The environmental pollutant, polychlorinated biphenyl 126, alters liver function in a rodent model of Alcohol-associated Liver Disease

Abstract

Background:

The prevalence of alcohol-associated liver disease (ALD), a subtype of fatty liver disease (FLD), continues to rise. ALD is a major cause of preventable death. Polychlorinated biphenyl (PCB) 126 is an environmentally relevant, dioxin-like pollutant whose negative metabolic effects have been well documented. In human and animal studies, PCB has been associated with the severity of nonalcoholic fatty liver disease (NAFLD). However, few studies have investigated whether exposures to environmental toxicants can worsen ALD. Thus, the objective of the current study was to develop an alcohol-plus-toxicant model to study how an environmental pollutant, PCB 126, impacts rodent ALD pathology.

Methods:

Briefly, male C57BL/6J mice were exposed to 0.2 mg/kg PCB 126 or corn oil vehicle four days prior to ethanol feeding using the chronic-binge (10-plus-one) model.

Results:

Concentrations of macromolecules, including hepatic lipids, carbohydrates, and protein (albumin) were impacted. Exposure to PCB 126 exacerbated hepatic steatosis and hepatomegaly in mice exposed to the chemical and fed an ethanol diet. Gene expression and the analysis of blood chemistry showed a potential net increase and retention of hepatic lipids and reductions in lipid oxidation and clearance capabilities. Depletion of glycogen and glucose was evident, which contributes to disease progression by generating systemic malnutrition. Granulocytic immune infiltrates were present but driven solely by ethanol feeding. Hepatic albumin gene expression and plasma levels were decreased by ~50% indicating a potential compromise of liver function. Finally, gene expression analyses indicated that the aryl hydrocarbon receptor and constitutive androstane receptor were activated by PCB 126 and ethanol, respectively.

Conclusions:

Various environmental toxicants are known to modify or enhance FLD in high-fat diet models. Findings from the present study suggest that they interact with other lifestyle factors such as alcohol consumption to reprogram intermediary metabolism resulting in exacerbated ethanol-associated systemic malnutrition in ALD.

INTRODUCTION

Fatty liver disease (FLD) is a clinical pathological state of the liver in which excess lipids are stored, and this may be accompanied by inflammation and fibrosis. FLD exists along a spectrum that often begins with simple steatosis which may progress to inflammation (steatohepatitis) and if not treated, can further result in tissue scarring (fibrosis/cirrhosis) and fulminant liver failure (Schlageter et al., 2014). While this pathological spectrum has been extensively studied and is well-defined, etiological differences have been identified to help characterize the onset of FLD. For instance, cases of non-alcoholic fatty liver disease (NAFLD) have been attributed to excess diet or poor nutritional caloric intake that may be comorbid with other cardiometabolic diseases (Rosato et al., 2019). NAFLD is the most common chronic liver disease in developed countries affecting roughly 25% of adults in the United States and has a global prevalence estimated at 24% (Glass et al., 2019; Younossi et al., 2018). In contrast, alcohol-associated liver disease (ALD) which also manifests through the FLD spectrum is caused by excessive alcohol consumption (Crabb et al., 2020). ALD is a progressive disease that persists with excess alcohol intake and is characterized by hepatic steatosis, disrupted energy metabolism, oxidative stress, and leaky gut (Louvet and Mathurin, 2015). Death from excessive alcohol use and ALD is also one of the most common forms of preventable deaths in the United States (Axley et al., 2019; Witkiewitz et al., 2019). According to the World Health Organization (WHO), alcohol-related deaths accumulated to 5.3% of all global deaths in 2016 (WHO, 2018). In addition to NAFLD and ALD, toxicant-associated steatohepatitis (TASH) and toxicant-associated fatty liver disease (TAFLD) are other subtypes of FLD that result from exposures to environmental chemicals either in the surrounding environment, occupational settings, or accidental exposures (Al-Eryani et al., 2015; Wahlang et al., 2013). These environmental chemicals include distinct categories of contaminants such as persistent organic pollutants (POPs) like polychlorinated biphenyls (PCBs). Using epidemiologic data and experimental animal models, several studies have demonstrated that exposure to PCBs are associated with liver injury, steatohepatitis, and obesity (Cave et al., 2022; Clair et al., 2018; Deng et al., 2019). Irrespective of etiology, each subtype of FLD can further progress to more severe pathological states which are dictated by other factors including sex, diet, alcohol or tobacco use, and genetic susceptibility (Crabb et al., 2020; Younossi et al., 2018).

PCBs are environmental toxicants that were initially manufactured exclusively by Monsanto Corporation in the United States from the 1930’s until their subsequent ban in the late 1970’s (Markowitz and Rosner, 2018). Based on the chlorine atom substitution on the biphenyl ring, there are up to 209 PCB congeners; many of these congeners were produced and sold as mixtures under trade name ‘Aroclor’ (Clair et al., 2018). These chemicals were commercially used for electrical equipment, motor oil, insulators, plastics, and adhesives (EPA, 2022). Poor hazardous waste management coupled with PCBs’ resistance to breakdown resulted in environmental pollution and human exposures (Cusack et al., 2020). PCBs have long half-lives and are not readily metabolized, hence popularly labeled as “forever chemicals.” One congener, 3,3',4,4',5-Pentachlorobiphenyl (PCB 126) is a dioxin-like, coplanar PCB that has demonstrated carcinogenesis and organ toxicity (NTP, 2006). Specifically, mixtures and individual PCB congeners cause a wide range of adverse health effects across the body including endocrine disruption and reproductive, cardiovascular, and importantly, liver toxicity (Hardesty et al., 2018; Mesnage et al., 2018; Petriello et al., 2018; Shirota et al., 2006). Previous toxicological studies have also shown PCB 126’s ability to enhance diet-induced non-alcoholic steatohepatitis (NASH) and cardiometabolic disease (Deng et al., 2019; Petriello et al., 2016; Wahlang et al., 2017). Importantly, PCB 126 is a metabolism disrupting chemical (MDC) and potent aryl hydrocarbon receptor activator, and studies have demonstrated that Ahr activation is likely a primary driver of toxicity and metabolism disruption (Shi et al., 2019a; Shi et al., 2019b; Wang et al., 2019; Zhang et al., 2012b; Heindel et al., 2017). Some PCBs are classified as obesogens and postulated mechanisms include receptor-based modes of action and epigenetic changes leading to altered adipocyte differentiation and oxidative stress (Ghosh et al., 2014; Klinge et al., 2021). Furthermore, many PCBs are also considered an endocrine disrupting chemical (EDC), which may disrupt or influence endogenous hormone activity (Heindel and Blumberg, 2019). While some PCBs are considered obesogens, dioxin-like PCBs at high doses, can cause wasting syndrome like lipodystrophy, causing redistribution of lipids from adipose stores to accumulate in the liver.

To date, there are few reports that study the complex interaction of environmental chemical exposure and alcohol consumption in FLD development. We postulate that a complex interaction occurs between environmental toxicants and ethanol to enhance toxicity that may be dependent on dose, duration, sequence of exposure, sex, and other factors. This theory resides in the fact that the liver is a primary target of xenobiotic toxicity where both alcohol and PCBs have demonstrated negative health outcomes. Currently, there is limited knowledge on how environmental pollutant exposures may enhance hepatic disease induced by alcohol consumption. It is becoming increasingly apparent that lifestyle factors such as alcohol or tobacco use coupled with chemical exposure may impact not only hepatic health and function, but also contribute to worsening disease prognosis (Al-Dayyat et al., 2018; Bailey et al., 2009; Nivukoski et al., 2020; Wahlang et al., 2019). The objective of the current study is to characterize a novel, alcohol-plus-toxicant model to understand if PCB 126 exposure can disrupt energy metabolism to enhance liver disease endpoints caused by ethanol feeding. Therefore, this study tests the hypothesis that PCB 126 exposure enhances ALD in an animal model.

MATERIALS AND METHODS

Animals, Diets, and Chemical Use

The University of Louisville Institutional Animal Care and Use Committee (IACUC) approved animal use protocols and procedures. Sixty, ten-week-old male C57BL/6J mice were purchased from Jackson Laboratory (strain#: 000664; Bar Harbor, Maine). Five mice per cage were housed in a pathogen-free, temperature controlled (23.9°C) room with a twelve-hour light-dark cycle that is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). Mice were allowed to acclimate from facility transfer for one week and fed autoclaved laboratory rodent chow diet containing 28.7%, 13.1%, and 58.2% kcal/g of protein, fat, and carbohydrates, respectively (5010; LabDiet, St. Louis, Missouri) and had ad libitum access to food and water. 3,3',4,4',5-Pentachlorobiphenyl (PCB 126) was purchased from AccuStandard (catalog: C-126N-5MG; New Haven, CT). Corn oil vehicle was purchased from Sigma Aldrich (catalog: C8267-500ML). Lieber-DeCarli ’82 Shake and Pour ethanol (catalog: F1258SP) and control (catalog: F1259SP) diets were purchased from Bio-Serv (Flemington, NJ). Caloric profile of the ethanol diet contains 355, 135, 359, and 151 kcal/L of ethanol, carbohydrates, fat, and protein, respectively. Caloric profile of the control diet contains 490, 359, and 151 kcal/L of carbohydrates, fat, and protein, respectively. 50 mL glass feeding tubes (catalog: 9019) were employed to facilitate ethanol and control diet feeding within rodent cages and were also purchased from Bio-Serv (Flemington, NJ).

Study and Exposure Paradigm

The commonly used chronic-binge (ten-plus-one) rodent alcohol model was employed as a foundational model for the duration of study (Bertola et al., 2013). Rodent cages were divided into four groups: pair-fed + vehicle, pair-fed + PCB 126, ethanol + vehicle, and ethanol + PCB 126. Mice were orally gavaged with either corn oil vehicle or 0.2 mg/kg rodent body weight of PCB 126 four days prior to ramp-up ethanol feeding. This dose was selected as a median exposure level based on doses that our research group had previously used to elicit toxicological hepatic effects (Jin et al., 2021; Shi et al., 2019a; Shi et al., 2019b). Two, 50 mL feeding tubes were placed in each rodent cage during ethanol- or pair-feeding. Over the course of five days, ethanol-fed mice were provided the diet in increasing concentrations of ethanol to acclimate them to the liquid feeding. Pair-fed mice were fed an isocaloric diet where ethanol was substituted with maltose dextrin and fed as described in Bertola et al. (Bertola et al., 2013). Ethanol-fed mice were then maintained on 5.0% ethanol diet for ten days. On the day of the binge, ethanol-fed mice were removed from the housing room at ZT 22 and brought to the euthanasia and tissue collection room. Mice were then administered 5 g/kg ethanol binge beginning at ZT 23 approximately every 8 minutes. Pair-fed mice were orally gavaged nano-pure water instead of the established maltose dextrin binge to avoid carbohydrate intake differences between the two dietary groups. It has previously been shown that maltose dextrin may cause distinct metabolic abnormalities between experimental groups (Soto et al., 2017). Thus, we wanted to control for these differences as ethanol is broken down into acetate and maltose dextrin is broken down into various saccharide molecules. Euthanasia and tissue collection started approximately at ZT 3. The chronic-binge model (Bertola et al., 2013) recommends nine hours post binge for tissue harvest; however, several mice (n=7) in the ethanol + vehicle group experienced premature mortality three hours post binge. To avoid quantitative differences in data within this group, the remaining experimental and control animals were euthanized at a similar time point which was equivalent to approximately three hours post binge. We have reviewed and accounted for the premature mortality in our statistical analysis and no differences between groups were observed. Besides the premature mortality experienced three hours post-binge, one mouse experienced mortality one day prior to planned euthanasia. Body weights were recorded daily during ramp-up ethanol feeding (0% - 4%) and 5% ethanol feeding. Blood collection from the inferior vena cava was performed followed by euthanasia via exsanguination and tissue collection.

Liver Histology

Liver tissues were fixed in 10% neutral buffered formalin for at least 48 hours and then placed in 75% ethanol until tissue processing and paraffin embedding. Tissues were sectioned at 5 μm with Leica Biosystem’s Histocore Autocut Automated Rotary Microtome (Leica Biosystem; Deer Park, IL). Tissue sections were stained with hematoxylin (catalog: FS72804; Epredia; Kalamazoo, MI) and eosin (catalog: HT110316-500ML; Sigma Aldrich; St. Louis, MO) (H&E) to assess hepatic morphology. Liver sections were also stained with Naphthol AS-D Chloroacetate-specific esterase purchased from Sigma Aldrich to assess granulocytic leukocyte infiltrates (catalog:91c-1kt; St. Louis, MO). Images were captured using an Aperio GT 450 – Automated, high-capacity pathology slide scanner (Leica Biosystems; Deer Park, IL). H&E and CAE images were captured on Aperio ImageScope software at 10 and 20X magnification, respectively (v12.4.3.5008) (Leica Biosystems Pathology Imaging; Deer Park, IL). CAE scoring was performed where total granulocytes were counted in ten 20X fields of view for five animals per group. Finally, liver sections were also stained according to manufacturer protocol with periodic acid (catalog: P7875-25g) and Schiff reagent (catalog: 3952016-500mL) to determine glycogen content (Sigma Aldrich; St. Louis, MO). All histological staining was performed according to manufacturer’s protocol. PAS images were captured at 10X using an Olympus BX43 microscope, DP74 digital camera, and CellSens software package (Olympus America, Breinigsville, PA, USA). Quantification for positively stained Periodic Acid Schiff (PAS) regions was performed using ImageJ (v1.53k) software (National Institute of Health; Bethesda, MD). To quantify glycogen content, we analyzed up to seven fields of view of all hepatic sections at 10X magnification and percent area was averaged respective to grouping. Finally, Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was performed to detect hepatocytes undergoing apoptosis. TUNEL staining was accomplished using Apoptag Peroxidase In Situ Apoptosis Detection kit from Sigma Aldrich (catalog: S7100; St. Louis, MO). TUNEL staining was performed according to manufacturer’s protocol and up to ten fields of view for each animal at 20X magnification were used to quantify the relative abundance of apoptotic cells. TUNEL positive cell counts were normalized by dividing the number of TUNEL positive cells by the number of fields of view.

Blood Chemistry Analyses

Liver enzymes, triglycerides, glucose, cholesterol, and lipoproteins (HDL, LDL, VLDL) were measured with Lipid Panel Plus diskettes (catalog:400-0030; Abaxis Inc.; Union City, CA) on the Piccolo Xpress Chemistry Analyzer (Abbott; Abbott Park, IL). Plasma adipokines were measured with the Milliplex Map Mouse Adipokine Magnetic Bead Panels (catalog: MADKMAG-71K; EMD Millipore) on a Luminex^® 100 system (Luminex Corp.; Austin, TX). Adipokines measured included Insulin, PAI-1, Resistin, and Leptin. Analysis for each assessment was performed according to the manufacturer’s protocol. Plasma albumin levels were measured via ELISA from MyBioSource Inc. and performed according to the manufacturer’s standard protocol (catalog: MBS564063; San Diego, CA).

Hepatic Triglyceride and Cholesterol Measurement

Hepatic triglycerides and cholesterol were extracted based on the Bligh and Dyer method utilizing a 2:1 chloroform:methanol solution (Bligh and Dyer, 1959). Triglyceride and cholesterol standards were used to generate a standard curve to quantify extracted lipids (catalog: T7531-STD, C7509-STD; Point Scientific; Canton, MI). Extracted lipids were colorimetrically measured with a microplate absorbance reader (BioTek Gen 5; Winsooki, VT).

qRT-PCR Analysis

Liver and ileum tissues were processed and homogenized in RNA-STAT60, and total RNA was extracted (catalog: CS-502; Tel-test Inc.; Friendswood, TX). RNA quantity and purity was assessed with a Nanodrop One^C spectrometer purchased from ThermoFisher Scientific (catalog: 701-058112; Madison, MI). cDNA was reverse transcribed from 3 μg RNA to yield 60 μl of cDNA using single step cDNA synthesis reagent, QScript (catalog: 95048-500; Quantabio; Beverly, MA). qRT-PCR was performed on a CFX384^™ Real-Time System (Bio-Rad; Hercules, CA). iTaq Universal Probe Supermix from Bio-Rad was used during qRT-PCR set up (catalog: 1725134; Hercules, CA). Relative mRNA expression was calculated based on the 2^−ΔΔCt method with Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH) as the housekeeping gene for liver tissues and 18S rRNA for ileal tissues (catalog: 4351309; catalog:4319413E; Applied Biosystems; Waltham, MA). Fold induction was calculated and normalized to the control group, pair-fed + vehicle, which was set to 1. All genes of interest (TaqMan probes) were purchased from ThermoFisher Scientific and are cataloged in Supporting Table S1.

Statistical Analysis

Statistical significance was based on two-way analysis of variance where our two factors were diet (pair-fed v. ethanol-fed) and exposure (vehicle v. PCB 126) using GraphPad Prism (v9.2) for Windows unless otherwise noted (GraphPad Software Inc.; La Jolla, CA). Alpha level of 0.05 (P ≤ 0.05) was used to determine significance. Tukey’s post-hoc test was used to compare individual groups for multiple comparison analyses. A P-value table is supplemented with all figures which displays the two-way ANOVA and Tukey’s multiple comparisons results to designate significance between our groups. Repeated measures ANOVA test with generalized linear models were used to analyze body weight changes over study days (within-subjects) or irrespective of study days (between-subjects) using SAS (v9.4). Repeated measures ANOVA results are listed in Supporting Tables S2 and S3 for an alpha level set to 0.05.

RESULTS

Ethanol and PCB 126 alter Organ and Body Composition

Diet consumption was recorded daily, and a pair-feeding figure is provided in Supporting Figure S1. No major diet intake differences were observed between dietary groups or due to PCB 126 exposure. Initial characterization of the mouse phenotype post PCB 126 exposure and ethanol feeding was performed using anthropometric measurements (Figure 1). Animal body weight changes were evaluated because the body weight of ethanol-fed mice may initially change during acclimatization to the ethanol diet then stabilize for the duration of the study. Body weight changes were stable in the present study during ethanol acclimatization and feeding, irrespective of PCB 126 exposure (Figure 1A). A repeated measures ANOVA (RM-ANOVA) was performed to test animal body weight changes over time and if weight changes were an effect due to dietary feeding, exposure, or the combination of feeding and exposure and are listed in Supporting Tables S2 and S3. It was observed that there was a change in mean weight over time (within-subjects) dependent on day, dietary feeding, PCB 126 exposure, and combination of day, diet, and PCB 126. We also analyzed body weight changes in respect to between-subjects (ignoring time effects) and observed no significant differences between pair-fed v EtOH-fed, vehicle v PCB 126, or interacting diet and exposure. Our cumulative results indicate that body weight changes are dependent on time as a factor of day, exposure, and/or diet and not solely due to the feeding or exposure paradigm. Liver-to-body weight ratio was significantly increased due to PCB 126 exposure in pair-fed mice (P=0.0004). Ethanol feeding alone significantly increased liver-to-body weight ratio, which was further enhanced in the ethanol + PCB 126 group to account for ~6% of body weight (Figure 1B). Quantified hepatic triglycerides were significantly increased due to ethanol feeding (P<0.0001) and PCB 126 exposure (P<0.0001) (Figure 1C). Importantly, hepatic triglycerides were further increased in the ethanol + PCB 126 group (74.5 ± 11.5 mg/g liver weight), which is ~1.4-fold higher than the ethanol + vehicle group (53.1 ± 9.1 mg/g liver weight) (Figure 1C). Furthermore, PCB 126 exposure led to decreased white adipose-to-body weight ratios by approximately 20% in both feeding groups (Figure 1D).

Figure 1 – PCB 126 and ethanol effects on body composition and hepatic triglycerides.

(A) Body weights were recorded daily during acclimatization to the alcohol diet and the duration of study. (B) After extracting the gallbladder, whole liver weight was recorded. Relative liver weight was calculated based on body weight giving a relative percent of liver to body weight. (C) Hepatic triglycerides were isolated from liver tissue and were quantified using a colorimetric assay. (D) After isolating the white adipose tissue, it was weighed, and relative white adipose weight was calculated based on body weight to provide a relative percent of white adipose to body weight. Values are represented as mean ± SD with an alpha level set to 0.05. A complete list of P-values, as determined by two-way ANOVA and Tukey’s post-hoc test, is provided in the accompanying table. EtOH, ethanol; Veh., vehicle; TGs, triglycerides; WAT, white adipose tissue.

PCB 126 exacerbated Ethanol-induced Steatosis

Histological staining (H&E) of liver sections demonstrated mild lipid accumulation due to PCB 126 exposure in the pair-fed group, while all ethanol-fed mice developed varying degree of steatosis (Figure 2A). Importantly, the ethanol + PCB 126 group displayed enhanced steatosis relative to the ethanol + vehicle group (Figure 2A). These results provide evidence for increased lipid accumulation in the liver, indicating developing steatosis.

Figure 2 – PCB 126 and ethanol effects on liver morphology.

(A) 5μm thin liver sections were analyzed using H&E staining to assess liver morphology and the presence of lipid droplet accumulation. Images were captured at 10X magnification. EtOH, ethanol.

Ethanol-feeding primarily drives Immune Response and Injury

Histological analysis of liver sections by CAE staining indicated granulocyte infiltration responding to stress in the experimental groups (Figure 3A). All ethanol-fed mice had varying degrees of positively stained granulocytes in liver tissue. CAE scoring revealed ~2.2-fold increase in granulocytes due to ethanol feeding (Figure 3B). Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were not significantly different, likely due to the discussed study limitations and are listed in Supporting Figure S2. Additionally, cytokines/chemokines involved in inflammation and granulocyte trafficking namely tumor necrosis factor-alpha (TNFα), C-X-C Motif Chemokine Ligand 2 (Cxcl2), and C-C Motif Chemokine Ligand 2 (Ccl2) were examined. However, there were no differences in gene expression for Cxcl2, and only a mildly significant increase of expression for Ccl2 due to ethanol feeding (Supporting Figure S2). Tnfα gene expression was increased in mice exposed to PCB 126; however, its expression was contradictorily decreased with ethanol feeding (Supporting Figure S2). Furthermore, we observed that ethanol feeding was also the primary driver of hepatocellular apoptosis by TUNEL staining and is listed in Supporting Figure S3. TUNEL positive cells were counted to be relatively two times higher than pair-fed controls. Markers of endoplasm reticulum (ER) stress and apoptosis was also evaluated by gene expressional changes and are listed in Supporting Figure S3. DNA Damage Inducible Transcript 3 (Chop) is a gene activated in response to ER stress to induce apoptosis was observed to be increased by ethanol feeding and PCB 126 exposure (Supporting Figure S3). Analysis of Chop mRNA levels also indicate an interaction effect where the combination of ethanol feeding and PCB 126 exposure increased its abundance up to ~13-fold. Activating Transcription Factor 3 and 4 (Atf3, Atf4) expression were also evaluated for their role in cellular and ER stress and are listed in Supporting Figure S3. Atf3 expression was significantly increased by ethanol feeding and PCB 126 exposure and robustly increased ~46-fold due to interacting ethanol and PCB 126 exposure. Atf4 mRNA levels were significantly increased due to ethanol feeding but, interestingly, relative mRNA levels were slightly mitigated in the ethanol + PCB 126 group.

Figure 3 – PCB 126 and ethanol effects on hepatic inflammation and Injury.

(A) 5μm thin liver sections were analyzed using CAE staining to detect the presence of granulocyte infiltration. (B) Total granulocytes were counted in ten fields of view at 20X magnification for five animals per group. Counted granulocytes were then graphed in the accompanied scatterplot and statistical analysis was performed. Values are represented as mean ± SD with an alpha level set to 0.05. A complete list of P-values, as determined by two-way ANOVA and Tukey’s post-hoc test, is provided in the accompanying table. EtOH, ethanol; Veh., vehicle; CAE, chloroacetate esterase.

PCB 126 and Ethanol alter Plasma Lipids, Adipokines, Glucose, Insulin, and Albumin

Evaluating the abundance of plasma lipids, adipokines, glucose, insulin, and albumin are important to assess the extent of altered liver function imposed by ethanol feeding or PCB 126 exposure and are listed in Table 1. Plasma triglycerides were overall significantly increased ~1.7-fold in the ethanol-fed groups compared to pair-fed groups. Interestingly, the ethanol + PCB 126 group (42.6 ± 9.8 mg/dL) showed attenuation of plasma triglyceride levels compared to the ethanol-fed group but were still significantly elevated relative to the pair-fed + PCB 126 group (26.8 ± 1.9 mg/dL). Plasma cholesterol, high density lipoprotein (HDL), and low density lipoprotein (LDL) were significantly decreased upon PCB 126 exposure in pair-fed mice and upon ethanol feeding in the ethanol + vehicle group. Plasma cholesterol, HDL, and LDL were decreased in the ethanol + PCB 126 but were not statistically significant. Plasma very low density lipoprotein (VLDL) was significantly elevated in the ethanol-fed mice, but VLDL levels were attenuated in the ethanol + PCB 126 group (8.5 ± 2.0 mg/dL), although still significantly higher than pair-fed mice (5.3 ± 0.5 mg/dL). PCB 126 exposure significantly decreased Leptin and Resistin levels; however, ethanol only increased Leptin values which were then mitigated in the ethanol + PCB 126 group. Plasma levels of plasminogen activating inhibitor 1 (PAI-1) were significantly elevated in the ethanol-fed mice but were not different when accounting for PCB 126 exposure, indicating that ethanol feeding may be the dominate driver of this injury marker. Glucose was found to be significantly decreased due to PCB 126 and ethanol feeding; notably, there was an interaction in the ethanol + PCB 126 group where plasma glucose levels were further decreased by about 22%. Insulin on the other hand, was significantly increased due to ethanol feeding but levels were mitigated due to PCB 126 exposure (197.9 ± 68.3 pg/mL) which is concordant with our previous observations in rodent and human studies (Clair et al., 2018; Wahlang et al., 2014b). Plasma albumin was significantly decreased by ethanol feeding and PCB 126 exposure. Importantly, the mean albumin concentration in the pair-fed + vehicle group was decreased ~50% from 2.75 g/dL to 1.36 g/dL in the ethanol + PCB 126 group.

Table 1 -. Plasma lipids, adipokines, glucose, insulin, and albumin.

Values are described as mean ± SD for an alpha level of 0.05. A complete list of P-values, as determined by two-way ANOVA and Tukey’s post-hoc test, is provided in the accompanying table. EtOH, ethanol; Veh., vehicle; TGs, triglycerides; CHOL, cholesterol.

Analyte	Pair-Fed		EtOH-Fed
	Vehicle	PCB 126	Vehicle	PCB 126
Triglycerides (mg/dL)	32.5 ± 5.7	26.8 ± 1.9	68.7 ± 24.3	42.6 ± 9.8
CHOL (mg/dL)	73.8 ± 15.3	51.9 ± 6.4	56.7 ± 12.5	44.1 ± 8.6
HDL (mg/dL)	56.4 ± 13.3	42.0 ± 3.3	42.2 ± 12.5	31.7 ± 8.3
LDL (mg/dL)	11.0 ± 4.4	6.25 ± 1.7	4.3 ± 1.7	4.8 ± 3.3
VLDL (mg/dL)	6.6 ± 1.2	5.3 ± 0.5	12.2 ± 3.6	8.5 ± 2.0
Leptin (pg/mL)	4350.5 ± 3957.9	1014.7 ± 647.3	6695.9 ± 3769.0	4104.4 ± 2769.5
Resistin (pg/mL)	1296.6 ± 319.5	909.1 ± 318.7	1236.9 ± 447.8	1129.4 ± 337.3
PAI-1 (pg/mL)	1276.1 ± 554.1	1708.8 ± 609.1	2712.8 ± 1056.9	3332.8 ± 1464.4
Glucose (mg/dL)	241.7 ± 59.5	154.2 ± 47.0	123.7 ± 46.3	96.8 ± 26.9
Insulin (pg/mL)	139.3 ± 54.1	108.2 ± 29.9	376.8 ± 172.1	197.9 ± 68.3
Albumin (g/dL)	2.75 ± 0.69	2.02 ± 0.38	1.62 ± 0.72	1.36 ± 0.41

Outcome	EtOH	PCB 126	Interaction	Pair-fed (Veh. vs. PCB 126)	EtOH-fed (Veh. vs. PCB 126)	Veh. (Pair-fed vs. EtOH-fed)	PCB 126 (Pair-fed vs. EtOH-fed)
TGs	<0.0001	<0.0001	0.0064	n.s.	0.0001	<0.0001	0.0051
CHOL	0.0012	<0.0001	n.s.	<0.0001	n.s.	0.0207	n.s.
HDL	0.0006	0.0005	n.s.	0.0056	n.s.	0.0363	n.s.
LDL	0.0038	n.s.	n.s.	0.0068	n.s.	0.0225	n.s.
VLDL	<0.0001	0.0002	0.0483	n.s.	0.0016	<0.0001	0.0007
Leptin	0.0044	0.0021	n.s.	0.0333	n.s.	n.s.	n.s.
Resistin	n.s.	0.0190	n.s.	0.0259	n.s.	n.s.	n.s.
PAI-1	<0.0001	n.s.	n.s.	n.s.	n.s.	0.0129	0.0009
Glucose	<0.0001	0.0004	0.0471	<0.0001	n.s.	<0.0001	0.0139
Insulin	<0.0001	0.0001	0.0046	n.s.	0.0001	<0.0001	0.0367
Albumin	<0.0001	0.0041	n.s.	0.0072	n.s.	0.0002	0.0177

Ethanol and PCB 126 alter Albumin and Lipid Metabolism gene expression

First, we wanted to briefly evaluate how our exposure paradigm altered albumin’s expression, which its related protein is vital to whole organism homeostasis and macromolecule transport. Albumin (Alb) gene expression was decreased by ethanol feeding and PCB 126 exposure and ~50% in the ethanol + PCB 126 group (Figure 4A). Transporter genes were first examined in our array of lipid-related genes important to this characterization study. The expression of Cd36, fatty acid translocase, was upregulated due to PCB 126 exposure and ethanol feeding which was further increased in the ethanol + PCB 126 group about 6-fold (Figure 4B). The expression of hepatic fatty acid binding protein, Fabp1, a cytoplasmic lipid chaperone protein, was downregulated 5-fold due to ethanol feeding (Figure 4C). Further, genes involved in lipid synthesis were assessed. Fatty acid synthase (Fasn) expression was downregulated 2-fold in both diet groups due to PCB 126 exposure (Figure 4D). Scd-1, a gene involved in synthesizing monounsaturated fatty acids, exhibited decreased expression due to PCB 126 in pair-fed mice and ~60% decreased due to ethanol feeding (Figure 4E). Additionally, Srebf1, a gene involved in the regulation of sterol synthesis was examined. Srebf1 mRNA level was downregulated in the ethanol + PCB 126 group and displayed an interaction effect where its expression was decreased ~40% (Figure 4F). We also evaluated genes involved in lipoprotein assembly and export. Apolipoprotein B (Apob) is the primary gene encoding for the protein involved in carrying and exporting LDL and VLDL from the liver. Apob expression was observed to be increased due to PCB 126 exposure in pair-fed mice; however, an interaction effect was noted in the ethanol + PCB 126 where Apob expression was decreased ~15% (Figure 4G). mRNA levels of microsomal triglyceride transfer protein (Mttp), which engages in apolipoprotein assembly, were slightly increased due to PCB 126 exposure and ethanol feeding (Supporting Figure S4). In addition, genes involved in lipid breakdown were also investigated. Cpt-1α and Cpt-2 are genes that encode fatty acid oxidation transferases that are located at the outer and inner membrane of the mitochondria, respectively. Cpt-1α expression was discovered to be significantly increased due to PCB 126 exposure in pair-fed mice; however, its expression was decreased about 50% due to ethanol feeding in both ethanol-fed groups (Figure 4H). Similarly, Cpt-2 expression was downregulated in response to ethanol feeding, highlighting decreased lipid oxidation capabilities when mice are consuming an ethanol diet (Supporting Figure S4).

Figure 4 – Hepatic qRT-PCR analysis of albumin and lipid metabolism related genes.

qRT-PCR was performed to measure hepatic mRNA levels of (A) Alb (B) Cd36, (C) Fabp1, (D) Fasn, (E) Scd-1, (F) Srebf1, (G) Apob, and (H) Cpt-1α. Values are represented as mean ± SD with an alpha level set to 0.05. A complete list of P-values, as determined by two-way ANOVA and Tukey’s post-hoc test, is provided in the accompanying table. EtOH, ethanol; Veh., vehicle; Alb, albumin; Cd36, fatty acid translocase; Fabp1, fatty acid binding protein; Fasn, fatty acid synthase; Scd-1, stearoyl-CoA desaturase 1; Srebf1, sterol regulatory element binding transcription factor 1; Apob, apolipoprotein b; Cpt-1α, carnitine palmitoyltransferase 1α.

Ethanol and PCB 126 deplete Glycogen stores and disrupts Carbohydrate Metabolism

Characterization of additional energy metabolism pathways is important since disrupted metabolism, including carbohydrate metabolism, is often manifested in FLD. It was hypothesized that this model would also impair carbohydrate stores and activity as both ethanol and PCB 126 have independently demonstrated previously (Steiner et al., 2015; Wahlang et al., 2017; Zhang et al., 2012b). PAS staining was performed to assess glycogen content, and captured images are displayed in Figure 5A. Quantitative assessment demonstrated that glycogen storage was heavily depleted due to PCB 126 exposure in pair-fed mice and due to ethanol in alcohol fed mice (Figure 5B). The relative abundance of positive staining was found to be decreased about 60% due to PCB 126 exposure and ethanol feeding. A statistical interaction effect was observed in the ethanol + PCB 126 group where glycogen levels were depleted by an additional 10% relative to the pair-fed + vehicle group (Figure 5B). In addition, hepatic expression of various genes involved in glycolysis and gluconeogenesis were evaluated to further identify key mediators that were altered by PCB 126 and ethanol feeding. Pyruvate kinase (Pklr), the rate-limiting enzyme in glycolysis, expression was decreased due to PCB 126 exposure and ethanol feeding and our statistical analysis indicated an interaction in the ethanol + PCB 126 group where expression was decreased by ~50% (Figure 5C). Phosphoenolpyruvate carboxykinase 1 (Pck-1), a gene regulating gluconeogenesis, was downregulated due to ethanol feeding and PCB 126 exposure (Figure 5D). Interestingly, Pck-1 expression was downregulated ~40% due to interacting PCB 126 and ethanol (Figure 5D). Expression of Glucose-6-phosphatase (G6pc), which catalyzes the last step of gluconeogenesis, was overall downregulated ~70% in response to ethanol feeding (Figure 5E). We also evaluated glycogen synthase (Gys2) mRNA levels and observed a ~60% decrease of expression due to ethanol feeding (Figure 5F). Interestingly, Gys2 expression was moderately increased due to PCB 126 exposure and an interaction effect was observed. Finally, glycogen phosphorylase (Pygl) expression was unchanged due to ethanol feeding or PCB 126 exposure (Figure 5G).

Figure 5 – Analysis of glycogen content and carbohydrate-related gene expression.

(A) 5μm thin liver sections were analyzed using PAS staining to detect glycogen presence in hepatic tissue at 10x magnification. (B) PAS stain scoring was performed using ImageJ and is displayed in the accompanying figure. qRT-PCR was performed to measure hepatic mRNA levels of (C) Pklr, (D) Pck-1, (E) G6pc, (F) Gys2, and (G) Pygl. Values are represented as mean ± SD with an alpha level set to 0.05. A complete list of P-values, as determined by two-way ANOVA and Tukey’s post-hoc test, is provided in the accompanying table. EtOH, ethanol; Veh., vehicle; PAS, Periodic Acid Schiff; Pklr, pyruvate kinase L/R; Pck-1, phosphoenolpyruvate carboxykinase 1; G6pc, glucose-6-phosphatase catalytic subunit 1; Gys2, glycogen synthase; Pygl, glycogen phosphorylase.

Ethanol and PCB 126 altered Hepatic Receptor Expression and Activation

Ethanol has been shown to perturb hepatic receptor activity such as peroxisomal-proliferator activated receptor’s (Ppar’s), constitutive androstane receptor (Car), and Ahr (Gyamfi and Wan, 2010; Zhang et al., 2012a). In addition, it is well-established that PCBs interact with Ahr and other xenobiotic receptors such as Pregnane X Receptor (Pxr) in a dose-dependent manner (Wahlang et al., 2014a; Zhang et al., 2012b). The expression of three hepatic receptors and respective target genes of interest were evaluated in Figure 6. Ahr expression was significantly increased due to PCB 126 exposure and ethanol feeding while its target gene, Cyp1a2, was induced 30-fold due to PCB 126 exposure in both feeding groups (Figure 6A). Car expression was significantly elevated due to PCB 126 exposure in pair-fed mice but was mitigated in the ethanol + PCB 126 mice (Figure 6B). Cyp2b10, a Car target, expression was found to be significantly induced (~350-fold) due to ethanol feeding alone, highlighting Car’s activity in response to ethanol (Figure 6B). Finally, Pparα’s expression was found to be decreased in response to PCB 126 exposure and ethanol feeding; however, the expression of its target gene, Cyp4a10, was not perturbed in this model (Figure 6C).

Figure 6 – Hepatic qRT-PCR analysis of hepatic receptors and target genes.

qRT-PCR was performed to measure hepatic mRNA levels of hepatic receptor Ahr and its target gene Cyp1a2 (A), Car and its target gene Cyp2b10 (B), Pparα and its target gene Cyp4a10 (C). Values are represented as mean ± SD with an alpha level set to 0.05. A complete list of P-values, as determined by two-way ANOVA and Tukey’s post-hoc test, is provided in the accompanying table. EtOH, ethanol; Veh., vehicle; Ahr, aryl hydrocarbon receptor; Cyp1a2, cytochrome P450 family 1, subfamily a, member 2; Car, constitutive androstane receptor; Cyp2b10, cytochrome P450 family 2, subfamily b, polypeptide 10; Pparα, Peroxisomal Proliferator Receptor α; Cyp4a10, cytochrome P450 family 4, subfamily a, polypeptide 10.

DISCUSSION

Although alcohol manufacture and use have thousands of years of history, the scientific understanding of its effects and abuse are relatively recent (Khaderi, 2019). Researchers have attempted to recapitulate human ALD in animal models; however, the full extent of injury and metabolic response is not apparent (Louvet and Mathurin, 2015). A secondary insult, such as high fat supplementation, is generally required to progress rodent ALD pathogenesis to study hepatic injury and metabolic disruption (Brandon-Warner et al., 2012). Toxicological effects of alcohol consumption are well known and characterized in experimental models and human populations (Beier and McClain, 2010; Plaza-Díaz et al., 2020). Similarly, the adverse effects of PCBs are extensively documented, and their mechanism of action have been postulated with high certainty (NTP, 2006; Deng et al., 2019; Wahlang et al., 2017). Importantly, studies have previously demonstrated that obesity and lifestyle factors can impact PCB bodily distribution and enhance disease prognosis (Domazet et al., 2020; Rey-Cadilhac et al., 2020; Wahlang et al., 2019). For instance, we have previously published that diet, a lifestyle factor, can modify toxicity imposed by environmental pollutants (Jin et al., 2020; Klinge et al., 2021). Overall, this model demonstrated an impact on several major macromolecules including lipids, carbohydrates, and proteins. The key findings from this study revealed that PCB 126 exposure and ethanol feeding altered energy metabolism and liver function to worsen ALD endpoints and are illustrated in Figure 7.

Figure 7 – Schematic diagram of the major findings in the present study.

A schematic diagram outlining the major findings from the present study in terms of hepatic or systemic changes where arrows indicate the increase (↑) or decrease (↓) of expression, processes, or marker levels. Red arrows indicate ethanol effects, blue arrows indicate PCB 126 effects, and green indicates an interaction effect (ethanol + PCB 126). EtOH, ethanol; Veh., vehicle; Cd36, fatty acid translocase; Fabp1, fatty acid binding protein; Apob, apolipoprotein b; Alb, albumin Pklr, pyruvate kinase L/R; Pck-1, phosphoenolpyruvate carboxykinase 1; G6pc, glucose-6-phosphatase catalytic subunit 1; Ahr, aryl hydrocarbon receptor; Car, constitutive androstane receptor.

A major finding in this study is that steatosis was exacerbated in mice exposed to PCB 126 that consumed the ethanol diet. Both PCB 126 and ethanol have independently shown to influence lipid storage and drive disease progression along the FLD spectrum to more severe states. Published literature suggests that ethanol promotes steatosis by increasing the ratio of NADH:NAD⁺ to impede β-oxidation pathways while also impairing Pparα activity, a major lipid metabolism regulator (Lieber, 2004; You and Arteel, 2019). Upon the development of steatosis, sustained alcohol use further disrupts energy metabolism and to promote pathology (Donohue, 2007). PCB 126 has also been reported to induce hepatic lipid accumulation and enhance diet-induced steatosis via transcriptional reprogramming and altering bodily distribution of lipids (NTP, 2006; Shi et al., 2019b). Contrasting hepatic triglycerides, plasma triglyceride levels were mitigated by PCB 126 in ethanol-fed mice, although still statistically higher than controls. PCBs are thought to impede hepatic triglyceride release via Apob by disrupting the expression of Mttp (Shan et al., 2020). Mttp plays a critical role in lipoprotein assembly; however, Apob’s mitigated expression in the ethanol + PCB 126 group may explain exacerbated steatosis in these animals. Furthermore, plasma VLDL levels were attenuated by PCB 126 in the ethanol-fed mice. These findings are consistent with previously reported PCB 126 results in a time course study where plasma triglycerides were elevated 36 hours post PCB 126 exposure followed by a gradual decline until the end of study (Gadupudi et al., 2016). PCB’s influence on lipid distribution may be an indication as to why hepatic lipid stores of our study mice were elevated but plasma triglycerides were mitigated at the euthanasia time point.

Previous toxicokinetic studies have alluded that PCBs, due to their high lipophilicity, are concentrated in adipose tissue and redistributed via plasma lipids to eventually access the liver and other target organs (ATSDR, 2000). While the mean body weights of our mice were stable during the study, white adipose tissue weight was decreased implying this phenomenon may have occurred in our study animals. Previous studies have implied that wasting syndrome may occur in chronic, high dose PCB exposures. Adipocyte lipolysis is a phenomenon that may have occurred in the current study; however, based on our observed upregulation of Cd36 and hepatic and plasma triglyceride data, its believed that PCB 126 may transcriptionally influence the liver to import more adipose-derived blood lipids to increase steatosis. Overall, at the transcriptional level, our data suggests a potential net accumulation of lipids where lipid import (Cd36) was increased, and lipid oxidation (Cpt-1α and Cpt-2) was decreased in the ethanol + PCB 126 group. These expressional changes coupled with increased hepatic triglycerides and liver-to-body weight ratio indicates exacerbated steatosis. Several mechanisms are known to promote steatosis and our data suggests this model recapitulates some of these mechanisms including increased lipid import, reduced lipid oxidation, and reduced lipid clearance (Nassir et al., 2015).

Malnutrition is a major factor in human ALD development where ethanol can directly disrupt glucose homeostasis (Steiner et al., 2015). Poor nutrition due to alcohol consumption may also be a result of the absence of protein, micronutrients, and vitamins (McClain et al., 2011). PAS scoring indicated that glycogen stores were depleted due to PCB 126 exposure and ethanol feeding accompanied by decreased gluconeogenic gene expression (G6pc and Pck-1) in the ethanol-fed groups. Glycogen synthase (Gys2) expression was also decreased due to ethanol feeding; however, the release of glycogen from storage via Pygl mRNA analysis was unperturbed. Additionally, Atf3 and Atf4 have demonstrated induction in the presence of glucose depravation and ER stress (Iurlaro et al., 2017; Ku and Cheng, 2020). It may be that carbohydrate metabolism was reprogrammed at the transcriptional level resulting in decreased glucose and glycogen levels observed in the present study to induce Atf3, Atf4, and Chop related apoptosis pathways. The disruption of these hepatic genes with the loss of hepatic glycogen implies that de novo glucose synthesis and storage is disrupted in the liver, but the exact mechanism and contribution to disease development is unclear. It was also noted that the gene expression of Pklr was downregulated in the PCB 126 exposed and ethanol-fed mice. Taken together, both gluconeogenesis and glycolysis were impeded, signifying disruption of carbohydrate metabolism and depleted glucose. Previous alcohol studies have suggested that ethanol consumption impairs both glycolysis and gluconeogenesis by impacting enzyme transcription and activity (Steiner et al., 2015; Young et al., 2006). With carbohydrate metabolism impaired, a shift from utilizing carbohydrates as a primary energy source to lipids occurs during excess ethanol consumption, which contributes to disease development (Berk et al., 2005). It should be noted that nutritional differences exist where pair-fed mice receive 490 kcal/L of carbohydrates and ethanol-fed mice receive 135 kcal/L of carbohydrates and 355 kcal/L ethanol. However, previous work has demonstrated that chronic ethanol feeding is responsible for decreased glycogen content rather than dietary carbohydrate intake differences (Van Horn and Cunningham, 1999). Nonetheless, ethanol-fed mice still exhibited decreased glucose levels and glycogen stores even though each dietary group were euthanized after the same amount of time post water or ethanol binge.

Albumin’s mRNA and protein levels were decreased ~50% due to ethanol and PCB 126 exposure, a finding that is a traditional hallmark of altered liver function and malnutrition in humans (Sun et al., 2019). In the present in vivo model, PCB 126 exposure additively decreased albumin with ethanol feeding. The observed albumin reduction appears to be a class effect for dioxin-like molecules as similar results have been published for the prototypical Ahr ligand, 2,3,7,8-tetrachlorodibenzo-p-dioxin, in mice fed a standard rodent diet (Nault et al., 2017). However, hepatic albumin protein abundance was increased in vivo in Ahr knockout mice fed a control diet in another secondary analysis of our previously published proteomics data (P=0.0006, data not shown) (Jin et al., 2021). While environmental dioxin exposures appear to regulate key liver functions such as albumin production in model systems, more mechanistic and confirmatory human data are required. We postulate this model’s toxic mode of action to be attributed to the alteration of liver functionality to induce a metabolic nutritionally deficit state (loss of carbohydrates, reduced lipid oxidation and export, and reduced albumin levels) to overall promote disease status.

Investigating hepatic receptor activation is imperative for identifying the major mode(s) of action of PCB 126 and ethanol and how they elicited some of the observed altered molecular processes. PCB 126 is considered the most potent dioxin-like PCB congener and demonstrates high affinity binding to Ahr (Hestermann et al., 2000; Shi et al., 2019a). The postulated molecular initiating event for PCB 126 toxicity begins with its binding and activation of Ahr followed by dysregulation of target pathways to induce injury and disease (NTP, 2006; Deng et al., 2019). It has been previously noted that Ahr may play a role in not only xenobiotic metabolism, but various energy metabolic pathways (Girer et al., 2020; Tanos et al., 2012). Ahr ligand structure and composition are also vital to consider because it has become apparent that specific ligands can also influence what metabolic pathways are acted upon (Safe et al., 2020). The current model demonstrated that PCB 126 exposure activated Ahr, via Cyp1a2 induction, and ethanol feeding activated Car, via Cyp2b10 induction. These perturbations in energy and xenobiotic metabolism suggests that both PCB 126 and ethanol may influence different pathways to jointly promote steatosis and immune modulation. Our group has previously demonstrated in an Ahr knock-out model that PCB 126’s effects on the hepatic proteome are Ahr dependent (Jin et al., 2021). Moreover, the study described by Jin et al. 2021 demonstrated Ahr’s critical role in regulating hepatic lipid metabolism. Interestingly the expression of Pparα, a nuclear receptor known to be heavily tied to lipid metabolism, was downregulated in response to ethanol feeding. Ethanol feeding also decreased the expression of other Pparα target genes (Cpt-1α, Cpt-2, and Fabp1). Indeed, the downregulated expression of Pparα and its target genes can potentially influence lipid oxidation and translocation to promote steatosis. Nonetheless, it is known that Pparα’s downregulation is a key step in the development and progression of steatosis and possibly fibrosis (Nakajima et al., 2004). Future analyses must confirm these observations as transcriptional modifiers may influence protein activity and abundance.

While this study is novel, in that it is one of few studies to address an alcohol-plus-toxicant model in FLD, it is not without limitations. A major limitation of this study is the mortality experienced prior to the expected euthanasia time point. The chronic-binge model suggests waiting nine hours post-binge to begin euthanasia and tissue collection to obtain peak liver injury enzyme levels. Furthermore, study animals were removed from their housing room approximately two hours (ZT22) before the end of the official ‘dark’ cycle, which may impact some metabolic endpoints in this study. Nonetheless, pair-fed mice were treated similarly, and we still observed significant differences between ethanol-fed and pair-fed mice. To address these limitations in future studies, mice will be removed from their housing room during the ‘light’ phase and a more time-controlled period prior to ethanol binge will be introduced to address the mortality issue. Future investigations will also incorporate more specific endpoints pertinent to ALD to better recognize injury mechanisms that PCB 126 may interact with. For instance, expression of several intestinal integrity genes was decreased in the current study due to ethanol feeding, but unperturbed due to PCB 126 exposure (Supporting Figure S5). Despite these limitations, the characterization of this model is necessary towards future model designs that are holistic in nature to understand the complex interaction between lifestyle factors and environmental toxicant exposures.

CONCLUSIONS

Environmental pollutant exposures have increasingly been recognized as disease modifiers in NAFLD, but almost no data exist for environmental health and ALD. However, as a proof of concept, cigarette smoking has previously been associated with ALD severity (Bailey et al., 2009; Lu et al., 2013). At this point in time, there are no data to suggest that drinkers are exposed to different types of pollutants than non-drinkers. Therefore, we investigated food and water contaminant, PCB 126, in ALD based on its human relevance, prior association with NAFLD, and the emerging role of its targeted receptor, Ahr, in ALD. In the present model, PCB 126 disrupted hepatic intermediary and xenobiotic metabolism. Hepatic lipid, carbohydrate, and protein (albumin) metabolism were impacted in association with hypoglycemia and attenuation of ethanol-induced hypertriglyceridemia. The net impact was worsened steatosis and hepatomegaly despite a reduction in liver glycogen stores. The PCB-exposed ALD liver did not appear capable of making and releasing glucose, albumin, and lipids appropriately, consistent with PCB’s previously established role as a metabolism disrupting chemical. We postulate that PCB 126 reprogramed intermediary metabolism thereby compounding ethanol-associated malnutrition impacting the liver-adipose and liver-muscle axes. Thus, it appears possible that PCB pollutant exposures could exacerbate the protein-calorie malnutrition associated with ALD, not by impacting food intake, but rather by transcriptional reprograming of systemic energy metabolism. Most importantly, this study suggests a potential role for environmental pollutants as understudied disease modifying factors in ALD. More data on the environmental factors influcencing ALD are required to further understand how toxicants can impact lifestyle related diseases.