Lipid droplets in the liver are coated with the perilipin family of proteins, notably adipocyte differentiation-related protein (ADRP) and tail-interacting protein of 47 kDa (TIP47). ADRP is increased in hepatic steatosis and is associated with hyperlipidemia, insulin resistance, and glucose intolerance. We have shown that reducing ADRP in the liver via antisense oligonucleotide (ASO) treatment attenuates steatosis and improves insulin sensitivity and glucose tolerance. We hypothesized that TIP47 has similar effects on hepatic lipid and glucose metabolism. We found that TIP47 mRNA and protein levels were increased in response to a high-fat diet (HFD) in C57BL/6J mice. TIP47 ASO treatment decreased liver TIP47 mRNA and protein levels without altering ADRP levels. Low-dose TIP47 ASO (15 mg/kg) and high-dose TIP47 ASO (50 mg/kg) decreased triglyceride content in the liver by 35% and 52%, respectively. Liver histology showed a drastic reduction in hepatic steatosis following TIP47 ASO treatment. The high dose of TIP47 ASO significantly blunted hepatic triglyceride secretion, improved glucose tolerance, and increased insulin sensitivity in liver, adipose tissue, and muscle. These findings show that TIP47 affects hepatic lipid and glucose metabolism and may be a target for the treatment of nonalcoholic fatty liver and related metabolic disorders.
- lipid droplet
nonalcoholic fatty liver disease (NAFLD) has reached epidemic proportions in the United States and worldwide and is closely associated with obesity and type 2 diabetes (1, 4, 20, 31). As in humans, mice fed a high-fat diet develop fatty liver characterized by excessive accumulation of lipid droplets, consisting of a neutral lipid core, mainly triglycerides, surrounded by a phospholipid monolayer (2, 20, 30). The surface of lipid droplets is coated by a perilipin family of proteins, metabolic enzymes, and vesicle trafficking proteins (2–5). Adipose differentiation-related protein (ADRP) (also named perilipin 2) and tail-interacting protein (TIP) 47 (also named perilipin 3) are the major lipid droplet proteins in hepatocytes (2–5, 17). Ablation of Adfp gene decreased hepatic steatosis, increased very low density lipoprotein (VLDL) secretion, and improved insulin sensitivity in ob/ob mice (7, 8). We have shown that an antisense oligonucleotide (ASO) against ADRP reduced steatosis and VLDL secretion, and enhanced hepatic insulin sensitivity in ob/ob and diet-induced obese (DIO) mice (16, 30).
As with ADRP, TIP47 is widely expressed in hepatocytes, enterocytes, macrophages, and other tissues and is increased in response to lipid loading (2, 6, 12, 13, 18, 26). In Adfp-null adipocytes, TIP47 compensates for the loss of ADRP and maintains triglyceride accumulation in lipid droplets (26). In HeLa cells, TIP47 is recruited to lipid droplets and displays apolipoprotein-like properties by inducing the formation of synthetic lipid droplet-like particles (6). Inhibition of TIP47 in HeLa cells decreases triglyceride incorporation and maturation of lipid droplets (6). However, the specific in vivo functions of TIP47 are unknown. The aim of this study was to determine whether TIP47 ASO treatment would reduce hepatic steatosis and circulating lipid levels and improve glucose homeostasis.
MATERIALS AND METHODS
Mouse model of hepatic steatosis.
Experiments were performed according to protocols reviewed and approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania, Perelman School of Medicine. Eight-week-old male C57BL/6J mice (Jackson Laboratory, Bar Harbor, ME) were housed (n = 5 per cage) under a 12:12-h light-dark cycle (lights on at 0700) and an ambient temperature of 22°C and allowed free access to water and food. We first determined whether 4 wk of a high-fat diet (HFD) would increase TIP47 and ADRP expression in liver. Mice were fed regular rodent chow diet (Lab Diet, Richmond, IN; catalog no. 5001, containing 4.5% fat, 49.9% carbohydrate, 23.4% protein; 4 kcal/g), or HFD (Research Diets, New Brunswick, NJ; catalog no. D12451 containing 45% fat, 35% carbohydrate, 20% protein, 4.7 kcal/g) (15, 24, 27, 28, 30). Body composition was measured with nuclear magnetic resonance (Echo MRI, Houston, TX) (16, 24, 30). VLDL secretion was measured after Poloxamer treatment, as described below (16, 24, 30). Mice were euthanized 3 days later, and livers were harvested for measurement of TIP47, ADRP, and triglyceride levels.
TIP47 antisense oligonucleotide (ASO) treatment.
Eight-week-old male C57BL/6J mice were fed HFD ad libitum, received saline vehicle, 15 mg/kg TIP47 ASO (low dose), 50 mg/kg TIP47 ASO (high dose), or control ASO via intraperitoneal injection twice a week for 4 wk, and they continued on HFD throughout the treatment. Chimeric second-generation ASOs were synthesized by ISIS Pharmaceuticals (Carlsbad, CA) and formulated in PBS (10, 16, 23, 30, 33). TIP47 ASO, ISIS 409003 (5′-CACAGTGTTGTCTAGGGCCT-3′), is a 20-mer phosphorothioate oligonucleotide complementary to the mRNA for mouse TIP47, and has 2′-O-methoxyethyl-modified ribonucleosides (2′-MOE) at the 3′- and 5′-ends with 2′-deoxynucleosides in between. A control oligonucleotide, ISIS 141923 (5′-CCTTCCCTGAAGGTTCCTCC-3′), contains similar chemical modifications, with no complementarity to known genes, including lipid droplet proteins. The cytosines in all of these molecules are methylated. The efficacy of TIP47 ASO was first tested in primary mouse hepatocytes, as described for other ASOs (10, 32). Food intake and body weight were measured twice a week.
Lipid and glucose testing.
To determine the rate of VLDL secretion from the liver, mice were fasted for 4 h in the morning (0800–1200), 3 days after ASO or saline treatment. A detergent, Polaxamer-407 (1 g/kg ip), was administered to inhibit lipases, and tail blood was obtained at time 0, and 1, 2, and 4 h later (16, 24, 30). Triglyceride levels were measured in serum with an enzyme assay, as previously described (16, 24, 30).
To determine glucose tolerance, mice were fasted overnight, glucose solution (2 g/kg) was injected intraperitoneally, and tail blood glucose was measured at time 0 (before glucose injection) and 15, 30, 60, 90, and 120 min later with a glucometer (One Touch Ultra, Johnson & Johnson, New Brunswick, NJ) (16, 24, 30). To determine tissue-specific insulin sensitivity, a hyperinsulinemic-euglycemic clamp was performed as described previously (30). An indwelling catheter was inserted in the right internal jugular vein and extended to the right atrium. Four days after surgery, the mice had regained their presurgery weight, and they were fasted for 6 h; then, they were administered a bolus injection of 5 μCi of [3-3H]glucose followed by continuous intravenous infusion at 0.05 μCi/min. Baseline glucose kinetics was measured for 120 min followed by hyperinsulinemic clamp for 120 min. A priming dose of regular insulin (16 mU/kg; Humulin; Eli Lilly, Indianopolis, IN) was given intravenously, followed by a continuous infusion at 2.5 mU·kg−1·min−1. Blood glucose was maintained at 120–140 mg/dl via a variable infusion rate of 20% glucose. 2-Deoxy-d-[1-14C]glucose (10 μCi) was injected 45 min before the end of the clamp, and blood samples were collected to estimate glucose uptake. The mice were euthanized, and liver, perigonadal white adipose tissue (WAT), and soleus muscle were excised, frozen immediately in liquid nitrogen, and stored at −80°C for subsequent analysis of glucose uptake (30).
Three days after the last TIP47 ASO, control ASO, or saline treatment mice were euthanized between 12 PM and 1 PM, blood was obtained via cardiac puncture, and liver samples were rapidly dissected, frozen in liquid nitrogen, and stored at −80°C. Serum insulin concentration was measured with an ELISA (Crystal Chem, Evanston, IL). Total and high-molecular-weight adiponectin concentrations were measured with an ELISA (Alpco, Salem, NH). Lipids were extracted from livers for triglyceride measurement (16, 24, 28, 30). RNA was extracted from livers, muscle and WAT of TIP47 using TRIzol reagent (Invitrogen, Carlsbad, CA), and expression of mRNA levels of TIP47, ADRP, and enzymes involved in lipogenesis and lipolysis were measured using real-time PCR (ABI Prism; Applied Biosystems, Foster City, CA) (16, 24, 28, 30). The level of mRNA expression was normalized to phosphoriboprotein (36B4).
TIP47 and ADRP protein levels in liver were determined by immunoblotting (16, 30). Liver samples were homogenized in lysis buffer containing 1% NP-40, 0.5% Triton, 10% glycerol, 0.15 M NaCl, 0.001 M EDTA, 0.5 M Tris·HCl, at pH 7.4, supplemented with complete protein inhibition cocktail tablet from Roche (Penzberg, Germany). Protein extracts (10–30 μg) were separated by 4–12% NuPAGE Bis-Tris gel (Invitrogen) and transferred to nitrocellulose membranes using wet transfer at 4°C. Membranes were then blocked with 5% nonfat dried milk for 1 h at room temperature, and then incubated with an anti-rabbit polyclonal antibody against TIP47 (Abcam, Cambridge, MA) at 1:250 dilution in 3% milk TBST or guinea pig polyclonal antibody against ADRP (Fitzgerald, Concord, MA) at 1:1,000 dilution overnight at 4°C on an orbital shaker. After washing 3 times with TBS with 0.1% (vol/vol) Tween 20, 1:5,000 dilution of horseradish peroxidase-conjugated goat anti-guinea pig, and goat anti-rabbit horseradish peroxidase antibody (Santa Cruz Technology, Santa Cruz, CA) was applied to the ADRP and TIP47 blot, respectively, for 1 h at room temperature. The membranes were washed for 1 h at room temperature and visualized with enhanced chemiluminescence (GE Healthcare, Piscataway, NJ). The membranes were stripped and blotted for GPDH (1:1,000) (Cell Signaling, Beverly, MA). Film autoradiograms were analyzed using laser densitometry and Image J (National Institutes of Health, Bethesda, MD) (16, 24, 30).
Liver samples were fixed in 10% buffered formalin overnight, washed with PBS, incubated with 30% sucrose, and frozen with optimal cutting temperature compound. Tissue sections were cut with a Leica cryostat and stained with Oil Red O to visualize neutral lipids (16, 28, 30). Other liver samples were embedded in paraffin and sectioned for immunohistochemical detection of TIP47 and ADRP. Tissue sections were deparaffinized with xylene and decreasing concentrations of ethanol. Antigen retrieval was done by microwaving, and slides were immersed in methanol/H2O2 solution, washed in tap water, 0.1 M Tris·HCl buffer, pH 7.6, blocked in 0.1 M Tris/2% FBS for 5 min, and incubated overnight at 4°C in either anti-TIP47 guinea pig polyclonal antibody (Progen, Heidelberg, Germany) or anti-ADRP guinea pig polyclonal antibody (Fitzgerald, Concord, MA) diluted 1:500 in 0.1 M Tris/2% FBS. After washing and then blocking in Tris/FBS, biotinylated IgG anti-rabbit diluted 1:1,000 in Tris/FBS was applied for 1 h at room temperature, then washed and blocked in Tris/FBS. Vector peroxidase diluted at 1:1,000 was applied for 1 h at room temperature. The slides were washed, and immunoreactive signal was detected using diaminobenzidine solution (Vector, Burlingame, CA). Slides were counterstained with hematoxylin and examined under bright field with Nikon 80i microscope. Images were captured with DS-Qi1MC camera and image analysis system.
Changes in various parameters were determined with t-test or ANOVA and post hoc Newman Keuls multiple-comparison test (GraphPad Prism, La Jolla, CA). P < 0.05 is considered significant.
HFD increases hepatic steatosis, TIP47 and ADRP expression, and triglyceride secretion.
Body weight and total fat content were increased slightly but not significantly in C57BL/6J mice fed a HFD for 4 wk (Fig. 1, A and B). On the other hand, there was a three-fold increase in liver triglyceride content (Fig. 1C), and triglyceride secretion (Fig. 1D) was significantly increased after 4 wk on HFD compared with the normal chow diet. The relative mRNA and protein levels of TIP47 and ADRP were significantly increased in the livers of mice on a HFD compared with those on a normal chow diet (Fig. 1, E–G).
TIP47 ASO reduces hepatic TIP47 expression without compensatory increase in ADRP.
Compared with saline and control ASO groups, TIP47 ASO treatment reduced hepatic TIP47 mRNA by ∼70% without changing ADRP mRNA level (Fig. 2, A and B). Immunoblotting of liver lysates showed that treatment with TIP47 ASO decreased the TIP47 protein level without changing ADRP liver protein level (Fig. 2C). This was confirmed by immunohistochemistry of liver sections, which showed a reduction in TIP47 immunostaining following TIP47 ASO treatment, compared with saline or control ASO (Fig. 3, A–C). TIP47 ASO treatment decreased hepatic steatosis, as evidenced by a drastic reduction in Oil Red O staining in liver sections of TIP47 ASO-treated mice compared with saline or control ASO (Fig. 3, D–F).
TIP47 ASO did not significantly change the relative expression of TIP47 and ADRP mRNA in white adipose tissue (Fig. 4, A and B). Furthermore, there was no apparent change in ADRP immunostaining in the liver following TIP47 ASO treatment (Fig. 4C).
TIP47 ASO reduces hepatic triglyceride content and triglyceride secretion.
The low and high doses of TIP47 ASO significantly decreased hepatic triglyceride content by 35% and 52%, respectively (Fig. 5A). Hepatic triglyceride secretion was measured to determine whether the reduction in hepatic triglyceride content following TIP7 ASO treatment was due to an increase in triglyceride secretion. The low-dose TIP47 ASO had no effect on hepatic triglyceride secretion, while the high-dose TIP47 ASO significantly reduced hepatic triglyceride secretion (Fig. 5B). Food intake and body weight were not affected by TIP47 ASO treatment (Table 1). The low-dose TIP47 ASO did not change the body composition; however, the high-dose TIP47 ASO decreased total body fat compared with the control ASO (Table 1).
We examined the effects of TIP47 ASO on enzymes involved in lipid metabolism (Fig. 6, A–E). TIP47 ASO treatment significantly reduced the mRNA expression of Dgat2 and Gpat, suggesting a major effect on lipogenesis.
Reduction of TIP47 improves glucose homeostasis.
Glucose levels were similar in the four groups of mice. Insulin levels were significantly lower, and total and high-molecular-weight (HMW) adiponectin levels were higher in the TIP47 ASO group (Table 1). We performed a glucose tolerance test to determine whether the reduction in hepatic steatosis in TIP47 ASO-treated mice affected glucose homeostasis. The high dose of TIP47 ASO significantly improved glucose tolerance (Fig. 5C). There was a trend toward improved glucose tolerance in the low-dose TIP47 ASO group, but this did not reach statistical significance, suggesting a dose-dependent effect of TIP47 ASO treatment (Fig. 5C).
To evaluate the effects of TIP47 ASO on insulin sensitivity in the liver and other organs, we assessed glucose kinetics under basal (fasting) and hyperinsulinemic-euglycemic clamp (30). The basal hepatic glucose production after 6 h of fasting was similar among all groups (Fig. 7A). Under hyperinsulinemc clamp, the high dose of TIP47 ASO significantly increased the glucose infusion rate (GIR) needed to reach the target blood glucose concentration of ∼140 mg/dl (Fig. 7B), decreased hepatic glucose production (HGP) (Fig. 7C), and increased the glucose disposal rate (Rd) (Fig. 7D). The high dose of TIP47 ASO increased glucose uptake in adipose tissue and skeletal muscle (Fig. 7, E and F). Together, these results demonstrated that TIP47 ASO treatment improved both hepatic and peripheral insulin sensitivity.
Mammals normally store most of their energy in the form of triglycerides in perilipin-coated lipid droplets in adipocytes (2, 12). However, obesity results in ectopic accumulation of lipid droplets in nonadipose tissues (e.g., liver, skeletal, and cardiac muscle, and pancreatic islets), and this has been linked to insulin resistance and organ dysfunction (29). Fatty liver is characterized by excessive accumulation of triglyceride-rich lipid droplets coated with ADRP, TIP47, OXPAT, and other proteins (12, 16, 19, 22, 25, 30). In the current study, we found that HFD increased triglyceride, ADRP, and TIP47 levels in livers of C57BL/6J mice. Previous studies showed that fatty acids increased ADRP and triglyceride levels in hepatocytes (11, 13, 22). Conversely, a reduction in ADRP levels decreased triglyceride accumulation in hepatocytes (11, 13, 22). Ablation of the Adfp gene in mice reduced hepatic lipid droplets and triglyceride content in lean, diet-induced obese (DIO) and ob/ob mice, and increased VLDL secretion and insulin sensitivity in ob/ob mice (7, 8). We have also shown that reducing ADRP via ASO treatment attenuated hepatic steatosis and improved insulin sensitivity in DIO and ob/ob mice (16, 30). However, unlike genetic ADRP deficiency, ASO treatment reduced VLDL secretion in obese mice (16, 30).
We predicted that TIP47 would exert similar actions on hepatic lipids and glucose homeostasis as ADRP. We successfully reduced TIP47 mRNA and protein levels in HFD liver using ASO treatment. The low (15 mg/kg) and high (50 mg/kg) doses of TIP47 ASO both decreased hepatic TIP47 mRNA and protein without changing ADRP levels. TIP47 ASO did not significantly change TIP47 or ADRP expression in adipose tissue. TIP47 ASO decreased hepatic triglycerides in a dose-dependent manner. Similar to our previous results in ADRP, ASO-treated mice, TIP47 ASO suppressed Dgat2 and Gpat expression in the livers of mice on a HFD, suggesting inhibition of lipogenesis (9, 16, 30). The high dose of TIP47 ASO suppressed hepatic triglyceride secretion and improved glucose tolerance and insulin sensitivity in liver, adipose tissue, and muscle.
The ability of TIP47 ASO treatment to regulate hepatic lipids is consistent with studies in other tissues. In HeLa cells, TIP47 is recruited to lipid droplets and is actively involved in organizing and expanding lipid droplets (6). Conversely, siRNA knockdown of TIP47 prevents lipid droplet maturation and triglyceride accumulation in HeLa cells (6). In human macrophages, TIP47 is localized in the plasma membrane and aggregates into clusters surrounding lipid droplets in response to oleic acid loading (5). TIP47 maintains triglyceride levels when macrophages are depleted of ADRP; however, siRNA knockdown of TIP47 in ADRP-deficient macrophages reduces triglyceride accumulation (6). These results demonstrate that TIP47 can compensate for ADRP deficiency during conversion of macrophages into foam cells (6). A similar relationship exists between ADRP and TIP47 in murine adipocytes (26). Adfp-null adipocytes form lipid droplets to the same extent as wild-type adipocytes, but siRNA knockdown of TIP47 reduces triglyceride accumulation and lipid droplets in ADRP-deficient adipocytes (26). In our study, TIP47 ASO treatment did not change ADRP expression in liver and adipose tissue. In primary rat hepatocytes, TIP47 is increased rapidly by fatty acids followed by a slow decline, while ADRP is increased steadily during fatty acid exposure (13). In small-intestine enterocytes, TIP47 levels are increased rapidly in enterocytes in response to acute ingestion of HFD, while ADRP levels are increased in response to chronic HFD (18). It is possible TIP47 and ADRP have different roles in the early response to excess lipids and long-term lipid storage.
Hepatic steatosis is commonly associated with insulin resistance, impaired glucose tolerance, and type 2 diabetes (1, 4, 20). Insulin resistance in hepatic steatosis is mediated by lipid metabolites, e.g., ceramides, long-chain fatty acyl CoAs, and diacylglycerol (5, 14, 20). Diacyl glycerol activates PKCε in the liver, resulting in inhibition of insulin signaling (20, 21). Reduction of activities of enzymes involved in hepatic lipid synthesis and pathogenesis of steatosis improves insulin sensitivity (9, 21, 23). We have previously reported that ADRP ASO treatment improved glucose tolerance and insulin sensitivity in diet-induced obese and ob/ob mice (16, 30). In our current study, TIP47 ASO improved glucose tolerance in a dose-dependent manner. Moreover, the high dose of TIP47 ASO significantly enhanced insulin sensitivity in the liver, adipose tissue, and skeletal muscle. Total and HMW adiponectin levels were higher in TIP47 ASO-treated mice, suggesting a potential role in the enhancement of insulin sensitivity (32).
Our results demonstrate that TIP47 is connected to hepatic lipid and glucose metabolism in vivo independently of ADRP. We propose that TIP47 is stimulated by HFD, coats the surface of lipid droplets, prevents hydrolysis of triglycerides, and promotes hepatic steatosis (22). Further research is needed to determine how TIP47 and other lipid droplet proteins in the liver are temporally and spatially regulated by dietary lipids. It is also crucial to determine how TIP47 interacts with other lipid droplet proteins to regulate lipid pools and metabolites implicated in insulin signaling in liver, adipose tissue, and muscle.
Perspectives and Significance
Lipid droplets are intracellular organelles that store neutral lipids in various organs. Studies over the past two decades have shown that lipid droplets regulate the storage and hydrolysis of neutral lipids, including triacylglycerol and cholesterol esters. Alterations in lipid droplet metabolism influence the risk of developing NAFLD, diabetes, and other metabolic diseases. Lipid droplets in liver are coated with ADRP (Plin2), TIP47 (Plin3), and other members of the perilipin family. In the current study, we demonstrate that TIP47 is increased in fatty liver and that treatment with TIP47 ASO reduces hepatic steatosis, improves glucose tolerance, and enhances hepatic and peripheral insulin sensitivity. We did not observe a compensatory increase in ADRP in livers of TIP47 ASO-treated mice, suggesting that TIP47 regulates lipids and glucose independently of ADRP. Further understanding of the molecular interactions between TIP47 and pathways involved in lipid metabolism will provide new insights into the pathogenesis of NAFLD and metabolic syndrome, and potentially lead to novel therapies for lipid disorders and diabetes.
Research support was provided by National Institutes of Health Grant P01-DK-049210 (to R. S. Ahima), Diabetes Research Center Mouse Metabolic Phenotyping Core (Grant P30-DK-19525), and Center for Molecular Studies of Digestive and Liver Disease Morphology Core (Grant P30-DK-50306). R. M. Carr was supported by National Institutes of Health Training Grant T32-DK-007066.
Mark J. Graham and Rosanne M. Crooke are employees of Isis Pharmaceuticals, which provided antisense oligonucleotides for experiments. There are no conflicts of interest for the other authors.
- Copyright © 2012 the American Physiological Society