Dietary resistant starch type 3 prevents tumor induction by 1,2-dimethylhydrazine and alters proliferation, apoptosis and dedifferentiation in rat colon.

Dietary resistant starch type 3 prevents tumor induction by 1,2-dimethylhydrazine

and alters proliferation, apoptosis and dedifferentiation in rat colon

Morana Bauer-Marinovic

, Simone Florian

,

Katrin Mu¨ller-Schmehl

, Hansruedi Glatt

and

Gisela Jacobasch

1

Research Group Food Chemistry and Preventive Nutrition and and 2_{Department of Nutritional Toxicology, German Institute of Human} Nutrition Potsdam-Rehbru¨cke, D-14558 Nuthetal, Germany

To whom correspondence should be addressed at: Department of Nutritional Toxicology, German Institute of Human Nutrition Potsdam-Rehbru¨cke, Arthur-Scheunert-Allee 114-116, D-14558 Nuthetal, Germany. Tel:+49 3320088 648; Fax:+49 3320088 500;

Email: florian@mail.dife.de

Some epidemiological and experimental studies suggest

that consumption of resistant starch is preventive against

colon cancer. Resistant starch leads to a

fermentation-mediated increase in the formation of short-chain fatty

acids, with a particularly high butyrate fraction in large

bowel. Butyrate is considered to be protective against colon

cancer because it causes growth arrest and apoptosis and

regulates expression of proteins involved in cellular

dedif-ferentiation in various tumor cell lines in culture. We

sought to investigate these processes under conditions of

a carcinogenicity experiment in vivo. In the present study,

1,2-dimethylhydrazine-treated Sprague–Dawley rats were

fed standard diet (n

¼ 12) or diet with 10% hydrothermally

modified Novelose 330

, a resistant starch type 3 (RS3),

replacing digestible starch (n

¼ 8). After 20 weeks tumor

number, epithelial proliferation, apoptosis,

immunoreact-ivity of carcinogenesis-related proteins [protein kinase C-d

(PKC-d), heat shock protein 25 (HSP25) and

gastrointest-inal glutathione peroxidase (GI-GPx)], as well as mucin

properties were evaluated in proximal and distal colon

in situ. No tumors developed under RS3 diet, compared

to a tumor incidence of 0.6 ± 0.6 (P

< 0.05) under the

standard diet. RS3 decreased the number of proliferating

cells, the length of the proliferation zone and the total

length of the crypt in the distal colon, but not proximal

colon, and enhanced apoptosis in both colonic segments. It

induced PKC-d and HSP25 expression, but inhibited

GI-GPx expression in the epithelium of distal colon. RS3

increased the number of predominantly acidic mucin

con-taining goblet cells in the distal colon, but had no effect on

the goblet cell count. We conclude that hydrothermally

treated RS3 prevented colon carcinogenesis, and that

this effect was mediated by enhanced apoptosis of damaged

cells accompanied by changes in parameters of

dediffer-entiation in colonic mucosa.

Introduction

The large intestine is a complex environment in which colonic

mucosa, mucus produced by the colonocytes, intestinal

micro-biota and their fermentation products, alimentary components

and bile acids interact. Based on epidemiological data it was

postulated that dietary fiber protects against the development

of colorectal neoplasia (1,2). Fermentable dietary fiber, like

starch, provides the majority of enzymatic substrate for

sac-charolytic colonic bacteria. Of particular interest is resistant

starch due to its resistance to digestion in the small intestine

and its use by cecal and colonic microbiota for anaerobic

fermentation productive of short-chain fatty acids (SCFA),

including acetate, propionate and butyrate. Butyrate is

con-sidered as an important factor in the maintenance of healthy

function in colorectal mucosa. Normal colonocytes gain 70–

80% of their energy from butyrate (3) and therefore epithelial

proliferation is supported (4). Conversely, intestinal tumor cell

lines respond to butyrate with growth arrest, differentiation

and apoptosis (5–10). In vitro studies demonstrated that

butyr-ate in non-toxic concentrations causes growth arrest in the G

/

G

phase of the cell cycle resulting from induction of p21/

Cip1, an inhibitor of cyclin D1 in the early G

phase (4).

Several rodent short-term studies have demonstrated an

increased acute apoptotic response to a genotoxic colorectal

carcinogen (6 h post-azoxymethane application) in vivo by

feeding resistant starch type 2 (11) or non-starch

polysacchar-ides such as wheat bran (12,13). This pathway may allow

removal of damaged cells prior to fixation of mutations. In

vitro experiments have shown induction of apoptosis by

butyr-ate through cytochrome c-medibutyr-ated caspase activation, and

partly by autophagic cell death (14). The former pathway is

suggested to be executed via caspase-3-dependent activation

of protein kinase C-d (PKC-d) (15). PKC-d belongs to a PKC

family of serin/threonin protein kinases, whose expression and

activity are downregulated in colorectal cancer (16,17). PKC-d

gene expression and protein levels were found to be increased

in HT-29 colon tumor cells under butyrate (18).

Heat shock protein 27 (HSP27) and its rodent homolog

HSP25 can inhibit apoptosis and may subsequently facilitate

malignant transformation (19,20). HSP25 is inducible by a

variety of stimuli in intestinal epithelial cells, e.g. oxidative

stress, cytokines, luminal bacteria and SCFA, in particular

butyrate (21–23).

Cytoprotective effects in the colonic mucosa are also

medi-ated via glutathione peroxidases, which are involved in cell’s

antioxidant system by reducing hydrogen peroxide and organic

hydroperoxides. Gastrointestinal glutathione peroxidase

(GI-GPx, GPx2), one of five selenium-dependent glutathione

per-oxidases (24), appears to be the major glutathione-dependent

peroxidase in intestinal mucosa (25). Mice lacking GI-GPx

and an additional glutathione peroxidase (cGPx, GPx1) are

highly susceptible to bacteria-associated inflammation and

cancer (26–28). GI-GPx upregulation is an early step during

Abbreviations::DMH, 1,2-dimethylhydrazine; GI-GPx, gastrointestinal glutathione peroxidase; HSP, heat shock protein; PKC, protein kinase C; RS3, resistant starch type 3; SCFA, short-chain fatty acids..

malignant transformation in the colon (29,30) and seems to be

an early neoplastic marker. Furthermore, GI-GPx expression

can be induced by luminal microbiota (31).

Mucus plays important protective and immunological roles

while providing the environment for colonic microbiota (32).

Mucins are high molecular mass glycoproteins responsible for

the physical properties of the intestinal mucus (33). Butyrate

has been shown to affect the expression of MUC genes, coding

the protein core of the mucins, in vitro (34), as well as mucin

production and composition in vivo (35–37). Mucus

composi-tion (38) and quantity (36,38,39) change due to dietary fiber

and microbiota, as well as their mutual interactions.

These data may indicate a preventive role for

butyrate-pro-ducing resistant starch in the development of colorectal

neo-plasia. Resistant starch can be classified into four types based

upon structural considerations and bacterial degradation

(40,41). Resistant starch type 3 (RS3) has shown the strongest

prebiotic properties (42). Commercially available RS3

(Nov-elose 330

) has been modified with a hydrothermal treatment

to further increase the fraction of resistant structures. This

hydrothermally treated RS3 shifts the fermentation towards

the distal colon and is suitable for use in humans (43).

Res-istant starches and other butyrogenic sources have been

invest-igated for protective effects against experimental colon

carcinogenesis in various animal studies (44). The results

depended on the type of butyrogenic material, the treatment

regime and the end point investigated. RS3 has only been used

in two short-term rodent experiments on initial stages of colon

carcinogenesis. Perrin et al. (45) found a protective effect,

whereas Maziere et al. (46) found no influence by RS3.

The aim of this study was to find out whether

hydrotherm-ally modified RS3 protects against chemichydrotherm-ally induced colon

tumorigenesis, using the actual tumors as the end point.

More-over, we have investigated whether RS3 leads to alterations in

morphologically normal colonic mucosa similar to those

observed in cells in culture after treatment with butyrate.

Materials and methods

Animals and diets

Male Sprague–Dawley rats (Charles River, MB, Denmark) (n¼ 20, weight on arrival 187 ± 6 g) were housed in stainless steel cages with splint bedding under a 12 h/12 h light/dark cycle and free access to chow and water. Animals were divided into two experimental groups. Control animals (n¼ 12) were fed a semisynthetic standard diet. Test animals were fed 10 g RS3, replacing 10 g starch/100 g chow (Table I). After a run in phase of 1 week animals received 20 mg 1,2-dimethylhydrazine (DMH) (pH 6.5) per kg body weight by s.c. injec-tion at 1-week intervals for 20 weeks.

Tissue preparation

Animals were killed by decapitation 1 week after the last DMH injection under ether anesthesia and autopsy was performed. Colon weight and length were recorded. The first 0.5 cm of the proximal colon and the final 0.5 cm of the distal colon were removed and washed in ice-cold phosphate-buffered saline. Tissue was fixed in 4% neutral buffered formalin for 24 h and washed for 24 h under tap water. Colon specimens were dehydrated and embedded in paraffin wax. Serial tissue sections (2mm) from eight animals per diet group were prepared for histology and immunohistochemical procedures. The remainder of the colon was cut open longitudinally, fixed flat in 4% formalin and stained with 0.2% methylene blue for the evaluation of tumors. Tumors were dissected and processed for histology and immunohistochemistry as described above. Immunohistochemistry and TUNEL assay

For the detection of Ki-67, active caspase-3, PKC-d, HSP25 and GI-GPx immunohistochemistry was performed on sections from the proximal and distal colon. Slides were immersed in boiling Target Retrieval Solution (ChemMateTM, DakoCytomation, Hamburg, Germany) and incubated with primary antibody for Ki-67 (MIB-5, Dianova, Hamburg, Germany), cleaved

caspase-3 (Asp175, Cell Signaling, Beverly, MA), PKC-d (C-17 sc-213, Santa Cruz Biotechnology, Santa Cruz, CA), HSP25 (Stressgen, Victoria, BC, Can-ada) and GI-GPx (47), diluted in antibody diluent (DakoCytomation), over-night at 4
_{C in a humid chamber. For active caspase-3 detection, secondary} horseradish peroxidase-conjugated anti-rabbit antibody (EnVision, DakoCyto-mation) was incubated for 35 min at room temperature. All other primary antibodies were treated with a biotin-spacer-conjugated secondary antibody IgG (Dianova) for 30 min at room temperature, followed by incubation with a streptavidin–biotin–horseradish peroxidase complex (StreptABComplex/HRP, DakoCytomation) and visualization via diaminobenzidine (DAB, DakoCyto-mation).

DNA strand brakes were detected by TUNEL assay (TACS*XL-DAB Trevigen, Inc., Gaithersburg, MD). Four micrometer tissue sections were pretreated with proteinase K (1:50 dilution) at 37
_{C for 45 min and the} assay was continued according to the manufacturer’s instructions.

Evaluation of proliferation, crypt length and apoptosis

Immunohistochemically detected Ki-67-positive cells, representing prolifera-tion, were counted in 50 crypt columns per specimen. Only crypts with an open longitudinal crypt axis were analyzed. Their lengths and that of their prolif-erative zone were measured. The crypt length was determined as a distance (mm) between the basal side of the lamina epithelialis at the bottom of the crypt and the apical side of the lamina epithelialis at the top of the crypt. The proliferative zone was defined as the distance (mm) between the furthest pos-itive Ki-67 cell in the bottom- and top-direction in the crypt.

Apoptotic cells, detected by active caspase-3 and TUNEL assay, were counted in four crypt quarters (from the lumen to the base of the crypt) of all longitudinally open crypts per tissue section (100–250) and expressed per 50 crypts.

Evaluation of PKC-d, HSP25 and GI-GPx immunoreactivity

The immunoreactivity was semiquantitatively described with score points standing for a product of occurrence and intensity of the staining. The occur-rence was assessed as percentage of positive cells among all cells of the same cell type (0¼ no positive cells; 1 ¼ 1–10%; 2 ¼ 11–50%; 3 ¼ 51–75%; 4¼ 76–100%). The intensity of the staining was evaluated in three grades between negative and intense (0¼ negative; 1 ¼ weak; 2 ¼ moderate; 3 ¼ intense). The crypts were divided into quarters from the lumen to the base of the crypt. HSP25-immunoreactive epithelial cells were counted in all longit-udinally open crypts per tissue section (100–250) and expressed per 50 crypts. Mucin histochemistry

To allow for evaluation of the mucin composition in the goblet cells, serial sections were stained by Periodic Acid-Schiff (PAS) reaction for neutral mucins (magenta), 1% alcian blue (AB), pH 2.5, for detection of acidic sialo- and sulfomucins (blue), and by high iron diamine (HID) staining to determine only sulfomucins (dark brown). PAS/AB staining was performed sequentially to differentiate neutral and acidic mucins. HID/AB staining was used to locate sialo- and sulfomucins. Dual staining allowed for determination of mucin type predominance. Goblet cells staining more than 50% of the vacuoles deep purple contained predominantly neutral mucins, whereas those staining more than 50% blue were classified as acidic mucins. The same criteria were used for HID/AB staining. Thirty well orientated, longit-udinal crypt columns per specimen were counted.

Table I. Composition of experimental diets (g/100 g dry matter)

Components Standard group RS group

Starcha 63 53 Caseinb 20 20 Sunflower oilc 5 5 Cellulosed _{5 5} RS3e 0 10 Mineral mixf _{5 5} Vitamin mixf 2 2 a

Waxy maize starch (National starch and Chemical Company, Bridgewater, USA) not containing any resistant starch. b_{Dauermilchwerk Peiting GmbH, Landshut, Germany.} c

Thomy GmbH, Karlsruhe, Germany. d_{Rettenmeier GmbH, Ellwangen, Germany.} e

Novelose 330 (National Starch & Chemical, Bridgewater, USA), hydrothermally treated.

f

Microscopy

Tumor scoring was performed using a stereomicroscope (SZH10, Olympus). All further microscopy and morphometry were done with the aid of a light microscope (Eclipse E1000, Nikon) in combination with a camera CCD-1300CB (Vosskuehler, Germany) and digital analysis software Lucia Image 4.61 (Nikon).

SCFA analysis

The SCFA were analyzed in contents of the proximal and distal colon by gas chromatography as described by Sembries et al. (48)

Statistical analysis

Data are presented in box plots, showing the distribution of the values within diet groups, by SPSS software, version 8.0, or as mean ± standard deviation (SD). Fisher’s exact test was used for the analysis of tumor incidence and Mann–Whitney U-test to determine the statistical significance of histological values.

Results

Weight progression

There was no significant influence of RS3 in the diet on the

weight progression through the entire experimental period

(Figure 1).

Tumor data

Six of eleven animals (a further animal was killed after 12

weeks after developing a cyst at the site of injection) in the

standard group developed tumors as compared to zero of eight

animals in the RS3 group (P

< 0.05) (Table II). Six tumors

were histologically classified as adenocarcinoma and one as

adenoma. They were located in the late proximal and early

distal large intestine (1 tumor after 30%, 1 after 40%, 1 after

50%, 3 after 60% and 1 after 70% of the length of the large

intestine).

Proliferation and crypt length

The proximal segment of the colon showed no significant

change in the number of proliferating cells, as assessed by

Ki-67 staining, with RS3 compared to standard diet

(Figure 2A and B). In the distal colon of the RS3 group

there were significantly fewer Ki-67-positive cells per crypt

present than in the standard group. The decreased proliferation

in the RS3 group was accompanied by a significantly

decreased proliferative zone (Figure 2C) and crypt length

(Figure 2D) compared with the standard group.

Apoptosis

The frequency and distribution of the apoptotic cells were

detected by the active caspase-3 immunohistochemistry

(Figure 3A and C) and TUNEL assay (Figure 3B and D).

Both methods showed markedly increased apoptotic activity

in both the proximal and distal colonic segments of the RS3

group compared with the standard diet group. This was the

case for active caspase-3-positive cells in the first and second

crypt quarter in the proximal colon and in the second and third

crypt quarter in the distal colon as well as for the

TUNEL-positive cells in the first crypt quarter in both colonic segments.

Unlike these findings, the RS3 group showed a decreased

number of apoptotic cells detected by the TUNEL assay in

the third crypt quarter in the distal colon, compared with the

standard diet group.

PKC-d expression pattern

PKC-d immunoreactivity was primarily detected in epithelial

cells (Figure 4). It was located in the luminal crypt quarter,

diffusely distributed in cytosol, apical of the nucleus

(Figure 4A). The immunoreactivity was slightly stronger in

the distal than in the proximal colon (Figure 4D). The PKC-d

expression strongly decreased in poorly differentiated tumor

epithelium (Figure 4C). Under the RS3 diet the

immunoreact-ivity of the epithelial PKC-d was significantly increased in the

first crypt quarter in the distal, but not in the proximal colon

(Figure 4D). In this diet group the staining was more

continu-ous and localized apical as well as basal of the nucleus

(Figure 4B). The immunoreactivity in lamina propria cells

was very weak, cytosolic and irrespective of the colonic

seg-ment, diet and neoplastic features.

HSP25 expression pattern

Cytosolic HSP25 immunoreactivity was detected in cells of all

layers of the colon wall in both colonic segments excluding the

lamina epithelialis in proximal colon. Strongest HSP25 protein

expression was found in submucosal blood vessels and in the

lamina muscularis mucosae (Figure 5A). There was a slight

increase in HSP25 expression in stroma cells, mainly in blood

vessels, and in some epithelial cells in tumor compared with

the morphologically normal mucosa. Under RS3 diet the

num-ber of HSP25-positive epithelial cells (Figure 5B) was

signi-ficantly increased in the first crypt compartment in the distal

colon compared with the standard diet (Figure 5D). In the

proximal colon RS3 diet increased the expression of HSP25

significantly in lamina propria cells (Figure 5C).

GI-GPx expression pattern

Expression and cellular distribution of GI-GPx differed

between the proximal and distal colon. The enzyme was

Table II. Incidence and multiplicity of tumors in the colon of DMH-treated rats

Diet group No. of animals No. of tumor bearing animals No. of tumors/ animal No. of tumors/tumor bearing animal Standard 12a _{6 0.6 ± 0.6 1.2 ± 0.4} RS 8 0 0 0 a

One animal was killed after 12 weeks.

P < 0.05, compared with the standard diet (Fisher’s exact test).

Fig. 1. Body weight development of DMH-treated rats fed standard (ST) and RS3 diet (RS). Values (means ± SD) between the groups were not significantly different.

most abundant in the distal colon in cytosol of luminal

epi-thelial cells (Figure 6A and B). In proliferating cells, GI-GPx

was also detected in the nuclei—in the mid crypt zone in the

proximal colon and in the crypt base in the distal colon

(Figure 6A and D). In the proximal colon an additional

faint membrane-bound staining in the crypt base was

detect-able. Tumors showed a higher cytosolic and nuclear GI-GPx

expression than morphologically normal crypts (Figure 6C).

RS3

diet

significantly

decreased

cytosolic

GI-GPx

Fig. 2.Influence of RS3 on proliferation in the proximal and distal colonic mucosa of DMH-treated rats (n¼ 8). (A) Representative photo of proliferating cells stained with Ki-67 antibody in morphologically normal colon of the standard diet group. (B) The count of Ki-67-positive cells, (C) the length of the proliferative zone and (D) crypt length in the standard (ST) and RS3 diet group (RS) (P< 0.05, compared with ST). The middle line of the box plot indicates the median for each group, the box edges mark the distribution between the 25th and 75th percentile, the whiskers represent the maximum and the minimum values without being outliers (open inverted triangles) or extreme values (closed inverted triangles).

Fig. 3.Influence of RS3 on apoptosis in the proximal and distal colonic mucosa of DMH-treated rats (n¼ 8). Representative photo of apoptotic cells detected with (A) active caspase-3 antibody and (B) TUNEL assay in morphologically normal colon of the standard diet group. The number of (C) active caspase-3- and (D) TUNEL-positive cells (crypts were divided into quarters from the lumen to the base of the crypt) in the standard (ST) and RS3 diet group (RS) (P< 0.05, P < 0.01, compared with ST). The box plots are explained in the legend to Figure 2.

immunoreactivity in the third and fourth crypt quarter of the

distal colon (Figure 6E). Nuclear immunoreactivity was

decreased as well and restricted to the third and fourth

crypt compartment in this colonic segment (Figure 6F).

Goblet cell count and composition of the mucins in the goblet

cells

RS3 diet did not affect the goblet cell count in either colonic

segment, independent of whether the values were expressed

per crypt or per 100

mm crypt length (proximal colon, standard

diet: 24.1 ± 1.9 (mean ± SD), RS3 diet: 24.4 ± 3.9; distal colon,

standard diet: 29.8 ± 3, RS3 diet: 27.8 ± 1.7 goblet cells per

crypt; proximal colon, standard diet: 17.7 ± 0.9, RS3 diet: 18.6

± 1.4; distal colon, standard diet: 13.4 ± 0.8, RS3 diet: 14.5 ±

1.2 goblet cells per 100

mm crypt length). The composition of

the mucins was not affected by the RS3 diet in the proximal

colon. However, in the distal colon significantly fewer goblet

cells containing neutral mucins in favor of those with acidic

mucins were observed in the RS3 compared with the standard

group (Figure 7, P

< 0.05). This distribution change was not

affected by a slightly different goblet cell count between the

two diet groups since this could be confirmed when the goblet

cell count was set to 100% (P

¼ 0.001). We could also observe

that in the distal colonic segment of both groups significantly

more acidic mucins containing goblet cells were detected than

in the proximal colon (Figure 7,

P

< 0.05). When

differen-tiating the acidic mucins, the crypts in the standard and RS3

group contained 59.5 ± 7.9% and 66.5 ± 6.4% (mean ± SD)

sulfomucin-positive goblet cells in the proximal colonic

seg-ment in contrast to 98.5 ± 1.2% and 99.8 ± 0.3% (mean ± SD)

in the distal colon. There were no significant differences

between the diet groups.

Luminal butyrate and total SCFA

Total SCFA and butyrate concentrations in intestinal contents

of the proximal and distal colon showed no significant

differ-ences between the diet groups. However, the values scattered

over a wide range within each dietary group and colonic

seg-ment (Figure 8).

Discussion

We found that hydrothermally treated RS3 prevents the

devel-opment of tumors in colon of rats treated with the genotoxic

colon carcinogen DMH. Perrin et al. (45) observed a decreased

number of early preneoplastic lesions, aberrant crypt foci, after

azoxymethane application by feeding a diet containing 19%

commercially available RS3 (retrograded amylose corn starch,

Cerestar) with 30% indigestible starch. In our study the

modi-fied Novelose 330 provided 10% of the diet, but contained

75% resistant structures due to hydrothermal treatment (43). In

further studies, other types of RS (49–52) and other

butyro-genic sources such as soluble non-starch polysaccharides (53–

56) were used. These substrates differ in fermentation

prop-erties and in many studies the level of resistant structures is not

clear. This complicates comparison between studies and may

explain some controversial results. While Young et al. (49)

found increased incidence, multiplicity and size of tumors with

resistant starch type 2 from potato starch in chemically

induced colon cancer in rats, Thorup et al. (51) observed a

lower number of aberrant crypt foci, using also resistant starch

Fig. 4. Influence of RS3 on PKC-d expression pattern in the proximal and distal colonic mucosa of DMH-treated rats (n ¼ 4). (A) Typical immunostaining against PKC-d in morphologically normal distal colon in the standard diet group showing diffuse cytosolic localization of the enzyme in the luminal epithelial cells. (B) Under the RS3 diet the immunoreactivity was enhanced. (C) In tumor epithelium PKC-d immunoreactivity was diminished. (D) PKC-d immunoreactivity evaluated semiquantitatively [in score points (sp)] in four crypt quarters (from the lumen to the base of the crypt) in the standard (ST) and RS3 diet group (RS) (P< 0.05, compared with ST). The box plots are explained in the legend to Figure 2.

type 2 from potato starch. Nevertheless, in the majority of the

studies some protection has been detected with resistant

starches and other soluble non-starch polysaccharides.

Generally, these anti-carcinogenic effects of functional

diet-ary fiber, including resistant starch, were attributed to a

fermentation-mediated increase in luminal SCFA, particularly

butyrate. However, it is difficult to correlate the preventive

effect to the luminal butyrate concentration as various in vivo

studies showed (55,56). It should be considered that butyrate

concentration in the colonic content is a result of its production

and absorption. It may also vary with time and these temporal

changes may be affected by the eating behavior of the

indi-vidual animals.

Butyrate is involved in the homeostatic maintenance of the

colonic mucosa. It inhibits the proliferation of various

neo-plastic colonocytes (4). Our results demonstrate that similar

effects occur in vivo under RS3 diet. They include reduced

proliferation, smaller proliferation zone and shorter crypts in

morphologically normal mucosa of the distal colon. The

growth arrest leading to decreased proliferation could enhance

DNA repair. Similar studies with other butyrogenic sources

showed variable results regarding proliferation rate. Soluble

dietary fiber, such as the fructans, reduced proliferation under

conditions of carcinogen induction and displayed preventive

effects (53). A RS3 did not affect proliferation although it

decreased the number of aberrant crypt foci (45), and a

res-istant starch type 2 caused increased proliferation correlating

with other tumor-promoting effects (49).

The effects of the chemical carcinogens DMH and

azoxy-methane upon colonic proliferation are well described. They

include an increased proliferation rate, increased size of

proliferation zone and deeper crypts (57,58). In normal colonic

epithelium, luminal butyrate would mediate an increase in

proliferation in the crypt (4), but under hyperproliferative

con-ditions this is not expected to occur. The protective effect

found in our study as a result of RS3 diet is hypothesized

to arise via counteraction of hyperproliferative effects in the

damaged colonic mucosa.

In previous studies it was shown that butyrate could induce

apoptosis in various colon carcinoma cell lines (10,15,59–61).

In our study apoptosis was significantly increased in both

colonic segments of the RS3 group, contributing to

anti-tumorigenic effects. Few in vivo studies on tumorigenesis

with intervention by RS3 or other butyrogenic diets have rarely

investigated the mucosal apoptotic response. Femia et al. (53)

did not observe a significant increase in apoptosis after feeding

of azoxymethane-treated rats with fructans for 31 weeks. On

the contrary, increased apoptosis was detected with diets

enriched in fructans only 24 h after DMH treatment (62).

Along with proliferation, the treatment of rodents with a

geno-toxic carcinogen influences apoptotic cell death in the colonic

crypts. Cells in the lower half of the crypt are most susceptible

to acute apoptotic response 6–8 h after genotoxic treatment

(13) as a physiological removal of damaged cells prior to

mutation accumulation. In this context, studies with

butyrate-producing fiber including resistant starch type 2 (11) and

non-starch polysaccharides as wheat bran (12,13) demonstrated an

increased acute apoptotic response after a genotoxic treatment.

Although we did not study the acute response phase as tissue

was harvested 7 days following the final DMH injection,

tosis was still significantly increased. The majority of

apop-totic cells were localized in the first and second crypt quarter

Fig. 5.Influence of RS3 on HSP25 expression pattern in the proximal and distal colonic mucosa of DMH-treated rats (n¼ 4). (A) Typical immunostaining against HSP25 in morphologically normal distal colon in the standard diet group showing immunoreactivity in cells of the lamina propria and lamina muscularis mucosae. (B) Discrete HSP25-positive epithelial cells within luminal mucosa under RS3. (C) HSP25 immunoreactivity in the lamina propria cells [evaluated semiquantitatively in score points sp)] and (D) the number of HSP25-positive epithelial cells (crypts were divided in quarters from the lumen to the base of the crypt) in the standard (ST) and RS3 diet group (RS) (P< 0.05, compared with ST). The box plots are explained in the legend to Figure 2.

from the luminal side. Since this is the place where apoptosis

normally takes place, the increase could implicate that RS3

ensures the reestablishment of homeostasis of the colonic

mucosa by removal of damaged cells.

Highly associated with growth suppression (63) and

apop-tosis induction (64) is PKC-d. We could show a significantly

higher PKC-d expression with the modified RS3 in

morpho-logically normal distal colonic mucosa. On the contrary,

with-out RS3, in tumor epithelium from the standard group

markedly lower levels of PKC-d were observed than in

mor-phologically normal colonic mucosa. PKC-d is known to be

downregulated in human colorectal tumors (16,17). These data

emphasize the beneficial effects of the RS3 diet and implicate

that RS3 in vivo exerts effects comparable to butyrate in vitro.

The PKC-d immunoreactivity was markedly stronger in the

first quarter of the crypts. This is the crypt compartment where

apoptosis was detected by TUNEL assay. In vitro PKC-d is

involved in mitochondria-mediated apoptosis induction. It is

cleaved to an active catalytic fragment by the caspase-3

fol-lowed by a further activation of caspase-3 (64). Interestingly,

although PKC-d is proposed to be activated by caspase-3 under

butyrate in human adenoma cells (15), we found almost no

co-localization of the PKC-d and active caspase-3. The

localiza-tion along the crypt clearly differed between these enzymes.

The increase in the active caspase-3 in the RS3 group was

highest in the second and third crypt quarter, while PKC-d was

located in the first crypt quarter in the distal colon.

Further-more, we observed active caspase-3 expression in a small

number of individual cells, whereas increased PKC-d

immun-oreactivity was continuously present in all crypts. Therefore,

active caspase-3 and PKC-d may not be involved in the same

apoptotic pathway in vivo, except possibly in the first

com-partment.

Butyrate can induce HSP25/27 expression in vitro and in

vivo (23,65,66). HSP expression can modulate the fate of cells

in response to stress or a death stimulus. RS3 diet resulted in

Fig. 6. Influence of RS3 on GI-GPx expression pattern in the proximal and distal colonic mucosa of DMH-treated rats (n¼ 4). (A) Typical immunostaining against GI-GPx in morphologically normal distal colon in the standard diet group showing (B) a cytosolic localization of the enzyme in luminal epithelial cells and (D) a predominantly nuclear staining in cells of the crypt base. (C) In tumor epithelium cytosolic and nuclear GI-GPx immunoreactivity was increased. (E) Cytosolic and (F) nuclear GI-GPx immunoreactivity evaluated semiquantitatively [in score points (sp)] in four crypt quarters (from the lumen to the base of the crypt) in the standard (ST) and the RS3 diet group (RS) (P< 0.05, compared with ST). The box plots are explained in the legend to Figure 2.

upregulation of HSP25 in single epithelial cells exclusively in

the first crypt compartment in distal colon. The number and

localization of HSP25-positive epithelial cells we detected

after feeding RS3 do not support the hypothesis that

butyrate-producing fiber improve the general defense against oxidative

injury via HSP25 in the luminal epithelium: pectin-enriched

diet, without carcinogen treatment, resulted in a continuous

upregulation of HSP25 in all luminal epithelial cells (23). Our

results suggest it is more likely that upregulated HSP25 in the

epithelium is involved in single cellular, e.g. apoptotic

pro-cesses, rather than in a general barrier function. Other studies

showed that HSP27 interacts with released cytochrome c in the

cytosol, preventing formation of the apoptosome and

sub-sequently inhibiting cleavage of procaspase-9 (67,68).

Inter-estingly, HSP25 upregulation was restricted to discrete cells in

the first crypt compartment whereas active caspase-3 was

mainly confined to single cells in the second and third crypt

compartment in distal colon, making possible that HSP25

indeed could inhibit activation of caspase-3 in vivo. We

also detected an upregulation of HSP25 in lamina propria

cells reflecting that RS3 not only affects gene expression of

epithelial cells, but also of connective tissue cells. However,

HSP25 upregulation might have different effects in epithelial

and lamina propria cells, since lamina propria cells already

showed a strong HSP25 expression. Whether the induction of

HSP25 is mediated only by a higher butyrate level or also by

changes in composition of colonic microbiota remains to be

further investigated.

GI-GPx expression, suggested to play a role in mucosa

homeostasis and to protect from ileocolitis and cancer (26),

was also influenced under RS3 diet in distal colon. This is the

first report demonstrating changes in the expression level of

GI-GPx in situ after dietary intervention. We and others could

detect an upregulation of GI-GPx already in very early stages

of malignant transformation in colorectal tumors in humans

(29,30) and also in the rat tumor model used. Even though

subcellular localization of the enzyme slightly differed in man

and rat, nuclear and diffuse cytosolic fraction show similar

distribution pattern in morphologically normal distal mucosa

and neoplastic tissue. It was supposed that increased GI-GPx

may reflect increased cell proliferation (27). Interestingly, RS3

decreased the cytosolic GI-GPx fraction in the third and fourth

crypt quarter and restricted the nuclear fraction to these

compartments in morphologically normal appearing mucosa

in distal colon. This probably reflects the decrease in

prolif-eration zone and anti-hyperplastic effects, since predominant

nuclear GI-GPx is known to be confined to the proliferating

cells at the crypt base. It still remains to be elucidated whether

this is a direct effect on GI-GPx expression due to RS3 diet or

whether the downregulation is secondary due to

anti-hyper-proliferative effects. Secondary effects are more likely since

cytosolic GI-GPx expression in the luminal crypt compartment

was not significantly affected by RS3.

There are various physiological differences between the

proximal and distal rat colon that must be considered. An

important observation in our study is that PKC-d, HSP25

and GI-GPx have a generally higher expression in epithelium

of the distal colon compared with the proximal colon.

More-over, RS3 influence on the epithelial expression of PKC-d,

HSP25 and GI-GPx was confined to the distal colon. The

position of the stem cells also differs between segments,

which leads to localization of the proliferating cells in the

proximal colon to the mid-crypt region and in the distal

colon to the bottom of the crypt (69). Furthermore, the distal

colon has a higher proliferation rate than the proximal colon,

as also showed in the present study. The application of DMH

increases the proliferation zone in both colonic segments (57),

but RS3 diet significantly inhibited proliferation only in the

distal colon. The distal rat colon is considered to be more

comparable with human colon than the proximal (58).

Although the observations suggest that proliferation has less

predictive value for colon tumorigenesis compared with

apop-tosis (58), a decrease in proliferation may still be linked to the

preventive effect of RS3 found. In contrast, enhanced

apop-tosis was observed in both colonic compartments in the RS3

Fig. 7.Influence of RS3 on the distribution of neutral and acidic mucin containing goblet cells (PAS/AB staining) along the crypt column in the proximal and distal colonic mucosa of DMH-treated rats (n¼ 8) in the standard (ST) and RS3 diet group (RS) (mean ± SD, P< 0.05, compared with ST in the distal colon;#P< 0.05, compared with the proximal colon).

Fig. 8.Concentrations of (A) short-chain fatty acids and (B) butyrate in the luminal content of the proximal and distal colon of DMH-treated rats (n¼ 8) in the standard (ST) and RS3 diet group (RS). The box plots are explained in the legend to Figure 2.

group. Therefore, we suspect that enhanced apoptosis was a

more sensitive process in tumor prevention.

Colonic microbial composition and activity is known to be

regio-specific as well. The immediate environment of the

intestinal bacteria is the colonic mucus layer as a part of

the extracellular barrier, which is produced by the goblet

cells of the crypts. The MUC protein composition, and even

more the oligosaccharide side chains vary between intestinal

regions showing dependence on bacterial colonization (70). As

in previous studies, we observed an increase in acidic and

sulfated mucins from the proximal to the distal colon

regard-less of diet. In our study the quantity of mucin producing

goblet cells was not affected by diet in any colonic segment.

However, we found a significant increase in acidic mucins and

a reduction of neutral mucins with RS3 compared with

stand-ard diet in the distal colon. This might be beneficial, since

acidic mucins form a better barrier against bacterial

translo-cation than neutral mucins (39). We assume that the increase in

the acidic mucins to be partly a consequence of the prebiotic

effect of the RS3 by changing the microbial population in the

colon in favor of butyrate-producing species. RS3 stimulates

the growth of bifidobacteria in non-associated (71) and human

microbiota-associated rats (72), and thus may disrupt the

growth of pathogenic bacteria and therefore contribute to

the homeostasis of the microbiota in the colonic lumen.

Par-allel to microbial-derived influence of RS3 on the mucin

com-position in the colon, it also remains possible for butyrate to

directly affect the expression of MUC and/or

glycosylation-related genes. An upregulation of MUC2, MUC3 and MUC5B

gene expression by butyrate has been demonstrated in vitro in

a colonic goblet cell line (34). In particular, an upregulation of

MUC2 would be considered beneficial since it has been

involved in suppression of colorectal cancer (73).

In summary, we demonstrated that the hydrothermally

trea-ted RS3 is preventive of tumor development in vivo. Moreover,

we observed enhanced apoptosis and decreased cell

prolifera-tion under the RS3 diet. Therefore, we suspect that an

enhanced removal of damaged cells and an increased repair

efficiency owing to lower proliferation may be involved in the

tumor prevention by RS3. This coincided with diminished

GI-GPx expression. Apparently, these effects can be ascribed to

the used RS3 preparation providing a stable butyrate supply for

the colonic mucosa, since we could also show further known

butyrate-mediated effects like upregulation of PKC-d, HSP25

and changes in mucus composition in situ.

Acknowledgements

We thank Prof. Regina Brigelius-Flohe´ for providing the GI-GPx antiserum, for helpful discussion and critical comments on the manuscript. We appreciate the skillful technical assistance of Elisabeth Meyer, Elke Thom and Ines Bebert. The authors also thank Colleen McClean for reading the manuscript. Conflict of Interest statement: None declared.

References

1. Martinez,M.E. (2005) Primary prevention of colorectal cancer: lifestyle, nutrition, exercise. Recent Results Cancer Res., 166, 177–211.

2. Cummings,J.H. and Bingham,S.A (1987) Dietary fibre, fermentation and large bowel cancer. Cancer Surv., 6, 601–621.

3. Roediger,W.E (1980) Role of anaerobic bacteria in the metabolic welfare of the colonic mucosa in man. Gut, 21, 793–798.

4. Blottiere,H.M., Buecher,B., Galmiche,J.P. and Cherut,C (2003) Molecular analysis of the effect of short-chain fatty acids on intestinal cell proliferation. Proc. Nutr. Soc., 62, 101–106.

5. Scheppach,W., Bartram,H.P. and Richter,F (1995) Role of short-chain fatty acids in the prevention of colorectal cancer. Eur. J. Cancer, 31A, 1077–1080.

6. Archer,S.Y., Meng,S., Shei,A. and Hodin,R.A (1998) p21(WAF1) is required for butyrate-mediated growth inhibition of human colon cancer cells. Proc. Natl Acad. Sci. USA, 95, 6791–6796.

7. Siavoshian,S., Segain,J.P., Kornprobst,M., Bonnet,C., Cherbut,C., Galmiche,J.P. and Blottiere,H.M. (2000) Butyrate and trichostatin A effects on the proliferation/differentiation of human intestinal epithelial cells: induction of cyclin D3 and p21 expression. Gut, 46, 507–514.

8. McBain,J.A., Eastman,A., Nobel,C.S. and Mueller,G.C. (1997) Apoptotic death in adenocarcinoma cell lines induced by butyrate and other histone deacetylase inhibitors. Biochem. Pharmacol, 53, 1357–1368.

9. Coradini,D., Pellizzaro,C., Marimpietri,D., Abolafio,G. and Daidone,M.G. (2000) Sodium butyrate modulates cell cycle-related proteins in HT29 human colonic adenocarcinoma cells. Cell Prolif, 33, 139–146.

10. Mariadason,J.M., Rickard,K.L., Barkla,D.H., Augenlicht,L.H. and Gison,P.R. (2000) Divergent phenotypic patterns and commitment to apoptosis of Caco-2 cells during spontaneous and butyrate-induced differentiation. J. Cell Physiol, 183, 347–354.

11. Le Leu,R.K., Brown,I.L., Hu,Y. and Young,G.P. (2003) Effect of resistant starch on genotoxin-induced apoptosis, colonic epithelium, and lumenal contents in rats. Carcinogenesis, 24, 1347–1352.

12. Le Leu,R.K., Hu,Y. and Young,G.P. (2002) Effects of resistant starch and nonstarch polysaccharides on colonic luminal environment and genotoxin-induced apoptosis in the rat. Carcinogenesis, 23, 713–719.

13. Hu,Y., Martin,J., Le Leu,R. and Young,G.P. (2002) The colonic response to genotoxic carcinogens in the rat: regulation by dietary fibre. Carcinogenesis, 23, 1131–1137.

14. Shao,Y., Gao,Z., Marks,P.A. and Jiang,X. (2004) Apoptotic and autophagic cell death induced by histone deacetylase inhibitors. Proc. Natl Acad. Sci. USA, 101, 18030–18035.

15. McMillan,L., Butcher,S.K., Pongracz,J. and Lord,J.M. (2003) Opposing effects of butyrate and bile acids on apoptosis of human colon adenoma cells: differential activation of PKC and MAP kinases. Br. J. Cancer, 88, 748–753.

16. Craven,P.A. and DeRuertis,F.R. (1994) Loss of protein kinase C delta isozyme immunoreactivity in human adenocarcinomas. Dig. Dis. Sci, 39, 481–489.

17. Kahl-Rainer,P., Sedivy,R. and Marian,B. (1996) Protein kinase C tissue localization in human colonic tumors suggests a role for adenoma growth control. Gastroenterology, 110, 1753–1759.

18. Della Ragione,F., Criniti,V., Della Pietra,V., Borriello,A., Oliva,A., Indaco,S., Yamamoto,T. and Zappia,V. (2001) Genes modulated by histone acetylation as new effectors of butyrate activity. FEBS Lett, 499, 199–204.

19. Garrido,C., Gurbuxani,S., Ravagnan,L. and Kroemer,G. (2001) Heat shock proteins: endogenous modulators of apoptotic cell death. Biochem. Biophys. Res. Commun, 286, 433–442.

20. Lee,Y.J., Lee,D.H., Cho,C.K., Chung,H.Y., Bae,S., Jhon,G.J., Soh,J.W., Jeoung,D.I., Lee,S.J. and Lee,Y.S. (2005) HSP25 inhibits radiation-induced apoptosis through reduction of PKCdelta-mediated ROS production. Oncogene, 24, 3715–3725.

21. Kojima,K., Musch,M.W., Ren,H., Boone,D.L., Hendrickson,B.A., Ma,A. and Chang,E.B. (2003) Enteric flora and lymphocyte-derived cytokines determine expression of heat shock proteins in mouse colonic epithelial cells. Gastroenterology, 124, 1395–1407.

22. Arvans,D.L., Vavricka,S.R., Ren,H., Musch,M.W., Kang,L., Rocha,F.G., Lucioni,A., Turner,J.R., Alverdy,J. and Chang,E.B. (2005) Luminal bacterial flora determines physiological expression of intestinal epithelial cytoprotective heat shock proteins 25 and 72. Am. J. Physiol, 288, G696–G704.

23. Ren,H., Musch,M.W., Kojima,K., Boone,D., Ma,A. and Chang,E.B. (2001) Short-chain fatty acids induce intestinal epithelial heat shock protein 25 expression in rats and IEC 18 cells. Gastroenterology, 121, 631–639.

24. Kryukov,G.V., Castellano,S., Novoselov,S.V., Lobanov,A.V., Zehtab,O., Guigo,R. and Gladyshev,V.N. (2003) Characterization of mammalian selenoproteomes. Science, 300, 1439–1443.

25. Esworthy,R.S., Swiderek,K.M., Ho,Y.S. and Chu,F.F. (1998) Selenium-dependent glutathione peroxidase-GI is a major glutathione peroxidase

activity in the mucosal epithelium of rodent intestine. Biochim. Biophys. Acta, 1381, 213–226.

26. Chu,F.F., Esworthy,R.S., Chu,P.G., Longmate,J.A., Huycke,M.M., Wilczynski,S. and Doroshow,J.H. (2004) Bacteria-induced intestinal cancer in mice with disrupted Gpx1 and Gpx2 genes. Cancer Res, 64, 962–968.

27. Chu,F.F., Esworthy,R.S. and Doroshow,J.H. (2004) Role of Se-dependent glutathione peroxidases in gastrointestinal inflammation and cancer. Free Radic. Biol. Med, 36, 1481–1495.

28. Esworthy,R.S., Yang,L., Frankel,P.H. and Chu,F.F. (2005) Epithelium-specific glutathione peroxidase, Gpx2, is involved in the prevention of intestinal inflammation in selenium-deficient mice. J. Nutr, 135, 740–745.

29. Mork,H., al-Taie,O.H., Bahr,K., Zierer,A., Beck,C., Scheurlen,M., Jakob,F. and Ko¨hrle,J. (2000) Inverse mRNA expression of the selenocysteine-containing proteins GI-GPx and SeP in colorectal adenomas compared with adjacent normal mucosa. Nutr. Cancer, 37, 108–116.

30. Florian,S., Wingler,K., Schmehl,K., Jacobasch,G., Kreuzer,O.J., Meyerhof,W. and Brigelius-Flohe´,R. (2001) Cellular and subcellular localization of gastrointestinal glutathione peroxidase in normal and malignant human intestinal tissue. Free Radic. Res, 35, 655–663. 31. Esworthy,R.S., Binder,S.W., Doroshow,J.H. and Chu,F.F. (2003)

Microflora trigger colitis in mice deficient in selenium-dependent glutathione peroxidase and induce Gpx2 gene expression. Biol. Chem, 384, 597–607.

32. Forstner,J.F. (1978) Intestinal mucins in health and disease. Digestion, 17, 234–263.

33. Creeth,J.M. (1978) Constituents of mucus and their separation. Br. Med. Bull, 34, 17–24.

34. Gaudier,E., Jarry,A., Blottiere,H.M., de Coppet,P., Buisine,M.P., Aubert,J.P., Laboisse,C., Cherbut,C. and Hoeler,C. (2004) Butyrate specifically modulates MUC gene expression in intestinal epithelial goblet cells deprived of glucose. Am. J. Physiol., 287, G1168–G1174. 35. Meslin,J.C., Bensaada,M., Popot,F. and Andrieux,C. (2001) Differential

influence of butyrate concentration on proximal and distal colonic mucosa in rats born germ-free and associated with a strain of Clostridium paraputrificum. Comp. Biochem. Physiol. A. Mol. Integr. Physiol., 128, 379–384.

36. Kleessen,B., Hartmann,L. and Blaut,M. (2003) Fructans in the diet cause alterations of intestinal mucosal architecture, released mucins and mucosa-associated bifidobacteria in gnotobiotic rats. Br. J. Nutr., 89, 597–606. 37. Fontaine,N., Meslin,J.C., Lory,S. and Andrieux,C. (1996) Intestinal mucin

distribution in the germ-free rat and in the heteroxenic rat harbouring a human bacterial flora: effect of inulin in the diet. Br. J. Nutr, 75, 881–892. 38. Sharma,R., Schumacher,U., Ronaasen,V. and Coates,M. (1995) Rat intestinal mucosal responses to a microbial flora and different diets. Gut, 36, 209–214.

39. Deplancke,B. and Gaskins,H.R. (2001) Microbial modulation of innate defense: goblet cells and the intestinal mucus layer. Am. J. Clin. Nutr, 73, 1131S–1141S.

40. Englyst,H.N., Kingman,S.M. and Cummings,J.H. (1992) Classification and measurement of nutritionally important starch fractions. Eur. J. Clin. Nutr, 46, S33–S50.

41. Brown,I.L., McNaught,K.J. and Moloney,E. (1995) Hi-maize. TM.: new directions in starch technology and nutrition. Food Aust., 47, 272–275. 42. Asp,N.-G., van Amelsvoort,J.M.M. and Hautvast,J.G.A.J. (1996)

Nutritional implications of resistant starch. Nutr. Res. Rev., 9, 1–31. 43. Jacobasch,G., Dongowski,G., Schmiedl,D. and Mu¨ller-Schmehl,K. (2006)

Hydrothermal treatment of Novelose 330 results in high yield of resistant starch type 3 with beneficial prebiotic properties and decreased secondary bile acid formation in rats.. Br. J. Nutr., in press.

44. Topping,D.L. and Clifton,P.M. (2001) Short-chain fatty acids and human colonic function: roles of resistant starch and nonstarch polysaccharides. Physiol. Rev, 81, 1031–1064.

45. Perrin,P., Pierre,F., Patry,Y., Champ,M., Berreur,M., Pradal,G., Bornet,F., Meflah,K. and Menanteau,J. (2001) Only fibres promoting a stable butyrate producing colonic ecosystem decrease the rate of aberrant crypt foci in rats. Gut, 48, 53–61.

46. Maziere,S., Meflah,K., Tavan,E., Champ,M., Narbonne,J.F. and Cassand,P. (1998) Effect of resistant starch and/or fat-soluble vitamins A and E on the initiation stage of aberrant crypts in rat colon. Nutr. Cancer, 31, 168–177. 47. Bocher,M., Boldicke,T., Kiess,M. and Bilitewski,U. (1997) Synthesis of mono- and bifunctional peptide–dextran conjugates for the immobilization of peptide antigens on ELISA plates: properties and application. J. Immunol. Methods, 208, 191–202.

48. Sembries,S., Dongowski,G., Jacobasch,G., Mehrlander,K., Will,F. and Dietrich,H. (2003) Effects of dietary fibre-rich juice colloids from apple pomace extraction juices on intestinal fermentation products and microbiota in rats. Br. J. Nutr., 90, 607–615.

49. Young,G.P., McIntyre,A., Albert,V., Folino,M., Muir,J.G. and Gison,P.R. (1996) Wheat bran suppresses potato starch—potentiated colorectal tumorigenesis at the aberrant crypt stage in a rat model. Gastroenterology, 110, 508–514.

50. Sakamoto,J., Nakaji,S., Sugawara,K., Iwane,S. and Munakata,A. (1996) Comparison of resistant starch with cellulose diet on 1,2-dimethylhydrazine-induced colonic carcinogenesis in rats. Gastro-enterology, 110, 116–120.

51. Thorup,I., Meyer,O. and Kristiansen,E. (1995) Effect of potato starch, corn starch and sucrose on aberrant crypt foci in rats exposed to azoxymethane. Anticancer Res, 15, 2101–2105.

52. Cassand,P., Maziere,S., Champ,M., Meflah,K., Bornet,F. and Naronne,J.F. (1997) Effects of resistant starch- and vitamin A-supplemented diets on the promotion of precursor lesions of colon cancer in rats. Nutr. Cancer, 27, 53–59.

53. Femia,A.P., Luceri,C., Dolara,P., Giannini,A., Biggeri,A., Salvadori,M., Clune,Y., Collins,K.J., Paglierani,M. and Caderni,G. (2002) Antitumorigenic activity of the prebiotic inulin enriched with oligofructose in combination with the probiotics Lactobacillus rhamnosus and Bifidobacterium lactis on azoxymethane-induced colon carcinogenesis in rats. Carcinogenesis, 23, 1953–1960.

54. Wijnands,M.V.W., Schoterman,H.C., Bruijntjes,J.P., Hollanders,V.M.H. and Woutersen,R.A. (2001) Effect of dietary galacto-oligosaccharides on azoxymethane-induced aberrant crypt foci and colorectal cancer in Fischer 344 rats. Carcinogenesis, 22, 127–132.

55. Zoran,D.L., Turner,N.D., Taddeo,S.S., Chapkin,R.S. and Lupton,J.R. (1997) Wheat bran diet reduces tumor incidence in a rat model of colon cancer independent of effects on distal luminal butyrate concentrations. J. Nutr, 127, 2217–2225.

56. McIntyre,A., Gibson,P.R. and Young,G.P. (1993) Butyrate production from dietary fibre and protection against large bowel cancer in a rat model. Gut, 34, 386–391.

57. Ma,Q.Y., Williamson,K.E. and Rowlands,B.J. (2002) Variability of cell proliferation in the proximal and distal colon of normal rats and rats with dimethylhydrazine induced carcinogenesis. World J. Gastroenterol., 8, 847–852.

58. Chang,W.C., Chapkin,R.S. and Lupton,J.R. (1997) Predictive value of proliferation, differentiation and apoptosis as intermediate markers for colon tumorigenesis. Carcinogenesis, 18, 721–730.

59. Buda,A., Qualtrough,D., Jepson,M.A., Martines,D., Paraskeva,C. and Pignatelli,M. (2003) Butyrate downregulates alpha2beta1 integrin: a possible role in the induction of apoptosis in colorectal cancer cell lines. Gut, 52, 729–734.

60. Medina,V., Edmonds,B., Young,G.P., James,R., Appleton,S. and Zalewski,P.D. (1997) Induction of caspase-3 protease activity and apoptosis by butyrate and trichostatin A (inhibitors of histone deacetylase): dependence on protein synthesis and synergy with a mitochondrial/cytochrome c-dependent pathway. Cancer Res, 57, 3697–3707.

61. Avivi-Green,C., Polak-Charcon,S., Madar,Z. and Schwartz,B. (2002) Different molecular events account for butyrate-induced apoptosis in two human colon cancer cell lines. J. Nutr, 132, 1812–1818.

62. Hughes,R. and Rowland,I.R. (2001) Stimulation of apoptosis by two prebiotic chicory fructans in the rat colon. Carcinogenesis, 22, 43–47. 63. Saxon,M.L., Zhao,X. and Black,J.D. (1994) Activation of protein kinase C

isozymes is associated with post-mitotic events in intestinal epithelial cells in situ. J. Cell Biol., 126, 747–763.

64. DeVries,T.A., Neville,M.C. and Reyland,M.E. (2002) Nuclear import of PKCdelta is required for apoptosis: identification of a novel nuclear import sequence. EMBO J., 21, 6050–6060.

65. Tan,S., Seow,T.K., Liang,R.C., Koh,S., Lee,C.P., Chung,M.C. and Hooi,S.C. (2002) Proteome analysis of butyrate-treated human colon cancer cells (HT-29). Int. J. Cancer, 98, 523–531.

66. Williams,E.A., Coxhead,J.M. and Mathers,J.C. (2003) Anti-cancer effects of butyrate: use of micro-array technology to investigate mechanisms. Proc. Nutr. Soc., 62, 107–115.

67. Garrido,C., Bruey,J.M., Fromentin,A., Hammann,A., Arrigo,A.P. and Solary,E. (1999) HSP27 inhibits cytochrome c-dependent activation of procaspase-9. FASEB J., 13, 2061–2070.

68. Bruey,J.M., Ducasse,C., Bonniaud,P. et al. (2000) Hsp27 negatively regulates cell death by interacting with cytochrome c. Nat. Cell Biol., 2, 645–652.

69. Sato,M. and Ahnen,D.J. (1992) Regional variability of colonocyte growth and differentiation in the rat. Anat. Rec., 233, 409–414.

70. Robbe,C., Capon,C., Coddeville,B. and Michalski,J.C. (2004) Structural diversity and specific distribution of O-glycans in normal human mucins along the intestinal tract. Biochem. J., 384, 307–316.

71. Kleessen,B., Stoof,G., Proll,J., Schmiedl,D., Noack,J. and Blaut,M. (1997) Feeding resistant starch affects fecal and cecal microflora and short-chain fatty acids in rats. J. Anim. Sci., 75, 2453–2462.

72. Silvi,S., Rumney,C.J., Cresci,A. and Rowland,I.R. (1999) Resistant starch modifies gut microflora and microbial metabolism in human

flora-associated rats inoculated with faeces from Italian and UK donors. J. Appl. Microbiol., 86, 521–530.

73. Velcich,A., Yang,W., Heyer,J., Fragale,A., Nicholas,C., Viani,S., Kucherlapati,R., Lipkin,M., Yang,K. and Augenlicht,L. (2002) Colorectal cancer in mice genetically deficient in the mucin Muc2. Science, 295, 1726–1729.

Received October 13, 2005; revised February 27, 2006; accepted March 27, 2006