Transgenic overexpression of CTRP3 prevents alcohol-induced hepatic triglyceride accumulation

Abstract

This study tested the ability of a novel adipose tissue derived cytokine, C1q TNF-related protein-3 (CTRP3), to prevent alcohol-induced hepatic lipid accumulation, or alcoholic fatty liver disease (ALD). Previous work has demonstrated that CTRP3 is effective at preventing high-fat diet-induced fatty liver; however, the potential of CTRP3 to inhibit ALD has not been explored. To test the potential protective effects of CTRP3, transgenic mice overexpressing CTRP3 (Tg) or wild-type littermates (WT) were subjected to one of two different models of ALD. In the first model, known as the NIAAA model, mice were fed control or alcohol-containing liquid diets (5% vol/vol) for 10 days followed by a single gavage of ethanol (5 g/kg). In the second model, the chronic model, mice were fed control or alcohol-containing diets for 6 wk with no gavage. This study found that CTRP3 reduced triglyceride accumulation in the chronic model of alcohol consumption by ~50%, whereas no reduction was observed in the NIAAA model. Further analysis of isolated primary hepatocytes from WT and Tg mice demonstrated that CTRP3 increased oxygen consumption in the presence of fatty acids, indicating that CTRP3 increases hepatic fatty acid utilization. In conclusion, this study indicates that CTRP3 attenuates hepatic triglyceride accumulation in response to long-term chronic, but not short-term, alcohol consumption.

INTRODUCTION

Alcoholic fatty liver disease (ALD) is one of the leading causes of morbidity and mortality from hepatic complications in the United States and the world (11, 16, 19, 24, 25). Fatty liver, or hepatic steatosis, results from an abnormal retention of lipids (>5% total liver weight) and is primarily caused by either excessive alcohol consumption (ALD) or a high-fat diet, the latter of which is termed nonalcoholic fatty liver disease (NAFLD). Although the initial stages of both ALD and NAFLD are asymptomatic, the steatotic (fatty) liver is more susceptible to secondary insults (11), which result in hepatitis and, if left unchecked, liver cirrhosis. Liver cirrhosis is the 12th leading cause of death in the United States, with approximately one-half of all causes attributed to ALD (16). Currently, there are no approved pharmaceutical treatments to treat ALD (11, 24); the main treatment option for ALD is cessation of alcoholic consumption. While abstinence from alcohol is the long-term goal of management of all forms of ALD, the development of novel pharmaceutical treatments would enhance the ability of clinicians to control this disease.

Adipose tissue secretes many bioactive molecules that circulate in blood, collectively termed adipokines (24, 28, 32, 41, 42). A newly identified family of such secreted proteins has been designated as C1q/tumor necrosis factor (TNF)-related proteins (32, 33, 41, 42). C1q TNF-related protein-3 (CTRP3) is a unique member of the C1q TNF superfamily, and CTRP3 represents a potential novel treatment option for ALD. Previous work has demonstrated that transgenic overexpression of CTRP3 as well as daily injections of CTRP3 reduced high-fat diet-induced fatty liver (27), although the mechanisms responsible for this reduction are yet to be elucidated. Furthermore, in human patients with NAFLD there is a significant reduction in circulating CTRP3 levels (43). This indicates that restoring CTRP3 levels is a potential treatment strategy for NAFLD.

The primary cause of excessive hepatic lipid accumulation in NAFLD is excessive lipid intake. In ALD, however, the excessive hepatic lipid accumulation is directly linked to ethanol-induced activation of lipogenesis and suppression of lipid oxidation (7, 10, 11, 24). Therefore, the purpose of this project was to test the hypothesis that overexpression of CTRP3 would prevent alcohol-induced hepatic steatosis (i.e., ALD).

METHODS

Animals.

The CTRP3 transgenic overexpression (Tg) mouse strain was developed on the C57Bl/6J background with the carboxy-terminal FLAG epitope (DYKDDDDK)-tagged CTRP3 expression driven by the CMV early enhancer/chicken β-actin (CAG) promoter as described previously (26). Mice were housed in polycarbonate cages on a 12:12-h light-dark photocycle with ad libitum access to water and food, except as specified. Littermates without the CTRP3 transgene were used as wild-type (WT) controls. Throughout the experimental protocols, mice were inspected and weighed daily. Three WT mice from the chronic ethanol-fed group were removed from the study on the advice of the veterinary staff due to the presence of moribundity (loss of body weight and inability to remain standing). No mice from the other experimental groups exhibited signs of moribundity, and Table 1 reports the final animal numbers. Food intake and body weight were measured daily. At the time points indicated, animals were anesthetized with isoflurane and euthanized via exsanguination. Serum samples were prepared according to the manufacturer’s instructions (Sarstedt, cat. no. 41.1500.005). Tissue samples were excised, snap-frozen in liquid nitrogen (unless otherwise indicated), and stored at −80°C until further analysis. All animal procedures were conducted in accordance with institutional guidelines, and ethical approval was obtained from the University Committee on Animal Care (Animal Welfare Assurance no. A3203-01).

Table 1.

Numbers of animals in each experimental model

Model	WT-Con	WT-ETOH	Tg-ETOH
NIAAA model	5	7	9
Chronic Model	5	8	8

Tg, C1q tumor necrosis factor-related protein-3 (CTRP3) transgenic; WT, wild type; ETOH, ethanol fed; Con, fed matched control diet without ETOH; NIAAA, National Institute on Alcohol Abuse and Alcoholism.

Ethanol feeding.

Two independent ethanol-feeding models were employed: First, the 11-day chronic plus binge model, also known as the National Institute on Alcohol Abuse and Alcoholism (NIAAA) model, was used as described by Bertola et al. (5). Briefly, 12-wk-old male mice were acclimatized to a liquid diet without the addition of alcohol (Bio-serv; cat. no. F1259SP), for 4 days followed by 10 days on the Lieber-DeCarli ethanol diet (5% vol/vol ethanol; Bio-serv; cat. no. F1258SP) ad libitum. On the morning of the 11th day (1 h into light cycle), food was removed and replaced with water, and the mice were given a single gavage of ethanol (5 g/kg). Nine hours post-gavage, mice were euthanized, and tissue/serum samples were collected and processed for analysis. This model mimics the hepatic steatosis and liver injury and inflammation that occur in many alcoholic hepatitis patients (5).

In the second model (chronic model), 12-wk-old male mice were acclimatized to a liquid diet ad libitum without the addition of alcohol for 1- wk and then gradually transitioned from 1 to 5% Lieber-DeCarli ethanol diet (vol/vol ethanol) over the course of the next 2 wk and then maintained with 5% ethanol (vol/vol ethanol) for the remaining 4 wk. This feeding protocol is believed to reflect chronic ethanol abuse, beginning with low volumes and increasing over time (7), resulting in hepatic steatosis with only moderate inflammation (5). On the morning of the final day, food was removed, and mice were fasted for 9 h before euthanasia. Following euthanasia, tissue/serum samples were collected and processed for analysis. The 9-h time point was selected to be consistent with the NIAAA model protocol.

Additional WT littermates were placed on an ethanol-free isocaloric control diet (Bio-serv; cat. no. F1259SP) supplemented with maltose dextrin for use as experimental controls. Food intake in the ethanol-fed mice was measured daily, and mice on the control diet had their food intake limited to match the daily intake for the previous day of the corresponding ethanol-fed mice. Overview of the final numbers of animals in each experimental group is listed in Table 1.

Biochemical and hormonal variables.

Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and triglyceride concentrations were determined using commercially available assays according to the manufacturers’ directions (ALT, Fisher Diagnostics cat. no. TR71121; AST, Pointe Scientific, cat. no. A7561; triglycerides, Fisher Diagnostics, cat. no. TR22421). Ghrelin, insulin, resistin, plasminogen activator inhibitor 1 (PAI-1), gastric inhibitory polypeptide (GIP), leptin, adiponectin, interleukin-6 (IL-6), glucagon-like peptide-1 (GLP-1), and tumor necrosis factor-α (TNFα) were determined with commercially available kits using the Bio-Plex Multiplex Immunoassay System (Bio-Rad cat. nos. 171F7001M, 171G5023M, 171I50001, 171G5007M). The intra- and interassay coefficients of variation for all assays were ≤8% and ≤12%, respectively. Any value below the detectable assay working range was assigned the value equivalent of the lower limit of quantification.

Liver histology.

A section of the lateral left lobe of the liver was fixed in 10% formalin for 2 days. Following fixation, the tissue was dehydrated and embedded in paraffin using standard embedding techniques. Each sample was sectioned at 5 μm with a Microm HM325 microtome, mounted, and stained using standard hemotoxylin and eosin (H&E) staining procedures. The tissue was examined and photographed using a Zeiss Axioskop 40 microscope equipped with a Cannon Powershot A640 camera.

Hepatic lipid analysis.

Lipids were extracted as described by Bligh and Dyer and as previously performed (6, 26). Briefly, liver samples were weighed and homogenized in phosphate-buffered saline (10 ml/g tissue), followed by the addition of 1:2 (vol/vol) chloroform-methanol (3.75 ml/ml sample homogenate). Next, chloroform was added (1.25 ml/ml sample homogenate), followed by a final addition of distilled water (1.25 ml/ml sample homogenate). Samples were vortexed for 30 s between each step. Samples were then centrifuged (1,100 g for 10 min at room temperature) for phase separation (aqueous phase on top and organic phase below). The lower phase was collected with a glass pipette with gentle positive pressure (so as not to disturb the upper phase). Samples were then divided into two aliquots and dried under nitrogen gas at 60°C. To measure total triglyceride levels, one aliquot from each sample was dissolved in tert-butyl alcohol-Triton X-100 (3:2 vol/vol) solution. Triglycerides were quantified via colorimetric assay according to the manufacturer’s directions (Infinity Triglycerides, Fisher Diagnostics, cat. no. TR22421). The remaining aliquot was prepared as described for fatty acid methyl ester (FAME) analysis.

FAME preparation and analysis.

Fatty acids extracted from snap-frozen and lyophilized liver tissue samples were esterified and quantified by gas chromatography (GC) using a flame ionization detector (Shimadzu GC-2010; Shimadzu, Kyoto Japan). Briefly, boron trifluoride-methanol reagent (B1252; Sigma-Aldrich, St. Louis, MO) was added to a prepared aliquot of isolated liver lipids. The tube was then closed and heated in a water bath at 100°C for 1 h before being returned to room temperature. Distilled water (1.5 ml) was added, and the samples were then centrifuged for 1 min at 4,000 g. FAMEs were extracted in the hexane phase, dried under nitrogen gas, and suspended in 275 μl of hexane. Five microliters of C17 internal standard (1:9 hexane dilution) was added, and samples were analyzed by GC-FID equipped with a capillary column (Zebron ZB-WAX, 30 m length, 0.25 mm i.d., 0.25 μm film thickness; Phenomenex, Torrance, CA). The peaks were identified by comparison with a Supelco 37-component FAME mix fatty acid standard (Sigma-Aldrich cat. no. 47885-U).

Immunoblot analysis.

Liver samples were powdered in liquid nitrogen, and ~100 mg was suspended in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris·HCl, pH 8.0, 150 mM NaCl, 0.1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS) with protease and phosphatase inhibitors (Bimake, cat. nos. B14001 and B15001). The samples were incubated for 30 min at 4°C with gentle rotation and then centrifuged at 16,000 g for 10 min at 4°C. The supernatant containing the soluble fraction was collected and prepared for protein quantification. Protein concentrations were determined by commercial assay according to the manufacturer’s directions (Thermo, cat. no. 23238). Samples were diluted to equal protein concentrations with RIPA buffer and denatured at 95°C for 5 min after the addition of equal parts 2× SDS loading buffer (4% SDS, 10% 2-mercaptoethanol, 20% glycerol, 0.004% bromophenal blue, 0.125 M Tris·HCl, pH 6.8), followed by cooling for 5 min on ice. Denatured proteins were separated by SDS-polyacrylamide gel electrophoresis, according to the manufacturer’s directions (Bio-Rad, cat. no. 456-1046), with the addition of a protein ladder (Bio-Rad, cat. no. 161-0374). After gel electrophoresis, proteins were transferred to a nitrocellulose membrane according to standard procedures (Bio-Rad, cat. no. 1620115). Equal loading was confirmed by Ponceau S stain (0.1% Ponceau S, 5% acetic acid). Membranes were blocked with 4% nonfat milk solution for 1 h at room temperature and then probed with primary antibodies (Table 2) overnight at 4°C, followed by appropriate horseradish peroxidase (HRP)-conjugated secondary antibody (Table 2). Chemiluminescent signals were detected (chemiluminescent HRP substrate, Millipore, cat. no. WBKLS0100) and quantified using the Alphaview software and FluorChem M Western Imaging System (Proteinsimple). Band intensities were measured densitometrically and expressed as fold changes relative to those of normal control livers.

Table 2.

Antibodies

Protein	Source	RRID
p-Akt (Ser437)	Cell Signaling Technology cat. no. 4051	AB_331158
Akt	Cell Signaling Technology cat. no. 4685	AB_2225340
SREBP-1	Santa Cruz Biotechnology cat. no. sc-365514	AB_10841768
ACC	Cell Signaling Technology cat. no. 4190S	AB_10547752
p-ACC (Ser79)	Cell Signaling Technology cat. no. 11818	AB_2687505
p-ERK1/2 (Thr202/Tyr204)	Cell Signaling Technology cat. no. 4370	AB_2315112
p-AMPK (Thr172)	Cell Signaling Technology cat. no. 2535	AB_331250
AMPK	Cell Signaling Technology cat. no. 2793	AB_915794
ERK1/2	Cell Signaling Technology cat. no. 9102	AB_330744
DGAT2	Abcam cat. no. ab59493	AB_941282
Caspase-3	Cell Signaling Technology cat. no. 9662P	AB_10839261
Cleaved caspase-3 (Asp175)	Cell Signaling Technology cat. no. 9664	AB_2070042
PARP	Cell Signaling Technology cat. no. 9542	AB_2160739
β-Tubulin	Cell Signaling Technology cat. no. 86298	AB_2715541
Rabbit IgG, HRP	Thermo Fisher Scientific cat. no. 31460	AB_228341
Mouse IgG, HRP	Cell Signaling Technology cat. no. 7076	AB_330924

Akt, protein kinase B; p-, phosphorylated; SREBP-1, sterol regulatory element-binding protein-1; ACC, acetyl-CoA carboxylase-1; AMPK, AMP-activated protein kinase; ERK1/2, extracellular signal-regulated kinases; DGAT2, diglyceride acyltransferase-2; PARP, poly (ADP-ribose) polymerase; HRP, horseradish peroxidase-conjugated secondary antibodies; RRID, Research Resource Identifier (https://scicrunch.org/resources).

Lipid peroxidation assay.

Lipid peroxidation was determined in 10-mg sections of liver samples by quantifying the concentration of malondialdehyde (MDA) using a lipid peroxidation (MDA) assay kit according to the manufacturer’s directions (Sigma-Aldrich, cat. no. MAK085).

Primary hepatocytes.

Mouse primary hepatocytes were isolated and cultured by a modified two-step perfusion method as described by Zhang et al. (44). Briefly, 8-wk-old mice were anesthetized with isoflurane, the portal vein was cannulated with a 23-gauge needle and the inferior vena cava was cut. Livers were then perfused with Hank’s balanced salt solution (Corning, cat. no. 21-020-CM) supplemented with 20 mM HEPES at 9 ml/min speed for 6 min, followed by digestion with DMEM (Corning, cat. no. 15-017-CV) supplemented with 15 mM HEPES and 100 U/ml collagenase (Sigma, cat. no. C-5138) for another 6–8 min. The liver was then excised, and the gall bladder was removed. The cells in suspension were washed three times at 50 g for 2 min at 4°C and allowed to attach to collagen-coated XF24 cell culture microplate plates (Seahorse Bioscience, cat. no. 100777-004) for 1 h at 37°C in DMEM supplemented with 5 mM HEPES, 10 nM dexamethasone, and 10% fetal bovine serum. Medium was then replaced with serum-free DMEM after 3 h, and cells were maintained for 24 h before treatment.

Hepatocyte fatty acid oxidation.

To measure oxygen consumption, the hepatocyte cell culture medium was replaced with XF assay medium (Seahorse Bioscience, cat. no. 100965-000) supplemented with 0.5 mM sodium pyruvate and 5 mM glucose and placed into the XF24 Extracellular Flux Analyzer according to the manufacturer’s directions (XF24 Extracellular Flux Assay Kit; Seahorse Bioscience, cat. no. 100850-001; XF Calibrant Seahorse Bioscience, cat. no. 100840-000). Palmitate (200 μM) conjugated to fatty acid-free bovine serum albumin (BSA) or BSA alone (vehicle control) was added at the time-points indicated. For the addition of palmitate, a 6:1 palmitate-to-BSA ratio of 1 mM palmitate (Sigma, cat. no. P5585) conjugated to 0.17 mM fatty acid-free BSA (Sigma, cat. no. A8806) was prepared according to the manufacturer’s directions (Seahorse Bioscience). Briefly, 50 ml of a BSA solution (0.34 mM BSA, 150 mM NaCl, pH 7.4) was heated to 37°C, and 40 ml of a palmitate solution (2.5 mM palmitate, 150 mM NaCL) was heated to 70°C and added in 5-ml increments. The combined solution was incubated at 37°C for 1 h under constant agitation; afterward, the pH was adjusted to 7.4 and final volume adjusted to 100 ml with 150 mM NaCl). Aliquots were stored until use at −20°C in glass vials.

Statistical analysis.

Descriptive statistics (means and SDs) were calculated for all measured variables. Since the feeding models were not performed or analyzed concurrently, each model was analyzed as an independent experiment. Body weight, food intake, and hepatocyte oxygen consumption data were analyzed by two-way repeated-measures ANOVA followed by Tukey's multiple comparisons test. Because of limitations in immunoblotting (14 samples per gel, one lane was required for protein ladder), six samples from the WT-ETOH and six from the Tg-ETOH groups were analyzed. An unpaired t-test was used to compare immunoblot blot data between Tg-ETOH and WT-ETOH groups (the two WT-Con samples on each blot were used for visualization and normalization, but not for statistical comparisons). For all other assays, all samples were included and analyzed by one-way ANOVA followed by Tukey's multiple comparisons test. All statistical analyses were performed with GraphPad Prism 6.

RESULTS

Effects of ethanol with or without CTRP3 on body weight and food intake.

Body weight and food intake were measured daily; there were no differences between the groups (Fig. 1).

Fig. 1.

Transgenic overexpression of C1q tumor necrosis factor-related protein-3 (CTRP3) (Tg) did not affect body weight or food intake in either the NIAAA (National Institute on Alcohol Abuse and Alcoholism) or Chronic model. Body weight (A) and food intake (B) were measured daily. For the NIAAA model, data were reported as grams of diet consumed per mouse each day, and daily body weights were also reported as raw values and normalized to day 0 (C). Day 0 measurements (also indicated by the vertical dashed bar) were taken before mice started on the ethanol (ETOH)-containing diet. For the chronic model, daily food intake for each mouse was measured, and the average daily intakes per mouse were recorded over the course of the week (D). Additionally, daily body weights were also obtained and averaged for each mouse over the week time period and reported in absolute grams (E) or normalized to week 0 (F). All data are reported as means ± SD. Genotype and diet interactions were analyzed using a 2-way ANOVA for repeated measures followed by Tukey's multiple comparisons test. Numbers of animals analyzed in each group are reported in Table 1.

CTRP3 altered ethanol-induced effects on serum analytes.

To determine the impact of the ethanol feeding protocols on inflammation and metabolic status of the animal, a number of serum analytes were examined. The NIAAA model of ethanol feeding increased the circulating concentrations of PAI-1, TNFα, and ALT regardless of the presence of CTRP3, with no changes to IL-6, ghrelin, GIP, GLP-1, insulin, leptin, resistin, glucagon, or glucose levels (Fig. 2). However, serum triglycerides in the Tg-ETOH mice were elevated compared with WT-Con and WT-ETOH groups, with no differences with ethanol feeding alone in the WT mice (Fig. 2L). In the chronic ethanol feeding model, ALT levels were elevated in the WT-ETOH mice, with no differences in serum ALT levels between the WT-Con and Tg-ETOH mice (Fig. 3K). No other differences were observed among the measured serum analytes (Fig. 3). It is worth noting that TNFα was below the detection limit in almost all WT-Con samples and within the range of the assay for most ethanol-fed mice.

Fig. 2.

Effect of the NIAAA model of ethanol feeding on serum analytes. Concentrations of circulating analytes were measured by Bio-Plex Multiplex assay (A–I). Aspartate aminotransferase (AST; J), alanine aminotransferase (ALT; K), and triglycerides (L) were measured by individual assays. WT-con, wild-type littermate control-fed mice; WT-ETOH, WT littermates that were ethanol fed; Tg-ETOH, C1q tumor necrosis factor-related protein-3 (CTRP3) transgenic overexpressing mice that were ethanol fed; IL-6, interleukin-6; TNFα, tumor necrosis factor-α; GIP, gastric inhibitory polypeptide; GLP-1, glucagon-like peptide-1; PAI-1, plasminogen activator inhibitor 1. Data reported as mean ± SD *P < 0.05 vs. WT-Con. Numbers of animals analyzed in each group are reported in Table 1.

Fig. 3.

Effect of chronic ethanol feeding on serum analytes. Concentrations of circulating analytes were measured by Bio-Plex Multiplex assay (A–I). Aspartate aminotransferase (AST; J), alanine aminotransferase (ALT; K), and triglycerides (L) were measured by individual assays. WT-con, wild-type littermate control-fed mice; WT-ETOH, WT littermates that were ethanol fed; Tg-ETOH, C1q tumor necrosis factor-related protein-3 (CTRP3) transgenic overexpressing mice that were ethanol fed; IL-6, interleukin-6; TNFα, tumor necrosis factor-α; GIP, gastric inhibitory polypeptide; GLP-1, glucagon-like peptide-1; PAI-1, plasminogen activator inhibitor 1. Data reported as mean ± SD *P < 0.05 vs. WT-Con. Number of animals analyzed in each group are reported in Table 1.

CTRP3 attenuates ethanol-induced hepatic triglyceride accumulation.

Transgenic overexpression of CTRP3 reduced hepatic triglyceride accumulation in the chronic but not in the NIAAA model of ethanol consumption. Livers of untreated mice appeared histologically normal, without steatosis, ballooning, or inflammation. At the end of the NIAAA model of ethanol feeding, livers had limited evidence of steatosis that was similar between WT-ETOH and Tg-ETOH, with no difference in total hepatic triglyceride accumulation (Fig. 4A). However, in the chronic model of ethanol feeding, there was a significant attenuation of hepatic triglyceride accumulation in the Tg-ETOH mice compared with WT-ETOH (Fig. 4B). Additionally, histological examination of WT-ETOH livers revealed obvious steatosis and ballooning after 6 wk of ethanol feeding, whereas livers of the Tg-ETOH mice had no obvious morphological changes compared with control mice (Fig. 4, C–E).

Fig. 4.

Effect of transgenic overexpression of C1q tumor necrosis factor-related protein-3 (CTRP3) on hepatic triglycerides. WT-con, wild-type littermate control-fed mice; WT-ETOH, WT littermates that were ethanol fed; Tg-ETOH, CTRP3 transgenic overexpressing mice that were ethanol fed; Hepatic triglycerides in response to the NIAAA model (A) or chronic model (B) of ethanol feeding were quantified. Representative hematoxylin and eosin-stained images (×20) from liver sections from the chronic model are also shown (C–E). *P < 0.05 vs. WT-Con. Numbers of animals analyzed in each group are reported in Table 1.

CTRP3 attenuates ethanol-induced changes in hepatic fatty acid composition.

To further characterize changes in hepatic metabolism, fatty acid in the diet and in the livers were analyzed by GC-FID. The major fatty acid in the ethanol and control diets were similar, with the composition as follows: 13% palmitic (C16:0), 63% linoleic (C18:2^Δ9,12), 22% α-linolenic (C18:3^Δ9,12,15), and 2% other acids. Ethanol feeding significantly decreased the concentration of palmitic acid and increased both linoleic and α-linolenic acid in liver, resulting in an overall increase in the ratio of C18 to C16 fatty acid. Transgenic overexpression of CTRP3 attenuated the changes in fatty acid composition due to ethanol feeding in the chronic but not in the NIAAA model (Fig. 5).

Fig. 5.

Fatty acid methyl ester (FAME) analysis. Hepatic lipids were isolated, extracted, and converted to FAME and quantified with GC-FID. A and B: hepatic FAMEs in WT and Tg mice using the NIAAA (A) or chronic (B) ethanol feeding protocol. C and D: comparison of the ratio of total hepatic saturated to unsaturated fatty acids in WT and Tg mice using the NIAAA (C) or chronic (D) ethanol feeding protocol. WT-con, wild-type littermate control-fed mice; WT-ETOH, WT littermates that were ethanol fed; Tg-ETOH, C1q tumor necrosis factor-related protein-3 (CTRP3) transgenic overexpressing mice that were ethanol fed. Data are reported as means ± SD. *P < 0.05 vs. WT-Con, **P < 0.05 vs. WT-ETOH. Numbers of animals analyzed are reported in Table 1.

CTRP3 alters hepatic signaling in response to chronic ethanol feeding.

Since we observed significant changes to hepatic triglyceride accumulation between the WT-ETOH and Tg-ETOH groups on the chronic ethanol feeding protocol, we also evaluated the changes in expression and phosphorylation status of Akt, ACC, ERK1/2, and AMPK between the two groups (Fig. 6, A–D). We observed that transgenic overexpression of CTRP3 in Tg-ETOH had no effect on Akt and ACC expression and phosphorylation status. However, we observed a significant reduction in the phosphorylation status of both ERK1/2 and AMPK in the Tg-ETOH compared with the WT-ETOH mice.

Fig. 6.

Effects of transgenic CTRP3 overexpression on liver protein immunoblot analysis of liver samples from WT-Con (n = 2), WT-ETOH (n = 6), and Tg-ETOH (n = 6) mice exposed to the chronic (6-wk) ethanol-containing diet protocol. WT-con, wild-type littermate control-fed mice; WT-ETOH, WT littermates that were ethanol fed; Tg-ETOH, C1q tumor necrosis factor-related protein-3 (CTRP3) transgenic overexpressing mice that were ethanol fed; Akt, protein kinase B; SREBP-1, sterol regulatory element-binding protein-1; ACC, acetyl-CoA carboxylase-1; AMPK, AMP-activated protein kinase; ERK1/2, extracellular signal-regulated kinases; DGAT2, diglyceride acyltransferase-2. A–D: data are reported as optical density (OD) of phosphorylated/total protein normalized to WT-Con values. E and F: data are reported as OD of indicated protein/loading control (β-tubulin) and then normalized to WT-Con values. G: OD of β-tubulin for each group normalized to WT-Con. A–G: representative blots are shown. Bar graphs represent means ± SE of quantitative densitometry from 3 independent experiments. For each blot inset, the first 2 bands are WT-Con, the next 3 are WT-ETOH, and the last 3 are Tg-ETOH. H: representative Ponceau red staining of full blot indicates equal loading between wells. Immunoblot densitometry data analysis was performed using an unpaired Student’s t-test with * denoting significance (P < 0.05) between Tg-ETOH and WT-ETOH. Data are reported as means ± SD.

CTRP3 increases fatty acid oxidation.

Oxygen consumption rates were measured in primary hepatocytes isolated from WT and Tg mice with and without the addition of 200 μM palmitic acid. No difference in oxygen consumption was noted between hepatocytes from WT and Tg animals under control conditions (Fig. 7A). However, when palmitic acid was added to the medium (at time 0), Tg hepatocytes had a significant elevation of oxygen consumption (Fig. 7B). These data indicate that the hepatocytes from the Tg animals had increased fatty acid oxidation (9).

Fig. 7.

C1q tumor necrosis factor-related protein-3 (CTRP3) induces fatty acid oxidation in primary hepatocytes. Primary hepatocytes were isolated from wild-type (WT) and CTRP3 transgenic overexpressing (Tg) mice, and oxygen consumption rates were measured (n = 5; Seahorse Bioscience XF24 cellular flux analyzer). In the absence of lipids, CTRP3 had no effect on hepatocyte oxygen consumption rate (A), whereas when 200 μm of palmitic acid was added (time 0, first vertical line), CTRP3 dramatically increased oxygen consumption (B). A and B: second vertical line represent the addition of a mitochondrial uncoupler (FCCP) to visualize maximal oxygen consumption. All data are reported as means ± SE. *P < 0.05, significant between Tg and WT.

Markers of cell stress and damage.

A marker of oxidative stress, MDA, and two inducers of apoptosis, caspase-3 and PARP, were measured in the liver to determine whether the CTRP3-induced attenuation of hepatic steatosis was related to a reduction in oxidative stress or cellular apoptosis. Ethanol feeding caused a similar and significant elevation of MDA levels in both WT and Tg mice (Fig. 8A). Full-length inactive PARP and caspase-3 were detected in the liver samples (Fig. 8, B and C). No difference was observed in caspase-3 levels; however, there was a significant elevation in PARP in the WT-ETOH compared with the Tg-ETOH mice. The cleaved active forms of PARP (89 kDa) and caspase-3 (17 and 19 kDa) could not be detected in any samples examined (data not shown), indicating that the livers were not undergoing active apoptosis.

Fig. 8.

Oxidative stress and apoptosis. WT-Con, wild-type littermate control-fed mice; WT-ETOH, wild-type littermates that were ethanol fed; Tg-ETOH, C1q tumor necrosis factor-related protein-3 (CTRP3) transgenic overexpressing mice that were ethanol fed; MDA, malondialdehyde; PARP, poly (ADP-ribose) polymerase. A: lipid peroxidation as a marker of oxidative stress was determined by MDA levels in liver tissue from mice exposed to the chronic (6-wk) ethanol-containing diet protocol. B and C: immunoblot analysis of liver samples from WT-Con (n = 4), WT-ETOH (n = 5), and Tg-ETOH (n = 5) mice exposed to the chronic ethanol-containing diet protocol. For each blot inset, the first 2 bands are WT-Con, the next 3 are WT-ETOH, and the last 3 are Tg-ETOH. Data are reported as means ± SD. *P < 0.05 vs. WT-Con.

DISCUSSION

The primary findings of this study are that transgenic overexpression of CTRP3 reduced triglyceride accumulation in response to 6 wk of ethanol feeding but not the 10-day chronic plus binge protocol (the NIAAA model). Our previous work had demonstrated that CTRP3 suppressed diet-induced fatty liver (26); therefore, we hypothesized that CTRP3 would also attenuate alcohol-induced hepatic lipid accumulation. However, our initial results with the NIAAA model resulted in no significant protection from alcohol consumption in the livers of the Tg-ETOH mice. Moreover, in the NIAAA model, transgenic overexpression of CTRP3 appeared to exacerbate ethanol-induced changes to the liver fatty acids. These findings initially led us to conclude that CTRP3 did not protect the liver from alcohol-induced steatosis. However, as noted by Bertola et al. (5), there are different animal models of ALD that mimic different aspects of the disease. Triglyceride accumulation in the NIAAA model is secondary to acute hepatic inflammation (5, 14), and it appears that CTRP3 was not sufficient to ameliorate the initial alcohol-induced inflammation and associated triglyceride accumulation. However, we speculated that although CTRP3 might not prevent the initial hepatic steatosis, as induced by the NIAAA model, CTRP3 might still protect the liver from secondary injury and associated accumulation of hepatic lipids over time. In stark contrast to the findings of the NIAAA model, transgenic overexpression of CTRP3 completely abolished ethanol-induced hepatic steatosis in response to chronic ethanol feeding (Fig. 3B), including reversing the ethanol-induced changes to the hepatic fatty acid profile (Fig. 5C). These findings would indicate that high levels of CTRP3 protects the liver over time. Furthermore, the combined reduction in the proportion of the fatty acid palmitate (16:0) and the elevation of linoleic acid (18:2^Δ9,12) and α-linolenic acid 18:3^Δ9,12,15) indicated an overall reduction in fatty acid oxidation with ethanol consumption. Although it is possible that the lower proportion of saturated fatty acids were due to increase fatty acid elongation and desaturation, it appears unlikely, as ethanol consumption is usually associated with a decreased desaturase activity (22, 29, 38–40). Nevertheless, CTRP3 (Tg) restored the ratio of C18 to C16 fatty acids in the liver, indicating that CTRP3 attenuated the effects of ethanol consumption on either fatty acid oxidation or fatty acid elongation and desaturation.

There are two major differences between the NIAAA model and the chronic model, which could account for the observed differences in hepatic lipid accumulation. The first is that the NIAAA model has a final acute bolus of ethanol 9 h before tissue collection. As acute ethanol consumption significantly increases hepatic lipid accumulation (4), it may be that CTRP3 cannot block the effects of acute ethanol consumption. The second major difference is that there is a gradual transition from 1 to 5% ethanol (vol/vol ethanol) over the course of 2 wk in the chronic model, which could indicate that CTRP3 helps the liver develop a tolerance to alcohol consumption. Nevertheless, it is clear that under certain circumstances CTRP3 protects the liver from ALD.

To explore the mechanism behind the CTRP3-induced attenuation of hepatic steatosis, we examined the phosphorylation status or protein content, as a marker of activity, for many key enzymes responsible for the regulation of lipid metabolism. Lack of change in Akt or ACC phosphorylation/activation between the WT-ETOH and Tg-ETOH groups indicates that overexpression of CTRP3 did not activate these enzymes. Furthermore, changes to the concentrations of DGAT2 and SREBP1 were not observed, indicating that CTRP3 did not activate lipid-regulatory pathways. Surprisingly, a decrease in ERK1/2 and AMPK phosphorylation in the Tg-ETOH liver compared with the WT-ETOH liver was observed. Activation of ERK1/2 by excessive alcohol consumption has been previously observed (15). Moreover, treatments that block ethanol-induced activation of ERK1/2 are associated with reduced alcohol-induced hepatic damage (15). However, the mechanism by which decreased ERK1/2 activation could protect the liver is yet to be elucidated.

On the other hand, attenuation of ERK1/2 activation in Tg-ETOH livers could theoretically sensitize hepatocytes to oxidative stress-induced cell death/damage (8, 31). To test this oxidative stress, MDA and apoptosis (cleaved caspase-3 and PARP) were measured. As expected, ethanol feeding significantly increased MDA levels in the livers, with no difference between Tg-ETOH and WT-ETOH. Furthermore, the active forms of the apoptotic inducers caspase-3 and PARP were not observed in any of the livers examined, indicating that chronic ethanol feeding alone was not sufficient to induced hepatic apoptosis. It has been hypothesized that the accumulation of hepatic triglycerides is the first-hit of ALD, which although asymptomatic, sensitizes the liver to second-hits, such as inflammation or infection. These second-hits then lead to more severe, symptomatic forms of ALD (3, 17, 18, 23). The reduction in the concentration of full-length PARP in the Tg-ETOH livers indicates that these mice are potentially more resistant to a secondary hepatic insult (20). Conversely, the reduction in phosphorylated ERK1/2 in the Tg-ETOH livers may also increase the sensitivity of the liver to injury on exposure to additional stressors. These conflicting data should be examined further (i.e., lipopolysaccharide exposure in ethanol-fed mice ± CTRP3) to fully understand whether CTRP3 prevents not only hepatic steatosis but also the second hits of ALD.

The effects of ethanol consumption on AMPK phosphorylation/activation appears to be mixed, with in vivo studies demonstrating that ethanol decreases (1, 2, 35, 36), increases (34, 35), or has no effect (4) on the phosphorylation/activation of AMPK. It has been previously reported that 6 wk of ethanol consumption can increase AMPK phosphorylation (34), without a concurrent increase ACC phosphorylation, and our results independently confirm these previous findings. Increased phosphorylation of AMPK indicates increased AMPK activity (13), and under normal conditions, AMPK activation promotes fatty acid oxidation through inactivation of ACC (12). However, in our experiments WT-ETOH have significantly elevated AMPK phosphorylation compared with the Tg-ETOH, with no difference in the phosphorylation status of ACC. It has been suggested that the increase in AMPK in response to ethanol feeding, without concurrent elevations in ACC are indicative of the inability of AMPK to inhibit lipogenesis and increase β-oxidation; thus resulting in increased hepatic fatty acid accumulation (12, 34). Although the mechanism remains elusive, our data support this hypothesis as CTRP3 overexpression attenuated the rise in AMPK phosphorylation and resulted in lower total triglyceride accumulation. To examine this further, additional experiments would need to be pursued in which ACC phosphorylation/inactivation is examined with AMPK stimulation (pharmaceutical AMPK activator) or inhibition (i.e., postprandial state) in WT-ETOH and Tg-ETOH livers. On the basis of the data from this study, we speculate that ACC phosphorylation/inactivation is disassociated with AMPK activation status in the WT-ETOH mice, but normal in the livers of the Tg-ETOH mice. This would further support the hypothesis of Shearn et al. (34, 35), and identifies a potential mechanism of action for CTRP3.

To test whether CTRP3 overexpression stimulates fatty acid oxidation, we examined the effects of CTRP3 in isolated hepatocytes. In the absence of fatty acids, CTRP3 had no effect on hepatocyte oxygen consumption, indicating that CTRP3 does not increase basal metabolism. However, when 200 μm of palmitic acid was added to the medium, hepatocytes from CTRP3 transgenic mice had increased oxygen consumption compared with WT hepatocytes. Combined, these data indicate that CTRP3 stimulates lipid metabolism in the liver. These findings support our previous work, where we were unable to detect any difference between the livers of WT and CTRP3 transgenic mice fed a low-fat diet but observed significant changes in high-fat-fed mice (26).

Study limitations.

Due to budgetary limitations, we were able to study only male mice, so it is unclear whether female mice would demonstrate similar results, especially since female mice are more susceptible to the deleterious effects of alcohol consumption (21, 30, 37). In addition, all of our data are from a 9-h-fasted time point, which is sufficient to demonstrate baseline data but does not demonstrate the metabolic changes to the postprandial state of these animals. Future studies should examine the effects of CTRP3 on female mice and in both male and female mice in response to refeeding stimulus. Finally, these studies examined mice with chronic overexpression of circulating CTRP3 levels. It remains to be determined whether an infusion of recombinant CTRP3 can prevent or treat ALD.

Conclusion.

Elevated CTRP3 levels protect against long-term, but not short-term, alcohol-induced lipid accumulation. These findings indicate that CTRP3 may be beneficial in combination with cessation of alcohol consumption in treating ALD.

GRANTS

This work was supported by National Institute on Alcohol Abuse and Alcoholism grant R03AA023612 (to J. M. Peterson) and funding from East Tennessee State University (ETSU) Research Development Committee (to W. A. Clark and J. M. Peterson) and ETSU College of Public Health Department of Health Sciences (to G. Trogen, J. Bacon, Y. Li, G. L. Wright, A. Degroat, K. L. Hagood, Z.Warren, A. Forsman, and J. M. Peterson).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

G.T., J.B., Y.L., A.D., K.L.H., Z.W., A.F., W.A.C., and J.M.P. performed experiments; G.T., J.B., Y.L., G.L.W., A.K., W.A.C., and J.M.P. analyzed data; J.B., Y.L., G.L.W., K.L.H., A.F., A.K., W.A.C., and J.M.P. interpreted results of experiments; Y.L., A.F., and J.M.P. conceived and designed research; G.L.W. and J.M.P. drafted manuscript; A.F. and J.M.P. prepared figures; G.T., G.L.W., A.D., K.L.H., A.K., W.A.C., and J.M.P. edited and revised manuscript; G.T., J.B., Y.L., G.L.W., A.D., K.L.H., Z.W., A.F., A.K., W.A.C., and J.M.P. approved final version of manuscript.