The adenovirus expression vector was generated using the AdMax™ Expression System (Microbix Biosystems, Inc., Mississauga, ON, Canada). The shuttle and adenoviral backbone plasmid pDC315 carrying a loxP expression cassette in the E₁ region was purchased from Addgene, Inc. (Cambridge, MA, USA). The recombinant plasmid pDC315-Angptl4 was constructed by restriction enzyme (EcoRI and BamHI) cleavage and connection. The plasmid pBHG lox ΔE_1,3 Cre containing the Cre/loxP element was purchased from Microbix Biosystems, Inc. The recombinant adenoviral-Angptl4 (adv-Angptl4) virus was packaged, amplified in HEK293 cells and then purified by CsCl density gradient centrifugation at 100,000 × g for 16 h at 4°C.

A total of 24 male C57BL/6 healthy mice between 8 and 10 weeks old (~20 g) were obtained from the Shanghai Experimental Animal Center, Chinese Academy of Sciences (Shanghai, China) and were housed in a room under controlled temperature (23±1°C) with free access to water. Mice were fed standard chow for the first week to allow them to adjust to the new environment. Subsequently, mice were divided into three groups randomly. As a control, mice in the first group (normal control) were fed with standard mouse chow. The other two groups were provided with a HFD for 6 weeks, and then adenoviral (adv) and adv-Angptl4 virus were injected via the tail vein in the second (model control) and third (Angptl4⁺) groups of mice, respectively. The normal control group continued to be fed with standard chow, whereas the model control and Angptl4⁺ groups were fed with HFD for an additional two weeks. The mice in the three groups were weighed once a week subsequent to injection with the virus. Two weeks after viral injection, a series of experiments were performed. All of the experiments were conducted under institutional guidelines for the humane treatment of laboratory animals, and the study was approved by the ethics committee of Dahua Hospital in Shanghai Xuhui (Shanghai, China).

Serum Angptl4 concentrations of the three groups of mice were measured using a sandwich enzyme-linked immunosorbent assay, with the experimental kit purchased from JRDUN Biotechnology, Co., Ltd. (Shanghai, China).

All mice were placed into other clean cages with free access to water, however no food, for 6 h. Subsequent to fasting, the mice were weighed, the tip of the tails were clipped in order to obtain blood, and the fasting blood glucose levels were measured with the OneTouch small glucometer. Mice were then rapidly administered with an intraperitoneal injection of glucose (1 g/kg of body weight). Blood was drawn from the mice tail tip at various time points: T15, T30, T60, T90, T120 and T150 min for measurement of glucose concentration.

The serum levels of TG and TC were detected by an enzyme-coupled assay, the ALT and AST levels were detected by the Reitman-Frankel method (14), FFA levels were detected using the Copper Color Method with a semi-automatic or fully-automatic spectrophotometer. All of the serum indices were measured using the kit from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). The serum levels of fasting FIN were detected using the ¹²⁵I-insulin radio immunoassay kit from the North Biotechnology Institute (Beijing, China). The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated using the following formula: Fasting blood glucose (FBG; mmol/l) x FIN (µIU/ml)/22.5. Subsequently, the data were logarithmically transformed in order to obtain a more accurate value.

Tissue blocks of livers were resected and fixed in 10% formalin for 48 h, then were embedded in paraffin. Subsequently, they were sectioned and stained with HE. Finally, the degree of steatosis and inflammation were analyzed under a light microscope (magnification, ×100).

Mice were sacrificed by CO₂ inhalation, then fresh liver tissue samples were obtained from the mice in the process of dissecting the mice. The liver tissues (~100 mg) were rapidly weighed, and the tissues were pulverized in a mortar using liquid nitrogen. Radioimmunoprecipitation assay (RIPA) lysis buffer was then added (3 ml RIPA/gram of tissue) premixed with proteinase and phosphatase inhibitors. The concentration of liver tissue protein was quantified using the Bicinchoninic Acid kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA). A total of 80 µg protein for each sample was loaded and underwent 10/15% sodium dodecyl sulfate-polyacrylimide gel electrophoresis for separation. The blots were blocked with 5% non-fat milk at room temperature for 1 h and were incubated overnight at 4°C with the following primary antibodies: Polyclonal rabbit Angptl4 (bs-1087R; 1:100; Bioss, Inc., Woburn, MA, USA), monoclonal rabbit janus kinase 2 (JAK2) (#3230; 1:1,000; Cell Signaling Technology, Inc., Danvers, MA, USA), rabbit monoclonal phosphorylated (P-) JAK2 (#8082; 1:1,000; Cell Signaling Technology, Inc.), rabbit polyclonal insulin receptor substrate 1 (IRS-1) (#2382; 1:1,000; Cell Signaling Technology, Inc.), rabbit polyclonal P-IRS-1 (#2381; 1:1,000; Cell Signaling Technology, Inc.), mouse monoclonal signal transducer and activator of transcription 3 (STAT3) (ab119352; 1:5,000; Abcam, Cambridge, MA, USA); rabbit monoclonal P-STAT3 (ab76315, 1:200,000; Abcam), rabbit polyclonal protein kinase B (AKT) (#9272; 1:1,000; Cell Signaling Technology, Inc.); rabbit polyclonal P-AKT (#9271; 1:1,000; Cell Signaling Technology, Inc.), rabbit monoclonal glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (#5174; 1:1,000; Cell Signaling Technology, Inc.). In order to ensure equal amount of protein loaded, levels were normalized to that of GAPDH. The blots were incubated for 1 h at 37°C temperature with the goat anti-rabbit (A0208) and goat anti-mouse (A0216) poly-clonal secondary antibodies (conjugated with horseradish peroxidase; 1:1,000; Beyotime Institute of Biotechnology, Shanghai, China) following three washes with Tris-buffered saline with Tween-20 (TBST) buffer. The signals were visualized using enhanced chemiluminescence (EMD Millipore, Billerica, MA, USA) following an additional wash with TBST. Intensities were measured using ImageJ software, version 1.4 (National Institutes of Health, Bethesda, MD, USA) and were reported as the relative pixel normalized to that of GAPDH.

Experiments were performed with eight mice per group with values expressed as the mean ± standard deviation. Statistical analysis was determined using one-way analysis of variance with Graphpad Prism software, version 6.0. P<0.05 was considered to indicate a statistically significant difference.

In order to verify the role of Angptl4 as a circulating hormone in mice subsequent to viral injection, a sandwich enzyme-linked immunoassay was conducted for measurement of the protein in mouse serum. Serum Angptl4 concentrations in the normal and model control groups were not significantly different, with values of 2.44±0.32 ng/ml and 2.24±0.21 ng/ml, respectively. The serum concentration in the Angptl4⁺ group was increased compared with that of the control groups, with a concentration of 3.86±0.24 ng/ml (Fig. 1). These results suggested that the recombinant adv-Angptl4 virus had been constructed successfully, and that the Angptl4 gene was highly expressed in HFD mice.

A week subsequent to the injection of the virus, no significant differences were observed between the model control and Angptl4⁺ groups. After two weeks, in the model control group, a significant increase in body weight was observed, with the mice weighing 28±1.07 g at the end of the experiment, with an increase ratio of 20.4%, markedly higher than that of the normal control group. However, Angptl4 reduced the weight growth rate in HFD mice, with an increase of 10.5%, weighing 26.13±0.64 g (Fig. 2A). The levels of serum indices TG, TC and FFA associated with blood lipids were detected, and the results indicated that all index levels in the Angptl4⁺ group were significantly increased compared with that of the control groups. The levels in the model control were significantly higher than that of the normal control group (Fig. 2B–D).

Angptl4 overexpression reduced the weight growth rate almost by half in HFD mice, and significantly increased the levels of the major serum indices in the serum of HFD mice, suggesting that subsequent to infection with Angptl4 in HFD mice, the blood lipid levels and lipolysis were increased, which may have directly induced hyperlipidemia.

In order to investigate the influence of Angptl4 on liver function in HFD mice, HE staining was conducted in paraffin-embedded liver tissue sections from the three groups. The levels of ALT and AST, which are associated with liver function, were detected. The results indicated that hepatocyte swelling occurred in the livers of model control mice, together with increased vacuolization in the liver cytoplasm. It is notable that the swelling was more marked with larger fat vacuoles visible in the liver cytoplasm in the Angptl4⁺ group (Fig. 3A). The ALT and AST levels in the model control group were marginally upregulated, whereas were significantly increased in the Angptl4⁺ group, with the normal control group as the control (Fig. 3B and C).

All results indicated that overexpression of Angptl4 serves a negative role in liver steatosis in HFD, and resulted in damage to liver function of HFD mice.

To validate the effect of Angptl4 on blood glucose levels, the glucose tolerance of the three groups was investigated. With the normal control group mice as a control, it was identified that the model control group mice exhibited a rise in blood glucose levels throughout the glucose tolerance curve and a maximum glucose concentration at 60 min. This increase was present, but less marked in the Angptl4⁺ group, with results similar to thos of the normal control. These results indicated that Angptl4 overexpression can improve glucose tolerance in HFD mice (Fig. 4).

The effects of Angptl4 on FIN levels were observed, and insulin resistance was assessed in HFD mice. The results indicated that overexpression of Angptl4 markedly reduced FIN (Fig. 5A) and FBG levels (Fig. 5B). HOMA-IR was increased in the model control group, together with a lower insulin sensitivity compared with the normal control group, while these levels were reduced back to levels similar to the normal control in the Angptl4⁺ group (Fig. 5C), suggesting that Angptl4 can alleviate insulin resistance and enhance insulin sensitivity in HFD mice.

IRS-1 links the insulin receptor kinase and its downstream serine kinases, including phosphoinositide 3-kinase (PI3K) and AKT phosphorylation. Under the insulin resistant state, the effects of the PI3K-AKT signal transduction pathway on insulin stimulation are reduced (15). The JAK2-STAT3 signaling pathway is involved in the lipid metabolism through mediation by leptin (16). In the current study, western blot analysis was used to detected the phosphorylation levels of IRS-1, AKT, JAK2 and STAT3, in addition to expression of Angptl4, in the liver samples. The results indicated that with the normal control group as the control, the phosphorylated proteins and Angptl4 were downregulated in the model control group, however were upregulated in the in Angptl4⁺ group to levels similar to those of the normal control group (Fig. 6).