Antisense oligonucleotide reduction of DGAT2 expression improves hepatic steatosis and hyperlipidemia in obese mice

Authors


  • Potential conflict of interest: Nothing to report.

Abstract

In this study, we investigated the role of acyl-coenzyme A:diacylglycerol acyltransferase 2 (DGAT2) in glucose and lipid metabolism in obese mice by reducing its expression in liver and fat with an optimized antisense oligonucleotide (ASO). High-fat diet-induced obese (DIO) C57BL/6J mice and ob/ob mice were treated with DGAT2 ASO, control ASO, or saline. DGAT2 ASO treatment reduced DGAT2 messenger RNA (mRNA) levels by more than 75% in both liver and fat but did not change DGAT1 mRNA levels in either of these tissues, which resulted in decreased DGAT activity in liver but not in fat. DGAT2 ASO treatment did not cause significant changes in body weight, adiposity, metabolic rate, insulin sensitivity, or skin microstructure. However, DGAT2 ASO treatment caused a marked reduction in hepatic triglyceride content and improved hepatic steatosis in both models, which was consistent with a dramatic decrease in triglyceride synthesis and an increase in fatty acid oxidation observed in primary mouse hepatocytes treated with DGAT2 ASO. In addition, the treatment lowered hepatic triglyceride secretion rate and plasma triglyceride levels, and improved plasma lipoprotein profile in DIO mice. The positive effects of the DGAT2 ASO were accompanied by a reduction in the mRNA levels of several hepatic lipogenic genes, including SCD1, FAS, ACC1, ACC2, ATP-citrate lyase, glycerol kinase, and HMG-CoA reductase. In conclusion, reduction of DGAT2 expression in obese animals can reduce hepatic lipogenesis and hepatic steatosis as well as attenuate hyperlipidemia, thereby leading to an improvement in metabolic syndrome. (HEPATOLOGY 2005;42:362–371.)

Fatty liver disease has become an epidemic in Western countries, with an estimated prevalence of over 14% in the entire population.1 It is caused by an excess accumulation of triglycerides in the liver and is often associated with obesity due to a long-term positive energy balance (energy input > energy output) with extra energy stored as triglycerides in various tissues.

One key enzyme involved in the synthesis of triglycerides is acyl-coenzyme A:diacylglycerol acyltransferase (DGAT), which catalyzes the last step in mammalian triglyceride synthesis via the covalent binding of the acyl moiety with diacylglycerol. DGAT is found in most tissues of the body and has a high expression in fat, liver, and small intestine.2, 3 Two DGATs that are encoded by two different gene families have been identified: DGAT1 and DGAT2.4, 5 Studies indicate that both DGAT1 and DGAT2 play important roles in triglyceride synthesis. DGAT1 knockout mice were found to be resistant to high-fat diet-induced obesity because of an increased metabolic rate, which was at least partially attributable to increased physical activity.2 DGAT1 null mice showed decreased levels of tissue triglycerides and increased insulin and leptin sensitivity.6 Furthermore, these mice generally survived well with a residual amount of triglycerides in adipose tissue and normal triglyceride levels in plasma. In contrast, DGAT2 null mice are lipopenic and die soon after birth as a result of profound reductions in substrates for energy metabolism and impaired skin permeability.7 These studies indicate that DGAT2 may play a more important role than DGAT1 in metabolism during the early stages of development. However, the function of DGAT2 at the adult stage or the effect of pharmacological reduction of this enzyme in key metabolic tissues (e.g., liver and fat) is unknown.

To define the precise role of DGAT2 in glucose and lipid metabolism, we employed antisense technology to inhibit the expression of DGAT2 in two different obese mouse models. Pharmacokinetic studies have shown that antisense oligonucleotides (ASOs) are active in a variety of tissues in vivo, including liver and adipose tissue, but are not active in muscle.8 Therefore, antisense technology was employed to specifically reduce DGAT2 expression in liver and fat tissues, and its effects were evaluated. The present study demonstrates that DGAT2 ASO treatment improves hepatic steatosis and blood lipid levels independent of adiposity in obese mice. Data indicate that reduced hepatic steatosis and improved blood lipid levels are related to decreased hepatic lipid synthesis. These results support the conclusion that DGAT2 plays an important role in hepatic lipid metabolism and that therapeutic interventions aimed at reducing DGAT2 function in lipogenic tissues may provide clinical benefit for hepatic steatosis and cardiovascular diseases.

Abbreviations

DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; ASO, antisense oligonucleotide; DIO mice, diet-induced obese mice; mRNA, messenger RNA; TG, triglyceride; BW, body weight; GTT, glucose tolerance test; ITT, insulin tolerance test; FFA, free fatty acid.

Materials and Methods

Selection of DGAT2 ASO.

Screens for identifying a DGAT2 messenger RNA (mRNA)-selective ASO inhibitor were performed with 20-base chimeric ASOs where the first five bases and last five bases have a 2′-O-(2-methoxy)-ethyl (2′-MOE) modification. Initial in vitro screens identified several potent and specific ASOs, all of which targeted a binding site within the coding region of the DGAT2 mRNA. Subsequent characterization in primary mouse hepatocytes identified ISIS-217376 to be the most potent and specific oligonucleotide for reducing DGAT2 mRNA levels. A control ASO, ISIS-141923, which is chemically identical to ISIS-217376 but not complementary to any known gene sequences, was used as a negative control for the studies.

Isolation and Transfection of Primary Mouse Hepatocytes.

Primary hepatocytes were isolated from C57BL/6J mice as previously described.9 The cells with more than 85% viability were seeded onto 60-mm culture plates or 25-cm2 flasks at 106 cells per plate or flask in culture medium (Williams' Medium E with 10% fetal bovine serum and 10 nm insulin plus penicillin and streptomycin) and cultured overnight. For transfection, the hepatocytes were incubated for 6 hours with 1 mL transfection mixture, which contained 150 nmol/L control ASO or DGAT2 ASO and 4.5 μg lipofectin (Invitrogen Life Technology Inc., Carlsbad, CA) in Williams' Medium E. The mixture was then aspirated and replaced with the culture medium for 24 hours to allow the cells to recover.

Determination of Triglyceride Synthesis and Fatty Acid Oxidation in Transfected Mouse Hepatocytes In Vitro.

TG synthesis in transfected mouse hepatocytes was determined by measuring the incorporation of [3H]glycerol into TG in the presence or absence of oleate as previously described.7 Fatty acid oxidation was determined by measuring the oxidation of [14C]oleate into acid-soluble products and CO2 as previously described.10, 11

Animal Care and Treatments.

The following investigations were conducted in accordance with the Institutional Animal Care and Use Committee guidelines. Male C57BL/6J mice (6 weeks of age) were purchased from Jackson Laboratories (Bar Harbor, ME) and housed 4 animals per cage at 22°C to 25°C with a 12-hour light/dark cycle. After 5 days of acclimation to rodent chow (Harlan-Teklad 8604, Madison, WI), the mice were then switched to a high-fat diet containing 60% fat calories (Research Diet D12492; Research Diets, New Brunswick, NJ). Eight weeks after being fed the high-fat diet, body weight (BW) and tail snip baseline glucose and insulin levels were measured. The mice (diet-induced obese mice, or DIO mice) were then randomly divided into different treatment groups based on BW. They were treated with DGAT2 ASO or the negative control ASO at a dose of 25 mg/kg BW twice a week or the same volume of saline for 5 or 7 weeks, or at 37.5 mg/kg BW twice a week or the same volume of saline for 2 weeks. All ASOs were dissolved in 0.9% saline and administered subcutaneously. For 5- and 7-week treatment studies, a group of lean C57BL/6J mice (n = 8) fed rodent chow and treated with saline served as normal controls. At the end of the treatment period, the animals were sacrificed. Blood samples were collected by cardiac puncture, and tissues were dissected and then frozen immediately in liquid N2 for further analysis. A small piece of liver, fat, and skin was fixed in 10% formalin for histological examination.

A study was also conducted with C57BL6J-Lepob/Lepob (ob/ob) mice treated with the DGAT2 ASO. Six-week-old ob/ob mice were purchased from Jackson Laboratories. After 7 days of acclimation to rodent chow (Labdiets 5015; Purina, St. Louis, MO), the animals were randomized to two treatment groups based on BW and plasma glucose levels and were treated with DGAT2 ASO at 25 mg/kg BW twice a week or saline for 4 weeks. Because the studies in DIO mice demonstrated that the negative control ASO had no effect on any parameters measured in the mice, it was not included in the experiment in ob/ob mice.

Glucose and Insulin Tolerance Tests.

Glucose tolerance test (GTT) and insulin tolerance test (ITT) were performed in DIO mice after overnight and 3-hour food withdrawal, respectively. A baseline (0 minute) tail blood glucose measurement was conducted followed by intraperitoneal administration of glucose for the GTT (1.0 g/kg BW) or intraperitoneal injection of insulin (0.5 U/kg BW) (R-Insulin; Lilly Research Laboratories, Indianapolis, IN) for the ITT. Tail blood glucose was then measured at 15- or 30-minute intervals for 2 hours after glucose or insulin challenge using a glucometer (Abbott Laboratories, Bedford, MA).

Metabolic Rate Measurement.

Metabolic rate in mice was measured using indirect calorimetry (Oxymax System; Columbus Instruments, Columbus, OH). Animals were acclimated to metabolic chambers for 2 to 3 days before initiation of the measurement. For each treatment group, 4 to 6 animals were measured for a 24-hour period.

Biochemical Analysis.

Plasma glucose, insulin, leptin, and free fatty acid (FFA) levels were measured with a YSI Biochemistry Analyzer (YSI Inc., Yellow Spring, OH), an insulin ELISA kit (ALPCO Diagnostics, Windham, NH), a leptin ELISA kit (Crystal Chem, Downers Grove, IL), and Wako's NEFA C reagents (Wako Chemicals GmbH, Neuss, Germany), respectively. Plasma TG and total cholesterol were measured with a chemistry analyzer Hitachi 717 (Roche Diagnostics, Indianapolis, IN). Liver TG and glycogen contents were analyzed as previously described.12 Tissue DGAT activity was assayed by measuring the incorporation of [14C]acyl-coenzyme A into TG at 20 mmol/L MgCl2 as described by Cases et al.5

Plasma Lipoprotein Analysis and Hepatic TG Secretion Measurement.

Plasma lipoprotein and cholesterol profiling in pooled samples of each group was performed with a Beckman System Gold 126 HPLC system, 507e refrigerated antosampler, 126-photodiode array detector (Beckman Instruments, Fullerton, CA) and a Superose 6 HR 10/30 column (Pfizer, Chicago, IL) as described by Crooke et al.13 Hepatic TG secretion was measured in fasted DIO mice as previously described.14

Hepatic Lipid Analysis.

Lipids were extracted from liver homogenates in hexane/isopropanol (3:2) and separated via thin-layer chromatography (TLC) on Silica Gel G-60 TLC plates with the solvent system mexane:ether:MeOH:CH3COOH (90:20:2:3) for neutral lipids, or CHCl3:CH3OH:H2O:NH4OH (56:47:10:2.5) for phospholipids. Lipids were visualized through exposure to iodine vapor, and individual bands were quantified by ImageQuant software (Molecular Dynamics).

Histological Analysis.

Liver, epididymal fat, and skin were fixed in 10% buffered formalin and embedded in paraffin wax. Multiple adjacent 4-μm sections were cut and mounted on glass slides. After dehydration, the sections were stained with hematoxylin-eosin. Images of the histological sections were analyzed.

Real-Time Quantitative Reverse-Transcriptase Polymerase Chain Reaction (RT-PCR) Analysis.

Total RNA extracted from harvested hepatocytes or tissues was used for real-time quantitative RT-PCR analysis, which was performed with a custom-made enzyme and reagent kit (Invitrogen Life Technology Inc.) and ABI Prism 7700 Sequence Detector (PE Applied Biosciences, Foster City, CA). Primers and probes for analysis of the expression of different genes (Table 1) were designed using Primer Express Software (PE Applied Biosciences).

Table 1. Sequences for Primers and Probes for Quantitative RT-PCR Analysis on the Expression of Different Mouse Genes
GenesForward Primer (5′ to 3′)Probe (5′ to 3′)Reverse Primer (5′ to 3′)
  1. NOTE. Primers and probes were designed using Primer Express Software (PE Applied Biosciences, Foster City, CA).

  2. Abbreviations: ACC, acetyl-coenzyme A carboxylase; ATP, adenosine triphosphate; DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; FAS, fatty acid synthase; G-6-P; glucose-6-phosphatase; GLUT, glucose transporter; GP, glycogen phosphorylase; HMG-CoA, 3-hydroxy-3-methylglutaryl-coenzyme; HSL, hormone-sensitive lipase; LPL, lipoprotein lipase; PEPCK, phosphoenolpyruvate carboxykinase; PPAR-γ, peroxisomal proliferator-activated receptor γ; SCD, stearoyl-coenzyme A desaturase; SREBP, sterol regulatory element-binding protein.

ACC1ATT GGG CAC CCC AGA GCT ACCC CAC TGA GCG GAA GGA GCT GCCC GCT CCT TCA ACT TGC T
ACC2GGG CTC CCT GGA TGA CAA CCTC TGA TGA GGA CCC TAG TGC CGG CTTC CGG GAG GAG TTC TGG A
ATP-citrate lyaseGAT GAA GTG GCA CCT GCA AAGCCA GCT ATG CCC CAA GGA AAG AGT GCGGT ATG TCG GCT GAA GAG GGT
DGAT1GTT CCG CCT CTG GGC ATTCAG CCA TGA TGG CTC AGG TCC CAC TGAA TCG GCC CAC AAT CCA
DGAT2ACT CTG GAG GTT GGC ACC ATCAG CGT TGC TCT GGC GCAGGG TGT GGC TCA GGA GGA T
FASCAT GAC CTC GTG ATG AAC GTG TCCG TCA CTT CCA GTT AGA GCA GGA CAA GCCGG GTG AGG ACG TTT ACA AAG
Glycerol kinaseCAC GTA TGG AAC AGG GTG CTTTAT TGT GCA ACA CGG GCC ATA AGT GTG TCAG GAG GCC ATG TTC AGA AAA
G-6-PTGC AAG GGA GAA CTC AGC AATCG TTC CCA TTC CGC TTC GCCGGA CCA AGG AAG CCA CAA TG
GLUT2ACG GAT GCC AAT TAC CGA CACCC ATC CTC GTG GCG CTG ATGTGC TGG GCC ATG TGC A
GLUT4CAT CCC ACA AGG CAC CCT CCTA CGC TCT GGG CTC TCT CCG TGGCAT GCC ACC CAC AGA GAA GA
GPGCA GAT GAC TGG CTC AGG CCCC TTG GGA GAA GGC TCG CCCTAG AAA TGC ACG GGC AGC A
HMG-CoA reductaseCCT GAC ACT GAA CTG AAG CGCTA ACAT GCA ATC TGT GAA TTA AAA ATC TCA ATGTCT TTC CAG AAC ACA GCA CG
HSLGAG CGC TGG AGG AGT GTT TTTGC TTG ACA GCC CAG CAG TAG GCACAG TTG AAC CAA GCA GGT CAC A
LPLTGG AGA AGC CAT CCG TGT GTGC AGA GAG AGG ACT CGG AGA CGT GGTCA TGC GAG CAC TTC ACC AG
PEPCKTTG AAC TGA CAG ACT CGC CCTTGT GGT GGC CAG CAT GCG GAGAT ATG CCC ATC CGA GTC ATG
PPAR-γAGT GGA GAC CGC CCA GGTTG CTG AAC GTG AAG CCC ATC GAGGCA GCA GGT TGT CTT GGA TGT
SCD1AGA TCT CCA GTT CTT ACA CGA CCA CTCC GGA AAT GAA CGA GAG AAG GTG AAG AGAC GGA TGT CTT CTT CCA GGT G
SREBP1CAG CTC AGA GCC GTG GTG AAAG CGC ACA GCC CAC AAT GCCTTG ATA GAA GAC CGG TAG CGC

Statistical Analysis.

Values presented represent the mean ± SEM of three in vitro or at least five in vivo independent measures per treatment. Statistical difference across treatment groups was determined using the Student t test or one-way ANOVA and Tukey HSD multiple comparisons. A P value of less than .05 was considered significant.

Results

Reduction of DGAT2 Gene Expression Lowers TG Synthesis and Increases Fatty Acid Oxidation in Mouse Hepatocytes.

To investigate whether reduction of DGAT2 expression affects TG synthesis and fatty acid oxidation in liver, hepatocytes isolated from C57BL/6J mice were cultured and transfected with DGAT2 ASO or control ASO, and the rate of incorporation of [3H]glycerol into TG or the rate of [14C]oleate oxidation was determined. As compared with control ASO, DGAT2 ASO treatment decreased DGAT2 mRNA levels by more than 75%, but did not change DGAT1 mRNA levels (Fig. 1A). The decrease in DGAT2 mRNA levels was associated with more than 55% inhibition of TG synthesis in the absence of oleate and more than 75% inhibition in the presence of oleate (Fig. 1B), demonstrating an important role of this enzyme in hepatic TG synthesis. In addition, DGAT2 ASO treatment caused an increase in fatty acid oxidation by more than 30% (Fig. 1C).

Figure 1.

Inhibition of DGAT2 expression reduced triglyceride synthesis and increased fatty acid oxidation in primary mouse hepatocytes. Isolated mouse hepatocytes were seeded in 60-mm plates or 25-cm2 flasks. The cells were then transfected with an optimized mouse DGAT2 ASO or a control ASO for 6 hours. After recovery, incorporation of [3H]glycerol into triglycerides or oxidation of [14C]oleate into CO2 and acid-soluble products was determined. (A) DGAT2 ASO reduced DGAT2 expression but did not change DGAT1 expression. (B) Reduction of DGAT2 expression by DGAT2 ASO caused decreased triglyceride synthesis in the absence and presence of oleate, respectively. (C) Reduction of DGAT2 expression by DGAT2 ASO was associated with an increase in fatty acid oxidation. Data are expressed as the mean ± SEM (n = 3). **P < .01 versus controls. ASO, antisense oligonucleotide; DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; mRNA, messenger RNA.

Reduction of DGAT2 Gene Expression Does Not Change Adiposity, Insulin Sensitivity, or Skin Structure in DIO Mice.

As compared with saline controls, DGAT2 ASO treatment decreased DGAT2 mRNA levels by more than 75% in both liver and fat tissues but did not affect mRNA levels in the small intestine (Fig. 2A). DGAT2 ASO treatment did not change DGAT1 mRNA levels in any of these tissues (data not shown). Enzymatic assay indicated that the decreased DGAT2 mRNA levels were associated with reduced total DGAT activity in the liver but not in fat (Fig. 2B), indicating that DGAT1 was able to compensate for DGAT2 reduction in fat but not in liver. DGAT2 ASO treatment did not affect BW or fat pad weights compared with control groups (Table 2), nor did it affect food intake, O2 consumption rate, or respiratory quotient (data not shown). The treatment also did not change plasma leptin levels (41.8 ± 12.4 vs. 45.4 ± 11.0 ng/mL in control ASO-treated mice after 7 weeks of treatment), suggesting unchanged leptin resistance.

Figure 2.

DGAT2 antisense reduced DGAT2 expression and DGAT activity in DIO mice. (A) DGAT2 mRNA levels were dramatically reduced in the liver, white adipose tissue, and brown adipose tissue but were not changed in the small intestine from DIO mice treated with DGAT2 ASO for 7 weeks. (B) Reduced DGAT2 mRNA levels were associated with decreased DGAT activities in the liver but not in WAT. Total RNA was prepared from the liver, epididymal white fat, interscapular brown fat, and duodenum and used for real-time quantitative reverse-transcriptase polymerase chain reaction analysis to evaluate the expression of DGAT2. DGAT activity was determined by measuring the incorporation of [14C]acyl-coenzyme A into triglycerides. Data are expressed as the mean ± SEM (n ≥ 6). **P < .01. mRNA, messenger RNA; ASO, antisense oligonucleotide; DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; WAT white adipose tissue; BAT, brown adipose tissue; TG, triglyceride; CoA, coenzyme A.

Table 2. DGAT2 ASO Treatment Did Not Change Body Weight or Adiposity in Either DIO Mice or ob/ob Mice
AnimalBW/Fat Weight (g)SalineControl ASODGAT2 ASO
  1. NOTE. Data are expressed as the mean ± SEM.

DIO miceBW, week 036.6 ± 0.636.0 ± 1.437.2 ± 1.0
 BW, week 538.7 ± 1.038.9 ± 1.837.4 ± 0.9
 Epididymal fat weight2.39 ± 0.112.46 ± 0.212.32 ± 0.11
 Peri-renal fat weight0.94 ± 0.090.90 ± 0.080.88 ± 0.06
ob/ob miceBW, week 043.9 ± 0.843.6 ± 1.1
 BW, week 452.4 ± 0.951.5 ± 0.9
 Epididymal fat weight4.34 ± 0.164.44 ± 0.22

Plasma glucose and insulin concentrations were measured during and at the end of the studies. DGAT2 ASO treatment did not change fed glucose levels (262.7 ± 5.0 vs. 252.4 ± 10.9 mg/dL in saline group after 7 weeks of treatment). However, treatment lowered insulin levels in both fed and fasted states (Fig. 3A-B). To evaluate whether the DGAT2 ASO treatment improved insulin sensitivity, both GTT and ITT were performed in DIO mice. Neither insulin sensitivity (Fig. 3C) nor glucose tolerance (Fig. 3D) was improved with DGAT2 ASO treatment.

Figure 3.

DGAT2 antisense treatment lowered plasma insulin levels in DIO mice but did not improve insulin sensitivity. (A) Fed-state plasma insulin concentration (n = 8) in lean mice treated with saline or DIO mice treated with saline, control ASO, or DGAT2 ASO at a dose of 25 mg/kg twice a week for 7 weeks. (B) Four-hour-fasted plasma insulin concentration after 5 weeks of treatment. (C) ITT and (D) GTT curves in DIO mice after 4 and 5 weeks of treatment, respectively. All data are expressed as the mean ± SEM (n = 6-8). *P < .05. ASO, antisense oligonucleotide; DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; ITT, insulin tolerance test; GTT, glucose tolerance test.

To investigate whether DGAT2 ASO treatment caused any changes in skin structure as reported for DGAT2-deficient mice,7 histological analysis was performed on skin from DGAT2 ASO treated DIO mice or controls. Microstructure examination with hematoxylin-eosin staining found no change in the skin from the DGAT2 ASO-treated group compared with the saline-treated group (Fig. 4A).

Figure 4.

DGAT2 antisense treatment did not change skin microstructure but markedly improved hepatic steatosis and reduced liver triglyceride content in DIO mice. (A) Representative section of skin from DIO mice treated with saline or DGAT2 ASO (n = 5). (B) Representative section of liver from lean mice treated with saline, and from DIO mice treated with saline, control ASO, or DGAT2 ASO (n = 4). All skin and liver sections were stained with hematoxylin-eosin and photographed at an original magnification of ×20. (C) Liver triglyceride and glycogen contents in lean and DIO mice (n = 8). Data are expressed as the mean ± SEM. **P < .01. DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; ASO, antisense oligonucleotide.

Hepatic Steatosis Is Improved in DIO Mice Treated With DGAT2 ASO.

Hepatic steatosis was markedly improved after DGAT2 ASO treatment as determined by histological examination (Fig. 4B). Biochemical analysis of TG content confirmed histological observations. Compared with saline controls, TG decreased by 62% (27.0 ± 1.9 vs. 70.8 ± 6.6 mg/g liver; P < .01) (Fig. 4C) and 56% (27.5 ± 5.3 vs. 62.1 ± 9.0 mg/g liver; P < .01) in DIO mice after 5 and 7 weeks of treatment, respectively. The treatment did not change hepatic glycogen content (Fig. 4C) or muscle TG content (7.71 ± 0.69 vs. 7.45 ± 0.89 mg/g tissue in saline group). TLC analysis also showed decreased TG content in DGAT2 ASO-treated mice compared with control ASO-treated mice (P < .01) (Fig. 5A). Furthermore, it demonstrated that both FFA and diacylglycerol levels decreased in the DGAT2 ASO-treated group (P < .01) (Fig. 5A). However, the levels of the major hepatic phospholipids, including phosphatidic acid, phosphatidylcholine, phosphatidylethanolamine, and phosphatidylinositol, were not changed by DGAT2 ASO treatment (Fig. 5B).

Figure 5.

DGAT2 antisense treatment lowered hepatic levels of neutral metabolic lipids (A) but did not change the levels of structural lipids (phospholiplids). Lipids were extracted from liver homogenates, separated via thin-layer chromatography with the solvent system (A) mexane:ether:MeOH:CH3COOH (90:20:2:3) for neutral lipids or (B) CHCl3:CH3OH:H2O:NH4OH (56:47:10:2.5) for phospholipids, and individual lipid bands were visualized through exposure to iodine vapor and then quantified. Data are expressed as the mean ± SEM (n = 6). **P < .01. ASO, antisense oligonucleotide; DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; TG, triglyceride; CE, cholestarylester; FFA, free fatty acid; DAG, diacylglycerol; PA, phosphatidic acid; PE, phosphatidylethanolamine; PI, phosphatidylinositol; PC, phosphatidylcholine.

Blood Lipid Levels Are Also Improved in DIO Mice Treated With DGAT2 ASO.

DGAT2 ASO treatment also decreased the blood lipid levels in DIO mice. As compared with saline-treated lean mice, saline-treated DIO mice displayed significantly elevated levels of total plasma cholesterol (P < .01). DGAT2 ASO treatment lowered the levels in DIO mice (P < .05) (Fig. 6A). The treatment also lowered plasma TG and FFA levels (P < .05) (Fig. 6A). To confirm these findings, lipoprotein analysis was performed using high-performance liquid chromatography. The results from this analysis correlated well with the data on total plasma cholesterol (Fig. 6B). Furthermore, the decrease in total cholesterol levels in the DGAT2 ASO-treated group was attributable to the decrease in both low-density lipoprotein and high-density lipoprotein cholesterol levels (Fig. 6B). To investigate whether the decreased plasma TG levels in the DGAT2 ASO-treated group were associated with reduced hepatic TG secretion, the hepatic TG secretion rate was measured in DIO mice treated with ASO at 37.5 mg/kg twice a week for 2 weeks. A 20% to 25% reduction in secretion rate was found in the ASO-treated group versus controls (P < .01) (Fig. 6C). In addition, decreased hepatic triglyceride content and total plasma cholesterol levels were also observed in these short-term DGAT2 ASO-treated mice (data not shown).

Figure 6.

DGAT2 antisense treatment improved blood lipid levels and lipoprotein profile in DIO mice (n = 8). (A) DIO mice treated with saline showed a significantly higher plasma total cholesterol level than lean mice and treatment with DGAT2 ASO at a dose of 25 mg/kg twice a week for 7 weeks significantly lowered the level. The treatment also significantly lowered plasma triglyceride and FFA levels. (B) High-performance liquid chromatography analysis of lipoprotein in pooled samples of each group. DGAT2 ASO–treated group showed decreased low-density lipoprotein and high-density lipoprotein cholesterol levels. (C) Hepatic triglyceride secretion rate was decreased in DIO mice treated with DGAT2 ASO at 37.5 mg/kg BW twice a week for 2 weeks (n = 5). Data are expressed as the mean ± SEM. *P < .05. **P < .01. ASO, antisense oligonucleotide; DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; FFA, free fatty acid; VLDL, very low-density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein.

Downregulation of Lipogenic Gene Expression in Liver From DIO Mice Treated With DGAT2 ASO.

Because DGAT2 ASO-treated DIO mice showed improved hepatic steatosis and decreased hepatic neutral lipids and blood lipid levels, the expression of genes involved in lipid and glucose metabolism were analyzed. In addition to reducing hepatic DGAT2 expression, DGAT2 ASO treatment resulted in a dramatic reduction in the expression of genes involved in TG synthesis (e.g., glycerol kinase), de novo fatty acid synthesis (ATP-citrate lyase, ACC1, ACC2, and FAS), fatty acid desaturation (SCD1), and cholesterol synthesis (HMG-CoA reductase) in the liver (Table 3). These observations are consistent with the decrease in liver TG, diacylglycerol, and FFA contents, and the improvement in blood lipid levels observed in DGAT2 ASO-treated DIO mice.

Table 3. Downregulation of Hepatic Lipogenic Gene Expression in DIO Mice Treated With DGAT2 ASO for 7 Weeks
GenesSalineControl ASODGAT2 ASO
  • NOTE. The analysis was performed using TaqMan quantitative RT-PCR assay (ABI 7700 system). Total RNA was prepared from tissues; 100 ng of the RNA was used for the assay as described in Materials and Methods. For each gene, 6 to 8 samples from each treatment group were analyzed in duplicate, and the average value was used. Data are expressed as the mean ± SEM.

  • Abbreviations: ATP, adenosine triphosphate; ACC, acetyl-coenzyme A carboxylase; FAS, fatty acid synthase; SCD, stearoyl-coenzyme A desaturase; HMG-CoA, 3-hydroxy-3-methylglutaryl-coenzyme; LPL, lipoprotein lipase; SREBP1, sterol regulatory element-binding protein 1; PPAR-γ, peroxisomal proliferator-activated receptor γ; G-6-P, glucose-6-phosphatase; PEPCK, phosphoenolpyruvate carboxykinase; GLUT 2, glucose transporter 2; GP, glycogen phosphorylase.

  • *

    P < .05, †P < .01 when compared with saline-treated group.

  • P < .01 when compared with control ASO-treated group.

ATP-citrate lyase100 ± 9.587.4 ± 4.253.4 ± 6.7†,
ACC1100 ± 8.882.4 ± 8.733.5 ± 1.1†,
ACC2100 ± 12.395.0 ± 23.010.4 ± 1.0†,
FAS100 ± 9.680.9 ± 5.927.2 ± 2.4†,
SCD1100 ± 11.883.0 ± 15.63.8 ± 0.5†,
Glycerol kinase100 ± 3.387.1 ± 5.4*62.9 ± 3.0†,
HMG-CoA reductase100 ± 15.380.5 ± 1.855.3 ± 3.6†,
LPL100 ± 10.1128.3 ± 11.0115.5 ± 11.6
SREBP1100 ± 8.978.4 ± 6.2*56.6 ± 5.2†,
PPAR-γ100 ± 13.179.9 ± 5.064.9 ± 9.2†
G-6-P100 ± 29.273.4 ± 8.091.1 ± 17.7
PEPCK100 ± 13.5113.9 ± 15.6123.3 ± 20.8
GLUT2100 ± 3.393.8 ± 5.0108.4 ± 6.7
GP100 ± 3.797.7 ± 4.381.1 ± 2.7

DGAT2 ASO treatment was also associated with a decreased expression of sterol regulatory element-binding protein 1 (SREBP1), a transcription factor involved in the regulation of lipid metabolism (Table 3). No significant difference in the expression of genes related to gluconeogenesis (G-6-Pase and PEPCK) was observed in the liver (Table 3). In adipose tissue, no changes were seen in the expression of genes associated with glucose uptake (e.g., glucose transporter 4: 108.4% ± 18.1% vs. 100% ± 27.0% in the saline-treated group) and lipid homeostasis (e.g., lipoprotein lipase: 75.4% ± 15.3% vs. 100% ± 23.2% in the saline-treated group; hormone-sensitive lipase: 141.8% ± 29.1% vs. 100% ± 17.5% in the saline-treated group).

Hepatic Steatosis Is Also Markedly Improved in ob/ob Mice Treated With DGAT2 ASO.

To investigate whether reduction of DGAT2 expression could improve hepatic steatosis in a leptin-deficient obese mouse model, ob/ob mice were treated with the DGAT2 ASO for 4 weeks. Similar to findings in DIO mice, DGAT2 ASO treatment lowered DGAT2 mRNA levels by more than 80% but did not change DGAT1 mRNA in either liver or fat compared with controls (Fig. 7A-B). DGAT2 ASO treatment did not affect BW, fat depot weight (Table 2), food intake, metabolic rate, or insulin sensitivity (data not shown). However, DGAT2 ASO treatment markedly improved hepatic steatosis (Fig. 7C) and reduced liver TG content (Fig. 7D). DGAT2 ASO treatment also lowered plasma FFA (1.08 ± 0.09 vs. 1.38 ± 0.10 mEq/L in controls; P < .05) and insulin levels (32.0 ± 9.7 vs. 58.5 ± 5.2 ng/mL in controls; P < .01). Quantitative RT-PCR analysis on some representative genes related to hepatic lipogenesis and gluconeogenesis found that DGAT2 ASO treatment lowered the expression of ACC2 and FAS but not G-6-Pase (Fig. 7E).

Figure 7.

DGAT2 antisense treatment inhibited DGAT2 expression and improved hepatic steatosis in ob/ob mice (n = 5). Mice were treated with saline or DGAT2 ASO at a dose of 25 mg/kg twice a week for 4 weeks. (A) DGAT2 ASO reduced DGAT2 expression by more than 80% in the liver and white adipose tissue. (B) DGAT2 ASO did not change DGAT1 expression in the liver or white adipose tissue. (C) Representative section of liver from ob/ob mice treated with saline or DGAT2 ASO. Hematoxylin-eosin–stained liver sections were photographed at an original magnification of ×20. (D) Liver triglyceride content was significantly reduced in the mice treated with DGAT2 ASO. (E) DGAT2 ASO treatment caused decreased gene expression of hepatic ACC2 and FAS but not G-6-P. Data are expressed as the mean ± SEM. **P < .01. DGAT, acyl-coenzyme A:diacylglycerol acyltransferase; ASO, antisense oligonucleotide; mRNA, messenger RNA; WAT, white adipose tissue.

Discussion

The key finding of this study is that hepatic reduction of DGAT2 expression in obese mice can significantly improve hepatic steatosis and attenuate hyperlipidemia, which may be due to inhibition of hepatic lipogenesis and activation of hepatic fatty acid oxidation. Because DGAT2-null mice die shortly after birth due to severe reduction in substrate availability,7 the precise physiological role of this enzyme during the postdevelopmental stage is undetermined. Our data indicate that reduction of DGAT2 in lipogenic tissues in adult obese mice abrogates many abnormalities of the metabolic syndrome and is not associated with any side effects or mortality. It was reported that because DGAT2-deficient mice lack energy substrates, they likely consume their available glycogen stores rapidly, thus hypoglycemia and decreased liver glycogen levels ensue.7 However, no such problems are encountered after pharmacological inhibition of this target and liver glycogen levels were unchanged in DGAT2 ASO-treated mice. Finally, we did not find any abnormalities in skin microstructure, appearance or any dehydration, which was a key finding observed in DGAT2-deficient mice.

The findings in this study are also in contrast to those observed in DGAT1-deficient mice, in which increased insulin and leptin sensitivity as well as increased metabolic rate were observed when the mice were fed a high-fat diet. However, DGAT1 deficiency in ob/ob mice was not associated with an improvement in insulin sensitivity, which was ascribed to a compensatory increase in DGAT2 expression in the absence of leptin.6 In contrast, the ASO effects on DGAT2 reduction are leptin-independent and are not associated with any compensatory upregulation of DGAT1 mRNA levels, improved insulin sensitivity, or increased whole body metabolic rate. The ASO-caused reduction of DGAT2 expression resulted in decreased total DGAT (DGAT1 + DGAT2) activity in liver but not in fat. These data, coupled with improved steatosis but unchanged adiposity and insulin sensitivity in DGAT2 ASO-treated mice, suggested that DGAT2 plays a more important role in lipid metabolism in liver than in fat, and/or DGAT1 is only able to compensate for DGAT2 reduction in fat but not in liver through posttranscriptional regulation. These data extend previous findings and further indicate that the two DGAT enzymes play fundamentally distinct roles in glucose and lipid metabolism in different tissues, and that lipid homeostasis in adipose tissue is critical for insulin sensitivity (see more discussion below). Thus, although both these enzymes catalyze the same final step in TG synthesis, there is marked diversity in the specific function of the enzymes in the body.

Overfeeding results in the storage of excess energy as lipid in the body, which is a result of increased lipid synthesis in the liver where the dietary carbohydrate is converted into long-chain fatty acids by hepatic de novo fatty acid synthesis. The synthesized fatty acids are further used for TG synthesis catalyzed by a series of enzymes, including DGAT for the final step. Long-term overfeeding, especially with a high-fat diet, can cause increased blood FFA and insulin concentrations.15–18 Consequently, the liver accelerates uptake of FFA, synthesis of TG and cholesterol, and assembly and output of very low-density lipoprotein,19–21 which leads to hepatic steatosis, hyperlipidemia, and increased cardiovascular risk. Inhibition of DGAT2 in vitro reduced TG synthesis, indicating a direct regulatory role of this enzyme. However, in addition to this direct effect, reduced hepatic TG synthesis and secretion in vivo could be partly attributable to an attenuation of hyperinsulinemia and a consequent decrease in de novo hepatic lipogenesis after DGAT2 ASO treatment. For example, insulin stimulates hepatic lipogenesis through stimulating SREBP1 expression,22–24 which in turn stimulates expression of additional lipogenic genes such as FAS, ACCs, and SCD1.25–27 Thus, a reduction in insulin levels could contribute to decreased lipogenesis. This may also help explain the difference between our study and DGAT2 knockout mice in which the levels of these genes were unchanged. In addition, a substantial decrease in the amount of hepatic FFA and diacylglycerol, two substrates for TG synthesis, were also observed in DGAT2-treated mice. Furthermore, although we did not measure fatty acid oxidation in vivo, data from hepatocytes indicate that DGAT2 ASO leads to an increase in fatty acid oxidation, which could be due to inhibition of ACCs as seen both in vitro and in vivo, thereby causing decreased malonyl-CoA concentrations and increased CPT I activities, after DGAT2 ASO treatment. Thus, the marked reduction of hepatic steatosis and improved blood lipid levels observed after DGAT2 reduction are likely due to decreased hepatic lipid synthesis and increased fatty acid oxidation.

An interesting observation that surfaced was that although DGAT2 reduction caused a lowering of circulating insulin levels, there was no improvement in insulin sensitivity. A decrease in circulating insulin levels in the presence of similar plasma glucose concentrations is often considered to be indicative of an improvement in insulin sensitivity. However, decreased insulin sensitivity in overfed animals is usually caused by increased adiposity and atopic lipid accumulation, especially in muscle, which is accompanied by compensatory hyperinsulinemia due to insulin oversecretion by the pancreas.28, 29 In addition, changes in circulating FFAs have been shown to modulate insulin secretion, and prolonged elevation in FFAs can lead to hyperinsulinemia.30 In DGAT1-deficient DIO mice, both improved insulin sensitivity and decreased body fat content were found.6 In contrast, in DGAT1-deficient ob/ob mice, unimproved insulin sensitivity was found to be associated with unchanged body fat content,6 suggesting that a reduction in adiposity may underlie the positive effects of DGAT1 on insulin sensitivity. Thus, lack of an improvement in insulin sensitivity in DGAT2 ASO-treated mice could be due to the unchanged adiposity, leptin resistance, and muscle triglyceride content, whereas the decreased plasma insulin levels may be a result of decreased plasma FFA levels.

Such a reduction in FFA, however, would also be expected to cause an improvement in insulin sensitivity. Furthermore, a reduction in hepatic steatosis would be expected to increase hepatic insulin sensitivity and consequently cause an improvement in glycemic control. If the latter is the case, current systemic GTT and ITT may not be sensitive enough to detect this local (hepatic) improvement in insulin sensitivity. The precise reasons underlying this disconnect remain undetermined; however, similar observations have been made previously. For example, liver-specific peroxisomal proliferator-activated receptor γ disruption lead to an improvement in steatosis and hepatic insulin action but caused a worsening of hyperglycemia, which was found to be due to an increase in fat and muscle insulin resistance.31, 32 Furthermore, rosiglitazone aggravated the fatty liver yet improved glucose tolerance in these mice,33, 34 suggesting that adipose tissue was the likely mediator of the glucose-lowering effects of rosiglitazone in this model. Because ASOs do not accumulate in muscle,8 and because we did not find changes in the expression of key lipid and glucose metabolism genes, or total DGAT (DGAT1 + DGAT2) activity in adipose tissue and body adiposity, or changes in TG content in muscle, it is likely that the effects of the DGAT2 ASO in these mice are primarily limited to the liver. Further studies looking at hepatic metabolic fluxes under basal and clamp conditions will help us to better understand the mechanisms underlying this surprising observation.

In conclusion, long-term overfeeding results in obesity, which is associated with liver steatosis, nonalcoholic steatohepatitis, hyperlipidemia, and other cardiovascular problems due to increased lipogenesis. The data from the present study demonstrate that inhibition of DGAT2 can significantly improve hepatic steatosis and cardiovascular risk profile by inhibiting hepatic lipogenic pathways. Therefore, DGAT2 could be a promising therapeutic target for the treatment of these obesity-related disorders. Future studies will help elucidate the effect of long-term DGAT2 inhibition on obesity and the mechanisms underlying its effects on different metabolic pathways.

Acknowledgements

We thank Xiaokun Xiao, Cathie York-DeFalco, and Gene Hung for assistance in histological analysis; Kannan Subramaniam and Kristina Lemonidis for high-performance liquid chromatography analysis; Sara Petok for technical assistance; and Kim Alexis for assistance in manuscript preparation.

Ancillary

Advertisement