The present study was performed at the Physical Examination Center of Hebei General Hospital (Shijiazhuang, China) and was approved by the Hebei General Hospital Ethics Committee (2018 Scientific Research Ethics Review; approval no. 39; Shijiazhuang, China). All of the clinical samples were obtained from the Physical Examination Center of Hebei General Hospital. A total of 36 subjects with elevated blood lipids in the physical examination population between May 1, 2021 and November 30, 2021 were randomly selected as the hyperlipidemia group (27 men, 9 women; mean age, 65.50±7.71 years; age range, 50-83 years). Another 36 healthy subjects matched for age and sex with the subjects in the hyperlipidemia group were randomly selected during the same period from the Physical Examination Center of Hebei General Hospital as the normal control group (CG group; 27 men, 9 women; mean age, 69.11±8.81 years; age range, 43-83 years). All participants provided written informed consent. The diagnostic criteria for hyperlipidemia were in accordance with the Guidelines for the Prevention and Treatment of Dyslipidemia in Chinese Adults (Revised Edition 2016) (2), which are as follows: Plasma TC ≥6.2 mmol/l (240 mg/dl), TG ≥2.3 mmol/l (200 mg/dl), LDL-C ≥4.1 mmol/l (160 mg/dl) or HDL-C <1.0 mmol/l (40 mg/dl) in adults after a 12-h fast. Hyperlipidemia was diagnosed if any one criterion was met. Participants taking aspirin, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers or statins within the previous 2 months were excluded from the study. Participants with chronic liver disease (including hepatitis B virus carriers), kidney disease, thyroid insufficiency or abnormalities, hypertension, diabetes, blood system disorders, mental disorders, acute and chronic infectious diseases, autoimmune diseases, tumors, pregnancy, lactation, long-term oral contraceptive use, and/or recent surgical history were excluded. The inclusion criteria for the CG group were as follows: No history of hypertension, diabetes mellitus and other chronic diseases; blood glucose 3.9-6.1 mmol/l; and the following blood lipid concentration levels: TC <5.2 mmol/l, TG <1.7 mmol/l and LDL-C <3.4 mmol/l.

Fasting blood samples (5 ml) were collected from each participant and placed in a BD Vacutainer SST tube (Becton, Dickinson and Company). Peripheral blood mononuclear cells (PBMCs) were isolated from fasting blood samples by Ficoll-Paque density gradient centrifugation (20˚C; 500 x g; 25 min) for miR-33 and sirtuin 6 (SIRT6) detection by reverse transcription-quantitative PCR (RT-qPCR). Biochemical tests [TC, TG, HDL-C, LDL-C, aspartate aminotransferase (AST), alanine aminotransferase (ALT) and fasting blood glucose (FBG)] were performed using an automatic biochemical detection instrument at the Clinical Laboratory of Hebei General Hospital. Glycated hemoglobin (HbA1c) was analyzed at Hebei Key Laboratory of Metabolic Diseases (Shijiazhuang, China) using an automatic glycohemoglobin analyzer (ADAMS A1c HA-8180; ARKRAY, Inc.) at 25˚C.

A total of 24 C57BL/6J mice (male; age, 8 weeks; weight, 22.0±2.0 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. The mice were housed in the animal laboratory at the Hebei Key Laboratory of Metabolic Diseases (temperature, 21-23˚C; humidity, 40-60%; and 12/12-h light/dark cycle) with constant access to food and water. All experimental procedures were approved (2022 Scientific Research Ethics Review; approval no. 217) by the Animal Care and Use Committee of Hebei General Hospital (Shijiazhuang, China) and complied with the Animal (Scientific Procedures) Act 1986 and associated guidelines (39).

After 1 week of adaptive feeding, the mice were randomly divided into three groups, with 8 mice in each group. The diet for the normal diet (ND) group was an ordinary diet (D12450J formula, consisting of 20% protein, 70% carbohydrate, 10% fat and 3.85 kcal/g). The model mice were fed a HFD (D12492 formula, consisting of 20% protein, 20% carbohydrate, 60% fat and 5.24 kcal/g). The Res mice were fed a HFD and a Res-based dietary supplement (60 mg/kg). All feed was purchased from Beijing Huafukang Biotechnology Co., Ltd. All mice were fed the respective diets for 6 weeks, after which they were euthanized by CO₂ asphyxiation (flow rate, 4 l/min; 30% vol/min) (40,41), and cervical dislocation was performed when the mice exhibited respiratory arrest and unconsciousness. Complete death was confirmed by cardiac arrest and dilated pupils. Blood samples were collected in tubes containing ethylenediaminetetraacetic acid (1.5 mg/ml) and centrifuged at 1,375 x g at 4˚C for 15 min. The plasma was collected and stored at -80˚C. Mice livers were quickly removed. Part of the liver tissues were snap-frozen in liquid nitrogen after washing with cold phosphate-buffered saline and stored at -80˚C for further analysis. Part of the liver tissues were fixed in 4% paraformaldehyde at 25˚C for 24 h for H&E staining.

The body weight and food intake of the mice in each group were measured at baseline and weekly thereafter until 6 weeks after baseline.

Serum glucose levels were determined using a glucose assay kit (cat. no. 60408ES60; Shanghai Yeasen Biotechnology Co., Ltd.). The TG content assay kit (cat. no. D799796-0100; Sangon Biotech Co., Ltd.) was used to detect TG levels. The TC assay kit (cat. no. A111-1-1), LDL-C assay kit (cat. no. A113-2-1), HDL-C assay kit (cat. no. A112-1-1), ALT assay kit (cat. no. C009-3-1) and AST assay kit (cat. no. C010-3-1) were purchased from Nanjing Jiancheng Bioengineering Institute, and were used to detect the concentrations of TC, LDL-C, HDL-C, ALT and AST. Serum malondialdehyde (MDA) concentration levels were measured using a lipid peroxidation MDA assay kit (cat. no. S0131S; Beyotime Institute of Biotechnology). All protocols were performed in accordance with the manufacturers' instructions.

After feeding for 6 weeks, glucose (1 g/kg) was given to each mouse via an orogastric tube for the OGTT. Blood glucose was measured immediately after glucose administration and 15, 30, 60 and 120 min after administration. A total of 24 h after the OGTT, the ITT was performed after a 12-h fast. The mice were injected intraperitoneally with insulin (1.5 IU/40 g; Tonghua Dongbao Pharmaceutical Co., Ltd.), and blood glucose was measured immediately after injection and 15, 30, 60 and 120 min after injection. The area under the receiver operating characteristic curve (AUC) for the OGTT was calculated using the trapezoidal method. The quantitative insulin sensitivity check index (QUICKI) was used to assess insulin sensitivity, as follows: QUICKI=1/[(log fasting blood glucose (mmol/l) + log fasting plasma insulin (µU/ml)].

Parts of the liver tissues were taken and fixed in 4% paraformaldehyde at 25˚C for 24 h. Subsequently, the tissues were embedded in paraffin wax, cut into 5-µm-thick sections, deparaffinized in xylene at 25˚C and rehydrated in a reverse-gradient series of ethyl alcohol (100, 95, 80 and 75%). The sections were stained with hematoxylin at 25˚C for 10 min and stained with eosin at 25˚C for 3 min, and visualized under a light microscope.

Oil Red O staining. Parts of the fresh liver tissues were taken and embedded in optimum cutting temperature compound, quickly frozen, and then sliced into 6-µm tissue sections. The sections were washed with PBS, stained with Oil red O working solution (6:4, oil red stock solution:distilled water; Oil red O: WSIG20100803; Sinopharm Chemical Reagent Co., Ltd.) at room temperature for 15 min and washed three times with PBS to remove the excess Oil red O dye. Subsequently, the sections were stained with Harris's hematoxylin (20151216; Nanjing Jiancheng Bioengineering Institute) for 3 min at 25˚C. The morphological features of the liver sections were observed under a light microscope.

HepG2 cells (human liver cancer cells) were purchased from Procell Life Science & Technology Co., Ltd., and cultured in complete DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Sangon Biotech Co., Ltd.) and 1% penicillin/streptomycin (Sangon Biotech Co., Ltd.) at 37˚C with 5% CO₂. HepG2 cells were immersed in normal medium and medium containing 0.25 mmol/l palmitate (PA) for 24 h. At the end of the stimulation period, the cells were washed three times with PBS and fixed with 4% paraformaldehyde for 10 min at 37˚C. Subsequently, cells were washed twice with PBS, then stained with 0.5% Oil red O for 30 min at 37˚C. After staining, the cells were washed once with 60% isopropanol, washed with PBS until a colorless solution was obtained, and observed under a fluorescence inverted microscope at a magnification of x50. Short tandem repeat profiling was used for authentication of HepG2 cells. HepG2 cells cultured in normal medium and transfected with miR-33 mimics, and HepG2 cells cultured in medium containing 0.25 mmol/l PA for 24 h after transfection with miR-33 inhibitor or SIRT6-pcDNA 3.1 were used to analyze the effect of transfection on lipid metabolism-related genes and lipid deposition.

miR-33 mimics, inhibitor and the corresponding controls were synthesized by Shanghai GenePharma Co., Ltd. For miR-33 mimics transfection, the HepG2 cells were seeded in 6-well plates at a density of 5x10⁵ cells/well. When 70-80% confluence was reached, cells were divided into three groups: CON (liposome), NC mimics (liposome + mimics control sequence) and miR-33 mimics (liposome + miR-33 mimics). The CON group was the control group, in which cells were transfected without any sequence. The NC mimics group was the scrambled negative control. The sequence of the corresponding controls (100 nmol/l) was 5'-GGUCUUACGUCAGUCACAAUAUCUG-3'. Cells were transfected using Lipofectamine^® 3000 Transfection Reagent (cat. no. L3000015; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. The cells were transfected for 6 h at 37˚C in a cell incubator with 5% CO₂, and then the medium was replaced with fresh DMEM. In the miR-33 mimics group, the cells were transfected with 100 nmol/l miR-33 mimics (5'-GUGCAUUGUAGUUGCAUUGCA-3') using Lipofectamine^® 3000 Transfection Reagent (cat. no. L3000015; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Cells were transfected for 6 h at 37˚C in a cell incubator with 5% CO₂, and then the medium was replaced with fresh DMEM. Subsequently, cells were incubated for 24 h in an incubator with 5% CO₂ at 37˚C, and the cells were collected for subsequent experiments.

To investigate the effect of miR-33 inhibitor transfection, HepG2 cells were divided into three groups: CON (liposome), NC inhibitor (liposome + inhibitor control sequence) and miR-33 inhibitor (liposome + miR-33 inhibitor). The CON group was the control group, in which cells were transfected without any sequence. The NC inhibitor group was transfected with 100 nmol/l scrambled negative controls (5'-GGUCUUACGUCAGUCACAAUAUCUG-3') using Lipofectamine^® 3000 Transfection Reagent (cat. no. L3000015; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Cells were transfected for 6 h at 37˚C in a cell incubator with 5% CO₂, and then the medium was replaced with fresh DMEM. In the miR-33 inhibitor group, cells were transfected with 100 nmol/l inhibitor (5'-UGCAAUGCAACUACAAUGCAC-3') using Lipofectamine^® 3000 Transfection Reagent (cat. no. L3000015; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. Cells were transfected for 6 h at 37˚C in a cell incubator with 5% CO₂, and then the medium was replaced with fresh DMEM. Subsequently, cells were incubated for 48 h in an incubator with 5% CO₂ at 37˚C, and the cells were collected for subsequent experiments.

To investigate the effect of miR-33 inhibitor on intracellular lipid metabolism, cells were divided into three groups: PA + lipo (liposome), PA + NC inhibitor (liposome + inhibitor control sequence) and PA + miR-33 inhibitor (liposome + miR-33 inhibitor). miR-33 inhibitor or NC inhibitor (scramble control) transfection was performed as aforementioned. After transfection for 6 h at 37˚C, cells were incubated with PA (0.25 mmol/l) for 24 h in an incubator with 5% CO₂ at 37˚C and then collected for the subsequent experiments.

To investigate the effect of SIRT6 overexpression, HepG2 cells were divided into the pcDNA 3.1 group (transfected with 500 ng pcDNA 3.1) and the SIRT6-pcDNA 3.1 group (transfected with 500 ng SIRT6-pcDNA 3.1). Cells were transfected in an incubator with 5% CO₂ at 37˚C for 6 h with the aforementioned plasmids using Lipofectamine^® 3000 Transfection Reagent (cat. no. L3000015; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol, and then the medium was replaced with fresh DMEM. Subsequently, cells were incubated for 24 h in an incubator with 5% CO₂ at 37˚C, and then collected for subsequent experiments.

To investigate the effect of SIRT6 overexpression on lipid metabolism, HepG2 cells divided into three groups: PA + lipo (liposome), PA + pcDNA 3.1 (liposome + 500 ng pcDNA 3.1) and PA + SIRT6-pcDNA 3.1 (liposome + 500 ng SIRT6-pcDNA 3.1). SIRT6-pcDNA 3.1 or pcDNA 3.1 transfection was performed as aforementioned. After transfection for 6 h at 37˚C, cells were incubated with PA (0.25 mmol/l) for 24 h in an incubator with 5% CO₂ at 37˚C and then collected for subsequent experiments. The human SIRT6-pcDNA 3.1 (cat. no. V38520) was purchased from Thermo Fisher Scientific, Inc.

RNAs from PBMCs of participants with hyperlipidemia, three randomly selected mouse liver tissues or cultured HepG2 cells were isolated using a total RNA purification kit (Sangon Biotech Co., Ltd.). Complementary DNA synthesis was performed using the Goscript Reverse Transcriptase System (Promega Corporation) and All-in-One^™ miRNA First-Strand cDNA Synthesis kit (GeneCopoeia, Inc.). The aforementioned operations were carried out in strict accordance with the manufacturer's instructions. qPCR was performed using an Applied Biosystems 7500 system (Thermo Fisher Scientific, Inc.) to detect mRNA levels using GoTaq^® qPCR Master Mix (Promega Corporation) and miRNA levels using the All-in-One^™ miRNA RT-qPCR detection kit (fluorophore, SYBR^® Green I; GeneCopoeia, Inc.). The thermocycling conditions were as follows: Polymerase activation for 1 cycle at 95˚C for 2 min; followed by 40 cycles of 95˚C for 15 sec and 60˚C for 1 min. Relative fold changes in RNA expression were calculated using the 2^-ΔΔCt method (42). mRNA levels were normalized to GAPDH gene expression, and miRNA levels were normalized to U6 small nuclear RNA levels. The primer sequences used were as follows: Mouse (m-)SIRT6 forward, 5'-CCGGGACCTGATGCTCGCTGATGA-3' and reverse, 5'-AGCCGTGGATGCGCAGGTCAG-3'; m-FASN forward, 5'-CGGTCCCTGTGCGCCTTCC-3' and reverse, 5'TGGGGTTGTGGAAGTGCAGGTTAGG-3'; m-PPARγ forward, 5'-CCGAAGAACCATCCGATTGAAGC-3' and reverse, 5'-CCGCCAACAGCTTCTCCTTCTCG-3'; m-PGC1α forward, 5'-AAGCGAAGAGCATTTGTCAACAGCA-3' and reverse, 5'-GCGGTTGTGTATGGGACTTCTTTTT-3'; m-CPT1 forward, 5'-AGCGCTGGCAAATGACTTCCTGAG-3' and reverse, 5'-CCTGCAGCGGTGTGGGGGTGAC-3'; m-SREBP1 forward, 5'-CGCAAGGCCATCGACTACATCCG-3' and reverse, 5'-CGGCGTCTGAGGGTGGAGGGGTAA-3'; m-ACC forward, 5'-GCCCCCGAGCCAGAGGACAGTAT-3' and reverse, 5'-CCGGGAGGAGTTCTGGAAGGAGC-3'; human (h-)SIRT6 forward, 5'-CGGCCCACGCAGACCCACATG-3' and reverse, 5'-TGGGGAAGCCTGAGCGCACAT-3'; h-FASN forward, 5'-GCGGCTGCTGCTGGAAGTCACCTAT-3' and reverse, 5'-GCCGCTCACGCCCACCCAGA-3'; h-PPARγ forward, 5'-GGCCGAGAAGGAGAAGCTGTTGG-3' and reverse, 5'-CGCCCTCGCCTTTGCTTTGGT-3'; h-PGC1α forward, 5'-CCCAGAACCATGCAAATCACAATCA-3' and reverse, 5'-GACGTCTTTGTGGCTTTTGCTGTTG-3'; h-CPT1 forward, 5'-CCCGGCAAGCCCCTCCAGTT-3' and reverse, 5'-GGACATGCAGTTGGCCGTTTC-3'; h-SREBP1 forward, 5'-CGCCCTCACCCCTGTCCCCTCC-3' and reverse, 5'-GGGGCTGTGGGGTGGGGGTC-3'; h-ACC forward, 5'-CCCCACTATGAGGCCGAGCA-3' and reverse, 5'-AGCGGGAGAAGCCACGGTAAAGT-3'; m/h-GAPDH forward, 5'-TGAACGGGAAGCTCACTG-3' and reverse, 5'-GCTTCACCACCTTCTTGATG-3'; m/h-miR-33 forward, 5'-GTGCATTGTAGTTGCATTGC-3' and reverse, 5'-GTCGTATCCAGTGCAGGGT-3'; m/h-U6 forward, 5'-CTCGCTTCGGCAGCACA-3' and reverse, 5'-AACGCTTCACGAATTTGCGT-3'.

Liver tissues of three randomly selected mice or cultured HepG2 cells were lysed in radioimmunoprecipitation assay lysis buffer (Thermo Fisher Scientific, Inc.; 25 mM Tris, HCl pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate and 0.1% SDS), and the total soluble protein was quantified using a BCA Protein Assay kit (Beijing Solarbio Science & Technology Co., Ltd.). Protein (20 µg/lane) from cell lysates was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Following transfer of the proteins onto polyvinylidene fluoride membranes, the membranes were blocked at room temperature for 60 min in 5% skim milk and probed with the primary antibodies overnight at 4˚C: Acetyl-CoA carboxylase (ACC; dilution, 1:2,000; cat. no. 3676; Cell Signaling Technology, Inc.), fatty acid synthase (FASN; dilution, 1:1,000; cat. no. ab128870; Abcam), SREBP1 (dilution, 1:2,000; cat. no. 557036; BD Biosciences), peroxisome proliferator-activated receptor-γ (PPARγ; dilution, 1:1,000; cat. no. 16,643-1-AP; Proteintech Group, Inc.), anti-PPARγ-coactivator 1 α (PGC1α; dilution 1:1,000; cat. no. 66,369-1-Ig; Proteintech Group, Inc.), carnitine palmitoyltransferase 1 (CPT1; dilution 1:1,000; cat. no. AF6558; Beyotime Institute of Biotechnology), SIRT6 (dilution, 1:1,000; cat. no. ab191385; Abcam) and anti-β-actin (dilution, 1:1,000; cat. no. 60008-1; Proteintech Group, Inc.). The membranes were incubated with the secondary antibodies for 2 h at room temperature. The secondary antibodies included the HRP-conjugated goat anti-rabbit antibody (dilution, 1:5,000; cat. no. ZDR-5306; OriGene Technologies, Inc.) and the HRP-conjugated goat anti-mouse antibody (dilution, 1:10,000, cat. no. ZDR-5307; OriGene Technologies, Inc.). Protein bands were visualized using enhanced chemiluminescent substrate (Pierce ECL Western Blotting substrate; Thermo Fisher Scientific, Inc.), and the band intensities were evaluated using Image J software (V1.8; National Institutes of Health).

The synthesized SIRT6 3'-untranslated region (UTR) was inserted into the pmirGLO vector (Promega Corporation). The mutation in the miR-33 seed-matching sequences was designed using the SIRT6 wild-type (WT) sequence generated by overlap extension PCR. SIRT6 WT and SIRT6 mutant-type (MT) reporter plasmids were designed and constructed by Guangzhou RiboBio Co., Ltd. 293T cells (Shanghai GeneChem Co., Ltd.) were cultured in High Glucose DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Sangon Biotech Co., Ltd.) and 1% penicillin/streptomycin at 37˚C with 5% CO₂. The WT and MT sequences were co-transfected with the miR-33 mimic (5'-GUGCAUUGUAGUUGCAUUGCA-3'; 100 nM) or corresponding control (5'-GGUCUUACGUCAGUCACAAUAUCUG-3'; 100 nM) into 293T cells using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) at 37˚C for 6 h. After transfection for 24 h, the cells were lysed and subjected to a Dual-Luciferase Reporter Assay (Promega Corporation). Luciferase activity was measured and calculated as the ratio of firefly luciferase activity to Renilla luciferase activity. The experiment was repeated three times.

All experimental data are presented as the mean ± SD. All experiments were repeated at least three times to verify the trends. One-way ANOVA with Tukey's post hoc test was used for comparisons among multiple groups. Comparisons between groups were performed using an unpaired Student's t-test. Sex differences were compared using the Pearson χ² test. SPSS (version 25.0; IBM Corp.) was used for all analyses. P<0.05 was considered to indicate a statistically significant difference.

Demographic, clinical and biochemical data were obtained from 36 participants with hyperlipidemia and 36 healthy control participants. As shown in Table I, BMI, weight, TC, TG, LDL-C, FBG, HbA1c, ALT and AST concentration levels were significantly higher, while HDL-C levels were significantly lower, in participants with hyperlipidemia compared with CG participants. No significant differences in sex, age or height were observed between the groups.

PBMCs from the hyperlipidemia group and CG were tested for miR-33 and SIRT6 expression levels and it was identified that miR-33 expression levels were significantly higher (Fig. 1A), and SIRT6 expression was significantly lower in the hyperlipidemia group compared with the CG (Fig. 1B).

Before investigating the underlying mechanism of Res in lipid metabolism, the mice in the HFD group were used to investigate lipid metabolism and miRNA expression in blood or liver tissues, respectively. After 6 weeks, body weights were significantly higher in the HFD group compared with the ND group. From 5 weeks, body weights were decreased significantly in the HFD + Res group compared with the HFD group (Fig. 2A). There was no significant difference in the daily food intake among the three groups (Fig. 2B). TC, TG, LDL-C, MDA, ALT and AST concentration levels were significantly higher in mice in the HFD group compared with mice in the ND group, and these levels were decreased significantly in the HFD + Res group compared with the HFD group (Fig. 2C and D). By contrast, HDL-C concentration levels were significantly lower in the HFD group compared with the ND group, and significantly increased in the HFD + Res group compared with the HFD group (Fig. 2C). Blood glucose levels were also recorded, and OGTT and ITT results are shown in Fig. 2E and G. In the OGTT, there was a significant decrease in the AUC in the HFD + Res group compared with the HFD group (Fig. 2F). Consistently, there was a statistically significant difference in QUICKI values between the HFD + Res and HFD groups (Fig. 2H).

To investigate the effect of Res on hepatic lipid deposition in mice, H&E staining and oil red O staining were performed using mouse liver tissues. H&E staining of mouse liver tissues revealed uniform cell cytoplasm in ND mice and fewer lipid droplets (Fig. 3A). In comparison, hepatocyte staining in the HFD group revealed disordered cellular structure and more lipid droplets (Fig. 3B). In the HFD + Res group, the morphology of the liver tissue and the number of lipid droplets were intermediate to those of the ND and HFD groups (Fig. 3C). Liver cell staining with oil red O showed that cells from the ND group contained blue nuclei with a small number of orange-red lipid droplets (Fig. 3D). The HFD group exhibited numerous orange-red lipid droplets (Fig. 3E), whereas after Res treatment, the number of lipid droplets decreased (Fig. 3F).

To gain further insights, the effect of Res on lipid metabolism and gene expression was investigated. The results revealed that miR-33 expression was significantly higher (Fig. 4A) and SIRT6 mRNA expression was significantly lower (Fig. 4B) in liver tissue of the HFD group compared with the ND group. Western blot analysis to assess SIRT6 expression in tissues revealed a significant decrease in the HFD group (Fig. 4C and D). In the HFD + Res group, Res reversed the increase in miR-33 expression and the decrease in SIRT6 expression. It was also found that mRNA expression levels (Fig. 4E) and protein expression levels (Fig. 4F and G) of ACC, FASN and SREBP1 were increased in the HFD group compared with the ND group, whereas PPARγ, PGC1α and CPT1 mRNA and protein expression levels were decreased. However, these changes in the expression levels of liver genes and proteins were reversed in the HFD + Res group (Fig. 4E-G). These findings indicated that Res improved basic metabolic parameters and changed the expression levels of metabolism-related genes in mice fed a HFD supplemented with Res.

To further examine the underlying mechanism, PA-induced HepG2 cells were used to investigate the effect of Res on lipid metabolism and expression levels of miR-33 and SIRT6 in vitro. First, a high-fat model was constructed by inducing HepG2 cells with PA, and changes after Res treatment were observed. It was found that lipid deposition in HepG2 cells improved after the addition of Res (Fig. 5A). Next, changes in miR-33 and SIRT6 mRNA expression were detected in PA-induced HepG2 cells. RT-qPCR analysis demonstrated that miR-33 expression was increased significantly (Fig. 5B) and mRNA levels of SIRT6 decreased significantly (Fig. 5C) in PA-induced HepG2 cells compared with CON cells. Additionally, treatment with Res decreased miR-33 expression and increased SIRT6 expression in PA-induced HepG2 cells (Fig. 5B and C). Furthermore, western blotting indicated that, with Res supplementation, protein expression levels of SIRT6, PPARγ, PGC1α and CPT1 (Fig. 5F and G) were increased compared with those in PA-induced HepG2 cells, whereas the expression levels of ACC, FASN and SREBP1 were decreased (Fig. 5F and G). Additionally, RT-qPCR results revealed that incubation with Res reversed mRNA expression levels of the aforementioned genes in PA-induced HepG2 cells (Fig. 5H). These results indicated that Res significantly changed the expression of metabolism-related genes in vitro.

100 nmol/l miR-33 mimics or mimic controls were transfected into HepG2 cells. The results demonstrated that, after transfection with miR-33 mimic, the expression levels of miR-33 in cells were significantly increased (Fig. 6A). Transfection of miR-33 mimics (but not a negative control miRNA) led to a significant decrease in SIRT6 mRNA (Fig. 6B) and protein expression levels (Fig. 6C and D), and promoted lipid deposition in HepG2 cells (Fig. 6E).

miR-33 inhibitor or inhibitor controls (100 nmol/l) were transfected into HepG2 cells. The results demonstrated that, after transfection with miR-33 inhibitor, the expression levels of miR-33 in cells were significantly decreased (Fig. 7A). Similarly, miR-33 inhibition significantly increased SIRT6 mRNA expression (Fig. 7B) and protein expression compared with those in the inhibitor control group (Fig. 7C and D). HepG2 cells were induced by PA for 24 h after transfection, and then RT-qPCR was used to analyze the effect of miR-33 inhibitor transfection on lipid metabolism-related genes, and Oil Red O staining was used to analyze the effect of miR-33 inhibitor on lipid deposition. RT-qPCR results revealed that transfection with the miR-33 inhibitor significantly increased the mRNA expression levels of PPARγ, PGC1α and CPT1, decreased the mRNA expression levels of ACC, FASN and SREBP1 (Fig. 7E), Simultaneously, Oil Red O staining revealed decreased lipid deposition in HepG2 cells after miR-33 inhibitor transfection (Fig. 7F).

To further analyze the effect of SIRT6 overexpression on intracellular lipid metabolism, SIRT6-pcDNA 3.1 was transfected into HepG2 cells. The results demonstrated that, after transfection with SIRT6-pcDNA 3.1, the expression levels of SIRT6 mRNA were significantly increased (Fig. 8A) and the protein expression levels of SIRT6 were also significantly increased compared with those in the pcDNA 3.1 group (Fig. 8B and C). HepG2 cells were induced by PA for 24 h after transfection, and RT-qPCR and western blotting were used to analyze the effect of SIRT6 overexpression on its downstream lipid metabolism-related genes, and Oil Red O staining was used to analyze the effect of SIRT6 overexpression on lipid deposition. The results revealed that transfection of HepG2 cells with SIRT6 overexpression vector increased PPARγ, PGC1α and CPT1 expression, and decreased ACC, FASN and SREBP1 mRNA (Fig. 8D) and protein expression levels (Fig. 8E and F). Simultaneously, Oil Red O staining revealed decreased lipid deposition in HepG2 cells after SIRT6 overexpression (Fig. 8G).

To confirm the direct binding of miR-33 to the 3'-UTR of SIRT6 mRNA, luciferase reporter constructs were generated containing the miR-33 binding site (SIRT6-WT) and its mutant sequence (SIRT6-MT). The results of the luciferase reporter assay demonstrated that the miR-33 mimic significantly decreased the luciferase activity in the SIRT6-WT group (Fig. 9A and B) and had no obvious effect in the SIRT6-MT group (Fig. 9B), indicating that SIRT6 was a direct target of miR-33.