The metabolism of linseed lignans in rumen and its impact on ruminal metabolism in male goats

To investigate the metabolism of linseed lignans in the rumen and its impact on ruminal metabolism, four healthy male Huainan goats were used in a crossover study. Following secoisolariciresinol diglucoside (SDG) crude extract treatment, the concentrations of SDG, enterodiol (END) and enterolactone (ENL) in both ruminal fl uid and serum were signifi cantly increased (P<0.01). Concomitantly, the pH value and ammonia-nitrogen concentration of ruminal fl uid were decreased (P<0.05), and the concentrations of microbial crude protein and total volatile fatty acids were increased (P<0.01). Furthermore, testosterone and 3, 3′, 5-triiodothyronine levels in serum and ruminal fl uid were both signifi cantly increased, while the concentration of E2 in ruminal fl uid was decreased. Based on these observations, it is suggested that ruminal microorganisms were able to effi ciently convert SDG to END and ENL, which were absorbed immediately. In return, SDG and/or its metabolites may facilitate the utilization of non-protein nitrogen and carbohydrates


INTRODUCTION
Phytoestrogens are oestrogen-like compounds of plants origin that comprise essentially 2 families-the isofl avones and the lignans.Lignans, which are ubiquitous in many edible plants, are the main source of dietary phytoestrogen.
Dietary lignans, particularly secoisolariciresinol diglucoside (SDG), are of interest because they have been proposed to play a role in the prevention of breast and colon cancer, atherosclerosis and diabetes (Wang, 2002).Previous studies in non-ruminant animals have indicated that the plant lignans per se are devoid of any biological properties, and that only the mammalian lignans enterodiol (END) and enterolactone (ENL), produced from SDG by intestinal bacteria, have interesting biological properties such as oestrogen agonism and antagonism as well as antioxidative and enzyme-inhibiting activities (Clavel et al., 2005).Earlier studies in rats and humans had established that SDG could be metabolized to END, probably through hydrolysis, dehydroxylation and then demethylation by facultative bacteria, and then ENL was produced from END through oxidation by faculatative bacteria (Wang, 2002); however the bacterial convertation of SDG in rumen have not been reported.
As reviewed by Han et al. (2006), injection of daidzein, one of the main types of isofl avone, via duodenal cannulae increased serum testosterone, rumen bacterial protein, ammonia nitrogen and total volatile fatty acids (TVFA).Previous studies have suggested that the absorbed daidzein may facilitate the growth of male animals via the neuroendocrine pathway.However, the absorption of lignans in ruminants and the impact of lignans on ruminal metabolism remain unclear.Since both isofl avones and lignans have, as their common denominator, a phenolic group that they share with oestrogenic steroids, it was presumed that lignans may have approximate physiological effect on ruminal metabolism.
Therefore, the aims of the current study were to elucidate the metabolism and absorption of SDG in ruminants and the possible impact of SDG and/or its metabolites (END and ENL) on ruminal metabolism.

Animals and management
Male Huainan goats (4 animals, fi tted with permanent rumen fi stulae and temporary catheters in the jugular vein) were housed in well-ventilated, cementfl oored individual pens and maintained under conditions of strict hygiene and uniform management.The animals were fed twice per day (08.00 and 18.00 h) at a maintenance energy level [55 g dry matter (DM) per kg body weight (BW) per day] on a basal diet (BD) consisting of 700 g•kg -1 hay, and 300 g•kg -1 cracked maize (DM basis).In order to control the interference of the other kinds of phytoestrogens, the hay was supplemented without forage legumes, which may contain high level of isofl avones.The animals also had free access to fresh water.ZHOU W. ET AL.

Experimental design and sampling
The goats were randomly assigned to 2 groups of 2 animals each on the basis of BW (20 ± 2.5 kg), in a crossover study with 2 periods (control or treatment period).Each period was of 4 weeks' duration, comprising 14 d of adaptation to the SDG crude extract (SDGCE) or 14 d of washout, followed by 14 d of sampling.During the treatment period, using a cylinder, 50 mg•kg body weight -1 SDGCE (20% SDG, w:w; LinumLife, Taiwan) was infused into the rumen of the goats via the fi stulae when the BD diet was offered.
On days 14, 21, and 28, samples of ruminal contents obtained via ruminal cannula and 2 ml blood (collected through a catheter in the jugular vein) were obtained from each goat at 7.00 a.m., and thereafter at 2-h intervals over the next 22 h.The blood samples were centrifuged (1000 g, 15 min) in order to obtain sera, which was stored at -20°C until analysis.Ruminal fl uid pH and the total activity of dehydrogenase (TDHA) were determined from fresh ruminal samples.Other rumen samples were strained through 4 layers of cheesecloth.Two ml ruminal samples were used for microbial crude protein (MCP) analyses.A 10-ml aliquot of the ruminal fl uid was acidifi ed with 1 ml of 6N HCl, and stored in a freezer (-20°C) for ammonia-nitrogen (NH 3 -N) analysis.Two millilitres of freshly prepared 25% (250 ml•l -1 ) metaphosphoric acid was added to 8 ml of strained ruminal fl uid.Samples were then centrifuged (17,000 g, 10 min), and the supernatant fl uid was stored at -20°C for volatile fatty acids (VFA) analysis.And another 2 ml of ruminal fulid was centrifuged (10,000 g for 10 min), the supernatant was stored at -20°C until hormone determination.

Chemical analysis
In this study, the levels of mammalian lignans in the serum and ruminal fl uid of the 4 male goats were used to assess the bacterial activation and absorption of SDG.The concentrations of SDG, END and ENL in both ruminal fl uid and serum were measured by high-performance liquid chromatography (HPLC) (Agilent 1100 Series; Agilent, USA) using the respective authentic standards [SDG (Chromadex, USA); END and ENL (Sigma, USA)] as described by Morton et al. (1997).As Adlercreutz et al. (1995) found that 92% of END and 98% ENL exist in glucuronide conjugates in serum, the analytical procedure for END and ENL in serum included a hydrolysis step using NaOH (1 mol•l -1 , 20°C, 48 h).
TDHA may refl ect the activities of ruminal microorganisms.In this study, TDHA was analysed following the method described by Dror et al. (1969).Briefl y, 4.5 ml rumen liquor and 0.5 ml of triphenyl tetrazolium chloride (TTC, 1.5%, w:w) were incubated at 38° for 5 min.The reaction was stopped by 4.5 ml isopropanol (50%, v:v).After centrifuged, chromatometry were used to determine the total dehydrogenase activity.To examine the effect of SDGCE on ruminal metabolism, some metabolism parameters, such as MCP, in rumianl fl uid were analysed.NH 3 -N was measured by the indophenol method (Weatherburn, 1967).The VFA (acetate, propionate and butyrate) were determined using a gas chromatograph (GC-9APTF; Shimadzu, Kyoto, Japan) equipped with a fl ame ionization detector (FID) and a capillary column (HP-INNOWAX, 1909N-133; Hewlett-Packard, USA).The temperatures of the detector and column were 220 and 170°C, respectively.Nitrogen was used as a carrier, and total fl ow and column fl ow were both 63.8 ml•min -1 .MCP was determined by the method of Zinn and Owens (1986), based on purine, and estimated from the ratio of purines to nitrogen in isolated microbes.

Hormones determination
The concentrations of testosterone (T); estradiol (E 2 ); 3, 3′, 5-triiodothyronine (T 3 ); and thyroxine (T 4 ) in the serum and ruminal fl uid were measured by radioimmunoassay (RIA) using commercial kits purchased from the North Institute of Biological Technology (Beijing, China).This method was based on the binding of antibody (in agent) and specifi ed hormones (the antigen) in samples.These binding have high degree of specifi city, for examples: the cross reaction rate of the antibody of E 2 to trihydroxyestrin (E 3 ), progesterone (P) and testosterone (T) were 0.016%, <0.01 and 0.01%, respectively.
For ruminal samples, a 100-µl aliquot of each ruminal fl uid sample were mixed in a tube with activated charcoal (10 ml : 1 g), homogenized, incubated for 3 h with continuous shaking at ambient temperature, and then centrifuged for 10 min (10000 rpm, 4°C).The supernatant was used to correct for the non-specifi c adsorption of ruminal fl uid.The inter-and intra-assay coeffi cients of variation of the 4 kits were 10 and 15%, respectively.The binding rates of SDG, END, and ENL to E 2 and T antibodies were also determined in the same manner using their respective authentic standards.The antibodies were demonstrated to be specifi c for T or E 2 and did not cross-react with SDG or the 2 mammalian lignans (END and ENL).

Statistical analyses
For metabolism parameter and hormones concentrations, such as the MCP and T levels, in order to control the infl uences of other factors (circadian, the sampling time, feed, water intake, and so on), all of the samples were analysed separately, and then the mean levels of each metabolism parameter were calculated on each sampling day (14, 21 and 28-d) for each goat.Based on random factor (goat) and fi xed factor (treatment and period), the differences between BD and SDGCE treatment were tested for signifi cance using the analysis of variance by SPSS system (V 11.5, SPSS Inc.)The concentrations of SDG, END and ENL in ruminal fl uid and serum were analysed by the analysis of variance for repeated measures.Data are expressed as means ± standard error (S.E.).

The lignan concentrations in ruminal fl uid and serum
There were low levels of END (0.018±0.018 g•l -1 ) and ENL (0.024±0.021 g•l -1 ) in ruminal fl uid in the ruminal fl uid under the basal diet condition (Figure 1).This may indicate that ruminal microorganisms effi ciently converted SDG to mammalian lignans; however, those of SDG, END and ENL in the serum were under the detected level.
As shown in Figure 1-B and C, on the fi rst day when SDGCE was offered, the concentrations of END and ENL (09.00 a.m.) were remarkably increased.The levels of SDG, END and ENL in both ruminal fl uid and serum were increased by SDGCE treatment, and their average levels in serum during the treatment period were 0.128±0.018g•l -1 , 0.246±0.068g•l -1 and 0.451±0.130g•l -1 , respectively.
Table 1 summarizes the impacts of SDGCE on ruminal metabolism.Following SDGCE treatment, the pH value of ruminal fl uid was signifi cantly decreased by 0.15; however, there was no marked change in the TDHA.In contrast, the TVFA level (P<0.01) and the (acetate + butyrate): propionate ratio [(A+B): P ratio, P<0.05] were signifi cantly increased by SDGCE.A positive effect on rumen bacterial protein synthesis was also observed; the NH 3 -N concentration was remarkably reduced (P<0.05),whereas the MCP level in ruminal fl uid was signifi cantly increased by SDGCE (P<0.01)., g• l -1 0.75 1.63 0.10 <0.001 1 BD -basal diet; 2 BD + SDGCE -basal diet + SDGCE; 3 effect of SDGCE supplementation TDHA -total dehydrogenase activeity; TVFA -total volatile fatty acids; NH 3 -N -ammonia nitrogen; MCP -microbial crude protein The impacts of SDGCE on concentrations of hormones in male goats are presented in Figure 2. T concentrations were signifi cantly increased at every time point throughout the sampling day in both serum (P<0.01) and ruminal fl uid (P<0.01).The level of E 2 in serum was below the level of detection; however, its concentration in ruminal fl uid was 5.431±2.209ng•l -1 on the BD and was remarkably decreased (P<0.01) by SDGCE.Positive effects of SDGCE on thyroid hormones were also observed.The T 3 levels were increased in both the rumen (P<0.05) and serum (P<0.01) by SDGCE; however, T 4 concentrations in serum were not signifi cantly changed, and remained below the level of detection in ruminal fl uid throughout this experiment.

Transformation and absorption of SDG in goats
The transformation and absorption of plant lignans have been studied in humans and rats, since ENL and END were fi rst isolated from the urine of humans in 1980.It appears that the metabolic fate of lignans in non-ruminant animals may include several steps.Firstly, after deglycosylation, SDG is transformed to secoisolariciresinol (SCEO), and then demethylated and dehydroxylated in reactions catalysed by intestinal bacteria.SCEO is subsequently converted to END, which is further dehydroxylated to form ENL.
In our study, the signifi cant increases in mammalian lignans were observed in both the rumen and blood circulation following SDGCE supplementation, suggested that ruminal microorganisms effi ciently converted SDG to enterolignans, which then were absorbed.Moreover, the observation that after a 14-d exposure to SDGCE, the area under the curve for ENL in serum was nearly twice that of END, may indicate that ENL is the main circulatory lignan in goats.Bowey et al. (2003) and Knust et al. (2006) have demonstrated that when SDG was offered to human or human fl ora-associated rats, a larger amount of ENL was formed, whereas Tan et al. (2004) found that the urinary END concentration in rats was nearly 10 times that of ENL.Lignan profi les may therefore be dependent on the composition of the dominant intestinal microbiota in different animals.
Thus far, there is still some disagreement about whether SDG can be absorbed directly into the circulation.Our data demonstrates that SDG was present in serum; therefore, it might be absorbed directly through rumen wall.This is agreement with previous study in non-ruminant animals (Knust et al., 2006).
When the data of SDGCE treatment period for 21 and 28 d were compared with that for 14 d of sampling, the levels of both SDG and the 2 mammalian lignans in the rumen and circulation were similar; this may indicate that after a 14-d exposure to SDGCE, the transformation and absorption of the SDG in goats had reached equilibrium of a relatively stable state.
The NH 3 -N concentration in ruminal fl uid was signifi cantly reduced by SDGCE, whereas the MCP level was remarkably increased.This contrasting response indicates that SDGCE may promote the utilization of non-protein nitrogen in the rumen.Simultaneously, the TVFA level and the (A+ B): P ratio in ruminal fl uid were signifi cantly increased by SDGCE, suggested that SDGCE may facilitate the anabolic metabolism of carbohydrates and change the fermentation pattern.These positive impacts on nitrogenous and anabolic metabolism were confi rmed by an in vitro study carried out by Wang et al. (2007).
As the present study is the fi rst to assess the effect of lignan supplementation on ruminants, there is currently a paucity of direct data relating to the physiological effects of SDG on rumen metabolism.Mao et al. (2007) demonstrated that daidzein, another important phytoestrogen, could dramatically change the acetate: propionate ratio (A:P), and that the amount of NH 3 -N in the rumen was signifi cantly reduced, whereas the MCP level tended to be higher.The causal mechanism appears to be the direct effect of daidzein on rumen microorganisms.As reviewed by Han (2006), in a study using water buffaloes fi tted with permanent rumen and intestine fi stulas, it was demonstrated that injection of daidzein (500 mg•d -1 , 12 d) via duodenal cannulae increased serum T, rumen bacterial protein, ammonia nitrogen and total VFA.Thus, the effects of SDGCE on ruminal metabolism may be due, in part at least, to linseed lignans and/or their metabolites.
Previous studies have demonstrated that isofl avones, particularly daidzein and genistein, could infl uence the hypothalamus-pituitary-sexual gland axis in male animals (Liu et al., 1996) and the hypothalamus-pituitary-thyroid axis in male chickens (Han and Wang, 1993), and increases the serum T and thyroid hormones levels, respectively.Since the 3 types of lignans were found in serum in this study, the elevated T, T 3 and T 4 levels in the serum might be the outcome of a similar series of reactions.Previous research has demonstrated that the T and thyroid hormones in circulation could enter the rumen with saliva or via the rumen epithelium (Zhengkang, 2006).The increase in T and T 3 levels in the rumen observed in this study may have been caused by the elevated T and T 3 levels in serum.Even though serum T 4 levels were almost 30 times that of T 3 , and low T 3 concentrations were found in the rumen, the concentrations of T 4 remained below the level of detection in ruminal fl uid throughout this experiment; these observations are consistent with the fi ndings of Hua and Han (1990).It has been reported that in fi lamentous fungi the transformation of T to E 2 may be catalysed by cytochrome P450 (Ahmed et al., 1996).This could explain the low level of E 2 found in the ruminal fl uid of male goats.The reduced E 2 concentration and increased T level in ruminal fl uid may have been caused by the inhibition of aromatase by END and ENL (Adlercreutz et al., 1993).

Figure 1 .
Figure 1.The concentrations of lignans in ruminal fl uid and serum in male goats.A, B, and Csecoisolariciresinol diglucoside (SDG), enterodiol (END) and enterolactone (ENL) concentrations, respectively, in the ruminal fl uid of goats fed a basal diet (BD) or BD + SDG crude extract (SDGCE).D -serum levels of SDG, END and ENL in goats with SDGCE treatment

Figure 2 .
Figure 2. Effects of secoisolariciresinol diglucoside crude extract (SDGCE) on the concentrations of hormones in male goats.A -effects of SDGCE on the concentrations of testosterone (T) in serum and ruminal fl uid.B -effect of SDGCE on the concentration of estradiol (E 2 ) in ruminal fl uid.Ceffects of SDGCE on the concentrations of 3, 3′, 5-triiodothyronine (T 3 ) in serum and ruminal fl uid.D -effect of SDGCE on the concentration of thyroxine (T 4 ) in serum