Apolipoprotein CIII Overexpressing Mice Are Predisposed to Diet-Induced Hepatic Steatosis and Hepatic Insulin Resistance

Plasma Parameters

Blood samples were taken from the tail vein after an overnight fast. Plasma triglyceride and nonesterified fatty acid concentrations were determined using standard commercial kits according to the manufacturer's instructions (Diagnostic Chemicals and Wako Chemicals, respectively). Plasma insulin and adiponectin levels were measured by radioimmunoassays (Linco, Billerica, MA). Plasma glucose was measured with a glucose analyzer (Beckman Coulter, Brea, CA). Human APOC3 concentrations were measured with Cobas Mira plus (Roche Diagnostic, Indianapolis, IN). Mouse cytokine array (RayBio) was conducted as described in the manufacturer's description.

Liver Histology

Liver sections for histology were obtained after overnight fasting, fixed in 10% formalin, and stained with hematoxylin-eosin or Masson trichrome. Histology was read by two independent pathologists blinded to experimental design and treatment groups. Briefly, steatosis (0-3), lobular inflammation (0-3), ballooning (0-3), and fibrosis (0-4) were scored separately on a scale of 0-4 as described.^{14, 15} NAFLD activity score is expressed as the sum of each scoring.¹⁵

Hyperinsulinemic-Euglycemic Clamp Study

The mice were maintained on an HFD for ≍2 months. After an overnight fast, [3-³H]-glucose (high-performance liquid chromatography [HPLC] purified; PerkinElmer, Waltham, MA) was infused at a rate of 0.05 μCi/min for 2 hours to assess a rate of basal glucose turnover, followed by a 140-minute hyperinsulinemic-euglycemic clamp with a primed/continuous infusion of human insulin (31.5 mU/kg prime for 3 minutes, 4.5 mU/(kg-min) infusion; Novo Nordisk, Copenhagen, DK) as described.^{16, 17} Rates of basal and insulin-stimulated whole-body glucose fluxes and tissue glucose uptake were determined by bolus injection of 10 μCi of 2-deoxy-D-[1-¹⁴C] glucose (PerkinElmer).

Tissue Lipid Measurements

Tissue triglyceride was extracted using the method of Bligh and Dyer¹⁸ and measured using a DCL Triglyceride Reagent (Diagnostic Chemicals). Hepatic ceramide species and hepatic cytosolic DAG species were measured after overnight fasting using liquid chromatography and tandem mass spectrometry as described.¹⁹ Total DAG and ceramide are expressed as the sum of individual species.

Membrane Translocation of Protein Kinase C-ε (PKCε) and Phosphorylated AKT

PKCε membrane translocation in the liver protein extracts after overnight fasting and insulin phosphorylation of AKT^Ser473 were assessed both after overnight fasting and after the hyperinsulinemic euglycemic clamp study using the methods described.^{20, 21} Images were analyzed and quantified with ImageJ (NIH). Membrane translocation of PKCε was expressed as the ratio of membrane to cytosol bands.

Tissue Lipid Clearance and Uptake

Plasma lipid clearance and tissue uptake were assessed using mouse [³H]-labeled triolein and endogenously dual labeled chylomicrons (CM), including [³H]-retinyl ester and [¹⁴C]-triglyceride. Animals fed HFD for ≍3 months were cannulated (jugular vein) 7 days before the injection day. After collection of overnight fasting blood samples, a bolus injection of Liposyn (20%, 0.75 mL/kg, Abbott Laboratories, North Chicago, IL) conjugated with 10 μCi of [9,10-³H(N)]-triolein was administered over 1 minute through the jugular vein. Blood samples were collected at 2.5, 5, 7.5, 10, 15, 20, 30, and 60 minutes from tail vein. Endogenously dual-radiolabeled CM were prepared as described.²² After collection of blood at 0 minutes, 2 × 10⁵ dpm of [¹⁴C]-triglyceride-CM and 6 × 10⁵ dpm of [³H]-RE-CM were injected. Blood was collected at 2.5, 5, 10, and 15 minutes after injection. Tissue and plasma organic phase were extracted using the method of Bligh and Dyer¹⁸ and ³H/¹⁴C radioactivity was measured by beta-counter and the plasma and tissue triglyceride concentration was determined using a DCL Triglyceride Reagent (Diagnostic Chemicals). Tissue-specific triglyceride uptake rates [mg/(kg tissue-min)] were calculated by the following equation: Triglyceride uptake rate [mg/(kg tissue-min)] = plasma triglyceride (mg/ml) × {tissue radio-activity [dpm/(kg tissue-min)]/area under curve (AUC) of plasma radioactivity (dpm/ml)}.

Liver Triglyceride Production

Mice were maintained on RC or HFD for 3 months. After an overnight fast, plasma triglyceride levels were assayed prior to (0 hours) and 1, 2, 3, and 4 hours after poloxamer407 injection using an enzymatic method (Diagnostic Chemicals). The VLDL-triglyceride production rate was calculated by the increase in plasma triglyceride level from baseline to 4 hours after 1 g/kg of poloxamer407 intraperitoneal injection and the data were expressed as micromoles of triglyceride produced per hour per kg of body weight. Plasma apolipoprotein B (apoB) levels were determined by western blotting using rabbit polyclonal antibody (Abcam) in 4%-12% gradient gel (Invitrogen). The same membrane was stained with Coomassie-blue to show equal loading.

Intraperitoneal Glucose Tolerance Test (IPGTT)

Weight-matched mice were maintained on an HFD for 6 weeks. After an overnight fast mice were weighed and received a bolus intraperitoneal injection of glucose (1 g/kg). Blood samples were obtained from a tail vein at baseline (0) and 15, 30, 45, 60, 90, and 120 minutes after glucose challenge for determination of plasma glucose and insulin concentrations.

Reverse-Transcription Polymerase Chain Reaction (RT-PCR)

Tissue samples were obtained from mice maintained on RC or HFD for 3 months. The detailed methods, gene expression data, and sequences of PCR primers used are described in Supporting Table 2.

ApoC3Tg Mice Manifest Hepatic Insulin Resistance

In order to assess the tissue-specific contributions of liver, skeletal muscle, and adipose tissue to whole-body insulin resistance in the HFD-fed ApoC3Tg mice, we performed hyperinsulinemic-euglycemic clamp studies combined with ³H/¹⁴C-labeled glucose infusions.^{16, 17} Overnight fasting plasma glucose and insulin concentrations were similar between the two groups (Table 1) as were basal rates of endogenous glucose production (EGP) (Fig. 2E). During the hyperinsulinemic-euglycemic clamp the glucose infusion rates required to maintain euglycemia (Fig. 2A) in the ApoC3Tg mice were 34% lower than those of WT mice (Fig. 2B), indicating that the ApoC3Tg mice were more insulin resistant than the age-weight-matched WT mice. The ApoC3Tg mice displayed marked hepatic insulin resistance as reflected by the lack of suppression of EGP during the hyperinsulinemic-euglycemic clamp compared with the WT mice (Fig. 2E,F). Rates of insulin-stimulated whole-body glucose uptake were also slightly decreased by 13% in the ApoC3Tg mice compared with the WT mice (Fig. 3C), whereas muscle 2-deoxy-glucose uptake was not significantly different (Fig. 2D). Fasting plasma fatty acid concentrations decreased in a similar fashion during the hyperinsulinemic-euglycemic clamp study in both groups (% suppression: WT, 42% ± 7%; ApoC3Tg, 45% ± 6%), indicating that adipose tissue insulin responsiveness was similar between the two groups. Plasma insulin concentrations were similar between WT and ApoC3Tg mice during the clamp study (Supporting Table 1).

Open in a separate window

Fig. 2

Hepatic steatosis in ApoC3Tg mice is associated with hepatic insulin resistance. (A) Plasma glucose concentration and glucose infusion rate in WT (n = 6) and ApoC3Tg (n = 7) mice during hyperinsulinemic-euglycemic clamp study following 2 months of HFD. (B,C) Glucose infusion rate and whole-body glucose uptake rate during steady status. (D) Insulin stimulated muscle 2-deoxy-D-[1-¹⁴C] glucose (2DOG) uptake in gastrocnemius muscle. (E) Basal EGP and insulin suppressed (clamped) EGP and (F) percentage of suppression during the hyperinsulinemic-euglycemic clamp study. Data are expressed as mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 by Student's t test except (A) (2-way ANOVA with post-hoc analysis).

Open in a separate window

Fig. 3

Hepatic insulin resistance in the ApoC3Tg mice is associated with increased hepatic diacylglycerol content and PKCε activation. Animals fed an RC and HFD for 2 months and fasted overnight prior to experiments. (A) Intracellular cytosolic diacylglycerol after overnight fasting. (B) Membrane translocation of PKCε was analyzed in the liver of WT and ApoC3Tg mice after hyperinsulinemic-euglycemic clamp study. (C) Akt phosphorylation on Ser⁴⁷³ was analyzed in the liver of overnight-fasted and hyperinsulinemic-euglycemic clamped WT and ApoC3Tg mice after 2 months of HFD feeding. Data are expressed as mean ± SEM. (A), n = 6; (B), n = 4; (C), n = 6-7 per group. *P < 0.05, **P < 0.01 by Student's t test.

ApoC3Tg Mice Have Increased Hepatic DAG Content, PKCε Activation, and Hepatic Insulin Resistance

Increased hepatic triglyceride content and hepatic insulin resistance was associated with ≍45% increase in hepatic cytosolic DAG content in the liver from ApoC3Tg mice compared with the WT mice fed the HFD (Fig. 3A), whereas hepatic ceramide content was not different between genotypes (Supporting Fig. 3C,D). Furthermore, this increase in liver DAG content was associated with an ≍90% increase in the membrane/cytosol ratio of PKCε (Fig. 3B), reflecting an increase in PKCε activity, and a 32% reduction in insulin-stimulated AKT2 activity in the livers of ApoC3Tg mice compared with WT mice (Fig. 3C). Taken together, these data are consistent with the hypothesis that increased plasma levels of APOC3 predispose animals to NAFLD and hepatic insulin resistance, which in turn can be attributed to DAG-induced PKCε activation resulting in decreased insulin signaling.

ApoC3Tg Mice Display Increased Hepatic Triglyceride Uptake

To address the underlying mechanism for the increased hepatic triglyceride content in the ApoC3Tg mice we assessed hepatic triglyceride uptake by intravenous injection of dual-labeled chylomicrons containing [³H]-retinyl ester (RE) and [¹⁴C]-triglyceride into ApoC3Tg and age-weight-matched WT littermate mice after 3 months of HFD feeding. Plasma decay of tracers in WT mice was 0.55 ± 0.15 pools/min for triglyceride and 0.19 pools/min for RE. In ApoC3Tg mice, the plasma clearance of triglyceride was reduced to 0.27 pools/min, whereas that of RE was not significantly altered (0.17 pools/min). These data suggest that the major effect of the APOC3 transgene was on the initial clearance of triglyceride.

We then assessed the uptake of tracer into tissues with a focus on the ratio of the two labels. The uptake of both [³H]-RE-CM and [¹⁴C]-triglyceride-CM was increased by approximately twofold in the liver of ApoC3Tg mice, and the liver was the major site of TGRL uptake, compared with skeletal muscle and white adipose tissues (Fig. 4A,B). Although, in WT mice, liver triglyceride/RE uptake was 1.00 ± 0.05, indicative of greater fatty acid uptake due to lipolysis, this ratio was reduced to 0.74 ± 0.11 in the ApoC3Tg mice. Triglyceride/RE uptake was 19.4 ± 9.67 in skeletal muscle from WT mice, as would be expected due to greater fatty acid uptake after lipolysis. This ratio decreased to 6.1 ± 2.12 in the ApoC3Tg mice. APOC3 has been reported to alter TGRL metabolism either by inhibiting lipoprotein lipase (LpL) activity or interfering with receptor-mediated hepatic uptake of remnant lipoproteins.²³ Our data suggest a block in peripheral triglyceride lipolysis consistent with LpL inhibition.

Open in a separate window

Fig. 4

Amount of hepatic triglyceride uptake was increased in ApoC3Tg mice both on RC and HFD. Animals fed an HFD for 3 months and fasted overnight prior to experiments (A,B). After collection of blood at 0 minutes, then endogenously dual-radiolabeled chylomicron, 6 × 10⁵ dpm of [³H]-RE-CM and 2 × 10⁵ dpm of [¹⁴C]-triglyceride-CM were injected. Blood was collected at 2.5, 5, 10, and 15 minutes after injection (A,B). With a separate batch of animals fed RC or HFD for 3 months, after 6 hours fasting a bolus injection of 9 μCi of [³H]-triolein was intravenously administered and blood samples were collected at 2.5, 5, 7.5, 10, 15, 20, 30, and 60 minutes from tail vein (C). Tissue-specific uptake rate of (A) [³H]-RE-CM (B) [¹⁴C]-triglyceride-CM were assessed in various tissues and (C) hepatic [³H]-triolein uptake rate were assessed after RC and HFD. N = 4-5. *P < 0.05; **P < 0.01; ns, not significant by Student's t test. CM, chylomicron. RE, retinyl ester. WAT, epididymal white adipose tissue.

Because ApoC3Tg mice developed hepatic steatosis when fed an HFD diet but not when fed a regular chow diet, we also compared hepatic triglyceride uptake in ApoC3Tg and age-weight-matched WT littermates fed the RC and HFD. Surprisingly, we found that hepatic triglyceride uptake was increased in ApoC3Tg mice fed either the RC or HFD compared with WT mice, and that there was no further increase in hepatic triglyceride uptake in the HFD-fed ApoC3Tg mice compared with the RC-fed ApoC3Tg mice (Fig. 4C).

Hepatic VLDL-Triglyceride Secretion Was Decreased in ApoC3Tg Mice During HFD

Next, we assessed VLDL-triglyceride secretion in WT and ApoC3Tg mice fed either the RC or HFD by injection of the lipoprotein lipase inhibitor poloxamer407. APOC3 has been shown to stimulate VLDL synthesis and secretion in cultured cells⁷ and elevated circulating APOC3 levels correlate with VLDL production⁸ and postprandial hyperlipidemia in humans.⁹ Consistent with these findings, we observed an ≍70% increase in VLDL-triglyceride production in ApoC3Tg mice fed a RC diet (Fig. 5A,B), which was associated with marked increases in circulating apoB48 concentrations in ApoC3Tg mice (Fig. 5C). In contrast, we observed a 53% reduction in VLDL-triglyceride production rate in HFD-fed ApoC3Tg mice compared with the HFD-fed WT mice (Fig. 5B). The decreased VLDL-triglyceride production was accompanied by a marked decrease in circulating apoB100 level as well as a slight decrease in apoB48 concentrations in ApoC3Tg mice during HFD feeding (Fig. 5C). Liver microsomal triglyceride transfer protein (MTP) expression (Supporting Fig. 4) and messenger RNA (mRNA) expression of MTP (Supporting Table 2) were not significantly different between WT and ApoC3Tg mice on HFD. Thus, the greater uptake of TGRLs in the ApoC3Tg mice was not compensated by increased liver secretion of apoB-containing lipoproteins.

Open in a separate window

Fig. 5

Hepatic VLDL-triglyceride secretion was decreased in ApoC3Tg mice during HFD. Animals fed an RC and HFD for 3 months and fasted overnight prior to experiments. (A) Net increase of plasma triglyceride concentration after poloxamer407 (p-407) injection from basal (0 minutes, before p-407 injection) (n = 4-6). (B) Triglyceride production rate during 4 hours after p-407 injection, expressed as micromoles of triglyceride produced per hour per kg of body weight (n = 4-6). (C) Plasma apoB determined by western blotting using 4%-12% gradient gel. The same membrane stained with Coomassie-blue was provided as loading control (n = 3, equal amount of 4 mice plasma were pooled into each sample). Data are expressed as mean ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001 by two-way ANOVA post-hoc analysis (A) and **P < 0.01 by two-tailed t test (B,C). ns, not significant.

Other possible contributors to greater hepatic triglyceride content are de novo synthesis and reduced fatty acid oxidation. Interestingly, mRNA expression of hepatic sterol regulatory element binding protein-1c (SREBP-1c) and LpL were significantly higher in ApoC3Tg mice compared with WT on HFD (Supporting Table 2), indicating a possible contribution of increased hepatic lipogenesis to the increased hepatic lipid accumulation in ApoC3Tg mice during HFD feeding. There were no differences in whole-body oxygen consumption, respiratory quotients (Table 1), and liver mRNA expression of peroxisome proliferator-activated receptor-α (PPAR-α) and carnitine palmitoyltransferase 1 (CPT1), between WT and ApoC3Tg mice both on RC and HFD (Supporting Table 2), suggesting that alterations in hepatic fatty acid oxidation were not likely responsible for the observed net increases in hepatic triglyceride content observed in the HFD-fed ApoC3Tg mice.

We thank to Irena Ignatova-Todorova, Xiaoxian Ma, Mario Kahn, Christopher M. Carmean for expert technical assistance withthe studies and Dr. Jan Breslow (Rockefeller University, New York, NY) for providing the ApoC3Tg mice.