PIO and ALO were provided by Takeda Pharmaceutical Company, Ltd. (Osaka, Japan). The CD diet and standard chow were purchased from Oriental Yeast Co., Ltd. (Tokyo, Japan). Eight-week-old 30 male KK-A^y mice were purchased from CLEA Japan, Inc. (Tokyo, Japan). These mice demonstrate ectopic overexpression of agouti peptide, an endogenous melanocortin-4 receptor (MC4R) antagonist, and social isolation promotes obesity due primarily to decreased energy expenditure and secondarily to increased food consumption. In addition, such isolation leads to insulin-independent diabetes associated with increased hepatic gluconeogenic gene expression (9). KK-A^y mice were housed individually in stainless-steel cages, and provided with unrestricted access to chow and tap water throughout the duration of the present study. KK-Ay mice were housed in the room of 23±3°C and 50±10% humidity and maintained on 12 h light/dark cycle. Following a 4-week acclimation period (12 weeks of age), the mice (45.5±4.9 g) were randomly separated into five groups of 6 mice each and administered the following diets: Control group, standard chow; CD group, a CD diet; PIO group, the CD diet containing 0.02% PIO; ALO group, the CD diet containing 0.03% ALO; and the PIO+ALO group, the CD diet with 0.02% PIO and 0.03% ALO. Following 8 weeks of the diet, the mice received no food for 18 h and were administered ether anesthesia and euthanized by exsanguination from the inferior vena cava. Serum and hepatic tissue samples were obtained and frozen at −80°C until assayed. Certain hepatic tissues were fixed in 10% formalin for 48 h at room temperature and 3 µm paraffin-embedded hepatic tissues were stained with hematoxylin for 10 min and eosin for 5 min at room temperature. Then the histology of samples was analyzed under a light microscope (magnification, ×40 or ×100). The present study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (Bethesda, MD, USA). The protocol was approved by the Institute for Animal Experimentation of Shimane University (Shimane, Japan; protocol no. IZ23-164).

Serum levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyl transpeptidase (γ-GTP), triglyceride (TG), total cholesterol (T-Chol) and high-density lipoprotein-cholesterol (HDL-C), in addition to fasting serum glucose were measured using a Spotchem 4430 Benchtop Biochemistry Analyzer (Arkray; SCIL, Holtzheim, France). Serum levels of fasting insulin and adiponectin were measured using ELISA kits (AKRIN-011T and AKMAN-011, respectively; Shibayagi Co., Ltd., Tokyo, Japan). Homeostasis model of assessment-insulin resistance (HOMA-IR) values were calculated using the following formula: HOMA-IR=[fasting serum glucose (mmol/l) × fasting serum insulin (ng/ml)]/22.5 (10).

Total RNA was isolated from hepatic tissues with RNAlater RNA Stabilization Reagent and purified using an RNeasy Mini kit (both from Qiagen GmbH, Hilden, Germany). Samples of cDNA were generated using a cDNA synthesis kit according to the manufacturer's protocols (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA). RT-qPCR was then performed using an ABI PRISM 7700 sequence detection system (Thermo Fisher Scientific, Inc.) with SYBR^®-Green PCR master mix (Applied Biosystems; Thermo Fisher Scientific, Inc.). The PCR protocol used was as follows: Enzyme activation: 94°C for 10 min; thermocycling, 94°C for 15 sec, 60°C for 1 min, repeat for 43 cycles. The relative expression of each gene was normalized to that of GAPDH. qPCR was replicated two times for each sample of cDNA. Primers for mouse fatty acid transport protein (FATP)1-4, fatty acid synthase (FAS), liver-type fatty acid binding protein (L-FABP), acyl-coenzyme A oxidase 1 (AOX) and long-chain acyl-coenzyme A dehydrogenase (LCAD) genes were prepared as previously described (11). Primers for mouse peroxisome proliferator-activated receptor α (PPARα), α smooth muscle actin (αSMA), superoxide dismutase 1 (SOD1), carnitine palmitoyltransferase 1a (CPT1a), microsomal triglyceride transfer protein (MTP), monocyte chemoattractant protein 1 (MCP1) and GAPDH genes were prepared using NCBI blast tool and Primer3 Input (version 0.4.0, http://bioinfo.ut.ee/primer3-0.4.0/) which is most commonly used for primer design (12). The primers are presented in Table I. All the mRNA levels were normalized to GAPDH mRNA in the same preparation.

Protein extraction was performed using PRO-PREP^™ protein extraction solution (Intron Biotechnology, Inc., Seongnam, Korea), according to the manufacturer's protocol. A total of 10 mg of liver tissue was suspended in 500 µl PRO-PREP^™ solution mixed with 5 µl Phosphatase-Inhibitor Mix II (Cosmo Bio Co., Ltd., Tokyo, Japan). Subsequently, the tissue was disrupted using a Tissue Lyser II (Qiagen GmbH) for 3 min at 30 Hz and homogenized with a syringe equipped with a 23-G needle, followed by incubation on ice for 30 min and sonication with Bioruptor UCD-200 TM (Cosmo Bio Co., Ltd.). In water cooled with ice, sonication for 5 min at 20 KHz was performed for 30 sec every 60 sec. The resulting liver tissue lysate was centrifuged at 14,000 × g for 5 min at 4°C, and the supernatant was collected and subjected to downstream processing. The protein concentration was estimated using a Pierce^™ Bicinchoninic Acid Protein Assay kit (Takara Bio, Inc., Otsu, Japan), for which 50 µg protein/lane was loaded and processed for SDS-PAGE fractionation (4–12% SDS-PAGE mini; Tefco Technical Frontier Co., Tokyo, Japan) and transferred to a polyvinylidene difluoride membrane (Hybond-P; GE Healthcare Life Sciences, Little Chalfont, UK). Following blocking of the membrane for 1 h using 5% skimmed milk (Difco; BD Biosciences, Franklin Lakes, NJ, USA) in TBS-Tween-20 (TBST; TBS and 0.05% Tween-20, pH 7.4) at room temperature, it was incubated for 24 h with a phosphorylated-AMP-activated protein kinase (AMPK)α (Thr172) rabbit monoclonal antibody (cat. no. 2535S; Cell Signaling Technology, Inc., Danvers, MA, USA) at a 1:500 dilution at 4°C, reacted with peroxidase-conjugated polyclonal goat anti-rabbit immunoglobulins (cat. no. P7-0448; Dako; Agilent Technologies Inc., Santa Clara, CA, USA) at a 1:1,000 dilution at room temperature for 2 h, and washed three times in TBST. The resulting signals were imaged using electrochemiluminescence (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA). Goat anti-β-actin antibody (catalogue no. SC-47778; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) was used as an internal control. Band densitometry was performed using ImageJ software of version 1.51 (National Institutes of Health, Bethesda, MD, USA).

All data are presented as the mean ± standard deviation (n=3). Mann-Whitney's U test was used for comparisons between two groups, while Steel-Dwass test following a Kruskal-Wallis test was used for comparisons among multiple groups. P<0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using Statistical Analysis Software (BellCurve for Excel, version 2.14; Social Survey Research Information Co., Ltd., Tokyo, Japan).

The Control group consumed a greater amount of food compared with the CD group (23.5±1.6 and 16.8±2.5 kcal/body/day, respectively; P=0.006). No significant differences were identified in the amounts of food consumed among the CD, PIO, ALO and PIO+ALO groups (Table II). In the CD, PIO, ALO and PIO+ALO groups, mean body weights prior to starting the diet were 49.0±1.4, 48.2±2.4, 48.5±2.5, 48.2±2.3 and 48.9±2.3 g, respectively, while those following 8 weeks of diet consumption were 52.5±2.3, 52.6±4.9, 59.9±7.5, 53.3±4.1 and 63.8±2.5 g, respectively, illustrating that the body weight of the PIO+ALO group was significantly increased compared with that of the CD (P=0.03) and ALO (P=0.03) groups at the end of the 8-week period (Table II).

Following 8 weeks of the dietary treatment, the liver/body weight ratio of the CD group was significantly greater compared with the Control group (P=0.006). By contrast, no significant differences were identified among the CD, PIO, ALO and PIO+ALO groups (Table II). Mild microvesicular hepatic steatosis was recognized in the Control group and severe macrovesicular hepatic steatosis was recognized in the CD group. Macrovesicular hepatic steatosis areas in the PIO, ALO and PIO+ALO groups were smaller compared with the CD group (Fig. 1).

No evident histological inflammatory alterations in the liver were observed in the CD group; however, the level of MCP1 expression in the livers of the CD group was significantly increased compared with in the Control group. Conversely, no significant differences were identified between MCP1 expression levels in the livers of the PIO, ALO and PIO+ALO groups compared with in the CD group. There was no evident hepatic fibrosis observed in the CD group and the expression levels of αSMA in the liver were not significantly different compared with in the Control group (Table III).

The serum levels of AST (P=0.006) and ALT (P=0.006) were significantly higher in the CD group compared with the Control group, whereas those levels in the PIO, ALO and PIO+ALO groups were not statistically different from the CD group (Table IV). Serum adiponectin in the CD group was not statistically different compared with the Control group, while its level in the PIO group demonstrated a tendency to be higher compared with the CD group (P=0.09; Table IV). Fasting serum insulin in the CD group demonstrated a tendency to be lower compared with the Control group (P=0.055), while no significant differences were identified among the CD diet groups. On the other hand, fasting serum glucose in the CD group tended to be higher compared with the PIO+ALO (P=0.082) group and HOMA-IR in the PIO+ALO (P=0.082) group also tended to be lower compared with the CD group (Table IV). Serum TG in the CD group was almost equal to that in the Control group and did not alter with PIO/ALO administration (Table IV). However, serum T-Cho and HDL-C in the CD group were significantly higher compared with the Control group, while serum HDL-C in the PIO+ALO group was higher compared with the ALO group (Table IV).

The gene expression levels of enzymes associated with lipid metabolism in liver tissue were examined using RT-qPCR, including genes associated with fatty acid uptake (FATP1, 2, 3 and 4), fatty acid synthesis, fatty acid oxidation (PPARα, L-FABP, CPT1a, AOX and LCAD) and very low density lipoprotein export. The expression levels of FATP3 (P=0.006) and FAS (P=0.006) in the liver were significantly increased in the CD group compared with the Control group (Table V). On the other hand, the expression levels of PPARα (P=0.017), L-FABP (P=0.018), LCAD (P=0.011) and MTP (P=0.006) in the liver were significantly decreased in the CD group compared with the Control group (Table V). Macrovesicular hepatic areas were decreased in the PIO, ALO and PIO+ALO groups compared with the CD group (Fig. 1). The expression levels of lipid metabolic genes in the livers of mice in the PIO, ALO and PIO+ALO groups were also compared with those in the CD group to examine the mechanism of prevention of fatty liver by PIO and/or ALO administration. The expression of FAS in the liver demonstrated a tendency to be decreased in the PIO group compared with the CD group (P=0.051), while the expression of CPT1a in the ALO group was significantly increased compared with the CD (P=0.031) and PIO (P=0.031) groups, whereas the expression of AOX did not alter following administration of PIO and/or ALO. The antioxidant enzyme SOD1 was additionally examined. Its expression in the liver was significantly decreased in the CD group compared with the Control group (P=0.045), while it demonstrated a tendency to increase in the PIO+ALO group compared with the CD group (P=0.051) (Table V).

Western blotting demonstrated that ALO promoted the phosphorylation of AMPK on Thr-172 in the livers of the ALO group mice compared with the CD group (Fig. 2).