The sulfate metabolite of 3,3’-dichlorobiphenyl (PCB-11) impairs Cyp1a activity and increases hepatic neutral lipids in zebrafish larvae (Danio rerio)

Abstract

The environmental contaminant 3,3’-dichlorobiphenyl (PCB-11) is widely detected in environmental samples, and this parent compound along with its metabolites 4-OH-PCB-11 and 4-PCB-11-Sulfate are detected in human serum. Our previous research in zebrafish (Danio rerio) embryos shows exposure to 20 μM PCB-11 inhibits Cyp1a enzyme activity and perturbs lipid metabolism pathways. In this study, wildtype AB embryos underwent acute exposures from 1–4 days post fertilization (dpf) to 0.002–20 μM 4-OH-PCB-11 or 0.2–20 μM 4-PCB-11-Sulfate, with and without co-exposures to 100 μg/L benzo[a]pyrene (B[a]P) or 5 nM 3,3’,4,4’,5-pentachlorobiphenyl (PCB-126), and were assessed for in vivo EROD activity and morphometrics. Chronic exposures from 1–15 dpf to assess lipid accumulation using Oil-Red-O staining were also conducted with 0.2 μM parent or metabolite compounds, alongside a co-exposure experiment of 0.002–0.2 μM 4-PCB-11-Sulfate and 10 μg/L B[a]P. For acute experiments, 2 and 20 μM 4-OH-PCB-11 was lethal but no Cyp1a or morphological effects were observed at lower concentrations; 20 μM 4-PCB-11-Sulfate significantly lowered the Cyp1a activity of B[a]P and PCB-126 but did not alter morphological development. For chronic experiments, 0.2 μM 4-PCB-11-Sulfate significantly increased lipid accumulation 30% in single exposures and 44% in co-exposures with B[a]P. Further long-term studies would better elucidate the effects of this contaminant, particularly in the context of environmentally-relevant mixtures.

Graphical Abstract

1. Introduction

The lower-chlorinated polychlorinated biphenyl (PCB) 3,3’-dichlorobiphenyl (PCB-11) is an unintentional byproduct of pigment and sealant manufacturing that can volatilize into air, leach into water sources (1, 2), and is now one of the most frequently-detected and abundant PCB congeners in these environmental sample mediums (3, 4). PCB-11 has also been detected in human serum (5), including in pregnant women (6). Our previous work in zebrafish (Danio rerio) embryos has shown that in single exposures PCB-11 impedes liver development, causes differential expression in lipid metabolism-related genes, interacts with the aryl hydrocarbon receptor (Ahr) pathway, and affects hepatic lipid accumulation. Further, in co-exposures with the Ahr agonist PCB-126, we found that PCB-11 can act as a partial Ahr antagonist to rescue toxicological outcomes, while in co-exposures with the model polycyclic aromatic hydrocarbon (PAH) beta-naphthoflavone (BNF) it can act as a Cyp1a enzyme inhibitor to exacerbate toxicological outcomes (7). However, it is not clear whether these observed effects are due to the parent PCB-11 compound or to one or more of its metabolites.

PCB-11 can undergo biotransformation into several common metabolites (8). For instance, in male Sprague-Dawley rats PCB-11 is transformed into hydroxylated (4-OH-PCB-11) and sulfated (4-PCB-11-Sulfate) metabolites within hours (9, 10). 4-OH-PCB-11 and 4-PCB-11-Sulfate are two of the most prevalent PCB-11 metabolites and have both been detected in human serum (11, 12). In this study, we tested whether these PCB-11 metabolites contribute to the Ahr interactions we observed with the parent compound, and also tested whether these metabolites contribute to hepatic lipid accumulation in the zebrafish model.

The Ahr pathway is conserved between humans and zebrafish, and is characterized as a basic-helix-loop-helix/per-Arnt-sim (bHLH/PAS) transcription factor family member that is activated by dioxin-like PCBs, many PAHs, and several other environmental chemicals (13–15). Upon Ahr ligand binding, the Ahr dissociates from co-chaperones heat shock protein 90 (HSP90) interacting protein p23 and hepatitis B virus X-associated protein 2 (XAP2) in the cytosol and translocates to the nucleus with its HSP90 dimer (16). There, it forms a heterodimer with the Ahr nuclear translocator (Arnt) (17) and binds to the xenobiotic response element to upregulate metabolism-related genes like cyp1a, which is then translated into its enzyme counterpart, Cyp1a, to metabolize the ligand (18). Genome duplication in zebrafish has given rise to several Ahr and Cytochrome p450 (Cyp) 1 family genes. Ahr gene isoforms include ahr1a, ahr1b, and ahr2, though pathway activation for the PCBs and PAHs in this study has been characterized to occur primarily through ahr2 (19, 20).

Upregulation of cyp1a in zebrafish is generally the most highly inducible Cyp1 member (21) and is similar to human Cyp1a1 induction through the Ahr (22, 23). Cyp1a enzymatic activity can be measured using the well-established in vivo ethoxyresorufin-O-deethylase (EROD) bioassay, where the substrate 7-ethoxyresorufin-O-deethylase (7-ER) binds to Cyp1a induced by the chemical of interest and can then be visualized under a fluorescence microscope with a red fluorescence protein (RFP) filter (24, 25). Ligands like BNF that activate the Ahr pathway and subsequent Cyp1a activity are good substrates for subsequent biotransformation into more hydrophilic substances for excretion (22). However, ligands like PCB-126 that upregulate the Ahr pathway and Cyp1a activity, but are not metabolized by Cyp1a, exert toxicity in zebrafish that results in pericardial edema, yolk sac edema, craniofacial malformations, and decreased swim bladder inflation when exposures begin at 1 day post fertilization (dpf) (24, 26); these outcomes are Ahr-dependent (27). Since our previous work showed PCB-11 can act as both a partial agonist and antagonist of the Ahr pathway, and since biotransformation can occur fairly quickly, this raises the question as to whether the effects we observed were direct actions of the parent compound, its metabolites, or some combination of both.

Both 4-OH-PCB-11 and 4-PCB-11-Sulfate fall along the same metabolic pathway, with PCB-11 either directly hydroxylated by Cytochrome p450 monooxygenases or indirectly via an arene oxide intermediate to form 4-OH-PCB-11, then undergoing sulfation to form 4-PCB-11-Sulfate (8) (Figure 1). The hydroxylated forms of lower-chlorinated PCBs have been shown to be more transient, but often more potent activators of the Ahr pathway than their parent compounds (28). Sulfated congeners such as 4-PCB-11-Sulfate have been shown to rapidly distribute to the liver to undergo further biotransformation by Cyps, sulfotransferases (SULTs), or Uridine 5’-diphospho-glucuronosyltranserases (UGTs) and then redistribute to other tissues, potentially to adipose tissues (11, 29). Parent PCB congeners have been shown to affect hepatic lipid accumulation (30, 31), however, few studies have investigated this outcome after direct exposure to PCB metabolites. In this study, we hypothesized that 4-OH-PCB-11 would have more amplified effects on Cyp1a activity than PCB-11 in the acute 4-day EROD bioassay, and we tested whether either of the PCB-11 metabolites would affect hepatic lipid accumulation under a 15-day chronic exposure paradigm compared to the parent compound.

Figure 1: PCB-11 Metabolism Schematic.

PCB-11 can volatilize or leach from the products it is in. Once inhaled or ingested, PCB-11 is transformed by Cyp enzymes into 4-OH-PCB-11 and then by sulfotransferase enzymes into 4-PCB-11-Sulfate.

2. Materials and Methods

2.1. Animal Care

Adult wildtype AB zebrafish (Danio rerio) were housed on a 14 h light:10 h dark cycle in a recirculating Aquaneering system (San Diego, CA) maintained at 28.5°C and fed twice daily with GEMMA Micro 300 (Skretting, Westbrook, ME). Embryos were collected from breeding groups with a 2:1 female:male ratio. Embryos were collected within 3 hours post-breeding, washed, and stored at low density in 0.3x Danieau’s media [17 mM NaCl, 2 mM KCl, 0.12 mM MgSO₄, 1.8 mM Ca(NO₃)₂, 1.5 mM HEPES, pH 7.6] in an incubator with the same temperature and light conditions as the adult fish. At 1 dpf, embryos were manually dechorionated using Watchmakers forceps and screened for normal development before use in experiments. All animal care and experiments were conducted in accordance with protocols approved by the University of Massachusetts Amherst Institutional Animal Care and Use Committee (IACUC; Protocol Number 2019–0067). Animals were treated humanely with due consideration to the alleviation of stress and discomfort.

2.2. Chemicals

3,3’,4,4’,5-pentachlorobiphenyl (PCB-126) was purchased from Ultra Scientific (North Kingstown, RI), benzo[a]pyrene (B[a]P) from Sigma-Aldrich (St. Louis, MO), and 7-ethoxyresorufin-O-deethylase (7-ER) from MP Biomedicals (Solon, OH). 3,3’-dichlorobiphenyl (PCB-11), 3,3’-dichlorobiphenyl-4-ol (4-OH-PCB-11) and 3,3′-dichloro-4-sulfooxy-biphenyl (4-OH-PCB-11-Sulfate) were synthesized at the Iowa Superfund Research Program (Iowa City, IA) via the Suzuki coupling reaction as described (32, 33). The authentication of these compounds follow guidelines established previously (34) and is documented elsewhere (6). The solubility of PCB-11 in water is 354 μg/L (35) and its octanol-water partition coefficient is log 5.28 (36). All chemicals were dissolved in 100% dimethyl sulfoxide (DMSO) from Fisher Scientific (Fair Lawn, NJ), stored as stock solutions in glass amber vials at −20°C, and were fully thawed and vortexed before use.

2.3. Chemical Exposures

The final concentrations of PCB-11, 4-OH-PCB-11, and 4-PCB-11-Sulfate used in this study ranged from 0.002–20 μM for all experiments. Higher exposure concentrations were used in the EROD bioassay to compare the effects of the metabolites to parent compound observed in our previous work (7), and are comparable to concentrations used in other rodent and cell culture studies (6, 37). Lower concentrations were used in chronic exposure experiments to mimic environmental and tissue concentrations observed in other studies (5, 11). For all experiments, wildtype AB zebrafish embryos were used, and at 1 dpf were manually dechorionated using Watchmaker forceps and screened for normal development.

For EROD and RT-qPCR experiments, zebrafish were statically exposed from 1–4 dpf to 4-OH-PCB-11 or 4-PCB-11-Sulfate, either alone for single exposure experiments or in combination with either of the Ahr agonists PCB-126 or B[a]P for co-exposure experiments, with exposures to DMSO and either of the Ahr agonists as negative and positive controls, respectively. During this period of development, ahr2 and cyp1a expression in zebrafish does not become consistent until 1 dpf (38), when exposures to both 100 μg/L B[a]P (396 nM) and 5 nM PCB-126 activate the Ahr pathway (27, 39); for PCB-126, 5 nM causes malformations in zebrafish (40), but for B[a]P, 100 μg/L exposures starting at 1 dpf induces Cyp1a activity which metabolizes the B[a]P and does not result in any morphological deformities (39). For each exposure group, five embryos were placed in 5 mL of 0.3x Danieau’s media in a 20 mL glass scintillation vial at 1 dpf, each vial containing 2 μL PCB-11 metabolite, Ahr agonist, or DMSO, and 0.5 μL 7-ER for a final DMSO concentration of 0.05% v/v and final 7-ER concentration of 0.5 μg/L. Each exposure group contained at least 2 vials, and exposure water was kept static (not changed) from 1–4 dpf. At 4 dpf, zebrafish were rinsed 3 times with Danieau’s media before use in microscopy or collection for RT-qPCR. For RT-qPCR experiments, 3–4 vials (15–20 fish) were pooled together per exposure group as 1 sample before storage in RNAlater. Each experiment was repeated at least 3 times.

For chronic exposure experiments, at 1 dpf dechorionated embryos were exposed in 5 mL of 0.3x Danieau’s media in 20 mL glass scintillation vials, with 5 embryos per vial and 4 vials per exposure group. Each vial was dosed with either DMSO, 0.2 μM PCB-11, 0.2 μM 4-OH-PCB-11, or 0.2 μM 4-PCB-11-Sulfate, with a final DMSO concentration of 0.05% v/v. Each day from 2–4 dpf, 50% of the media solution was changed. At 5 dpf, zebrafish were moved to individual vials (1 fish per vial) containing 5 mL of the appropriate Danieau’s-based media solution. From 6–8 dpf, exposure group solutions were made in system water obtained from the recirculating Aquaneering system housing the adult fish, and 5 mL of the appropriate solution was added to each vial, so that at 8 dpf each vial contained 20 mL of solution. From 9–14 dpf, exposure group solutions were made in system water and 50% of the solution (10 mL) was changed per vial. After solution changes, fish were fed standard juvenile GEMMA Micro 75 (Skretting, Westbrook, ME) fish starting at 5 dpf. In a separate experiment, the exposure paradigm described above was repeated for fish exposed to either DMSO, 0.002 μM 4-PCB-11-Sulfate + 10 μg/L B[a]P, 0.02 μM 4-PCB-11-Sulfate + 10 μg/L B[a]P, 0.2 μM 4-PCB-11-Sulfate + 10 μg/L B[a]P, or 10 μg/L B[a]P alone. At 15 dpf, all zebrafish were rinsed 3 times and moved into different vessels for live imaging before fixation for further analysis. Each experiment was repeated 3 times.

2.4. Oil-Red-O (ORO) Staining

At 15 dpf, chronically-exposed zebrafish that had undergone live imaging were put into separate 1 dram glass vials, fixed with 4% paraformaldehyde overnight covered on a rocker, rinsed 2x with 1x PBS, and stored in 70% ethanol at 4°C until processing. Briefly for ORO staining, zebrafish were rinsed 2x with 1x PBS, submerged in 85% propylene glycol for 10 min (once removed, this was saved post-removal for the next day), then 100% propylene glycol for 10 min (once removed, this was saved post-removal for the next day), and then 0.5% ORO made in 100% propylene glycol overnight on a rocker (41); the next day the ORO was removed and each fish was submerged in 100% propylene glycol (saved from the day before) for 30 min, 85% propylene glycol (saved from the day before) for 50 min, and then 85% propylene glycol again for 40 min. Equal amounts of 1x PBS were added to each vial, swirled and left for 5 min, then removed. Each fish was rinsed 3x with 1x PBS, and then submerged with 100% glycerol for storage prior to imaging. All ORO steps took place at room temperature.

2.5. Microscopy and Image Analysis

For all imaging, live larvae were sedated by a 10 second exposure to 2% v/v MS-222 solution (prepared as 4 mg/mL tricaine powder in water, pH buffered, and stored at −20°C until thawed for use) before being mounted on individual 3% methylcellulose drops in a left-lateral orientation. All imaging took place on a Zeiss Stereo Axio Zoom.V16 equipped with a HXP 200 C light source (Carl Zeiss Inc.). For EROD experiments with 4 dpf zebrafish, whole organism images were captured under brightfield settings at 20x total magnification, and the gut region was captured at 100x under both a brightfield setting and using a red fluorescent protein (RFP) filter. For chronic exposure experiments with 15 dpf zebrafish, whole organism brightfield images were captured at 20x total magnification for both live zebrafish and ORO stained zebrafish. For each individual experiment capturing EROD or ORO light intensity, the same exposure time was used for each fish across all exposure groups within the same experimental replicate. In addition to the representative images in the figures, additional representative images for each experiment are provided in the supplementary material.

All measurements for zebrafish body length, EROD light intensity, ORO light intensity, and pericardial area were measured with the Zen Lite program (Carl Zeiss Inc.). To quantify light intensity for EROD (black and white) images, the software was used to find the maximum intensity out of the entire image. To calculate background light intensity, the circle tool was used in a consistent place in each image and the mean intensity was recorded. To compare fluorescent images between experimental replicates, both background light intensity and dark pixel intensity (a calculated constant, (42)) were subtracted from the recorded light intensity for every image, and then this figure was divided by the exposure time. Each experimental replicate was normalized to its DMSO control, and then data from experimental replicates was merged. To quantify light intensity for ORO (color) images, the images were blinded and the circle tool was used to find and record the highest average light intensity in sub-regions of the liver that excluded pigment. Mean background light intensity was recorded from the same area of every image accounted for in light intensity calculations. Each experimental replicate was normalized to its DMSO control, and then data from experimental replicates was merged. We measured pericardial area on all 4 dpf zebrafish as a quantitative measurement to represent overall deformities since pericardial edema is a characteristic outcome of fish embryos exposed to PCB-126 (43) and other compounds that suppress Cyp1a activity (44). Swim bladder inflation was scored as 0=not inflated or 1=inflated. All four measured outcomes for 4 dpf fish were averaged per vial so that each vial of fish was n=1. All measured outcomes for 15 dpf fish were analyzed so that each fish was n=1 since fish in these experiments were raised in individual vials.

2.6. RT-qPCR

For all RT-qPCR experiments, at 4 dpf zebrafish larvae were rinsed 3 times and pooled in groups of 15–20 larvae, and preserved in RNAlater (Thermo Fisher Scientific, Waltham, MA) in 1.5 mL Eppendorf tubes at −80°C. Zebrafish larvae were thawed, transferred to lysis buffer, and sonicated by pulsing 3–5 times with an Emerson Industrial Branson Sonifier® (Danbury, CT). RNA isolation was performed using 2-Mercaptoethanol (MP Biomedicals) and a GeneJET RNA Purification Kit (Thermo Fisher Scientific) following manufacturer instructions. RNA quantity and quality was assessed using a BioDrop μLITE spectrophotometer (Cambridge, United Kingdom). Sample cDNA was prepared using an iScript reaction mix kit (Bio-Rad, Hercules, CA), diluted 1:9 with nuclease-free water, and stored at −80°C until processing. Each RT-qPCR sample was prepared using 10 μL of 2X iQ SYBR® Green Supermix (Bio-Rad), 5 pmol (250 nM) each of forward and reverse primers (1 μL total), 5 μL of nuclease-free water, and 4 μL (10 ng) of cDNA. Samples were run on 96-well plates in a CFX Connect Real-Time PCR Detection System (Bio-Rad), and samples were analyzed using the CFX Manager software (Bio-Rad). RT-qPCR was carried out in duplicate for the aryl hydrocarbon receptor 2 (ahr2) and cytochrome p4501A1 (cyp1a) genes. The β-actin (actb) gene was used as a housekeeping gene, and its transcription did not change significantly across exposure groups. Gene primer sequences can be found in Supplementary Table ST1 (45).

2.7. Statistical Analyses

For 4 dpf experiments, embryos damaged as a result of dechorionation error were excluded from analyses (on average one fish per exposure group per experiment). Data was then analyzed for outliers (0.25 quartile tails) for body length, EROD, and pericardial area endpoints across all experimental replicates, and individual samples were removed completely from all analyses if they were flagged as outliers for at least one endpoint. For each type of experiment, 1–3% of samples were considered outliers. For 15 dpf experiments, data was analyzed for outliers (0.25 quartile tails) for ORO light intensity and 1–2% of individual fish were removed from analysis. A one-way Analysis of Variance (ANOVA) with a Tukey-Kramer post-hoc statistical test was performed for all experiments with JMP® Pro software version 14.1.0 (Cary, NC). The unit of replication for 4 dpf experiments was the vial and the unit of replication for 15 dpf experiments was individual fish. Statistical significance was considered using a 95% confidence interval (α=0.05). For RT-qPCR experiments, gene transcription fold-changes were calculated using the ΔΔC_T method (46).

3. Results

3.1. PCB-11 inhibits Cyp1a enzyme activity induced by B[a]P but not cyp1a gene expression

We previously showed that 20 μM PCB-11 can inhibit Cyp1a enzyme activity induced by the model PAH BNF (7). To verify that this effect was consistent with a more environmentally-relevant PAH, zebrafish were co-exposed under static conditions from 1–4 dpf to 100 μg/L B[a]P in combination with the same concentrations of PCB-11 used in previous experiments: 0.2, 2, or 20 μM, and at 4 dpf fish were imaged to analyze cumulative growth, morphology and EROD activity. Fish that were exposed to B[a]P grew normally (Figure 2A), with a significant 270% increase in EROD activity (Figure 2B), maintaining normal pericardial area (Figure 2C) and swim bladder inflation (Figure 2D). Similar to previous results, 20 μM PCB-11 significantly reduced the EROD activity induced by 100 μg/L B[a]P, and fish in this co-exposure group had significantly decreased overall body length, increased pericardial area, and decreased swim bladder inflation (Figure 2D) compared to fish exposed to either DMSO or B[a]P alone; these morphological and EROD activity outcomes can be seen in Figure 2E, and additional representative images can be found in Supplementary Figure SF1. To verify whether Cyp1a activity reduction also occurred at the gene transcript level, RT-qPCR was performed on pooled groups of fish exposed either to DMSO, 20 μM PCB-11 + 100 μg/L B[a]P, or 100 μg/L B[a]P alone. No significant changes were observed for ahr2 (Figure 2F). The co-exposure group and B[a]P increased cyp1a expression 70-fold and 38-fold (Figure 2G), respectively, suggesting a synergistic response for the co-exposure group, though no statistically significant differences were observed between the co-exposure group and B[a]P alone.

Figure 2: Zebrafish exposed to 0, 0.2, 2, or 20 μM PCB-11 in combination with 100 μg/L B[a]P.

A) Body length, B) EROD quantification, C) pericardial area, and D) swim bladder inflation (mean±SEM, n=6 vials of 3–5 pooled larvae per exposure group across 3 experiments, ANOVA with Tukey’s post-hoc test, p<0.05). E) Representative images of whole zebrafish and EROD activity. The black arrow indicates significantly increased pericardial area and the white arrow indicates a significant loss of EROD activity and lack of swim bladder inflation. Gene expression for F) ahr2 and G) cyp1a are shown against β-actin as a housekeeping gene (mean±SEM, n=3 sets of 4 pooled vials of 4–5 larvae per exposure group across 3 experiments, ANOVA with Tukey’s post-hoc test, p<0.05).

3.2. 4-OH-PCB-11 and 4-PCB-11-Sulfate alone do not affect the Ahr pathway

To investigate whether PCB-11 metabolites behave similarly to the parent compound, concentrations of 4-OH-PCB-11 and 4-PCB-11-Sulfate were used in single static exposures to zebrafish from 1–4 dpf, and at 4 dpf morphology and EROD activity was measured. Initially, we exposed zebrafish to both metabolites in concentrations of 0.2, 2, and 20 μM, however, 20% and 0% survival was observed for fish exposed to 2 μM and 20 μM 4-OH-PCB-11, respectively (Supplementary Figure SF2). A concentration one order of magnitude lower (0.02 μM, or 20 nM) was added for subsequent single exposure experiments, shown here. No changes in length (Figure 3A), EROD activity (Figure 3B), or pericardial area (Figure 3C) were observed for fish exposed to either of the metabolites compared to fish exposed to DMSO, however, swim bladder inflation was significantly decreased 40% for fish exposed to 20 μM 4-PCB-11-Sulfate compared to fish exposed to DMSO (Figure 3D). Fish that were exposed to the PCB-126 positive control experienced a 6% decrease in length, a 159% increase in EROD activity, and a 142% increase in pericardial area. These morphological and EROD activity outcomes can be seen in Figure 3E, and additional representative images can be found in Supplementary Figure SF3. In a separate trial to probe whether lower concentrations of 4-OH-PCB-11 could elicit different responses, we exposed zebrafish to single static exposures of 4-OH-PCB-11 in concentrations of 0.00002, 0.0002, and 0.002 μM, though we did not observe any differences in morphology or EROD activity compared to fish exposed to DMSO (Supplementary Figure SF4). However, the 0.002 μM concentration was added for subsequent co-exposure experiments.

Figure 3: Zebrafish exposed to either 0.02–0.2 μM 4-OH-PCB-11, 0.2–20 μM 4-PCB-11-Sulfate, or DMSO or 5 nM PCB-126 controls from 1–4 dpf.

A) Body length, B) EROD quantification, C) pericardial area, and D) swim bladder inflation (mean±SEM, n=8 vials of 3–5 pooled larvae per exposure group across 3–5 experiments, ANOVA with Tukey’s post-hoc test, p<0.05). E) Representative images of whole zebrafish and EROD activity. For the PCB-126 positive control fish, the black arrow indicates significantly increased pericardial area and the white arrow indicates significantly increased EROD activity and lack of swim bladder inflation.

3.3. 4-PCB-11-Sulfate inhibits Cyp1a enzyme activity but not cyp1a gene expression

To investigate whether 4-OH-PCB-11 or 4-PCB-11-Sulfate interact with the Ahr pathway in the presence of other Ahr agonists, we co-exposed zebrafish to either DMSO, 5 nM PCB-126, or 5 nM PCB-126 in combination with either 0.002–0.2 μM 4-OH-PCB-11 or 0.2–20 μM 4-PCB-11-Sulfate from 1–4 dpf and then measured morphological development and EROD activity. At 4 dpf, zebrafish exposed to 5 nM PCB-126 experienced a 7% reduction in body length (Figure 4A), a 404% increase in EROD activity (Figure 4B), a 132% increase in pericardial area (Figure 4C), and no swim bladder inflation (Figure 4D). These morphological and EROD activity outcomes can be seen in Figure 4E, and additional representative images can be found in Supplementary Figure SF5. We observed that neither metabolite prevented the morphological deformities induced by PCB-126, however, 20 μM 4-PCB-11-Sulfate significantly reduced EROD activity to similar levels as in the DMSO exposure group. To investigate whether this reduction also occurred at the gene transcript level, RT-qPCR to look at ahr2 and cyp1a was performed on pooled groups of 4 dpf zebrafish that had been exposed to either DMSO, 20 μM 4-PCB-11-Sulfate, 20 μM 4-PCB-11-Sulfate + 5 nM PCB-126, or 5 nM PCB-126. Both exposure groups containing PCB-126 induced about a 1.5-fold change increase in ahr2 and a 150-fold change increase in cyp1a transcript levels (Figure 4F), with no significant differences between the co-exposure group and the group containing PCB-126 alone; thus, 20 μM 4-PCB-11-Sulfate did not inhibit the expression of ahr2 or cyp1a induced by PCB-126.

Figure 4: Zebrafish exposed to 5 nM PCB-126 and either 0.002–0.2 μM 4-OH-PCB-11 or 0.2–20 μM 4-PCB-11-Sulfate from 1–4 dpf.

A) Body length, B) EROD quantification, C) pericardial area, and D) swim bladder inflation (mean±SEM, n=6 vials of 3–5 pooled larvae per exposure group across 3 experiments, ANOVA with Tukey’s post-hoc test, p<0.05). E) representative images of whole zebrafish and EROD activity. The white arrow indicates a significant loss of EROD activity. F) Gene expression for ahr2 and cyp1a is shown for zebrafish exposed to either DMSO, 20 μM 4-PCB-11-Sulfate, 20 μM 4-PCB-11-Sulfate + 5 nM PCB-126, or 5 nM PCB-126, using β-actin as a housekeeping gene (mean±SEM, n=3–4 sets of 4 pooled vials of 4–5 larvae per exposure group across 4 experiments, ANOVA with Tukey’s post-hoc test, p<0.05).

In another experiment to probe whether either metabolite would inhibit Cyp1a activity and result in toxicological outcomes in co-exposures with B[a]P, zebrafish were exposed under static conditions from 1–4 dpf to either DMSO, 100 μg/L B[a]P, or 100 μg/L B[a]P in combination with either 0.002–0.2 μM 4-OH-PCB-11 or 0.2–20 μM 4-PCB-11-Sulfate, and then morphology and EROD activity were measured. At 4 dpf, zebrafish exposed to 100 μg/L B[a]P developed normally (Figure 5A) with a 140% increase in EROD activity (Figure 5B), and normal pericardial area (Figure 5C) and swim bladder inflation (Figure 5D). 20 μM 4-PCB-11-Sulfate significantly reduced the B[a]P-induced EROD activity to levels observed in the DMSO exposure group, however, this did not result in morphological deformities for the fish; these morphological and EROD activity outcomes are shown in Figure 5E, and additional representative images can be found in Supplementary Figure SF6. In separate experiments we also verified that EROD activity induced by BNF could be inhibited by 20 μM 4-PCB-11-Sulfate without resulting in morphological deformities (Supplementary Figure SF7). To investigate whether inhibition of Cyp1a induced by B[a]P also occurred at the gene transcript level, RT-qPCR to look at ahr2 and cyp1a was performed on pooled groups of 4 dpf zebrafish that had been exposed to either DMSO, 20 μM 4-PCB-11-Sulfate, 20 μM 4-PCB-11-Sulfate + 100 μg/L B[a]P, or 100 μg/L B[a]P. Both exposure groups containing B[a]P induced about a 1.6-fold change increase in ahr2 and a 37-fold change increase in cyp1a transcript levels (Figure 5F), with no significant differences between the co-exposure group and the group containing B[a]P alone; thus, 20 μM 4-PCB-11-Sulfate did not inhibit the expression of ahr2 or cyp1a induced by 100 μg/L B[a]P.

Figure 5: Zebrafish exposed to 100 μg/L B[a]P and either 0.002–0.2 μM 4-OH-PCB-11 or 0.2–20 μM 4-PCB-11-Sulfate from 1–4 dpf.

A) Body length, B) EROD quantification, C) pericardial area, and D) swim bladder inflation (mean±SEM, n=5–6 vials of 3–5 pooled larvae across 3 experiments per exposure group, ANOVA with Tukey’s post-hoc test, p<0.05). E) representative images of whole zebrafish and EROD activity. The white arrow indicates a significant loss of EROD activity and lack of swim bladder inflation. F) Gene expression for ahr2 and cyp1a is shown for zebrafish exposed to either DMSO, 20 μM 4-PCB-11-Sulfate, 20 μM 4-PCB-11-Sulfate + 100 μg/L B[a]P, or 100 μg/L B[a]P, using β-actin as a housekeeping gene (mean±SEM, n=3–4 sets of 4 pooled vials of 4–5 larvae per exposure group across 4 experiments, ANOVA with Tukey’s post-hoc test, p<0.05).

3.4. 4-PCB-11-Sulfate increases hepatic neutral lipid accumulation

In previous work, we found that zebrafish exposures to 20 μM PCB-11 from 1–4 dpf misregulated genes involved with lipid metabolism, and appeared to increase vacuole formation in liver tissue (7). In this study, zebrafish were exposed to either DMSO or 0.2 μM concentrations of PCB-11, 4-OH-PCB-11, or 4-PCB-11-Sulfate under a chronic exposure paradigm from 1–15 dpf and then collected for ORO staining. We found that 0.2 μM 4-PCB-11-Sulfate significantly increased hepatic lipid accumulation in zebrafish by 30% compared to zebrafish exposed to DMSO (Figures 6A–B). In a subsequent experiment, zebrafish were exposed to either DMSO, 10 μg/L B[a]P, or 10 μg/L B[a]P in combination with either 0.002, 0.02, or 0.2 μM 4-PCB-11-Sulfate under chronic exposure conditions from 1–15 dpf. We found that zebrafish exposed to 10 μg/L B[a]P experienced a non-significant 18% increase in hepatic neutral lipid accumulation, and a dose-dependent increase in hepatic lipid accumulation for the co-exposure groups of 11%, 20%, and 44%, with significance for the highest concentration exposure group of 10 μg/L B[a]P + 0.2 μM 4-PCB-11-Sulfate (Figures 6C–D). Additional representative images for these 15 dpf experiments can be found in Supplementary Figure SF8. Lipid accumulation was also analyzed for the brain region of each fish, but these increases were not significant for either experiment (Supplementary Figure SF9).

Figure 6:

A) Representative images and B) quantification of ORO staining of zebrafish exposed to either DMSO, 0.2 μM PCB-11, 0.2 μM 4-OH-PCB-11, or 0.2 μM 4-PCB-11-Sulfate from 1–15 dpf (mean±SEM, n=45–51 fish per exposure group across 3 experiments, ANOVA with Tukey’s post-hoc test, p<0.05). The black arrow points to the liver area of the exposure group where significance was observed. C) Representative images and D) quantification of ORO staining of zebrafish exposed to either DMSO, 10 μg/L B[a]P, or 10 μg/L B[a]P in combination with 0.002–0.2 μM 4-PCB-11-Sulfate from 1–15 dpf (mean±SEM, n=37–42 fish per exposure group across 3 experiments, ANOVA with Tukey’s post-hoc test, p<0.05). The black arrows point to the liver area of the exposure group where significance was observed. Background contrasting (+20% for single exposure images and +40% for co-exposure images) was made in the same way for all images for better clarity.

4. Discussion

In this study, two prevalent PCB-11 metabolites, 4-OH-PCB-11 and 4-PCB-11-Sulfate, were investigated using the zebrafish model to understand how they affect the Ahr pathway in larval zebrafish under an acute exposure paradigm, and how they affect hepatic lipid accumulation in juvenile zebrafish under a chronic exposure paradigm. PCB-11 is a lower-chlorinated congener that is a byproduct in the manufacturing of many consumer products, most notably diarylide azo-type pigments, but has only recently undergone investigation as a potential public health risk (47). While PCB-11 has been detected in many water bodies downstream of wastewater treatment plants or industrial manufacturing (1, 48), it is thought that most of human exposures to PCB-11 occur through inhalation due to its semi-volatile nature (2, 49). Once inhaled, PCB-11 travels to the bloodstream and undergoes metabolism in the liver, where metabolites are either excreted, distributed throughout the body, or further metabolized (50) (Figure 1). We previously used the zebrafish model to conduct waterborne PCB-11 exposures where we were able to measure PCB-11 reduction in exposure water, uptake into larval tissue, and toxicological effects (7). In this study, we conducted similar exposures of the 4-OH-PCB-11 and 4-PCB-11-Sulfate metabolites to understand whether these metabolites drive the Ahr pathway effects we observed previously, and to understand whether there would be longer term hepatic effects after lower-concentration chronic exposures.

In our previous work, we observed that in combination with 50 μg/L BNF, 20 μM PCB-11 partially inhibited both cyp1a gene transcription and Cyp1a enzyme activity and resulted in severe toxicological outcomes (7). In this study we used B[a]P as a more environmentally relevant chemical to probe PCB-11 and metabolite interactions with Cyp1a activity. In contrast with 50 μg/L BNF co-exposures, 20 μM PCB-11 did not significantly reduce cyp1a gene expression in co-exposures with 100 μg/L B[a]P, however, it inhibited the Cyp1a activity (as shown via EROD) induced by B[a]P and still resulted in severe morphological deformities (Figure 2). This could be due to differences in concentrations used for these two PAHs, where 20 μM PCB-11 may have been sufficient to inhibit Cyp1a enzyme function induced by 100 μg/L B[a]P but not sufficient to inhibit cyp1a gene expression; or, there may have been direct inhibition at the enzyme level rather than inhibition of transcription via the receptor. Differences could also exist in receptor-mediated binding between PCB-11 and the two PAHs, as this can vary greatly between PCB congeners (51).

The PCB-11 metabolites we tested in acute 1–4 dpf single exposures did not induce changes in Cyp1a activity or morphological outcomes (Figure 3). We initially hypothesized that the 4-OH-PCB-11 metabolite would result in more dramatic effects on Cyp1a activity, since hydroxylated PCB congener metabolites, including metabolites of lower-chlorinated and non-dioxin-like congeners can exhibit more potent effects on the Ahr pathway (28). Pěnčíková et al. reported in a cell culture model that while the parent PCB-11 compound can have mild effects on Ahr activity, 4-OH-PCB-11 is a significant inhibitor of Ahr activation (52). Hydroxylated PCB metabolites are potentially more toxic than their parent compounds (53) and have been reported in human samples in the same concentration range as parent PCB compounds (8, 54). However, for the concentrations we tested in our zebrafish model, we did not see any effects on Cyp1a activity in either single or co-exposure settings, though we did observe increased larvae lethality for 4-OH-PCB-11 concentrations greater than 0.2 μM, confirming reports that it is more overtly toxic than the parent compound (53).

While no differences were observed for single exposures of PCB-11 metabolites, in co-exposure experiments, 20 μM 4-PCB-11-Sulfate inhibited Cyp1a activity, but not cyp1a gene transcription induced by either 5 nM PCB-126 (Figure 4) or 100 μg/L B[a]P (Figure 5). As with the parent PCB-11 compound co-exposure experiments, this could again be a PAH concentration-specific outcome. However, 5 nM PCB-126 was also used as an Ahr agonist in co-exposure experiments and we observed that while 20 μM PCB-11 significantly reduced both cyp1a gene transcription and Cyp1a enzyme activity (7), 20 μM 4-PCB-11-Sulfate only significantly reduced Cyp1a activity. The main difference between PCB-11 and 4-PCB-11-Sulfate co-exposure outcomes was that PCB-11 Cyp1a activity inhibition corresponded to severe morphological deformities for zebrafish larvae, whereas 4-PCB-11-Sulfate Cyp1a activity inhibition did not. From these results it is clear that toxicity outcomes differ between the parent and metabolite PCB-11 compounds, and that Cyp1a inhibition is not a requirement for toxicological outcomes.

Interestingly, when we subsequently tested co-exposures of PCB-11 metabolites with 50 μg/L BNF in this study, we also observed that 20 μM 4-PCB-11-Sulfate inhibited Cyp1a activity without causing toxicity (Supplementary Figure SF7). One possibility is that the enzyme kinetics between the parent PCB-11 compund and sulfate metabolite are different, where an irreversible inhibition occurs with the parent compound, leading to toxicological outcomes, but a reversible inhibition occurs with the metabolite (55), which allows for normal development. Another possibility is that research groups have clearly demonstrated the toxicological outcomes that occur via Ahr-dependent Cyp1a enzyme inhibition (24, 44), but that despite Cyp1a inhibition, other enzymes such as Cyp1b1 or Cyp1c1, could play a role in metabolizing PAH Ahr agonists (56). For instance, in a cell culture model PCB-11 has been shown to be a constitutive androstane receptor (CAR) agonist, increasing Cyp2b6 mRNA levels (52). In addition, inhibition of Cyp1a activity in the presence of non-PAH Ahr agonists such as PCB-126 or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) has been shown not to influence toxicity, since neither of these substrates is metabolized by Cyp1a. Overall, more research is needed to understand how PCB-11 and its metabolites activate and affect hepatic enzyme function.

Not many studies have yet explored how lower-chlorinated PCBs like PCB-11, and their metabolites, interact with the Ahr pathway and affect lipid biology. Alam et al. investigated how exposures of 4-OH-PCB-11 in cells demonstrated a protective role of functional sirtuin 3 (SIRT3), without which cells exposed to low μM concentrations 4-OH-PCB-11 had increased mitochondrial respiration and increased expression of genes controlling fatty acid biosynthesis, metabolism, and transport (37). In the present study, we saw an increase in hepatic neutral lipids in fish that were exposed to both single exposures of 20 μM 4-PCB-11-Sulfate and co-exposures with 10 μg/L B[a]P (Figure 5). Dioxin-like PCBs like PCB-126 misregulate adipose biology by increasing free fatty acids and triglycerides, leading to a non-alcoholic fatty liver disease (NAFLD) phenotype. In a rodent model some of the root causes of NAFLD after exposures to dioxin-like PCBs is differential regulation of genes like patatin-like phospholipase domain-containing protein 3 (pnpla3) and the hepatokine fibroblast growth factor 21 (fgf21). However, non-dioxin-like PCBs have also disrupted the regulation of these genes (57), perhaps through crosstalk with other transcription factors that affect lipid metabolism and lead to downstream changes in physiological function (58). Specifically, an AHR-FGF21 regulatory axis has been established, showing that a properly-functioning Ahr controls fgf21 signaling, which in turn partially controls the regulation of the peroxisome proliferator-activator receptor α (pparα) gene, among other metabolically important genes (59, 60). Other research has also shown that Cyp1a activation by B[a]P inhibits 17β-estradiol (E2), which subsequently reduces the expression of PPARα and increases the expression of adipogenic genes (61). We observed from our previous RNAseq experiment that single exposures to 20 μM PCB-11 significantly increased cyp1a gene transcription 2.35-fold change, and this corresponded to significant downregulation 0.57-fold of sterol regulatory element-binding transcription factor 1 (srebf1) and upregulation 1.38-fold of fatty acid binding protein 10a (fabp10a), as well as overall downregulation of the KEGG fatty acid metabolism pathway when we did a bioinformatics analysis (7),

PCB-11 and its metabolites have also been shown to affect endocrine activity in cell culture models, demonstrating full or partial agonist and antagonist activity against the androgen receptor (AR) and estrogenic receptor (ER) (52, 62, 63). While our group has not specifically focused on hormone receptor effects, in our previous RNAseq experiment we found that one of the most significantly misregulated genes after zebrafish exposures to 20 μM PCB-11 was the parathyroid hormone 1a (pth1a) gene (7), orthologous to human parathyroid hormone and important for calcium ion homeostasis. Other research groups have shown that 4-PCB-11-Sulfate is a high affinity ligand of the thyroid hormone transport protein transthyretin (TTR) and can cross into brain tissue (29, 33). TTR is a thyroid hormone binding plasma protein that is primarily synthesized in the liver and plays an important role in transporting thyroxine (T₄) to tissues throughout the body, including the brain (64) and the placenta (65), where it is converted into the active thyroid hormone triiodothyronine (T₃). Many parent PCB congeners, and particularly their hydroxylated metabolites, have been established as high-affinity TTR ligands (66) to cause reductions in T₄ concentrations in rodent plasma and forebrain samples (67). Thus, it is interesting that 4-PCB-11-Sulfate also has the ability to do this. While we originally did not plan to investigate brain-specific outcomes in this study, we noticed potential differences between exposure groups in brain lipid accumulation, and while we did observe an increase in lipid accumulation to nearly the same degree in fish exposed to 0.2 μM 4-PCB-11-Sulfate alone, and a dose-dependent increase in lipid accumulation in fish co-exposed to 4-PCB-11-Sulfate + 10 μg/L B[a]P, these increases were not statistically significant (Supplementary Figure SF9).

In this present study, we tested a wide range of PCB-11, 4-OH-PCB-11, and 4-PCB-11-Sulfate concentrations. For the 15 dpf chronic exposure experiments, we observed significant effects on hepatic lipid accumulation for 4-PCB-11-Sulfate at 0.2 μM. For the 4 dpf acute exposure experiments, we observed significant effects on Cyp1a activity for 4-PCB-11-Sulfate at 20 μM. These concentrations at which we observed significant effects are the concentrations of the exposure water and do not reflect internal body concentrations. Our goal was to emulate internal body concentrations in zebrafish that would be observed in human serum. While in this study we did not measure how much of the PCB-11 metabolites in the water were taken up into fish tissue, we did measure parental PCB-11 uptake in our previous study, where only 0.61% of the PCB-11 in the exposure water was observed in whole body fish tissue at the 4 dpf collection time (7). If 4-PCB-11-Sulfate uptake into zebrafish tissue is similar to the parental PCB-11 compound, the resulting internal body concentrations from 20 μM water exposures is 2x the 4-PCB-11-Sulfate concentration reported in human serum (61 nM), and the resulting internal body concentration from 0.2 μM 4-PCB-11-Sulfate, at which we observed significant hepatic lipid accumulation, is well within this reported range (11). However, further research is needed to understand how larval and juvenile stages of zebrafish metabolize lower-chlorinated PCB congeners.

This present research highlights the importance of conducting mixtures assessments of environmental chemicals, as exposures to lower-chlorinated PCB congeners such as PCB-11 are not likely to occur in isolation. In addition, while it is important to understand how acute exposures to PCB-11 can affect hepatic enzyme function in the presence of other chemicals, it is perhaps more important to understand how lower-concentration chronic exposures to PCB-11 in combination with other environmentally relevant compounds can affect adipose biology and other endpoints over time. For instance, NAFLD affects around 25% of the world population and has an enormous economic burden in many countries (68, 69). It is now recognized that environmental chemicals play a role in toxicant-induced steatohepatitis (TASH) (70). Additional work to gather gene-level and tissue-level data on how PCB-11 and its metabolites affect hepatic lipid biology at different developmental time points would better illustrate PCB-11’s potential contributions to liver disease development.

5. Conclusions

In this study, two prevalent PCB-11 metabolites, 4-OH-PCB-11 and 4-PCB-11-Sulfate, were tested in the zebrafish model under an acute exposure paradigm to assess their effects on Cyp1a activity, and under a lower-concentration chronic exposure paradigm to assess their effects on hepatic lipid accumulation. Our results show that 20 μM 4-PCB-11-Sulfate inhibited Cyp1a activity induced by Ahr agonists, and that 0.2 μM 4-PCB-11-Sulfate increased hepatic lipid accumulation. Further long-term studies with the 4-PCB-11-Sulfate metabolite would be beneficial to better understand its public health risk.

Highlights.

• 4-OH-PCB-11 is more acutely toxic than both the parent PCB-11 compound and the 4-PCB-11-Sulfate metabolite

• 4-OH-PCB-11 and 4-PCB-11-Sulfate metabolites alone do not affect the Ahr pathway or morphological development

• 20 μM 4-PCB-11-Sulfate acute 4-day exposures inhibits Cyp1a activity in the presence of Ahr agonists

• 0.2 μM 4-PCB-11-Sulfate chronic 15-day exposures increases hepatic neutral lipid accumulation

Abbreviations

Ahr

Aryl hydrocarbon receptor

ANOVA

analysis of variance

B[a]P

benzo[a]pyrene

BNF

beta naphthoflavone

EROD

Ethoxyresorufin-O-deethylase

Cyp1a

Cytochrome p450 1a

dpf

days post fertilization

ORO

Oil-red-O

PCB

Polychlorinated biphenyl

PCB-11

3,3’-dichlorobiphenyl

PCB-126

3,3’,4,4’,5-pentachlorobiphenyl

RT-qPCR

reverse transcriptase quantitative polymerase chain reaction