Dietary daidzein, but not genistein, has a hypocholesterolemic effect in non-ovariectomized and ovariectomized female Sprague-Dawley rats on a cholesterol-free diet

Abstract

We compared the effects of two major isoflavones, daidzein and genistein, on lipid metabolism in rats. Daidzein (150 mg/kg diet), genistein (150 mg/kg diet), daidzein and genistein (1:1, 300 mg/kg diet), or control diets were fed to 4 groups of 6-week-old ovariectomized (Ovx) and non-Ovx Sprague Dawley rats for 4 weeks. Dietary daidzein, but not genistein, reduced serum and hepatic total cholesterol levels significantly relative to that by the control group, regardless of whether the rats had undergone ovariectomy. Genistein did not exhibit any physiological effects on lipid levels, but did affect genes involved in cholesterol metabolism. These results indicate that daidzein and genistein may influence lipid regulation via differing modes of action.

Dietary daidzein, but not genistein, reduced serum and hepatic total cholesterol levels significantly relative to that by the control group, regardless of whether the rats had undergone ovaritectmy.

Key words:

• daidzein
• hypocholesterolemic effect
• ovariectomy
• gene expression
• ApoB

Hypercholesterolemia is associated with cardiovascular disease. Animal and human nutrition studies have shown that soy intake lowers blood lipid levels and reduces atherosclerosis.^1–3) The United States FDA has reported that a diet containing 25 g of soy per day may reduce the risk of coronary heart disease, and associated risks.^4,5) However, the soy components responsible for these effects and their underlying mechanisms of action are still unclear.

Dietary soy contains proteins, lipids, dietary fibers, and bioactive compounds including soy isoflavones. Daidzein and genistein classes of phytoestrogens are the predominant isoflavones in the soy. Daidzein is further metabolized into equal by intestinal enterobacteria. In fact, daidzein and genistein, are the major isoflavones, and their metabolites have been detected in the blood and urine of humans and animals.^6,7) Isoflavones^8–11) or soy protein^12–14) containing intact isoflavones have hypocholesterolemic effects in animal models. Soy protein containing isoflavones also affect hepatic lipid metabolism in experimental animals.^15,16) In a previous study of Sprague Dawley (SD) rats, we found that an AIN76 diet containing isoflavone-rich fermented soybean extract (FSBE) exhibited hypocholesterolemic effects in female rats with or without ovariectomy; we observed no such effect in male rats.¹⁷⁾ Lee et al. reported that female Golden Syrian hamsters fed a diet containing pure synthetic daidzein, genistein, or glycitein (0.9 mmol/kg diet) presented significantly lower total plasma and non-HDL cholesterol than control animals.¹⁸⁾ Balmir et al. reported that male SD rats and hamsters fed a diet containing isoflavone-rich soy protein isolate presented decreased total cholesterol and LDL levels. On feeding casein with supplemental isoflavones to the hamster group, the same authors found decreased levels of total and LDL cholesterol. These results indicate that isoflavones alone may lower cholesterol.¹⁹⁾

Isoflavones are estrogen receptor agonists, and genistein is more potent than daidzein.²⁰⁾ However, after digestion, enterobacteria convert daidzein into equol, which exhibits stronger estrogenic effects than genistein.^21,22) The present study investigated the soy extract components, daidzein and genistein, to confirm that they are responsible for the hypocholesterolemic effect on Ovx and non-Ovx female rats. In the previous study, we have found female-specific hypocholesterolemic effect in the rats fed the diet containing 155 mg/kg diet of genistein, 127 mg/kg diet of daidzein, and 18 mg/kg diet of glycitein.¹⁷⁾ Thus, we examined the hypocholesterolemic effect of dietary daidzein and genistein at a concentration of 150 mg/kg diet in Ovx and non-Ovx female rats. We also investigated gene expression of the key regulators of cholesterol and bile acid metabolism in the liver and small intestinal mucosa (upper and lower).

Materials and methods

Animals and diets

We purchased female Sprague-Dawley rats (Japan SLC, Hamamatsu, Japan; 6 weeks old). The experimental protocol was approved by the Laboratory Animal Care Committee of Ehime University, and the rats were maintained in accordance with the Guidance for the Care and Use of Laboratory Animals of Ehime University. All animals were housed individually in a room controlled by a 12-h light-dark cycler (dark phase: 15:00–3:00) at constant temperature (23 ± 1 °C) and humidity (55–65%). Animals were fed regular tap water and an isoflavone-free diet based on AIN-93G during the acclimatization and recovery period.

Experiment 1

After 7 d of acclimatization, the female rats were divided into 4 experimental groups with similar mean body weights (156 g) composed of 8 rats. Each group was given free access to the control diet (C diet) or a diet containing 150 mg daidzein, 150 mg genistein, or 300 mg mixture of daidzein and genistein per kg diet (D, G, or D + G diet) for 4 weeks. The C diet comprised (in g/kg): casein (New Zealand Dairy Board, Wellington, New Zealand), 200; cellulose (Danisco Japan., Inc., Tokyo, Japan), 50; soybean oil (J-oil Mills., Inc., Tokyo, Japan), 70; AIN-93 mineral mixture, 35; AIN-93 vitamin mixture, 10; l-cystine (Nacalai tesque, Kyoto, Japan), 3; sucrose (Nippon Beet Sugar Manufacturing, Tokyo, Japan), 100; and α-corn starch (Sanwa Starch, Nara, Japan), 532. The two major isoflavones, daidzein and genistein, were purchased from the LC laboratories (MA, USA). The additions of daidzein and genistein were performed at the expense of α-corn starch. During the experimental period, the body weight and amount of food consumed by weight were recorded each morning, and the respective diet was replenished. The rats were killed by decapitation, and a blood sample corresponding to a non-fasting state was collected from the neck at 20:00 h on the last day of the experimental period, identically to the condition in the previous study.¹⁷⁾ Subsequently, the serum was separated and extracted by centrifugation at 1,500 × g for 10 min at 4 °C, and stored at –50 °C until analysis. The liver was removed immediately after blood collection, weighed, and stored at –50 °C until analysis. Similarly, the small intestine (upper and lower) was excised immediately. After washing with ice-cold saline, the mucosa was scraped out and homogenized in Sepasol-RNA I Super G (Nacalai tesque, Kyoto, Japan), kept at room temperature for 5 min, and stored at –80 °C until further analysis. The perirenal, mesentery, and ovary fat tissues were removed and weighed: the sum of the weights is considered as the total white fat weight. During the final 3 d of the experimental period, feces were collected from each rat and freeze-dried, weighed, and stored in a desiccator until analysis.

Experiment 2

After 2 d of acclimatization, 35 rats were selected for the bilateral ovariectomy operation under sodium pentobarbital anesthesia (Kyoritsu Pharmaceutical Co., Ltd., Tokyo, Japan, 30 mg/kg of body weight, intraperitoneal injection). During this period, the operated rats were fed the control diet with casein. After 5 d recovery period, the Ovx rats were divided into 4 groups with a mean weight of 144 g (n = 8), and each group was given free access to the C diet or D, G, or D + G diets and water for 4 weeks. The present study did not have a sham-operated group, because serum and hepatic cholesterol concentrations of sham-operated female rats were not appreciably different from those of intact female rats in our previous study.¹⁷⁾ At the end of the experimental period, the rats were decapitated, as described in experiment 1.

Lipid analysis

Total serum cholesterol and triglyceride concentrations were measured by absorbance using commercially available colorimetric cholesterol and triglyceride reagents (Cholesterol E-test Wako and TG E-test Wako, respectively; Wako Pure Chemical Industries Co., Osaka, Japan).

Total liver lipid was extracted from 1 g liver with 25 mL of chloroform:methanol (2:1) and gravimetrically measured by the method of Folch et al.²³⁾ Liver cholesterol (total, free, and esterified) and triglyceride concentrations were enzymatically determined with commercial diagnostic kits (Cholesterol E-test Wako, Free-Cholesterol E-test Wako, TG E-test Wako, respectively; Wako Pure Chemical Industries Co., Osaka, Japan.) by taking 5 mL sample from the adjusted 25 mL liver lipid extract. For the cholesterol test, 200 μL of this extract was evaporated, and the residue was mixed with 200 μL of propane-2-ol containing triton X 100 (9:1 v/v). From this mixture, 100 μL was taken and mixed with 900 μL coloring solution and incubated at 37 °C for 30 min. After which, 400 μL chloroform was added, the solution was vortexed and centrifuged at 1160 × g for 5 min at 4 °C, and the upper layer was decanted for cholesterol profiling. Cholesterol esters were quantitated by subtracting free cholesterol from total cholesterol.

Fecal bile acid was extracted from 0.05 g feces with 5 mL of chloroform:methanol (1:1) at 70 °C for 60 h. The bile acid level was enzymatically analyzed by the 3-β-hydroxysteroid dehydrogenase assay method of Sheltaway and Losowsky, using taurocholic acid as the standard.²⁴⁾

Total fecal lipids were extracted from 0.5 g feces with 20 mL of chloroform:methanol (2:1) and gravimetrically measured, as described by Folch et al.²³⁾

RNA isolation and cDNA synthesis

Total RNA was isolated from the liver and small intestinal mucosa (upper and lower) as per the method described by Chomczynski and Sacchi.²⁵⁾ mRNA was isolated from total RNA using Oligotex-dT30 (Takara Bio, Inc., Shiga, Japan) to synthesize cDNA using reverse transcriptase (Reverse Transcriptase XL [AMV] and RT-PCR, 5 U/μL, Takara Bio Inc., Shiga, Japan) a thermal cycler (ABI GeneAmp2400; PerkinElmer, Inc., Waltham, USA). mRNA expression was determined by quantitative PCR using the StepOnePlus real-time PCR system (Applied Biosystems, Carlsbad, California, USA). The primer sets of cholesterol 7α-hydroxylase (CYP7A1), sterol 12-alpha-hydroxylase (CYP8B1), sterol 27hydroxylase (CYP27), low-density lipoprotein receptor (LDLR), hydroxymethylglutaryl-coenzyme reductase (HMG-CoA-R), apolipoprotein-B (Apo-B), liver X receptor (LXR), farnesoid X receptor (FXR), acyl-CoA:cholesterol acyl transferase (ACAT), ileal sodium-dependent bile acid transporter (IBAT), microsomal triglyceride transfer protein (MTP), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and cyclophilin (Table 1) were used with the THUNDERBIRD SYBR qPCR mix (TOYOBO, Osaka, Japan). The basic amplification program was set to 50 cycles: denaturizing at 95 °C for 15 s, annealing and primer extension at 60 °C for 1 min. Fluorescence was recorded at 530 nm during extension. Relative mRNA expression was calculated using the crossing point of each target gene, with the GAPDH or cyclophilin gene as the reference, and the corresponding real-time PCR efficiency of respective primer sets as per the method previously reported.²⁶⁾

Table 1. The primers used for reverse transcription-polymerase chain reaction (RT-PCR) analysis.

Genes		Primer sequence	PCR product (bp)
CYP7A1	Sense	CTG CAA ACT GAT GGG GAA AT	145
	Antisense	TGG GTC AAT GCT TCT GTG TC
CYP8B1	Sense	ATG AAG GCT GTG CGA GAG	127
	Antisense	TCT CTT CCA TCA CGC TGT C
CYP27	Sense	CTC AAG AGA CTG TCG GCA C	114
	Antisense	CTT CCT GGA TCT CTG GGT T
LDLR	Sense	CTT GCC CTG ATG GTA TGC TA	137
	Antisense	CTT GCG CTT CAG TGA CAC A
HMG-CoA-R	Sense	GCT GGT GAG TTG TCC TTG AT	117
	Antisense	CTT CTT GGT GCA TGT TCC CT
Apo-B	Sense	TTG ACA CAC TGA AGT TCC TA	124
	Antisense	ACA TCA AAC CCT GGT ATT AG
LXR	Sense	TGC TAA TGA AGC TGG TGA	144
	Antisense	AGA GGA CAC GGA GAA ACA GT
FXR	Sense	CGT TCA GCG GAG ATT TTC AA	140
	Antisense	ATT TTC AGC TCC CCG ACA CT
ACAT2	Sense	CGGTCATGCTGATCCTCTTT	204
	Antisense	GTGTCACCAGCTCCCAAAAT
IBAT	Sense	TGC CTG GAT CAT TGA ACC	149
	Antisense	GAG TGT TCT GCA TTC CAG TT
MTP	Sense	AGCGACATCACAGTGGACTC	223
	Antisense	GGCTCTCTTTTCTTCTCCGA
GAPDH	Sense	GCCATCAACGACCCCTTCATT	240
	Antisense	CGCCTGCTTCACCACCTTCTT
Cyclophilin	Sense	ATGGTCAACCCCACCGTGTT	206
	Antisense	CGTGTGAAGTCACCACCCT

Notes: CYP7A1, cholesterol 7α-hydroxylase; CYP8B1, sterol 12-alpha-hydroxylase; CYP27, sterol 27-hydroxylase; LDLR, low density lipoprotein receptor; HMG-CoA-R, hydroxymethylglutaryl-coenzyme reductase; Apo-B, apolipoprotein-B; LXR, liver X receptor; FXR, farnesoid X receptor; ACAT, acyl-CoA:cholesterol acyl transferese; IBAT, ileal sodium-dependent bile acid transporter; MTP, Microsomal triglyceride transfer protein; GAPDH, Glyceraldehyde 3-phosphate dehydrogenase.

Statistical analysis

All data are presented as the mean ± standard error (SEM). The main effects and associations between daidzein administration (dietary daidzein), genistein administration (dietary genistein), and the operation (Ovx) were assessed by three-way analysis of variance (ANOVA). If the ANOVA result indicated a significant interaction, we proceeded to post hoc tests. Comparisons between each group were made using an unpaired Student’s t-test with Bonferroni corrections. All statistical tests were performed using the IBM SPSS Statistics software package (SPSS Japan Inc., an IBM company). Statistical tests with p < 0.05 were considered significant.

Results

Growth parameters, serum lipid concentration, and fat tissue weight

The final body weight, body weight gain, and food intake in rats fed the D diet were significantly lower than those in rats fed the C diet only in the Ovx group; a significant three-way interaction was observed (Table 2). Dietary daidzein significantly decreased serum cholesterol and triglyceride concentrations in the female rats, regardless of whether they had undergone ovariectomy, although Ovx significantly increased serum triglyceride concentrations (Table 2). Furthermore, Ovx significantly decreased the ovary fat weight, and increased the perirenal fat weight and total white fat weight, while dietary daidzein significantly reduced the perirenal fat weight and the total white fat weight (Table 3). The mesentery fat weight in rats fed D diets was significantly lower than that in rats fed C diets only in the Ovx group; a significant interaction between dietary daidzein and the operation was observed (Table 3).

Table 2. Effect of daidzein and genistein on body weight, body weight gain, and food intake and serum lipid concentration.^a

		Final body weight^f	Body weight gain^f	Food intake^f	Serum lipid concentration
					Total cholesterol	Triglycerides
		(g)	(g/28)	(g/28)	(mM)
Non-Ovx^b	Control	238 ± 7	82 ± 6	421 ± 10	2.58 ± 0.08	0.94 ± 0.17
	Daidzein	242 ± 6	86 ± 6	431 ± 10	2.31 ± 0.08	0.86 ± 0.15
	Genistein	247 ± 5	91 ± 4	446 ± 10	2.56 ± 0.10	0.95 ± 0.15
	Daidzein and Genistein	239 ± 7	82 ± 6	415 ± 16	2.20 ± 0.09	0.87 ± 0.14
Ovx^c	Control	316 ± 10	178 ± 9	627 ± 25	2.72 ± 0.09	1.32 ± 0.12
	Daidzein	272 ± 7*	133 ± 6*	534 ± 19*	2.24 ± 0.15	1.20 ± 0.07
	Genistein	306 ± 8	168 ± 8	604 ± 20	2.87 ± 0.16	1.42 ± 0.13
	Daidzein and Genistein	289 ± 8	151 ± 9	583 ± 14†	2.39 ± 0.14	0.99 ± 0.10
Three-way ANOVA^d^,^e	p-value
Main effects	OP	<0.01	<0.01	<0.01	NS	<0.01
	D	<0.01	<0.01	<0.01	<0.01	<0.05
	G	NS	NS	NS	NS	NS
Interaction	OP*D	<0.01	<0.01	<0.05	NS	NS
	OP*G	NS	NS	NS	NS	NS
	D*G	NS	NS	NS	NS	NS
	OP*D*G	<0.05	<0.05	<0.05	NS	NS

^a Data are expressed as mean ± SEM (n = 8).

^b Non-Ovx, non-ovariectomized.

^c Ovx, ovariectomized.

^d D, effect of dietary daidzein; G, effect of dietary genistein; OP, effect of operation (Non-Ovx and Ovx).

^e NS, Not significant.

^f Asterisks show a significant difference compared with the control group, and daggers show a significant difference when compared with daidzein group, determined by three-way ANOVA with Bonferroni’s test (p < 0.05).

Table 3. Effect of daidzein and genistein on total white fat (ovary, mesentery, perirenal fat).^a

		Ovary fat	Mesentery fat^f	Perirenal fat	Total white fat
		(g)
Non-Ovx^b	Control	5.7 ± 0.4	4.7 ± 0.3	5.8 ± 0.4	16.3 ± 1.0
	Daidzein	5.6 ± 0.6	4.7 ± 0.2	5.2 ± 0.4	15.5 ± 1.0
	Genistein	6.7 ± 0.4	5.4 ± 0.3	6.2 ± 0.4	18.3 ± 0.8
	Daidzein and Genistein	5.8 ± 0.8	4.7 ± 0.4	5.2 ± 0.8	15.8 ± 1.9
Ovx^c	Control	4.9 ± 0.3	7.9 ± 0.5	8.7 ± 0.6	21.5 ± 0.9
	Daidzein	3.5 ± 0.6	5.4 ± 0.4*	6.1 ± 0.7	15.0 ± 1.5
	Genistein	4.3 ± 0.5	7.7 ± 0.6	9.4 ± 1.3	21.4 ± 2.2
	Daidzein and Genistein	4.6 ± 0.4	6.7 ± 0.6†	7.7 ± 0.7	19.0 ± 1.4
Three-way ANOVA^d^,^e	p-value
Main effects	OP	<0.01	<0.01	<0.01	<0.01
	D	NS	<0.01	<0.01	<0.01
	G	NS	NS	NS	NS
Interaction	OP*D	NS	<0.05	NS	NS
	OP*G	NS	NS	NS	NS
	D*G	NS	NS	NS	NS
	OP*D*G	NS	NS	NS	NS

^a Data are expressed as mean ± SEM (n = 8).

^b Non-Ovx, non-ovariectomized.

^c Ovx, ovariectomized.

^d D, effect of dietary daidzein; G, effect of dietary genistein; OP, effect of operation (Non-Ovx and Ovx).

^e NS, Not significant.

^f Asterisks show a significant difference compared with the control group, and daggers show a significant difference when compared with daidzein group, determined by three-way ANOVA with Bonferroni’s test (p < 0.05).

Liver lipid concentration

Along with lowering serum lipid levels, daidzein also decreased hepatic lipid concentrations (Table 4). Total liver wet weight and phospholipid content in rats fed the D diet was significantly lower than that in rats fed the C diet only in the Ovx group; a significant interaction between dietary daidzein and the operation was observed (Table 4). Ovx resulted in a significant increase in the total liver lipid content, hepatic triglyceride content, and hepatic total and esterified cholesterol contents, and a nonsignificant increase in free cholesterol content (main effect, p = 0.05). Dietary daidzein significantly decreased hepatic triglycerides content, and total and free cholesterol contents, regardless of whether the rats had undergone ovariectomy (Table 4). On the other hand, genistein decreased the hepatic free cholesterol and increased the esterified cholesterol, regardless of ovariectomy status (Table 4). Daidzein did not exhibit any apparent effect on liver esterified cholesterol concentration.

Table 4. Effect of daidzein and genistein on liver wet weight, total lipids, cholesterol and triglyceride concentration; fecal bile and lipid excretion.^a

						Cholesterol				Fecal excretion^f
		Total liver wet weight^f	Total lipid	Phospholipid^f	Triglycerides	Total	Free	Esterified		Bile acid		Lipid
		(g)	(mg/liver)	(μmol/liver)					(μmol/day)		(mg/day)
Total liver wet weight^f	Total lipid	Phospholipid^f	Triglycerides	Total	Free	Esterified		Bile acid		Lipid
(g)	(mg/liver)	(μmol/liver)					(μmol/day)		(mg/day)
Non Ovx^b	Control	7.1 ± 0.3	390 ± 33	236 ± 11	110 ± 27	42 ± 2	38 ± 1	3.2 ± 0.8		10.5 ± 1.5		18.1 ± 1.3
	Daidzein	7.4 ± 0.3	317 ± 16	231 ± 9	56 ± 9	37 ± 2	32 ± 2	4.4 ± 0.6		21.1 ± 2.3*		25.2 ± 1.7*
	Genistein	7.5 ± 0.3	354 ± 16	227 ± 6	89 ± 9	43 ± 1	36 ± 2	7.0 ± 2.0		12.8 ± 2.4		21.9 ± 1.5
	Daidzein and genistein	7.1 ± 0.3	316 ± 19	210 ± 12	81 ± 13	38 ± 2	30 ± 1	7.3 ± 1.6		26.9 ± 4.7#		24.5 ± 2.2
Ovx^c	Control	10.3 ± 0.6	547 ± 66	314 ± 18	197 ± 25	51 ± 3	43 ± 3	8.7 ± 1.0		2.9 ± 0.6		31.9 ± 2.0
	Daidzein	8.4 ± 0.4*	431 ± 51	248 ± 11*	149 ± 27	43 ± 2	34 ± 2	8.3 ± 1.6		16.3 ± 0.8*		24.8 ± 1.4*
	Genistein	9.8 ± 0.5	518 ± 59	291 ± 15	208 ± 54	45 ± 4	37 ± 2	8.6 ± 2.0		11.3 ± 1.0 *		30.8 ± 2.3
	Daidzein and genistein	8.9 ± 0.4	469 ± 40	255 ± 13 #	165 ± 23	44 ± 3	33 ± 1	11.2 ± 2.3		30.4 ± 2.2#,†		34.2 ± 2.3†
Three way ANOVA^d^,^e					p-value
Main effects	OP	<0.01	<0.01	<0.01	<0.01	<0.01	0.05	<0.01		NS		<0.01
	D	<0.01	NS	<0.01	<0.05	<0.01	<0.01	NS		<0.01		NS
	G	NS	NS	NS	NS	NS	<0.05	<0.05		<0.01		<0.05
Interaction	OP*D	<0.05	NS	<0.05	NS	NS	NS	NS		NS		<0.01
	OP*G	NS	NS	NS	NS	NS	NS	NS		<0.05		NS
	D*G	NS	NS	NS	NS	NS	NS	NS		NS		NS
	OP*D*G	NS	NS	NS	NS	NS	NS	NS		NS		<0.01

^a Data are expressed as mean ± SEM (n = 8).

^b Non-Ovx, non-ovariectomized.

^c Ovx, ovariectomized.

^d D, effect of dietary daidzein; G, effect of dietary genistein; OP, effect of operation (Non-Ovx and Ovx).

^e NS, Not significant.

^f Asterisks show a significant difference compared with the control group, daggers show a significant difference when compared with daidzein group, and number signs show a significant difference when compared with genistein group, determined by three-way ANOVA with Bonferroni’s test (p < 0.05).

Fecal bile acid and lipid excretion

Dietary daidzein significantly increased the fecal bile acid excretion in both Ovx and non-Ovx rats. In contrast, dietary genistein significantly increased the fecal bile acid in Ovx rats only with operational effect (Table 4). The amount of fecal lipid excretion was significantly higher in non-Ovx rats, whereas, the amount was significantly lower in the Ovx rats fed with dietary daidzein; we observed a three-way interaction (Table 4).

Lipid metabolism and transportation-mediated gene expression in the liver and small intestine

The present study confirmed that dietary daidzein significantly reduces serum and hepatic lipid levels in both Ovx and non-Ovx female rats. To determine the underlying mechanism of the cholesterol lowering effect of daidzein, we investigated the expression of genes responsible for cholesterol metabolism in the liver and small intestine. We found that daidzein intake was not associated with any changes in the mRNA expression, except for the expression of ApoB mRNA. Daidzein intake decreased ApoB expression in the liver, regardless of whether the rats had undergone ovariectomy (Table 5). However, genistein influenced some of the genes involved in cholesterol regulation. Dietary genistein intake tended to increase the ACAT2 mRNA expression (main effect, p = 0.05) and CYP7A1 mRNA expression (main effect, p = 0.07), and significantly upregulated the FXR mRNA expression in the liver (Table 5). Dietary genistein significantly suppressed MTP mRNA expression in the upper intestine of Ovx rats; a significant interaction between dietary genistein intake and Ovx was observed (Table 6). Genistein intake exhibited an operational effect by upregulating lower small intestinal FXR gene expression in non-Ovx rats (Table 6).

Table 5. Effect of daidzein and genistein on the mRNA expression for lipid metabolism genes in the liver.^a,f

		Apo-B	ACAT2	LDLR	HMG-CoA-R	CYP7A1	CYP8B1	CYP27	LXR	FXR
		(×10^–2)	(×10^–2)	(×10^–3)	(×10^−3)	(×10^−2)	(×10^−1)	(×10^−1)	(×10^−1)	(×10^−2)
Non OVX^b	Control	0.9 ± 0.2	2.5 ± 0.4	1.5 ± 0.2	8.5 ± 0.9	1.1 ± 0.2	1.2 ± 0.2	1.2 ± 0.2	1.4 ± 0.2	3.1 ± 0.4
	Daidzein	0.7 ± 0.1	2.5 ± 0.2	1.7 ± 0.2	7.7 ± 0.8	0.9 ± 0.1	0.9 ± 0.1	1.2 ± 0.2	1.2 ± 0.1	3.2 ± 0.2
	Genistein	0.8 ± 0.1	3.1 ± 0.3	1.3 ± 0.3	7.5 ± 0.6	1.3 ± 0.2	1.1 ± 0.1	1.4 ± 0.1	1.4 ± 0.1	3.7 ± 0.3
	Daidzein and genistein	0.8 ± 0.1	3.5 ± 0.2	1.7 ± 0.2	7.7 ± 0.5	1.4 ± 0.2	0.8 ± 0.1	1.3 ± 0.2	1.3 ± 0.1	3.8 ± 0.2
OVX^c	Control	2.2 ± 0.3	4.5 ± 0.5	3.1 ± 0.6	9.7 ± 1.4	1.9 ± 0.4	3.0 ± 0.7	2.3 ± 0.1	2.2 ± 0.2	5.5 ± 0.6
	Daidzein	1.8 ± 0.2	4.6 ± 0.9	4.0 ± 0.8	11.6 ± 1.3	2.0 ± 0.3	2.4 ± 0.2	2.1 ± 0.1	2.1 ± 0.1	5.1 ± 0.5
	Genistein	1.9 ± 0.1	5.0 ± 0.7	3.0 ± 0.4	10.2 ± 2.3	2.2 ± 0.5	2.0 ± 0.3	2.4 ± 0.1	2.4 ± 0.1	5.8 ± 0.6
	Daidzein and genistein	1.5 ± 0.1	5.4 ± 0.6	3.1 ± 0.5	9.6 ± 1.5	2.7 ± 0.5	2.0 ± 0.3	2.2 ± 0.2	2.2 ± 0.1	6.7 ± 0.8
Three-way ANOVA^d^,^e	p-value
Main effects	OP	<0.01	< 0.01	<0.01	<0.05	<0.01	<0.01	<0.01	<0.01	<0.01
	D	<0.05	NS	NS	NS	NS	NS	NS	NS	NS
	G	NS	0.05	NS	NS	0.07	NS	NS	NS	<0.05
Interaction	OP*D	NS	NS	NS	NS	NS	NS	NS	NS	NS
	OP*G	NS	NS	NS	NS	NS	NS	NS	NS	NS
	D*G	NS	NS	NS	NS	NS	NS	NS	NS	NS
	OP*D*G	NS	NS	NS	NS	NS	NS	NS	NS	NS

^a Data are expressed as mean ± SEM (n = 8).

^b Non-Ovx, non-ovariectomized.

^c Ovx, ovariectomized.

^d D, effect of dietary daidzein; G, effect of dietary genistein; OP, effect of operation (Non-Ovx and Ovx).

^e NS, Not significant.

^f Reference gene was cyclophilin. CYP7A1, cholesterol 7α-hydroxylase; CYP8B1, sterol 12-alpha-hydroxylase; CYP27, sterol 27-hydroxylase; LDLR, low density lipoprotein receptor; HMG-CoA-R, hydroxymethylglutaryl-coenzyme reductase; Apo-B, apolipoprotein-B; LXR, liver X receptor; FXR, farnesoid X receptor; ACAT, acyl-CoA:cholesterol acyl transferese.

Table 6. Effect of daidzein and genistein on the mRNA expression for lipid metabolism genes in the small intestinal mucosa.^a,g

		IBAT	HMG-CoA-R	MTP^f	FXR	LXR
Upper intestinal mucosa		(×10^−3)	(×10^−3)	(×10^−1)	(×10^−2)	(×10^−2)
Non Ovx^b	Control	2.7 ± 0.5	5.0 ± 0.3	4.9 ± 0.7	0.9 ± 0.1	5.9 ± 0.3
	Daidzein	2.0 ± 0.5	5.7 ± 0.7	6.7 ± ± 2.5	1.0 ± 0.2	8.1 ± 2.1
	Genistein	3.1 ± 0.5	5.4 ± 0.6	4.8 ± 0.4	0.9 ± 0.1	5.6 ± 0.2
	Daidzein and Genistein	3.1 ± 0.6	5.1 ± 0.4	5.4 ± 0.8	1.1 ± 0.2	6.4 ± 0.2
Ovx^c	Control	0.9 ± 0.1	4.3 ± 1.4	2.1 ± 0.5	1.0 ± 0.6	3.4 ± 0.7
	Daidzein	0.8 ± 0.1	5.0 ± 2.0	2.2 ± 0.6	0.9 ± 0.6	2.8 ± 0.5
	Genistein	0.6 ± 0.1	2.5 ± 0.4	1.2 ± 0.2*	0.3 ± 0.1	4.2 ± 2.2
	Daidzein and Genistein	1.0 ± 0.2	3.3 ± 0.4	1.2 ± 0.2^†	0.4 ± 0.1	2.5 ± 0.2
Three-way ANOVA^d^,^e	p-value
Main effects	OP	<0.01	<0.05	<0.01	NS	<0.01
	D	NS	NS	NS	NS	NS
	G	NS	NS	<0.05	NS	NS
Interaction	OP*D	NS	NS	NS	NS	NS
	OP*G	NS	NS	<0.05	NS	NS
	D*G	NS	NS	NS	NS	NS
	OP*D*G	NS	NS	NS	NS	NS
		IBAT	HMG-CoA-R	MTP	FXR^f	LXR
Lower intestinal mucosa		(×10^−3)	(×10^−3)	(×10^−1)	(×10^−2)	(×10^−2)
Non Ovx^b	Control	2.3 ± 0.4	14.7 ± 3.1	26.4 ± 5.7	4.0 ± 0.4	30.7  ± 4.1
	Daidzein	2.0 ± 0.3	13.0 ± 2.1	26.7 ± 3.6	4.1 ± 0.5	29.6  ± 2.6
	Genistein	2.4 ± 0.3	14.5 ± 2.5	34.8 ± 3.2	5.2 ± 0.4	39.4  ± 4.2
	Daidzein and Genistein	3.4 ± 0.6	18.4 ± 3.1	33.4 ± 5.1	6.0 ± 0.9^†	34.2  ± 4.5
Ovx^c	Control	0.3 ± 0.1	3.4 ± 0.5	10.3 ± 1.8	1.6 ± 0.3	18.2 ± 3.9
	Daidzein	0.4 ± 0.1	4.8 ± 0.9	8.6 ± 1.2	1.4 ± 0.3	16.5 ± 4.3
	Genistein	0.4 ± 0.1	4.2 ± 0.5	7.0 ± 1.5	1.4 ± 0.3	15.3 ± 3.0
	Daidzein and Genistein	0.4 ± 0.1	4.0 ± 0.7	7.0 ± 1.6	1.5 ± 0.4	15.3 ± 2.5
Three-way ANOVA^d^,^e	p-value
Main effects	OP	<0.01	<0.01	<0.01	<0.01	<0.01
	D	NS	NS	NS	NS	NS
	G	NS	NS	NS	<0.05	NS
Interaction	OP*D	NS	NS	NS	NS	NS
	OP*G	NS	NS	<0.05	<0.05	NS
	D*G	NS	NS	NS	NS	NS
	OP*D*G	NS	NS	NS	NS	NS

^a Data are expressed as mean ± SEM (n = 8).

^b Non-Ovx, non-ovariectomized.

^c Ovx, ovariectomized.

^dD, effect of dietary daidzein; G, effect of dietary genistein; OP, effect of operation (Non-Ovx and Ovx).

^e NS, Not significant.

^fAsterisks show a significant difference compared with the control group, daggers show a significant difference when compared with daidzein group, and number signs show a significant difference when compared with genistein group, determined by three-way ANOVA with Bonferroni’s test (p < 0.05).

^g Reference gene was GAPDH. IBAT, ileal sodium-dependent bile acid transporter; HMG-CoA-R, hydroxymethylglutaryl-coenzyme reductase; MTP, Microsomal triglyceride transfer protein; FXR, farnesoid X receptor; LXR, liver X receptor; GAPDH, Glyceraldehyde 3-phosphate dehydrogenase.

All hepatic genes were relatively highly expressed in Ovx rats (Table 5), whereas the expression of upper and lower small intestinal genes were relatively low in Ovx rats, except for the upper small intestinal FXR (Table 6). Significant interactions were observed between dietary genistein and the operation in the upper small intestinal MTP and lower small intestinal MTP and FXR gene expression (Table 6).

Discussion

Dietary daidzein, but not genistein, significantly reduced serum cholesterol in rats, regardless of ovariectomy status (Table 2). Our previous experiments on isoflavone-rich FSBE showed that FSBE decreased serum cholesterol in Ovx and non-Ovx rats, but not in males.¹⁷⁾ Dietary daidzein decreased the food intake only in Ovx rats (Table 2). We reported previously that the serum cholesterol concentration of rats pair-fed a diet containing FSBE was not significantly different from that in rats fed the control diet.¹⁷⁾ The decreased food intake in rats fed the D diet might not simply be associated with the hypocholesterolemic effect. Relative to daidzein, genistein reportedly has greater affinity for the estrogen receptors and stronger inhibitory effect on the activity of tyrosine kinase, which affects the regulatory capacity of lipids.^20,27) Takahashi et al. showed that genistein was more effective than daidzein at lowering serum triglycerides when administered to SD male rats at dietary doses of 1 or 2 g/kg.²⁸⁾ However, concentrations of genistein in these literal data might be much higher than the nutritional levels used in the present study. Firstly, these could be approximately 5 or 10-fold higher than that of the AIN-93G-based diet that casein were replaced by soy protein isolate, which used as isoflavones source in many nutritional reports, as the sole protein source.²⁹⁾ Secondly, these are much higher than the value estimated from Japanese consumption of soy isoflavones, which is estimated at very high level compared to other countries. For example, for daidzein, mean intake estimated from food-frequency questionnaires and dietary records was 18.3 mg/d and for genistein, that was 31.4 mg/d.³⁰⁾ The daily total isoflavones to total nutrient intake (426.9 g/d) is recalculated as 43 and 74 mg/kg nutrient.

The hypocholesterolemic effect might be exhibited by the metabolite of daidzein, i.e. equol. Equol has higher binding affinity for ERα and antioxidant activity than genistein and daidzein.^21,31) In female rats fed a diet containing FSBE, the serum equol concentration was several fold higher than that of genistein both in Ovx and non-Ovx rats.³²⁾ Furthermore, serum equol concentration was more than threefold higher than that of daidzein in non-Ovx rats fed a diet containing 150 mg/kg daidzein.³³⁾ Unpublished results of experiments on Ovx and non-Ovx rats showed that the serum equol concentration was more than 2000 nmol/L in rats fed a diet containing daidzein, regardless of whether the rats had undergone ovariectomy. Thus, the stronger estrogenic and antioxidative activities and higher concentrations could be responsible for the physiological effects of equol compared to genistein in rats. In addition, dietary equol administration also reportedly reduced plasma total cholesterol in Ovx rats,³⁴⁾ although the dose of equol seems to be too high (400 mg/kg diet) to explain the contribution of equol to hypocholesterolemic effect of daidzein in the present study.

The reduction of serum cholesterol is associated with stimulation of bile acid excretion and bile acid synthesis in the liver, catalyzed by the cholesterol 7-α hydroxylase (CYP7A1), the rate-limiting enzyme in the classical bile acid synthesis pathway.^35,36) In the present study, however, daidzein increased fecal bile excretion in both Ovx and non-Ovx rats but did not affect the hepatic mRNA for cytochrome P450 enzymes (CYP7A1, CYP27, and CYP8B1) and their promoter, FXR (Table 5). Increased levels of bile acids decrease mRNA expression of HMG-CoA-R, the rate-limiting enzyme in cholesterol biosynthesis, and increase mRNA expression of LDLR.^36,37) However, daidzein did not affect the mRNA expression of HMG-CoA-R and LDLR either in the present study (Table 5). Our previous study showed that dietary FSBE increased the hepatic LDLR gene expression in non-Ovx rats but not Ovx rats.¹⁷⁾ Other functional components of FSBE might be associated with an increased expression of LDLR mRNA. It is not clear whether increased hepatic LDLR gene expression contributed to hypocholesterolemic effect of daidzein. As such, further studies will be needed. The increased fecal bile excretion can also be correlated with the lower reabsorption rate of bile salts from the intestine back into the bloodstream for hepatic uptake, mediated by the IBAT. Ross-Viola et al. found that the IBAT mRNA levels were reduced by 55% in both liver and intestinal samples by feeding mice isoflavone at high and low doses.¹⁶⁾ In the present study, however, daidzein did not change the IBAT mRNA expression in the small intestine. In addition, we reported previously that dietary FSBE did not affect the fecal bile acid excretion, despite decreasing serum cholesterol concentration.¹⁷⁾ In the present study also, there was no significant correlation between the serum total cholesterol and fecal bile excretion (data not shown), suggesting that an increase in the fecal bile excretion caused by daidzein may not contribute to the decrease in cholesterol synthesis and clearance of serum cholesterol. As we did not analyze the fecal steroids of all feeding period (only 3 days), we could not fully discuss about the fecal steroids balance. Further studies are needed to determine whether the fecal bile excretion caused by daidzein contribute to the decreases in serum and hepatic cholesterol by examining the time-dependent changes in the fecal steroid levels, and the activity and gene expressions of enzymes and transporters involved in the regulation of whole body cholesterol balance.

Another potential mechanism of reducing serum and hepatic cholesterol would involve the decreasing ApoB mRNA expression in the liver.³⁸⁾ The reduced ApoB secretion would reduce the assembly and secretion of VLDL in the liver, and ultimately lower the LDL cholesterol level. Clarkson et al. also showed that an increased serum LDL-cholesterol concentration in Ovx rats was due to increased VLDL assembly and secretion caused by increased lipoprotein lipase activity.³⁹⁾ In the present study, daidzein significantly decreased ApoB mRNA expression (Table 5), which may be linked to decreased serum cholesterol levels in both Ovx and non-Ovx female rats. However, we do not have evidence to support the hypothesis that ApoB is the main factor in reducing the serum and hepatic cholesterol levels.

On the other hand, genistein did not show any lowering effect on serum and hepatic lipid levels (Tables 2 and 4). However, genistein significantly decreased hepatic free cholesterol, and significantly increased esterified cholesterol, in both Ovx and non-Ovx female rats (Table 4). These effects could be explained by an increased trend of hepatic ACAT2 gene expression in both Ovx and non-Ovx female rats (main effect, p = 0.05; Table 5). ACAT-2 is a major cholesterol esterification enzyme and may control the amount of hepatic-free cholesterol available for secretion into bile acid pool or into the VLDL formation.^40,41) Dietary genistein significantly increased fecal bile acid excretion (Table 4). Consistent with an increased fecal bile acid excretion, an increased trend by genistein intake was observed in hepatic CYP7A1 gene expression (main effect, p = 0.07; Table 5). It may be due to the acceleration of the bile acid excretion and bile acid synthesis induced by genistein intake. However, dietary genistein did not alter serum and hepatic total cholesterol levels (Tables 2 and 4). Therefore, we speculated that the balance of steroids could be maintained by the acceleration in de novo synthesis of cholesterol. This speculation does not necessarily contradict the result that HMG-CoA-R gene expression was not changed by dietary genistein in this study, because regulation of this enzyme is achieved by not only transcriptional control, but also control of phosphorylation.⁴²⁾ We consider that FXR gene expression might be up-regulated in response to an increase in bile acid levels.⁴³⁾ The accelerating effect of genistein on bile acid synthesis, caused by the up-regulation of CYP7A1 gene expression, might surpass FXR responses.

In conclusion, dietary daidzein, but not genistein, reduced the serum and hepatic total cholesterol levels of Ovx and non-Ovx female rats. The decrease in serum cholesterol by daidzein could be explained by the decreasing effect of ApoB mRNA expression. However, we found no evidence to support this hypothesis. Genistein, on the other hand, significantly did not show any lowering effect on serum and hepatic lipid levels. The hypocholesterolemic effect might be exhibited by the metabolite of daidzein, equol, which has strong estrogenic and antioxidative activities and high bioavailability. Further investigation at the molecular level is required to understand the effects of dietary daidzein and equol on cholesterol regulation.

Author contribution

KB conducted the experiments, analyzed the results, and wrote the manuscript with the aids by SA. MF and TK contributed to analysis and interpretation of data, and assisted in the preparation of the manuscript. All the authors reviewed the results and approved the final version of the manuscript.

Funding

This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology, Japan, under a Monbukagakusho scholarship.

Disclosure statement

No potential conflict of interest was reported by the authors.

Acknowledgement

We also thank the members of the Fujitani and Kishida groups for their helpful contributions in dissection.