The Journal of Nutrition Vol. 128 No. 3 March 1998, pp. 615-623

Chronic Exposure to Secoisolariciresinol Diglycoside Alters Lignan Disposition in Rats^1,2,3

Sharon E. Rickard and Lilian U. Thompson⁴

Department of Nutritional Sciences, University of Toronto, Toronto, Ontario, Canada M5S 3E2

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Mammalian lignans from colonic bacterial action on secoisolariciresinol diglycoside (SDG) may mediate the anticarcinogenic effect of prolonged SDG feeding in rats. To elucidate lignan bioactivity, our objective was to determine ³H-SDG disposition in rats with acute or chronic SDG treatment over 48 h. After food deprivation overnight, female Sprague-Dawley rats (70-72 d old) were given a single gavage of ³H-SDG (3.7 kBq/g body weight) either immediately (acute, n = 12) or after 10 d of gavage with 1.5 mg unlabeled SDG/d (chronic, n = 12). Rats were killed at 12, 24, 36 and 48 h after gavage, and samples collected and analyzed for radioactivity by liquid scintillation counting. Radioactivity was 1- to 16-fold higher at 12 vs. 48 h for tissues, blood and gastrointestinal contents (P < 0.05). By 48 h, >80% of the recovered dose was excreted in both groups (feces > 50%, urine = 28-32%). Tissue radioactivity was highest (by 0.5- to 176-fold) in the cecum (P < 0.05). Levels in the liver, kidney and uterus (12 h) were 0.2- to 7.5-fold higher than in other nongastrointestinal tissues. At 12 h, fecal radioactivity was negligible, and cecal content, liver and adipose radioactivity were one- to threefold greater in rats with chronic SDG exposure than in those acutely exposed (P < 0.05). Blood radioactivity, present mostly in the plasma fraction (0.4% of dose), suggested that lignan concentrations could be 3000 times higher than peak estrogen levels in rats. Thus, SDG metabolites accumulated in the liver, kidney, intestinal tissues and uterus. Chronic SDG exposure delayed fecal excretion while increasing liver and adipose lignan levels.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Nearly 20 years ago, mammalian lignans, primarily enterolactone (EL)⁵ and enterodiol (ED), were first discovered in various biological samples of humans and animals (Axelson and Setchell 1981, Dehennin et al. 1982, Setchell et al. 1980 and 1981a, Stich et al. 1980). Secoisolariciresinol diglycoside (SDG) is a plant lignan present in high levels in flaxseed (Axelson et al. 1982, Obermeyer et al. 1995) and a major precursor of mammalian lignans. The SDG is metabolized to ED, which is then converted to EL by the action of colonic bacteria (Axelson and Setchell 1981, Borriello et al. 1985, Setchell et al. 1981b). Once formed, ED and EL undergo enterohepatic circulation, eventually being excreted in the urine mainly as glucuronide conjugates (Axelson and Setchell 1981).

Much evidence accumulated to date suggests that consuming a diet containing mammalian lignan precursors would have a protective effect against cancer. Higher urinary lignan levels have been found in humans and animals at lower risk of developing cancer (Adlercreutz et al. 1992b). The mammalian lignan EL is more biologically active than ED in vitro (Martin et al. 1996, Sathyamoorthy et al. 1994, Wang et al. 1994). Cytotoxic effects have been observed with EL in the estrogen-sensitive MCF-7 and ZR-75-1 breast cancer cell lines (Hirano et al. 1990, Mousavi and Adlercreutz 1992). In addition, EL moderately inhibits in vitro angiogenesis, the generation of new capillaries necessary for aggressive tumor growth (Fotsis et al. 1993).

By exhibiting weak estrogenic/antiestrogenic activities, lignans also appear to antagonize the metabolism and biological effects of estrogen, a known promoter of mammary tumorigenesis (Miller and O'Neill 1990). Depression of estrogen-stimulated RNA synthesis in the rat uterus was observed when EL was given 22 h before estradiol (Waters and Knowler 1982). Both ED and EL have exhibited dose-dependent inhibition of estradiol binding to the type II rat uterine estrogen receptor (Adlercreutz et al. 1992a) and of estrone binding to rat -fetoprotein, an oncofetal plasma protein with high affinity for steroid estrogens (Garreau et al. 1991). The inhibition of the enzyme aromatase by EL may be a potential mechanism whereby lignans deny tumors an endogenous source of estrogen (Adlercreutz et al. 1993, Wang et al. 1994). In addition, EL has been shown to stimulate sex hormone binding globulin synthesis in HepG2 human liver cancer cells, which would reduce the level of free estrogen available (Adlercreutz et al. 1992a).

Of a variety of plant-based foods tested, flaxseed is the highest producer of mammalian lignans both in vitro (Thompson et al. 1991) and in vivo (Axelson et al. 1982). For this reason, flaxseed has been used as a model to test the effects of lignans on carcinogenesis. Flaxseed supplementation to a high fat diet significantly reduced the epithelial cell proliferation and aberrant crypt foci formation in the colon (Serraino and Thompson 1992a) and epithelial cell proliferation and nuclear aberration in the mammary gland of rats (Serraino and Thompson 1991). Mammary tumor size of dimethylbenz[a]anthracene (DMBA)-treated animals was shown to be significantly reduced at both early and late promotional stages of carcinogenesis with flaxseed feeding (Serraino and Thompson 1992b, Thompson et al. 1996a). With the purified SDG isolated from flaxseed (Rickard et al. 1996), it was demonstrated that lignans are in part responsible for the anticancer effect of flaxseed. Significant reductions in mammary tumor number and size (Thompson et al. 1996a and 1996b) and in the size and multiplicity of colonic aberrant crypts (Jenab and Thompson 1996) were observed with long-term feeding of SDG.

Despite the observed positive effect of lignans, few studies have examined the metabolic fate of lignans in vivo and how this plays a role in their mechanisms of action. Urinary lignan levels stabilize after 1 wk of continuous consumption of SDG (Rickard 1994). Rickard et al. (1996) found that urinary lignan excretion is linearly related to dose from 0 to 1.5 mg SDG/d with no further increase to 3.0 mg SDG/d in rats. However, because fecal lignans were not measured, it is not known if this plateau was related to increased fecal excretion of lignan metabolites or increased tissue retention. Acute as opposed to prolonged exposure to a lignan precursor such as SDG may affect its disposition (i.e., metabolism, distribution and elimination), which may in turn affect the actions of lignans on tumor growth. Consumption of 25-g raw flaxseed for 8 d resulted in significantly higher plasma and urinary lignans in comparison to a 1-d consumption in premenopausal females (Nesbitt 1997), suggesting enhanced absorption and higher tissue exposure with chronic exposure. Increased rates of absorption of the structurally similar isoflavones genistein and daidzein have been observed with chronic intakes of soybean milk (Lu et al. 1996).

Thus, the specific objectives of this experiment were to determine the disposition of ³H-SDG metabolites in rats over a 48-h period and whether this differed with acute vs. chronic SDG treatment. It is hypothesized that SDG metabolites will accumulate in tissues that are estrogen sensitive, in light of the anticancer effects of SDG in the mammary gland (Thompson et al. 1996a and 1996b) as well as evidence of estrogen receptor binding in tissues such as the uterus (Adlercreutz et al. 1992b, Waters and Knowler 1982). Because of the production of lignans by intestinal bacteria and expected liver conjugation of lignans with glucuronide (Axelson and Setchell 1981), it is anticipated that the liver and intestinal tissues will have high concentrations of lignan metabolites. In addition, chronic treatment with SDG is expected to increase tissue levels of lignan metabolites as a result of increased exposure from higher plasma lignan levels. And finally, it is hypothesized that chronic SDG consumption will increase urinary excretion and decrease fecal excretion.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Preparation of ³H-SDG radioisotope.  The SDG was extracted from defatted flaxseed by dioxane/ethanol (1:1, v/v) and the extract purified using silica gel columns as described by Rickard et al. (1996). The benzyl methylenes of the purified SDG were labeled with tritium by Amersham International (Little Chalfont, Buckinghamshire, UK) using a gas exchange method resulting in a specific radioactivity of 999 GBq/mmol and a radiochemical purity of 98.5% (Fig. 1). The benzyl methylene groups were chosen because they are present in the known metabolites ED and EL and tritium is more stable at these sites in comparison to the benzene ring. Tritium labeling has previously been used in metabolic studies of many compounds, including the structurally similar antiestrogens tamoxifen and toremifene (Simberg et al. 1990, Sipila et al. 1990). The ³H-SDG was stored in a solution of water/ethanol (1:1, v/v = 8.6 mol/L ethanol) in 74 MBq aliquots (37 GBq/L) at 20°C until use. One aliquot of ³H-SDG was diluted 50 times with 8.6 mol/L ethanol to form a stock solution of 740 MBq/L and stored at 20°C in 10-mL aliquots.

View larger version (15K): [in this window] [in a new window]   Fig 1. Structure of 3H-secoisolariciresinol diglycoside.

Animals and diets.  Twenty-four female Sprague-Dawley rats (age 56 d, n = 12; age 64-65 d, n = 12; Charles River Canada, Montreal, Canada) were housed in pairs in plastic cages, with sawdust bedding, in a room with a 12-h light:dark cycle at 22-24°C with 50% humidity. Animal care and use conformed to the Guide to the Care and Use of Experimental Animals (Canadian Council on Animal Care 1984), and the experimental protocol was approved by the University of Toronto Animal Care Committee. The basal diet, prepared by Dyets (Bethlehem, PA), was based on the AIN-93G formulation (Reeves et al. 1993) except that a higher fat concentration (200 g/kg soybean oil) was used at the expense of cornstarch and dextrose (Table 1). Fresh diet, stored at 4°C, was provided every 2 d.

Experimental design.  After 5-6 d of adaptation to the AIN-93G basal diet and environment, older rats were deprived of food overnight and then given a single gavage of ³H-SDG [3.7 kBq/g body weight (BW)] (acute group, n = 12). The younger rats were similarly treated after a 10-d period of daily gavaging with 1.5 mg unlabeled SDG dissolved in 1 mL of distilled water (chronic group, n = 12). Thus, at the time of administration of the radioisotope, rats in both the acute and chronic groups were 70-72 d of age and 200-250 g in weight. The average total dose given per rat was 835 ± 9 kBq (or 835 ± 9 pmol equivalents of ³H-SDG). The syringe was rinsed with 1 mL distilled water, which was then administered to the rat to ensure that all radioactivity was delivered. Rats were transferred to metabolic cages for the collection of urine and feces. At 12, 24, 36 and 48 h postgavage, rats were killed by CO₂ asphyxiation for a total of three rats per time point. The following samples were collected, weighed (or volumes recorded) and stored at 20°C until analysis: trunk blood, urine, feces, gastrointestinal tissues and contents, mammary gland, adipose tissue, skin, muscle and all major organs. Total weight of skin and muscle was estimated as 40 and 25% of BW, respectively (Mainland 1945, Rikimaru et al. 1989). Weight estimates of gastrointestinal contents were obtained by weighing tissues before and after removal of contents. Tissues were rinsed with saline to remove any residual contents or blood. Before storage at 20°C, an aliquot of blood (0.5 mL) was removed and the residual sample centrifuged (1000 × g, 10 min, 25°C) to obtain the plasma and RBC components. Total blood and plasma volumes of the animal were estimated using the value of 64.1 and 40.4 mL/kg BW, respectively (Baker et al. 1980). The metabolic cages were rinsed with a minimal amount of water and the washings collected to reduce the loss of residual urine radioactivity. Experiments were started at the same time in the morning (0900-0930 h) to minimize effects of diurnal variation. During the observation period, the diet and water were freely available.

Liquid scintillation counting.  Radioactivity of all collected tissues, blood, urine, feces and gastrointestinal contents was measured with the LKB Wallac 1217 Rackbeta liquid scintillation counter (Fisher Scientific, Ottawa, Canada) with a mean counting efficiency of 63% for tritium (based on unquenched standards). Corrections for chemical and color quenching of samples were done automatically by the liquid scintillation counter by using the sample channels ratio technique. The ratio of counts for the sample in each channel, i.e., section of the energy spectrum for tritium, varies linearly with efficiency for quenched solutions (Evans 1974). Extensive color quenching of samples was avoided by bleaching colored samples with hydrogen peroxide (described below). Samples were counted for 10-min periods.

Analytical procedures described below for blood, urine, feces, gastrointestinal contents and tissues were based on the methods described by Sakamoto et al. (1993) and by ICN Biomedicals, Radiochemical Division (Irvine, CA). Distilled water (1 mL) and ³H-SDG counting controls were included in each sample run. The ³H-SDG stock solution was diluted with 8.6 mol/L ethanol (1:20, v/v) for the counting controls. The ³H-SDG counting control consisted of distilled water (blood, 40 µL; tissues, 190 µL; urine, 990 µL; and feces and gastrointestinal contents, 90 µL) spiked with 10 µL of the counting control solution (total radioactivity = 370 Bq). To account for possible losses in radioactivity due to heating or other chemicals, ³H-SDG counting controls were processed in the same manner as the respective samples. All samples were analyzed in duplicate except for the ovaries because of their small size. Detected radioactivity in samples was converted to picomole equivalents of ³H-SDG by dividing by the specific activity of the radioisotope. Values were then adjusted to 100% recovery to allow for appropriate comparisons between time points. Mean recoveries of administered radioactivity over the 48 h period were the following: 1) acute group, 80.0 ± 9.3% (12 h), 71.5 ± 3.9% (24 h), 65.3 ± 10.1% (36 h) and 62.9 ± 4.6% (48 h); 2) chronic group, 84.0 ± 8.1% (12 h), 73.4 ± 5.4% (24 h), 79.6 ± 4.9% (36 h) and 70.6 ± 4.1% (48 h).

In addition to accounting for radioactivity losses during processing, the counting efficiencies for each sample were also determined. Counting efficiencies were calculated by spiking both the sample aliquots from a nonradioactive rat and the reagents alone with 370 Bq of ³H-SDG followed by processing as described above. Unspiked samples typically had radioactivity levels <100 Bq. The counting efficiency of a particular sample was then determined by dividing the spiked sample activity by the spiked reagent activity. Mean counting efficiencies were determined to be 31% for tissues containing blood (heart, liver, kidney, spleen and lung), 63% for all other tissues, 79% for GI contents and feces, and 77% for blood samples.

Analysis of whole blood and its components.  Whole blood, plasma or red blood cells (50 µL) were placed into 20-mL glass scintillation vials with plastic-lined caps (Fisher Scientific), and 300 µL of hyamine hydroxide (ICN Biomedicals, Aurora, OH) in ethanol (1:2, v/v) was added. Vials were capped tightly and incubated in a shaking water bath at 55-60°C for 1 h. Immediately after the incubation, 0.5 mL of 9.7 mol/L hydrogen peroxide was added to prevent color quenching and the vials incubated, loosely capped, for 30 min at 55-60°C. After the vials were cooled to room temperature, 15 mL of the aqueous cocktail scintillator Cytoscint ES (ICN Biomedicals) was added and the solution mixed vigorously. To eliminate chemiluminescence, 0.5 mL of 0.5 mol/L glacial acetic acid was added and the vials were shaken and stored at room temperature in the dark for 4 d before counting.

Urine analysis.  Aliquots of urine (100 µL) were placed directly into glass scintillation vials and diluted with 900 µL of distilled water. Cytoscint ES (15 mL) was added and the vials capped and shaken vigorously before counting.

Analysis of feces and gastrointestinal contents.  Samples were homogenized with 30 mL distilled water using a Polytron homogenizer (Brinkman Instruments Canada, Mississauga, Canada). Aliquots (100 µL) were placed in glass scintillation vials with the addition of 1 mL of Cytoscint ES. Vials were tightly capped and incubated for 2 h at 55-60°C in a shaking water bath. After the samples were cooled to room temperature, 15 mL of Cytoscint ES was added. Each vial was then mixed vigorously followed by further mixing with 0.5 mL of 0.5 mol/L glacial acetic acid. Vials were counted after being in the dark for 4 d to eliminate chemiluminescence from fecal porphyrins.

Tissue analysis.  Preweighed tissue samples (150-200 mg) were minced with scissors in glass scintillation vials containing 0.5 mL (0.75 mL for mammary gland and adipose tissues) of a 9:1 v/v solution of hyamine hydroxide and distilled water. Vials were capped tightly and incubated in a 55-60°C shaking water bath overnight until clear. Tissues with residual blood required bleaching to minimize color quenching of samples and were further treated with 100 µL of 9.7 mol/L hydrogen peroxide followed by heating at 55-60°C for 30 min with vials loosely capped. After the samples were cooled to room temperature, 15 mL of Cytoscint ES was added to each sample followed by vigorous mixing. The addition of 0.5 mL of 0.5 mol/L glacial acetic acid minimized chemiluminescence. Vials were then shaken and allowed to equilibrate in the dark for 30-60 min before counting.

Statistical analyses.  All statistical analyses were done using SigmaStat Version 2.0 by Jandel Scientific (San Rafael, CA). Within a treatment group, comparisons of radioactivities within a sample over time or radioactivities between samples at a particular time point were done using one-way ANOVA. Comparisons between acute and chronic groups at particular time points were done using unpaired Student's t test. Data not normally distributed were analyzed nonparametrically using either the Kruskal-Wallis one-way ANOVA on ranks or the Mann-Whitney Rank Sum test (Rosner 1990). Post-hoc multiple comparison tests included Tukey's test (parametric) or Dunn's method (nonparametric) (Rosner 1990). For all analyses, the acceptable level of significance was P < 0.05. Results are expressed as means ± SEM.

	RESULTS

Abstract Introduction Methods Results Discussion References

Total radioactivity in the feces and urine.  In rats with acute SDG exposure, maximum fecal radioactivity, representing >50% of the recovered dose, was reached by 12 h with no further increase to the 48-h time point (Fig. 2). Although fecal radioactivity in chronically treated rats was negligible at 12 h (P = 0.006 vs. acute), fecal levels approached that of acutely treated animals by 24 h after ³H-SDG administration (Fig. 2). An unusual increase in fecal radioactivity at 36 h was observed in the chronic group. Although the 36-h fecal radioactivity level was significantly higher than that of the acute group (P = 0.045), it was not significantly different than the 24- and 48-h fecal radioactivity levels determined in the chronic group (Fig. 2).

View larger version (29K): [in this window] [in a new window]   Fig 2. Cumulative radioactivity excreted in feces and urine over 48 h by rats that were not exposed to secoisolariciresinol diglycoside (SDG) (acute) or that were gavaged daily with unlabeled SDG (1.5 mg/d) for 10 d (chronic). Each symbol represents a mean ± SEM, n = 3 rats. Symbols within a line with different letters are significantly different by Tukey's test (parametric) or Dunn's method (nonparametric), P < 0.05; * a significant difference between acute and chronic groups at that time point by either unpaired Student's t test or Mann-Whitney Rank Sum test, P < 0.05.

The urinary radioactivity increased from 3 or 7% of the recovered dose at 12 h in the chronic and acute groups, respectively, to 28-32% of the recovered dose at 48 h (cumulative) (Fig. 2). The increase in urinary radioactivity in the acute group over the 48-h period was not significant (P = 0.088). The urinary radioactivity was 5.8-fold higher at 48 h compared with 12 h in the chronic group (P < 0.05) (Fig. 2). No significant differences in urinary radioactivity were observed between acute and chronic groups at the four time points (Fig. 2).

Total radioactivity in the gastrointestinal contents.  At 12 h, a large proportion of the administered radioactivity was found in the contents of the cecum and colon (Fig. 3). When combined, this radioactivity represented 27.5% of the recovered dose in the acute group and 57% of the recovered dose in the chronic group. Except for the stomach contents, radioactivity levels in the gastrointestinal contents significantly decreased over the 48-h period to <40 pmol equivalents of ³H-SDG (1-2% of the recovered dose) (P < 0.05). Compared with acutely treated rats at 12 h, rats with chronic SDG treatment had significantly higher cecal content radioactivity by nearly threefold (P = 0.006) (Fig. 3). The colon content radioactivity increased in the acute group at 36 h and was significantly greater than that in the chronic group (P = 0.010).

View larger version (27K): [in this window] [in a new window]   Fig 3. Level of radioactivity over 48 h in the gastrointestinal contents of rats that were not exposed to secoisolariciresinol diglycoside (SDG) (acute) or that were gavaged daily with unlabeled SDG (1.5 mg/d) for 10 d (chronic). Each symbol represents a mean ± SEM, n = 3 rats. Symbols within a line with different letters (nonbold = acute, bold = chronic) are significantly different by Tukey's test (parametric) or Dunn's method (nonparametric), P < 0.05; * a significant difference between acute and chronic groups at that time point by either unpaired Student's t test or Mann-Whitney Rank Sum test, P < 0.05.

Radioactivity per gram of tissue.  The tissue with the highest radioactivity throughout the 48-h experimental period was the cecum (P < 0.05) (Fig. 4). The radioactivity per gram of tissue was significantly higher at 12 h vs. the latter time points (36 or 48 h) for most gastrointestinal tissues (P < 0.05), except the small intestine in the acute group (P = 0.068). This decrease represented an overall drop from an initial 3 to 7% to <1% of the recovered dose in the gastrointestinal tissues. Some differences in gastrointestinal radioactivity between treatment groups included higher stomach radioactivity in the chronic group at 24 h (P = 0.038) and a tendency for higher cecal tissue radioactivity in the chronic group at 12 h (P = 0.067) (Fig. 4). By 36 h, gastrointestinal tissue radioactivity levels did not differ between groups.

View larger version (17K): [in this window] [in a new window]   Fig 4. Distribution of radioactivity over 48 h in gastrointestinal tissues of rats that were not exposed to SDG (acute) or that were gavaged daily with unlabeled SDG (1.5 mg/d) for 10 d (chronic). Each bar represents a mean + SEM, n = 3 rats. * Indicates a significant difference between acute and chronic groups for that tissue by either unpaired Student's t test or Mann-Whitney Rank Sum test, P < 0.05. Bars between panels with different letters (nonbold = acute, bold = chronic) represent significant differences between time points for a tissue within a treatment group by Tukey's test (parametric) or Dunn's method (nonparametric), P < 0.05. Significant differences among tissues in the acute group using Tukey's test (12 h only) or Dunn's method, P < 0.05: cecum vs. stomach (12, 24, 36 and 48 h); cecum vs. colon (12 and 48 h), cecum vs. small intestine (12 h); colon vs. stomach (12 h). Significant differences among tissues in the chronic group using Dunn's method, P < 0.05: cecum vs. stomach (12, 24, 36 and 48 h), cecum vs. colon (12, 24 and 48 h); cecum vs. small intestine (36 h); small intestine vs. stomach (24 h).

Overall, tissues other than gastrointestinal tissues represented ~4-6% of the recovered dose at 12 h, which declined to about 3% of the recovered dose by 48 h after ³H-SDG administration (Fig. 5). Significant decreases in radioactivity per gram of tissue were observed in the following tissues over time: liver (acute: 12 h vs. 48 h, chronic: 12 and 24 h vs. 36 and 48 h), kidney (acute: 12 h vs. 24, 36 and 48 h), adipose (chronic: 12 h vs. 36 h), uterus (chronic: 12 h vs. 36 and 48 h), and muscle (chronic: 12 h vs. 24, 36 and 48 h) (P < 0.05). Compared with other nongastrointestinal tissues, radioactivity was highest in the liver (0.5- to 6-fold greater than other tissues), kidney (0.4- to 7.5-fold greater), and uterus (0.2- to 4.4-fold greater).

View larger version (39K): [in this window] [in a new window]   Fig 5. Distribution of radioactivity over 48 h in tissues of rats that were not exposed to secoisolariciresinol diglycoside (SDG) (acute) or that were gavaged daily with unlabeled SDG (1.5 mg/d) for 10 d (chronic). Each bar represents a mean + SEM, n = 3 rats. Bars with an * indicate a significant difference between acute and chronic groups for that tissue by either unpaired Student's t test or Mann-Whitney Rank Sum test, P < 0.05. Bars between panels with different letters (nonbold = acute, bold = chronic) represent significant differences between time points for a tissue within a treatment group by Tukey's test (parametric) or Dunn's method (non-parametric), P < 0.05. Significant differences among tissues within the acute group using Tukey's test (48 h) or Dunn's method were as follows, P < 0.05: 1) 12 h, kidney and liver vs. adipose; 2) 24 h, uterus, kidney and liver vs. mammary gland and adipose; 3) 36 h, uterus, kidney, liver and brain vs. adipose; and 4) 48 h, kidney and uterus vs. adipose. Significant differences among tissues in the chronic group using Tukey's test or Dunn's method (12 h) were as follows, P < 0.05: 1) 12 h, liver, uterus and kidney vs. mammary gland, brain and heart; 2) 24 h, (a) liver vs. all except uterus and kidney, (b) uterus vs. mammary gland; 3) 36 h, (a) kidney vs. all except heart, spleen, and liver, (b) liver and spleen vs. adipose, mammary gland, ovaries, and skin, (c) heart vs. adipose and mammary gland, (d) lung, brain, uterus, and muscle vs. adipose; 4) 48 h, (a) kidney vs. adipose, mammary gland, and heart, (b) liver vs. adipose and mammary gland, (c) uterus vs. adipose. Mam. G. = mammary gland.

Some significant differences in tissue radioactivity levels were observed between acute and chronic groups at particular time points (Fig. 5). Chronic treatment enhanced the radioactivity per gram of liver by 50-80% compared with the acutely treated rats over the 48-h period, except at the 36 h time point (P < 0.03). Tissue radioactivity per gram adipose was threefold higher in the chronic group at 12 h (P = 0.042), but the reverse occurred, i.e., acute higher than chronic, at 36 h (P = 0.041). In the spleen, chronic treatment resulted in significantly greater radioactivity per gram at 36 h (P = 0.019), although similar levels were observed in both treatment groups at all other times. The ovary radioactivity was significantly higher in the acute group vs. the chronic group at 36 h (P = 0.035). Chronic treatment tended to increase radioactivity per gram of tissue in the uterus at 12 h (P = 0.106).

Total radioactivity in blood and blood components.  The level of radioactivity in the blood was always <1% of the recovered dose throughout the experimental period. The highest radioactivity levels in the blood and blood components were observed at 12 h (Fig. 6). The plasma fraction appeared to account for most of the whole-blood radioactivity with levels of 200-350 pmol/L (~0.4% of the recovered dose). Much higher and more variable plasma and whole-blood radioactivity levels were seen with chronic treatment at 12 h, but this was not significantly different than the acute group (Fig. 6). In the acute group, plasma radioactivity at 12 and 24 h was twice that measured at 48 h (P < 0.05). In the chronic group, significantly lower whole-blood and RBC radioactivity was found at 48 vs. 12 h (P < 0.05, Fig. 6).

View larger version (23K): [in this window] [in a new window]   Fig 6. Level of radioactivity over 48 h in blood and blood components of rats that were not exposed to secoisolariciresinol diglycoside (SDG) (acute) or that were gavaged daily with unlabeled SDG (1.5 mg/d) for 10 d (chronic). Each symbol represents a mean (±SEM) of n = 3 rats. Symbols within a line with different letters are significantly different by Tukey's test (parametric) or Dunn's method (nonparametric), P < 0.05.

The contribution of residual blood content radioactivity as a percentage of the tissue radioactivity was calculated on the basis of the radioactivity detected in the whole-blood sample. The residual blood content values (mL/g tissue) for the heart (0.243 ± 0.021), liver (0.182 ± 0.007), kidney (0.209 ± 0.011) and spleen (0.157 ± 0.007) were based on the data of Smith (1970). In most tissues over the 48-h experimental period, the residual blood radioactivity was 15% of the tissue radioactivity (data not shown). The residual blood content radioactivity in the heart was higher, between 20 and 30% at most time points. However, the heart residual blood radioactivity in the chronic group accounted for the majority of the total tissue radioactivity at 12 h (70.3 ± 11.2%) and was significantly higher than that observed in the acute group (24.7 ± 5.9%, P = 0.011, data not shown).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

This study is the first to demonstrate both body distribution and excretion of lignan metabolites after oral administration of a purified precursor compound. Many human trials examining lignan levels in plasma, urine or feces have used flaxseed because it is a rich source of SDG (Atkinson et al. 1993, Kurzer et al. 1995, Lampe et al. 1994, Shultz et al. 1991). The SDG levels in flaxseed have been shown to be approximately 2000 µg/g (Rickard et al. 1996) with the use of a method developed by Obermeyer and colleagues (Obermeyer et al. 1995). However, flaxseed appears to contain other precursors that produce the mammalian lignans ED and EL (Rickard et al. 1996), thus making it difficult to determine the proportion of the consumed dose recovered in various biological samples. The pharmacokinetics of lignans can be determined only after the administration of a pure precursor compound or, at the very least, after feeding a food containing known amounts of all precursor compounds. After feeding rats purified SDG from flaxseed, we found that total urinary levels of the lignan metabolites, i.e., ED, EL and secoisolariciresinol (SECO), were 16.8% of the daily gavaged dose of 1.5 mg SDG (Rickard et al. 1996). However, because plasma and fecal measurements were not made, the body retention of lignan metabolites could not be determined. In addition, tissue lignan concentrations are unknown due to the lack of a suitable lignan extraction method, necessitating the use of a labeled compound.

In this study, oral administration of ³H-SDG resulted in high fecal and urinary radioactivity with relatively low blood and tissue radioactivity, indicating low retention of lignan metabolites after consumption. By 48 h, >50% of the recovered dose was present in the feces and at least 70% of the total absorbed dose (32-35% of the recovered dose) was found in the urine. The highest tissue radioactivity was found in the cecum. Of the nongastrointestinal tissues, the liver, kidney and uterus had the highest radioactivity levels. Over the 48 h period, radioactivity levels were greatest in the gastrointestinal contents, tissues and blood at 12 h and lowest at the 48-h time point.

Some important differences were observed between rats exposed to SDG for the first time (acute group) and those treated with 1.5 mg SDG for a 10-d period (chronic group). The SDG dose in the chronic group was chosen because it has been used previously in this laboratory in examining urinary lignan metabolites (Rickard et al. 1996) and has been an effective dose in carcinogenesis studies (Thompson et al. 1996a and 1996b). The length of SDG treatment for the chronic group was considered adequate because previous studies have shown that urinary lignan levels stabilize within 1 wk (Rickard 1994). In rats with chronic exposure to SDG, maximal fecal radioactivity levels were seen at 24 h after oral administration. In contrast, maximal radioactivity levels in the acute group were observed by 12 h. In addition, the 12-h cecal content radioactivity was significantly higher, with a strong tendency for higher cecal tissue radioactivity in the chronic group compared with the acute group. Furthermore, tissue radioactivity levels, particularly in the liver and adipose, were greater with chronic treatment.

The fecal results seem to indicate that chronic treatment with SDG delays but does not reduce fecal lignan excretion. This delay may be due to differences in the metabolic activity of the cecal bacteria in the acute and chronic groups. Although fermentation in humans occurs largely in the colon, rats are mainly cecal fermenters. Many bacterial enzymes are inducible to cope with increased substrate (Reddy et al. 1992). The enzyme -glucuronidase is responsible for removing the glucuronide conjugate and allowing reabsorption of compounds from the gastrointestinal tract. The activity of this enzyme tended to be higher in cecal and colon contents of rats exposed to a high fat diet with SDG compared with those fed a high fat diet alone (Jenab and Thompson 1996). Mammalian lignans re-entering the small intestine are in the conjugated form and must be acted upon by -glucuronidase for reabsorption (Axelson and Setchell 1981). In unexposed animals, more lignans may therefore be excreted in the feces initially because of inefficient metabolism of SDG and absorption of lignan metabolites. The lower gastrointestinal content radioactivity, especially that in the cecum, in acutely treated rats may be the result of smaller amounts of lignans undergoing enterohepatic circulation. In rats with chronic exposure to SDG, the probable increase in bacterial activity would result in faster metabolism of the labeled SDG as it entered the cecum and hence, increased absorption of the mammalian lignans formed. The combination of recycled lignan metabolites (from enterohepatic circulation) and perhaps residual levels of the initial SDG dose would increase lignan levels in the cecal contents at 12 h. Between 12 and 24 h, the excess substrate may have eventually exceeded the metabolic capacity of the bacteria. This would result in decreased absorption and increased fecal lignan excretion, hence the rise in fecal radioactivity of the chronic group to levels seen in the acute group by 24 h. The -glucuronidase enzyme has been shown to be saturable with estrogen substrates (Fishman 1939), but no kinetic studies have been done with lignans to date.

As mentioned previously, a substantial proportion of the radioactivity was present in the urine. Urinary radioactivity in the chronic group was significantly higher at 48 h compared with 12 h, suggesting that this may be the result of lignan removal from tissues. Chronic treatment resulted in greater liver and adipose radioactivity levels compared with acute treatment (Fig. 5). The proportion of radioactivity present in the urine at 24 h in the chronic group (12%) is consistent with our previous report of 11.4% for the mammalian lignans ED and EL in rats fed 1.5 mg SDG/d for 2 wks (Rickard et al. 1996). However, this is slightly lower than the 16.8% reported for total lignans (ED + EL + SECO) in the same study. Because only total radioactivity was measured in this study, we cannot comment on the types or amounts of metabolites present in the urine or indeed any of the samples tested. Nevertheless, the similarity in urinary recovery between this study and our previous one in rats chronically exposed to SDG (Rickard et al. 1996) suggests that the majority of the recovered dose in the chronic group is ED and EL. This provides some evidence that the radioactivity detected, at least in the urine, is due to actual metabolites of SDG and not the result of hydrogen exchange. Work is presently in progress to identify and quantify labeled urinary lignan metabolites by HPLC and gas chromatography/mass spectometry. This should clarify whether acute vs. chronic treatment changes the profile of ³H-SDG metabolites in urine.

The type and amount of metabolite may also be dependent upon the dose of the parent compound. Rickard et al. (1996) found proportionately smaller amounts of mammalian lignans in the urine in rats fed 3.0 mg SDG/d compared with those fed either 0.75 or 1.5 mg SDG/d. More research into dose-dependent pharmacokinetics of lignans is warranted.

The higher radioactivity levels in the liver, kidney and uterus compared with other nongastrointestinal tissues were not surprising. The liver is believed to be the site for lignan conjugation with glucuronide, although glucuronidation by the intestinal mucosal cells has not been ruled out (Axelson and Setchell 1981). A large proportion of the lignan metabolites is excreted via the kidney. Appreciable radioactivity in the uterus was expected because mammalian lignans antagonize estrogen action in this tissue (Adlercreutz et al. 1992a, Waters and Knowler 1982). The greater liver radioactivity with chronic SDG exposure may be due to enhanced absorption, suggested by the tendency for higher cecal tissue radioactivity within the first 12 h. Increased lignan binding in the liver may have occurred because the liver is a tissue rich in antiestrogen binding sites that bind nonsteroidal antiestrogens such as tamoxifen (Sudo et al. 1983). However, the binding of lignans to these sites have not been examined to date.

In acutely treated rats, the lowest radioactivity was generally found in the adipose tissue. Chronic treatment resulted in significantly greater adipose radioactivity at 12 h by threefold compared with acute treatment. This early increase with chronic exposure to SDG may be the result of transient storage of the excess mammalian lignans. Adipose tissue is also a primary site for peripheral production of estrogen by the action of aromatase on androgens. The mammalian lignan EL has exhibited antiaromatase activity in vitro (Adlercreutz et al. 1993, Wang et al. 1994), with a K_i = 14.4 µmol/L (androstenedione, K_m = 30 nmol/L) in human preadipocytes (Wang et al. 1994). With a recovered dose of 0.03%/g adipose tissue in the chronically treated group at 12 h, potential mammalian lignan levels at the SDG level fed (~2.2 µmol/d) would be ~0.6 µmol/L of adipose tissue (estimating the specific gravity of adipose tissue at 900 g/L). Because this value cannot be compared directly with the inhibitory concentration of EL observed in vitro, its physiologic importance is still unclear. However, the result does show that levels in adipose tissue are not negligible.

For some tissues, the radioactivity level at the 36 h time point did not follow expected trends. Levels in spleen were greater in the chronic group than in the acute group, whereas the groups had similar levels at all other time points. Radioactivity levels in adipose tissues at 36 h were higher in the acute group, whereas the general trend was for higher levels in the chronic group at 12, 24 and 48 h. Although the reason for this change in trend is unknown, we do not believe it to be of physiologic importance. Higher radioactivity in the colon contents of rats in the acute group at 36 h compared with those in the chronic group was also observed, but this may be explained by higher colon content weights in the acutely treated rats (data not shown). The higher fecal radioactivity at 36 h in the chronic group is considered to be an anomalous result without physiologic relevance.

High radioactivity levels were expected in the rat mammary gland because SDG has been shown to be antitumorigenic in the DMBA model (Thompson et al. 1996a and 1996b). The relatively low levels observed suggest that the anticancer effect of lignans may be through direct binding to the tumor itself and not to the mammary gland. Antiestrogen-binding sites have been detected in 98% of DMBA-induced tumors in rats (Hwang and How 1990). Yet, an earlier study by Mehta et al. (1984) did find the number of cytoplasmic sites in the mammary gland to be threefold higher than that in tumors induced by N-methyl-N-nitrosourea. Nevertheless, the function of these binding sites in chemically induced tumor growth is still unknown. Other possibilities are that lignan concentrations peak in the mammary gland at times other than those chosen for this study or that the low radioactivity may be the result of a constant flux of lignans into and out of the tissue with no accumulation. Examination of isolated tissues in culture may provide an explanation.

From the analyses of the blood samples, the lignan metabolites appear to be present in the plasma fraction. Although a significant drop in plasma radioactivity over time was observed in the acute group, the change in plasma radioactivity in the chronic group was not significant because of the large variability at 12 h. However, whole-blood radioactivity did decrease significantly in the chronic group, indicating that blood clearance rates are faster initially (i.e., between 12 and 24 h). Total elimination of lignans from the blood apparently takes longer than 48 h because appreciable levels were still detected at this time point.

The SDG picomole equivalents per liter in plasma were similar to peak estrogen levels (300 pmol/L) recorded in rats (Butcher et al. 1974). Because this value represents only 0.4% of the recovered labeled dose, actual unlabeled plasma lignan concentrations in rats with chronic exposure to SDG would be higher. On a molar basis, the labeled dose given was >2500 times less than doses used (e.g., 1.5 mg/d or 2.2 µmol/d) in our carcinogenesis studies (Thompson et al. 1996a and 1996b). Using the estimates of 0.4% of dose and a plasma volume 40.4 mL/kg BW (8 mL in a 200-g rat), plasma lignan concentrations would be about 1 µmol/L in rats fed 1.5 mg/d. This value is over 3000 times greater than peak estrogen levels in rats.

The mammalian lignans are weaker estrogens than steroidal estrogens. Concentrations required for sex hormone binding globulin stimulation were only 10 times that of estradiol (Adlercreutz et al. 1992a). Subcutaneous administration of EL (0.3-30 µg) reduced estradiol-stimulated uterine RNA synthesis in the immature rat fourfold (Waters and Knowler 1982). In addition, elimination of the stimulatory effect of estradiol on MCF-7 breast cancer cell growth was achieved at EL concentrations (0.5-2 µmol/L) that were 1000 times higher (Mousavi and Adlercreutz 1992). Furthermore, levels of ED and EL (1 µmol/L) that reduced estradiol binding to the rat uterine type II receptor by 40% were only 25 times greater (Adlercreutz et al. 1992a). However, the affinity of lignans for other estrogenic activities, such as stimulation of the estrogen-responsive protein pS2 (Sathyamoorthy et al. 1994) or binding to rat -fetoprotein (Garreau et al. 1991) is 5000 to 10,000 times less than steroidal estrogens. Nevertheless, the relatively higher plasma levels achieved with SDG feeding can have a significant effect on the biological activity of steroidal estrogens in vivo.

The residual blood content radioactivity calculated for some tissues was quite low. The unusually high 12-h radioactivity levels observed for the heart residual blood in the chronic group may be due to an underestimation of heart tissue radioactivity. Nevertheless, tissue radioactivity does not appear to be associated with its blood supply alone but may be some evidence of tissue binding. Autoradiography of formalin-fixed tissues would verify lignan binding to tissues and whether this binding is targeted to certain cell types.

In conclusion, regardless of previous exposure to SDG, >50% of the lignans are excreted in the feces and 30% is present in the urine with low tissue retention by 48 h after administration. The tissues with the greatest concentration of lignan metabolites are involved in SDG metabolism (i.e., intestinal, hepatic and renal). Appreciable levels of lignan metabolites also occur in certain estrogen-sensitive tissues such as the uterus. Blood lignan concentrations with chronic exposure to SDG are much higher than peak estrogen levels recorded in the rat, suggesting that lignans can have a significant effect on estrogen action in vivo. Chronic exposure to SDG resulted in a delay, as opposed to a reduction, in fecal lignan excretion and increased uptake of lignan metabolites in the liver and adipose tissues one- to threefold. This may have been the result of increased bacterial activity and hence more efficient absorption of lignans initially.

	FOOTNOTES

Manuscript received 11 July 1997. Initial reviews completed 11 September 1997. Revision accepted 17 November 1997.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Adlercreutz H., Bannwart C., Wahala K., Makela T., Brunow G., Hase T., Arosemena P. J., Kellis J. T. Jr., Vickery L. E. Inhibition of human aromatase by mammalian lignans and isoflavonoid phytoestrogens. J. Steroid Biochem. Mol. Biol. 1993; 44:147-153 [Medline]
• Adlercreutz H., Mousavi Y., Clark J., Hockerstedt K., Hamalainen E., Wahala K., Makela T., Hase T. Dietary phytoestrogens and cancer: in vitro and in vivo studies. J. Steroid Biochem. Mol. Biol. 1992a; 41:331-337 [Medline]
• Adlercreutz H., Mousavi Y., Hockerstedt K. Diet and breast cancer. Acta Oncol. 1992b; 31:175-181 [Medline]
• Axelson M., Setchell K.D.R. The excretion of lignans in ratsevidence for an intestinal bacterial source for this new group of compounds. FEBS Lett. 1981; 123:337-342 [Medline]
• Axelson M., Sjovall J., Gustafsson B. E., Setchell K.D.R. Origin of lignans in mammals and identification of a precursor from plants. Nature (Lond.) 1982; 298:659-660 [Medline]
• Baker, H. J., Lindsey, J. R. & Weisbroth, S. W., eds. (1980) The Laboratory Rat, Research Applications, vol. 2, pp. 257-258. Academic Press, New York, NY.
• Borriello S. P., Setchell K.D.R., Axelson M., Lawson A. M. Production and metabolism of lignans by the human faecal flora. J. Appl. Bacteriol. 1985; 58:37-43 [Medline]
• Butcher R. L., Collins W. E., Fugo N. W. Plasma concentrations of LH, FSH, prolactin, progesterone and estradiol-17 throughout the 4-day estrous cycle of the rat. Endocrinology 1974; 94:1704-1708 [Abstract/Free Full Text]
• Canadian Council on Animal Care (1984) Guide to the Care and Use of Experimental Animals. Ottawa, Canada.
• Dehennin L., Reiffsteck A., Jondet M., Thibier M. Identification and quantitative estimation of a lignan in human and bovine semen. J. Reprod. Fert. 1982; 66:305-309 [Abstract/Free Full Text]
• Evans, E. A., ed. (1974) Tritium and Its Compounds, pp. 489-490. Halsted Press, Toronto, Canada.
• Fishman W. H. Studies on -glucuronidase. J. Biol. Chem. 1939; 131:225-232 ^{[Free Full Text]}
• Fotsis T., Pepper M., Adlercreutz H., Fleischmann G., Hase T., Montesano R., Schweigerer L. Genistein, a dietary-derived inhibitor of in vitro angiogenesis. Proc. Natl. Acad. Sci. U.S.A. 1993; 90:2690-2694 [Abstract/Free Full Text]
• Garreau B., Vallette G., Adlercreutz H., Wahala K., Makela T., Benassayag C., Nunez E. A. Phytoestrogens: new ligands for rat and human alpha-fetoprotein. Biochim. Biophys. Acta 1991; 1094:339-345 [Medline]
• Hirano T., Fukuoka K., Oka K., Naito T., Hosaka K., Mitsuhashi H., Matsumoto Y. Antiproliferative radioactivity of mammalian lignan derivatives against the human breast carcinoma cell line, ZR-75-1. Cancer Invest. 1990; 8:595-602 [Medline]
• Hwang P. L., How B. E. Antiestrogen-binding sites in 7,12-dimethylbenz[a]anthracence-induced rat mammary tumors: relation to growth. Oncology 1990; 47:495-500 [Medline]
• Jenab M., Thompson L. U. The influence of flaxseed and lignans on colon carcinogenesis and beta-glucuronidase radioactivity. Carcinogenesis 1996; 17:1343-1348 [Abstract/Free Full Text]
• Lu L. J., Lin S. N., Grady J. J., Nagamani M., Anderson K. E. Altered kinetics and extent of urinary daidzein and genistein excretion in women during chronic soya exposure. Nutr. Cancer 1996; 26:289-302 [Medline]
• Mainland, D. (1945) The body as a wholevariation and development. In: Anatomy (Mainland, D., ed.), pp. 31-69. Paul B. Boeber, New York, NY.
• Martin M. E., Haourigui M., Pelissero C., Benassayag C., Nunez E. A. Interactions between phytoestrogens and human sex steroid binding protein. Life Sci. 1996; 58:429-436 [Medline]
• Mehta R. G., Cerny W. L., Moon R. C. Distribution of antiestrogen-specific binding sites in normal and neoplastic mammary gland. Oncology 1984; 41:387-392 [Medline]
• Miller W. R., O'Neill J. S. The significance of steroid metabolism in human cancer. J. Steroid Biochem. Mol. Biol. 1990; 37:317-325 [Medline]
• Mousavi Y., Adlercreutz H. Enterolactone and estradiol inhibit each other's proliferative effect on MCF-7 breast cancer cells in culture. J. Steroid Biochem. Mol. Biol. 1992; 41:615-619 [Medline]
• Nesbitt, P. D. (1997) Mammalian Lignan Production from Flaxseed: Studies In Vitro and in Humans. Master's thesis, University of Toronto, Toronto, Canada.
• Obermeyer W. R., Musser S. M., Betz J. M., Casey R. E., Pohland A. E., Page S. W. Chemical studies of phytoestrogens and related compounds in dietary supplements: flax and chaparral. Proc. Soc. Exp. Med. Biol. 1995; 208:6-12 [Medline]
• Reddy B. S., Engle A., Simi B., Goldman M. Effect of dietary fiber on colonic bacterial enzymes and bile acids in relation to colon cancer. Gastroenterology 1992; 102:1475-1482 [Medline]
• Reeves P. G., Nielsen F. M., Fahey G. C. Jr. AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet. J. Nutr. 1993; 123:1939-1951
• Rickard, S. E. (1994) Flaxseed Lignans and Mammary Carcinogenesis. Master's thesis, University of Toronto, Toronto, Canada.
• Rickard S. E., Orcheson L. J., Seidl M. M., Luyengi L., Fong H. H., Thompson L. U. Dose-dependent production of mammalian lignans in rats and in vitro from the purified precursor secoisolariciresinol diglycoside in flaxseed. J. Nutr. 1996; 126:2012-2019
• Rikimaru T., Oozeki T., Ichikawa M., Ebisawa H., Fujita Y. Comparisons of urinary creatine, skeletal muscle mass, and indices of muscle protein catabolism in rats fed ad libitum with restricted food intake, and deprived of food. J. Nutr. Sci. Vitaminol. 1989; 35:199-209
• Rosner, B., ed. (1990) Fundamentals of Biostatistics, 3rd ed., pp. 293-347, 474-526. Duxberry Press, Belmont, CA.
• Sakamoto K., Vucenik I., Shamsuddin A. M. [³H]Phytic acid (inositol hexaphosphate) is absorbed and distributed to various tissues in rats. J. Nutr. 1993; 123:713-720
• Sathyamoorthy N., Wang T.T.Y., Phang J. M. Stimulation of pS2 expression by diet-derived compounds. Cancer Res. 1994; 54:957-961 [Abstract/Free Full Text]
• Serraino M., Thompson L. U. The effect of flaxseed supplementation on early risk markers for mammary carcinogenesis. Cancer Lett. 1991; 60:135-142 [Medline]
• Serraino M., Thompson L. U. Flaxseed supplementation and early markers of colon carcinogenesis. Cancer Lett. 1992a; 63:159-165 [Medline]
• Serraino M., Thompson L. U. The effect of flaxseed supplementation on the initiation and promotional stages of mammary tumorigenesis. Nutr. Cancer 1992b; 17:153-159 [Medline]
• Setchell K.D.R., Lawson A. M., Borriello S. P., Harkness R., Gordon H., Morgan D. M., Kirk D. N., Adlercreutz H., Anderson L. C., Axelson M. Lignan formation in manmicrobial involvement and possible roles in relation to cancer. Lancet 1981a; 2:4-7 ^[Medline]
• Setchell K.D.R., Lawson A. M., Conway E., Taylor N. F., Kirk D. N., Cooley G., Farrant R. D., Wynn S., Axelson M. The definitive identification of the lignans trans-2,3-bis(3-hydroxybenzyl)-gamma-butyrolactone and 2,3-bis(3-hydroxybenzyl)butane-1,4-diol in human and animal urine. Biochem. J. 1981b; 197:447-458 ^[Medline]
• Setchell K.D.R., Lawson A. M., Mitchell F. L., Adlercreutz H., Kirk D. N., Axelson M. Lignans in man and in animal species. Nature (Lond.) 1980; 287:740-742 ^[Medline]
• Simberg N. H., Murai J. T., Siiteri P. K. In vitro and in vivo binding of toremifene and its metabolites in rat uterus. J. Steroid Biochem. 1990; 36:197-202 [Medline]
• Sipila H., Kangas L., Vuorilehto L., Kalapudas A., Eloranta M., Sodervall M., Toivola R., Antila M. Metabolism of toremifene in the rat. J. Steroid Biochem. 1990; 36:211-215 [Medline]
• Smith B.S.W. A comparison of ¹²⁵I and ⁵¹Cr for measurement of total blood volume and residual blood content of tissues in the rat: evidence for accumulation of ⁵¹Cr by tissues. Clin. Chim. Acta 1970; 27:105-108 ^[Medline]
• Stich S. R., Toumba J. K., Groen M. B., Funke C. W., Leemhuis J., Vink J., Woods G. F. Excretion, isolation and structure of a new phenolic constituent in female urine. Nature (Lond.) 1980; 287:738-740 ^[Medline]
• Sudo K., Monsma F. J., Katznellenbogen B. S. Antiestrogen binding sites distinct from the estrogen receptor: subcellular localization, ligand specificity, and distribution in tissues of the rat. Endocrinology 1983; 112:425-434 [Abstract/Free Full Text]
• Thompson L. U., Rickard S. E., Orcheson L. J., Seidl M. M. Flaxseed and its lignan and oil components reduce mammary tumor growth at a late stage of carcinogenesis. Carcinogenesis 1996a; 17:1373-1376 [Abstract/Free Full Text]
• Thompson L. U., Robb P., Serraino M., Cheung F. Mammalian lignan production from various foods. Nutr. Cancer 1991; 16:43-52 [Medline]
• Thompson L. U., Seidl M., Rickard S., Orcheson L., Fong H. Antitumorigenic effect of a mammalian lignan precursor from flaxseed. Nutr. Cancer 1996b; 26:159-165 [Medline]
• Wang C., Makela T., Hase T., Adlercreutz H., Kurzer M. S. Lignans and flavonoids inhibit aromatase enzyme in human preadipocytes. J. Steroid Biochem. Mol. Biol. 1994; 50:205-212 [Medline]
• Waters A. P., Knowler J. T. Effect of a lignan (HPMF) on RNA synthesis in the rat uterus. J. Reprod. Fert. 1982; 66:379-381 [Abstract/Free Full Text]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences