点击显示 收起
1 From the Departments of Viticulture and Enology (SEE, CAB, WTJ, MRW, LC-R, EAM, ACC, AL, and JDE), Nutrition (G-SK and AJC), and Animal Science (AI-T), University of California, Davis, and the Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha (AK and SHH).
2 Supported by the US Department of Agriculture (Regional Research W-143); the California Agricultural Experiment Station (Hatch 2850); the University of California, Davis, Clinical Nutrition Research Unit (NIH P30 DK 35747); an Institutional Research Grant (IRG-205) awarded to the University of California, Davis, by the American Cancer Society; the Wine Institute; and the California Prune Board. MRW was funded by a National Institute of Environmental Health and Safety Fellowship (NIH 5T32E507059). 3 Address reprint requests to SE Ebeler, Department of Viticulture & Enology, One Shields Avenue, University of California, Davis, Davis, CA 95616. E-mail: seebeler{at}ucdavis.edu.
ABSTRACT
Background: Evidence exists that red wine, which contains a large array of polyphenols, is protective against cardiovascular disease and possibly cancer.
Objective: We tested the hypothesis that catechin, the major monomeric polyphenol in red wine, can delay tumor onset in transgenic mice that spontaneously develop tumors.
Design: Mice were fed a nutritionally complete amino acid–based diet supplemented with (+)-catechin (0–8 mmol/kg diet) or alcohol-free solids from red wine. Mice were examined daily; the age at which a first tumor appeared was recorded as the age at tumor onset. Plasma catechin and metabolite concentrations were quantified at the end of the study.
Results: Dietary catechin significantly delayed tumor onset; a positive, linear relation was observed between the age at tumor onset and either the amount of dietary catechin (r2 = 0.761, P < 0.001) or plasma catechin and metabolite concentrations (r2 = 0.408, P = 0.003). No significant effects on tumor onset were observed when mice consumed a diet supplemented with wine solids containing <0.22 mmol catechin/kg diet, whereas a previous study showed that wine solids with a similar total polyphenol concentration but containing 4 times more catechin significantly delayed tumor onset by 30 d compared with a control diet. The catechin composition of the wines is directly related to processing conditions during vinification.
Conclusions: Physiologic intakes of specific dietary polyphenols, such as catechin, may play an important role in cancer chemoprevention. Wines have different polyphenol concentrations and compositions; therefore, the overall health benefits of individual wines differ.
Key Words: Amino acid–based diet • transgenic mice • catechin • cancer • tumor onset • wine
INTRODUCTION
Numerous epidemiologic studies show that fruit- and vegetable-rich diets reduce the risk of many types of cancer (1–4). Although the nature of the bioprotective compounds is unclear, components such as ascorbic acid (vitamin C), carotenoids (ß-carotene and lycopene), -tocopherol (vitamin E), fiber, and trace minerals that are present in fruit and vegetables are often associated with a reduced cancer risk. The proposed roles of glucosinolates, plant sterols, saponins, terpenes, and phytoestrogens have also been evaluated.
There is a growing interest in dietary polyphenols and flavonoids because they have been shown to have numerous chemopreventive properties, both in vitro and in vivo (reviewed in 5). Not only do polyphenols have strong antioxidant activity, but they can also modify the activity of many metabolic enzymes, including those involved in eicosanoid synthesis, and they can prevent or halt the spread of nascent tumors by altering platelet aggregation and minimizing angiogenesis (6–13). However, the role of individual polyphenols in cancer prevention remains unclear, and there is inconclusive evidence about the specific nature by which polyphenolics exert their bioprotective properties in vivo.
Red wine is a rich source of dietary polyphenols, with concentrations as high as 2.5 mmol/L (14, 15). The actual polyphenol concentration depends on the grapes used, vinification and processing procedures, and storage conditions. Moderate wine consumption, as opposed to the consumption of other types of alcoholic beverages, has been associated with a decreased risk of cardiovascular disease and some types of cancer (3, 16). Animal studies have also indicated that the nonalcoholic components of red wine may have a tumor-preventive effect, both in a chemically induced model of carcinogenesis and in a transgenic model of tumor progression (17, 18).
These epidemiologic and experimental animal studies lead to the hypothesis that red wine contains nonalcoholic components that can delay the onset or incidence of tumors. However, the nature of the chemopreventive agents, the effective concentration range, and the mechanism of action are still unknown. We conducted 3 experiments to test the hypothesis that catechin, the major monomeric polyphenol in red wine, can delay tumor onset in HTLV-1 tax transgenic mice. These mice develop tumors spontaneously without treatment with a carcinogen. In the first 2 experiments, the age at tumor onset as a function of the dietary catechin content and plasma catechin concentrations were monitored. In the third experiment, the effects of vinification treatments on total and individual polyphenol concentrations in the wine were evaluated and discussed with respect to the overall health effects associated with wine consumption.
MATERIALS AND METHODS
Materials
The amino acid–based diet was a nutritionally complete dry powder purchased from Dyets Inc (no. 517802; Bethlehem, PA). Gallic acid was purchased from Nutrition Biochemical (Cleveland). Malvidin-3-glucoside (oenin chloride) was obtained from Indofine Chemical Co (Somerville, NJ). Taxifolin was purchased from Apin Chemicals, Ltd (Abingdon, United Kingdom). N,O-bis (trimethylsilyl)-trifluoroacetamide was purchased from Pierce (Rockford, IL). Acetonitrile (American Chemical Society grade), o-phosphoric acid, hydrochloric acid, ethyl acetate (HPLC grade), pyridine, and the tax primers were obtained from Fischer Scientific (Pittsburgh). The catechin metabolites 3'-O-methylcatechin (3'MC) and 4'-O-methylcatechin (4'MC) were synthesized as previously described (19). All other reagents and chemicals were purchased commercially at the highest grade available from Fluka/Sigma Chemical Co (St Louis).
Diets and experimental treatments
(+)-Catechin was incorporated into the standard amino acid–based diet at 0.0, 0.5, 1.0, or 2.0 mmol/kg diet in experiment 1 and at 0.0, 1.0, 2.0, 4.0, or 8.0 mmol/kg diet in experiment 2. In experiment 3, the wine was a 1994 Sangiovese (Sonoma County, CA) made at the Pilot Winery (University of California, Davis). The wine was made in 1-ton lots and fermented in 2000-L stainless steel fermenters at 24–27°C. One volume of the fermenting must was pumped over twice daily to ensure uniform fermentation rates. The wine was pressed after 7 d, and sulfur dioxide concentrations of 25 mg/L were maintained throughout subsequent processing. The wine was oak aged for 6 mo and then placed into a stainless steel tank (10°C) until bottled in September 1997. At the time of bottling, 30 mg SO2/L was added before screw-cap closures were attached to the bottles; the wine was stored at 10°C until used.
Wine solids were prepared by bubbling nitrogen through a bottle of the Sangiovese wine (750 mL) for 2 h at 20°C to remove some of the ethanol. The wine was then poured into stainless steel trays to a depth of 1 cm and lyophilized (model 50SRC; Virtis Co, Gardiner, NY). The residue was scraped from the trays, dissolved in 5 mL water for each 750 mL of the original wine, transferred to a plastic container that was purged with argon, and stored in a vacuum dessicator at 10°C until it was incorporated into the standard diet. The solids from 750 mL of the wine were incorporated into 1 kg standard diet by using a food processor (model DLC-7P; Cuisinart, Greenwich, CT) to yield a homogeneous mixture. The total phenol content of the diet was 1950 mg phenols/kg diet as gallic acid equivalents (11.46 mmol GAE/kg).
For all studies, small amounts of the supplemented diet (0.5–1 kg) were prepared at a time to ensure that it was fresh when presented to the mice. The bulk diets were stored in sealed plastic containers at 3°C until they were fed to the mice. Individual food cups were filled with fresh diet each day. Mice had free access to the diet, and the diet consumption was monitored daily by weighing the cups before and after they were refilled.
Transgenic mice
Mice carrying the HTLV-1 transactivator (tax1) gene in their germ line under control of its own long terminal repeat, the transcriptional regulatory region of the virus, were described previously (20). This strain was originally derived by microinjection of the long terminal repeat tax1 gene construct into fertilized eggs from superovulated CD1 females crossed with C57BL/6-DBA2 F1 males. Mice derived from this original founder line are maintained as a breeding colony at the University of California, Davis. A heterozygous line was maintained by crossing transgenic mice from each generation with nontransgenic CD-1 mice (Charles River, Wilmington, MA).
Mice were genotyped at 7–10 d of age. A small (1 mm) portion of the tail of each mouse was snipped, frozen on dry ice, and stored at -80°C until analyzed. In experiments 1 and 3, mice were genotyped as previously described (17). In experiment 2, a polymerase chain reaction (PCR) protocol was used. The use of tax-A and tax-B primers (5'GTC AGG GCC CAG ACT AAG GCT and 5'CTT CCC GGA GGT CTG AGC TTA TG, respectively), which anneal to an upstream sequence of the tax insert, provides an accurate and rapid determination of the mouse genotype. For the PCR protocol, DNA was isolated from the tail snips by incubation with digestion buffer [1X PCR buffer, 0.5% NP40 detergent (Amersham Biosciences Inc, Piscataway, NJ), 0.5% Tween (Amersham Biosciences Inc), and 0.2 µg/µL proteinase K] for 3 h at 55°C. The proteinase K was denatured by heating at 95°C for 10 min and then cooling to 15°C. The samples were centrifuged at 1000 x g for 5 min at 4°C, and 1 µL supernatant fluid was used for PCR. DNA was amplified for 33 cycles at 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s. Bands were separated (85V; 45 min) on 2% agarose gel and stained with ethidium bromide. Positive and negative tax controls were run simultaneously with the samples.
At weaning, mice were housed in individual stainless steel wire-bottom cages, and heterozygous transgenic mice from different litters were randomly assigned (within litter and sex) to the treatment groups. The mean weight was similar for each treatment group at the onset of each study. All mice were initially weaned onto the powdered amino acid diet (control diet) at 16–21 d of age. When the animals were 20–27 d of age, the control diet was replaced with the appropriate treatment diet. The mice were housed in a room with a temperature of 20–23°C, a relative humidity of 50%, and a12-h light, 12-h dark cycle.
The mice were weighed daily. The weights of the individual animals were graphed, and the slope of the line for each animal and the average growth rate for each treatment group were compared by regression analysis (EXCEL; Microsoft Corp, Seattle).
The mice were also examined daily for the appearance of a first externally visible tumor. These tumors appear on the snout, ears, tail, forelimbs, and hindlimbs. The date of tumor onset (tumor latency) for each individual animal was defined as the day when a first tumor had been positively identified on 3 consecutive d. Tumor latency was chosen as the endpoint rather than the alternative endpoint tumor multiplicity (number of tumors divided by the number of mice), because tumor multiplicity is typically defined from a specific date and some animals may not have a tumor at the selected time.
After tumor onset was confirmed, one female and one male mouse—both of which had been fed the highest catechin dose in experiment 2 (8 mmol/kg)—were bred with nontransgenic CD-1 mice. These mice reproduced successfully (12–16 pups/litter) and no apparent effects of long-term consumption of the catechin-supplemented diets were observed in the offspring.
All mice were killed and tissues were collected before wasting (weight loss associated with anorexia cachexia) became apparent and before the tumor burden was >1% of total body weight (cumulative tumor volume/body weight, assuming that each tumor was spherical). In general, the mice were killed within 60 d of the time when a first tumor appeared. This schedule was chosen to avoid complications of cachexia and tissue wasting.
The mice were killed with an overdose of carbon dioxide or ether and then bled by cardiac puncture with the use of a 1-cc tuberculin syringe. The whole blood was transferred to a 2-mL Vacutainer (Becton-Dickinson, Franklin Lakes, NJ) blood collection tube containing sodium heparin as an anticoagulant. The whole blood was centrifuged at 2000x g for 15 min at room temperature, and the plasma fraction was separated and frozen at -80°C. In experiment 2 only, the plasma was weighed and mixed with 25 µL phosphate-buffered ascorbic acid (PBA; 200 g ascorbic acid/L, 0.4 mol NaH2PO4/L, pH 3.6) before being frozen to prevent oxidation of the catechin and metabolites.
In addition to whole blood, the following tissues were obtained: heart, liver, kidney, spleen, muscle, brain, tumor, testes, and feet. The tissues were weighed, flash frozen in liquid nitrogen, and stored at -80°C. These tissues were not used as part of this study. The study was approved by the Institutional Animal Care and Use Committee at the University of California, Davis.
Plasma catechin and catechin metabolites
Catechin and the catechin metabolites 3'MC and 4'MC in the plasma from experiment 2 only were measured with the use of a modification of the method of Donovan et al (19). Because of limited sample availability, glucuronide and sulfate conjugates of catechin and metabolites were enzymatically hydrolyzed and quantified with the unconjugated forms for all samples. To prevent oxidation of the analytes, all solvents and reagents were deoxygenated by purging with nitrogen and kept on ice. Duplicate weighed aliquots of plasma (50–200 µL) with added PBA were thawed and brought to a total volume of 275 µL by adding 25 µL PBA and, if necessary, water. An aliquot (120 µL) of a solution of 0.6 mol CaCl2/L was also added to each sample. Taxifolin was used as an internal standard, and 25 µL of a 4-ng/µL (13.2-µmol/L) solution in PBA was added to all plasma samples to achieve a final concentration of 0.24 ng/µL (0.795 µmol/L). The plasma was incubated at 37°C in a shaking water bath for 45 min in nitrogen-flushed tubes containing 50 U sulfatase (S-3009; EC 3.1.6.1) and 1200 U ß-glucuronidase (G-0251; EC 3.2.1.31) dissolved in 120 µL water. After incubation, the plasma was placed in an ice bath and extracted with 2 mL ethyl acetate and 1 mL water by vortex mixing for 1 min. The sample was then centrifuged at 4500 x g for 10 min at 4°C, and the aqueous layer was extracted a second time with 1.5 mL ethyl acetate. The combined ethyl acetate extracts were passed through anhydrous sodium sulfate packed in a Pasteur pipette, dried under nitrogen gas, and then redissolved in 40 µL pyridine and derivatized with 60 µL N,-O-bis (trimethylsilyl)-trifluoroacetamide at 70°C for 2 h. The derivatized samples were analyzed by gas chromatography–mass spectrometry (GC-MS) on a model 6890 gas chromatograph (Hewlett-Packard, Palo Alto, CA) equipped with a 5972 quadrupole mass spectrometer with the use of a DB-5 capillary GC column (30 m x 0.25 mm internal diameter, 0.25-µm film thickness; J&W Scientific, Folsom, CA). Helium was the carrier gas at 35 cm/s; splitless injections of 2 µL were made and the column temperature was programmed from 100°C (3 min) to 260°C at 30°C/min and held at this temperature for 30 min. Major fragmentation ions for catechin [mass-to-charge ratio (m/z) = 355 and 368], 3'MC (m/z = 310 and 355), 4'MC (m/z = 310 and 280), and taxifolin (m/z = 368 and 296) were monitored in selective ion monitoring mode with the use of a dwell time of 100 ms/channel.
For quantitation and peak confirmation, pure compounds were purchased commercially or synthesized as previously described (19). Standard curves were prepared in 7% bovine serum albumin (BSA) in PBA that contained undetectable concentrations of catechin or metabolites. Catechin, 3'MC, and 4'MC were added to the blank plasma at final concentrations of 0, 17.2, 34.0, 84.4, 126.8, and 169.6 ng/mL (0–585 nmol catechin/L and 0–558 nmol 3'MC and 4'MC/L). Taxifolin was added to all plasma standards at a final concentration of 0.24 ng/µL (0.795 µmol/L). The plasma standards were incubated, extracted, and analyzed exactly as described for the samples. Calibration curves were established by plotting the ratio of the peak areas of the analyte (catechin, 3'MC, or 4'MC) to the internal standard (taxifolin) versus the amount of analyte added to the blank BSA solution. The standard curves for all compounds were linear and had average r2 values >0.95. The limit of detection (signalto-noise ratio = 3) was 0.1 ng (0.3 nmol) injected onto the column for all analytes. The standard curves were used to calculate catechin and metabolite concentrations in the plasma samples, and the measured concentrations were corrected for initial plasma weight to determine analyte concentrations in the original plasma samples.
Determination of wine polyphenol composition
The total polyphenol concentration of the wine was determined by Folin-Ciocalteu analysis (21). Gallic acid standards at concentrations of 0–500 mg/L (0–2.9 mmol/L) were prepared in 10% (by vol) ethanol and used to prepare the standard curve. Absorbance at 765 nm was measured with a Spectronic 1201 spectrophotometer (Milton-Roy, Rochester, NY), and the total phenol concentration of the wine sample was determined by linear regression. Results are expressed as mg GAE/L or mmol GAE/L.
Malvidin-3-glucoside, caffeic acid, quercetin, myricetin, gallic acid, catechin, epicatechin, and resveratrol in the wine were measured with an HPLC system (model HP 1090) fitted with an ultraviolet-visible diode array detector and CHEMSTATION integration software (Hewlett-Packard). The separation was performed with a reversed-phase LiChrospher 100 RP18e (250 mm x 5 mm internal diameter, 4-µm particle size column; Merck KGaA, Darmstadt, Germany). A binary solvent gradient was used (22): mobile phase A consisted of 200 mmol o-phosphoric acid/L with a pH adjusted to 1.5 with concentrated hydrochloric acid; mobile phase B consisted of 80% (by vol) acetonitrile and 20% (by vol) mobile phase A. The gradient consisted of increasing the percentage of acetonitrile in solvent A from 0% at the start of the analysis to 8% at 8 min, 14% at 20 min, 16.5% at 25 min, 21.5% at 35 min, and 50% at 70 min, followed by a 10-min hold. The solvent flow rate was 0.5 mL/min, and the total run time was 80 min. Wavelengths of 280, 316, 365, and 520 nm were monitored throughout the analysis.
Mixed standard solutions of malvidin-3-glucoside, caffeic acid, quercetin, myricetin, epicatechin, resveratrol, catechin, and gallic acid were prepared (Table 1). The concentration of each polyphenol in the wine was calculated by linear regression by using the peak area measured at 280 nm.
View this table:
TABLE 1 . Standards used for quantitation of polyphenols in wine: experiment 3
Data analysis
Means (±SDs or SEMs) were calculated for all response variables. Differences in the age at first tumor onset between mice fed the different treatments were evaluated by analysis of variance and Bonferroni-Dunn means-comparison tests (STATVIEW; Abacus Concepts, Berkeley, CA). Dose-response relations were evaluated by using regression analysis (STATVIEW). The data for experiments 1 and 2 were first analyzed separately. The data from both experiments were then combined to determine whether there was a significant experiment-by-diet group interaction for the age at tumor onset. In the absence of a significant interaction, the experimental effect could be controlled for and the overall effect (main effect) of the dietary treatments on the age at tumor onset could be tested.
RESULTS
The age at tumor onset was significantly delayed in mice fed catechin-supplemented diets in a linear dose-response relation up to 4 mmol/kg diet (experiments 1 and 2; Figure 1). Although the latest time at which a first tumor appeared was greater in experiment 1 than in experiment 2, no significant experiment-by-diet group interaction effect on the age at tumor onset was observed. Compared with the control diet, 4 mmol catechin/kg diet delayed the mean age at tumor onset by 32 d—a 45% extension of the tumor-free period. The highest concentration of catechin (8 mmol/kg diet) did not result in a further delay in tumor onset (Figure 1).
FIGURE 1. . Correlation between dietary catechin levels and age at first tumor onset in transgenic mice for experiment 1 (y = 76.37 + 42.06x; r2 = 0.513, P = 0.001) and experiment 2 (y = 73.2 + 8.36x; r2 = 0.761, P < 0.001). Note that the axes of the 2 panels are different.
Measurable concentrations of catechin and 3'MC were observed in mice fed the catechin-supplemented diets (Table 2). The metabolite 4'MC was present in most samples, but the concentrations were generally below the limit of quantitation. Total plasma catechin (catechin + metabolites) and 3'MC concentrations increased with increasing amounts of dietary catechin. A significant correlation was observed between plasma 3'MC concentrations and age at first tumor onset for individual mice (Figure 2). Similar results were observed for total catechin (catechin + metabolites) concentrations in plasma. Because the 3'MC metabolite makes up 84% of total plasma catechin concentrations (range: 74.0–89.5%), only data for the metabolite are shown.
View this table:
TABLE 2 . Plasma catechin and metabolite concentrations in mice fed the control and catechin-supplemented diets: experiment 21
FIGURE 2. . Correlation between age at tumor onset and plasma 3'O-methylcatechin concentrations for individual transgenic mice in experiment 2. y = 81.30 + 5.76x; r2 = 0.408, P = 0.003.
When the mice were fed a diet supplemented with the ethanol-free residue of red wine (wine solids; experiment 3), no significant delay in tumor onset was observed compared with the mice fed the control diet (81.5 ± 6.4 and 78.1 ± 4.0 d, respectively; P > 0.05). Catechin and gallic acid were the predominant polyphenols in the wine, at concentrations of 85 mg/L each (293 and 500 µmol/L, respectively). When solids from 750 mL of this wine were added to 1 kg diet, the resulting catechin concentration was 63.8 mg catechin/kg diet (223 µmol/kg), which was below the catechin concentration that was shown to significantly delay tumor onset in experiments 1 and 2. Concentrations of the other polyphenols in the wine in decreasing order of concentration were as follows: epicatechin, 48 mg/L (166 µmol/L); caffeic acid, 18 mg/L (100 µmol/L); quercetin, 17 mg/L (56 µmol/L); malvidin-3glucoside, 8 mg/L (16 µmol/L); myricetin, 6 mg/L (19 µmol/L); and resveratrol, 2 mg/L (8.8 µmol/L). The total polyphenol content of the wine used in this experiment was 2741 ± 227 mg GAE/L (16 mmol GAE/L).
Initial body weight, mean growth rates (rate of body weight gain), and final body weight of mice fed the catechin or wine-solids supplemented diets were not significantly different from those of sibling mice fed the control diet (Figure 3). The mice grew well and remained healthy throughout the study, even when consuming 8 mmol catechin/kg diet. Diet consumption in all treatment groups averaged 5 g/d and was not significantly different between the various treatments.
FIGURE 3. . Growth of transgenic mice fed a nutritionally adequate amino acid–based diet (control) or the same diet supplemented with catechin (experiment 2) or the solids from 750 mL red table wine/kg diet (wine solids) (experiment 3). Growth curves for experiment 1 were similar to those of experiment 2; only data from experiment 2 with the higher levels of catechin supplementation are shown.
DISCUSSION
Epidemiologic, animal, and in vitro studies have consistently linked the consumption of fruit and vegetables with a decreased risk of many degenerative diseases, including cancer (1–4). However, establishing the exact nature of the chemopreventive compounds and their mechanism of action has proven elusive.
Many available transgenic animal models offer promise for evaluating the molecular events involved in cancer initiation, promotion, and progression (23–28). Transgenic models generally provide kinetics of tumor development that are well defined and consistent because all animals with a given gene modification will develop tumors of a specific type. In addition, animals with a known predisposition to neoplasia can be readily compared with sibling controls that do not contain the gene modification. Transgenic models are not complicated by possibly confounding metabolic changes that can occur with the application of high doses of carcinogens, which are commonly used in chemical carcinogen models. Finally, in the case of the HTLV-1 model described here, the neoplasms develop externally and can be easily viewed and assessed. This allows tumor growth to be monitored over several time points, and the effects of experimental variables on tumor growth and progression can be easily monitored.
Because many of the dietary components with chemopreventive and biological activities are present in trace amounts in natural food sources, it is critical that dietary protocols for studying chemopreventive nutrients be rigorously controlled (27). In the present study we used an amino acid–based diet in which the composition was well defined (29) and that promotes normal mouse growth and reproduction (17, 29–31). Pure compounds (eg, catechin) or food fractions (eg, wine solids) can be added to the diet in known concentrations and the bioavailability and biological activity of the test compounds can be readily evaluated. This was shown in the current, ie, mice that consumed added catechin or wine solids in the diet were healthy, grew well, and reproduced successfully (Figure 3). Plasma catechin concentrations (free catechin and 3'MC) increased with increasing amounts of added dietary catechin, indicating that the catechin was absorbed from the diet (Table 2).
The major catechin metabolite in the plasma was 3'MC, a finding that was consistent with the results of previous studies that showed this to be the major methylated metabolite in humans and rats (19, 32–34). In plasma, catechin and methylcatechin metabolites occur predominantly as glucuronide and sulfonate conjugates (19, 32–34). In the current study, limited sample availability required that all conjugates be hydrolyzed before analysis, thus it was not possible to determine whether the extent of conjugation was constant for all doses of catechin. Hackett et al (35) observed that the extent of enzymatic conjugation can vary with the flavonoid dose, and Kim et al (36) showed that the long-term consumption of polyphenols from green tea (28 d) can alter the bioavailability of the individual flavonoids in rats and mice. Therefore, further studies are needed to fully evaluate the absorption and elimination of catechin and metabolites at doses that represent a range of normal dietary intakes over long periods of time.
We observed that catechin, when incorporated into a nutritionally complete amino acid–based diet, significantly delayed tumor onset in a transgenic mouse model of neurofibromatosis. The delay in onset varied in a dose-dependent manner at a range of 0.5–4 mmol catechin/kg diet (Figure 1) and with plasma catechin concentrations (Figure 2). In a previous study we observed that alcohol-free red wine solids, of which catechin was the major monomeric polyphenol (1.2 mmol/L wine added to the diet to yield 0.90 mmol/kg diet), also significantly delayed tumor onset in this same model (17). In contrast, wine solids that contained a similar concentration of total phenols, but very low catechin concentrations (0.22 mmol catechin/kg diet), did not have a significant effect on tumor onset when added to the diet. Our results suggest that specific dietary phenols, rather than total polyphenol concentrations, may be important for cancer prevention and that the flavonoid catechin may have important biological effects in vivo.
Our results are consistent with those of a recent study showing that catechin can inhibit intestinal tumor formation in the transgenic min/+ (multiple intestinal neoplasia) mouse (37). Similarly, Suganuma et al (38) showed that the structurally related flavonoid (-)-epigallocatechin gallate from green tea can inhibit tumor formation in the min/+ model. Related studies with breast, prostate, hepatoma, colorectal, and oral squamous cell cultures also showed growth-inhibitory effects of structurally related catechins from red wine and green tea (39–44).
Some human epidemiologic studies have also shown that wine may contain components that can decrease the risk of some cancers. For example, consumption of distilled spirits has been shown to increase the risk of many types of cancer, including those of the upper gastrointestinal tract, whereas consumption of wine results in no increased risk or even a slightly decreased risk of these cancers (16, 45–48). At the same time, the relation between alcohol and wine consumption and breast cancer risk is unclear and remains controversial (49–53). Therefore, further studies are needed to better understand the mechanisms involved in cancer development and prevention in humans.
Our results also suggest that winemaking conditions that affect the polyphenol composition can have an effect on the overall health benefits of the resulting wine. Although the wine used in the current study was matched to provide total phenol concentrations equivalent to those we used previously (17), catechin concentrations differed by a factor of 4. No significant differences in other monomeric phenols were observed between the 2 wines; concentrations of polymeric phenols and tannins were not determined and it is not known how they may have differed between the 2 wines. Our results do not indicate whether synergistic or additive effects between different polyphenols or between polyphenols and other dietary components may influence the overall health benefits.
The predominant source of catechin in the grape berry is from seed extraction; consequently, wines undergoing extended maceration would be expected to have higher concentrations of phenolic compounds. The wine used previously underwent an extended maceration whereby grape must was exposed to seeds for 6 wk after completion of the primary fermentation (17). In contrast, the wine used in the current study was only in contact with the skins and seeds for 7 d, resulting in an overall lower extraction of catechin into the final wine. These results indicate that vinification methods can have a significant effect on the overall health benefits of wine.
In summary, we showed that cancer onset due to a genetic predisposition can be affected by environmental factors such as diet. Catechin, at 0.5–4 mmol/kg diet, was absorbed by mice, had no adverse health effects, and delayed tumor onset in a linear dose-dependent manner in a transgenic animal model of neurofibromatosis. Plasma concentrations of total catechin and the metabolite 3'MC increased with increasing catechin concentrations in the diet and were positively correlated with the delay in time of tumor onset. Concentrations of specific dietary polyphenols, such as catechin, may play a more important role in cancer prevention than does the total polyphenol concentration in the diet. Processing conditions can have a significant effect on the polyphenol composition of many foods and beverages, including red wine. Therefore, further studies are needed to explore the complex interrelations between food processing, dietary polyphenol composition, cancer chemopreventive activity, and cancer risk in humans.
ACKNOWLEDGMENTS
We thank Jennifer Donovan for technical advice and for providing the analytic standards.
REFERENCES