A total of 42 male Sprague-Dawley rats (3-weeks-old, 54–67 g) were purchased from Japan SLC, Inc. (Hamamatsu, Japan). The animals were maintained according to the ‘Guide for the Care and Use of Laboratory Animals’ established by Fuji Women's University (Ishikari, Japan), and the study was approved by the university's ethics committee. The rats were individually housed in a room with a controlled temperature (22–24°C), relative humidity (55–65%) and under a 12-h light/dark cycle (light from 08:00 a.m. to 08:00 p.m.). Following ad libitum access to a non-purified commercial rodent powder diet (CE-2; CLEA Japan, Inc., Tokyo, Japan) for 3 days, the rats (78–90 g) were assigned to weight-matched groups (n=6 rats per group) and given ad libitum access to experimental diets and deionized water. Food intake and body weight were measured daily at 08:00 a.m. In experiment 1, the rats from four groups were fed an HSR or HSC diet with or without 1.02% dodecasodium phytate (sodium PA; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) for 12 days. In experiment 2, the rats from three groups were fed the HSC diet or HSC diets with 1.02% sodium PA or 0.2% MI (Wako Pure Chemical Industries, Ltd., Osaka, Japan) for 12 days. Table I lists the compositions of the experimental diets (17). The molar concentration of the added MI was equivalent to that of the added PA. Feces were collected over the last 3 days of the experiments, stored at −20°C, and subsequently freeze-dried and milled. At the end of the feeding period, the rats were anesthetized with 3.0% isoflurane (Wako Pure Chemical Industries, Ltd.) and sacrificed to collect whole blood from the abdominal aorta. The serum was obtained by centrifugation at 2,000 × g for 20 min at 4°C and then stored at −80°C. The liver and cecum were removed and weighed. A part of the liver was suspended in RNAlater stabilization reagent (Ambion; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and stored at −80°C until RNA extraction. The remaining portions of liver and cecum were immediately frozen using liquid nitrogen and stored at −80°C until subsequent analyses.

Liver lipids were extracted using chloroform-methanol (2:1) solution as previously described (5–7). The solvent was removed by evaporation, and the total lipid (TL) was measured gravimetrically (5–7). The TG and cholesterol (CH) concentrations were measured using enzymatic kits (total cholesterol C-Test Wako, cat. no. 439–17501 and triglyceride G-Test Wako, cat. no. 432-40201; Wako Pure Chemical Industries, Ltd.). The frozen liver was homogenized in nine volumes of 0.14 M KCl and centrifuged at 20,000 × g for 30 min at 4°C. The supernatant was used as the enzyme source. The activities of liver glucose-6-phosphate dehydrogenase (G6PD) and malic enzyme (ME) were determined spectrophotometrically by monitoring the formation of NADPH at 340 nm, as previously described (5–7). The reaction mixtures were as follows: 0.017 M glycylglycine, 3.5×10⁻³ M MgSO₄, 0.33×10⁻³ M NADP, 0.67×10⁻³ M glucose 6-phosphate for G6PD activity; and 0.017 M glycylglycine, 7×10⁻³ M MgSO₄, 0.7×10⁻³ M MnSO₄, 0.33×10⁻³ M NADP, 0.33×10⁻³ M malate for ME activity. All reagents were obtained from Wako Pure Chemical Industries, Ltd.

It has been recently demonstrated that dietary MI increases the plasma level of adiponectin in C57BL/6 mice fed a high-fat diet (18). Thus, the level of serum adiponectin was measured by using an enzyme-linked immunosorbent assay (cat. no. TI 25462309; Otsuka Pharmaceutical Co., Ltd., Tokushima, Japan) in experiment 2.

mRNA expression levels were determined by RT-qPCR (14,19). Total RNA was isolated using a NucleoSpin RNA kit (Takara Bio, Inc., Otsu, Japan) according to the manufacturer's instructions. Purified total RNA (400 ng) was converted to cDNA using a PrimeScript RT Master Mix (Takara Bio, Inc.). Real-time PCR was performed using SYBR Premix Ex Taq II (Takara Bio, Inc.), and samples were amplified in a LightCycler 480 Instrument (Roche Diagnostics GmbH, Mannheim, Germany) under the following cycle conditions: 95°C for 30 sec followed by 40 cycles of 95°C for 5 sec and 60°C for 30 sec. The primer sets for G6PD forward, GATGACATCCGCAAACAGAGTGA and reverse, GCTACA TAGGAGTTACGGGCAAAGA; ME1 forward, CAGGCCTC CGTTAGCTTTGTTC and reverse, GCACAAGGACTGTAA ACAGCAGTGA; and fatty acid synthetase (FAS) forward, GCTGCTACAAACAGGACCATCAC and reverse, TCTT GCTGGCCTCCACTGAC were purchased from Takara Bio, Inc. (14,19). Melting curve analysis was performed following amplification to distinguish the targeted PCR product from the non-targeted PCR product using the LightCycler 480 Basic Software v1.5. Relative expression levels were calculated for each sample following normalization to those of the reference gene GAPDH (forward, GGCACAGTCAAGGCTGAGAATG and reverse, ATGGTGGTGAAGACGCCAGTA), which was used as the endogenous control gene (19).

Fecal and cecal microbiota were analyzed as previously described (20). Briefly, bacterial genomic DNA was isolated from the freeze-dried and milled feces or the frozen cecal content using an UltraClean Fecal DNA extraction kit (MO BIO Laboratories; Qiagen, Inc., Valencia, CA, USA) according to the manufacturer's instructions. DNA was quantified using a Qubit dsDNA BR Assay kit (Thermo Fisher Scientific, Inc.). Bacterial groups were quantified by qPCR using the LightCycler 480 Instrument (Roche Diagnostics GmbH) and the group-specific primers, which have been provided previously (20). Following initial denaturation at 95°C for 30 sec, 40 PCR cycles were performed with denaturation at 95°C for 5 sec, annealing at 55°C [total bacteria, Lactobacillus spp., Bifidobacterium spp., C. coccoides, C. Clostridium leptum (C. leptum) and Bacteroides] for 30 sec, and extension at 72°C for 15 sec (total bacteria, Lactobacillus spp., Bifidobacterium spp. and Bacteroides) or 1 min (C. coccoides and C. leptum). Melting curve analysis was performed following amplification to distinguish the targeted PCR product from the non-targeted PCR product. Data were analyzed using the second derivative maximum method of the LightCycler 480 Basic Software v1.5. The amplification efficiencies (e) of real-time PCR for each primer set were estimated using a linear regression of the crossing point for each fecal DNA dilution versus the log dilution using the formula: e = × - 1/slope, where ‘x’ was fold dilution. Efficiencies were between 1.94 and 1.99 (optimum value of 2.0). The relative abundance of microbial populations was expressed as the proportion of total bacterial 16S rRNA genes.

The pH of cecal digesta was measured directly using a compact pH meter (B-212; Horiba, Ltd., Kyoto, Japan). The empty cecum was weighed and the weight of cecal digesta was calculated as the difference between the weight of the cecum and the weight of the empty cecum. Cecal OAs were measured by the internal standard method using high-performance liquid chromatography (HPLC; L-2130; Hitachi, Ltd., Tokyo, Japan) equipped with an Aminex HPX-87H ion exclusion column (7.8 mm i.d. × 30 cm; Bio-Rad Laboratories, Inc., Hercules CA, USA) as previously described (20,21). Briefly, 500 mg cecal digesta was homogenized in 5 ml of 50 mmol/l H₂SO₄ containing 10 mmol/l 2,2-dimethyl butyric acid as an internal standard and subsequently centrifuged at 17,000 × g at 2°C for 20 min. The supernatant was ultrafiltered, and the filtrate was examined by HPLC (column at 60°C). The mobile phase (5 mM H₂SO₄) was delivered at a flow rate of 0.7 ml/min. Cecal OAs were detected at 210 nm using a variable wavelength detector (L-2400; Hitachi, Ltd.).

All values were expressed as means ± standard error of the mean. Normally distributed data were analyzed using two-way analysis of variance (ANOVA), whereas non-normally distributed data were analyzed using the non-parametric Kruskal-Wallis one-way ANOVA. The Tukey-Kramer post hoc test was performed when a significant effect was detected by two-way ANOVA. The Steel-Dwass post hoc test was conducted when a significant effect was detected by the Kruskal-Wallis one-way ANOVA. Associations between liver TGs and other parameters (liver lipogenic enzyme activity and gene expression, serum adiponectin, fecal and cecal microbiota and cecal OAs) were analyzed using the Spearman rank correlation analysis. Data analysis was performed using Excel Statistics 2012 for Windows (SSRI Co., Ltd., Tokyo, Japan). P<0.05 was considered to indicate statistical significance.

In experiment 1, body weight gain was not affected by supplementation of the dietary carbohydrate source with PA (Table II). Compared with the HSR diet, the HSC diet marginally increased food intake (P<0.05), whereas PA intake had no effect on food intake (Table II). The HSC diet induced significant increases in the liver weight, hepatic TL and TG concentrations, hepatic activities of G6PD and ME, and hepatic expression of G6PD, ME and FAS (P<0.05; Table II). The supplementation of PA in the diet effectively prevented the increases in hepatic TL and TG concentrations and significantly decreased the hepatic expression of G6PD and ME1 in rats fed the HSC diet (P<0.05). Similarly, PA intake decreased the hepatic activities of G6PD and ME, and the hepatic expression of FAS in rats fed the HSC diet, though these changes were determined as non-significant. In rats fed the HSR diet, PA intake did not alter these lipid metabolic parameters except for marginally decreasing the hepatic TG concentration (P<0.05).

Dietary PA significantly increased fecal weight in rats fed the HSC diet (P<0.05; Table III). The supplementation of PA in the HSC diet significantly increased the fecal ratio of Lactobacillus spp. and reduced the ratio of C. coccoides (P<0.05; Table III). In rats fed the HSR diet, dietary PA did not have a significant effect on the fecal ratios of Lactobacillus spp. and C. coccoides. The fecal ratio of Bacteroides was higher in rats fed the HSC diet with PA compared with that in rats fed the HSR diet with PA (P<0.05). PA intake did not have a significant effect on the fecal ratios of Bifidobacterium spp. regardless of the dietary carbohydrate source (Table III).

The weight of cecal tissue was lower in rats fed the HSC diet compared with that in rats fed the HSR diet, and the weight of cecal digesta was significantly enhanced by PA intake in rats fed the HSR diet (P<0.05; Table III). The cecal pH was higher in rats fed the HSC diet compared with that in rats fed the HSR diet. PA intake significantly lowered the cecal pH regardless of the dietary carbohydrate source (P< 0.05; Table III). In rats fed the HSC diet, dietary PA significantly increased not only the fecal ratio but also the cecal ratio of Lactobacillus spp. and significantly decreased the cecal ratio of C. coccoides (P<0.05; Table III). In addition, in rats fed the HSC diet, dietary PA significantly increased the cecal ratio of Bifidobacterium spp. (P<0.05). In rats fed the HSR diet, dietary PA had no significant effect on the cecal microbiota composition (Table III).

In rats fed the HSR diet or the HSC diet, the cecal succinate level was markedly enhanced by PA intake (P<0.05; Table III). The cecal acetate level was significantly higher in rats fed the HSR diet with PA compared with that in rats fed the HSC diet without PA. Furthermore, the cecal propionate level was significantly higher in rats fed the HSR diet without PA compared with that in rats fed the HSC diet with PA (P<0.05; Table III).

In rats fed the HSC diet (experiment 2), dietary PA and MI did not affect body weight gain or food intake (Table IV). Liver weight was significantly reduced by MI intake (P<0.05), and moderately by PA intake (Table IV), and PA or MI intake decreased the liver TL and TG concentrations (P<0.05; Fig. 1A). The activity of hepatic ME was significantly decreased by dietary PA intake (P<0.05), and moderately by MI intake (Fig. 1B). Although the activity and expression of hepatic G6PD were not significantly affected by dietary PA or MI, the expression of hepatic ME1 and FAS in rats fed the HSC diet with PA or MI was decreased (P<0.05; Fig. 1B and C). The serum adiponectin level was significantly increased by dietary MI (P<0.05), whereas it was moderately increased by PA in rats fed the HSC diet (Fig. 2).

In experiment 2, the fecal and cecal weights were not significantly affected by dietary PA or MI (Table IV). In accordance with experiment 1, PA intake significantly increased the weight and lowered the pH of the cecal digesta (P<0.05). Similarly, MI intake moderately increased the digesta weight and significantly decreased the digesta pH (Table IV). The supplementation of PA in the HSC diet significantly increased the fecal ratio of Lactobacillus spp. and decreased the fecal ratio of C. coccoides compared with the HSC diet alone (P<0.05; Fig. 3). Similar to PA intake, MI intake decreased the ratio of C. coccoides (P<0.05) and moderately increased the ratio of Lactobacillus spp. (Fig. 3). No significant differences were detected in the fecal ratio of Bifidobacterium spp. among the three groups (Fig. 3). As presented in Table IV, dietary PA or MI significantly increased the cecal ratio of Lactobacillus spp. The cecal ratio of C. coccoides was decreased by dietary PA (P<0.05), but was not affected by dietary MI (Table IV). Dietary PA significantly increased the cecal ratio of Bifidobacterium spp. (P<0.05; Table IV); dietary MI also increased the ratio, but its effect was not significant (Table IV). Dietary PA or MI did not affect the cecal ratio of Bacteroides (Table IV). The cecal succinate level was increased by the supplementation of PA in the HSC diet, as in experiment 1; whereas it was unaffected by MI intake. MI intake also marginally increased the cecal butyrate level (P<0.05). The levels of all other cecal OAs were unaffected by PA or MI intake.

The data from all rats were used for correlation analysis. As depicted in Table V, the liver TG concentration was positively correlated with the activity and expression of hepatic lipogenic enzymes in experiments 1 and 2, and negatively correlated with the serum adiponectin level in experiment 2 (P<0.05). Fecal and cecal Lactobacillus spp. correlated negatively with the liver TG concentration in experiments 1 and 2 (P<0.05). A similar association was observed between the abundance of Bifidobacterium spp. and liver TG concentration, although this correlation was not significant for fecal abundance in experiment 2. The liver TG concentration exhibited a significant positive correlation with the fecal and cecal abundance of C. coccoides in experiment 2 (P<0.05); while this correlation was not significant in experiment 1. The cecal levels of succinate, lactate, acetate and propionate exhibited a significant negative correlation with the liver TG concentration in experiment 1 (P<0.05): while this correlation was not significant in experiment 2 (Table V).