All animal protocols were approved by the Animal Care Committee of Southeast University (approval no. 2017-AN-1). All in vivo experiments described in the present study were in accordance with institutional guidelines for the care and use of animals (28).

Male Sprague Dawley rats (aged 8–12 weeks, 230–260 g) purchased from Kay Biological Technology Co., Ltd. (Shanghai, China) were maintained at a temperature of 23±3°C and a humidity of 35±5% under a 12 h dark/light cycle in a specific pathogen-free animal facility. A total of 14 rats were randomly separated into two groups, an HFD group and a control diet (CD) group. The HFD group was fed with D12451 formula (Research Diets, Inc., New Brunswick, NJ, USA), and the CD group received a standard diet. All rats were provided free access to water. Body weight and random glucose blood levels were measured every 5 days. On day 30, all rats were sacrificed. The livers, gastrocnemius (GAS) and epididymal white adipose tissue (eWAT) were immediately excised and placed in liquid nitrogen and stored at −80°C. Blood was also collected from the heart to investigate the components present in the serum.

The glucose tolerance test (GTT) and pyruvate tolerance test (PTT) were performed by administering an intraperitoneal injection of glucose (2 g/kg) or sodium pyruvate (2 g/kg), respectively, into rats following a 12–16-h period of starvation. The blood glucose levels of the rats were measured at 0, 15, 30, 60, 90 and 120 min following treatment. In the insulin tolerance test (ITT), rats were starved for 6 h prior to intraperitoneal injection of insulin (0.8 U/kg), and the blood glucose levels of the rats were measured 0, 15, 30, 60, 90 and 120 min later (Abbott Diabetes Care; Abbott Pharmaceutical Co., Ltd., Lake Bluff, IL, USA) (29).

Triglycerides and free fatty acids (FFAs) in the liver were analysed according to the manufacturer's protocols (Wako Pure Chemical Industries, Ltd., Osaka, Japan). Blood samples were centrifuged at 3,000 × g at 4°C for 15 min to isolate the serum, and insulin levels were detected using an ELISA kit (EZRMI-13K, EMD Millipore, Billerica, MA, USA).

The rat hepatocarcinoma cell line, H4IIE (ml-cs-0524; American Type Culture Collection, Manassas, VA, USA) was cultured in Dulbecco's Modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (Lonsera Science, Canelones, Uruguay), 1 mM glutamine, 100 U/ml penicillin and 100 mg/ml streptomycin at 37°C.

Mimics for miR-215 (5′-AUGACCUAUGAUUUGACAGACA-3′) and inhibitors for miR-215 (Ant-215; 5′-UGUCUGUCAAAUAGGUCAU-3′) were obtained from Shanghai GenePharma Co., Ltd. (Shanghai, China). Nonsense sequences were used as mimics negative control (NC; 5′-UCACAACCUCCUAGAAAGAGUAGA-3′) and Ant-NC (5′-UCUACUCUUUCUAGGAGGUUGUGA-3′). Cells were transfected with mimics or inhibitors (100 nM) using Lipofectamine^® 2000 (Thermo Fisher Scientific, Inc.) when the cells reached 70–80% confluence and treated for 48 h prior to RNA and protein isolation. Cells were starved for 4 or 12 h in serum-free DMEM containing 0.5% bovine serum albumin (BSA, Sangon Biotech Co., Ltd., Shanghai, China) prior to FFA treatment. FFAs, including palmitic acid (PA), linoleic acid (LA) and oleic acid (OA) were obtained from Sigma-Aldrich (Merck KGaA, Darmstadt, Germany) and diluted to 0.5 mM; BSA (0.5 mM) was used as a control. Cells were treated with FFA for 48 h at 37°C according to the manufacturer's protocols.

For the overexpression of Rictor, Rictor was cloned without its 3′-UTR into a pcDNA3.1 plasmid (VT1001, YouBio, Hunan, China) using the forward primer 5′-TTGCGGCCGCATGAGAAAGCTGGGCCATCTG-3′ and reverse primer 5′-CCGCTCGAGTCAGGATTCGGCAGATTCGT-3′, and the restriction enzymes NotI (R3189S, New England BioLabs, Inc., Ipswich, MA, USA) and XhoI (R0146S, New England BioLabs, Inc.). Cells were transfected with pcDNA3.1-Rictor using Lipofectamine 2000 as aforementioned and lysed 48 h later for subsequent experiments.

H4IIE cells were transfected with mimics or plasmids for 36 h, starved for 12 h, and treated with 100 nM insulin for 5–20 min at 37°C prior to protein collection for western blot analysis, as described below.

Putative target genes of mir-215 were screened using TargetScan 7.2 (http://www.targetscan.org/vert_72/).

To construct reporter plasmids, the 3′-UTR sequence of Rictor containing the miR-215 regulatory elements (MRE) was cloned into the pRL-TK plasmid (YouBio). Mutations in the MRE were made using the KOD-Plus mutagenesis kit (Toyobo Life Science, Osaka, Japan) according to the manufacturer's protocols. H4IIE cells were cultured in a 48-well plate and co-transfected with miR-215 mimics (100 or 200 nM) or mimics-NC (100 nM), and pRL-TK-Rictor-3′-UTR-wild-type (wt) or pRL-TK-Rictor-3′-UTR-mutant (mut) by Lipofectamine 2000; pGL3basic (YouBio) served as an internal reference. Following culture for 48 h, cells were lysed and relative luciferase activity was analysed with the Dual Luciferase Reporter Assay System (Promega Corporation, Madison, WI, USA) on a luminometer (Promega Corporation).

RNA was isolated using TRIzol^® reagent (Thermo Fisher Scientific, Inc.) from cells and rat livers, and 500 ng of RNA was reverse-transcribed into cDNA using the PrimeScript RT reagent Kit (Takara Bio, Inc., Otsu, Japan) according to the manufacturer's protocols. qPCR was performed on an Applied Biosystems™ 7500 Fast Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.) using 4.6 µl cDNA, 5 µl SYBR^® Green (Roche Diagnostics, Basel, Switzerland) and 0.4 µl (10 µM) primers. qPCR was performed under the following conditions: 95°C for 10 min, then 40 cycles of 95°C for 25 sec, 60°C for 30 sec and 72°C for 30 sec. The relative expression of genes was calculated using the 2^−∆∆Cq method and normalised to GAPDH (30). The primer sequences used were as follows: miR-215, forward 5′-ACACTCCAGCTGGGATGACCTATGATTTGA-3′, reverse, 5′-CTCAACTGGTGTCGTGGAGTCGG-3′; glucose 6-phosphatase (G6Pase), forward 5′-AAGCCAACGTATGGATTCCG-3′, reverse, 5′-ACAGCAATGCCTGACAAGACT-3′; phosphoenolpyruvate carboxykinase (PEPCK), forward 5′-GTGCTGGAGTGGATGTTCGG-3′, reverse, 5′-CTGGCTGATTCTCTGTTTCAGG-3′; peroxisome proliferator-activated receptor γ co-activator 1α (PGC-1α), forward 5′-CAATGAATGCAGCGGTCTTA-3′, reverse, 5′-GTGTGAGGAGGGTCATCGTT-3′; and GAPDH, forward 5′-AGGTCGGTGTGAACGGATTTG-3′ and reverse, 5′-TGTAGACCATGTAGTTGAGGTCA-3′.

Cells or tissues were lysed in radioimmunoprecipitation lysis buffer [10 mM Tris-HCl (pH 7.5), 1% SDS, 1 mM Na₃VO₄, 10 mM NaF] and protease inhibitor cocktail (Roche Diagnostics), and total protein was quantified using the Bradford assay. Protein samples (20 µg) were separated via 80 V under constant pressure 10% SDS-PAGE. Proteins were then transferred onto nitrocellulose membranes with 300 mA for 3 h and blocked with 5% skimmed milk for 1 h at room temperature. Membranes were probed with the following primary antibodies (diluted with 5% BSA, Sangon Biotech Co., Ltd.) at 4°C overnight: Anti-AKT (1:1,000; cat. no. 4685, Cell Signaling Technology, Inc., Danvers, MA, USA), anti-phosphorylated (p)-AKT (Ser473; 1:1,000; cat. no. 4060, Cell Signaling Technology, Inc.), anti-Rictor (1:1,000; cat. no. 5379, Cell Signaling Technology, Inc.), anti-GSK3β (1:1,000; BM2974, Boster Biological Technology, Pleasanton, CA, USA), anti-p-GSK3β (1:1,000; cat. no. 9323, Cell Signaling Technology, Inc.), anti-actin (1:1,000; ab8227, Abcam, Cambridge, UK) and anti-GAPDH (1:5,000; HPA040067, Sigma-Aldrich; Merck KGaA). Membranes were then washed with TBS-0.1% Tween 20 buffer and probed with horseradish peroxidase-conjugated secondary antibodies (HAF008, Novus Biologicals, LLC, Littleton, CO, USA) diluted with 5% skim milk (1:2,000) for 1 h at 37°C. Protein bands were visualised using SuperSignal Ultra (36209ES01, Shanghai Yeasen Biotechnology Corporation Co., Ltd., Shanghai, China). ImageJ V1.49 (National Institutes of Health, Bethesda, MD, USA) was employed for densitometry analysis.

The livers were extracted from rats and fixed in 4% paraformaldehyde at 37°C for 24–36 h. Paraffin-embedded sections (10 µm) were subjected to H&E and Oil-Red staining for 10 min at room temperature with aid from Wuhan Google (Wuhan, China). Images of Oil-Red and H&E staining were acquired using an inverted microscope (magnification, ×20; Olympus Corporation, Tokyo, Japan); three fields per view were analysed.

Data were presented as the mean ± standard deviation of at least three independent experiments. Animal sample size for each study was selected on the basis of literature documentation of similar experiments, and no statistical method was used to pre-determine the sample size. Excel (Microsoft Corporation, Redmond, WA, USA) and GraphPad Prism 5.0 (GraphPad Software, Inc., La Jolla, CA, USA) were used for statistical analysis. Differences between multiple groups were determined using one-way analysis of variance followed by a Dunnett's post-hoc test. Differences between two groups were analysed using Student's t-tests. Samples were not excluded during the animal experiments. P<0.05 was considered to indicate a statistically significant difference.

A rat model of short-term fatty liver was generated to investigate the association between fatty liver and hepatic insulin resistance. Rats were randomly separated into two groups; one group was fed a CD and the other was fed a HFD. The blood glucose levels and liver condition of rats were evaluated after 30 days. As presented in Fig. 1A, HFD rats exhibited significantly increased body weight following 25 days compared with the CD group. Oil-Red and H&E staining of livers following 30 days on the respective diets revealed that the lipid content of the livers of HFD rats was markedly increased compared with CD rats (Fig. 1B). Additionally, the liver weight: Body weight ratio of HFD rats was significantly increased compared with CD rats (Fig. 1C). The lipid content of the livers was determined via an ELISA. As presented in Fig. 1D, the lipid content of livers from the HFD group was significantly increased compared with that of the CD group. The results suggest that the HFD treatment successfully induced fatty liver symptoms in rats.

A number of studies reported an association between fatty liver and T2D (31,32). To investigate whether the hepatic insulin signalling pathway was impaired in rats with fatty liver, blood glucose levels were regularly monitored (Fig. 2A). Random blood glucose levels in HFD rats were significantly elevated on days 25 and 30 compared with in the CD group. Furthermore, HFD treatment resulted in significantly impaired glucose, pyruvate and insulin tolerance compared with CD treatment (Fig. 2B-D). As presented in Fig. 2B and D, there were no significant differences in glucose and insulin tolerance reported between the two groups 60 and 90 min following injection. This may be due to two reasons; the HFD group was not fed for a sufficient amount of time to develop severe insulin resistance, or the serum glucose level of rats may be highly variable. The number of rats in each group may be insufficient to counteract this variability; however, the maximum possible number of rats were included in each group. On the contrary, the area under the curves across the 120-min period indicate significant differences in the glucose and insulin tolerances between the two groups.

Furthermore, the expression of the major molecular nodes of the insulin signalling pathways in the livers of HFD and CD rats were determined by western blot analysis. It was revealed that the expression levels of p-AKT and p-GSK3β protein in the liver of HFD rats were significantly decreased compared with CD rats (Fig. 2E and F). The liver modulates glucose homeostasis of the body via the regulation of gluconeogenesis (33). Therefore, the expression of gluconeogenesis-associated genes, including G6PASE, PEPCK and PGC-1α, in the liver of the two groups was investigated by RT-qPCR. It was demonstrated that the expression levels of gluconeogenesis-associated genes in HFD rat livers were significantly increased compared with control rats (Fig. 2G). In addition to the liver, muscle and adipose tissues are involved in glucose homeostasis (34). The levels of p-AKT and p-GSK3β expression in rat GAS and eWAT were determined. It was demonstrated that the levels of p-AKT and p-GSK3β expression in HFD GAS (Fig. 2H) and eWAT (Fig. 2I) were not markedly altered. Additionally, it was revealed that the serum expression levels of insulin in HFD rats were significantly increased compared with in CD rats (Fig. 2J). The results indicated that HFD rats may have developed hepatic insulin resistance and hyperglycaemia.

Numerous studies reported that miRNAs were involved in the regulation of glucose homeostasis (29,35), and that the expression of miR-215 in fatty liver was aberrant, as observed via array analysis (36,37). The contribution of miR-215 to the development of T2D and fatty liver requires further study. The involvement of miR-215 in hepatic insulin resistance was investigated and the expression levels of miR-215 in the livers of the CD and HFD groups were determined. It was revealed that miR-215 expression in the livers of the HFD group was significantly increased compared with in the CD group (Fig. 3A). Via bioinformatics analysis, an miR-215 binding site was reported in the 3′-UTR of Rictor that is conserved across several species (Fig. 3B). Rictor is an important regulatory protein in the insulin signalling pathway (25). By transfecting H4IIE cells with mimics-215, miR-215 was successfully overexpressed and significantly reduced the levels of Rictor protein expression compared with the control (Fig. 3C and D). Conversely, following transfection of H4IIE cells with Ant-NC or Ant-215, it was demonstrated that Ant-215 significantly suppressed the expression of miR-215 and increased the levels of Rictor protein expression compared with the control (Fig. 3E and F). To investigate whether Rictor is a direct target for miR-215, the 3′-UTR of Rictor, including the MRE, was cloned; luciferase reporter assays demonstrated that miR-215 mimics significantly reduced the activity of the reporter gene containing Rictor 3′-UTR-wt but not Rictor 3′-UTR-mut (Fig. 3G), suggesting that Rictor was a target of miR-215. Additionally, Rictor protein expression was determined in the liver of HFD and CD rats; it was revealed that the levels of Rictor protein expression were significantly decreased in the livers of the HFD group compared with the CD group (Fig. 3H). The results indicated that Rictor is a direct target gene of miR-215.

Insulin signalling was investigated in H4IIE cells following overexpression of miR-215 (Fig. 4A). It was revealed that overexpression of miR-215 markedly reduced the levels of Rictor expression and decreased the insulin-stimulated phosphorylation of AKT compared with the NC. The expression of gluconeogenesis-associated genes was also evaluated; it was demonstrated that overexpression of miR-215 significantly increased the expression levels of G6PASE, PEPCK and PGC-1α compared with the NC (Fig. 4B). Conversely, knockdown of miR-215 by Ant-215 markedly increased the levels of Rictor and p-AKT expression following treatment with insulin compared with the NC (Fig. 4C), and significantly decreased the expression levels of gluconeogenesis-associated genes (Fig. 4D). Rictor was subsequently cloned without its 3′UTR into the pcDNA3.1 plasmid; as presented in Fig. 4E, transfection of H4IIE cells with pcDNA3.1-Rictor significantly increased the expression levels of Rictor mRNA and protein compared with empty vector. Following restoration of Rictor expression, the suppressive effects of mimics-215 on the phosphorylation of AKT were eliminated (Fig. 4F), indicating that regulating the expression of miR-215 affects the insulin signalling pathway. Then, the mechanisms promoting enhanced miR-215 expression in fatty liver were investigated. As triglyceride levels were significantly increased in livers from HFD rats, it was hypothesised that fatty acids may regulate the expression of miR-215. H4IIE cells were stimulated with various fatty acids; it was revealed that PA, LA and OA markedly upregulated the levels of miR-215 expression compared with control BSA treatment (Fig. 4G). Furthermore, the protein expression levels of Rictor were markedly decreased in H4IIE cells following treatment with LA, PA and OA compared with BSA treatment (Fig. 4H). The levels of FFAs in the livers of HFD and CD rats were determined via an ELISA. It was demonstrated that the levels of FFAs in the liver were significantly increased in HFD rats compared with CD rats (Fig. 4I).