuse of animal testing, the 3R conceptstrives to replace, reduce or refine the ex-isting in vivo animal models (Russell andBurch, 1959). In the area of colon cancerprevention the rat model for chemieallyinduced colon cancer and the Apc mutantmouse model for spontaneous intestinalcancer remain the most extensively usedpreclinical models (Corpet and Pierre,2003). However, unlike the clinical FAPsyndrome for colon cancer, the Apcmouse models exhibits carcinogenesispredominantly in the small intestinerather than in the colon (Moser et al.,1990; Fodde et al., 1994), thus, requiringextrapolation of the data to determinetheir elinical relevanee. These aspeetsemphasize a need to develop alternativemodels that express clinical1y relevantgenetic defects (Apc mutation) at theappropriate target organ site (colon), andexhibit a quantifiable risk for carcinogen-esis.Most of the existing cell culture mod-

els relevant to genetically predisposedsyndromes of colon cancer are developedfrom elinical colon cancer cell lines thatexhibit genetic defects in Apc (Ilyas etal., 1997) or DNA mismatch repair genes(Meyers et al., 2001). Fully transformedtumor cell phenotype of these models of-fers only a limited advantage in studiesfocused to identify early occurringmolecular/genetic events that predisposeto the multi-step carcinogenic process byenhaneing the risk of cancer develop-ment. Existing animal models for the ge-netie predisposition exhibit carcinogene-sis predominantly in the small intestine,rather than in the colon (Moser et al.,1990, Fodde et al., 2001), thus differingin the target organ site from that in theearly onset clinical colon cancer. In con-trast, the present cell culture model ex-

19

LINZ 2006 m....---------------------------------------~---

presses a clinically relevant genetic de-fect (Apc mutation) in cells establishedfrom the appropriate target organ site(histopathologically normal colon), andexhibit a quantifiable risk for carcinogen-esis (AICF in vitro and tumorigenicity invivo). Thus, these observations stronglysuggest that the present cell culture mod-el uniquely facilitates identification ofclinically translatable mechanistic leadswith minimal extrapolation.

The outcome of experiments designedto characterize the present model clearlydemonstrated that the clonally derivedApc mutant preneoplastic 1638N COL-Cl, and the tumor derived 1638N COL-Pr! cell lines exhibit a progressive de-crease in PDT, increase of the cellpopulation in the S+GVM phases of thecell cycle, and decrease in confluency-dependent spontaneous apoptosis. In thiscontext it is noteworthy that, similar tothe present data, a previously publishedstudy on the aberrantly proliferativeparental 1638N COL cells (Katdare etal., 2002) has demonstrated that Apc mu-tant cells exhibit aneuploidy, increasedpopulation doublings and decreasedGo/G!:S+G2/M ratio. Similar alterationshave also been reported in mammary ep-ithelial cells transfected with ras, myc orHer-2/neu oncogenes (Telang et al.,1991; Telang et al., 1990; Katdare et al.,2003). Thus, these observations suggestthat down-regulation of the tumor sup-pressor Apc gene expression or aberrantover-expression of ras, myc or HER-2/neu oncogenes may represent a prima-ry genetic defect responsible for aberrantproliferation.

The experiments conducted to evaluatethe risk of carcinogenesis utilized AICFin vitro and tumor incidence upon trans-plantation as the quantitative end points.The data generated from these experi-ments demonstrated that the Apc mutantcells exhibit a higher incidence of AICFin vitro and tumor formation in vivo incomparison with the wild type Apc [+/+]C57 COL cells that tested negative forboth these end points. These data areconsistent with those from previous stud-ies on mammary epithelial cells wheretargeted over-expression of ras, myc orHER-2/neu oncogenes results in AICF invitro and tumor formation in vivo (Telang

20

et al., 1990; Telang et al., 1991; Zhai etal., 1993). It is noteworthy that clonaIlyselected 1638N COL-Cl1 cells exhibiteda high incidence of tumors upon in vivotransplantation. Thus, the 1638N COL-Cli represent a preneoplastic phenotype,and AICF, a modulatable surrogate endpoint biomarker for the risk of carcino-genic transformation, similar to that re-ported for mammary epithelial cells (Jin-no et al., 1999; Katdare et al., 2002;Katdare et al., 2003).

In a previously published study theApc mutant parental 1638N COL cellsdemonstrated susceptibility to growthinhibition by several mechanisticaIlydistinct chemopreventive agents used assingle agents (Katdare et al., 2002). ThemechanisticaIly distinct preventive/thera-peutic compounds CLX and 5-FU havedocumented efficacy as single agents inthe clinical management of colon cancer(Steinbach et al., 2000; Takimoto et al.,1999; Diasio and Harris, 1989). Long-term high dose therapy with these com-pounds, however, is associated with ad-verse toxicity that compromises patientcompliance (Solomon et al., 2005; Dia-sie and Harris, 1989; Takimoto et al.,1999). The ICso values obtained in thepresent cell culture model compare fa-vorably with clinicaIly or experimentallyachievable plasma levels, thereby provid-ing clinical relevance for the responsive-ness of the present cell culture model toCLX and 5-FU.

Preclinical animal studies with lowdose combinations of mechanisticallydistinct compounds have demonstratedenhanced efficacy with minimal toxicity(Torrance et al., 2000; Swamy et al.,2006). Consistent with the above obser-vations, data from the present experi-ments provide proof-of-principle evi-dence that combination of low doseCLX+5-FU exhibits greater efficacy forcytostatic growth arrest and for inhibitionof AICF, relative to that exhibited bythese compounds used independently.Furthermore, the higher efficacy of theCLX+5-FU combination is predominant-ly due to enhanced G! phase arrest andinhibition of S+G2/M phase of the cellcycle. The molecular mechanisms re-sponsible for the observed cell cyclemodulation and synergistic/additive in-

teractions of the CLX+5-FU combina-tion, however, remain to be identified.

In conclusion, the data generated fromthis study demonstrate that aberrantlyproliferative Apc mutant ceIls display anincreased risk for carcinogenic transfor-mation, and that the risk is modifiableby combination of mechanistically dis-tinct preventive/therapeutic compounds.These data therefore validate a novelalternative cell culture approach to rapid-ly prioritize efficacious preventive/therapeutic agents and, thereby, reduceand refine tradition al animal studiesidentifying interventional strategiesfor future clinical trials on preventionltherapy of colon cancer.

ReferencesCorpet, D. E. and Pierre, F. (2003). Point:

From animal models to prevention ofcolon cancer: Systemic review ofchemoprevention in Min mice andchoice of the model systems. Cancer.Epid. Biomark. Prevent. 12, 391-400.

Diasio, .B. and Harris B. E. (1989). Clin-ical pharmacology of 5-Fluorouracil.Clin. Pharmacokinetics 6, 215-237.

Fearon, E. R. and Vogelstein. B. (1990).A genetic model for colorectal tumori-genesis. Cell61, 759-767.

Fodde, R., Edelmann, w., Yang, K. et al.(1994). A targeted chain terminationmutation in the mouse APC gene re-sults in intestinal tumors. Prac. Natl.Acad. Sei. USA. 91, 8969-8973.

Fodde, R., Srnits, R. and Cleavers, H.(2001). APC signal transduction andgenetic instability in colorectal cancer.Nat. Rev. Cancer. 1, 55-67.

Ilyas, M., Tomlinson, 1. P. M., Rowan, A.et al. (1997). ß-catenin mutations incell lines established from human col-orectal cancers. Prac. Natl. Acad. Sei.(USA).94,10330-10334.

Jacoby, R. F., Siebert, K., Cole, C. E. etal. (2000). The cyclo-oxygenase in-hibitor Celecoxib is a potent preven-tive and therapeutic agent in the Minmouse model of adenomatous polypo-sis. Cancer Res. 60, 5040-5044.

Jinno, H., Steiner, M. G., Mehta, R. G. etal. (1999). Inhibition of aberrant pro-liferation and induction of apoptosis inHER-2/neu oncogene transformed hu-

ALTEX 24, 1/07

~ ~~-~-------------

man mammary epithelial cells by N-(4-hydroxyphenyl) retinamide. Car-cinogenesis 20, 229-236.

Katdare, M. , Kopelovich, L. and Telang,N. (2002). Efficacy of chemopreven-tive agents for growth inhibition ofAPC [+1-] 1.638N colonic epithelialcells. Int. J. Mol. Med. 10, 427-432.

Katdare, M., Osbome, M. P and Telang,N. T (2002). Soy isoflavone genesteinmodulates cell cycle progression andinduces apoptosis in HER-2/neu ex-pressing human breast epithelial cells.Int. J. Oncol. 21, 809-815.

Katdare, M., Osborne, M. P and Telang,N. T (2003). Novel cell culture modelsfor prevention of human breast cancer(Review). Int. J. Oncol. 22, 509-515.

Meyers, M., Wagner, M. W., Hwang, M.S. et al. (2001). Role of hMLHl DNAmismatch repeir protein in fluoropy-rimidine-rnediated cell death and cellcycle responses. Cancer Res. 61, 5193-5201.

Moser, A. R., Pitot, H. and Dove, W. F(1990). A dominant mutation that pre-disposes to multiple intestinal neopla-sia in the mouse. Science 247, 322-324.

Paulson, J. E., Steffensen, J-L., Loberg,E. M. et al. (2001). Quantitative andqualitative relationship between dis-plastic aberrant crypt foci and tumori-genesis in the Min/+ mouse colon. Can-cer Res. 61,5010-5015.

Potter, J. D. (1996). Risk factors forcolon neoplasia: epidemiology and bi-ology. Eur. 1. Cancer 31A, 1033-1038.

Rao, C. v..Yang, Y.M., Swamy, M. V. etal. (2005). Colonic tumorigenesis inBubRl [+I-]/Apc Min/+ compound mu-tant mice is linked to premature sepa-ration of sister chromatid and en-hanced genomic instability. Proc. Natl.Acad. Sei. (USA) 102,4365-4370.

Russell, W. M. S. and Burch, R. L.(1959). The Principles of Humane Ex-perimental Techniques. London:Methuen.

ALTEX 24, 1/07

Saez, E., Tontonoz, P, Nelson, M. C. etal. (1998). Activators of nuclear recep-tor PPAR-y enhance colon polyp for-mation. Nat. Med. 4, 1058-1061.

Sodir, N. M., Chen, X., Park, P et al.(2006). Smad-3 deficiency prornotestumorigenesis in the distal colon ofApc Min/+ mice. Cancer Res. 66, 8430-8438.

Solomon, S. D., Mc. Murray, J. J., Pfef-fer, M. A. et al (2005). Cardiovascularrisk associated with Celecoxib in aclinical trial for colorectal adenomaprevention. N. Engl. J. Med. 352,1071-1080.

Steinbach, G., Lynch, PM., Phillips, R.K. S. et al. (2000). The effect of Cele-coxib, a cyclo-oxygenase inhibitor, infamilial adenomatous polyposis. N.Engl. J. Med. 342, 1946-1952.

Swamy, M. v., Patlolla, J. M. R., Steele,V. E. et al. (2006). Chemopreventionof familial adenomatous polyposis bylow doses of Atorvastatin and Celecox-ib, given individually and in cornbina-tion to APC Min mice. Cancer Res. 66,7370-7377.

Takimoto, C. H., Yee, L. K., Venzon, D.L. et al. (1999). High inter- and intra-patient variation in 5-Fluorouracilplasma concentrations during a pro-longed drug infusion. Clin. CancerRes. 5, 1347-1352.

Telang, N. and Katdare, M. (2007). Com-binatorial prevention of carcinogenicrisk in a model for familial colon can-cer. Oncol. Rep. 17,909-914.

Telang, N., Li, G. and Katdare, M.(2006). Prevention of carcinogenesisin a model for familial colon cancer. J.Cancer. Res. Ther. 2, SIO.

Telang, N. T, Li, G. and Katdare, M.(2006). Prevention of early onset fa-milial/hereditary colon cancer: Newmodels and mechanistic biomarkers(Review). Int. 1. Oncol. 28, 1523-1529.

Telang, N. T, Narayanan, R., Bradlow,H. L. et al. (1991). Co-ordinated ex-pression of intermediate biomarkers

for tumorigenic transformation in ras-transfected mouse mammary epithelialcells. Breast Cancer Res. Treat. 18,155-163.

Telang, N. T, Osbome, M. P., Sweterl-itsch, L. A. et al. (1990). Neoplastictransformation of mouse mammary ep-ithelial cells by deregulated myc ex-pression. Cell Regulation, 1, 863-872.

Torrance, C. J., Jackson, PE., Mont-gomery, E. et al. (2000). Combinatori-al chemoprevention of intestinal neo-plasia. Nat. Med. 6, 1024-1028.

Zhai, Y. F,Beitttenmiller, H., Wang, B. etal. (1993). Increased expression ofspecific protein tyrosine phosphatasesin human breast epithelial cells trans-formed by the neu oncogene. CancerRes. 53, 2272-2278.

AcknowledgementsThe authors thank Milan Zvanovec andTonye Briggs for their expert technicalassistance. This research has been sup-ported in part by NIH grants CA-29502-SI (N.T.), CA-29502-20 (M.K.), andfunds from The Irving Weinstein Foun-dation (N.T and M.K.).

Correspondence toNitin Te1ang, Ph.D.Strang Cancer Prevention Centre428E, 72nd StreetNew York, NY 10021USAe-mail: entitytoo@cs.com

Meena Katdare, Ph.D.Department of SurgeryWeill Medical College of Comell Uni-versity1300 York Avenue, Box-287New York, NY 10021USAe-mail: msk2004@med.comell.edu

21