Dietary n-6 and n-3 polyunsaturated fatty acids and colorectalcarcinogenesis: results from cultured colon cells, animal models and

human studies�

Yvonne E.M. Dommels a,b,*, Gerrit M. Alink b, Peter J. van Bladeren c,Ben van Ommen c

a WUR/TNO Centre for Food Toxicology, The Netherlandsb Toxicology Group, Division of Toxicology, Wageningen University, Tuinlaan 5, PO Box 8000, 6700 EA Wageningen, The Netherlands

c Toxicology Division, TNO Nutrition and Food Research, Department of Explanatory Toxicology, Utrechtseweg 48, PO Box 360, 3700 AJ Zeist, The

Netherlands

Received 19 October 2001; received in revised form 18 December 2001; accepted 11 January 2002

Abstract

During the past few decades, many studies have been conducted to evaluate the effects of n-6 and n-3 polyunsaturated fatty acids

(PUFAs) on colorectal carcinogenesis. This report provides a brief overview of the recent studies that have been performed in

cultured colon cells, animal models as well as of the population-based and short-term biomarker studies with humans. No

differential effect between n-6 and n-3 PUFAs has been observed in vitro. Results from animal models indicate that n-6 PUFAs have

a tumor enhancing effect, predominantly during the post-initiation phase. n-3 PUFAs may protect against colorectal carcinogenesis

during both the initiation and post-initiation phase. Population-based human studies show little or no associations between n-6 or n-

3 PUFA intake and colorectal cancer. Short-term biomarker studies in humans suggest though that fish oil (FO) supplementation

with high amounts of n-3 PUFAs may protect against colorectal carcinogenesis and that n-6 PUFA supplementation may increase

the risk. # 2002 Elsevier Science B.V. All rights reserved.

Keywords: Colorectal carcinogenesis; Dietary n-6 and n-3 fatty acids; Human colorectal carcinoma cell lines; Animal models; Biomarkers;

Epidemiology

1. Introduction

Colorectal cancer is one of the most common causes

of cancer deaths in the industrialized Western countries.

It is the most prevalent cancer form for men after lung-

and prostate cancer. For females, colorectal cancer is the

second form after breast cancer (Woutersen et al. , 1999).

In the Netherlands the colon cancer incidence is 8600

new cases per year (Visser et al., 2001). In 1998, 4400

people died of colorectal cancer, which is around 12% of

Abbreviations: AA, arachidonic acid (20:4n-6); ACF, aberrant crypt foci; AgNOR, silver-stained nucleolar organizer region protein; AIN,

American Institute of Nutrition; ALA, a-linolenic acid (18:3n-3); AOM, azoxymethane; APC, adenomatous polyposis coli; BHT, butylated hydroxy

toluene; CO, corn oil; COX, cyclooxygenase; COX-1, cyclooxygenase 1; COX-2, cyclooxygenase 2; DAG, diacylglycerol; DBA, Dolichos biflorus

agglutinin; DHA, docosahexaenoic acid (22:6n-3); EPA, eicosapentaenoic acid (20:5n-3); FAP, familial adenomatous polyposis; FO, fish oil; FPTase,

farnesyl protein transferase; GJIC, gap junctional intercellular communication; GLA, g-linolenic acid (18:3n-6); HFCO, high-fat corn oil; HFFO,

high-fat fish oil; HFML, high-fat mixed lipids; HNPCC, Hereditary Non-Polyposis Colorectal Cancer; IM, indomethacin; LA, linoleic acid (18:2n-

6); LFCO, low-fat corn oil; MDA, malondialdehyde; MO, menhaden oil; O, olive oil; OA, oleic acid (18:1n-9); ODC, ornithine decarboxylase; P,

perilla oil (18:3n-3); PGE2, prostaglandin E2; PGE3, prostaglandin E3; PI-PLC, phosphatidylinositol-specific phospholipase C; PKC, protein kinase

C; PLA2, phospholipase A2; PUFAs, polyunsaturated fatty acids; S, safflower oil (18:2n-6); TPK, tyrosine protein kinase.�

PII of original article: S 1 3 8 2 - 6 6 8 9 ( 0 2 ) 0 0 0 0 6 - 6

* Corresponding author. Tel.: �/31-317-484357; fax: �/31-317-484931

E-mail address: yvonne.dommels@algemeen.tox.wau.nl (Y.E.M. Dommels).

Environmental Toxicology and Pharmacology 12 (2002) 233�/244

www.elsevier.com/locate/etap

1382-6689/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.

PII: S 1 3 8 2 - 6 6 8 9 ( 0 2 ) 0 0 0 9 5 - 9

the total cancer deaths (Health Council of the Nether-

lands, 2001). Many factors can be responsible for the

development of colon cancer. Genetic predisposition is

considered as an important risk factor. Two groups ofgenetic predisposition are Hereditary Non-Polyposis

Colorectal Cancer (HNPCC) and Familial Adenoma-

tous Polyposis (FAP). HNPCC is a genetic disease,

which is characterized by an early age of onset and

mutational inactivation of mismatch repair genes.

Patients with FAP may develop hundreds of polyps,

also at early age, due to an inactivation mutation of the

Apc gene (Kinzler and Vogelstein, 1998; Roncucci et al.,2000).

Besides genetic predisposition, also diet is an impor-

tant risk factor for colon cancer. Already in the early

1980s Doll and Peto (1981) estimated that 90% of the

risk of colorectal cancer could be attributed to environ-

mental factors, mostly dietary factors. Accumulating

evidence suggests an association between a high fat

intake and increased risk of colorectal cancer (Doll andPeto, 1981; Woutersen et al., 1999). Moreover, epide-

miological and experimental studies do provide much

evidence that not only the amount of fat consumed, but

also the type of fat is very important.

This mini-review will provide a brief overview of the

current knowledge on the differential effects between n-6

and n-3 polyunsaturated fatty acids (PUFAs) on color-

ectal carcinogenesis as determined in cultured coloncells, animal models and human studies.

2. Nomenclature, dietary sources and metabolism of n-6

and n-3 PUFAs

N-6 and n-3 fatty acids are PUFAs with two or more

double bonds in the carbon atom chain. N-6 and n-3

fatty acids are named after the position of the firstdouble bond from the methyl end of the molecule. For

example, linoleic acid (LA, 18:2n-6) has 18 C-atoms and

two double bonds, with the first double bond at the 6th

carbon atom counted from the methyl end. Linoleic acid

is the parent compound of the n-6 family, whereas a-

linolenic acid (ALA, 18:3n-3) is the parent compound of

the n-3 fatty acid family. These two PUFA families are

considered as essential and must be derived from the diet(Gurr, 1996). Linoleic acid is mostly found in vegetable

seeds and oils such as safflower, soybeans, corn and

sunflower oil. Perilla oil from the Asian beefsteak plant

(Perilla frutescens), linseed oil, rapeseed, walnuts and

blackcurrant oil are rich in a-linolenic acid. a-Linolenic

acid is also present in dark green leafy plants (Bartsch et

al., 1999).

Linoleic acid and a-linolenic acid can be metabolizedto more polyunsaturated fatty acids by a combination of

desaturation and elongation enzymes (Gurr, 1996).

Desaturation enzymes introduce a new double bond in

the carbon chain and elongase enzymes introduce two

new C-atoms. Arachidonic acid (AA, 20:4n-6) is the

major long-chain n-6 PUFA. Eicosapentaenoic acid

(EPA, 20:5n-3) and docosahexaenoic acid (DHA,

22:6n-3) are the major PUFAs of the n-3 family. Marine

fish such as salmon, tuna, herring, mackerel and

anchovy are rich sources of EPA and DHA (Rose and

Connolly, 1999). The marine food chain is based on n-3

fatty acids, which are present in plankton and algae on

which fish feed (Bartsch et al., 1999). Dietary LA is

considered to be the main source of tissue AA, although

lean meats and meat fat are direct dietary AA sources

(Rose and Connolly, 1999). Production of AA and EPA

from LA and ALA, respectively, is thought to proceed

preferentially by D6-desaturation followed by a two-

carbon atom chain elongation and D5-desaturation

(Grammatikos et al., 1994) (see Fig. 1). Metabolism of

n-6 and n-3 fatty acids follows a series of competitive

elongation and desaturation steps, which is limited by

the activity of D6-desaturase, with the n-3 fatty acids

having greater affinities for the enzyme. In consequence,

increasing dietary intake of n-3 fatty acids reduces the

desaturation of LA and so, the production of AA (Rose

and Connolly, 1999).

AA and EPA also show competition for cyclooxy-

genase (COX) enzyme activities. Cyclooxygenase 1 and -

2 (COX-1, COX-2, respectively) introduce two atoms of

oxygen into AA to form the hydroperoxy endoperoxide,

PGG2, which is then reduced by the endoperoxidase

moiety of the enzyme to the hydroxy endoperoxide

PGH2. PGH2 can be a substrate for several enzymes

including PGE synthase, which forms prostaglandin E2

(PGE2) (Fischer, 1997) (see Fig. 2). EPA is a precursor

of the prostaglandins of the 3-series, with three double

bonds. The n-3 fatty acid is converted by the same COX

enzyme to the 3-series endoperoxide PGH3, by way of

PGG3, which can be further metabolized to PGE3 (Rose

and Connolly, 1999; see Fig. 2). EPA supplementation

can thus lead to competitive inhibition of arachidonic

acid metabolism as well as the production of metabolites

such as PGE3. These metabolites are biologically less

active than the corresponding arachidonic acid metabo-

lites like PGE2 (Fischer, 1997). AA and EPA can besides

the cyclooxygenase pathway, also be metabolized by the

Fig. 1. Metabolism of n-6 and n-3 PUFAs.

Erratum234

lipoxygenase pathway and the cytochrome P450 mono-

oxygenase pathway (Krause and Dubois, 2000) to

prostaglandins, thromboxanes and leukotrienes, collec-

tively referred to as eicosanoids. In general, the potency

of eicosanoids derived from the n-3 PUFAs is less than

of those derived from the n-6 PUFAs (Gurr, 1996).

3. In vitro effects of n-6 and n-3 PUFAs on human

colorectal carcinoma cell lines

Several in vitro studies have been carried out to

investigate the effects of essential fatty acids on various

cells in culture (Grammatikos et al., 1994; Jiang et al.,

1998).

In the present review we will focus on the effects of

individual n-6 and n-3 PUFAs on human colorectal

carcinoma cell lines. Table 1 gives a summary of some of

the recent studies.

Biomarkers of carcinogenesis, which are mostly

determined in vitro are cell proliferation and apoptosis.

From the responses of the different cell lines, it can be

concluded that there is no obvious differential effect

between n-6 and n-3 PUFAs on colon cancer cell lines.

In most studies, LA (18:2n-6) and ALA (18:3n-3)

showed no effect on cell proliferation (Awad et al.,

1995; Chen and Istfan, 2000; Collett et al., 2001; Tsai et

al., 1998). Other PUFAs with more double bonds, such

as AA (20:4n-6), EPA (20:5n-3) and DHA (22:6n-3)

caused an overall decrease in cell proliferation (Men-

geaud et al., 1992; Tsai et al., 1998; Clarke et al., 1999;

Chen and Istfan, 2000; Collett et al., 2001; Kim et al.,

2000; Palozza et al., 2000) or increase in apoptosis

(Clarke et al., 1999; Chen and Istfan, 2000). These fatty

acids appear to act directly, because, indomethacin

(IM), an inhibitor of prostaglandin synthesis, did not

Fig. 2. Formation of prostaglandin E from arachidonic acid (AA;

20:4n-6) and eicosapentaenoic acid (EPA; 20:5n-3).

Table 1

Effects of n-6 and n-3 PUFAs on human colorectal carcinoma cell lines as shown by some recent studies

References Colon cell lines Fatty acids Concentration Exposure

time

Parameters Response

Mengeaud et

al., 1992

HT-29

HRT-18

GLA, EPA 0�/100 mg/ml 48 h Cell prolifera-

tion

GLA, EPA: decrease

(HRT18�HT29�Caco2)

Caco-2 Lipid peroxida-

tion

GLA, EPA: increase in MDA. Vitamin E

reduced MDA, IM did not

Awad et al.,

1995

HT-29

LS174T

LA, ALA 30 mM 0�/9 days Cell prolifera-

tion

HT-29: LA, ALA: better growth than 18:0,

no difference between LA and ALA

LS174T: no effect on growth

Tsai et al.,

1998

SIC oncogene

transformant cell

line

LA, GLA,

AA, ALA,

EPA, DHA

1�/50 mg/ml 48 h Cell prolifera-

tion

AA: no effect till 20 mg/ml, inhibitory at

higher concentrations

EPA, DHA: 5�/50 mg/ml inhibitory

LA, ALA, GLA: no significant inhibition

Clarke et al.,

1999

HT-29 EPA 0�/15 mg/ml 0�/8 days Cell prolifera-

tion

EPA: reduction of adherent cells and

increase in floating cells

Apoptosis EPA: increase in floating cells

Chen and Ist-

fan, 2000

HT-29 LA, ALA,

EPA, AA,

DHA

0�/200 mM 24 h Cell number

Apoptosis

LA, ALA: no effect; AA, EPA: no

significant decrease; DHA: inhibition

DHA: induction; BHT reversed apoptosis,

IM did not

Collett et al.,

2001

YAMC-ras DHA, LA 50 mM 72 hours Cell prolifera-

tion

LA: no effect; DHA: inhibition

Cell viability No effect

Apoptosis No effect

Ras membrane

localization

No effect on total Ras protein; LA

increased Ras membrane-to-cytosol ratio

Kim et al.,

2000

Caco-2 LA, EPA,

DHA

100 mM 0�/14 days Cell prolifera-

tion

LA: increase till day 14; EPA, DHA:

increase till day 4, decrease day 4�/14

Palozza et al.,

2000

WiDr EPA 0�/25�/50�/100�/

250 mM

12 and 24

h

Cell prolifera-

tion

EPA: dose-dependantly decrease

Lipid peroxida-

tion

b-carotene diminished EPA-induced MDA

release

Erratum 235

modify the effects. The decrease in cell proliferation was,

however, highly related to lipid peroxidation, as anti-

oxidants such as vitamin E (Mengeaud et al., 1992),

BHT (Chen and Istfan, 2000) and b-carotene (Palozza etal., 2000) diminished the effects.

To better compare the differential effects between n-6

and n-3 PUFAs, the effects of fatty acids with compar-

able chain lengths such as LA versus ALA and AA

versus EPA, should be determined together in one colon

cell line. Only Awad et al. (1995), Tsai et al. (1998) and

Chen et al. (2000) performed these kinds of experiments.

From this limited number of studies it can, however,also be concluded that there is no differential effect

between n-6 and n-3 PUFAs in colon cancer cell lines.

We investigated the effects of n-6 and n-3 PUFAs on

gap junctional intercellular communication (GJIC),

another biomarker of carcinogenesis, in human colon

adenocarcinoma Caco-2 cells. Also no differential

effects between LA versus ALA and AA versus EPA

were observed. All fatty acids inhibited GJIC after long-term incubation (Dommels et al., 2002).

4. Effects of n-6 and n-3 PUFAs in animal models of

colorectal carcinogenesis

Many in vivo studies have been performed to evaluate

the effects of n-6 and n-3 PUFAs on colorectal

carcinogenesis. In 1992, Reddy reviewed the knowledgeat that time on the relationship between different types

of dietary fat and colon carcinogenesis in laboratory

animal models with emphasis on n-3 fatty acids. He

concluded that lack of a colon tumor enhancing effect of

dietary fish oil (FO) was observed during both the

initiation and postinitiation phases. This was mediated

by an effect of FO on ornithine decarboxylase activity

(ODC), colonic secondary bile acids and/or prostaglan-din synthesis. Klurfeld and Bull wrote a review article in

1997 about fatty acids and colon cancer in experimental

animal models. In the present mini-review we will give

the state of the art knowledge on n-6 and n-3 PUFAs on

colorectal carcinogenesis in animal models till 2002.

In Table 2, several recent experimental animal studies

on the effects of n-6 and n-3 PUFAs on colorectal

carcinogenesis are summarized. From these studies itbecame indeed clear that the colon tumor enhancing

effect of corn oil (CO) is most prevalent during the

postinitiation phase and that the inhibiting effect of FO

is observed during both the initiation and postinitiation

phase (Reddy et al., 1991). Not only FO with a high

amount of EPA and DHA but also perilla oil (Onogi et

al., 1996) with high levels of ALA, and individual n-3

fatty acids such as DHA (Takahashi et al., 1997; Oshimaet al., 1995) can inhibit colorectal carcinogenesis.

Reddy and Sugie (1988) performed a study to

investigate the modulating effects of varying levels of

n-3 and n-6 fatty acids during the promotional phase of

colon carcinogenesis in order to determine the optimum

dietary levels of these fatty acids that elicit maximum

inhibition of colon tumors. Inhibition of colon tumorincidence by decreasing the level of dietary CO in the

high fat diets or increasing the ratio of n-3 to n-6 fatty

acids in the diet was, however, not dose-dependant. A

23.5% high- fat diet with only 5.9% Menhaden FO and

17.6% CO had no enhancing effect on the incidence of

total colon tumors and colon adenocarcinomas as

compared with a low-fat 5% CO diet, whereas the

incidence of total colon tumors and adenocarcinomaswas increased in animals fed a high-fat diet containing

23.5% CO. These results indicate that high fat intake is a

necessary but not a sufficient condition for colon tumor

promotion, and that the relative proportions of n-3 and

n-6 fatty acids in the diet are determinants of the high

fat effect (Reddy and Sugie, 1988).

Besides chemically-induced colon tumors, effects of n-

3 and n-6 PUFAs have been investigated on thedevelopment of aberrant crypt foci (ACF) and in

transgenic mice. Oshima et al. (1995) have investigated

the effects of DHA on mouse intestinal polyposis using

Apc gene knockout mice. Dietary DHA decreased

tumor number in female, but not in male mice. This

was the first study that demonstrated that DHA also

inhibits intestinal polyposis induced by an Apc muta-

tion. According to the authors, this may open apossibility for chemopreventive intervention of FAP

by dietary supplementation with DHA. In addition,

Petrik et al. (2000) were the first to report that dietary

EPA also has anti-tumorigenic properties in the

ApcMin/� mice.

ACF have been used as intermediate biomarkers of

colon cancer development in animal studies (Roncucci

et al., 2000). Aberrant crypts are crypts, which appear tobe larger, thicker and darker than normal crypts and

cluster in aggregates, foci. However, it is not definitely

proven that ACFs are true precursors of colon tumors

(Klurfeld and Bull, 1997), because, no correlation

between the number of ACF and the incidence of

carcinomas in rats was found or it was found only in

the left colon (Roncucci et al., 2000). On the other hand,

crypt multiplicity in ACF, i.e. the number of crypts perfocus, seems to be a predictor of tumor incidence. The

explanation of this apparent discrepancy is that most

ACF regress and that only larger foci progress towards

cancer (Roncucci et al., 2000). The prognostic value of

ACF is, however, still a matter of debate. Furihata et al.

(2001) did find a positive correlation between the

incidence of ACF and the incidence of tumors in F344

rats.There are relatively few studies on effects of n-6 and

n-3 PUFAs on ACF formation. Those that have been

performed show a decrease in total ACF incidence and

multiplicity by perilla oil (Onogi et al., 1996), DHA

Erratum236

Table 2

Effects of n-6 and n-3 PUFAs on animal colon cancer models as shown by some recent studies

References Animals, sex and

age

Control diet Experimental diet AOM Experimental diet

supplementation:

(post)initiation

Sacrificed Parameters Results

Reddy and

Sugie, 1988

F344 rats, male, 5

week weanlings

4 weeks modified

AIN-76A (5% LFCO)

5%CO, 23%CO,

1%CO�4%MO,

17.6%CO�5.9%MO,

11.8%CO�11.8%MO;

5.9%CO�17.6%MO

Once

weekly for

2 weeks

4 days after last

injection

38 weeks post

AOM treat-

ment

Tumor incidence

Multiplicity

ODC activity

5.9%MO inhibited

tumor incidence

Multiplicity only

decreased with 17.6%

MO

23.5%CO highest

activity

Reddy et

al., 1991

F344 rats, male, 5

week weanlings

Modified AIN-76A

(LFCO)

LFCO, HFCO, HFFO Once

weekly for

2 weeks

2 weeks before injec-

tion, 3 weeks after

injection subgroups

changed:

LFCO0HFCO,

HFCO0HFFO,

HFFO0HFCO

36 weeks post

AOM treat-

ment

Tumor incidence

Tumor multiplicity

HFCO increased

incidence+multiplicity

during postinitation

period,

HFFO reduced

incidence�multiplicity

during initiation and/or

postinitiation

Rao and

Reddy, 1993

F344 rats, male, 5

week weanlings

Modified AIN-76A

(LFCO)

LFCO, HFCO, HFFO Once

weekly for

2 weeks

After 1 week control

diet, 2 weeks before

injection

5 days post

AOM treat-

ment

ODC activity

TPK activity

PGE2, 6-keto PGF1

HFCO: increase;

HFFO: reduction

HFFO: slightly

reduction

Oshima et

al., 1995

ApcD716 mice (129/

Sv�C57BL/6J),

male�female, 3

weeks

AIN-76A 3% DHA as ethyl ester

�AIN-76A

7 weeks After 7 weeks

diet

Polyp number DHA: no effect in

males, decrease in

females

Onogi et al.,

1996

F344 rats, male, 5�/

6 weeks

Basal laboratory diet O12: 12% olive oil; S12: 12%

safflower oil; P12: 12% perilla

oil; P6O6: 6% perilla oil�6%

olive oil; P3O9: 3% perilla

oil�9% olive oil

Once

weekly for

3 weeks

1 week before first

injection

4 weeks after

first injection

ACF

PGE2

Cell proliferation

c-H-Ras expression

Perilla oil: less ACF

(dose-dependant)

P12: less PGE2

Reduced by perilla oil

Reduced by perilla oil

Reddy et al.,

1996

F344 rats, male,

weanlings

2 weeks modified

AIN-76A (LFCO)

LFCO, HFCO, HFFO Once

weekly for

2 weeks

1 day after last

injection

1, 12 and 36

weeks post

AOM treat-

ment

Bacterial 7a-dehy-

droxylase

PI-PLC

Fecal bile acids

LFCO, HFFO: decrease

HFFO: decrease

HFFO: increase in

primary and decrease in

secondary bile acids

DAG Kinase HFCO: increase

Total PKC HFFO: inhibition in

cytosol

Erra

tum

23

7

Table 2 (Continued )

References Animals, sex and

age

Control diet Experimental diet AOM Experimental diet

supplementation:

(post)initiation

Sacrificed Parameters Results

Rao et al.,

1996

Chang et al.,

1997

F344 rats, male, 5

week weanlings

Sprague-Dawley

rats, male, 21 days

weanlings

Modified AIN-76A

(LFCO)

Standard rat chow

LFCO, HFCO, HFFO

CO, FO

Once

weekly for

2 weeks

Once

weekly for

2 weeks

1 day after last

injection

1 week before

injection

1, 12 and 36

weeks post

AOM treat-

ment

16 and 34

weeks post

AOM treat-

ment

PLA2�PI-PLC

activity

COX activity

Cell proliferation

Cell differentiation

Apoptosis

PLA2: no difference;

PI-PLC: decrease by

LFCO, HFFO

HFFO: decrease

CO higher proliferation in

proximal colon

FO more differentiated in

proximal�distal colon

FO more apoptosis in

proximal�distal colon

Adenocarcinoma

incidence

CO, 70.3%; FO, 56.1%

Singh et al.,

1997b

F344 rats, male, 4

week weanlings

2 weeks modified

AIN-76A (LFCO)

LFCO, HFCO, HFFO Once

weekly for

2 weeks

1 day after last

injection

1, 12 and 36

weeks post

AOM treat-

ment

Tumor incidence

Multiplicity

COX-1 and COX-2

protein expression

LFCO, HFCO, HFFO:

57, 76, 40%

0.7390.78, 1.3891.24,

0.45 90.6

COX-1: no differences;

COX-2: higher in HFCO

Singh et al.,

1997a

F344 rats, male, 4

week weanlings

2 weeks modified

AIN-76A (LFCO)

LFCO, HFCO, HFFO Once

weekly for

2 weeks

1 day after last

injection

1, 12 and 36

weeks post

AOM treat-

ment

Tumor incidence

Multiplicity

Ras-p21 protein ex-

pression

LFCO, HFCO, HFFO:

57, 76, 40%

0.7390.78, 1.3891.24,

0.45 90.6

HFCO: increased ras-p21;

HFFO: increased cytoso-

lic ras

Takahashi et

al., 1997

F344 rats, male, 6

weeks

AIN-76A 1 ml DHA as ethyl ester/water

5� a week by gastric incuba-

tion

Once

weekly for

2 weeks

1 day before injection 4, 12 and 36

weeks post

AOM treat-

ment

ACF

PGE2, PUFAs

Cancer incidence

Cancer multiplicity

DHA: at 4 and 12 weeks;

76% and 62% of control

values

DHA: lower PGE2 � AA;

No difference

AOM�water: 3.6592.18;

AOM�DHA: 2.4191.58

Singh et al.,

1998

F344 rats, male, 4

week weanlings

2 weeks modified

AIN-76A (LFCO)

LFCO, HFCO, HFFO Once

weekly for

2 weeks

1 day after last

injection

1, 12 and 36

weeks post

AOM treat-

ment

Tumor incidence

Multiplicity

LFCO, HFCO, HFFO:

57, 76, 40%

0.7390.78, 1.3891.24,

0.4590.6

FPTase protein ex-

pression

HFFO reduced FPTase

expression

Davidson et

al., 1999

Sprague-Dawley

rats, male, weanl

ings

Standard rat chow CO, FO Once

weekly for

2 weeks

1 week before injec-

tion

34 weeks

post AOM

treatment

Ras protein

FPTase activity

ERK protein

Ras mutations

CO: higher membrane

No difference

No difference

Trend towards greater

incidence in CO

Erra

tum

23

8

Table 2 (Continued )

References Animals, sex and

age

Control diet Experimental diet AOM Experimental diet

supplementation:

(post)initiation

Sacrificed Parameters Results

Hong et al.,

2000

Sprague-Dawley

rats, male, weanl-

ings

CO, FO 1 injection 2 weeks before

injection

3, 6, 9 and 12

h post AOM

treatment

O6-methylguanine

adducts

O6-MG-DNA-

methyltransferase

Apoptosis

Lower levels at 6, 9 and 12

h in FO group

No difference

Increase in FO group

Petrik et al.,

2000

C57BL/6J

ApcMin/� mice,

male, 38�/43 days

Non-purified diet Fat-free AIN-93G diet�70 g/

kg soybean oil, supplemented

with: OA; OA�AA;

OA�EPA; AA�EPA

8 weeks After 8 weeks

diet

Tumor number

Fatty acids

Prostaglandins

EPA group had 68 and

54% fewer tumors than

OA and AA groups

EPA-group had less AA;

EPA-group had lower le-

vels of PGE2

Rao et al.,

2001

F344 rats, male, 5

week weanlings

Modified AIN-76A

(LFCO)

LFCO, HFML, HFFO Once

weekly for

2 weeks

One day after last

injection

8, 23 and 38

weeks post

AOM treat-

ment

ACF

Tumors

Apoptosis

Cox-2 activity

Cox-2 protein

2 to 3 fold increase in

multiplicity (]4) for

HFML

HFML: higher incidence

and multiplicity

HFML: less apoptosis

HFML: increased activity

HFML: increased

expression

Erra

tum

23

9

(Takahashi et al., 1997) and HFFO (Rao et al., 2001).

These studies only focussed on the protecting effect of n-

3 PUFAs during the postinitiation phase. This experi-

mental design, however, superimposes effects on initia-tion such as carcinogen metabolism. It has been

demonstrated that FO reduced the azoxymethane

(AOM)-induced k-Ras mutations and decreased mem-

brane Ras expression (Davidson et al., 1999) when given

for and after initiation. These results indicate that FO

may protect against colon carcinogenesis by either

decreasing DNA adduct formation and/or enhancing

DNA repair. Hong et al. (2000) determined the ability ofFO and CO to simultaneously modulate O6-methylgua-

nine DNA adduct formation (DNA damage), removal

by O6-methylguanine-DNA-methyltransferase (repair)

and deletion (apoptosis). No main effect of diet on O6-

methylguanine-DNA-methyltransferase was found.

However, FO enhanced apoptosis combined with a

reduction in adduct formation. This may account, in

part, for the observed protective effect of n-3 PUFAsagainst experimentally induced colon cancer during the

initiation phase (Reddy et al., 1991). The protective

effect may also be explained by modulation of bio-

transformation enzymes related to carcinogen activa-

tion, thereby altering the amounts and activities of

oxidative (Phase 1) and conjugative (Phase 2) xenobiotic

metabolizing enzymes (Hong et al., 2000; Reddy, 1992).

Additional experiments are needed to address thishypothesis.

Various mechanisms have been postulated to explain

the enhancing effect of a high fat CO diet and the

protecting effect of a high fat FO diet during the

promotion phase of carcinogenesis. These mechanisms

include as stated above modulation of colonic mucosal

ODC activity, colonic secondary bile acids and/or PGE2

synthesis. Secondary bile acids can increase ODCactivity and cell proliferation and act as tumor promo-

ters (Rao and Reddy, 1993). It has been shown that

arachidonic acid metabolites are involved in increased

secondary bile acid production and the induction of

tissue ODC activity (Reddy et al., 1996; Rao and Reddy,

1993). It is possible that diets rich in n-3 fatty acids

result in decreased levels of arachidonic acid and its

metabolites and thereby inhibit tissue ODC activity andcell proliferation.

n-3 PUFAs have been reported to inhibit the produc-

tion of the type-2 series of eicosanoids, including PGE2,

from arachidonic acid (Onogi et al., 1996; Rao and

Reddy, 1993; Takahashi et al., 1997). Endogenous PGE2

has been shown to promote rat colon tumors and COX-

inhibitors that prevent prostaglandin production such as

IM can block the development of colon carcinomas(Inaba et al., 1999). Therefore, the mechanism respon-

sible for the inhibitory effects of n-3 PUFAs on color-

ectal tumors may partly be related to inhibition of PGE2

synthesis from AA and reduction of the AA levels itself

(Takahashi et al., 1997). Singh et al. (1997b) observed

that n-3 PUFAs inhibit AOM-induced expression of

COX-2, whereas n-6 PUFAs enhance levels of AOM-

induced COX-2 expression. Overexpression of COX-2has been reported in 90% of colon tumors and pre-

malignant colorectal adenomas. Therefore, the assay of

COX-2 expression may be used to monitor the process

of colon carcinogenesis (Singh et al., 1997b). Also Rao

et al. (2001) found that n-3 PUFAs in a high-fat fish oil

(HFFO) diet inhibited the levels of COX-2 and AA

metabolites (eicosanoids). They suggest that modulation

of AA metabolism through COX activity play a role inapoptosis, since the HFFO diet also enhanced apopto-

sis.

Chang et al. (1997) investigated whether the protec-

tive effect of dietary FO is mediated through changes in

proliferation, differentiation or apoptosis, all intermedi-

ate biomarkers for colon tumor development during the

promotion phase of tumorigenesis. Dolichos biflorus

agglutinin (DBA) binding (marker of differentiation)was higher for FO versus CO fed animals in both

proximal and distal colon. There was also a greater

number of apoptotic cells/crypt column in both prox-

imal and distal colon after feeding with FO compared

with CO. However, changes in cell proliferation did not

predict the beneficial effect of FO versus CO. Though,

Onogi et al. (1996) found that perilla oil significantly

reduced silver-stained nucleolar organizer region protein(AgNORs) count/nucleus, suggesting that perilla oil

decreased the number of cells in S-phase and thus

decreased cell proliferation.

Another inconsistent relationship exists between FO

and farnesyl protein transferase (FPTase) expression.

FPTase is an enzyme which catalyses the biological

activation of Ras proteins. The first step in this process

is the transfer of a 15-carbon isoprene, farnesyl, to thecysteine residue of the C-terminal tetrapeptide

sequence*/CAAX, of Ras precursors, which is cata-

lysed by FPTase. Farnesylation of Ras precursors is a

critical step during post-translational modification of

Ras oncoproteins, thereby enabling their anchorage to

the plasma membrane. Singh et al. (1998) demonstrated

that consumption of high amounts of FO reduced the

levels of FPTase comparing to high CO levels, thusinhibiting post-translational processing of Ras precur-

sors resulting in decreased Ras functioning. However,

Davidson et al. (1999) reported that perturbation in the

farnesylation of Ras is not a decisive factor regulating

membrane localization during malignant transforma-

tion in the colon. They found no differences in farnesyl

protein transferase activity and prenylation state of Ras

between tumors and uninvolved mucosa.Overall results from the animal models suggest that n-

3 fatty acids may protect against colorectal carcinogen-

esis and that n-6 fatty acids may enhance the risk of

colorectal carcinogenesis. The effects of these fatty acid

Erratum240

families on the complex and multistage process of

carcinogenesis can, however, not easily be explained

by one mechanism.

5. Effects of n-6 and n-3 PUFAs on colon cancer in

humans

Epidemiological studies have provided much of the

information available about diet and cancer risk in

humans. Migration studies have shown that the low

mortality rates from colon cancer in Japan increase

when Japanese migrate to the US and adapt to aWestern diet, which contains for example higher levels

of n-6 PUFAs, such as linoleic acid. Like Japanese, also

Eskimos have a relatively low incidence of colon cancer

(Woutersen et al., 1999). Both populations consume

large quantities of fish (1000�/3000 mg per day), which is

rich in n-3 fatty acids, such as EPA and DHA, whereas

the average consumption of n-3 PUFAs in western

countries is only 100 mg per day (Hong et al., 2000). Adetailed correlation analysis of international data ex-

plored the relationship between estimates of per capita

disappearance of PUFAs (total, fish n-3 and n-6) and

the incidence of cancers of the breast, colon, prostate,

lung and cervix (Hursting et al., 1990). In this study, an

estimate of n-6 PUFA intake was obtained by subtract-

ing fish n-3 PUFAs from total PUFA intake. Fish n-3

PUFAs were used, because, the database used for n-3fatty acids in foods other than fish was substantially

incomplete. Fish n-3 PUFAs showed a nonsignificant

negative association with the cancer sites studied. The

estimate of n-6 PUFAs gave virtually the same results as

total PUFAs, namely a positive correlation with the

incidence of breast and prostate cancers but not for

colon cancer. Zock and Katan (1998) reviewed the

epidemiological literature on linoleic acid (18:2n-6)intake and cancer risk. From this meta-analysis of

case-control and prospective cohort studies, it seemed

unlikely that a high intake of linoleic acid substantially

raises the risk of colorectal cancer in humans but a small

increase in risk could not be excluded. Terry et al. (2001)

examined the relation between total fat, fat types,

specific fatty acids and the risk of colorectal cancer in

a population-based prospective cohort of Swedishwomen. They observed no associations for any of the

specific PUFAs, including linoleic acid, a-linolenic acid,

EPA, DHA or the sum of n-6 fatty acids or n-3 fatty

acids and colorectal cancer.

An estimate of the dietary intake of individual fatty

acids by for example food composition tables or food

questionnaires can, however, be biased in several ways.

Assessment of fatty acid composition in lipid subfrac-tions in relation to widely varying fatty acid intake

patterns will provide more precise indicators of recent

dietary intake (Bartsch et al., 1999). Bakker et al. (1997)

used adipose tissue fatty acid composition as a biomar-

ker for long-term dietary exposure to fatty acids. In this

study, no associations between n-3 and n-6 fatty acids in

adipose tissue and incidences of cancer of the breast,prostate and colon were found. Thus, despite the

findings in animal models, data from population-based

studies show little or no associations between PUFA

intake and colorectal cancer risk.

There have been a couple of short-term human

intervention studies on FO supplementation and bio-

markers of colon cancer risk. Bartram et al. (1993)

investigated the effect of FO on cell proliferation, ODCactivity and PGE2 release in 12 healthy volunteers. In

addition to a controlled basal diet, the test subjects

received either 550 mg FO (4,4 g n-3 fatty acids per day)

or CO supplements (4�/5 capsules per day) for two 4-

week periods in a double-blind crossover trial. A

washout period of 4 weeks was allowed between the

two study periods to maintain basal conditions for each

supplementation. All three biomarkers were signifi-cantly lower during the FO than the CO period,

suggesting an inhibiting effect of FO supplementation

on colon cancer development. In another short-term

intervention study, Bartram et al. (1996) investigated the

effect of FO consumption in healthy subjects on fecal

parameters. No significant differences were noted for

fecal activities of b-glucuronidase, b-glucosidase and

sulfatase, nor did FO or CO consumption significantlyaffect fecal bile acid excretion. However, daily excretion

of the putative colon carcinogen 4-cholesten-3-one was

significantly lower in the FO than in the CO period. The

authors suggest that this may be another mechanism by

which FO may inhibit colon cancer development.

The effect of n-3 fatty acids on patients who have had

a bowel neoplasm and thus at high risk of developing a

second neoplasm was studied by Huang et al. (1996).Patients with stage 1 or stage 2 colon carcinoma or

adenomatous polyps were randomized to consume

either 9 g per day n-3 fatty acid or placebo capsules.

Patients in the n-3 group experienced significant reduc-

tions in BrdU labeling index (cell proliferation) after 3

months of supplementation. The effect was only found

in patients with hyperproliferative baseline levels. There-

fore, characteristics of mucosal proliferation at baselinemay be a crucial factor for the effect of n-3 fatty acid

supplementation. Anti et al. (1994) attempted to identify

an optimum dose for FO supplementation and evalu-

ated the persistence of its effects during long-term

administration. Sixty patients with sporadic adenomas

received 2.5, 5.1 or 7.7 g FO per day or placebo for 30

days. Significantly reduced proliferation was also only

observed in patients with abnormal baseline patterns.The effects persisted during long-term (6 months) FO

treatment (2.5 g per day). The data stated above suggest

that if the n-3/n-6 fatty acid equilibrium indeed affects

colonic tumor development, its mode of action seems to

Erratum 241

be based on inhibition of cell proliferation, most likely

through the cyclooxygenase/prostaglandin pathways.

Other mechanisms may also be involved, but have not

been investigated yet in the human population.Moreover, there is a lack of consistency between the

various studies. Gee et al. (1999) determined the effect of

FO on the fatty acid composition of colonic mucosa and

on rectal crypt cell proliferation in patients (28 males

and 21 females) undergoing surgery for colonic carci-

noma. Consumption of FO capsules (1.4 g EPA and 1.0

g DHA) for an average of 12.3�//�/0.5 days lead to

incorporation of EPA into colonic mucosa but had nodetectable effect on epithelial cytokinetics. During a

long-term intervention trial of DHA-concentrated fish

oil capsules (2.2 g DHA and 0.6 g EPA per day), three

patients with FAP and two patients with more than 30

colorectal polyps received these FO capsules for 1 or 2

years. A marked increase or decrease in the number of

polyps was not observed, but the three patients with

FAP developed either endometrial or lung cancer after12 months or colon cancer after 24 months, respectively,

(Akedo et al., 1998). Both studies showed no protective

effect of FO consumption. In these studies, however,

beneficial effects of n-3 fatty acids may have been missed

due to either a short supplementation period or a limited

number of subjects.

Nair et al. (1997) performed a dietary study on n-6

PUFA intake and biomarkers of cancer risk. This studyrevealed that high intake of n-6 PUFAs increased

malonaldehyde-derived adducts in both male and female

subjects. Etheno adducts in white blood cells were not

elevated in males but about 40 times higher in females.

This might be a possible link between increased intake of

dietary n-6 PUFAs, DNA damage and elevated risk of

for example colon cancer.

Taken together, the data of the population-basedstudies suggest that n-6 and n-3 PUFA intake do not

influence colorectal cancer. If associations exist, they are

likely to be weak. Short-term intervention studies reveal

though that fish oil supplementation may protect

against colorectal carcinogenesis in humans and that

n-6 PUFA intake may increase the risk.

6. General discussion

In this review we provided a brief overview of

experimental and epidemiological studies that have

been performed in the past decade on the influence of

n-6 and n-3 PUFAs on colorectal carcinogenesis. No

differential effects between n-6 and n-3 PUFAs on in

vitro colonic cell proliferation and apoptosis, both

biomarkers of carcinogenesis, have been observed.Results from animal models indicate that n-6 PUFAs

have a tumor enhancing effect, predominantly during

the postinitiation phase. n-3 PUFAs may protect against

colorectal carcinogenesis during both the initiation and

postinitiation phase. Population-based studies show

little or no associations between dietary n-6 or n-3

PUFA intake and colorectal cancer. If associations existin the human population, they are likely to be weak.

Short-term biomarker studies in humans suggest though

that fish oil supplementation may protect against color-

ectal carcinogenesis and that n-6 PUFA intake may

increase the risk.

Several explanations could be given for the discre-

pancy in results between in vitro, animal and human

studies. Colon cell lines are used to investigate themechanistic effects of individual n-6 and n-3 PUFAs on

colorectal carcinogenesis in vitro. However, lipid per-

oxidation is a complication that must be considered in

PUFA-supplemented cell cultures with low levels of

antioxidants or enzymatic detoxification enzymes

(Grammatikos et al., 1994; Schonberg et al., 1997)

compared with the in vivo situation. This may be an

explanation why AA (20:4n-6) and EPA (20:5n-3) donot show a differential effect and are both cytotoxic for

human colon carcinoma cell lines. Furthermore, con-

clusions from in vitro experiments are obtained from

colorectal carcinoma cell lines and not from normal

cells, which develop to cancer cells, contrary to most

animal studies. Carcinoma cell lines may not be sensitive

anymore to fatty acid treatment, or react in a different

way due to mutations which turn cell signaling pathwayson or off. Therefore, it would be better to also

investigate the effects of n-6 and n-3 PUFAs on primary

cultures of normal colon cells or better differentiated

colon tumor cells. A good example may be the human

colonic adenocarcinoma cell line, HCA-7. These cells

retained some of the morphological and functional

polarity exhibited by normal colonic epithelium (Kirk-

land, 1985).Another reason for the discrepancy between the in

vitro and in vivo results may be the lack of interaction in

vitro with other cell systems and integrative functions

involved in the process of carcinogenesis. In general, cell

lines can be useful to study mechanistic pathways and

parts of cell biological and physiological processes, but

care should be taken in interpretation of the results.

As generally known, animal models have severaladvantages over in vitro models in studying colorectal

carcinogenesis. They can for example be used to test

hypotheses about mechanisms in a physiological con-

text. Another advantage, also comparing to human

studies, is the possibility to quantify the development

of tumors in a short time. Inherent to these animal

experiments is the use of high amounts of isolated

dietary constituents. In most animal experiments de-scribed in this review, rats were fed experimental diets,

supplemented with high amounts (9�/23.5 wt.%) of CO

or fish oil as the only fat source at the expense of starch.

Zock et al. (1998) revealed that in the population-based

Erratum242

studies they reviewed, the intake difference for linoleic

acid was between 5�/25 g per day, which is comparable

to a range intake of 4�/10% of daily energy. This is much

lower than the n-6 or n-3 PUFA intake in the animalexperiments, which is about 18�/48% of daily energy.

Therefore, a narrow range of n-6 or n-3 PUFA intake in

the population may be a reason why associations

between PUFA intake and colorectal cancer were not

found in population-based studies compared with the

animal data. It can thus not be excluded that cancer risk

within populations would be affected by larger differ-

ences in PUFA intake, which is the case in the humanshort-term biomarker studies on for example cell pro-

liferation. These studies revealed that extra supplemen-

tation of fish oil (2.5�/9 g n-3 PUFAs per day) in

addition to a normal diet may protect against colorectal

carcinogenesis in humans. Intervention studies on real

disease endpoints such as polyps or tumors have,

however, not been conducted yet for PUFAs intake

and colorectal cancer.The efficacy of dietary n-3 PUFAs can depend on the

levels of AA that are found in mixed diets and may

therefore, account for some of the variability of results

observed among studies using human subjects (Whelan,

1996). The efficacy of dietary n-3 PUFAs is also

dependant on the baseline patterns of for example cell

proliferation, as described in the short-term biomarker

studies. In two studies described, significantly reducedproliferation after fish oil supplementation was only

observed in patients at high risk of developing colon

cancer due to abnormal baseline patterns of cell

proliferation. This situation is comparable to the in

vivo studies in which the animals are also at higher risk

of developing colon cancer due to chemically induced

AOM mutations or germline Apc mutations. Therefore,

this may be another reason why there were no associa-tions found between n-6 or n-3 PUFA intake and

colorectal carcinogenesis in the population-based stu-

dies of the healthy population.

In conclusion, despite the lack of a differential effect

in vitro, animal studies and moreover, data from short-

term human biomarker studies indicate a promising

beneficial effect of fish oil supplementation, with high

amounts of n-3 PUFAs, on colorectal carcinogenesis.However, more human dietary intervention studies are

needed to evaluate the real implication of fish oil

supplementation on colon tumor development in hu-

mans at high and low risk of developing colon cancer.

References

Akedo, I., Ishikawa, H., Nakamura, T., Kimura, K., Takeyama, I.,

Suzuki, T., Kameyama, M., Sato, S., Matsuzawa, Y., Kakizoe, T.,

Otani, T., 1998. Three cases with familial adenomatous polyposis

diagnosed as having malignant lesions in the course of a long-term

trial using docosahexaenoic acid (DHA)-concentrated fish oil

capsules. Jpn. J. Clin. Oncol. 28, 762�/765.

Anti, M., Armelao, F., Marra, G., Percesepe, A., Bartoli, G.M.,

Palozza, P., Parrella, P., Canetta, C., Gentiloni, N., De Vitis, I.,

1994. Effects of different doses of fish oil on rectal cell proliferation

in patients with sporadic colonic adenomas. Gastroenterology 107,

1709�/1718.

Awad, A.B., Kamei, A., Horvath, P.J., Fink, C.S., 1995. Prostaglandin

synthesis in human cancer cells: influence of fatty acids and

butyrate. Prostaglandins Leukotrienes Essent Fatty Acids 53,

87�/93.

Bakker, N., Van’t Veer, P., Zock, P.L., 1997. Adipose fatty acids and

cancers of the breast, prostate and colon: an ecological study. Int.

J. Cancer 72, 587�/591.

Bartram, H.P., Gostner, A., Kelber, E., Dusel, G., Weimer, A.,

Scheppach, W., Kasper, H., 1996. Effects of fish oil on fecal

bacterial enzymes and steroid excretion in healthy volunteers:

implications for colon cancer prevention. Nutr. Cancer 25, 71�/78.

Bartram, H.P., Gostner, A., Scheppach, W., Reddy, B.S., Rao, C.V.,

Dusel, G., Richter, F., Richter, A., Kasper, H., 1993. Effects of fish

oil on rectal cell proliferation, mucosal fatty acids, and prosta-

glandin E2 release in healthy subjects. Gastroenterology 105,

1317�/1322.

Bartsch, H., Nair, J., Owen, R.W., 1999. Dietary polyunsaturated fatty

acids and cancers of the breast and colorectum: emerging evidence

for their role as risk modifiers. Carcinogenesis 20, 2209�/2218.

Chang, W.C., Chapkin, R.S., Lupton, J.R., 1997. Predictive value of

proliferation, differentiation and apoptosis as intermediate markers

for colon tumorigenesis. Carcinogenesis 18, 721�/730.

Chen, Z.Y., Istfan, N.W., 2000. Docosahexaenoic acid is a potent

inducer of apoptosis in HT-29 colon cancer cells. Prostaglandins

Leukotrienes Essent Fatty Acids 63, 301�/308.

Clarke, R.G., Lund, E.K., Latham, P., Pinder, A.C., Johnson, I.T.,

1999. Effect of eicosapentaenoic acid on the proliferation and

incidence of apoptosis in the colorectal cell line HT29. Lipids 34,

1287�/1295 (published erratum appears in Lipids 2000

Jan;35(1):113).

Collett, E.D., Davidson, L.A., Fan, Y.Y., Lupton, J.R., Chapkin,

R.S., 2001. n-6 and n-3 polyunsaturated fatty acids differentially

modulate oncogenic Ras activation in colonocytes. Am. J. Physiol.

Cell. Physiol. 280, C1066�/C1075.

Davidson, L.A., Lupton, J.R., Jiang, Y.H., Chapkin, R.S., 1999.

Carcinogen and dietary lipid regulate ras expression and localiza-

tion in rat colon without affecting farnesylation kinetics. Carcino-

genesis 20, 785�/791.

Doll, R., Peto, R., 1981. The causes of cancer: quantitative estimates of

avoidable risks of cancer in the United States today. J. Natl.

Cancer Inst. 66, 1191�/1308.

Dommels, Y.E.M., Alink, G.M., Linssen, J.P.H., van Ommen, B.

2002. Effects of n-6 and n-3 polyunsaturated fatty acids on gap

junctional intercellular communication (GJIC) during spontaneous

differentiation of the human colon adenocarcinoma cell line Caco-

2, Nutrition and Cancer (accepted for the publication).

Fischer, S.M., 1997. Prostaglandins and cancer. Front. Biosci. 2,

d482�/d500.

Furihata, T., Kawamata, H., Ohsugi, R., Sato, S., Kubota, K.,

Fujimori, T., 2001. Colitis enhances the colorectal carcinogenesis

in rats: correlation between the incidence of aberrant crypt foci and

the incidence of tumors. Nippon Shokakibyo Gakkai Zasshi 98,

525�/532.

Gee, J.M., Watson, M., Matthew, J.A., Rhodes, M., Speakman, C.J.,

Stebbings, W.S., Johnson, I.T., 1999. Consumption of fish oil leads

to prompt incorporation of eicosapentaenoic acid into colonic

mucosa of patients prior to surgery for colorectal cancer, but has

no detectable effect on epithelial cytokinetics. J. Nutr. 129, 1862�/

1865.

Erratum 243

Grammatikos, S.I., Subbaiah, P.V., Victor, T.A., Miller, W.M., 1994.

Diverse effects of essential (n-6 and n-3) fatty acids on cultured

cells. Cytotechnology 15, 31�/50.

Gurr, M., 1996. Fats. In: Garrow, J.S., James, W.P.T. (Eds.), Human

Nutrition and Dietetics. Churchill Livingstone, London, pp. 77�/

102.

Health Council of the Netherlands, 2001. Population screening for

colorectal cancer. The Hague: Health Council of the Netherlands

Publication number 2001/01.

Hong, M.Y., Lupton, J.R., Morris, J.S., Wang, N., Carroll, R.J.,

Davidson, L.A., Elder, R.H., Chapkin, R.S., 2000. Dietary fish oil

reduces O6-methylguanine DNA adduct levels in rat colon in part

by increasing apoptosis during tumor initiation [In Process

Citation]. Cancer Epidemiol. Biomarkers. Prev. 9, 819�/826.

Huang, Y.C., Jessup, J.M., Forse, R.A., Flickner, S., Pleskow, D.,

Anastopoulos, H.T., Ritter, V., Blackburn, G.L., 1996. n-3 Fatty

acids decrease colonic epithelial cell proliferation in high- risk

bowel mucosa. Lipids 31 (Suppl), S313�/S317.

Hursting, S.D., Thornquist, M., Henderson, M.M., 1990. Types of

dietary fat and the incidence of cancer at five sites. Prev. Med. 19,

242�/253.

Inaba, A., Uchiyama, T., Oka, M., 1999. Role of prostaglandin E2 in

rat colon carcinoma. Hepatogastroenterology 46, 2347�/2351.

Jiang, W.G., Bryce, R.P., Horrobin, D.F., 1998. Essential fatty acids:

molecular and cellular basis of their anti-cancer action and clinical

implications. Crit. Rev. Oncol. Hematol. 27, 179�/209.

Kim, E., Kim, W., Hang, Y., Ha, Y., Bach, L., Park, J., 2000.

Inhibition of Caco-2 cell proliferation by (n-3) fatty acids: possible

mediation by increased secretion of insulin-like growth factor

binding protein-6. Nutr. Res. 20, 1409�/1421.

Kinzler, K.W., Vogelstein, B., 1998. Colorectal Tumors. In: Vogel-

stein, B., Kinzler, K.W. (Eds.), The Genetic Basis of Human

Cancer. McGraw Hill, Baltimore.

Kirkland, S.C., 1985. Dome formation by a human colonic adeno-

carcinoma cell line (HCA-7). Cancer Res. 45, 3790�/3795.

Klurfeld, D.M., Bull, A.W., 1997. Fatty acids and colon cancer in

experimental models. Am. J. Clin. Nutr. 66, 1530S�/1538S.

Krause, W., Dubois, R.N., 2000. Eicosanoids and the large intestine.

Prostaglandins Lipid Mediat. 61, 145�/161.

Mengeaud, V., Nano, J.L., Fournel, S., Rampal, P., 1992. Effects of

eicosapentaenoic acid, gamma-linolenic acid and prostaglandin E1

on three human colon carcinoma cell lines. Prostaglandins

Leukotrienes Essent Fatty Acids 47, 313�/319.

Nair, J., Vaca, C.E., Velic, I., Mutanen, M., Valsta, L.M., Bartsch, H.,

1997. High dietary v-6 polyunsaturated fatty acids drastically

increase the formation of etheno-DNA base adducts in white blood

cells of female subjects. Cancer Epidemiol Biomarkers Prev. 6(8),

597�/601.

Onogi, N., Okuno, M., Komaki, C., Moriwaki, H., Kawamori, T.,

Tanaka, T., Mori, H., Muto, Y., 1996. Suppressing effect of perilla

oil on azoxymethane-induced foci of colonic aberrant crypts in

rats. Carcinogenesis 17, 1291�/1296.

Oshima, M., Takahashi, M., Oshima, H., Tsutsumi, M., Yazawa, K.,

Sugimura, T., Nishimura, S., Wakabayashi, K., Taketo, M.M.,

1995. Effects of docosahexaenoic acid (DHA) on intestinal polyp

development in Apc delta 716 knockout mice. Carcinogenesis 16,

2605�/2607.

Palozza, P., Calviello, G., Maggiano, N., Lanza, P., Raneletti, F.O.,

Bartoli, G.M., 2000. Beta-carotene antagonizes the effects of

eicosapentaenoic acid on cell growth and lipid peroxidation in

WiDr adenocarcinoma cells. Free Radic. Biol. Med. 28, 228�/234.

Petrik, M.B., McEntee, M.F., Chiu, C.H., Whelan, J., 2000. Antagon-

ism of arachidonic acid is linked to the antitumorigenic effect of

dietary eicosapentaenoic acid in ApcMin/� mice. J. Nutr. 130 (5),

1153�/1158.

Rao, C.V., Reddy, B.S., 1993. Modulating effect of amount and types

of dietary fat on ornithine decarboxylase, tyrosine protein kinase

and prostaglandins production during colon carcinogenesis in male

F344 rats. Carcinogenesis 14, 1327�/1333.

Rao, C.V., Simi, B., Wynn, T.T., Garr, K., Reddy, B.S., 1996.

Modulating effect of amount and types of dietary fat on colonic

mucosal phospholipase A2, phosphatidylinositol-specific phospho-

lipase C activities, and cyclooxygenase metabolite formation during

different stages of colon tumor promotion in male F344 rats.

Cancer Res. 56, 532�/537.

Rao, C.V., Hirose, Y., Indranie, C., Reddy, B.S., 2001. Modulation of

experimental colon tumorigenesis by types and amounts of dietary

fatty acids. Cancer Res. 61, 1927�/1933.

Reddy, B.S., 1992. Dietary fat and colon cancer: animal model studies.

Lipids 27, 807�/813.

Reddy, B.S., Sugie, S., 1988. Effect of different levels of omega-3 and

omega-6 fatty acids on azoxymethane-induced colon carcinogenesis

in F344 rats. Cancer Res. 48, 6642�/6647.

Reddy, B.S., Burill, C., Rigotty, J., 1991. Effect of diets high in omega-

3 and omega-6 fatty acids on initiation and postinitiation stages of

colon carcinogenesis. Cancer Res. 51, 487�/491.

Reddy, B.S., Simi, B., Patel, N., Aliaga, C., Rao, C.V., 1996. Effect of

amount and types of dietary fat on intestinal bacterial 7 a-

dehydroxylase and phosphatidylinositol-specific phospholipase C

and colonic mucosal diacylglycerol kinase and PKC activities

during stages of colon tumor promotion. Cancer Res. 56, 2314�/

2320.

Roncucci, L., Pedroni, M., Vaccina, F., Benatti, P., Marzona, L., De

Pol, A., 2000. Aberrant crypt foci in colorectal carcinogenesis. Cell

and crypt dynamics. Cell. Prolif. 33, 1�/18.

Rose, D.P., Connolly, J.M., 1999. Omega-3 fatty acids as cancer

chemopreventive agents. Pharmacol. Ther. 83, 217�/244.

Schonberg, S.A., Rudra, P.K., Noding, R., Skorpen, F., Bjerve, K.S.,

Krokan, H.E., 1997. Evidence that changes in Se-glutathione

peroxidase levels affect the sensitivity of human tumour cell lines

to n-3 fatty acids. Carcinogenesis 18 (10), 1897�/1904.

Singh, J., Hamid, R., Reddy, B.S., 1997a. Dietary fat and colon

cancer: modulating effect of types and amount of dietary fat on ras-

p21 function during promotion and progression stages of colon

cancer. Cancer Res. 57, 253�/258.

Singh, J., Hamid, R., Reddy, B.S., 1997b. Dietary fat and colon

cancer: modulation of cyclooxygenase-2 by types and amount of

dietary fat during the postinitiation stage of colon carcinogenesis.

Cancer Res. 57, 3465�/3470.

Singh, J., Hamid, R., Reddy, B.S., 1998. Dietary fish oil inhibits the

expression of farnesyl protein transferase and colon tumor devel-

opment in rodents. Carcinogenesis 19, 985�/989.

Takahashi, M., Fukutake, M., Isoi, T., Fukuda, K., Sato, H., Yazawa,

K., Sugimura, T., Wakabayashi, K., 1997. Suppression of azox-

ymethane-induced rat colon carcinoma development by a fish oil

component, docosahexaenoic acid (DHA). Carcinogenesis 18,

1337�/1342.

Terry, P., Bergkvist, L., Holmberg, L., Wolk, A., 2001. No association

between fat and fatty acids intake and risk of colorectal cancer.

Cancer Epidemiol. Biomarkers Prev. 10, 913�/914.

Tsai, W.S., Nagawa, H., Kaizaki, S., Tsuruo, T., Muto, T., 1998.

Inhibitory effects of n-3 polyunsaturated fatty acids on sigmoid

colon cancer transformants. J. Gastroenterol 33, 206�/212.

Visser, O., Coebergh, J., Schouten, L., Van Dijck, J., 2001. Incidence

of cancer in The Netherlands. In: (Eds.), Verening van Integrale

Kankercentra, Utrecht, 86.

Whelan, J., 1996. Antagonistic effects of dietary arachidonic acid and

n-3 polyunsaturated fatty acids. J. Nutr. 126 (Suppl. 4), 1086S�/

1091S.

Woutersen, R.A., Appel, M.J., Van Garderen-Hoetmer, A., Wijnands,

M.V., 1999. Dietary fat and carcinogenesis. Mutat. Res. 443, 111�/

127.

Zock, P.L., Katan, M.B., 1998. Linoleic acid intake and cancer risk: a

review and meta-analysis. Am. J. Clin. Nutr. 68, 142�/153.

Erratum244