Preventing High Fat Diet-induced Obesity and Improving Insulin Sensitivity through Neuregulin 4 Gene Transfer

Neuregulin 4 (NRG4), an epidermal growth factor-like signaling molecule, plays an important role in cell-to-cell communication during tissue development. Its function to regulate energy metabolism has recently been reported. This current study was designed to assess the preventive and therapeutic effects of NRG4 overexpression on high fat diet (HFD)-induced obesity. Using the hydrodynamic gene transfer method, we demonstrate that Nrg4 gene transfer in mice suppressed the development of diet-induced obesity, but did not affect pre-existing adiposity and body weight in obese mice. Nrg4 gene transfer curbed HFD-induced hepatic steatosis by inhibiting lipogenesis and PPARγ-mediated lipid storage. Concurrently, overexpression of NRG4 reduced chronic inflammation in both preventive and treatment studies, evidenced by lower mRNA levels of macrophage marker genes including F4/80, Cd68, Cd11b, Cd11c, and macrophage chemokine Mcp1, resulting in improved insulin sensitivity. Collectively, these results demonstrate that overexpression of the Nrg4 gene by hydrodynamic gene delivery prevents HFD-induced weight gain and fatty liver, alleviates obesity-induced chronic inflammation and insulin resistance, and supports the health benefits of NRG4 in managing obesity and obesity-associated metabolic disorders.

age-matched chow-fed mice. There was no difference in the Nrg4 transcript level in subcutaneous white adipose tissue (SubWAT) and BAT between normal-weight and obese mice. Results in Fig. 2a showed that hydrodynamic injections of pLIVE-NRG4 plasmids (on day 1 and day 28) were sufficient to protect mice from HFD-induced weight gain. After nine weeks of HFD feeding, pLIVE-NRG4-injected animals had an average body weight of 36 grams, 10 grams less than that of the control animals (~46 g) injected with control plasmid pLIVE-SEAP carrying the secreted embryonic alkaline phosphatase gene. The difference in body size was visually differentiable (Fig. 2c). Analysis of body composition revealed that the difference in body weight was primarily from fat mass (~49% reduction), not from lean mass (Fig. 2b). The average food intake between the two groups of animals was similar (Fig. 2d).
To test the effect of NRG4 increase on diet-induced obesity, we overexpress NRG4 by performing hydrodynamic gene delivery. Hydrodynamic-based gene transfer is an effective gene delivery method, which has been widely used for gene expression and functional analysis in whole animals [10][11][12][13][14] . Expression of NRG4 in animals receiving pLIVE-NRG4 plasmid transfer was confirmed by real-time PCR (RT-PCR) analysis and the results revealed that the mRNA level in the liver of the treated animals at the end of the experiment was approximately 97-fold higher than that of the control animals receiving two injections of pLIVE-SEAP control plasmids (Fig. 2f,  Fig. S1). A high level of Nrg4 gene transfer was also confirmed by regular PCR analysis (Fig. 2e). In addition, an approximate 2-3-fold increase of Nrg4 expression was detected in the adipose tissues including BAT, eWAT, and SubWAT, but not in the pancreas (Fig. 2f). These results prove that NRG4 overexpression was responsible for blocking high fat diet-induced weight gain.

NRG4 suppresses chronic inflammation in eWAT and stimulates BAT thermogenesis.
Consistent with the lower fat mass in pLIVE-NRG4-injected animals, H&E staining of adipocytes (Fig. 3a) and further quantitative analysis (Fig. 3b) revealed that Nrg4 gene transfer blocked the hypertrophy of adipocytes in eWAT and SubWAT by more than 50%, which was in line with the weight difference in the fat pads (Fig. 3c). Crown-like structures that are seen in the adipose tissue sections of the control animals are not detected in mice with the hydrodynamic delivery of pLIVE-NRG4 plasmids, indicating that NRG4 overexpression is capable of preventing diet-induced macrophage infiltration. Results from RT-PCR analysis (Fig. 3d) show that Nrg4 gene transfer suppressed HFD-induced expression of macrophage-specific marker genes including F4/80, Cd68, Cd11b and Cd11c by ~88%, ~87%, ~64% and ~83%, respectively. Nrg4 gene transfer also reduced monocyte chemotactic protein 1 (Mcp1) expression, a chemokine gene that regulates migration and infiltration of macrophages. Similarly, Nrg4 gene transfer enhanced M2 macrophage marker gene Cd163 (Supplemental Fig. S2). These results demonstrate the anti-inflammatory effect of NRG4 overexpression on diet-induced obesity. Nrg4 gene transfer also increased the expression of adiponectin (~3.2-fold) and adipose triglyceride lipase (Atgl, ~3.8-fold) genes.
Results in Fig. 3a (lowest panel) show fewer lipid droplets in BAT of pLIVE-NRG4-injected mice than those of the control animals, as determined by the density of vacuole-type structures. BAT is known as the thermogenic organ, facilitating utilization of extra glucose or lipids to generate heat 15,16 . To evaluate the effect of NRG4 overexpression on body temperature, we measured the rectal temperature of the animals with a digital thermometer. The results in Supplemental Fig. S1 showed that pLIVE-NRG4-injected mice had an increase of ~1.0 °C body temperature one day after the gene transfer and maintained this relatively higher body temperature for two weeks. Moreover, NRG4 overexpression also enhanced the expression of BAT thermogenic genes, including uncoupling protein Ucp1 (~ 2.2-fold), cell death-inducing DNA fragmentation factor-α -like effector A (Cidea, ~2.4-fold), and the type 2 deiodinase (Dio2, ~2.0-fold) (Fig. 3e), suggesting that Nrg4 gene transfer influences energy expenditure in BAT. Similar effects of Nrg4 gene transfer on browning marker genes were also seen in the inguinal WAT (Supplemental Fig. S4).
Nrg4 gene transfer prevents diet-induced hyperinsulinemia and insulin resistance. To assess the impact of Nrg4 gene transfer on insulin sensitivity and glucose homeostasis, glucose tolerance tests were performed and the results in Fig. 4a,b revealed a lower peak level and much higher clearance rate of intraperitoneally injected glucose in animals with Nrg4 gene transfer. pLIVE-NRG4-injected mice were more sensitive to insulin administration than the control mice (Fig. 4c). Nrg4 gene transfer dramatically suppressed an HFD-induced increase in the blood insulin level (7.3 μ g/L vs. 1.6 μ g/L, Fig. 4d). Compared to the control animals, pancreatic mRNA levels of Insulin1 and Insulin2 genes were markedly lower in mice with Nrg4 gene transfer (Fig. 4e). NRG4 overexpression also suppressed glucose 6-phosphate (G6p) gene expression, but not the phosphoenolpyruvate carboxykinase gene (Pepck) that is directly involved in glucogenesis (Fig. 4f). These results validate our hypothesis that NRG4 protects animals from obesity-associated insulin resistance.
Nrg4 gene transfer blocks hepatic lipogenesis and lipid storage. Obesity is associated with fatty liver disease 17 . Results in Fig. 5a show that Nrg4 gene transfer inhibits liver enlargement by 2.4-fold compared to that of the control animals (1.0 g vs. 2.4 g); a photograph of the liver discerned this outcome (Fig. 5b). H&E-stained liver sections showed a relatively normal liver structure in mice with Nrg4 gene transfer, compared to the extensive hepatocyte vacuolation in control animals (Fig. 5c). Oil Red O staining of liver sections also confirmed the difference (Fig. 5d). These results provide direct evidence in support that NRG4 overexpression is capable of blocking hepatic lipid accumulation.
NRG4 overexpression improves insulin sensitivity in obese mice without changing body weight. The preventive effect of NRG4 on HFD-induced obesity has been clearly demonstrated. Next, we assessed the effects of Nrg4 gene transfer on obese mice. Diet-induced obese mice (average body weight 56 g) were randomly divided into two groups and injected with pLIVE-NRG4 or pLIVE-SEAP plasmids using a hydrodynamics-based procedure. Results in Fig. 6a-c show that Nrg4 gene transfer failed to reduce body weight and did not change fat or lean mass. There was no difference in food intake between mice receiving pLIVE-NRG4 or pLIVE-SEAP plasmid DNA. However, NRG4-treated mice exhibited a much higher clearance rate of glucose in the glucose tolerance test (GTT) assay (Fig. 6d), which was better represented by the calculated area under the curve (AUC, Fig. 6e). Improvement of insulin sensitivity in obese mice was also seen in an insulin tolerance test. Taken together, these data demonstrate that Nrg4 gene transfer improves insulin sensitivity in obese mice.
NRG4 overexpression reduced chronic inflammation in obese mice. H&E staining was performed and the results in Fig. 7a-c show no structural change of the liver, BAT, and SubWAT among the treated or control animals. Large lipid droplets in the liver and adipocytes were found in both groups of animals, indicating that lipid storage in these tissues was not affected. Gene expression analysis showed that Nrg4 gene transfer did not change the mRNA levels of Srebp1c and Acc1, but did reduce Fas and Scd1 mRNA levels by ~30% and ~42%,  (Fig. 7d). In addition, NRG4 overexpression reduced mRNA levels of Pparγ1 (~22%) and Mgat1 (~28%), but not that of Pparγ2 and Cd36 (Fig. 7e).
Results in Fig. 8a show significant macrophage infiltration into EWAT of the obese animals, as reflected by multiple crown-like structures in H&E-stained EWAT sections. However, the crown-like structures were less obvious in animals with Nrg4 gene transfer. RT-PCR was carried out to confirm the difference in the macrophage activation in the WAT of the treated and controlled animals. Results in Fig. 8b show a significant reduction in the mRNA levels of macrophage marker genes in mice with Nrg4 gene transfer, including F4/80, Cd68, Cd11b, and Cd11c in EWAT of obese mice. In addition, the mRNA level of inflammatory factor gene Mcp1 was also reduced by ~40%. A similar effect of Nrg4 gene transfer on gene expression was detected in the liver and BAT of mice with NRG4 overexpression (Fig. 8c,d). These results suggest that NRG4 overexpression is capable of inhibiting obesity-related chronic inflammation.

Discussion
In this study, we systematically examined the effects of Nrg4 gene transfer on high fat diet-induced obesity and obesity-associated metabolic changes. The results show that Nrg4 gene transfer protects mice from diet-induced weight gain and adiposity (Figs 2 and 3), and inhibits fatty liver and insulin resistance (Figs 4 and 5). The beneficial effects are correlated with an inhibition of lipogenesis, lipid storage, and chronic inflammation (Figs 3 and 5). NRG4 overexpression in obese mice did not affect pre-existing fat and body weight, but reduced chronic inflammation and improved obesity-related insulin resistance (Figs 6-8).
Obesity increases the risk of type 2 diabetes, cardiovascular diseases, and certain cancers, which are among the leading causes of death worldwide 18 . Although a cure for obesity is important, treatments are costly, and success so far is limited due to a lack of safe and effective methods. Prevention that is capable of benefiting a larger majority of the population and halting obesity-related metabolic disorders is more desirable and effective. Weight control programs through dietary intervention and enhanced physical activities have been popular in the past, but the successes have been limited to a small percentage of people since persistence in restricting food intake and maintaining regular exercise can be difficult to accomplish 19 . Evidently, efforts to explore new methods and strategies are urgently needed in dealing with the obesity epidemic.
NRG4 was originally identified as a factor that plays a critical role in tissue development 1 . However, a high level of expression in BAT and cold-induced expression in WAT suggest that NRG4 may act as an adipokine with a distinct function 8 . Since BAT plays an essential role in regulating body temperature and thermogenesis 15,16 , as well as maintaining systemic lipid and glucose homeostasis [20][21][22] , an elevated level of NRG4 in BAT indicates that NRG4 may be involved in energy metabolism and maintenance of metabolic homeostasis. A lower mRNA level of Nrg4 (Fig. 1) in EWAT and in the liver of diet-induced obese mice appears to support such a prediction. In addition, a decrease in NRG4 levels was also observed in patients with non-alcoholic fatty liver disease (NAFLD) 23 . NRG4 significantly inhibited lipogenesis by suppressing Srebp1c and its target genes, which is consistent with the previous results obtained from NRG4-deficient mice 9 . Elevating SREBP1c expression promotes lipid accumulation in the liver and fat tissue 24,25 , and HFD feeding induces SREBP1c-mediated lipogenesis in obese mice 26,27 . However, inhibition of lipogenesis exerts a positive effect on metabolic diseases 28,29 . Therefore, suppression of lipogenesis by NRG4 plays, at least in part, an important role in preventing diet-induced obesity. In addition, Nrg4 gene transfer inhibited expression of Pparγ and its target genes (Fig. 5e). PPARγ is a transcription factor regulating the expression of genes responsible for lipid storage and adipocyte differentiation 30 . Previous studies have shown that elevated expressions of Pparγ, Cd36, and Mgat1 genes are directly linked to hepatic steatosis [31][32][33] . Therefore, by targeting the hepatic PPARγ pathway, NRG4 protects against obesity-induced hepatic steatosis.
Several studies have demonstrated the interaction of NRG4 with inflammation. NRG4 levels in animals with ulcerative colitis (UC) and Crohn's disease are reduced compared to un-inflamed controls. NRG4 treatment blocked inflammatory cytokine-induced apoptosis in colon epithelial cells and protected animals from experimental inflammatory bowel disease 3,4 . This conclusion agrees with our observation that NRG4 significantly reduced crown-like structures in EWAT and lowered expression of macrophage marker genes such as F4/80, Cd68, Cd11, and inflammatory factor Mcp1 (Figs 3 and 8). These results also align with our previous reports that suppressing HFD-induced inflammation by either antioxidant 34 or macrophage elimination 35 is sufficient to block HFD-induced obesity. However, results shown in Fig. 6a,b suggest that blockage of inflammation is not sufficient to reduce pre-existing body weight and fat mass in obese mice, indicating further strategy is needed to mobilize stored fat.
Reduction of inflammation by NRG4 appears to benefit insulin sensitivity in both prevention and treatment studies. It is known that chronic low-grade tissue inflammation is a major detrimental factor for obesity-associated metabolic disorders such as insulin resistance. Long-term nutrient overload in adipose tissue and the liver increases a stress response, leading to the recruitment and activation of macrophages and thus the activation of inflammatory pathways 36 . Consequently, insulin action is blunted in the stressed tissues by inhibiting insulin-signaling transduction, causing cellular insulin resistance 37,38 . Therefore, inhibiting inflammatory and macrophage infiltration by NRG4 contributes to improved insulin sensitivity in both the prevention and treatment studies. This conclusion is in agreement with the results of the studies where normal weight rats and db/db mice were used to assess whether acute injection of NRG1 protein improves glucose tolerance through   In conclusion, the current study demonstrates that Nrg4 gene transfer inhibits lipogenesis and reduces chronic inflammation, resulting in the prevention of HFD-induced adiposity and fatty liver, and the improvement of insulin sensitivity. These findings provide direct evidence in support of the potential health benefits of NRG4 in managing obesity and obesity-associated metabolic disorders.

Materials and Methods
Materials. C57BL/6 mice were purchased from Charles River (Wilmington, MA). High-fat diet was obtained

Plasmid construction.
A coding region of mouse Nrg4 (NCBI GenBank, NM_032002) was cloned from complementary DNA sequences of C57BL/6 mouse liver and inserted into a pLIVE vector at the NheI and XhoI cutting sites. The insertion in the new plasmid (pLIVE-NRG4) was confirmed by DNA sequencing. The control plasmid pLIVE-SEAP contains the same backbone and a secreted alkaline phosphatase gene. Both pLIVE-NRG4 and pLIVE-SEAP plasmids were purified by CsCl-ethidium bromide density-gradient ultracentrifugation and kept in saline (0.9% sodium chloride). Purity of the plasmids was verified by absorbency ratio at 260 and 280 nm and by 1% agarose gel electrophoresis.

Animals and treatment. All procedures performed on animals were approved by the Institutional Animal
Care and Use Committee at the University of Georgia, Athens, Georgia and the methods were carried out in accordance with the approved protocol #A2014-07-008-Y1-A0. For the prevention study, C57BL/6 mice (male, 8-week old) were housed under standard conditions with a 12 h light-dark cycle and fed an HFD for a total of 9 weeks. For the treatment study, mice were first fed an HFD for 25 weeks, reaching a body weight of 56 grams on average. Gene transfer was performed on the obese mice and the treated animals were continued on an HFD for additional 3 weeks. Hydrodynamic gene delivery was performed according to a pre-established procedure 10,41,42 . Briefly, an appropriate volume of saline solution (equivalent to 8% lean mass) containing 20 μ g pLIVE-NRG4 or pLIVE-SEAP was injected through the tail vein within 5-8 sec. The body weight of each mouse was measured on an electronic balance, and body composition was analyzed using an EchoMRI-100 TM (Echo Medical Systems, Houston, TX). Food intake per mouse was calculated based on the amount consumed divided by time and the number of mice per cage. Blood was collected at the desired time using a Microvette-CB300-LH. Rectal temperature of the mice was measured at the desired time using a specially designed Thermocouple Meter.
Histochemical analysis. After the mice were euthanized, the liver, EWAT, SubWAT, and BAT were collected and fixed in 10% formalin. Paraffined tissues were sectioned at a thickness of 6 μ m and stained with H&E solution. For Oil Red O staining of liver sections, freshly collected liver samples were frozen in liquid nitrogen. Frozen sections (8 μ m) were made and stained with 0.2% Oil Red O in 60% of isopropanol for 20 min and washed three times with phosphate buffered saline. A microscopic examination was performed and photographs were taken under a regular light microscope.

Glucose tolerance test (GTT) and insulin tolerance test (ITT).
The analyses were performed on week nine post gene transfer in the prevention study and on week three after gene transfer to obese mice in the treatment study. For GTT, mice were fasted for 6 h and injected intraperitoneally with glucose at 1.5 g/kg body weight. Blood samples were collected from the tail vein and glucose levels were measured using a glucometer. For ITT, mice were fasted for 4 h and intraperitoneally injected with insulin (0.75 U/kg, Eli Lilly Indianapolis, IN). Blood samples were collected from the tail vein and glucose levels were measured using a glucometer.
Gene expression analysis by real time PCR. Total RNA was isolated from the liver, pancreas, BAT, and EWAT using the TRIZOL reagent or an RNeasy kit. Two micrograms of total RNA were used for the first strand cDNA synthesis, as recommended by the manufacturer. Real time PCR was performed, using SYBR Green as an indicator, employing an ABI StepOne Plus Real Time PCR system. PCR was carried out for 40 cycles at 95 °C for 15 s and 60 °C for 1 min. Fluorescence was read during the reaction, allowing a continuous monitoring of the amount of PCR product. The data were normalized using Gapdh mRNA as an internal control. The primer sequences employed are summarized in Supplemental Table S1.

Statistical analysis
A statistical analysis was performed using the Student's t test. All data are reported as mean ± standard deviation (SD) with statistical significance set at P < 0.05.