Carcinogenesis vol.29 no.1 pp.113119, 2008 doi:10.

1093/carcin/bgm204 Advance Access publication September 24, 2007

Green tea polyphenols inhibit colorectal aberrant crypt foci (ACF) formation and prevent oncogenic changes in dysplastic ACF in azoxymethane-treated F344 rats
Hang Xiaoy, Xingpei Haoy, Barbara Simi, Jihyeung Ju, Heyuan Jiang, Bandaru S.Reddy and Chung S.Yang
Susan Lehman Cullman Laboratory for Cancer Research, Department of Chemical Biology, Ernest Mario School of Pharmacy, Rutgers, The State University of New Jersey, Piscataway, NJ 08854-8020, USA
To whom correspondence should be addressed. Tel: 1 732 445 3400 ext. 244; Fax: 1 732 445 0687; Email: csyang@rci.rutgers.edu

Green tea and its constituents have shown cancer-preventive activities in many animal models. In order to prepare for a human trial on the inhibition of colon carcinogenesis, we conducted a study with green tea polyphenols as the preventive agent in an azoxymethane (AOM)-induced rat colon cancer model using aberrant crypt foci (ACF) as an end point. F344 rats were given two weekly injections of AOM (15 mg/kg), and then fed a 20% highfat diet with or without 0.12 or 0.24% Polyphenon E (PPE, a standardized green tea preparation consisting 65% of (2)-epigallocatechin-3-gallate and 22% of other catechins) for 8 weeks. Colorectal ACF were analyzed under a microscope after methylene blue staining. Dietary PPE administration was found to signicantly and dose dependently decrease the total number of ACF per rat and the total number of aberrant crypt per rat. Moreover, treatment with 0.24% PPE also signicantly decreased the percentage of large ACF (four or more crypts) and the percentage of ACF with high-grade dysplasia in total ACF. The high-grade dysplastic ACF from 0.24% PPE-treated group had increased apoptosis and decreased nuclear expression levels of b-catenin and cyclin D1. Retinoid X receptor (RXR)a expression was reduced in high-grade dysplastic ACF, adenoma and adenocarcinoma during AOM-induced colon carcinogenesis, and the PPE treatment partially prevented the loss of RXRa expression in high-grade dysplastic ACF. Taken together, our results strongly suggest the colon cancer-preventive activity of PPE and identied possible molecular markers for future colon cancer prevention studies.

Introduction Tea, made from dried leaves of the plant Camellia sinensis, is the second most popular beverage worldwide next to water. In many different animal models, tea and its constituents have been demonstrated to inhibit tumorigenesis in different organ sites including the lung, oral cavity, esophagus, stomach, small intestine, colon, skin, prostate, mammary glands, liver, pancreas and bladder (15). Tea polyphenols, which account for about a third of the water extractable materials in dry weight, are the most abundant bioactive components in tea. The intestine may be a promising site for cancer chemoprevention with tea polyphenols because the intestine is exposed to high levels of tea polyphenols after ingestion (6,7). The inhibitory effects of tea and tea polyphenols on intestinal tumorigenesis in mice have been consistently observed in different laboratories. Yin et al. (8) showed that tea catechins inhibited 1,2dimethylhydrazine-induced colon carcinogenesis in mice. Suganuma et al. (9) and Orner et al. (10) both demonstrated an inhibition of tumorigenesis in the ApcMin/ mouse model by orally administered green tea and a greater inhibition by its combination with sulindac.
Abbreviations: AC, aberrant crypt; ACF, aberrant crypt foci; AOM, azoxymethane; CRC, colorectal cancer; EGCG, ()-epigallocatechin-3-gallate; IHC, immunohistochemistry; PPE, Polyphenon E; RXR, retinoid X receptor.
y

These authors contributed equally to this work.

Recently, we systematically investigated the effect of two key constituents of green tea, ()-epigallocatechin-3-gallate (EGCG) and caffeine, on intestinal tumorigenesis in ApcMin/ mice and found that intestinal tumor formation was inhibited in a dose-dependent manner by oral administration of EGCG, but not by caffeine (11). Our previous study also showed that oral administration of green tea inhibited the formation of azoxymethane (AOM)-induced aberrant crypt foci (ACF) in female CF-1 mice on a high-fat diet (12). The inhibitory effects of tea preparation on colon tumorigenesis in rats have also been demonstrated. Colon cancer formation induced by AOM in rats was inhibited by green tea polyphenols in the drinking uid (13). Green tea extract (2%) in drinking uid also signicantly inhibited ACF formation in 2-amino-3-methylimidazo[4,5-f]quinoline-treated F344 rats (14). In another study in AOM-treated F344 rats, however, green tea extract (1 or 2% in drinking uid) only had a marginal inhibitory effect on ACF formation (15). In a similar model, when different tea preparations were given to F344 rats after AOM treatment, EGCG was found to inhibit ACF formation, but the effect of green tea polyphenols was marginal (16). There are also studies that did not observe an inhibitory effect of tea polyphenol preparations in 1,2-dimethylhydrazine- or AOM-induced rat colon tumorigenesis models (1720). The reasons for the inconsistency among the different studies in rats may be related to several factors: (i) the diet used, (ii) the protocol of tumor initiation and (iii) the type and dose of tea or polyphenol preparations used, when they were given, and whether they were administered through drinking uid or the diet. Therefore, additional studies with well-dened tea polyphenol preparations are needed. The relationship between tea consumption and colorectal cancer (CRC) risk in human has been studied by many investigators, but no clear-cut conclusions can be drawn because of confounding factors (2,2124). Sun et al. (25) recently conducted a meta-analysis on 21 relevant papers and indicated a reduced risk of CRC risk with intake of green tea (summary odds ratio 0.82, 95% condence interval 5 0.690.98). A prospective study on women in Shanghai indicated a negative association between tea consumption and the risk of CRC (26). We participated in a nested casecontrol study within the Shanghai Cohort Study, involving 18 244 men, to investigate the association between prediagnostic urinary tea polyphenols and CRC risk (27). After 16 years of follow-up, the results showed that high prediagnostic urinary levels of ()-epigallocatechin and 4#-methyl ()-epigallocatechin, markers of tea consumption, were signicantly associated with a reduced risk of developing CRC in men. These results provide a strong rationale for conducting human intervention trials on the prevention of human CRC with well-dened tea polyphenol preparations. In such trials, ACF can be utilized as a surrogate end point biomarker, which are detected and characterized by magnifying chromoendoscopy (28,29). ACF, originally described by Bird (30) in AOM-treated murine colon, have been accepted as pre-neoplastic lesions of the colon. In humans, ACF are commonly observed in familial adenomatous polyposis patients, as well as in sporadic CRC patients (3134). Human ACF harbor many neoplastic alterations, such as increased proliferation activity, mutations in K-RAS and APC genes, as well as microsatellite instability (3336). Hao et al. (37) demonstrated that nuclear and cytoplasmic b-catenin expression was increased in human colonic ACF according to their degree of dysplasia, and moreover, nuclear and cytoplasmic b-catenin expression was increased from dysplastic ACF to adenoma and to adenocarcinoma in the human colon. Consistent with this notion, dysplastic ACF with nuclear b-catenin accumulation recently have been suggested to be more informative pre-neoplastic lesions of chemical carcinogen-induced colon cancer in rats (38).

Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

113

H.Xiao et al.

In preparation for a human intervention trial, we investigated in the current study the effect of a standardized tea polyphenol preparation, Polyphenon E (PPE), on ACF formation in AOM-treated F344 rats. The cellular and molecular changes in the high-grade dysplastic ACF caused by dietary PPE treatment were characterized. Our results provide evidence that PPE is a potential chemopreventive agent for CRC and suggest possible biomarkers for use in future cancer prevention studies in humans. Materials and methods
Animals Male F344 rats at 5 weeks of age were purchased from Taconic Farms (Germantown, NY). After 1 week of acclimation, animals were randomly distributed into control and experimental groups. All animals were housed three in a plastic cage with a lter top. The animal room was controlled at 20 2C, 50 10% humidity and a 12 h light/dark cycle. Animals had free access to food and water at all times. Food cups were replenished with fresh diet twice weekly. Experimental procedure At 7 weeks of age, all animals were subjected to two weekly subcutaneous injections of AOM (Midwestern Research Institute, Kansas City, MO) at a dose of 15 mg/kg each. One day after the second injection, animals were fed a modied 20% high-fat AIN-76A diet (39) containing 0, 0.12 or 0.24% PPE (nine animals per group). PPE was a gift from Dr Yukihiko Hara of the Mitsui Norin Co., Ltd (Tokyo, Japan). This is a standardized green tea polyphenol preparation, containing 65% EGCG, 7% epicatechin-3-gallate, 3% epigallocatechin, 9% epicatechin, 3% gallocatechin gallate and 0.6% caffeine. Dose selection for PPE was based on our previous study, which showed that oral administration of EGCG at doses of 0.08 or 0.16% in drinking uid signicantly decreased small intestinal tumor formation in ApcMin/ mice (11). To minimize possible degradation of tea polyphenols, all stock diets were stored at 4C for the entire period of the experiment. Body weight and food consumption were monitored once and twice a week, respectively, until the termination of the experiment at 8 weeks after the second AOM injection. All animals were killed by CO2 asphyxiation. After laparotomy, the spleen and entire large intestine were harvested. Spleen was washed with ice-cold saline, blotted and then weighed. The large intestine from cecum to anus was longitudinally opened, ushed with ice-cold saline and xed at between two pieces of lter paper in 10% buffered formalin for 24 h. In another experiment, F344 rats were treated with AOM as described above, and then fed with modied AIN-76A diet containing 20% fat. At 34 weeks after the second AOM injection, the entire large intestine was harvested and slit open. Colorectal tumors were dissected with adjacent normal mucosa and then xed in 10% buffered formalin for 24 h. Identication of ACF The formalin-xed colonic tissues were cut into 4 cm segments and stained in 0.2% methylene blue solution for 35 min. The total number of ACF and the number of aberrant crypts (ACs) in each focus were counted under a microscope (40 to 100). ACF were identied with the following morphological characteristics: (i) the enlarged and elevated crypts than normal mucosa and (ii) increased pericryptal space and irregular lumens. Histological analyses of ACF and tumors In order to harvest most of the ACF for histological analyses, methylene bluestained colonic tissues were examined under a microscope, and large ACF containing four or more AC were marked by permanent ink. At least 1012 pieces of colonic tissues containing the marked ACF were dissected from the mid-distal part of each colon, where most ACF occurred. These tissues were then processed and appropriately oriented in parafn blocks for longitudinal sectioning serially at 4 lM thickness. Because ACF may develop into corkscrew shape when they progress into high-grade dysplasia, only a few whole crypts were observed in one section. Sections 1, 10 and 20 were stained with hematoxylin and eosin for histological evaluation, which allowed ACF visualization at different section levels. Based on nuclear to cytoplasmic ratio, cell polarity, chromatin pattern, mitotic gures and mucin secretion, ACF were classied into three categories: (i) ACF with hyperplasia (no dysplasia); (ii) ACF with low-grade dysplasia (elongated, slightly crowded and pseudostratied nuclei, but polarity well preserved, normal or slightly reduced number of goblet cells); and (iii) ACF with high-grade dysplasia (elongated, crowded and pseudostratied nuclei; markedly increased nucleus to cytoplasm ratio, signicantly reduced number of goblet cells, back-to-back glands and markedly decreased interglandular stroma) (37). The sections adjacent to hematoxylin and eosin sections that contained histologically conrmed ACF were used

for immunohistochemistry (IHC). Fixed colorectal tumors were processed and stained as described above, and adenomas and adenocarcinomas were identied as described previously (37). IHC A standard avidinbiotin peroxidase complex method was employed as described previously (37). In brief, after dewaxing and rehydrating, the slides were heated in a pressure cooker in sodium citrate buffer (0.01 M, pH 6.0) for 3 min after reaching full pressure. Endogenous peroxidase was quenched using 3% hydrogen peroxide in methanol. Sections were then blocked for 1 h at room temperature in phosphate-buffered saline containing 3% normal horse or goat serum depending on the origin of the primary antibody. The sections were then immunostained with anti-b-catenin (1 : 3000, BD Biosciences, Franklin Lakes, NJ), cyclin D1 (1 : 100, DAKO, Carpinteria, CA), retinoid X receptor (RXR)a (1 : 1000, Santa Cruz Biotechnology, Santa Cruz, CA) or cleaved caspase-3 (1 : 200, Cell Signaling, Danvers, MA) antibodies overnight at room temperature. The antibodies were diluted in 10% goat or horse serum. The sections were rinsed in phosphate-buffered saline and incubated with a biotinylated secondary antibody and subsequently incubated in VECTORSTAIN ELITE ABC reagent for 30 min, using 3,3#-diaminobenzidine (Vector Laboratories, Burlingame, CA) as the chromogen. Sections were then counterstained for 23 min with hematoxylin (Sigma, St Louis, MO) and mounted with Permount.

Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

Evaluation of the staining Both positive and negative stained cells in high-grade dysplastic ACF were counted for cleaved caspase-3, nuclear expression of b-catenin, cyclin D1 and RXRa. Apoptotic index, nuclear positivity for b-catenin, cyclin D1 and RXRa were then calculated based on the percentage of positive cells in total ACF cells analyzed. The intensity of membranous staining for b-catenin was scored as follows: (i) 0 5 negative, (ii) 1 5 weak, (iii) 2 5 moderate and (iv) 3 5 strong (as intense as normal mucosa). Cytoplasmic staining for b-catenin was scored as follows: (i) 0 5 negative (no cytoplasmic staining), (ii) 1 5 weak, (iii) 2 5 moderate and (iv) 3 5 strong (37). Tumor sections were evaluated for b-catenin and RXRa staining as described above. Statistical analysis The data on ACF multiplicity, labeling index and intensity of staining were analyzed by Students t-test and analysis of variance. Effects of the treatments on the number of ACF in different categories were analyzed by chi-square test.

Results General observation As shown in Figure 1A, all animals had a steady body weight gain during the treatment, and the administration of AOM and 0.12 or 0.24% PPE did not affect the growth of the rats in all the groups measured at different time points. Food consumption of treatment groups of 0.12 or 0.24% PPE was $1215 gram per day per animal, which was not different from that of the AOM-treated positive control group (Figure 1B). During the entire period of experiment, there were no signs of toxicity or conditions, suggesting adverse effects caused by dietary administration of PPE. There was no difference among the mean spleen weights of positive control and treatment animals (average 0.640.67 g) at the time of killing. Dietary PPE treatment inhibited ACF formation After 8 weeks of dietary treatment with PPE, animals were killed and the entire colons were harvested. Colonic ACF were identied and analyzed under a light microscope after methylene blue staining. ACF mostly occurred in the distal colon. Table I summarizes the total number of ACF per rat, the total number of AC per rat, the number of AC per focus and the number (percentage) of foci containing one, two, three or four or more AC. All rats developed ACF in the colon 8 weeks after AOM treatment (100% incidence). Compared with the positive control group, dietary PPE treatment at levels of 0.12 and 0.24% dose dependently decreased the total number of ACF per rat by 16 and 37% (P , 0.01), respectively, and decreased the total number of AC per rat by 17 and 43% (P , 0.01), respectively. The rats treated with 0.24% PPE also showed a signicantly lower number of AC per focus compared with the positive control animals (P , 0.05). We subsequently categorized ACF based on the number of AC in each

114

Polyphenon E inhibits ACF in rat

Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

Fig. 1. Body weight and food consumption of AOM-treated F344 rats and histological classication of AOM-induced colorectal ACF in F344 rats. Male F344 rats were subject to two weekly injections of AOM (15 mg/kg, subcutaneous), and then fed modied AIN-76A diet containing 0, 0.12 or 0.24% PPE for 8 weeks before killing. Body weight (A) and food consumption (B) were monitored weekly. Based on the criteria described in Materials and Methods, ACF were categorized as those with hyperplasia (C), low-grade dysplasia (D) and high-grade dysplasia (E).

focus. The results showed that PPE treatments generally decreased the number of foci in all four categories compared with the positive control group, especially by 0.24% PPE treatment. Furthermore, the percentage of large ACF (containing four or more AC) in total ACF of 0.24% PPE-treated rats was signicantly lower than that of the positive control animals (27% reduction, P , 0.05). As a consequence, the percentage of the smallest ACF (containing one AC) in total ACF of 0.24% PPE-treated rats was signicantly higher than that of the positive control animals (21% increase, P 5 0.05). Dietary PPE treatment inhibited formation of dysplastic ACF ACF from different groups were subjected to histological analyses. A total of 63 ACF from the positive control rats, 76 ACF from 0.12% PPE-treated rats and 42 ACF from 0.24% PPE-treated rats were examined histologically and classied according to the criteria described in Materials and Methods and shown in Figure 1CE. The percentages of ACF with dysplasia were signicantly lower in both the 0.12 and 0.24% PPE treatment groups (Table II). The percentages of high-grade dysplastic ACF among total ACF were reduced .50% by PPE treatment at both dose levels. As a consequence, the percentages of hyperplastic ACF were higher in PPE treatment groups. Induction of apoptosis in high-grade dysplastic ACF by PPE To shed light on the mechanism of action of PPE, ACF with highgrade dysplasia from positive control and 0.24% PPE-treated rats were analyzed for apoptosis by IHC with anti-cleaved caspase-3 antibody (Figure 2H and I). Cleaved caspase-3-positive cells displaying nuclear staining were observed in ACF, but rarely in the normal mucosa. Apoptotic index was dened as the percentage of cleaved caspase-3 positively stained cells among total cells in high-grade

dysplastic ACF observed in each tissue section. About 1.3% cells from the positive control group were positively stained by anticleaved caspase-3 antibody, i.e. with an apoptosis index of 1.3%. The group that received dietary treatment with 0.24% PPE had a signicantly higher apoptotic index of 3.0% in high-grade dysplastic ACF (2.3-fold higher, P 5 0.05, Table III). PPE treatment did not cause any appreciable increase of cleaved caspase-3-positive cells in normal mucosa. Modulation of b-catenin expression in high-grade dysplastic ACF by PPE Aberrant b-catenin expression is a common feature of AOM-induced rat colon tumor and human CRC (40,41). We analyzed b-catenin expression in normal mucosa, high-grade dysplastic ACF, adenoma and adenocarcinoma from positive control rats. The staining patterns with anti-b-catenin antibody are shown in Figure 2AE. Normal mucosa showed a distinct membranous staining of b-catenin. High-grade dysplastic ACF, adenoma and adenocarcinoma displayed reduced membranous staining, accompanied by increased nuclear and cytoplasmic staining of b-catenin, as compared with the normal mucosa. Moreover, adenocarcinoma showed weaker membranous staining of b-catenin than high-grade dysplastic ACF and adenoma. We then analyzed b-catenin expression in high-grade dysplastic ACF of 0.24% PPE-treated animals and compared it to that of the positive control group (Figure 2C and Table III). In the positive control group, 26.2% of high-grade dysplastic ACF cells showed positive nuclear b-catenin staining. In the 0.24% PPE treatment group, only 7.2% ACF cells showed positive nuclear b-catenin staining (73% reduction, P , 0.01). Cytoplasmic b-catenin expression in high-grade dysplastic ACF was signicantly lower in 0.24% PPE-treated group by 73%

115

H.Xiao et al.

43.2 13.9 (21.1 6.4a,) 38.7 7.7 (22.8 4.7a,) 20.1 6.0 (15.5 4.0b,)

Table II. Effect of dietary PPE treatment on histology of AOM-induced colorectal ACF in F344 rats Total ACF Hyperplasia Low-grade High-grade Pa analyzed (%) dysplasia (%) dysplasia (%) Control 63 0.12% PPE 76 0.24% PPE 42
a

Four or more AC

25 (40.0) 49 (64.5) 28 (66.7)

19 (30.1) 17 (22.4) 8 (19.0)

19 (30.1) 10 (13.1) 6 (14.3)

0.01 0.025

Chi-square test was used to test the difference between positive control and treatment groups.

38.6 8.1 (19.0 4.4) 34.2 6.7 (20.0 3.3) 23.6 5.5 (18.3 4.1)

P , 0.01).

(P , 0.01), compared with the positive control. In contrast, membranous b-catenin expression level in high-grade dysplastic ACF was higher in the 0.24% PPE-treated group by 32% (P , 0.04). These results demonstrated that dietary PPE treatment prevented aberrant b-catenin expression in ACF with high-grade dysplasia. Inhibition of cyclin D1 expression in high-grade dysplastic ACF by PPE Cyclin D1 is an important downstream protein of b-catenin signaling and is involved in cell growth and proliferation. We analyzed cyclin D1 expression in high-grade dysplastic ACF to determine whether cyclin D1 expression would be reduced concomitantly to the prevention of aberrant b-catenin expression by PPE treatment observed above. ACF from both positive control and PPE-treated groups showed nuclear staining patterns for cyclin D1 (Figure 2F and G). In the positive control group, 59.2% of high-grade dysplastic ACF cells showed positive nuclear cyclin D1 staining, while only 31.2% high-grade dysplastic ACF cells showed positive nuclear cyclin D1 staining from 0.24% PPE-treated group (47% reduction, P , 0.01, Table III). These results further support the observation that PPE treatment prevented aberrant b-catenin function in ACF with highgrade dysplasia. Loss of RXRa expression and its prevention by PPE in high-grade dysplastic ACF Studies have shown that RXRs were associated with carcinogenesis in multiple organ sites (4246). To shed light on the role of RXRa in colorectal carcinogenesis, RXRa expression was analyzed in AOMinduced colorectal ACF with high-grade dysplasia, adenomas and adenocarcinomas (Figure 3). In positive control rats, normal epithelial cells showed strong nuclear staining of RXRa; whereas in adenomas, the nuclear RXRa staining was much weaker; and in adenocarcinomas, the nuclear RXRa expression was almost lost. It is interesting that a decreased nuclear level was even observed in high-grade dysplastic ACF. Using staining intensity of RXRa in normal epithelial cells as the internal standard, IHC evaluation showed that 34.2% highgrade dysplastic ACF cells from the positive control group showed positive nuclear RXRa staining. In the dietary PPE-treated group, positively stained cells in high-grade dysplastic ACF cells were 56.6% (65% increase, P , 0.01, Table III). No appreciable change was observed in the normal epithelia from the positive control and PPE-treated rats. Discussion In the present study, we investigated the effect of PPE, a standardized tea polyphenol preparation, on colorectal ACF formation in AOMtreated F344 rats. To mimic typical American dietary composition, the animals were fed a modied AIN-76A diet that contained 20% fat during the entire period of study. The composition of the 20% fat was a slight modication of the American blend fat developed by the Institute of Shortening and Edible Oils (39). PPE was fed to the animals through the diet. Dietary administration is a more convenient way of delivering PPE to the rat as compared with administration
Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

Three AC

93.2 11.3 (45.8 5.0) 74 13.2 (43.4 6.7) 63.6 8.9 (49.3 3.8) Control 0.12% PPE 0.24% PPE 203.3 11.2 170.2 12.0b, 128.8 13.3c,
a,

116

Values are means SDs. Different superscripts (a, b and c) indicate signicant statistical difference by one-way analysis of variance (P , 0.05;

Number (percentage) of foci containing

Two AC

Table I. Effect of dietary PPE treatment on ACF formation in AOM-treated F344 rats

Group

Total number of ACF per rat

560.1 63.6 464.6 41.0b, 318.9 43.9c,

Total number of AC per rat

a,

2.75 0.25 2.73 0.16a, 2.47 0.17b,

Number of AC per focus

a,

28.3 6.5 (13.96 3.2 ) 23.3 3.4 (13.8 2.5a,) 21.6 3.3 (16.8 2.7b,)

One AC

a,

Polyphenon E inhibits ACF in rat

Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

Fig. 2. b-Catenin expression in normal mucosa, high-grade dysplastic ACF, adenoma and adenocarcinoma in AOM-treated rats, and effect of dietary PPE treatment on b-catenin expression, cyclin D1 expression and apoptosis in high-grade dysplastic ACF in AOM-treated rats. Staining patterns of b-catenin expression in normal mucosa of the positive control rats (A), ACF with high-grade dysplasia of the positive control rats (B), ACF with high-grade dysplasia of 0.24% PPE-treated rats (C), adenoma of the positive control rats and (D), adenocarcinoma of the positive control rats (E). Adenoma and adenocarcinoma samples were harvested from another experiment as described in Materials and Methods. Staining patterns of cyclin D1 in ACF with high-grade dysplasia of the positive control (F) and 0.24% PPE-treated rats (G). Apoptosis was visualized by antibodies against activated (cleaved) caspase-3. More apoptotic cells were seen in ACF from rats treated with dietary PPE (I) than those from the positive control rats (H). Arrow indicates apoptotic cells. N, normal mucosa; AC, aberrant crypt.

Table III. Effects of dietary PPE treatment on b-catenin, cyclin D1 and RXRa expression and apoptosis in high-grade dysplastic ACF in AOM-treated F344 rats Treatment Number of ACF analyzed b-catenin Nuclear staining (%) Control 0.24% PPE P value 12 6 26.2 3.2 7.2 3.8 0.01 Membranous staining (intensity) 1.9 0.2 2.5 0.2 0.04 Cytoplasmic staining (intensity) 1.5 0.3 0.4 0.2 0.01 59.2 4.4 31.2 7.8 0.01 1.3 0.3 3.0 0.5 0.05 34.2 1.6 56.6 0.5 0.01 Cyclin D1 (%) Apoptotic index (%) RXRa (%)

b-Catenin was analyzed by IHC using antibody against b-catenin for nuclear (positivity), membranous and cytoplasmic (both for intensity). Cyclin D1 was analyzed by IHC using antibody against cyclin D1 and apoptotic index was determined by IHC using antibody against cleaved caspase-3 (Asp175). RXRa was analyzed by IHC using antibody against RXRa. The positivity was expressed as percentage of positive cells in all ACF cells analyzed. Two-tailed Students t-test was used for statistical analysis. Data represent mean SD.

through drinking uid because uid consumption may be affected by the bitterness of the PPE solution. Our results demonstrated that dietary PPE treatment inhibited the formation of ACF in AOM-treated rats, and this inhibition is characterized by (i) decrease in total number of ACF per rat, total number of AC per rat and number of AC per focus; (ii) reduction in the percentage of large ACF (with four or more AC) among total ACF; and (iii) reduction in the percentage of highgrade dysplastic ACF among total ACF. In the colon mucosa, mucin-depleted foci and b-catenin-accumulated crypts, all with dysplasia in nature, have been suggested to be more relevant pre-neoplastic lesions for colon cancer, whereas hyperplastic ACF often lack the potential to develop into neoplasia (4749). Herein, we categorized ACF into three types, i.e. hyperplastic, low-grade dysplastic and high-grade dysplastic. We prefer to use the term dysplastic ACF following that by Hao et al (37) and Siu et al (50), even this type of lesions are also known as microadenomas (51,52). Our results

showed that high-grade dysplastic ACF, like adenomas and adenocarcinomas, displayed reduced membranous expression, but increased nuclear and cytoplasmic expression, of b-catenin. We demonstrated, for the rst time, the inhibitory effect of dietary tea polyphenols on the formation of high-grade dysplastic ACF, and this effect was associated with increased apoptosis, decreased nuclear expression levels of bcatenin and cyclin D1 and increased expression level of nuclear RXRa. Aberrant expression of these molecules was not frequently observed in low-grade dysplastic or hyperplastic ACF, whereas high-grade dysplastic ACF showed relatively uniform aberrant expression of these molecules. Therefore, their responses to PPE treatment were not systematically analyzed in low-grade dysplastic or hyperplastic ACF in the current study. The biological activities of retinoids are mediated by two families of nuclear receptors: RXRs and retinoic acid receptors. Dysfunctions of RXRs have been associated with carcinogenesis in multiple organ

117

H.Xiao et al.

Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

Fig. 3. RXRa expression in normal mucosa, high-grade dysplastic ACF, adenoma and adenocarcinoma in AOM-treated rats, and effect of dietary PPE treatment on RXRa expression in high-grade dysplastic ACF in AOM-treated rats. Staining patterns of RXRa in ACF with high-grade dysplasia of the positive control rats (A), ACF with high-grade dysplasia of 0.24% PPE-treated rats (B), adenoma of the positive control rats (C) and adenocarcinoma of the positive control rats (D). N, normal mucosa; AC, aberrant crypt; Ad, adenoma; Ca, adenocarcinoma. Adenoma and adenocarcinoma samples were harvested from another experiment as described in Materials and Methods.

sites, such as the breast, lung, thyroid, prostate and pituitary (4246). In this work, we demonstrated that the expression of RXRa was decreased in ACF with high-grade dysplasia, and it was further decreased in adenomas and adenocarcinomas. These results suggest that decreased nuclear expression of RXRa occurs before the development of the adenoma and may be used as a biomarker for monitoring colon cancer chemopreventive activity. Recently, Xiao et al (53) demonstrated an APC-independent regulation of b-catenin degradation via RXRa-mediated pathway, in which RXRa repressed b-catenin-mediated transcription by inducing the proteasomal degradation of b-catenin. Consistent with this nding, our results showed that the decreased expression of RXRa occurred in parallel with the increased nuclear accumulation of bcatenin in colorectal lesions during AOM-induced rat colorectal carcinogenesis. More interestingly, we demonstrated that dietary PPE treatment prevented the loss of RXRa expression as well as the aberrant nuclear expression of b-catenin in high-grade dysplastic ACF. Nuclear expression of cyclin D1, an important downstream protein of b-catenin pathway, was also decreased in high-grade dysplastic ACF by dietary PPE treatment. Moreover, dietary PPE treatment signicantly induced apoptosis in high-grade dysplastic ACF. These ndings suggested that, besides dysfunctions of APC signaling pathways, the decrease in RXRa expression might represent a mechanism by which pre-neoplastic lesions, such as high-grade dysplastic ACF, escape the control of normal growth and differentiation and progress into colorectal tumors. Maintaining RXRa expression level in dysplastic ACF by PPE treatment may contribute to the lower nuclear b-catenin level through proteasomal degradation of b-catenin. It is possible that high-grade dysplastic ACF with high nuclear b-catenin level and low RXRa level may have more potential to develop into adenomas or adenocarcinomas than high-grade dysplastic ACF with lower nuclear b-catenin level and higher RXRa level. Further investigation is needed to establish a relationship between maintaining RXRa level and the prevention of aberrant nuclear b-catenin expression by PPE in this animal model. It is also possible that the decreased nuclear level of b-catenin is due to a decreased nuclear translocation of b-catenin into the nuclear as we reported previously (11).

Taken together, dietary PPE treatment inhibited ACF formation, especially high-grade dysplastic ACF, in AOM-treated F344 rats. We propose that this inhibition was associated with promoting apoptosis, preventing aberrant nuclear b-catenin accumulation, decreasing cyclin D1 level and maintaining RXRa expression. These may be useful biomarkers in colon cancer prevention studies in humans. Acknowledgements
Conict of Interest Statement: None declared.

References
1. Yang,C.S. et al. (1993) Tea and cancer. J. Natl Cancer Inst., 85, 10381049. 2. Yang,C.S. et al. (2002) Inhibition of carcinogenesis by tea. Annu. Rev. Pharmacol. Toxicol., 42, 2554. 3. Chung,F.L. et al. (2003) Tea and cancer prevention: studies in animals and humans. J. Nutr., 133, 3268S3274S. 4. Crespy,V. et al. (2004) A review of the health effects of green tea catechins in in vivo animal models. J. Nutr., 134, 3431S3440S. 5. Khan,N. et al. (2006) Targeting multiple signaling pathways by green tea polyphenol (-)-epigallocatechin-3-gallate. Cancer Res., 66, 2500 2505. 6. Kim,S. et al. (2000) Plasma and tissue levels of tea catechins in rats and mice during chronic consumption of green tea polyphenols. Nutr. Cancer, 37, 4148. 7. Lambert,J.D. et al. (2003) Epigallocatechin-3-gallate is absorbed but extensively glucuronidated following oral administration to mice. J. Nutr., 133, 41724177. 8. Yin,P. et al. (1994) Experimental studies of the inhibitory effects of green tea catechin on mice large intestinal cancers induced by 1,2-dimethylhydrazine. Cancer Lett., 79, 3338. 9. Suganuma,M. et al. (2001) Combination cancer chemoprevention with green tea extract and sulindac shown in intestinal tumor formation in Min mice. J. Cancer Res. Clin. Oncol., 127, 6972. 10. Orner,G.A. et al. (2003) Suppression of tumorigenesis in the Apc(min) mouse: down-regulation of beta-catenin signaling by a combination of tea plus sulindac. Carcinogenesis, 24, 263267.

118

Polyphenon E inhibits ACF in rat

11. Ju,J. et al. (2005) Inhibition of intestinal tumorigenesis in Apcmin/ mice by (-)-epigallocatechin-3-gallate, the major catechin in green tea. Cancer Res., 65, 1062310631. 12. Ju,J. et al. (2003) Effects of green tea and high-fat diet on arachidonic acid metabolism and aberrant crypt foci formation in an azoxymethane-induced colon carcinogenesis mouse model. Nutr. Cancer, 46, 172178. 13. Kim,M. et al. (1994) Preventive effect of green tea polyphenols on colon carcinogenesis. In Huang,M.-T.,Osaswa,T.,Ho,C.-T. and Rosen,R.T. (eds.) Food Phytochemicals in Cancer Prevention. Vol. 2, ACS Symposium Series 546 pp. 5155. 14. Xu,M. et al. (1996) Protection by green tea, black tea, and indole-3carbinol against 2- amino-3-methylimidazo[4,5-f]quinoline-induced DNA adducts and colonic aberrant crypts in the F344 rat. Carcinogenesis, 17, 14291434. 15. Challa,A. et al. (1997) Interactive suppression of aberrant crypt foci induced by azoxymethane in rat colon by phytic acid and green tea. Carcinogenesis, 18, 20232026. 16. Wargovich,M.J. et al. (2000) Efcacy of potential chemopreventive agents on rat colon aberrant crypt formation and progression. Carcinogenesis, 21, 11491155. 17. Hirose,M. et al. (2002) Lack of inhibitory effects of green tea catechins in 1,2-dimetylhydrazine-induced rat intestinal carcinogenesis model: comparison of the different formulations, administration routes and doses. Cancer Lett., 188, 163170. 18. Hirose,M. et al. (2001) Green tea catechins enhance tumor development in the colon without effects in the lung or thyroid after pretreatment with 1,2Dimethylhydrazine or 2,2#-dihydroxy-di-n-propylnitrosamine in male F344 rats. Cancer Lett., 168, 2329. 19. Weisburger,J.H. et al. (1998) Effect of tea extracts, polyphenols, and epigallocatechin gallate on azoxymethane-induced colon cancer. Proc. Soc. Exp. Biol. Med., 217, 104108. 20. Caderni,G. et al. (2000) Effects of black tea, green tea and wine extracts on intestinal carcinogenesis induced by azoxymethane in F344 rats. Carcinogenesis, 21, 19651969. 21. Blot,W.J. et al. (1996) cancer: a review of the epidemiological evidence. Eur. J. Cancer Prev., 5, 425438. 22. Kohlmeier,L. et al. (1997) Tea and cancer prevention: an evaluation of the epidemiologic literature. Nutr. Cancer., 27, 113. 23. Bushman,J.L. (1998) Green tea and cancer in humans: a review of the literature. Nutr. Cancer, 31, 151159. 24. Lambert,J.D. et al. (2003) Mechanisms of cancer prevention by tea constituents. J. Nutr., 133, 3262S3267S. 25. Sun,C.L. et al. (2006) Green tea, black tea and colorectal cancer risk: a meta-analysis of epidemiologic studies. Carcinogenesis, 27, 13011309. 26. Yang,G. et al. (2007) Prospective cohort study of green tea consumption and colorectal cancer risk in women. Cancer Epidemiol. Biomarkers Prev., 16, 12191223. 27. Yuan,J.M. et al. (2007) Urinary biomarkers of tea polyphenols and risk of colorectal cancer in the Shanghai Cohort Study. Int. J. Cancer, 120, 1344 1350. 28. Takayama,T. et al. (1998) Aberrant crypt foci of the colon as precursors of adenoma and cancer. N. Engl. J. Med., 339, 12771284. 29. Adler,D.G. et al. (2002) Endoscopic identication and quantication of aberrant crypt foci in the human colon. Gastrointest Endosc, 56, 657662. 30. Bird,R.P. (1987) Observation and quantication of aberrant crypts in the murine colon treated with a colon carcinogen: preliminary ndings. Cancer Lett., 37, 147151. 31. Pretlow,T.P. et al. (1991) Aberrant crypts: putative preneoplastic foci in human colonic mucosa. Cancer Res., 51, 15641567.

32. Roncucci,L. et al. (1991) Identication and quantication of aberrant crypt foci and microadenomas in the human colon. Hum. Pathol., 22, 287294. 33. Roncucci,L. et al. (2000) Aberrant crypt foci in colorectal carcinogenesis. Cell and crypt dynamics. Cell Prolif., 33, 118. 34. Takayama,T. et al. (2001) Analysis of K-ras, APC beta-catenin in aberrant crypt foci in sporadic adenoma, cancer, and familial adenomatous polyposis. Gastroenterology, 121, 599611. 35. Smith,A.J. et al. (1994) K-ras codon 12 mutations in aberrant crypt foci from human colons. Cancer Res., 54, 55275530. 36. Augenlicht,L.H. et al. (1996) Evidence for genomic instability in human colonic aberrant crypt foci. Oncogene, 12, 17671772. 37. Hao,X.P. et al. (2001) Beta-catenin expression is altered in human colonic aberrant crypt foci. Cancer Res., 61, 80858088. 38. Ochiai,M. et al. (2003) Characterization of dysplastic aberrant crypt foci in the rat colon induced by 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine. Am. J. Pathol., 163, 16071614. 39. Rao,C.V. et al. (2001) Modulation of experimental colon tumorigenesis by types and amounts of dietary fatty acids. Cancer Res., 61, 19271933. 40. Takahashi,M. et al. (1998) Beta-catenin is frequently mutated and demonstrates altered cellular location in azoxymethane-induced rat colon tumors. Cancer Res., 58, 4246. 41. Hao,X. et al. (1997) Reciprocity between membranous and nuclear expression of beta-catenin in colorectal tumours. Virchows Arch, 431, 167172. 42. Zhang,H. et al. (2004) Haploid inactivation of the amplied-in-breast cancer 3 coactivator reduces the inhibitory effect of peroxisome proliferatoractivated receptor gamma and retinoid X receptor on cell proliferation and accelerates polyoma middle-T antigen-induced mammary tumorigenesis in mice. Cancer Res., 64, 71697177. 43. Takiyama,Y. et al. (2004) Decreased expression of retinoid X receptor isoforms in human thyroid carcinomas. J. Clin. Endocrinol. Metab., 89, 58515861. 44. Huang,J. et al. (2002) Prostatic intraepithelial neoplasia in mice with conditional disruption of the retinoid X receptor alpha allele in the prostate epithelium. Cancer Res., 62, 48124819. 45. Martinet,N. et al. (2000) Retinoic acid receptor and retinoid X receptor alterations in lung cancer precursor lesions. Cancer Res., 60, 28692875. 46. Gittoes,N.J. et al. (1998) Retinoid X receptor expression in the normal pituitary and clinically non-functioning pituitary tumours. Clin. Endocrinol. (Oxf), 48, 425433. 47. Femia,A.P. et al. (2007) Frequent mutation of Apc gene in rat colon tumors and mucin-depleted foci, preneoplastic lesions in experimental colon carcinogenesis. Cancer Res., 67, 445449. 48. Hirose,Y. et al. (2003) Azoxymethane-induced beta-catenin-accumulated crypts in colonic mucosa of rodents as an intermediate biomarker for colon carcinogenesis. Carcinogenesis, 24, 107111. 49. Mori,H. et al. (2005) Signicance and role of early-lesions in experimental colorectal carcinogenesis. Chem. Biol. Interact., 155, 19. 50. Siu,I.M. et al. (1997) Identication of dysplasia in human colonic aberrant crypt foci. Am. J. Pathol., 150, 18051813. 51. Cheng,L. et al. (2003) Aberrant crypt foci as microscopic precursors of colorectal cancer. World J. Gastroenterol., 9, 26422649. 52. Roncucci,L. et al. (1991) Classication of aberrant crypt foci and microadenomas in human colon. Cancer Epidemiol. Biomarkers Prev., 1, 5760. 53. Xiao,J.H. et al. (2003) Adenomatous polyposis coli (APC)-independent regulation of beta-catenin degradation via a retinoid X receptor-mediated pathway. J. Biol. Chem., 278, 2995429962. Received July 8, 2007; revised August 31, 2007; accepted September 9, 2007

Downloaded from carcin.oxfordjournals.org by guest on October 12, 2010

119