Beneficial effects of an endogenous enrichment in n3-PUFAs on Wnt signaling are associated with attenuation of alcohol-mediated liver disease in mice

Abstract

Alcohol-associated liver disease (ALD) is a major human health issue for which there are limited treatment options. Experimental evidence suggests that nutrition plays an important role in ALD pathogenesis, and specific dietary fatty acids, for example, n6 or n3-PUFAs, may exacerbate or attenuate ALD, respectively. The purpose of the current study was to determine whether the beneficial effects of n3-PUFA enrichment in ALD were mediated, in part, by improvement in Wnt signaling. Wild-type (WT) and fat-1 transgenic mice (that endogenously convert n6-PUFAs to n3) were fed ethanol (EtOH) for 6 weeks followed by a single LPS challenge. fat-1 mice had less severe liver damage than WT littermates as evidenced by reduced plasma alanine aminotransferase, hepatic steatosis, liver tissue neutrophil infiltration, and pro-inflammatory cytokine expression. WT mice had a greater downregulation of Axin2, a key gene in the Wnt pathway, than fat-1 mice in response to EtOH and LPS. Further, there were significant differences between WT and fat-1 EtOH+LPS-challenged mice in the expression of five additional genes linked to the Wnt signaling pathway, including Apc, Fosl1/Fra-1, Mapk8/Jnk-1, Porcn, and Nkd1. Compared to WT, primary hepatocytes isolated from fat-1 mice exhibited more effective Wnt signaling and were more resistant to EtOH-, palmitic acid-, or TNFα-induced cell death. Further, we demonstrated that the n3-PUFA-derived lipid mediators, resolvins D1 and E1, can regulate hepatocyte expression of several Wnt-related genes that were differentially expressed between WT and fat-1 mice. These data demonstrate a novel mechanism by which n3-PUFAs can ameliorate ALD.

1 |. INTRODUCTION

Long-term consumption of as little as 12 g alcohol/day¹ increases the risk of alcohol-associated liver disease (ALD), with the highest risk at levels exceeding 40 g/day.² Most chronic consumers of alcohol will develop hepatic steatosis, and a subset of these individuals will proceed to more severe manifestations of the disease which include hepatic inflammation, fibrosis, and cirrhosis. In these more advanced cases, the disease can be complicated by infection and organ failure, which are significant contributors to morbidity, mortality, and the expenditure of health-care dollars.^3,4 The mechanisms and mediators of ALD progression and severity are not well understood, and most importantly, there are no FDA-approved treatments. Therefore, the identification of the specific physiological and cellular mechanisms by which alcohol damages the liver is of paramount importance for the development of rational, targeted treatment regimens. Underlying cofactors that contribute to ALD include nutritional status, gender, age, ethnicity, genetic polymorphisms, and, increasingly more prevalent in Western societies, obesity.⁵ Data from preclinical animal studies from our group and others strongly suggest that diet, specifically dietary fat, plays an important interactive role with alcohol consumption in ALD pathogenesis.^6–8 We have previously shown that a diet high in linoleic acid (LA, an n6 polyunsaturated fatty acid [n6-PUFA]) exacerbated ALD in mouse models in part through the generation of oxidized LA metabolites.⁶ Importantly, the American diet contains high levels of n6-PUFAs due to heavy reliance on corn and soybeans and low consumption of n3-PUFAs [reviewed in 7]. Consequently, this high n6/n3-PUFA ratio may increase the production of n6-PUFA metabolites (which are generally pro-inflammatory in nature) at the expense of n3-PUFA-derived metabolites (which are generally anti-inflammatory, eg, the specialized pro-resolving mediators, resolvins, protectins, and maresins).^9,10 An n6-PUFA-rich diet can lead to hepatic damage that manifests as steatosis, increased glycogen storage, increased NFκB activity, and apoptosis.¹¹ When combined with other toxic insults (eg, alcohol), an n6-PUFA-rich diet leads to exacerbated hepatic injury, as we have previously shown with mice-fed ethanol and a diet high in n6-PUFAs.⁶ Further, we have also demonstrated that transgenic fat-1 mice with endogenously elevated n3-PUFAs and subsequently decreased n6/n3-PUFA ratio are partially protected from ALD,¹² suggesting that n3-PUFAs and/or their metabolites activate hepatoprotective mechanisms and could be a promising nutritional intervention for ALD. These transgenic mice express the C elegans gene fat-1 which encodes an n3-fatty acid desaturase, and therefore, directly convert n6-PUFAs to n3, leading to an increase in n3-PUFAs with a decrease in the n6/n3-PUFA ratio in all tissues without the need for dietary intervention.¹³

It has been previously reported that the canonical Wnt pathway (ie, β-catenin dependent), an important signaling pathway for liver development and homeostasis,¹⁴ is compromised in rodent models of ALD.¹⁵ Wnt signaling is complex and is comprised of a number of gene products. There are 19 Wnt genes in both humans and mice and all function as paracrine signaling proteins that bind to cell-surface receptors (Frizzled, [FZD]) which then propagate the signal through a series of proteins including Dishevelled (DVL), Adenomatous Polyposis Coli (APC), and glycogen synthase kinase-3β (GSK-3β), ultimately leading to the cytoplasmic accumulation and nuclear translocation of β-catenin.¹⁶ Nuclear β-catenin functions as a transcription factor to regulate the expression of genes important for many liver functions such as hepatocyte proliferation.¹⁷ Signaling is terminated in part by AXIN2, a feedback inhibitor by virtue of its destabilizing effect on β-catenin.¹⁸ Importantly, the expression of Axin2 is induced by Wnt activation and has been shown to be a reliable indicator of active signaling through this pathway.¹⁹ The goal of the current study was to examine whether an endogenous increase in n3-PUFAs with a reduction in the ratio of n6/n3-PUFAs attenuates ALD through modulation of the Wnt signaling pathway in experimental mice.

2 |. MATERIALS AND METHODS

2.1 |. Animals and experimental design

Wild-type (WT) mice and transgenic fat-1 littermates were used in a preclinical animal model of ALD mimicking alcohol-associated hepatitis (AH), an advanced form of human ALD which is often complicated by severe inflammation and bacterial infection.³ fat-1 mice express the C elegans gene for the omega-3 fatty acid desaturase FAT-1, which confers on these mice the ability to directly convert n6-PUFAs to n3-PUFAs. As a result, fat-1 mice have elevated levels of tissue and systemic n3-PUFAs without the need for dietary n3-PUFA supplementation, and can lower the n6/n3 ratio from as high as 49:1 as found in WT mice to 0.7:1 in fat-1 mice.¹³ All animal experiments were conducted in accordance with the guidelines set forth by the University of Louisville Institutional Animal Care and Use Committee under an approved protocol (number 15423 to IAK). Mice were housed in a specific pathogen-free facility under constant temperature (23.9°C) with a 12-hour light:dark cycle. This facility is currently accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. For all experiments, male fat-1^+/− mice (herein referred to as fat-1) were bred to female WT C57BL/6J mice. Pups were genotyped by standard PCR using primers reported in Table S1. As a positive control for the PCR genotyping assay, we used primers to the Gdf5 gene. Prior to placing mice on an all-liquid diet, they were initially fed standard laboratory chow (autoclavable 5010 diet; LabDiet, St. Louis, MO) and water (ad libitum). The mice were switched to an all-liquid diet which was their sole source of nutrition and water [Lieber-DeCarli ‘82, BioServ, Flemington, NJ, part numbers: F1259SP (control), F1258SP (ethanol-EtOH)]. For the EtOH-fed group, the diet was supplemented with increasing concentrations of EtOH (v/v) as follows: 2 days each of 0%, 1%, and 2%, 1 week each for 4% and 5%, and finally, 6% for 3 weeks. The control, pair-fed (PF) groups received the same number of calories as the EtOH-fed group but in the form of maltose dextrin. Diets were prepared fresh, food consumption monitored daily, and body weights recorded weekly. In order to mimic acute inflammation, bacterial infection, and sepsis often found in advanced human ALD/AH,³ mice were challenged with a single dose of lipopolysaccharide (LPS, 5 mg/kg, i.p., InvivoGen, San Diego, CA) 24 hours prior to euthanasia. This dose of LPS was chosen because it is high enough to lead to a proinflammatory response but not cause significant mortality as is observed at higher doses (>15 mg/kg).²⁰ Sodium chloride (0.9% [w/v], APP Pharmaceuticals, LLC, Schaumburg, IL) was used as the vehicle for LPS and for control mice. Mice were divided into the following experimental groups:(a) WT-PF; (b) fat-1-PF; (c) WT-EtOH+LPS; (d) fat-1-EtOH+LPS. These studies were performed using male mice.

2.2 |. Plasma ALT measurement

Plasma ALT levels, as a measurement of liver injury, was determined with ALT/GPT reagent from Thermo Fisher Scientific (Waltham, MA) according the manufacturer’s instructions.

2.3 |. Liver tissue staining

Liver tissue staining and analysis was performed on formalin-fixed, paraffin-embedded tissue sections. Hematoxylin and Eosin (H&E) staining was performed to evaluate the overall liver tissue pathological changes. TUNEL (Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling) staining using a commercially available kit from Millipore (Burlington, MA) was used to determine the number of apoptotic/necrotic cells.²¹ Quantitation of TUNEL-positive cells was performed by manually counting TUNEL-positive cells in 10 fields/liver section (200× magnification, n = 6–7) in a blinded manner by two independent investigators. Immunohistochemistry for myeloperoxidase (MPO)-positive-cells was performed with anti-MPO antibodies (R&D Systems, Minneapolis, MN). F4/80-positive cells were visualized by immunohistochemistry using an anti-mouse F4/80 antibody (Abcam, Cambridge, MA). Ten fields per mouse liver (n = 6–10, 200× magnification) were analyzed for both MPO- and F4/80-positive cells by a macro written in Image J. The expression of AXIN2 in liver tissue was evaluated by immunohistochemistry using an anti-mouse AXIN2 antibody (ProSci, Poway, CA).

Oil Red O staining was performed as described previously, with minor modifications.²² Cryosections were incubated with 0.5% (w/v) Oil Red O in propylene glycol (Electron Microscopy Sciences, Hadfield, PA) at 60°C for 10 minutes. Staining was quantified by imaging 10 random fields/section (200× magnification), and then measuring Oil Red O-positive areas using a macro written in Image J.

2.4 |. Measurement of hepatic n3- and n6-PUFAs

Hepatic fatty acid analysis was performed by gas chromatography as described previously.²³

2.5 |. RNA isolation and real-time PCR

Total RNA was isolated from snap-frozen liver samples or cultured primary hepatocytes using TRIzol reagent (Thermo Fisher Scientific). Carryover genomic DNA was eliminated by digestion with RNase-free DNase I (Thermo Fisher Scientific), and then 500 ng RNA was used to synthesize cDNA with qScript cDNA SuperMix (Quanta Biosciences, Beverly, MA). Real-time, semiquantitative PCR (qPCR) was performed on the cDNA equivalent of 10 ng RNA (0.1 ng for 18S rRNA) using PerfeCTa SYBR Green Fast Mix (Quanta Biosciences) on the Applied Biosystems 7900HT or StepOne Plus platform (Foster City, CA). Primer sequences are provided in Table S1. Data were reduced using the ΔΔCt method.²⁴

2.6 |. Wnt PCR array

The expression of 84 genes in the Wnt signaling pathway was determined by a PCR array (RT² Profiler PCR Array, Qiagen, Germantown, MD) as outlined by the manufacturer. The data are presented as fold-change (FC) ± SEM, or as 2^−ΔCt (×1000) to visualize relative changes in expression across all four groups. Cytoscape software was used to visualize the relationship among differentially expressed genes.²⁵

2.7 |. Liver cytokine measurement

Analysis of liver cytokine levels were measured with the V-PLEX assay platform (Meso Scale Discovery, Rockville, MD). Briefly, liver tissue from 8 to 11 mice/group was homogenized in 20 mM of Tris-Cl, 2 mM of EDTA, 10 mM of EGTA, 1% of Triton X-100, 0.25 M of sucrose, and supplemented with protease and phosphatase inhibitors (Halt, Thermo Fisher Scientific), centrifuged for 10 minutes at 10 000× g. Supernatants were collected, assayed for protein concentration by the BCA method (Thermo Fisher Scientific) and 600 μg analyzed by immunoassay. Data were collected on the MESO Sector S 600 Instrument and analyzed using Discovery Workbench, v 4.0 (Meso Scale Discovery).

2.8 |. Western blot analysis

Liver samples were homogenized in RIPA buffer (20 mM Tris-Cl, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% IGEPAL, and 1% deoxycholate) supplemented with Halt protease/phosphatase inhibitor (Thermo Fisher Scientific), centrifuged for 10 minutes at 10 000× g and protein concentration in the supernatants measured by the BCA method (Thermo Fisher Scientific). Fifty μg protein was separated on CriterionTGX Any kD precast gels (Bio-Rad Laboratories, Hercules, CA) and electroblotted onto Immun-Blot PVDF membranes (Bio-Rad Laboratories). Blots were incubated overnight at 4°C with the indicated primary antibodies (see below), and then, for 1 hour with HRP-conjugated secondary antibodies (Thermo Fisher Scientific). Signals were developed with Clarity Max Western ECL substrate and collected using the ChemiDoc imaging system and quantitated with Image Lab software, version 6.0.1 (all from Bio-Rad Laboratories). Primary antibodies used: anti-FRA-1 (catalog no. PA5–76185; Thermo Fisher Scientific), anti-IL-1β (catalog no. ab9722; Abcam, Cambridge, MA), anti-phospho-JNK1 (catalog no. 9255S; Cell Signaling Technology, Danvers, MA), anti-JNK1 (catalog no. 370; BioVision, Milpitas, CA), anti-LSD1 (catalog no. 2139; Cell Signaling Technology), and anti-β-actin (catalog no. 4970; Cell Signaling Technology).

2.9 |. Primary hepatocyte cultures

Primary cultures of hepatocytes from WT and fat-1 mice were established using a two-step perfusion method as described previously.²⁶ Cells were suspended in Williams E medium containing 10% of FBS, 2 mM of glutamine, penicillin/streptomycin (all from Thermo Fisher Scientific), 1X ITS supplement (Corning, Inc, Corning, NY), and 0.1 μM of dexamethasone (Sigma-Aldrich, St. Louis MO), and then seeded into a 96-well plate (for MTT assay) or a 24-well plate (for qPCR analyses) at a density of ~65 000 cells/cm². Following overnight incubation, cells were washed two times with PBS (Thermo Fisher Scientific), and then, incubated with various treatments in Williams E medium containing 2 mM of glutamine and antibiotics. Palmitic acid (PA) treatment: PA (200 mM, Sigma-Aldrich) was prepared in 95% of EtOH, and then, diluted to a concentration of 400 μM in Williams E medium containing 1% of fatty acid-free BSA (final concentration of EtOH was 33 μM). Primary hepatocytes were treated with PA (400 μM) and incubated for 24 hours. TNF-α treatment: hepatocytes were first pretreated with 0.4 μg/mL actinomycin D (ActD, Sigma-Aldrich) for 30 minutes prior to the addition of 25 ng/mL TNF-α for 24 hours (R&D Systems, Minneapolis, MN). TGF-β treatment: hepatocytes were treated with 5 ng/mL recombinant TGF-β (Thermo Fisher Scientific) for 24–48 hours prior to MTT assay. Wnt stimulation: hepatocytes were incubated with 200 ng/mL Wnt-9b for 4 hours, and then, the expression of Axin2 determined by qPCR. Wnt-9b was selected because it has been previously shown that this isoform activates β-catenin in pericentral hepatocytes.²⁷ LiCl stimulation: hepatocytes were stimulated for 4 hours with 10 mM of LiCl (Sigma-Aldrich) prepared in PBS prior to analysis of gene expression by qPCR. LiCl inhibits GSK-3β activity, thereby preventing phosphorylation of β-catenin and leading to nuclear translocation and transcriptional activation.²⁸ Resolvin treatment: primary hepatocytes were incubated for 3 hours with 100 nM of RvD1 or 100 nM of RvE1 (Cayman Chemical, Ann Arbor MI) and gene expression measured by qPCR as described above.

2.10 |. Methylthiazolyldiphenyl-tetrazolium bromide (MTT) assay

After specific treatments, primary hepatocytes were incubated for ~4 hours with 0.5 mg/mL MTT (Sigma-Aldrich) prepared in culture medium. Medium was removed and the dye extracted with a solution of 4 mM of HCl and 0.1% of IGEPAL prepared in isopropanol. The absorbance at 630 nm was measured by spectrophotometry, and then, subtracted from the absorbance at 570 nm.

2.11 |. RAW264.7 cells

The mouse macrophage cell line RAW264.7 was obtained from ATCC (Manassas, VA). Cultures were grown in RPMI-1640 media (Sigma Chemical Co.) supplemented with 10% of FBS and penicillin/streptomycin (Thermo Fisher). Subconfluent cultures were treated for 3 hours with 100 nM of RvD1 or 100 nM of RvE1 (Cayman Chemical) and gene expression measured by qPCR as described above.

2.12 |. Statistical analyses

GraphPad Prism 8.0 software (GraphPad Software, Inc, La Jolla, CA) was used to perform statistical analyses (except for the Wnt PCR array). Differences among multiple groups were analyzed by the unpaired one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. The Wnt PCR array was analyzed and statistical analysis performed using the online data analysis tool at the GeneGlobe Data Analysis Center (http://www.qiagen.com/geneglobe). The data are expressed as the mean ± standard error of the mean (SEM). A P value of <.05 was considered statistically significant.

3 |. RESULTS

3.1 |. fat-1 mice with enhanced hepatic n3-PUFA levels are partially protected from liver injury induced by EtOH+LPS administration

To test the impact of increased endogenous n3-PUFA levels on ALD, we fed WT and fat-1 mice an EtOH-containing diet for 6 weeks followed by a single LPS challenge to mimic severe human ALD/AH which is characterized by acute inflammation and sepsis.³ The mice were euthanized 24 hours after LPS treatment. Interestingly, fat-1 mice had less LPS-induced mortality when compared to WT mice (Figure S1A). As shown in Figure 1A, EtOH+LPS treatment led to a large increase in plasma ALT activity in WT compared to fat-1 mice (~400 U/L vs ~200 U/L, respectively); therefore, fat-1 mice had significantly lower ALT levels, more than twofold less than WT in response to EtOH+LPS challenge. Hepatocyte cell death is a hallmark of ALD.²⁹ We observed a significant increase in the number of TUNEL-positive cells in response to EtOH+LPS treatment in both WT and fat-1 mice, with slightly lower levels in fat-1 mice (Figure S2A,B). Interestingly, however, we found that primary hepatocytes isolated from fat-1 mice and treated with cell death-inducing components were more resistant to cell death than hepatocytes isolated from WT mice (Figure S2C–E), suggesting that fat-1 hepatocytes may be inherently protected from these toxic insults. Further, H&E-staining of liver tissue revealed less hepatic steatosis in fat-1 compared to WT EtOH+LPS-treated mice (Figure 1B), which was confirmed by Oil Red O staining (Figure 1C,D). Of note, although WT and fat-1 mice consumed similar amounts of EtOH-containing diet, the latter tended to gain less weight once they were provided 6% of EtOH in the diet (see Figure S1B,C).

FIGURE 1.

Evaluation of liver injury in experimental animals. A, Plasma ALT activity demonstrating decreased levels in EtOH-fed, LPS-challenged fat-1 vs WT mice. B, H&E-stained liver sections, C-D, Oil Red O staining and quantitation, respectively. Scale bar is 20 μm. E, Total n6-PUFA levels. F, Total n3-PUFA levels. G, Ratio of hepatic n6/n3-PUFAs. Values are expressed as mean ± SEM, (n = 7–10), *P < .05, one-way ANOVA

Finally, analysis of the hepatic n3- and n6-PUFA levels revealed that the total n6-PUFAs were not appreciably altered either by genotype or EtOH+LPS treatment (Figure 1E). Both PF and EtOH+LPS-treated fat-1 mice had significantly more hepatic n3-PUFAs than their WT counterparts (Figure 1F). Interestingly, the level of total n3-PUFAs was significantly elevated in the EtOH+LPS group compared to PF mice in both WT and fat-1 mice (Figure 1F). Importantly, the n6/n3-PUFA ratio was decreased in both fat-1 PF and EtOH-fed mice compared to WT animals (Figure 1G). These data demonstrate that compared to WT littermates, fat-1 mice have less EtOH+LPS-induced liver injury, associated with decreased n6/n3-PUFA ratio due to increased n3-PUFA levels.

3.2 |. fat-1 mice have attenuated inflammatory responses induced by EtOH+LPS challenge

Hepatic macrophages and neutrophils play a critical role in ALD pathogenesis, specifically in inflammatory responses [reviewed in 30]. We observed that in WT mice, there was a significant decrease in hepatic macrophages following EtOH+LPS-treatment that was not observed in fat-1 mice (Figure 2A,B). fat-1 mice had significantly fewer hepatic neutrophils than WT mice in response to EtOH+LPS administration (Figure 2C,D). Compared to WT, fat-1 mice also had a significantly lower expression of EtOH+LPS-induced Ly6g, a marker for myeloid cells including monocytes, granulocytes, and neutrophils (Figure 2E), while Cxcl1, a chemokine involved in neutrophil recruitment and activation, was similarly induced by EtOH+LPS in both genotypes (Figure S3A,B). Of note, the expression of Pai-1, a gene which can promote neutrophil recruitment,³¹ was significantly lower in fat-1 vs WT EtOH+LPS-treated mice (Figure 2F). Further, TNF-α mRNA and protein levels were similarly increased by EtOH+LPS in both WT and fat-1 mice (Figure 3A,B). The expression of Il-6 mRNA was induced by EtOH+LPS to a greater extent in WT vs fat-1 mice (Figure 3C), and although a similar trend was found for IL-6 protein levels, there were no significant differences between WT and fat-1 EtOH+LPS-treated mice (Figure 3D). The expression of Il-1β mRNA (which represents total Il-1β) was increased approximately to the same levels in WT and fat-1 mice treated with EtOH+LPS (Figure 3E); however, compared to WT mice, both PF and EtOH+LPS-treated fat-1 mice had significantly less cleaved (mature) IL-1β as determined by Western blotting (Figure 3F). Finally, the expression of the anti-inflammatory cytokine Il-10 was increased significantly more in fat-1 vs WT mice following treatment with EtOH+LPS at the mRNA level (Figure 3G). Although there was an increase in IL-10 protein in EtOH+LPS-treated fat-1 mice vs WT, it was not statistically significant (Figure 3H).

FIGURE 2.

Hepatic neutrophils and macrophages in fat-1 and WT mice. A, Photomicrograph of immunohistochemistry for F4/80 (macrophages) expression (arrows indicate F4/80-positive cells) and (B) qPCR quantitation for F4/80 mRNA expression. C, Photomicrographs of immunohistochemistry for MPO expression (neutrophils) in liver sections (arrows indicate MPO-positive cells). D, Quantitation of MPO-positive cells. E-F, Hepatic expression Ly6g, and Pai-1, respectively. Six-10 animals/per group; *P < .05, one-way ANOVA. Scale bars are 50 μm

FIGURE 3.

Hepatic expression of TNF-α, IL-6, IL-1β, and IL-10 following chronic EtOH exposure and LPS challenge in WT and fat-1 mice. The expression of Tnf-α, Il-6, Il-1β, and Il-10 was measured by qPCR (panels A, C, E, and G, respectively) and by immunoassay [TNF-α (B), IL-6 (D), and IL-10 (H)] or Western blotting/densitometry [cleaved, 17 kDa form of IL-1β (F)]. *P < .05, one-way ANOVA, n = 3–10 animals/per group

3.3 |. fat-1 mice are partially protected from EtOH+LPS-mediated downregulation of canonical Wnt signaling: analysis of global changes in the Wnt pathway

The canonical Wnt signaling pathway (ie, β-catenin dependent) directly and indirectly regulates numerous biological processes necessary for liver development and homeostasis, including apoptosis, proliferation, metabolism, and inflammation.³² Importantly, the activity of this pathway has been previously shown to be downregulated in rodent models of ALD.¹⁵ We sought to determine if n3-PUFAs could modulate expression of this pathway as part of the mechanism by which n3-PUFAs exert their beneficial effect on ALD. Therefore, we performed a PCR array on liver mRNA isolated from PF and EtOH+LPS-treated WT and fat-1 mice to interrogate the expression of Wnt ligands and other Wnt pathway-related genes. The PCR array gene list (84 genes in total) is provided in Table S2. Treatment of WT mice with EtOH+LPS led to the differential expression of 25 genes when compared to the PF group, which included several Wnts (Wnt-2 and 7a), Wnt receptors (Fzd-5, −7, and −8), the transcription factor Tcf7l1, the inhibitor Dkk3, and the target gene Fbxw4 (Table 1). Nineteen genes were differentially expressed in EtOH+LPS-treated fat-1 mice (vs PF), including the inhibitor Sfrp1 and the transcription factor Jun (Table 2). A comparison of these two gene sets revealed 12 genes in common between WT and fat-1 mice, 13 unique to WT mice, and 7 unique to fat-1 mice. These data are illustrated in the Venn diagram (Figure 4A). The relationship among the differentially expressed genes in each group is illustrated below the Venn diagram and highlights that Wnt ligands and the Lef1 transcription factor are among the most highly changed genes within each cluster. When a direct comparison of the WT and fat-1 EtOH+LPS groups was performed, there were six genes that were significantly differentially regulated between WT and fat-1 mice (two upregulated and four downregulated), as illustrated in the volcano plot (Figure 4B) and listed in Table 3.

TABLE 1.

WT+EtOH+LPS vs WT-PF gene changes as determined by qPCR array

Gene	Fold change	P value
Mmp7	55.5	.000001
Fosl1	28.82	.000004
Frzb	5.02	.000055
Porcn	2.7	.000071
Wnt5a	2.32	.018154
Dkk3	2.15	.009429
Wnt4	1.94	.012746
Mapk8	1.74	.02531
Ppard	1.62	.012296
Wnt7a	−1.53	.002053
Fbxw4	−1.67	.001952
Fzd7	−1.72	.042764
Fzd5	−1.9	.007486
Sox17	−1.93	.017011
Prickle1	−1.98	.011169
Tcf7l1	−1.99	.00274
Rhou	−2.25	.008977
Daam1	−2.28	.018936
Tcf7	−3.34	.013213
Wnt2	−3.78	.006365
Nkd1	−4.93	.000271
Fzd8	−6.91	.007205
Axin2	−8.42	.004633
Ccnd1	−8.74	.036213
Wif1	−13.79	.04637

TABLE 2.

fat-1+EtOH+LPS vs fat-1-PF gene changes as determined by qPCR array

Gene	Fold change	P value
Mmp7	35.07	.000901
Fosl1	13.74	.019603
Frzb	6.92	.000987
Wnt5a	2.46	.003017
Jun	−2.2	.042767
Fzd1	−2.34	.020871
Axin2	−2.38	.030047
Dixdc1	−2.5	.033317
Lef1	−2.54	.018519
Nkd1	−2.59	.002037
Prickle1	−2.72	.010309
Rhou	−2.91	.004754
Fzd4	−2.95	.046333
Daam1	−3.62	.001746
Fzd3	−3.62	.024214
Tcf7	−5.31	.001982
Sfrp1	−7.88	.046557
Ccnd1	−9.65	.000006
Wif1	−10.71	.007232

FIGURE 4.

PCR array analysis of the hepatic expression of Wnt pathway genes. A, Venn diagram summarizing gene expression changes induced by EtOH+LPS in fat-1 and WT mice (Red, upregulated; Blue, downregulated). B, Volcano plot demonstrating gene expression changes between EtOH+LPS-treated fat-1 and WT mice. n = 4 (PF) or n = 3 (EtOH+LPS-treated)

TABLE 3.

fat-1+EtOH+LPS vs WT+EtOH+LPS gene changes as determined by qPCR array

Gene	Fold Change	P value
Axin2	3.17	.038144
Nkd1	1.69	.04775
Apc	−1.54	.025884
Mapk8	−1.75	.048314
Porcn	−2.1	.011555
Fosl1	−2.75	.017928

3.4 |. Analysis of differentially expressed genes of Wnt pathway in vivo and in vitro: effects of n3-PUFAs

To identify Wnt pathway genes potentially regulated by n3-PUFAs, we performed a direct comparison of the genes differentially regulated by EtOH+LPS in WT and fat-1 mice. This analysis identified six genes, which are presented in Figures 5 and 6 and in Table 3. Of note, there were no differences in the expression of these genes between control PF WT and fat-1 animals. The expression of Nkd1 was downregulated by EtOH+LPS in both genotypes (although to a greater extent in WT mice, Figure 5A). The expression of other genes, such as Apc, Porcn, Fosl1/Fra-1, and Mapk8/Jnk1 was significantly increased by EtOH+LPS, predominantly in WT mice, resulting in significantly higher levels in WT vs fat-1 EtOH+LPS-treated mice (Figure 5B–D,F, respectively). Changes in the expression of the selected genes, for example, Fosl11/Fra-1 and Mapk8/Jnk1 were confirmed at the protein level (Figure 5E,G, respectively), and had a similar trend as mRNA expression.

FIGURE 5.

Differential expression of Wnt pathway genes between EtOH+LPS-treated WT and fat-1 mice. PCR array data for genes whose expression is significantly different between treated groups, expressed as mean 2^−ΔCt ± SEM (panels A-D and F). Western blots and quantitative analysis of FOSL1/FRA-1 (E) and phospho-JNK1 and JNK-1 (G). *P < .05, one-way ANOVA, n = 6–10/group

FIGURE 6.

Axin2 expression in EtOH+LPS-treated mice and in primary hepatocytes from WT and fat-1 mice. A, PCR array data demonstrating larger downregulation of hepatic Axin2 expression in WT vs fat-1 mice following EtOH+LPS treatment. B, Immunohistochemistry of AXIN2 expression demonstrating expression around the central vein (CV). C, Axin2 expression via qPCR of primary hepatocytes treated with 50 mM of EtOH for 24 hours or with 200 ng/mL Wnt-9b for 4 hours. D, Axin2 expression in primary hepatocytes treated with 10 mM of LiCl for 4 hours. Data are presented as the mean ± SEM, *P < .05, one-way ANOVA

Similar to Nkd1, the expression of Axin2, a direct measure of Wnt signaling, was decreased by EtOH+LPS-treatment in both genotypes. However, this decrease was greater in WT mice resulting in ~3-fold higher Axin2 levels in fat-1 vs WT EtOH+LPS-treated mice (Figure 6A). This observation suggests that fat-1 mice were partially protected from EtOH+LPS-induced downregulation of the Wnt pathway. Loss of Wnt signaling has been previously shown to be associated with ALD in a rat model whereby rescuing activity ameliorates the severity of ALD.¹⁵ It has been demonstrated that hepatic expression of AXIN2 is high in the region immediately adjacent to the central vein and is lost following tissue injury.¹⁷ By immunohistochemistry, we found that the expression pattern of AXIN2 was similar in PF mice from both groups (Figure 6B) but observed that there was a reduced pericentral signal in EtOH+LPS-treated WT vs fat-1 mice. To directly test the effect of EtOH on Wnt signaling in vitro, we isolated primary hepatocytes from WT and fat-1 mice, incubated them with EtOH, then assayed the expression of Axin2 by qPCR. As shown in Figure 6C, EtOH treatment decreased the expression of Axin2 (albeit not significantly) in WT, but not in fat-1 primary hepatocytes. In fact, the expression of Axin2 in fat-1 hepatocytes was increased (although not significantly) compared to control levels by a mechanism that remains to be elucidated. Concurrently, we also tested the responsiveness of WT and fat-1 primary hepatocytes to Wnt-9b, which has been previously shown to activate Wnt signaling in pericentral hepatocytes.²⁷ Wnt-9b stimulation had no effect on the expression of Axin2 in WT hepatocytes but fat-1 hepatocytes showed an almost threefold increase (Figure 6C). As an independent assay to measure the responsiveness of primary hepatocytes to Wnt signaling, we stimulated WT and fat-1 primary hepatocytes with LiCl, which activates the canonical Wnt pathway downstream of the receptor by directly inhibiting the activity of GSK-3β, thus preventing phosphorylation and degradation of the transcription factor, β-catenin. In our studies, LiCl stimulated the expression of Axin2 in both WT and fat-1 hepatocytes; however, the effect was significantly greater in fat-1 cells (Figure 6D). Collectively, our data demonstrate that fat-1 mice are resistant to EtOH+LPS-mediated downregulation of Axin2 expression and that fat-1 primary hepatocytes are more responsive to Wnt pathway activation than those isolated from WT mice.

n3-PUFAs are precursors to specific lipid mediators, such as resolvins (eg, RvD1 and RvE1) that act through specific cell-surface receptors such as FPR2 (RvD1) and CHEMR23 (RvE1), which could mediate the effects seen in fat-1 mice. We assayed for the expression of Fpr2 and ChemR23 in liver tissue of WT and fat-1 mice by qPCR (Figure S4). Interestingly, the expression of Fpr2 was increased in EtOH+LPS-treated WT and fat-1 mice. There was a smaller effect of EtOH+LPS treatment on the expression of ChemR23. We performed an in vitro experiment to determine if resolvins can regulate expression of genes associated with the Wnt pathway. We found no effect of RvD1 or RvE1 on Axin2 expression in primary hepatocytes (Figure 7A); however, these lipid metabolites decreased the expression of both Apc and Porcn (Figure 7B,C, respectively). Although there was a downregulation of Fosl1/Fra-1 expression in response to both resolvins, it was not statistically significant (Figure 7D). We were unable to detect expression of Nkd1 and we did not assay for Mapk8/Jnk1 because it is generally not regulated at the transcriptional level.³³ These data suggest that although neither RvD1 nor RvE1 affected signaling upstream of Axin2, they could directly regulate downstream Wnt-related genes in primary hepatocytes in a manner consistent with that seen in vivo. There were no effects of RvD1 or RvE1 on the expression of Apc, Fosl1/Fra-1, or Porcn in RAW264.7 macrophages (Figure S5). The expression of Axin2 and Nkd1 in these cells was not detected.

FIGURE 7.

Effect of RvD1 and RvE1 on gene expression in primary hepatocytes. Primary hepatocyte cultures from WT mice were treated with 100 nM of RvD1 or 100 nM of RvE1 for 3 hours followed by qPCR analysis for the expression of genes within the Wnt signaling pathway, (A) Axin2, (B) Apc, and (C) Porcn and the Wnt target gene, (D) Fosl1/Fra-1. Data are presented as the mean ± SEM, *P < .05, one-way ANOVA

4 |. DISCUSSION

In the current study, we set out to explore the effects of endogenous n3-PUFA enrichment on liver injury in experimental ALD, and to examine the underlying mechanisms. We found that chronic EtOH feeding combined with an LPS challenge produced less severe liver injury in fat-1 mice (which have endogenously increased n3-PUFA levels) compared to WT littermates, as determined by a number of endpoints including decreased plasma ALT levels, reduced steatosis, fewer hepatic neutrophils, decreased levels of several pro-inflammatory cytokines, and increased expression of the pro-restorative cytokine Il-10. We found only a modest decrease in the number of TUNEL-positive cells in EtOH+LPS-treated fat-1 vs WT mice; however, we did find that in vitro, primary hepatocytes from fat-1 mice were more refractory to apoptosis-inducing signals that have relevance to ALD, such as TNF-α and palmitic acid.³⁴ Although the effect on hepatocyte cell death is small in fat-1 vs WT mice, even slight effects may have significant biological effects due to a decrease in the release of damage-associated molecular patterns (DAMPs), which are known drivers of ALD.³⁵ These data suggest that n3-PUFAs may have a pleiotropic effect, acting both on lipid metabolism and on the inflammatory response. In a mouse model of nonalcoholic fatty liver disease (NAFLD), Liebig et al found similar results after 6 weeks of a high-fat diet on plasma ALT, steatosis, and CAE-positive cells, but no effect on macrophage numbers in the liver.³⁶ Interestingly, we found that the expression of F4/80 was decreased in EtOH+LPS-treated WT but not fat-1 mice, suggesting a reduction in the number of macrophages in WT mice. The role of macrophages in ALD is complex because there are several phenotypes with distinct and often opposing functions.³⁷ However, it has been previously demonstrated that in both an acute-on-chronic and chronic mouse model of ALD, there was a decrease in the total number of F4/80-positive macrophages.^38,39 This loss of macrophages may exacerbate ALD by eliminating an immune cell essential for limiting the extent of tissue injury and promoting repair. Indeed, it has been previously shown that macrophages secrete Wnt-3a, an important signaling cue for tissue regeneration in mouse models of chronic liver disease.⁴⁰

The Wnt signaling pathway is important for both liver development and homeostasis,¹⁴ and it has been reported that rats chronically fed EtOH had decreased Wnt signaling, but that pharmacological activation of this pathway attenuated ALD.⁴¹ We hypothesized that Wnt signaling is one of the mechanisms by which n3-PUFAs protect mice from ALD. In our study, EtOH+LPS treatment led to the differential expression of 25 Wnt-associated genes in WT mice and 19 in fat-1 mice, compared to their own PF groups, with 12 of these genes common to both genotypes. Wnt-associated genes that were specifically downregulated by EtOH+LPS in WT mice included several Wnt ligands (Wnt-2 and 7a), receptors (Fzd5/7/8), and the transcription factor Tcf7l1. Conversely, the expression of Dkk3, a potent inhibitor of Wnt signaling, was significantly increased in WT mice. Taken together, these changes may lead to an overall deficit in Wnt signaling as a result of EtOH+LPS treatment. Also, in WT mice, the expression of Fbxw4 was downregulated by EtOH+LPS Fbxw4 has been demonstrated to be important for protein ubiquitination/degradation,⁴² and reduced expression of this gene could potentially exacerbate the accumulation of damaged or misfolded proteins (ie, endoplasmic reticulum [ER] stress). ER stress is a significant contributor to liver damage resulting from excessive alcohol consumption.⁴³ Of the differentially expressed genes identified in fat-1 mice, we found decreased expression of the Wnt signaling antagonist Sfrp1 in EtOH+LPS-treated fat-1 mice as compared to PF. Finally, the expression of the Wnt-regulated gene Jun was also decreased in EtOH+LPS-treated fat-1 mice (vs PF). JUN has been demonstrated to promote inflammation and fibrosis in rodent models of liver disease.⁴⁴ A direct comparison of the EtOH+LPS groups between genotypes revealed a significant difference in the expression of six genes (including Axin2, a readout of Wnt signaling) that may explain some of the underlying mechanisms by which n3-PUFAs mediate their effects. The expression of Apc, for example, was induced by EtOH+LPS in WT but not fat-1 mice. APC is a negative regulator of Wnt signaling through its role in forming the “destruction complex” that targets β-catenin for proteasomal degradation.¹⁴ Maintaining normal expression of Apc in fat-1 mice is, therefore, important for the liver’s response to the detrimental effects of EtOH. The expression of Mapk8 (Jnk1), a central player in the regulation of inflammation in the liver [in both parenchymal and non-parenchymal cells⁴⁵], was also decreased in fat-1 mice. The decrease in expression of this gene and decreased levels of phospho-JNK1 correlate with a decreased inflammatory response in fat-1 mice and may be an important mechanism by which n3-PUFAs regulate hepatic inflammation. Finally, the expression of Fosl1 was decreased in fat-1 mice when compared to WT. FOSL1 (also known as FRA-1) is part of the AP-1 transcription factor complex, and transgenic mice overexpressing this protein have increased expression of pro-inflammatory hepatic cytokines, increased neutrophil infiltration, and fibrosis.⁴⁶ However, the effect on overall inflammation may be somewhat limited in scope as we found that the expression of some cytokines (eg, TNF-α) was unaffected in fat-1 mice.

The mechanisms by which n3-PUFAs regulate Wnt signaling may be multifaceted, including direct effects mediated by n3-PUFA/receptor interactions (eg, with GPR120) or indirect effects mediated by specialized pro-resolving mediators (SPMs) such as the D- and E-series resolvins. In previous studies, it has been demonstrated fat-1 mice express higher levels of SPMs, including these resolvins.⁴⁷ We have also recently demonstrated that resolvin D1-treated mice have attenuated liver injury.¹² We did find that direct treatment of primary hepatocytes (but not a macrophage cell line) with RvD1 or RvE1 regulated several of the genes that were differentially regulated in fat-1 vs WT mice. However, we do not have any evidence that SPMs directly alter the expression of Wnt gene expression. Interestingly, it was recently shown that the n3-PUFAs α-linolenic acid and eicosapentaenoic acid can directly modulate the function of Wnt receptors (FZDs) by regulating their dimerization and activation.⁴⁸ Future studies are needed to investigate the underlying mechanisms by which n3-PUFAs can regulate Wnt signaling.

In conclusion, our data demonstrate that fat-1 mice were partially protected from EtOH+LPS-induced liver damage via the rescue of Wnt signaling (Figure 8) and provide a rationale to further investigate the use of n3-PUFA supplementation as an adjuvant therapy to treat ALD.

FIGURE 8.

Schematic of the hepatic effects of EtOH+LPS on Wnt gene expression in WT and fat-1 mice. WT mice have a high n6/n3 ratio and when treated with EtOH+LPS have reduced Wnt signaling and increased expression of pro-inflammatory genes (Mapk8/Jnk1 and Fosl1/Fra-1). fat-1 mice have a decreased n6/n3 ratio and partially rescued Wnt signaling following EtOH+LPS treatment, decreased expression of pro-inflammatory genes, and therefore, exhibit less liver injury than WT mice

Abbreviations:

ActD

actinomycin D

AH

alcohol-associated hepatitis

ALD

alcohol-associated liver disease

ALT

alanine aminotransferase

APC

adenomatous polyposis coli

BSA

bovine serum albumin

CAE

chloroacetate esterase

C elegans

Caenorhabditis elegans

DHA

docosahexaenoic acid

DVL

disheveled

EPA

eicosapentaenoic acid

EtOH

ethanol

fat-1

n-3 fatty acid desaturase-1

FC

fold-change

FDA

(United States) Food and Drug Administration

FZD

frizzled

GSK-3β

glycogen synthase kinase-3β

H&E

hematoxylin and eosin

IL-10

interleukin 10

LA

linoleic acid

LPS

lipopolysaccharide

MPO

myeloperoxidase

MTT

Methylthiazolyldiphenyl-tetrazolium bromide

n3-PUFA

n3-polyunsaturated fatty acid

n6-PUFA

n6-polyunsaturated fatty acid

PA

palmitic acid

PAI-1

plasminogen activator inhibitor-1

PF

pair fed

qPCR

semiquantitative real-time polymerase chain reaction

TG

triglyceride

TGF-β

transforming growth factor-β

TNF-α

tumor necrosis factor-α

TUNEL

terminal deoxynucleotidyl transferase dUTP nick end labelling

WT

wild-type