- Email: [email protected]
Enhanced estrogenic responses and sensitivity to azoxymethane following dietary soy isoflavone supplementation in older female rats
Food and Chemical Toxicology 45 (2007) 628–637 www.elsevier.com/locate/foodchemtox
Enhanced estrogenic responses and sensitivity to azoxymethane following dietary soy isoflavone supplementation in older female rats K.T. Daly a, A.C. Tracy a, M. Malik
a,d
, T. Wang b, S. Francke-Carroll c, B.A. Magnuson
a,*
a University of Maryland, College Park, MD, United States Phytonutrients Laboratory, USDA, Beltsville, MD, United States Pathology Branch, CFSAN, FDA, College Park, MD, United States d OBG, USUHS Bethesda, MD, United States
b c
Received 23 August 2005; accepted 22 October 2006
Abstract Soy isoflavones are popular supplements among middle-aged and older women based on their potential protection against cancer and their use as alternative hormone replacement therapy. The purpose of this study was to investigate the effects of dietary soy isoflavones on early stage colon cancer in various ages of female rats. Young (1 month), mature (11 month) and old (22 month) female Fisher 344 rats were fed either the control diet or a diet containing 0.4% soy isoflavone isolate for 1 week, injected once with 20 mg/kg azoxymethane (AOM) and maintained on the diets for another 15 weeks. The concentration of isoflavones in the diet was 2 g/kg diet, composed of 1.2 g/kg genistin, 0.7 g/kg daidzin and 0.1 g/kg other isoflavones including glycitin, acetylgenistin, acetyldaidzin, genistein, daidzein, and glycitein. There was no difference over all ages in the development of preneoplastic colonic aberrant crypt foci between rats fed the soy compared to the control diet, indicating that the soy diet did not provide protection against early stage colonic carcinogenesis. On the contrary, several adverse effects of soy supplementation in female AOM-treated rats were observed. Soy-supplemented rats had greater weight loss and a slower recovery of body weight following the AOM injection compared to rats fed the control diet and these changes increased with age. Five of the 21 rats fed the soy supplement died before the end of the experiment while all animals on the control diet survived to term. The density of normal crypts lining the colonic mucosa was reduced in rats fed the soy compared to control diet, indicating gastrointestinal damage. Uterine weights, serum estradiol and serum isoflavone levels were increased in mature and old female rats fed the soy-supplemented diets compared to age-matched controls, suggesting an increasing estrogenic response with age to isoflavone supplementation. These adverse effects of soy isoflavones in aged female animals need further examination because women, and particularly older women, are the prime target population for consumption of soy supplements. 2006 Elsevier Ltd. All rights reserved. Keywords: Soy isoflavones; Aberrant crypt foci; Gastrointestinal; Estradiol
1. Introduction Colon cancer is the second most common cause of cancer death in the United States and it is predicted that approximately 56,290 people will die each year from this disease (Jemal et al., 2005). Furthermore, among newly diagnosed cases, age is the primary risk factor with 90% of all colon cancer occurring in people over the age of *
Corresponding author. Tel.: +1 3014054523; fax: +1 3013143313. E-mail address: [email protected] (B.A. Magnuson).
0278-6915/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.fct.2006.10.021
50. Currently, incidences of colorectal cancer as well as the death rates are approximately equal in both men and women (Jemal et al., 2005). However, due to their increased susceptibility to chemical colon carcinogens, the majority of in vivo colon cancer studies are completed using male animals and limited research has been performed utilizing a female animal model (Moriya et al., 1978). In addition to the risk posed by increased age on colon cancer development, post-menopausal women also have decreased hormone levels, specifically estrogen. The steroid hormone, estrogen, has been shown to play a major role in
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
the development of cancer in women. For example, studies have shown that menopausal women taking hormone replacement therapy greatly decrease their risk of colon cancer suggesting that estrogen may play a role in preventing or protecting against colon cancer in aging women (Campbell-Thompson et al., 2001; Foley et al., 2000; Genazzani et al., 2001; Gustafsson, 2003; Jemal et al., 2005; Konstantinopoulos et al., 2003; Rossouw et al., 2002). The risk of developing colon cancer is also greatly affected by diet. Epidemiological studies suggest that soy foods and soy isoflavones may act to inhibit colon cancer (Spector et al., 2003). However, the effect of soy isoflavones, either in purified form or in combination with soy protein, on experimental colon cancer has not been consistent. Reduction of pre-cancerous lesions called aberrant crypt foci (ACF) (Bird et al., 1989) by soy protein isolate or soy flakes containing isoflavones has been reported (Murillo et al., 2004; Thiagarajan et al., 1998). A decrease in the size, but not incidence, of AOM-induced colon tumors in male rats consuming a diet containing 10% miso, a fermented soy product, was recently reported (Ohuchi et al., 2005). However, in other studies, dietary soy isoflavones either did not inhibit ACF or colonic tumors (Davies et al., 1999; Sorensen et al., 1998) or enhanced colon tumor incidence (Rao et al., 1997). All of these studies utilized male animals. Recently, Guo et al. (2004) reported that soy protein plus an isoflavone mix, but not either soy protein alone or soy plus genistein, reduced colon tumor incidence in young ovariectomized female mice. To our knowledge, no study has investigated the effect of soy isoflavones or estrogen on colon carcinogenesis in older female animals, having reduced estrogen levels due to aging. A recent study in our laboratory showed that young F344 male rats, injected subcutaneously with azoxymethane (AOM), a known colon carcinogen, and fed a semi-purified AIN-93 diet, had a statistically significant increase in ACF numbers and multiplicity compared to identically treated mature and old male rats (Kwon et al., 2004). In an earlier study, however, we found that old female Sprague–Dawley rats, injected intraperitoneally with AOM and fed a standard laboratory rat chow, had statistically significant increased ACF numbers and multiplicity compared to young and mature female rats (Magnuson et al., 2000). We hypothesized that the different results in these two studies may be attributed to animal gender; if estrogen has protective effects against early stages of colon cancer in female animals, then, as estrogen levels decrease with age, the older females would become more susceptible to colon carcinogens such as AOM. We believe that the effects of soy isoflavones on early stages of colon cancer need further investigation in an aging female animal model because women are taking increasing amounts of soy during and after menopause as an alternative to hormone replacement therapy (HRT) (Kok et al., 2005).
629
The purposes of this study were to determine the effect of age and soy isoflavone supplementation on ACF development in a female animal model. We choose F344 rats for this investigation because they are widely used in aging studies. We hypothesized that any effects of naturally produced estrogen on colon cancer development may be further enhanced by the phytoestrogens in soy. Therefore, we investigated the effect of a soy isoflavone-supplemented diet on ACF development in three age groups of female rats. Unexpectedly, we observed significant adverse effects in older rats fed the soy diet and subsequently, we evaluated the morphology of other tissues and the expression levels of estrogen receptor b. 2. Materials and methods 2.1. Study design and animal treatment With the approval of and in accordance with regulations of the United States Department of Agriculture Animal Use and Care Committee, 14 young (one month old), 14 mature (11 month old) and 14 old (22 month old) F344 female rats were obtained from the NIH aged animal colony (Bethesda, MD). They were housed singly in 8 · 8 · 12 in ONECAGETM cages with raised stainless steel wire bar floors and a ventilated EnviroGARD-B system under twelve hours reverse light/dark cycle. The rats were given free access to water and the AIN-93 powdered control diet, and allowed to acclimate for four days. Following this four-day acclimation period, rats from each age group were placed into either the AIN-93 control diet group or a soy isoflavone experimental diet group resulting in six groups with seven rats per group. The soy isoflavone diet was prepared by adding 4 g of Novasoy (Archer Daniels Midland, Decatur, Illinois) per kg AIN-93 powdered diet to create a final concentration of 1.9 mg isoflavone/g diet, based on reported concentration of 47.07% isoflavones in Novasoy. The amount of soy isoflavones added to the diet in this study was based on the effective dose used in a prostate chemopreventative study (Zhou et al., 1999). We opted to use the high end of the dose range for rat studies with isoflavones (Demonty et al., 2002; Nakai et al., 2005a,b; Wade et al., 2003) as female Wistar rats were reported to have lower plasma concentrations following dosing with genistein as compared to males (Coldham and Sauer, 2000), and because we thought it was possible that older rats would have a lower response due to lower estrogen and estrogen receptor levels. One week after being assigned to experimental diets, all rats were given one 20 mg/kg body weight s.c. injection of AOM (Sigma Chemicals, St. Louis, MO). We elected to use one injection at 20 mg/kg over the common protocol of two weekly injections of 15 mg/kg to reduce animal handling and animal stress as we observed that the older female rats tended to be more stressed by handling than the young female rats or the male older animals we have used in previous experiments. Dosing rats with 20 mg/kg of AOM is common practice and reported to be well tolerated (Caderni et al., 2002; Delker et al., 2000; Ealey et al., 2001; Khare et al., 2003; Morioka et al., 2005; Momen et al., 2002; Shimoji et al., 2003; Suzuki et al., 2004). Following the AOM injections, rats were weighed every day until body weights were restored and then weighed once a week for the remainder of the study. At the start of week six, the amount of food each animal consumed was measured over a three-day period.
2.2. Analysis of isoflavones Chemicals: Isoflavone standards (genistein, daidzein, glycitein, genistin, daidzin, glycitin, acetyldaidzin and acetylgenistin) for HPLC analyses were purchased from LC Laboratories (Woburn, Massachusetts). All of the isoflavone standards used were at least 99% pure, except for acetylgenistin, which was at least 98% pure. Acetonitrile, acetic acid, methanol, and flavone were HPLC grade reagents from Fisher Scientific (Fair Lawn,
630
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
NJ). Sodium acetate trihydrate was from J.T. Baker Inc. (Phillipsburg, NJ). Isoflavone extraction: Isoflavones were extracted as previously reported (Lin and Giusti, 2005) from 100 mg NovaSoy isoflavone concentrate using 10 ml of aqueous methanol solution (80%), in a 1:5 (w:v) ratio, and using flavone as internal standard. The sample was homogenized for 2 h at room temperature and vacuum filtered through Whatman No. 41 filter paper (Whatman International Ltd., Kent, England) using a Buchner funnel. The filtrate was evaporated using a rotary evaporator with a 40 C water bath. The extract was redissolved in 5 ml 16% aqueous acetonitrile solution and refrigerated for no more than 4 h before HPLC analysis. The extraction was conducted twice during the experiment, and each time the analyses were done in triplicate. HPLC analysis: Analysis of isoflavones was carried out using reversed-phase separation of the compounds on a C18 column (Waters Corp., Milford, MA) and using a linear gradient composed of acidified water (A: 0.1% acetic acid and 5% acetonitrile in water) and acidified acetonitrile (B: 0.1% acetic acid in acetonitrile) as previously reported (Lin and Giusti, 2005). After the injection of 25 ll of the extract sample, the linear gradient started from 10% B to 14% over 10 min, then increased to 20% over 2 min, maintained at 20% for 8 min, continued to increase to 70% over 10 min, maintained at 70% for 3 min, and returned to 10% at the end of the 34 min running time. The flow rate was maintained at 1 ml/ min throughout the running of the sample. A high-pressure liquid chromatography (Waters Delta 600 system) equipped with a photodiode array detector (Waters 996), autosampler (Waters 717plus), and Millennium32 software (Waters Corp.) was used. Elution was monitored at 254 nm, and spectral data from 200 to 450 nm were recorded and stored over the time of the run on all samples. Isoflavones were identified by comparing spectral data and retention times to those of pure standards. Calibration curves, prepared by using different concentrations of pure isoflavone standards and flavone (used as internal standard), were used for quantitation of the isoflavones in the samples. Quantification of isoflavones was based on the calibration curves and corrected with the peak area ratio of flavone and isoflavones.
2.3. Collection of tissues Fifteen weeks after AOM injections, rats were euthanized by CO2 inhalation. Blood was collected via cardiac puncture. Colons were removed and flushed with phosphate buffered saline (PBS). A 1 cm section was cut from the middle of each colon and fixed in 10% buffered formalin leaving two partial segments. These two segments were both opened via a longitudinal incision. A mucosal layer sample was collected by scraping a 3 cm section from the cecal end with a glass microscope slide. Scrapings were then fixed in RNAlater (Ambion, Austin, TX). The remaining colon sections were pinned flat on a balsa wood board and stored in formalin for at least 48 h. Uterine, ovary and mammary tissues were collected and also fixed in formalin. Liver and kidney tissues were removed, a section saved in formalin, and the remainder frozen in liquid nitrogen and stored at 80 C. All formalin-fixed tissues were subsequently embedded in paraffin, stained with hematoxylin and eosin (H&E) and evaluated for histopathological changes.
2.4. Identification of aberrant crypt foci Following fixation, the flattened colon tissues were stained with a 0.2% methylene blue/PBS solution. ACF were visualized by light microscopy and identified from normal crypts using a widely accepted set of criteria: they are larger, have an increased peri-cryptal area, have greater staining intensity due to a thickened layer of epithelial cells, are slightly raised above adjacent normal crypts, and have abnormally shaped lumina (Bird and Good, 2000). The total number of ACF and ACF multiplicity (number of crypts per focus) were quantified for each animal. ACF with a multiplicity of 2–3 were designated as small, 4–5 were designated as medium, and greater than five were designated as large. During quantifi-
cation, the scorer was not aware of corresponding animal numbers in order to eliminate bias.
2.5. Serum estrogen quantification To confirm that the older female rats had lower serum estrogen levels, estrogen levels were quantified. After clotting, blood samples were centrifuged for 30 min at 10,000 RPM. Serum was collected and stored at 20 C. Estradiol was measured in the serum of each animal using the Ultra-Sensitive Estradiol RadioImmunoassay (RIA) kit according to the manufacturer’s instructions (Diagnostic Systems Laboratories, Webster, TX).
2.6. Serum isoflavone quantification Serum samples (0.5 ml) were extracted as described above for the soy isolate, and redissolved in 0.3 ml aqueous acetonitrile. The isoflavone extract was mixed with 5 ml of 0.1 M sodium acetate buffer (pH 5.0) and 100 ll b-glucuronidase (from Helix pomatia, with 131,000 b-glucuronidase units/ml and 3.180 sulfatase units/ml, Sigma Chemical Co., St. Louis, MO), and the mixture was incubated at 37 C for 5 h. The reaction mixture was passed through Sep-Pak C18 cartridge previously activated with methanol and washed with 16% acetonitrile. Isoflavones were then recovered with 80% methanol. The elute was concentrated in a rotary evaporator with a 40 C water bath and redissolved in 16% acetonitrile solution and refrigerated until analysis.
2.7. Estrogen receptorb (ERb) mRNA quantification and ERb protein detection The presence of estrogen receptors in cells, including colon cells, is known to affect response to estrogen and phytoestrogens (English et al., 2001). To determine if difference in response to isoflavones by age were due to age-related differences in ERb, mRNA and protein levels in the colon were evaluated. Colonic ERb mRNA was quantified using real time reverse transcriptase-polymerase chain reaction (RT-PCR) using Estrogen Receptor Type 2 Assay-on-Demand kit (Applied Biosystems, Foster City, CA). ERb expression was measured relative to 18S rRNA expression. ERb protein was evaluated using immunohistochemistry. Formalin-fixed colon pieces were trimmed and embedded in paraffin using standard histological techniques. Antigen retrieval was performed using the pressure cooker method as described in the Trilogy Antigen Retrieval kit (Cell Marque, Hot Springs, AZ). Endogenous peroxidase was inactivated by incubating slides for 5 min in a 3% hydrogen peroxide solution. Sections were incubated for 30 min with an ERb polycolonal rabbit antibody diluted 1:200 in PBS (BioGenex, San Ramon, CA). ERb protein was detected using the LSAB2 kit, refined ABC method (DakoCytomation, Carpinteria, CA). Staining intensity and the distribution of staining within the tissue was compared among age groups visually by a pathologist (SFC) using light microscopy.
2.8. Quantifying normal colonic morphological structure Following fixation, two serial colon sections per animal were paraffinembedded and H&E stained (Washington Adventist Hospital Histology Laboratory, Takoma Park, MD). Visual examination of the structure of the colon sections illustrated that there appeared to be a difference in the density of normal crypts lining the mucosa among rats. A loss of normal crypts in the colon, or of villi in the intestine, may result from chronic gastrointestinal damage. To assess whether this difference was related to either age or diet, we devised a stereometric approach to quantitatively evaluate normal crypt density. Each colon section was digitally photographed using a Nikon Colorpix 995, 3.34 Mega Pixel digital camera through a microscope at 25· magnification and uploaded as a jpeg file on a standard PC. Using Image J software (National Institutes of Health, Bethesda MD) a 12,000-pixel grid containing two pixel lines was overlaid on the digital pictures. Only horizontal and vertical lines that were completely within the image of the mucosa were used for counting crypt
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637 density and the number of lines within this area was recorded. Crypts were only counted if they crossed these lines. Using these criteria, the number of crypt crossings per total number of lines was recorded for each section. This number was used to represent normal crypt density.
300 275 250 Grams
2.9. Statistical analyses
3.1. Diet composition The NovaSoy isoflavone isolate contained 515 mg/g isoflavones, consisting of the following specific isoflavones in mg/g (% total): genistin, 292.8 mg/g (57% total); daidzin, 182.7 mg/g (35% total); glycitin, 15.1 mg/g (2.9% total); acetylgenistin, 9.7 mg/g (1.9% total); acetyldaidzin, 6.4 mg/g (1.2% total); genistein, 4.3 mg/g (0.8% total); daidzein, 3.4 mg/g (0.7% total); and glycitein, 0.8 mg/g (0.2% total). Thus, the diet containing 4 g Novasoy isolate/kg diet had a concentration of soy isoflavones equal to 2.06 mg/g. 3.2. Animal health Within 96 h following AOM injections, three rats in the old rats fed soy group and one rat from the mature rats fed soy group were found dead or had to be euthanized. Histopathological evaluation of the tissues from these animals revealed changes in the liver, the kidney, the gastrointestinal tract, the adrenal gland and the spleen, which were characteristic of acute AOM toxicity. In addition, three of the four rats that died early had hepatocellular adenomas among other aging changes in the liver. 3.3. Animal body weights and food consumption By the end of the experiment, in all age groups the rats fed the soy diet had statistically significantly lower body weights than rats fed the control diet (Fig. 1). It was observed that older rats in the soy diet group lost three times as much weight and took twice as long to recover from weight loss as compared to old animals on the control diet (Table 1). Young rats consumed statistically significantly more food per kilogram body weight than mature or old rats; however, there were no statistically significant differences found between the diet groups at the time food consumption was measured. The young rats consumed 50.9 ± 2.5 g diet/kg body weight/day whereas the mature and old consumed 34.1 ± 2.4 and 38.6 ± 2.3 g diet/kg body weight/day, respectively. Based on this food intake and the amount of isoflavones in the diet, the estimated average
225 200 175 150 125
One-way analysis of variance (ANOVA) was performed for all data using the SAS System Version 8.0. Two-way ANOVA was also completed to investigate possible associative effects of diet and age for all data. The general linear models (GLM) procedure for unbalanced designs was used, due to the loss of animals during the experiment. Multiple comparison Student–Newman–Keuls test was performed when the ANOVA showed significance. Significance level was set at p < 0.05 for all analyses.
3. Results
631
100 75 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Week YC
YS
MC
MS
OC
OS
Fig. 1. Body weight of young, mature and old rats fed either the control or soy diets. The groups are: young control diet (YC), mature control diet (MC), old control diet (OC), young soy diet (YS), mature soy diet (MS), and old soy diet (OS) rats. At week 15, N = 7 for YC, MC, OC and YS, N = 6 for MS, N = 3 for OS.
Table 1 Effect of diet and age on time to regain body weight and weight loss following azoxymethane injection Group
n
Average days to begin gaining weight
Weight loss (g) from before injection to time when animals began gaining weight
Young control Young soy Mature control Mature soy Old control Old soy
7 7 7 6 7 4
0 ± 0.2a 0 ± 0.1 4 ± 0.2 6 ± 0.5 5 ± 0.6 11 ± 1.3
0.4 ± 0.2 4.6 ± 0.9 21.9 ± 2.3 39.1 ± 3.2 21.8 ± 2.8 62.7 ± 5.9
a
Means ± SE.
isoflavone intakes were 104 mg/kg for young, 70 mg/kg for mature and 79 mg/kg for old. 3.4. ACF quantification Young rats had a statistically significant greater number of ACF compared to mature and old rats (p = 0.0001, Fig. 2). The average total number of ACF was 53 ± 6 in
Fig. 2. Number of ACF in young, mature and old rats. Young rats had significantly more ACF compared to mature and old (p = 0.0001; N = 14 for young, 13 for mature, and 10 for old).
632
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637 Table 2 Serum levels of isoflavones in young, mature and old female rats fed soyisoflavone-supplemented diet
Fig. 3. Number of ACF in control and soy diet groups. There was no significant effect of diet on ACF number (N = 21 for control and 16 for soy diet groups).
Group
n
Isoflavone
lg isoflavone/ml serum
Young Mature Old Young Mature Old Young Mature Old
7 5 3 7 5 3 7 5 3
Diadzein Diadzein Diadzein Genistein Genistein Genistein Diadzein + Genistein Diadzein + Genistein Diadzein + Genistein
0.55 ± 0.06a 0.66 ± 0.07 0.74 ± 0.04 0.58 ± 0.056 0.70 ± 0.07 0.71 ± 0.09 1.13 ± 0.10 1.37 ± 0.14 1.45 ± 0.11
a Means ± SE. There were no statistically significant differences (p > 0.05) in serum isoflavone levels among age groups.
3.6. Serum isoflavones young animals, 26 ± 3 in mature, and 22 ± 5 in old rats. Specifically, younger animals had significantly more medium sized ACF (multiplicity of 4–5), 14 ± 1, compared to mature and old rats, with 4 ± 1 and 2 ± 1, respectively (p-value, 0.0001). There were no significant differences for small (multiplicity of 2–3) and large (multiplicity of 6 and over) ACF between age groups (data not shown). There was no statistically significant effect of diet on ACF number or ACF multiplicity (Fig. 3). 3.5. Serum estradiol concentration Serum estradiol concentration (pg/ml) decreased with age for animals on the control diet with old rats having a statistically significant decrease (p < 0.05) in serum estradiol concentrations compared to young (Fig. 4). Serum estradiol concentration increased with age for animals on the soy isoflavone diet with all three soy groups having statistically similar concentrations compared to young control rats (p > 0.05) (Fig. 4). Lack of cross-reactivity of the estradiol radioimmunoassay with isoflavones was confirmed with the isoflavone standards (data not shown).
Fig. 4. Average serum estradiol concentrations in young control diet (YC), mature control diet (MC), old control diet (OC), young soy diet (YS), mature soy diet (MS), and old soy diet (OS) rats. YC rats were found to have increased serum estradiol concentrations compared to MC and OC rats. Serum estradiol concentrations were significantly increased (p < 0.05) in MS and OS rats as compared to rats fed the control diet (N = 7 for YC, MC, OC and YS, N = 6 for MS, N = 3 for OS).
Isoflavones were not detected in serum from rats that were fed the AIN-93 control diet (data not shown). Although there were higher levels of genistin than diadizin in the diet, serum levels of genistein and daidzein were similar within each age group (Table 2). There was a trend for serum levels of both isoflavones to increase with age, but there were no statistically significant differences among age groups (Table 2). 3.7. ERb expression There were no statistically significant differences in ERb mRNA expression in the colonic mucosa among age or diet groups (data not shown). Following examination of immunhistochemically-stained colon sections, we found no differences in ERb protein staining intensity or in the distribution of ERb positively stained cells in colonic crypts (data not shown). 3.8. Uterine weights The ratios of uterine weight to body weight were similar among all age groups fed the control diet, but increased with age in rats fed the soy diet (Fig. 5).
Fig. 5. Average uterine weight (g)/body weight (g) ratios in young control diet (YC), mature control diet (MC), old control diet (OC), young soy diet (YS), mature soy diet (MS), and old soy diet (OS) rats. Old and mature rats fed soy isoflavones had slightly higher uterine/body weight ratios (p = 0.07; N = 7 for YC, MC, OC and YS, N = 6 for MS, N = 3 for OS).
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
633
3.9. Tissue histology Adverse histological effects attributable to the soy diet were not apparent in any tissues, including specifically the uterus, ovary or mammary gland. Changes noted were consistent with common aging changes in F344 rats, including dilation of the uterine lumen and/or uterine glandular dilation. Evidence of mild squamous metaplasia of scattered uterine glands was observed in uteri of mature and old animals independent of diet. One granular cell tumor was noted in the ovary of one old control rat. 3.10. Normal colonic crypt density Young rats had a greater density of the normal crypts lining the colonic mucosa as compared to the mature and old animals (p < 0.001, Fig. 6). Additionally, control fed rats had significantly more crypt density as compared to soy-fed animals (p < 0.001, Fig. 7). Furthermore the following differences in crypt morphology were observed: reduced numbers of epithelial cells, increased lamina propria between crypts with increased numbers of stromal cells indicative of crypt loss/crypt drop out (Fig. 8).
Fig. 8. Colonic crypt morphology illustrating normal colonic crypt density in mucosa of young (A) and decreased colonic crypt density in old (B) female rats. Magnification = 25·.
4. Discussion
Fig. 6. Effect of age on crypt density. The mean crypts per line were significantly higher in young rats compared to mature and old (p < 0.001; N = 14 for young, N = 13 for mature, N = 11 for old).
We initiated this study to assess the effect of age and soy isoflavones on the development of AOM-induced early lesions of colon cancer, ACF, in female rats, hypothesizing that phytoestrogens in soy may provide protection against colon cancer development. However, we report no evidence of a beneficial effect on the development of preneoplastic colonic aberrant crypt foci in female rats fed a soy isoflavone-supplement diet compared to rats fed the control diet, indicating that the soy isoflavone diet did not provide protection against early stage colonic carcinogenesis. On the contrary, several adverse effects of soy supplementation in mature and old female AOM-treated rats were observed. We believe this age- and diet-dependent increased sensitivity to azoxymethane is an important finding. Consequently, we investigated and will discuss below potential mechanisms for the severe acute toxicity experienced by older, soy fed, female rats. 4.1. Lack of beneficial effect of soy on ACF development
Fig. 7. Effect of diet on crypt density. The mean crypts per line were significantly higher in control fed rats compared to soy-fed rats (p < 0.001; N = 21 for control and 17 for soy).
Aberrant crypt foci (ACF) are pre-cancerous lesions, which can be chemically induced in rodents to assess the effect of dietary compounds on colon carcinogenesis (Bird, 1995). Animals possessing higher numbers of ACF with higher crypt multiplicity have an increased risk of developing colon cancer (Fenoglio-Preiser and Noffsinger, 1999; Magnuson et al., 1993). ACF have been found in humans and are considered to be an excellent biomarker for development of colon tumors (Corpet and Tache, 2002; Rafter et al., 2004).
634
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
As discussed in the introduction, reports on the effect of soy isoflavones, either in purified form or in combination with soy protein, on experimental colon cancer have not been consistent (Murillo et al., 2004; Thiagarajan et al., 1998; Ohuchi et al., 2005; Davies et al., 1999; Sorensen et al., 1998; Guo et al., 2004). In the apparently healthy young female rats, we report no evidence of a chemoprotective effect of the soy isoflavone diet, in agreement with the findings of Davies et al. (1999) and Sorensen et al. (1998). The potential chemoprotective effect of soy isoflavones in older rats is difficult to assess in this experiment due to acute toxicity and weight loss. 4.2. Lack of beneficial effect of estrogen on ACF development In two previous studies, we observed an increased number of ACF in old Sprague–Dawley female rats compared to young females (Magnuson et al., 2000) but the opposite result in male Fisher rats where young males had increased ACF numbers compared to old males (Kwon et al., 2004). Therefore, we hypothesized that the increase of ACF in older compared to young Sprague–Dawley female rats may have been due to an age-related decrease in naturally produced estrogen (Rossouw et al., 2002). Estrogen, in the form of HRT, decreases the risk of colon cancer in post-menopausal women (Campbell-Thompson et al., 2001; Gustafsson, 2003; Konstantinopoulos et al., 2003). We found a significant decrease in serum estradiol concentrations with age, indicating that this F344 female rat model successfully mimicked the natural decrease of estrogen that occurs in aging women. However, interestingly, young female rats had statistically significant more ACF compared to mature and old animals consistent with our previous findings in male F344 rats (Kwon et al., 2004), suggesting that age-related differences in ACF development may depend on rat strain, rather than gender. Another consideration is however, that differences in susceptibility to the acute and chronic effects of AOM among the three age groups of rats may have had an effect on ACF formation. Caderni et al. (2002) reported that while periods of fasting followed by refeeding promote AOM-induced ACF development in rats, a period of caloric restriction, or starvation inhibits colon tumor development. Moderate reduction in body weight gain due to treatments has also been observed to inhibit mammary tumor formation (Rodriguez-Burford et al., 2002). Although we found no differences in food intake during week six, we did not measure food consumption immediately following AOM injection. The mature and old rats may have had decreased food consumption during this time, which would have likely affected both the number of ACF, and the development of larger ACF. 4.3. Estrogenic responses to soy in older rats There have been conflicting reports on the effect of soy isoflavones on estrogenic responses in humans and animals
(Goldin et al., 2005; Harrison et al., 1999). We observed increased serum estrogen, increased serum isoflavones levels and increased uterine/body weight ratios in older, but not young rats fed the soy diet despite the fact that young rats consumed more isoflavones per kg body weight. These findings are in agreement with the reports of increased uterine weights in mature female rats fed a similar level of soy isoflavones for 3 months (Nakai et al., 2005a,b). These data suggest that changes occurring during aging may increase the estrogenic effect of soy isoflavones. ERb expression was measured in order to investigate any possible relationships among this predominant colonic estrogen receptor, ACF formation and/or aging. There were no significant differences in either ERb mRNA or protein expression among age groups indicating that changes in ERb expression are not related to ACF development in this animal model. However, we restricted our investigation to the expression of ERb in total colonic tissue rather than comparing ERb expression in normal colonic tissue to expression in individual ACF. In women, there is a down-regulation of ERb in colonic tumors compared to normal tissue (Konstantinopoulos et al., 2003). ERb expression in pre-cancerous lesions may differ from expression in normal surrounding tissue. It is also possible that estrogen and ERb play roles in colon cancer development during later stages of tumor formation, which occur after ACF development. Future studies may compare ERb expression in large ACF, colonic adenomas and adenocarcinomas to normal surrounding tissue. 4.4. Potential mechanisms of acute toxicity To begin the evaluation of possible mechanisms for the observed AOM/isoflavone sensitivity in the older rats, the dosages of AOM and soy isoflavones used in our study need to be considered. Several investigators have reported injecting young rats with the dose of 20 mg/kg of AOM used in this study (Caderni et al., 2002; Delker et al., 2000; Momen et al., 2002; Shimoji et al., 2003). When a dose of 15 mg/kg is used, this is usually administered twice, as previously reported by our laboratory in mature or older rats (Kwon et al., 2004; Magnuson et al., 2000). However, as older animals appeared more stressed during handling for injections, we elected to use the one 20 mg/kg injection protocol. The dietary isoflavone concentration used in this study was not intended to represent levels of isoflavones that may be achieved in the human diet through consumption of soy foods, but rather to investigate whether age would alter the effects of dietary supplements of isoflavones. The calculated doses of isoflavones that rats consumed in this study (ranging from 70 to 104 mg/kg body weight) were lower than the NOAEL of 120 mg/kg/day determined in a 28 day repeated dose study with genistein in male and female Crj:CD(SD)IGS rats (Okazaki et al., 2002). More recently, McClain et al. (2006) reported that in subchronic and chronic toxicity studies, genistein was well tolerated by male and female Wistar Crl:(WI)BR rats at doses up to
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
500 mg/kg/day. Therefore, the doses of AOM and isoflavones used in this study have been individually used previously without adverse effects, and it appears that it was the combination with aging changes that resulted in toxicity as young rats fed the isoflavones tolerated the AOM. Whether we would observe the described interaction at lower doses of AOM and/or lower doses of soy isoflavones needs to be determined. As we did not anticipate the severe acute toxicity associated with age and diet, food consumption was not measured immediately following AOM injections. In addition, serum isoflavone levels were only measured at the end of the study and not following AOM injections, as it was not possible to obtain sufficient blood for serum analysis without killing the young rats. The amount of serum needed to evaluate serum isoflavones was 0.5 ml. As plasma represents about 50–55% of blood, one needs approximately 2–2.5 times as much blood as the amount of plasma or serum needed. Therefore, we needed at least 1–1.25 ml blood for serum isoflavone analyses. The maximal volume of blood that can be safely collected from a rat without adverse effects is 10% of the total blood volume, which is 6% of the body weight (Weiss et al., 2000). As our young rats weighed between 80 and 120 g during the first two weeks of the experiment, the maximal amount of blood we could have collected was between 0.5 ml and 0.75 ml. This would not have provided sufficient serum for isoflavone analyses. Therefore, one limitation of this study is that the amount of isoflavones consumed or the serum isoflavone levels immediately following injections when the acute toxicity among mature and old rats was observed is unknown. In future experiments, food consumption and serum isoflavone levels should be measured immediately following administration of a carcinogen in case unexpected toxicity occurs. Acute toxic changes observed in the rats that died shortly after AOM injections affected mainly the liver and the gastrointestinal tract. The acute changes in the gastrointestinal tract following dimethylhydrazine and AOM injections have been well documented (Magnuson et al., 1994; Sunter et al., 1981). That AOM causes a variety of mucosal changes in grossly normal colonic tissue has also been reported (Pan et al., 1985). Older animals may be more susceptible to the acute side effects of AOM due to reduced repair capacities of the colonic tissue with age. Interestingly, the increased sensitivity to AOM of older animals was further exacerbated in rats fed the soy diet, suggesting that the soy isoflavones enhanced gastrointestinal damage by AOM. In addition, it is likely that the hepatocellular adenomas present in three of the rats that died early could have affected proper metabolism of AOM and in this way contributed to the increased sensitivity of these rats to AOM toxicity. We hypothesized that the acute cytotoxic effects of the AOM/soy combination possibly amplified morphological aging changes of the colon tissue. To test this hypothesis we conducted a detailed stereo-morphometric evaluation
635
of the colonic mucosa of all surviving animals 15 weeks after exposure to AOM. We report for the first time significant structural changes in the colonic epithelium of aged, AOM-treated female rats compared to young AOM-treated female rats. The changes included a statistically significant decrease in the crypt density per measuring unit and were characterized by reduced numbers of epithelial cells, increased lamina propria between crypts with increased numbers of stromal cells indicative of crypt loss/drop out. Young rats had a statistically significant increase in colonic crypt density as compared to the mature and old animals and control fed rats had a statistically significant increase in crypt density as compared to soy fed animals. Recently, Nakai et al. (2005b) observed the presence of extensive squamous metaplasia in the uterine gland of mature 3-month-old female Sprague–Dawley rats fed dietary soy protein supplemented with approximately 18 mg isoflavones/kg body weight for 3 months. These authors repeated the study using Fisher F344 rats, but did not observe any changes in the reproductive tract in these rats (Nakai et al., 2005a). Likewise, in this study a dose of 16 mg isoflavones/kg did not produce adverse effects, specifically squamous metaplasia, in the uterus, ovary or mammary glands of any rats. This observation is consistent with the hypothesis of Nakai et al. (2005a) of a strain variation regarding susceptibility of soy-related uterine glandular metaplasia. Changes in the uteri noted in this study were consistent with common aging changes in F344 rats (Brown and Leininger, 1992). The age, diet and gender combination used in our study revealed important interactions with the xenobiotic (azoxymethane) and soy isoflavones at doses well tolerated in young animals. These interactions resulted not only in increased immediate sensitivity but also in significant persistent morphological tissue alterations in the intestine, which in turn likely contributed to the reduced body weight gains and slower recovery from the original exposure in the older, soy-fed rats. This observation raises the question of the potential for adverse interactions between soy isoflavone supplements and other xenobiotics. In conclusion, we report an unexpected interaction between age, gender, dietary isoflavones and the acute effect of a colon carcinogen, azoxymethane resulting in increased immediate sensitivity of aged animals and significant persistent morphological mucosal changes. No beneficial effect of isoflavones on colonic ACF development was observed in any age group of female F344 rats. Increased estrogenic effects of isoflavones were observed with age. The adverse effects of soy isoflavones in aged female animals need further examination because women, and particularly older women, are the prime target population for consumption of soy supplements. Acknowledgements K.T. Daly and A.C. Tracy were Howard Hughes Undergraduate Research Fellows and received research
636
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
support from the Howard Hughes Medical Institute. The Joint Institute of Food Safety and Nutrition also supported this research through a research grant to B. Magnuson. The assistance of M. Giusti and F. Lin with the analysis of soy isoflavones and of T. Castonguay with the analysis of serum estrogen levels is gratefully acknowledged. References Bird, R.P., 1995. Role of aberrant crypt foci in understanding the pathogenesis of colon cancer. Cancer Lett. 93, 55–71. Bird, R.P., Good, C.K., 2000. The significance of aberrant crypt foci in understanding the pathogenesis of colon cancer. Toxicol. Lett. 112– 113, 395–402. Bird, R.P., McLellan, E.A., Bruce, W.R., 1989. Aberrant crypts, putative precancerous lesions, in the study of the role of diet in the aetiology of colon cancer. Cancer Surv. 8, 189–200. Brown, R.H., Leininger, J.R., 1992. Alterations of the uterus. In: Mohr, U., Dungworth, D.L., Capen, C.C. (Eds.), Pathobiology of the Aging Rat. International Life Sciences Institute Press, Washington, DC, p. 378. Caderni, G., Perrelli, M.G., Cecchini, F., Tessitore, L., 2002. Enhanced growth of colorectal aberrant crypt foci in fasted/refed rats involves changes in TGFbeta1 and p21CIP expressions. Carcinogenesis 23, 323–327. Campbell-Thompson, M., Lynch, I.J., Bhardwaj, B., 2001. Expression of estrogen receptor (ER) subtypes and ERbeta isoforms in colon cancer. Cancer Res. 61, 632–640. Coldham, N.G., Sauer, M.J., 2000. Pharmacokinetics of [(14)C]Genistein in the rat: gender-related differences, potential mechanisms of biological action, and implications for human health. Toxicol. Appl. Pharmacol. 164, 206–215. Corpet, D.E., Tache, S., 2002. Most effective colon cancer chemopreventive agents in rats: a systematic review of aberrant crypt foci and tumor data, ranked by potency. Nutr. Cancer 43, 1–21. Davies, M.J., Bowey, E.A., Adlercreutz, H., Rowland, I.R., Rumsby, P.C., 1999. Effects of soy or rye supplementation of high-fat diets on colon tumour development in azoxymethane-treated rats. Carcinogenesis 20, 927–931. Delker, D.A., Papanikolaou, A., Suhr, Y.J., Rosenberg, D.W., 2000. Diallyl sulfide enhances azoxymethane-induced preneoplasia in Fischer 344 rat colon. Chem. Biol. Interact. 124, 149–160. Demonty, I., Lamarche, B., Deshaies, Y., Jacques, H., 2002. Role of soy isoflavones in the hypotriglyceridemic effect of soy protein in the rat. J. Nutr. Biochem. 13, 671–677. Ealey, K.N., elSohemy, A., Archer, M.C., 2001. Conjugated linoleic acid does not inhibit development of aberrant crypt foci in colons of male Sprague–Dawley rats. Nutr. Cancer 41, 104–106. English, M.A., Stewart, P.M., Hewison, M., 2001. Estrogen metabolism and malignancy: analysis of the expression and function of 17betahydroxysteroid dehydrogenases in colonic cancer. Mol. Cell Endocrinol. 171, 53–60. Fenoglio-Preiser, C.M., Noffsinger, A., 1999. Aberrant crypt foci: a review. Toxicol. Pathol. 27, 632–642. Foley, E.F., Jazaeri, A.A., Shupnik, M.A., Jazaeri, O., Rice, L.W., 2000. Selective loss of estrogen receptor beta in malignant human colon. Cancer Res. 60, 245–248. Genazzani, A.R., Gadducci, A., Gambacciani, M., 2001. Controversial issues in climacteric medicine II Hormone replacement therapy and cancer. Maturitas 40, 117–130. Goldin, B.R., Brauner, E., Adlercreutz, H., Ausman, L.M., Lichtenstein, A.H., 2005. Hormonal response to diets high in soy or animal protein without and with isoflavones in moderately hypercholesterolemic subjects. Nutr. Cancer 51, 1–6. Guo, J.Y., Li, X., Browning Jr., J.D., Rottinghaus, G.E., Lubahn, D.B., Constantinou, A., Bennink, M., MacDonald, R.S., 2004. Dietary soy
isoflavones and estrone protect ovariectomized ERalphaKO and wildtype mice from carcinogen-induced colon cancer. J. Nutr. 134, 179– 182. Gustafsson, J.A., 2003. What pharmacologists can learn from recent advances in estrogen signalling. Trends Pharmacol. Sci. 24, 479–485. Harrison, R.M., Phillippi, P.P., Swan, K.F., Henson, M.C., 1999. Effect of genistein on steroid hormone production in the pregnant rhesus monkey. Proc. Soc. Exp. Biol. Med. 222, 78–84. Jemal, A., Murray, T., Ward, E., Samuels, A., Tiwari, R.C., Ghafoor, A., Feuer, E.J., Thun, M.J., 2005. Cancer statistics, 2005. CA Cancer J. Clin. 55, 10–30. Khare, S., Cerda, S., Wali, R.K., von Lintig, F.C., Tretiakova, M., Joseph, L., Stoiber, D., Cohen, G., Nimmagadda, K., Hart, J., Sitrin, M.D., Boss, G.R., Bissonnette, M., 2003. Ursodeoxycholic acid inhibits Ras mutations, wild-type Ras activation, and cyclooxygenase-2 expression in colon cancer. Cancer Res. 63, 3517–3523. Kok, L., Kreijkamp-Kaspers, S., Grobbee, D.E., Lampe, J.W., van der Schouw, Y.T., 2005. A randomized, placebo-controlled trial on the effects of soy protein containing isoflavones on quality of life in postmenopausal women. Menopause 12, 56–62. Konstantinopoulos, P.A., Kominea, A., Vandoros, G., Sykiotis, G.P., Andricopoulos, P., Varakis, I., Sotiropoulou-Bonikou, G., Papavassiliou, A.G., 2003. Oestrogen receptor beta (ERbeta) is abundantly expressed in normal colonic mucosa, but declines in colon adenocarcinoma paralleling the tumour’s dedifferentiation. Eur. J. Cancer 39, 1251–1258. Kwon, Y., Malik, M., Magnuson, B.A., 2004. Inhibition of colonic aberrant crypt foci by curcumin in rats is affected by age. Nutr. Cancer 48, 37–43. Lin, F., Giusti, M.M., 2005. Effects of solvent polarity and acidity on the extraction efficiency of isoflavones from soybeans (Glycine max). J. Agric. Food Chem. 53, 3795–3800. Magnuson, B.A., Carr, I., Bird, R.P., 1993. Ability of aberrant crypt foci characteristics to predict colonic tumor incidence in rats fed cholic acid. Cancer Res. 53, 4499–4504. Magnuson, B.A., Shirtliff, N., Bird, R.P., 1994. Resistance of aberrant crypt foci to apoptosis induced by azoxymethane in rats chronically fed cholic acid. Carcinogenesis 15, 1459–1462. Magnuson, B.A., South, E.H., Exon, J.H., Dashwood, R.H., Xu, M., Hendrix, K., Hubele, S., 2000. Increased susceptibility of adult rats to azoxymethane-induced aberrant crypt foci. Cancer Lett. 161, 185– 193. McClain, M.R., Wolz, E., Davidovich, A., Pfannkuch, F., Edwards, J.A., Bausch, J., 2006. Acute, subchronic and chronic safety studies with genistein in rats. Food Chem. Toxicol. 44, 56–80. Momen, M.A., Monden, Y., Houchi, H., Umemoto, A., 2002. Effect of ursodeoxycholic acid on azoxymethane-induced aberrant crypt foci formation in rat colon: in vitro potential role of intracellular Ca2+. J. Med. Invest. 49, 67–73. Morioka, T., Suzui, M., Nabandith, V., Inamine, M., Aniya, Y., Nakayama, T., Ichiba, T., Yoshimi, N., 2005. Modifying effects of Terminalia catappa on azoxymethane-induced colon carcinogenesis in male F344 rats. Eur. J. Cancer Prev. 14, 101–105. Moriya, M., Kato, K., Ohta, T., Watanabe, K., Watanabe, Y., Shirasu, Y., 1978. Detection of mutagenicity of the colon carcinogen 1,2dimethylhydrazine by the host-mediated assay and its correlation to carcinogenicity. J. Natl. Cancer Inst. 61, 457–460. Murillo, G., Choi, J.K., Pan, O., Constantinou, A.I., Mehta, R.G., 2004. Efficacy of garbanzo and soybean flour in suppression of aberrant crypt foci in the colons of CF-1 mice. Anticancer Res. 24, 3049– 3055. Nakai, M., Black, M., Jeffery, E.H., Bahr, J.M., 2005a. Dietary soy protein and isoflavones: no effect on the reproductive tract and minimal positive effect on bone resorption in the intact female Fischer 344 rat. Food Chem. Toxicol. 43, 945–949. Nakai, M., Cook, L., Pyter, L.M., Black, M., Sibona, J., Turner, R.T., Jeffery, E.H., Bahr, J.M., 2005b. Dietary soy protein and isoflavones have no significant effect on bone and a potentially negative effect on
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637 the uterus of sexually mature intact Sprague–Dawley female rats. Menopause 12, 291–298. Ohuchi, Y., Myojin, Y., Shimamoto, F., Kashimoto, N., Kamiya, K., Watanabe, H., 2005. Decrease in size of azoxymethane induced colon carcinoma in F344 rats by 180-day fermented miso. Oncol Rep. 14 (6), 1559–1564. Okazaki, K., Okazaki, S., Nakamura, H., Kitamura, Y., Hatayama, K., Wakabayashi, S., Tsuda, T., Katsumata, T., Nishikawa, A., Hirose, M., 2002. A repeated 28-day oral dose toxicity study of genistein in rats, based on the ‘Enhanced OECD Test Guideline 407’ for screening endocrine-disrupting chemicals. Arch. Toxicol. 76, 553–559. Pan, Q., Hamilton, S.R., Hyland, J., Boitnott, J.K., 1985. Effects of carcinogen dosage on experimental colonic carcinogenesis by azoxymethane: an ultrastructural study of grossly normal colonic mucosa. J. Natl. Cancer Inst. 74, 689–698. Rafter, J., Govers, M., Martel, P., Pannemans, D., Pool-Zobel, B., Rechkemmer, G., Rowland, I., Tuijtelaars, S., van Loo, J., 2004. PASSCLAIM–diet-related cancer. Eur. J. Nutr. 43 (Suppl 2), II47– II84. Rao, C.V., Wang, C.X., Simi, B., Lubet, R., Kelloff, G., Steele, V., Reddy, B.S., 1997. Enhancement of experimental colon cancer by genistein. Cancer Res. 57, 3717–3722. Rodriguez-Burford, C., Steele, V.E., Anderson, A.S., Stockard, C.R., Weiss, H.L., Eto, I., Johanning, G.L., Grizzle, W.E., Grubbs, C.J., 2002. Effects of body weight gain reduction resulting from chemopreventive agent treatment on mammary gland morphology. Nutr. Cancer 43, 67–75. Rossouw, J.E., Anderson, G.L., Prentice, R.L., LaCroix, A.Z., Kooperberg, C., Stefanick, M.L., Jackson, R.D., Beresford, S.A., Howard, B.V., Johnson, K.C., Kotchen, J.M., Ockene, J., 2002. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. JAMA 288, 321–333.
637
Shimoji, Y., Sugie, S., Kohno, H., Tanaka, T., Nanda, K., Tamura, Y., Nishikawa, Y., Hayashi, R., Uenakai, K., Ohigashi, H., 2003. Extract of vinegar ‘‘Kurosu’’ from unpolished rice inhibits the development of colonic aberrant crypt foci induced by azoxymethane. J. Exp. Clin. Cancer Res. 22, 591–597. Sorensen, I.K., Kristiansen, E., Mortensen, A., Nicolaisen, G.M., Wijnands, J.A., van Kranen, H.J., van Kreijl, C.F., 1998. The effect of soy isoflavones on the development of intestinal neoplasia in ApcMin mouse. Cancer Lett. 130, 217–225. Spector, D., Anthony, M., Alexander, D., Arab, L., 2003. Soy consumption and colorectal cancer. Nutr. Cancer 47, 1–12. Sunter, J.P., Appleton, D.R., Watson, A.J., 1981. Acute changes occurring in the intestinal mucosae of rats given a single injection of 1,2 dimethylhydrazine. Virchows Arch. B Cell. Pathol. Incl. Mol. Pathol. 36, 47–57. Suzuki, R., Kohno, H., Murakami, A., Koshimizu, K., Ohigashi, H., Yano, M., Tokuda, H., Nishino, H., Tanaka, T., 2004. Citrus nobiletin inhibits azoxymethane-induced large bowel carcinogenesis in rats. Biofactors 22, 111–114. Thiagarajan, D.G., Bennink, M.R., Bourquin, L.D., Kavas, F.A., 1998. Prevention of precancerous colonic lesions in rats by soy flakes, soy flour, genistein, and calcium. Am. J. Clin. Nutr. 68, 1394S–1399S. Wade, M.G., Lee, A., McMahon, A., Cooke, G., Curran, I., 2003. The influence of dietary isoflavone on the uterotrophic response in juvenile rats. Food Chem. Toxicol. 41, 1517–1525. Weiss, J., Taylor, G.R., Zimmermann, F., Nebendahl, K., 2000. Collection of body fluids. In: Krinke, G.J. (Ed.), The Handbook of Experimental Animals: The Laboratory Rat. Academic Press, San Diego, pp. 485–510. Zhou, J.R., Gugger, E.T., Tanaka, T., Guo, Y., Blackburn, G.L., Clinton, S.K., 1999. Soybean phytochemicals inhibit the growth of transplantable human prostate carcinoma and tumor angiogenesis in mice. J. Nutr. 129, 1628–1635.
Enhanced estrogenic responses and sensitivity to azoxymethane following dietary soy isoflavone supplementation in older female rats K.T. Daly a, A.C. Tracy a, M. Malik
a,d
, T. Wang b, S. Francke-Carroll c, B.A. Magnuson
a,*
a University of Maryland, College Park, MD, United States Phytonutrients Laboratory, USDA, Beltsville, MD, United States Pathology Branch, CFSAN, FDA, College Park, MD, United States d OBG, USUHS Bethesda, MD, United States
b c
Received 23 August 2005; accepted 22 October 2006
Abstract Soy isoflavones are popular supplements among middle-aged and older women based on their potential protection against cancer and their use as alternative hormone replacement therapy. The purpose of this study was to investigate the effects of dietary soy isoflavones on early stage colon cancer in various ages of female rats. Young (1 month), mature (11 month) and old (22 month) female Fisher 344 rats were fed either the control diet or a diet containing 0.4% soy isoflavone isolate for 1 week, injected once with 20 mg/kg azoxymethane (AOM) and maintained on the diets for another 15 weeks. The concentration of isoflavones in the diet was 2 g/kg diet, composed of 1.2 g/kg genistin, 0.7 g/kg daidzin and 0.1 g/kg other isoflavones including glycitin, acetylgenistin, acetyldaidzin, genistein, daidzein, and glycitein. There was no difference over all ages in the development of preneoplastic colonic aberrant crypt foci between rats fed the soy compared to the control diet, indicating that the soy diet did not provide protection against early stage colonic carcinogenesis. On the contrary, several adverse effects of soy supplementation in female AOM-treated rats were observed. Soy-supplemented rats had greater weight loss and a slower recovery of body weight following the AOM injection compared to rats fed the control diet and these changes increased with age. Five of the 21 rats fed the soy supplement died before the end of the experiment while all animals on the control diet survived to term. The density of normal crypts lining the colonic mucosa was reduced in rats fed the soy compared to control diet, indicating gastrointestinal damage. Uterine weights, serum estradiol and serum isoflavone levels were increased in mature and old female rats fed the soy-supplemented diets compared to age-matched controls, suggesting an increasing estrogenic response with age to isoflavone supplementation. These adverse effects of soy isoflavones in aged female animals need further examination because women, and particularly older women, are the prime target population for consumption of soy supplements. 2006 Elsevier Ltd. All rights reserved. Keywords: Soy isoflavones; Aberrant crypt foci; Gastrointestinal; Estradiol
1. Introduction Colon cancer is the second most common cause of cancer death in the United States and it is predicted that approximately 56,290 people will die each year from this disease (Jemal et al., 2005). Furthermore, among newly diagnosed cases, age is the primary risk factor with 90% of all colon cancer occurring in people over the age of *
Corresponding author. Tel.: +1 3014054523; fax: +1 3013143313. E-mail address: [email protected] (B.A. Magnuson).
0278-6915/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.fct.2006.10.021
50. Currently, incidences of colorectal cancer as well as the death rates are approximately equal in both men and women (Jemal et al., 2005). However, due to their increased susceptibility to chemical colon carcinogens, the majority of in vivo colon cancer studies are completed using male animals and limited research has been performed utilizing a female animal model (Moriya et al., 1978). In addition to the risk posed by increased age on colon cancer development, post-menopausal women also have decreased hormone levels, specifically estrogen. The steroid hormone, estrogen, has been shown to play a major role in
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
the development of cancer in women. For example, studies have shown that menopausal women taking hormone replacement therapy greatly decrease their risk of colon cancer suggesting that estrogen may play a role in preventing or protecting against colon cancer in aging women (Campbell-Thompson et al., 2001; Foley et al., 2000; Genazzani et al., 2001; Gustafsson, 2003; Jemal et al., 2005; Konstantinopoulos et al., 2003; Rossouw et al., 2002). The risk of developing colon cancer is also greatly affected by diet. Epidemiological studies suggest that soy foods and soy isoflavones may act to inhibit colon cancer (Spector et al., 2003). However, the effect of soy isoflavones, either in purified form or in combination with soy protein, on experimental colon cancer has not been consistent. Reduction of pre-cancerous lesions called aberrant crypt foci (ACF) (Bird et al., 1989) by soy protein isolate or soy flakes containing isoflavones has been reported (Murillo et al., 2004; Thiagarajan et al., 1998). A decrease in the size, but not incidence, of AOM-induced colon tumors in male rats consuming a diet containing 10% miso, a fermented soy product, was recently reported (Ohuchi et al., 2005). However, in other studies, dietary soy isoflavones either did not inhibit ACF or colonic tumors (Davies et al., 1999; Sorensen et al., 1998) or enhanced colon tumor incidence (Rao et al., 1997). All of these studies utilized male animals. Recently, Guo et al. (2004) reported that soy protein plus an isoflavone mix, but not either soy protein alone or soy plus genistein, reduced colon tumor incidence in young ovariectomized female mice. To our knowledge, no study has investigated the effect of soy isoflavones or estrogen on colon carcinogenesis in older female animals, having reduced estrogen levels due to aging. A recent study in our laboratory showed that young F344 male rats, injected subcutaneously with azoxymethane (AOM), a known colon carcinogen, and fed a semi-purified AIN-93 diet, had a statistically significant increase in ACF numbers and multiplicity compared to identically treated mature and old male rats (Kwon et al., 2004). In an earlier study, however, we found that old female Sprague–Dawley rats, injected intraperitoneally with AOM and fed a standard laboratory rat chow, had statistically significant increased ACF numbers and multiplicity compared to young and mature female rats (Magnuson et al., 2000). We hypothesized that the different results in these two studies may be attributed to animal gender; if estrogen has protective effects against early stages of colon cancer in female animals, then, as estrogen levels decrease with age, the older females would become more susceptible to colon carcinogens such as AOM. We believe that the effects of soy isoflavones on early stages of colon cancer need further investigation in an aging female animal model because women are taking increasing amounts of soy during and after menopause as an alternative to hormone replacement therapy (HRT) (Kok et al., 2005).
629
The purposes of this study were to determine the effect of age and soy isoflavone supplementation on ACF development in a female animal model. We choose F344 rats for this investigation because they are widely used in aging studies. We hypothesized that any effects of naturally produced estrogen on colon cancer development may be further enhanced by the phytoestrogens in soy. Therefore, we investigated the effect of a soy isoflavone-supplemented diet on ACF development in three age groups of female rats. Unexpectedly, we observed significant adverse effects in older rats fed the soy diet and subsequently, we evaluated the morphology of other tissues and the expression levels of estrogen receptor b. 2. Materials and methods 2.1. Study design and animal treatment With the approval of and in accordance with regulations of the United States Department of Agriculture Animal Use and Care Committee, 14 young (one month old), 14 mature (11 month old) and 14 old (22 month old) F344 female rats were obtained from the NIH aged animal colony (Bethesda, MD). They were housed singly in 8 · 8 · 12 in ONECAGETM cages with raised stainless steel wire bar floors and a ventilated EnviroGARD-B system under twelve hours reverse light/dark cycle. The rats were given free access to water and the AIN-93 powdered control diet, and allowed to acclimate for four days. Following this four-day acclimation period, rats from each age group were placed into either the AIN-93 control diet group or a soy isoflavone experimental diet group resulting in six groups with seven rats per group. The soy isoflavone diet was prepared by adding 4 g of Novasoy (Archer Daniels Midland, Decatur, Illinois) per kg AIN-93 powdered diet to create a final concentration of 1.9 mg isoflavone/g diet, based on reported concentration of 47.07% isoflavones in Novasoy. The amount of soy isoflavones added to the diet in this study was based on the effective dose used in a prostate chemopreventative study (Zhou et al., 1999). We opted to use the high end of the dose range for rat studies with isoflavones (Demonty et al., 2002; Nakai et al., 2005a,b; Wade et al., 2003) as female Wistar rats were reported to have lower plasma concentrations following dosing with genistein as compared to males (Coldham and Sauer, 2000), and because we thought it was possible that older rats would have a lower response due to lower estrogen and estrogen receptor levels. One week after being assigned to experimental diets, all rats were given one 20 mg/kg body weight s.c. injection of AOM (Sigma Chemicals, St. Louis, MO). We elected to use one injection at 20 mg/kg over the common protocol of two weekly injections of 15 mg/kg to reduce animal handling and animal stress as we observed that the older female rats tended to be more stressed by handling than the young female rats or the male older animals we have used in previous experiments. Dosing rats with 20 mg/kg of AOM is common practice and reported to be well tolerated (Caderni et al., 2002; Delker et al., 2000; Ealey et al., 2001; Khare et al., 2003; Morioka et al., 2005; Momen et al., 2002; Shimoji et al., 2003; Suzuki et al., 2004). Following the AOM injections, rats were weighed every day until body weights were restored and then weighed once a week for the remainder of the study. At the start of week six, the amount of food each animal consumed was measured over a three-day period.
2.2. Analysis of isoflavones Chemicals: Isoflavone standards (genistein, daidzein, glycitein, genistin, daidzin, glycitin, acetyldaidzin and acetylgenistin) for HPLC analyses were purchased from LC Laboratories (Woburn, Massachusetts). All of the isoflavone standards used were at least 99% pure, except for acetylgenistin, which was at least 98% pure. Acetonitrile, acetic acid, methanol, and flavone were HPLC grade reagents from Fisher Scientific (Fair Lawn,
630
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
NJ). Sodium acetate trihydrate was from J.T. Baker Inc. (Phillipsburg, NJ). Isoflavone extraction: Isoflavones were extracted as previously reported (Lin and Giusti, 2005) from 100 mg NovaSoy isoflavone concentrate using 10 ml of aqueous methanol solution (80%), in a 1:5 (w:v) ratio, and using flavone as internal standard. The sample was homogenized for 2 h at room temperature and vacuum filtered through Whatman No. 41 filter paper (Whatman International Ltd., Kent, England) using a Buchner funnel. The filtrate was evaporated using a rotary evaporator with a 40 C water bath. The extract was redissolved in 5 ml 16% aqueous acetonitrile solution and refrigerated for no more than 4 h before HPLC analysis. The extraction was conducted twice during the experiment, and each time the analyses were done in triplicate. HPLC analysis: Analysis of isoflavones was carried out using reversed-phase separation of the compounds on a C18 column (Waters Corp., Milford, MA) and using a linear gradient composed of acidified water (A: 0.1% acetic acid and 5% acetonitrile in water) and acidified acetonitrile (B: 0.1% acetic acid in acetonitrile) as previously reported (Lin and Giusti, 2005). After the injection of 25 ll of the extract sample, the linear gradient started from 10% B to 14% over 10 min, then increased to 20% over 2 min, maintained at 20% for 8 min, continued to increase to 70% over 10 min, maintained at 70% for 3 min, and returned to 10% at the end of the 34 min running time. The flow rate was maintained at 1 ml/ min throughout the running of the sample. A high-pressure liquid chromatography (Waters Delta 600 system) equipped with a photodiode array detector (Waters 996), autosampler (Waters 717plus), and Millennium32 software (Waters Corp.) was used. Elution was monitored at 254 nm, and spectral data from 200 to 450 nm were recorded and stored over the time of the run on all samples. Isoflavones were identified by comparing spectral data and retention times to those of pure standards. Calibration curves, prepared by using different concentrations of pure isoflavone standards and flavone (used as internal standard), were used for quantitation of the isoflavones in the samples. Quantification of isoflavones was based on the calibration curves and corrected with the peak area ratio of flavone and isoflavones.
2.3. Collection of tissues Fifteen weeks after AOM injections, rats were euthanized by CO2 inhalation. Blood was collected via cardiac puncture. Colons were removed and flushed with phosphate buffered saline (PBS). A 1 cm section was cut from the middle of each colon and fixed in 10% buffered formalin leaving two partial segments. These two segments were both opened via a longitudinal incision. A mucosal layer sample was collected by scraping a 3 cm section from the cecal end with a glass microscope slide. Scrapings were then fixed in RNAlater (Ambion, Austin, TX). The remaining colon sections were pinned flat on a balsa wood board and stored in formalin for at least 48 h. Uterine, ovary and mammary tissues were collected and also fixed in formalin. Liver and kidney tissues were removed, a section saved in formalin, and the remainder frozen in liquid nitrogen and stored at 80 C. All formalin-fixed tissues were subsequently embedded in paraffin, stained with hematoxylin and eosin (H&E) and evaluated for histopathological changes.
2.4. Identification of aberrant crypt foci Following fixation, the flattened colon tissues were stained with a 0.2% methylene blue/PBS solution. ACF were visualized by light microscopy and identified from normal crypts using a widely accepted set of criteria: they are larger, have an increased peri-cryptal area, have greater staining intensity due to a thickened layer of epithelial cells, are slightly raised above adjacent normal crypts, and have abnormally shaped lumina (Bird and Good, 2000). The total number of ACF and ACF multiplicity (number of crypts per focus) were quantified for each animal. ACF with a multiplicity of 2–3 were designated as small, 4–5 were designated as medium, and greater than five were designated as large. During quantifi-
cation, the scorer was not aware of corresponding animal numbers in order to eliminate bias.
2.5. Serum estrogen quantification To confirm that the older female rats had lower serum estrogen levels, estrogen levels were quantified. After clotting, blood samples were centrifuged for 30 min at 10,000 RPM. Serum was collected and stored at 20 C. Estradiol was measured in the serum of each animal using the Ultra-Sensitive Estradiol RadioImmunoassay (RIA) kit according to the manufacturer’s instructions (Diagnostic Systems Laboratories, Webster, TX).
2.6. Serum isoflavone quantification Serum samples (0.5 ml) were extracted as described above for the soy isolate, and redissolved in 0.3 ml aqueous acetonitrile. The isoflavone extract was mixed with 5 ml of 0.1 M sodium acetate buffer (pH 5.0) and 100 ll b-glucuronidase (from Helix pomatia, with 131,000 b-glucuronidase units/ml and 3.180 sulfatase units/ml, Sigma Chemical Co., St. Louis, MO), and the mixture was incubated at 37 C for 5 h. The reaction mixture was passed through Sep-Pak C18 cartridge previously activated with methanol and washed with 16% acetonitrile. Isoflavones were then recovered with 80% methanol. The elute was concentrated in a rotary evaporator with a 40 C water bath and redissolved in 16% acetonitrile solution and refrigerated until analysis.
2.7. Estrogen receptorb (ERb) mRNA quantification and ERb protein detection The presence of estrogen receptors in cells, including colon cells, is known to affect response to estrogen and phytoestrogens (English et al., 2001). To determine if difference in response to isoflavones by age were due to age-related differences in ERb, mRNA and protein levels in the colon were evaluated. Colonic ERb mRNA was quantified using real time reverse transcriptase-polymerase chain reaction (RT-PCR) using Estrogen Receptor Type 2 Assay-on-Demand kit (Applied Biosystems, Foster City, CA). ERb expression was measured relative to 18S rRNA expression. ERb protein was evaluated using immunohistochemistry. Formalin-fixed colon pieces were trimmed and embedded in paraffin using standard histological techniques. Antigen retrieval was performed using the pressure cooker method as described in the Trilogy Antigen Retrieval kit (Cell Marque, Hot Springs, AZ). Endogenous peroxidase was inactivated by incubating slides for 5 min in a 3% hydrogen peroxide solution. Sections were incubated for 30 min with an ERb polycolonal rabbit antibody diluted 1:200 in PBS (BioGenex, San Ramon, CA). ERb protein was detected using the LSAB2 kit, refined ABC method (DakoCytomation, Carpinteria, CA). Staining intensity and the distribution of staining within the tissue was compared among age groups visually by a pathologist (SFC) using light microscopy.
2.8. Quantifying normal colonic morphological structure Following fixation, two serial colon sections per animal were paraffinembedded and H&E stained (Washington Adventist Hospital Histology Laboratory, Takoma Park, MD). Visual examination of the structure of the colon sections illustrated that there appeared to be a difference in the density of normal crypts lining the mucosa among rats. A loss of normal crypts in the colon, or of villi in the intestine, may result from chronic gastrointestinal damage. To assess whether this difference was related to either age or diet, we devised a stereometric approach to quantitatively evaluate normal crypt density. Each colon section was digitally photographed using a Nikon Colorpix 995, 3.34 Mega Pixel digital camera through a microscope at 25· magnification and uploaded as a jpeg file on a standard PC. Using Image J software (National Institutes of Health, Bethesda MD) a 12,000-pixel grid containing two pixel lines was overlaid on the digital pictures. Only horizontal and vertical lines that were completely within the image of the mucosa were used for counting crypt
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637 density and the number of lines within this area was recorded. Crypts were only counted if they crossed these lines. Using these criteria, the number of crypt crossings per total number of lines was recorded for each section. This number was used to represent normal crypt density.
300 275 250 Grams
2.9. Statistical analyses
3.1. Diet composition The NovaSoy isoflavone isolate contained 515 mg/g isoflavones, consisting of the following specific isoflavones in mg/g (% total): genistin, 292.8 mg/g (57% total); daidzin, 182.7 mg/g (35% total); glycitin, 15.1 mg/g (2.9% total); acetylgenistin, 9.7 mg/g (1.9% total); acetyldaidzin, 6.4 mg/g (1.2% total); genistein, 4.3 mg/g (0.8% total); daidzein, 3.4 mg/g (0.7% total); and glycitein, 0.8 mg/g (0.2% total). Thus, the diet containing 4 g Novasoy isolate/kg diet had a concentration of soy isoflavones equal to 2.06 mg/g. 3.2. Animal health Within 96 h following AOM injections, three rats in the old rats fed soy group and one rat from the mature rats fed soy group were found dead or had to be euthanized. Histopathological evaluation of the tissues from these animals revealed changes in the liver, the kidney, the gastrointestinal tract, the adrenal gland and the spleen, which were characteristic of acute AOM toxicity. In addition, three of the four rats that died early had hepatocellular adenomas among other aging changes in the liver. 3.3. Animal body weights and food consumption By the end of the experiment, in all age groups the rats fed the soy diet had statistically significantly lower body weights than rats fed the control diet (Fig. 1). It was observed that older rats in the soy diet group lost three times as much weight and took twice as long to recover from weight loss as compared to old animals on the control diet (Table 1). Young rats consumed statistically significantly more food per kilogram body weight than mature or old rats; however, there were no statistically significant differences found between the diet groups at the time food consumption was measured. The young rats consumed 50.9 ± 2.5 g diet/kg body weight/day whereas the mature and old consumed 34.1 ± 2.4 and 38.6 ± 2.3 g diet/kg body weight/day, respectively. Based on this food intake and the amount of isoflavones in the diet, the estimated average
225 200 175 150 125
One-way analysis of variance (ANOVA) was performed for all data using the SAS System Version 8.0. Two-way ANOVA was also completed to investigate possible associative effects of diet and age for all data. The general linear models (GLM) procedure for unbalanced designs was used, due to the loss of animals during the experiment. Multiple comparison Student–Newman–Keuls test was performed when the ANOVA showed significance. Significance level was set at p < 0.05 for all analyses.
3. Results
631
100 75 0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Week YC
YS
MC
MS
OC
OS
Fig. 1. Body weight of young, mature and old rats fed either the control or soy diets. The groups are: young control diet (YC), mature control diet (MC), old control diet (OC), young soy diet (YS), mature soy diet (MS), and old soy diet (OS) rats. At week 15, N = 7 for YC, MC, OC and YS, N = 6 for MS, N = 3 for OS.
Table 1 Effect of diet and age on time to regain body weight and weight loss following azoxymethane injection Group
n
Average days to begin gaining weight
Weight loss (g) from before injection to time when animals began gaining weight
Young control Young soy Mature control Mature soy Old control Old soy
7 7 7 6 7 4
0 ± 0.2a 0 ± 0.1 4 ± 0.2 6 ± 0.5 5 ± 0.6 11 ± 1.3
0.4 ± 0.2 4.6 ± 0.9 21.9 ± 2.3 39.1 ± 3.2 21.8 ± 2.8 62.7 ± 5.9
a
Means ± SE.
isoflavone intakes were 104 mg/kg for young, 70 mg/kg for mature and 79 mg/kg for old. 3.4. ACF quantification Young rats had a statistically significant greater number of ACF compared to mature and old rats (p = 0.0001, Fig. 2). The average total number of ACF was 53 ± 6 in
Fig. 2. Number of ACF in young, mature and old rats. Young rats had significantly more ACF compared to mature and old (p = 0.0001; N = 14 for young, 13 for mature, and 10 for old).
632
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637 Table 2 Serum levels of isoflavones in young, mature and old female rats fed soyisoflavone-supplemented diet
Fig. 3. Number of ACF in control and soy diet groups. There was no significant effect of diet on ACF number (N = 21 for control and 16 for soy diet groups).
Group
n
Isoflavone
lg isoflavone/ml serum
Young Mature Old Young Mature Old Young Mature Old
7 5 3 7 5 3 7 5 3
Diadzein Diadzein Diadzein Genistein Genistein Genistein Diadzein + Genistein Diadzein + Genistein Diadzein + Genistein
0.55 ± 0.06a 0.66 ± 0.07 0.74 ± 0.04 0.58 ± 0.056 0.70 ± 0.07 0.71 ± 0.09 1.13 ± 0.10 1.37 ± 0.14 1.45 ± 0.11
a Means ± SE. There were no statistically significant differences (p > 0.05) in serum isoflavone levels among age groups.
3.6. Serum isoflavones young animals, 26 ± 3 in mature, and 22 ± 5 in old rats. Specifically, younger animals had significantly more medium sized ACF (multiplicity of 4–5), 14 ± 1, compared to mature and old rats, with 4 ± 1 and 2 ± 1, respectively (p-value, 0.0001). There were no significant differences for small (multiplicity of 2–3) and large (multiplicity of 6 and over) ACF between age groups (data not shown). There was no statistically significant effect of diet on ACF number or ACF multiplicity (Fig. 3). 3.5. Serum estradiol concentration Serum estradiol concentration (pg/ml) decreased with age for animals on the control diet with old rats having a statistically significant decrease (p < 0.05) in serum estradiol concentrations compared to young (Fig. 4). Serum estradiol concentration increased with age for animals on the soy isoflavone diet with all three soy groups having statistically similar concentrations compared to young control rats (p > 0.05) (Fig. 4). Lack of cross-reactivity of the estradiol radioimmunoassay with isoflavones was confirmed with the isoflavone standards (data not shown).
Fig. 4. Average serum estradiol concentrations in young control diet (YC), mature control diet (MC), old control diet (OC), young soy diet (YS), mature soy diet (MS), and old soy diet (OS) rats. YC rats were found to have increased serum estradiol concentrations compared to MC and OC rats. Serum estradiol concentrations were significantly increased (p < 0.05) in MS and OS rats as compared to rats fed the control diet (N = 7 for YC, MC, OC and YS, N = 6 for MS, N = 3 for OS).
Isoflavones were not detected in serum from rats that were fed the AIN-93 control diet (data not shown). Although there were higher levels of genistin than diadizin in the diet, serum levels of genistein and daidzein were similar within each age group (Table 2). There was a trend for serum levels of both isoflavones to increase with age, but there were no statistically significant differences among age groups (Table 2). 3.7. ERb expression There were no statistically significant differences in ERb mRNA expression in the colonic mucosa among age or diet groups (data not shown). Following examination of immunhistochemically-stained colon sections, we found no differences in ERb protein staining intensity or in the distribution of ERb positively stained cells in colonic crypts (data not shown). 3.8. Uterine weights The ratios of uterine weight to body weight were similar among all age groups fed the control diet, but increased with age in rats fed the soy diet (Fig. 5).
Fig. 5. Average uterine weight (g)/body weight (g) ratios in young control diet (YC), mature control diet (MC), old control diet (OC), young soy diet (YS), mature soy diet (MS), and old soy diet (OS) rats. Old and mature rats fed soy isoflavones had slightly higher uterine/body weight ratios (p = 0.07; N = 7 for YC, MC, OC and YS, N = 6 for MS, N = 3 for OS).
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
633
3.9. Tissue histology Adverse histological effects attributable to the soy diet were not apparent in any tissues, including specifically the uterus, ovary or mammary gland. Changes noted were consistent with common aging changes in F344 rats, including dilation of the uterine lumen and/or uterine glandular dilation. Evidence of mild squamous metaplasia of scattered uterine glands was observed in uteri of mature and old animals independent of diet. One granular cell tumor was noted in the ovary of one old control rat. 3.10. Normal colonic crypt density Young rats had a greater density of the normal crypts lining the colonic mucosa as compared to the mature and old animals (p < 0.001, Fig. 6). Additionally, control fed rats had significantly more crypt density as compared to soy-fed animals (p < 0.001, Fig. 7). Furthermore the following differences in crypt morphology were observed: reduced numbers of epithelial cells, increased lamina propria between crypts with increased numbers of stromal cells indicative of crypt loss/crypt drop out (Fig. 8).
Fig. 8. Colonic crypt morphology illustrating normal colonic crypt density in mucosa of young (A) and decreased colonic crypt density in old (B) female rats. Magnification = 25·.
4. Discussion
Fig. 6. Effect of age on crypt density. The mean crypts per line were significantly higher in young rats compared to mature and old (p < 0.001; N = 14 for young, N = 13 for mature, N = 11 for old).
We initiated this study to assess the effect of age and soy isoflavones on the development of AOM-induced early lesions of colon cancer, ACF, in female rats, hypothesizing that phytoestrogens in soy may provide protection against colon cancer development. However, we report no evidence of a beneficial effect on the development of preneoplastic colonic aberrant crypt foci in female rats fed a soy isoflavone-supplement diet compared to rats fed the control diet, indicating that the soy isoflavone diet did not provide protection against early stage colonic carcinogenesis. On the contrary, several adverse effects of soy supplementation in mature and old female AOM-treated rats were observed. We believe this age- and diet-dependent increased sensitivity to azoxymethane is an important finding. Consequently, we investigated and will discuss below potential mechanisms for the severe acute toxicity experienced by older, soy fed, female rats. 4.1. Lack of beneficial effect of soy on ACF development
Fig. 7. Effect of diet on crypt density. The mean crypts per line were significantly higher in control fed rats compared to soy-fed rats (p < 0.001; N = 21 for control and 17 for soy).
Aberrant crypt foci (ACF) are pre-cancerous lesions, which can be chemically induced in rodents to assess the effect of dietary compounds on colon carcinogenesis (Bird, 1995). Animals possessing higher numbers of ACF with higher crypt multiplicity have an increased risk of developing colon cancer (Fenoglio-Preiser and Noffsinger, 1999; Magnuson et al., 1993). ACF have been found in humans and are considered to be an excellent biomarker for development of colon tumors (Corpet and Tache, 2002; Rafter et al., 2004).
634
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
As discussed in the introduction, reports on the effect of soy isoflavones, either in purified form or in combination with soy protein, on experimental colon cancer have not been consistent (Murillo et al., 2004; Thiagarajan et al., 1998; Ohuchi et al., 2005; Davies et al., 1999; Sorensen et al., 1998; Guo et al., 2004). In the apparently healthy young female rats, we report no evidence of a chemoprotective effect of the soy isoflavone diet, in agreement with the findings of Davies et al. (1999) and Sorensen et al. (1998). The potential chemoprotective effect of soy isoflavones in older rats is difficult to assess in this experiment due to acute toxicity and weight loss. 4.2. Lack of beneficial effect of estrogen on ACF development In two previous studies, we observed an increased number of ACF in old Sprague–Dawley female rats compared to young females (Magnuson et al., 2000) but the opposite result in male Fisher rats where young males had increased ACF numbers compared to old males (Kwon et al., 2004). Therefore, we hypothesized that the increase of ACF in older compared to young Sprague–Dawley female rats may have been due to an age-related decrease in naturally produced estrogen (Rossouw et al., 2002). Estrogen, in the form of HRT, decreases the risk of colon cancer in post-menopausal women (Campbell-Thompson et al., 2001; Gustafsson, 2003; Konstantinopoulos et al., 2003). We found a significant decrease in serum estradiol concentrations with age, indicating that this F344 female rat model successfully mimicked the natural decrease of estrogen that occurs in aging women. However, interestingly, young female rats had statistically significant more ACF compared to mature and old animals consistent with our previous findings in male F344 rats (Kwon et al., 2004), suggesting that age-related differences in ACF development may depend on rat strain, rather than gender. Another consideration is however, that differences in susceptibility to the acute and chronic effects of AOM among the three age groups of rats may have had an effect on ACF formation. Caderni et al. (2002) reported that while periods of fasting followed by refeeding promote AOM-induced ACF development in rats, a period of caloric restriction, or starvation inhibits colon tumor development. Moderate reduction in body weight gain due to treatments has also been observed to inhibit mammary tumor formation (Rodriguez-Burford et al., 2002). Although we found no differences in food intake during week six, we did not measure food consumption immediately following AOM injection. The mature and old rats may have had decreased food consumption during this time, which would have likely affected both the number of ACF, and the development of larger ACF. 4.3. Estrogenic responses to soy in older rats There have been conflicting reports on the effect of soy isoflavones on estrogenic responses in humans and animals
(Goldin et al., 2005; Harrison et al., 1999). We observed increased serum estrogen, increased serum isoflavones levels and increased uterine/body weight ratios in older, but not young rats fed the soy diet despite the fact that young rats consumed more isoflavones per kg body weight. These findings are in agreement with the reports of increased uterine weights in mature female rats fed a similar level of soy isoflavones for 3 months (Nakai et al., 2005a,b). These data suggest that changes occurring during aging may increase the estrogenic effect of soy isoflavones. ERb expression was measured in order to investigate any possible relationships among this predominant colonic estrogen receptor, ACF formation and/or aging. There were no significant differences in either ERb mRNA or protein expression among age groups indicating that changes in ERb expression are not related to ACF development in this animal model. However, we restricted our investigation to the expression of ERb in total colonic tissue rather than comparing ERb expression in normal colonic tissue to expression in individual ACF. In women, there is a down-regulation of ERb in colonic tumors compared to normal tissue (Konstantinopoulos et al., 2003). ERb expression in pre-cancerous lesions may differ from expression in normal surrounding tissue. It is also possible that estrogen and ERb play roles in colon cancer development during later stages of tumor formation, which occur after ACF development. Future studies may compare ERb expression in large ACF, colonic adenomas and adenocarcinomas to normal surrounding tissue. 4.4. Potential mechanisms of acute toxicity To begin the evaluation of possible mechanisms for the observed AOM/isoflavone sensitivity in the older rats, the dosages of AOM and soy isoflavones used in our study need to be considered. Several investigators have reported injecting young rats with the dose of 20 mg/kg of AOM used in this study (Caderni et al., 2002; Delker et al., 2000; Momen et al., 2002; Shimoji et al., 2003). When a dose of 15 mg/kg is used, this is usually administered twice, as previously reported by our laboratory in mature or older rats (Kwon et al., 2004; Magnuson et al., 2000). However, as older animals appeared more stressed during handling for injections, we elected to use the one 20 mg/kg injection protocol. The dietary isoflavone concentration used in this study was not intended to represent levels of isoflavones that may be achieved in the human diet through consumption of soy foods, but rather to investigate whether age would alter the effects of dietary supplements of isoflavones. The calculated doses of isoflavones that rats consumed in this study (ranging from 70 to 104 mg/kg body weight) were lower than the NOAEL of 120 mg/kg/day determined in a 28 day repeated dose study with genistein in male and female Crj:CD(SD)IGS rats (Okazaki et al., 2002). More recently, McClain et al. (2006) reported that in subchronic and chronic toxicity studies, genistein was well tolerated by male and female Wistar Crl:(WI)BR rats at doses up to
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
500 mg/kg/day. Therefore, the doses of AOM and isoflavones used in this study have been individually used previously without adverse effects, and it appears that it was the combination with aging changes that resulted in toxicity as young rats fed the isoflavones tolerated the AOM. Whether we would observe the described interaction at lower doses of AOM and/or lower doses of soy isoflavones needs to be determined. As we did not anticipate the severe acute toxicity associated with age and diet, food consumption was not measured immediately following AOM injections. In addition, serum isoflavone levels were only measured at the end of the study and not following AOM injections, as it was not possible to obtain sufficient blood for serum analysis without killing the young rats. The amount of serum needed to evaluate serum isoflavones was 0.5 ml. As plasma represents about 50–55% of blood, one needs approximately 2–2.5 times as much blood as the amount of plasma or serum needed. Therefore, we needed at least 1–1.25 ml blood for serum isoflavone analyses. The maximal volume of blood that can be safely collected from a rat without adverse effects is 10% of the total blood volume, which is 6% of the body weight (Weiss et al., 2000). As our young rats weighed between 80 and 120 g during the first two weeks of the experiment, the maximal amount of blood we could have collected was between 0.5 ml and 0.75 ml. This would not have provided sufficient serum for isoflavone analyses. Therefore, one limitation of this study is that the amount of isoflavones consumed or the serum isoflavone levels immediately following injections when the acute toxicity among mature and old rats was observed is unknown. In future experiments, food consumption and serum isoflavone levels should be measured immediately following administration of a carcinogen in case unexpected toxicity occurs. Acute toxic changes observed in the rats that died shortly after AOM injections affected mainly the liver and the gastrointestinal tract. The acute changes in the gastrointestinal tract following dimethylhydrazine and AOM injections have been well documented (Magnuson et al., 1994; Sunter et al., 1981). That AOM causes a variety of mucosal changes in grossly normal colonic tissue has also been reported (Pan et al., 1985). Older animals may be more susceptible to the acute side effects of AOM due to reduced repair capacities of the colonic tissue with age. Interestingly, the increased sensitivity to AOM of older animals was further exacerbated in rats fed the soy diet, suggesting that the soy isoflavones enhanced gastrointestinal damage by AOM. In addition, it is likely that the hepatocellular adenomas present in three of the rats that died early could have affected proper metabolism of AOM and in this way contributed to the increased sensitivity of these rats to AOM toxicity. We hypothesized that the acute cytotoxic effects of the AOM/soy combination possibly amplified morphological aging changes of the colon tissue. To test this hypothesis we conducted a detailed stereo-morphometric evaluation
635
of the colonic mucosa of all surviving animals 15 weeks after exposure to AOM. We report for the first time significant structural changes in the colonic epithelium of aged, AOM-treated female rats compared to young AOM-treated female rats. The changes included a statistically significant decrease in the crypt density per measuring unit and were characterized by reduced numbers of epithelial cells, increased lamina propria between crypts with increased numbers of stromal cells indicative of crypt loss/drop out. Young rats had a statistically significant increase in colonic crypt density as compared to the mature and old animals and control fed rats had a statistically significant increase in crypt density as compared to soy fed animals. Recently, Nakai et al. (2005b) observed the presence of extensive squamous metaplasia in the uterine gland of mature 3-month-old female Sprague–Dawley rats fed dietary soy protein supplemented with approximately 18 mg isoflavones/kg body weight for 3 months. These authors repeated the study using Fisher F344 rats, but did not observe any changes in the reproductive tract in these rats (Nakai et al., 2005a). Likewise, in this study a dose of 16 mg isoflavones/kg did not produce adverse effects, specifically squamous metaplasia, in the uterus, ovary or mammary glands of any rats. This observation is consistent with the hypothesis of Nakai et al. (2005a) of a strain variation regarding susceptibility of soy-related uterine glandular metaplasia. Changes in the uteri noted in this study were consistent with common aging changes in F344 rats (Brown and Leininger, 1992). The age, diet and gender combination used in our study revealed important interactions with the xenobiotic (azoxymethane) and soy isoflavones at doses well tolerated in young animals. These interactions resulted not only in increased immediate sensitivity but also in significant persistent morphological tissue alterations in the intestine, which in turn likely contributed to the reduced body weight gains and slower recovery from the original exposure in the older, soy-fed rats. This observation raises the question of the potential for adverse interactions between soy isoflavone supplements and other xenobiotics. In conclusion, we report an unexpected interaction between age, gender, dietary isoflavones and the acute effect of a colon carcinogen, azoxymethane resulting in increased immediate sensitivity of aged animals and significant persistent morphological mucosal changes. No beneficial effect of isoflavones on colonic ACF development was observed in any age group of female F344 rats. Increased estrogenic effects of isoflavones were observed with age. The adverse effects of soy isoflavones in aged female animals need further examination because women, and particularly older women, are the prime target population for consumption of soy supplements. Acknowledgements K.T. Daly and A.C. Tracy were Howard Hughes Undergraduate Research Fellows and received research
636
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637
support from the Howard Hughes Medical Institute. The Joint Institute of Food Safety and Nutrition also supported this research through a research grant to B. Magnuson. The assistance of M. Giusti and F. Lin with the analysis of soy isoflavones and of T. Castonguay with the analysis of serum estrogen levels is gratefully acknowledged. References Bird, R.P., 1995. Role of aberrant crypt foci in understanding the pathogenesis of colon cancer. Cancer Lett. 93, 55–71. Bird, R.P., Good, C.K., 2000. The significance of aberrant crypt foci in understanding the pathogenesis of colon cancer. Toxicol. Lett. 112– 113, 395–402. Bird, R.P., McLellan, E.A., Bruce, W.R., 1989. Aberrant crypts, putative precancerous lesions, in the study of the role of diet in the aetiology of colon cancer. Cancer Surv. 8, 189–200. Brown, R.H., Leininger, J.R., 1992. Alterations of the uterus. In: Mohr, U., Dungworth, D.L., Capen, C.C. (Eds.), Pathobiology of the Aging Rat. International Life Sciences Institute Press, Washington, DC, p. 378. Caderni, G., Perrelli, M.G., Cecchini, F., Tessitore, L., 2002. Enhanced growth of colorectal aberrant crypt foci in fasted/refed rats involves changes in TGFbeta1 and p21CIP expressions. Carcinogenesis 23, 323–327. Campbell-Thompson, M., Lynch, I.J., Bhardwaj, B., 2001. Expression of estrogen receptor (ER) subtypes and ERbeta isoforms in colon cancer. Cancer Res. 61, 632–640. Coldham, N.G., Sauer, M.J., 2000. Pharmacokinetics of [(14)C]Genistein in the rat: gender-related differences, potential mechanisms of biological action, and implications for human health. Toxicol. Appl. Pharmacol. 164, 206–215. Corpet, D.E., Tache, S., 2002. Most effective colon cancer chemopreventive agents in rats: a systematic review of aberrant crypt foci and tumor data, ranked by potency. Nutr. Cancer 43, 1–21. Davies, M.J., Bowey, E.A., Adlercreutz, H., Rowland, I.R., Rumsby, P.C., 1999. Effects of soy or rye supplementation of high-fat diets on colon tumour development in azoxymethane-treated rats. Carcinogenesis 20, 927–931. Delker, D.A., Papanikolaou, A., Suhr, Y.J., Rosenberg, D.W., 2000. Diallyl sulfide enhances azoxymethane-induced preneoplasia in Fischer 344 rat colon. Chem. Biol. Interact. 124, 149–160. Demonty, I., Lamarche, B., Deshaies, Y., Jacques, H., 2002. Role of soy isoflavones in the hypotriglyceridemic effect of soy protein in the rat. J. Nutr. Biochem. 13, 671–677. Ealey, K.N., elSohemy, A., Archer, M.C., 2001. Conjugated linoleic acid does not inhibit development of aberrant crypt foci in colons of male Sprague–Dawley rats. Nutr. Cancer 41, 104–106. English, M.A., Stewart, P.M., Hewison, M., 2001. Estrogen metabolism and malignancy: analysis of the expression and function of 17betahydroxysteroid dehydrogenases in colonic cancer. Mol. Cell Endocrinol. 171, 53–60. Fenoglio-Preiser, C.M., Noffsinger, A., 1999. Aberrant crypt foci: a review. Toxicol. Pathol. 27, 632–642. Foley, E.F., Jazaeri, A.A., Shupnik, M.A., Jazaeri, O., Rice, L.W., 2000. Selective loss of estrogen receptor beta in malignant human colon. Cancer Res. 60, 245–248. Genazzani, A.R., Gadducci, A., Gambacciani, M., 2001. Controversial issues in climacteric medicine II Hormone replacement therapy and cancer. Maturitas 40, 117–130. Goldin, B.R., Brauner, E., Adlercreutz, H., Ausman, L.M., Lichtenstein, A.H., 2005. Hormonal response to diets high in soy or animal protein without and with isoflavones in moderately hypercholesterolemic subjects. Nutr. Cancer 51, 1–6. Guo, J.Y., Li, X., Browning Jr., J.D., Rottinghaus, G.E., Lubahn, D.B., Constantinou, A., Bennink, M., MacDonald, R.S., 2004. Dietary soy
isoflavones and estrone protect ovariectomized ERalphaKO and wildtype mice from carcinogen-induced colon cancer. J. Nutr. 134, 179– 182. Gustafsson, J.A., 2003. What pharmacologists can learn from recent advances in estrogen signalling. Trends Pharmacol. Sci. 24, 479–485. Harrison, R.M., Phillippi, P.P., Swan, K.F., Henson, M.C., 1999. Effect of genistein on steroid hormone production in the pregnant rhesus monkey. Proc. Soc. Exp. Biol. Med. 222, 78–84. Jemal, A., Murray, T., Ward, E., Samuels, A., Tiwari, R.C., Ghafoor, A., Feuer, E.J., Thun, M.J., 2005. Cancer statistics, 2005. CA Cancer J. Clin. 55, 10–30. Khare, S., Cerda, S., Wali, R.K., von Lintig, F.C., Tretiakova, M., Joseph, L., Stoiber, D., Cohen, G., Nimmagadda, K., Hart, J., Sitrin, M.D., Boss, G.R., Bissonnette, M., 2003. Ursodeoxycholic acid inhibits Ras mutations, wild-type Ras activation, and cyclooxygenase-2 expression in colon cancer. Cancer Res. 63, 3517–3523. Kok, L., Kreijkamp-Kaspers, S., Grobbee, D.E., Lampe, J.W., van der Schouw, Y.T., 2005. A randomized, placebo-controlled trial on the effects of soy protein containing isoflavones on quality of life in postmenopausal women. Menopause 12, 56–62. Konstantinopoulos, P.A., Kominea, A., Vandoros, G., Sykiotis, G.P., Andricopoulos, P., Varakis, I., Sotiropoulou-Bonikou, G., Papavassiliou, A.G., 2003. Oestrogen receptor beta (ERbeta) is abundantly expressed in normal colonic mucosa, but declines in colon adenocarcinoma paralleling the tumour’s dedifferentiation. Eur. J. Cancer 39, 1251–1258. Kwon, Y., Malik, M., Magnuson, B.A., 2004. Inhibition of colonic aberrant crypt foci by curcumin in rats is affected by age. Nutr. Cancer 48, 37–43. Lin, F., Giusti, M.M., 2005. Effects of solvent polarity and acidity on the extraction efficiency of isoflavones from soybeans (Glycine max). J. Agric. Food Chem. 53, 3795–3800. Magnuson, B.A., Carr, I., Bird, R.P., 1993. Ability of aberrant crypt foci characteristics to predict colonic tumor incidence in rats fed cholic acid. Cancer Res. 53, 4499–4504. Magnuson, B.A., Shirtliff, N., Bird, R.P., 1994. Resistance of aberrant crypt foci to apoptosis induced by azoxymethane in rats chronically fed cholic acid. Carcinogenesis 15, 1459–1462. Magnuson, B.A., South, E.H., Exon, J.H., Dashwood, R.H., Xu, M., Hendrix, K., Hubele, S., 2000. Increased susceptibility of adult rats to azoxymethane-induced aberrant crypt foci. Cancer Lett. 161, 185– 193. McClain, M.R., Wolz, E., Davidovich, A., Pfannkuch, F., Edwards, J.A., Bausch, J., 2006. Acute, subchronic and chronic safety studies with genistein in rats. Food Chem. Toxicol. 44, 56–80. Momen, M.A., Monden, Y., Houchi, H., Umemoto, A., 2002. Effect of ursodeoxycholic acid on azoxymethane-induced aberrant crypt foci formation in rat colon: in vitro potential role of intracellular Ca2+. J. Med. Invest. 49, 67–73. Morioka, T., Suzui, M., Nabandith, V., Inamine, M., Aniya, Y., Nakayama, T., Ichiba, T., Yoshimi, N., 2005. Modifying effects of Terminalia catappa on azoxymethane-induced colon carcinogenesis in male F344 rats. Eur. J. Cancer Prev. 14, 101–105. Moriya, M., Kato, K., Ohta, T., Watanabe, K., Watanabe, Y., Shirasu, Y., 1978. Detection of mutagenicity of the colon carcinogen 1,2dimethylhydrazine by the host-mediated assay and its correlation to carcinogenicity. J. Natl. Cancer Inst. 61, 457–460. Murillo, G., Choi, J.K., Pan, O., Constantinou, A.I., Mehta, R.G., 2004. Efficacy of garbanzo and soybean flour in suppression of aberrant crypt foci in the colons of CF-1 mice. Anticancer Res. 24, 3049– 3055. Nakai, M., Black, M., Jeffery, E.H., Bahr, J.M., 2005a. Dietary soy protein and isoflavones: no effect on the reproductive tract and minimal positive effect on bone resorption in the intact female Fischer 344 rat. Food Chem. Toxicol. 43, 945–949. Nakai, M., Cook, L., Pyter, L.M., Black, M., Sibona, J., Turner, R.T., Jeffery, E.H., Bahr, J.M., 2005b. Dietary soy protein and isoflavones have no significant effect on bone and a potentially negative effect on
K.T. Daly et al. / Food and Chemical Toxicology 45 (2007) 628–637 the uterus of sexually mature intact Sprague–Dawley female rats. Menopause 12, 291–298. Ohuchi, Y., Myojin, Y., Shimamoto, F., Kashimoto, N., Kamiya, K., Watanabe, H., 2005. Decrease in size of azoxymethane induced colon carcinoma in F344 rats by 180-day fermented miso. Oncol Rep. 14 (6), 1559–1564. Okazaki, K., Okazaki, S., Nakamura, H., Kitamura, Y., Hatayama, K., Wakabayashi, S., Tsuda, T., Katsumata, T., Nishikawa, A., Hirose, M., 2002. A repeated 28-day oral dose toxicity study of genistein in rats, based on the ‘Enhanced OECD Test Guideline 407’ for screening endocrine-disrupting chemicals. Arch. Toxicol. 76, 553–559. Pan, Q., Hamilton, S.R., Hyland, J., Boitnott, J.K., 1985. Effects of carcinogen dosage on experimental colonic carcinogenesis by azoxymethane: an ultrastructural study of grossly normal colonic mucosa. J. Natl. Cancer Inst. 74, 689–698. Rafter, J., Govers, M., Martel, P., Pannemans, D., Pool-Zobel, B., Rechkemmer, G., Rowland, I., Tuijtelaars, S., van Loo, J., 2004. PASSCLAIM–diet-related cancer. Eur. J. Nutr. 43 (Suppl 2), II47– II84. Rao, C.V., Wang, C.X., Simi, B., Lubet, R., Kelloff, G., Steele, V., Reddy, B.S., 1997. Enhancement of experimental colon cancer by genistein. Cancer Res. 57, 3717–3722. Rodriguez-Burford, C., Steele, V.E., Anderson, A.S., Stockard, C.R., Weiss, H.L., Eto, I., Johanning, G.L., Grizzle, W.E., Grubbs, C.J., 2002. Effects of body weight gain reduction resulting from chemopreventive agent treatment on mammary gland morphology. Nutr. Cancer 43, 67–75. Rossouw, J.E., Anderson, G.L., Prentice, R.L., LaCroix, A.Z., Kooperberg, C., Stefanick, M.L., Jackson, R.D., Beresford, S.A., Howard, B.V., Johnson, K.C., Kotchen, J.M., Ockene, J., 2002. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. JAMA 288, 321–333.
637
Shimoji, Y., Sugie, S., Kohno, H., Tanaka, T., Nanda, K., Tamura, Y., Nishikawa, Y., Hayashi, R., Uenakai, K., Ohigashi, H., 2003. Extract of vinegar ‘‘Kurosu’’ from unpolished rice inhibits the development of colonic aberrant crypt foci induced by azoxymethane. J. Exp. Clin. Cancer Res. 22, 591–597. Sorensen, I.K., Kristiansen, E., Mortensen, A., Nicolaisen, G.M., Wijnands, J.A., van Kranen, H.J., van Kreijl, C.F., 1998. The effect of soy isoflavones on the development of intestinal neoplasia in ApcMin mouse. Cancer Lett. 130, 217–225. Spector, D., Anthony, M., Alexander, D., Arab, L., 2003. Soy consumption and colorectal cancer. Nutr. Cancer 47, 1–12. Sunter, J.P., Appleton, D.R., Watson, A.J., 1981. Acute changes occurring in the intestinal mucosae of rats given a single injection of 1,2 dimethylhydrazine. Virchows Arch. B Cell. Pathol. Incl. Mol. Pathol. 36, 47–57. Suzuki, R., Kohno, H., Murakami, A., Koshimizu, K., Ohigashi, H., Yano, M., Tokuda, H., Nishino, H., Tanaka, T., 2004. Citrus nobiletin inhibits azoxymethane-induced large bowel carcinogenesis in rats. Biofactors 22, 111–114. Thiagarajan, D.G., Bennink, M.R., Bourquin, L.D., Kavas, F.A., 1998. Prevention of precancerous colonic lesions in rats by soy flakes, soy flour, genistein, and calcium. Am. J. Clin. Nutr. 68, 1394S–1399S. Wade, M.G., Lee, A., McMahon, A., Cooke, G., Curran, I., 2003. The influence of dietary isoflavone on the uterotrophic response in juvenile rats. Food Chem. Toxicol. 41, 1517–1525. Weiss, J., Taylor, G.R., Zimmermann, F., Nebendahl, K., 2000. Collection of body fluids. In: Krinke, G.J. (Ed.), The Handbook of Experimental Animals: The Laboratory Rat. Academic Press, San Diego, pp. 485–510. Zhou, J.R., Gugger, E.T., Tanaka, T., Guo, Y., Blackburn, G.L., Clinton, S.K., 1999. Soybean phytochemicals inhibit the growth of transplantable human prostate carcinoma and tumor angiogenesis in mice. J. Nutr. 129, 1628–1635.