3,3'-Dichlorobiphenyl (PCB 11) alters the hepatic expression of cytochrome P450 enzymes in the liver of mouse dams exposed orally during pregnancy and lactation
Crystal M. Roach, Nate R. Koester, Xueshu Li, Jeonghyeon Ahn, R. Marshall Pope, Rebecca J. Wilson, Rosalia Mendieta, Anthony Valenzuela, Weiguo Han, Xinxin Ding, Pamela J. Lein, Hans-Joachim Lehmler
TL;DR
This study shows that exposure to PCB 11 during pregnancy and lactation in mice disrupts liver enzymes important for detoxification and metabolism.
Contribution
The first comprehensive hepatic proteome analysis of PCB 11 exposure during pregnancy and lactation in mice.
Findings
PCB 11 exposure led to downregulation of cytochrome P450 enzymes and other metabolic proteins in the liver.
Metabolite screening identified OH-PCB 11 and related sulfate metabolites in serum.
Pathway analysis revealed disruptions in xenobiotic metabolism and endocrine pathways.
Abstract
Healthy maternal metabolism is critical during pregnancy and lactation to support fetal development and protect against environmental toxins. Polychlorinated biphenyl 11 (PCB 11), a lower-chlorinated, non-legacy congener, is detected in human serum, including pregnant women and children; however, its impact during these sensitive life stages remains poorly understood. This study presents the first comprehensive hepatic proteome analysis of mouse dams exposed to PCB 11 during pregnancy and lactation. Female C57BL/6 J mice received daily oral doses of PCB 11 (0, 1.0 or 6.0 mg/kg) prior to conception through lactation. At postpartum day 21, brain, liver, and serum samples were analyzed for PCB 11 and its metabolites, and hepatic proteomic changes were assessed. Low detection of PCB 11 and its metabolites was observed in tissues, suggesting rapid clearance. Metabolite screening revealed…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4- —http://dx.doi.org/10.13039/100000066National Institute of Environmental Health Sciences
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsToxic Organic Pollutants Impact · Effects and risks of endocrine disrupting chemicals · Carcinogens and Genotoxicity Assessment
Introduction
Polychlorinated biphenyls (PCBs) are synthetic persistent organic pollutants formerly used in industrial and consumer products for their chemical stability, thermal resistance, and insulating properties (Lallas 2001). Although production ceased in the United States in the late 1970s, PCBs persist in the environment and continue to pose adverse health risks. PCB congeners are classified as dioxin-like (DL) or non-dioxin-like (NDL) based on their mechanisms of action (Van den Berg et al. 2006). DL-PCBs exert their effects through binding to the aryl hydrocarbon receptor (AhR), leading to the induction of cytochrome P450 (P450) enzymes such as CYP1A1 and CYP1A2 in the liver (Jin et al. 2021). In contrast, NDL-PCBs interact with nuclear receptors such as the constitutive androstane receptor (CAR) and pregnane X receptor (PXR) (Wahlang et al. 2015), resulting in the induction of hepatic CYP2B isoforms (Al-Salman and Plant 2012). Notably, individual PCB congeners vary in their receptor affinities and induce specific P450 isoforms, contributing to congener-specific metabolism and toxicity. NDL-PCBs are a particular concern as they are the predominant PCB congeners detected in human tissues (Beyer and Biziuk 2009).
Recent concerns have emerged surrounding inadvertently produced lower-chlorinated PCB congeners (LC-PCBs) with four or fewer chlorine substituents (Grimm et al. 2015). 3,3'-Dichlorobiphenyl (PCB 11) is an LC-PCB that is a byproduct of diarylide yellow pigment manufacturing (Hu and Hornbuckle 2010). Despite its volatility and short half-life (Hu et al. 2013), PCB 11 is frequently detected in ambient air and dietary sources, contributing to ongoing human exposure through inhalation and ingestion (Grimm et al. 2015). Recent studies suggest that indoor inhalation has become a dominant route of human exposure to PCB 11 (Ampleman et al. 2015). The Markers of Autism Risk in Babies-Learning Early Signs (MARBLES) study detected PCB 11 in maternal serum from pregnant women, highlighting the importance of investigating its biological effects during this critical period (Sethi et al. 2019). Emerging evidence implicates LC-PCBs, including PCB 11, and their metabolites in a range of toxic outcomes, including endocrine disruption and neurotoxicity (Grimm et al. 2015). Importantly, PCB 11 and its metabolites have been linked to developmental neurotoxicity in cell culture studies (Sethi et al. 2017, 2018).
Hepatic cytochrome P450 enzymes facilitate PCB metabolism by transforming parent compounds into hydroxylated metabolites (OH-PCBs), which may exert greater toxicity than the parent PCBs (Grimm et al. 2015). In rodents, hepatic CYP1A, CYP2A, and CYP2B isoforms metabolize PCBs to OH-PCBs under the regulation of the CAR and PXR (Grimm et al. 2015). In humans, CYP2A6 and CYP2B6 are involved in the oxidation of PCBs (Uwimana et al. 2019). Further metabolism of OH-PCBs yields complex mixtures of dihydroxylated, dechlorinated, methyl sulfone, and phase II conjugates (Grimm et al. 2015). For example, PCB 11 is metabolized to OH-PCB 11 by CYP1A and CYP2B family enzymes (Kaminsky et al. 1981), and these hydroxylated metabolites can undergo additional modifications, including oxidation, sulfation, glucuronidation, and methylation, in cultured cells (Zhang et al. 2020).
The potential for PCB 11 to be metabolized into developmental neurotoxicants raises concerns about its impact on maternal liver function, particularly the expression of hepatic P450 enzymes, during pregnancy and lactation. The physiological adaptations that occur during these periods include significant changes in the expression of hepatic cytochrome P450 (Koh et al. 2011). These adaptations, along with PCB 11-induced alterations in the expression of P450 enzymes, may affect how the mother metabolizes PCB 11 into toxic metabolites and how these substances are transferred to the fetus or neonate. Furthermore, the fetus and neonate have limited detoxification capacity due to immature hepatic enzyme systems (Makri et al. 2004). While the effects of higher-chlorinated (HC-) PCBs on the liver proteome have been characterized in mice (Rignall et al. 2013), the hepatic effects of PCB 11 and other LC-PCBs remain understudied. To address this knowledge gap, this study investigated alterations in the hepatic proteome at weaning (postpartum day 21) in mouse dams exposed to PCB 11 through their diet during pregnancy and lactation.
Materials and methods
Chemicals and analytical standards
CAUTION: Handle PCBs, dichloromethane (Group 1 carcinogens), and diazomethane (toxic and explosive derivatization reagent) only in a chemical fume hood while wearing appropriate personal protective equipment. PCB 11 (99.97%) used to dose animals was synthesized from the corresponding benzidine and authenticated following a published guideline (Li et al. 2018). Sources of chemicals, abbreviations, and unique identifiers (Online Resource Table S1) of all analytical standards are summarized in the Online Resource.
Animal exposure and tissue collection
Eight-week-old wildtype dams were exposed daily to PCB 11 (1.0 or 6.0 mg/kg/day) via peanut butter, beginning two weeks before mating and continuing throughout pregnancy and lactation until postpartum day (PD) 21 as part of a developmental neurotoxicity study being conducted at the University of California, Davis. This exposure window was selected to model critical periods of human neurodevelopment, given that the first three postnatal weeks in rodents correspond to the human third trimester of brain development (Rice and Barone 2000; Semple et al. 2013). The dosing paradigm was based on previous studies demonstrating non-monotonic dose–response relationships in offspring brain outcomes following maternal exposure to a human-relevant PCB mixture containing PCB 11 (Sethi et al. 2021; Keil-Stietz et al. 2021). Control dams received vehicle (sterile peanut butter) alone. At PD21, dams were euthanized within 20 h of the final PCB 11 exposure. Blood samples and tissues were collected, snap-frozen in liquid nitrogen, and stored at -80 °C. All samples were shipped on dry ice to The University of Iowa for further analysis. No adverse effects were observed during animal exposure. Samples per exposure group per cohort (n = 6) were randomly collected from the larger developmental neurotoxicity study, with samples from the same animals used for all analyses.
Analysis of PCB 11 and its hydroxylated metabolites in tissues and serum
PCB 11 and metabolite levels were determined by gas chromatography-tandem mass spectrometry (GC–MS/MS) and liquid chromatography-high resolution mass spectrometry (LC-HRMS). For GC–MS/MS analyses, PCB 11 and its hydroxylated metabolites were extracted from brain (n = 18; 0.025 ± 0.002 g), liver (n = 18; 0.110 ± 0.014 g), and serum (n = 18; 0.058 ± 0.021 g) using a modified liquid–liquid extraction method (Hu et al. 2013). OH-PCB extracts were derivatized using diazomethane to form methoxylated PCBs prior to GC–MS/MS analysis, as described in the Online Resource. Precursor ions, product ions, and collision energies for PCB 11 and its metabolites, as well as quality assurance/quality control (QA/QC) information, are provided in Online Resource Tables S2–S6. For LC-HRMS analysis, serum aliquots (n = 18; 0.058 ± 0.002 g) were analyzed using a modified protocol (Li et al. 2024), with surrogate standards (3-F, 4'-OH-PCB 3 and 3-F, 4'-PCB 3 sulfate) to assess recovery, reproducibility, and precision. Detailed sample preparation, QA/QC, and instrument procedures are provided in the Online Resource.
Liver proteome analysis by liquid chromatography-mass spectrometry coupled with tandem mass spectrometry (LC–MS/MS)
Protein extraction, quality validation, and in-gel digestion
Liver tissue samples (n = 18; 0.050–0.054 g) were lysed using TRIzol® (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer’s instructions. Protein quality was validated by SDS-PAGE, and for in-gel digestion, targeted protein bands were excised, diced into ~ 1 mm^3^ pieces, washed with ammonium bicarbonate:acetonitrile, dried, reduced with dithiothreitol, alkylated with chloroacetamide, washed again, dehydrated, and quantified using the Pierce Micro BCA™ Protein Assay Kit (Thermo Fisher Scientific). Each sample (55 µg) was then digested with Lys-C followed by overnight trypsin incubation, and the resulting peptides were desalted using a Sep-Pak C18 Cartridge (Thermo Fisher Scientific), with salt removal confirmed by MALDI analysis for quality control. One sample per treatment group was removed based on low or high protein levels to give a sample number of n = 5 in all exposure groups, allowing for LC–MS/MS analysis in a single batch. For additional information, refer to the Online Resource.
Isobaric labeling and LC–MS/MS analysis
Following quality control validation, samples were labeled using tandem mass tags (TMT) as previously described (Zecha et al. 2019). Briefly, equal aliquots of protein were dissolved in 17.5 µL of triethylammonium bicarbonate and mixed at a 2:1 (w/w) ratio with TMTpro™ 16-plex Label Reagent Set (Thermo Fisher Scientific). The TMT reagent consists of an MS/MS reporter group, a spacer arm, and an amine-reactive group, and uses synchronous precursor selection (SPS)-based MS3 to minimize interference from chimeric precursors in MS2 spectra. Samples were incubated for 1 h at ambient temperature, quenched with hydroxylamine, and pooled. The combined sample was desalted and fractionized into eight fractions using the Pierce™ High pH Reversed-Phase Peptide Fractionization kit (Thermo Fisher Scientific), according to the manufacturer’s instructions. Fractions were lyophilized and reanalyzed by MALDI for additional quality control.
A global reference sample was prepared by pooling equal volumes from all fractionized samples and analyzed alongside experimental samples. The LC–MS/MS analysis was performed on a LUMOS Orbitrap LC–MS/MS system (Thermo Fisher Scientific), with the eight fractions analyzed in duplicates. The instrument operated in a data-dependent acquisition mode, capturing MS1 spectra of eluting peptides followed by MS2 fragmentation to generate reporter ions. Positively identified peptides in MS2 triggered an MS3 scan, isolating and fragmenting up to 10 sequence ions per precursor. Reporter ion intensities were quantified using 3 mmu peak integration around the most confident centroids, enhancing quantification accuracy and proteome coverage.
Proteomics profiling, quantification, and network analysis
Raw proteomics data were processed in Proteome Discoverer v3.0.4.388 (Thermo Fisher Scientific) using the “consensus workflow,” with PSMs searched against the Mus musculus UniProt and SwissProt databases under defined mass tolerances, enzymatic constraints, and modification settings. Peptide and protein identification was controlled at strict FDR thresholds (0.01), with quantification based on unique peptides normalized to the channel with the highest intensity. Bioinformatics analysis was performed in Amica v3.0.1, including filtering, normalization, imputation, and statistical testing. Statistical differential expression in the proteomics analysis was determined using Limma R package, with an FDR threshold of ≤ 0.05. In the volcano plots, a log2 fold-change cutoff of ≥ 1.0 or ≤ -1.0 comparing PCB-exposed and control groups was used to visualize a two-fold change in abundance. This threshold identifies proteins with biological relevance (Schork et al. 2021). Functional protein networks were analyzed using STRING v12.0 (Szklarczyk et al. 2015), clustered by Markov clustering (MCL, inflation = 3.0), with nodes color-coded by cluster and edges representing high-confidence associations (score ≥ 0.900). Differentially expressed proteins were annotated via UniProt and mapped to KEGG pathways for enrichment analysis (P ≤ 0.05) relative to the Mus musculus reference list. For additional details, refer to the Online Resource.
RT-PCR and immunoblot analysis of CYP2B10 in maternal liver
To assess CYP2B10 expression, liver mRNA and protein were analyzed by RT-PCR and western blotting, respectively. For quantitative RT-PCR, total RNA was extracted from the livers of the dams and reverse transcribed into cDNA. PCR reactions were performed in triplicate. Detailed protocols for sample preparation, RT-PCR conditions, and primer/probe sequences are provided in the Online Resource. For protein analysis, pooled liver microsomal proteins (n = 3; 5 µg per lane) were separated by electrophoresis and transferred for detection. Information on sample preparation, antibody usage, and controls is available in the Online Resource.
Statistical analysis
All statistical analyses were performed at the individual animal level. PCB 11 and metabolite levels are reported as mean ± SD. RT-PCR, immunoblot, and P450 abundance data were analyzed by one-way ANOVA using GraphPad Prism v10.1 (P ≤ 0.05). A Grubbs’ test was applied to identify potential single-point outliers within normally distributed datasets. One sample was identified as an outlier in CYP2C23 protein abundance and was excluded to prevent disproportionate results on the group means and comparison. For additional details on statistical procedures, refer to the Online Resource.
Results
Tissue distribution and detection of PCB 11 and its metabolites in maternal dams
PCB 11 is metabolized by P450 enzymes into OH-PCB 11 (Kaminsky et al. 1981), which undergoes further biotransformation, including oxidation, sulfation, glucuronidation, and methylation (Zhang et al. 2020). To assess the distribution of PCB 11 and its metabolites in mouse dams, levels of PCB 11 and OH-PCB 11 metabolites were measured in the brain, liver, and serum of female dams at PD21 using GC–MS/MS and LC-HRMS (Table 1). In brain tissue, the parent congener PCB 11 and its hydroxylated metabolites were below detection limits, with exception of 5-OH-PCB 11, which was detected in 33% of samples from the high-dose group (0.88 ± 0.03 ng/g wet weight [ww]). In the liver, PCB 11 was detected in a few samples in both exposure groups (33% of low-dose and 17% of high-dose samples). 4-OH-PCB 11 (119 ng/g ww) and 5-OH-PCB 11 (2.3 ng/g ww) were each observed with a detection frequency of 17% in low-dose samples. In the serum, PCB 11 was detected in 17% of low-dose samples, and several hydroxylated metabolites were detected. 4-OH-PCB 11 was the most frequently detected PCB 11 metabolite in serum, with a detection frequency of 33% in low-dose samples and 17% in high-dose samples. Additionally, 5-OH-PCB 11 and 6-OH-PCB 11 were detected in 17% and 67% of high-dose samples, respectively.Table 1. Levels (ng/g wet weight) and detection frequency (%) of PCB 11 and its metabolites determined by GC–MS/MS and LC-HRMS across matrices for the low and high PCB 11 exposure groups.^a^MatrixAnalyte classSpecific analyteAnalytical method^b^Limits of detection^c^Levels^d^ (ng/g wet weight)Number of detects/number of samples (detection frequency, %)LowHighLowHighBrainParent PCBPCB 11GC–MS/MS24.0 < LOD < LOD––OH-PCB2-OH-PCB 112.0 < LOD < LOD––4-OH-PCB 1113.0 < LOD < LOD––5-OH-PCB 111.0 < LOD0.88 ± 0.03–2/6 (33)6-OH-PCB 111.0 < LOD < LOD––LiverParent PCBPCB 116.024 ± 1472/6 (33)1/6 (17)OH-PCB2-OH-PCB 112.0 < LOD < LOD––4-OH-PCB 1110.0119 < LOD1/6 (17)–5-OH-PCB 110.32.3 < LOD1/6 (17)–6-OH-PCB 110.3 < LOD < LOD––SerumParent PCBPCB 115.05.9 < LOD1/6 (17)–OH-PCB2-OH-PCB 111.0 < LOD < LOD––4-OH-PCB 1143.044 ± 0.31892/6 (33)1/6 (17)5-OH-PCB 112.0 < LOD8-1/6 (17)6-OH-PCB 110.31.21.9 ± 1.21/6 (17)4/6 (67)SerumOH-PCBX_1_LC-HRMSS/N < 3NANA2/6 (33)5/6 (83)X_2_2/6 (33)5/6 (83)X_3_0/6 (0)2/6 (33)PCB sulfateY_1_3/6 (50)6/6 (100)Y_2_2/6 (33)6/6 (100)OH-PCB sulfateZ_1_3/6 (50)6/6 (100)Z_2_3/6 (50)6/6 (100)^a^Low = 1.0 mg/kg; High = 6.0 mg/kg; n = 6 per group^b^GC-MS/MS = gas chromatography mass spectrometry/mass spectrometry; LC-HRMS = liquid chromatography-high resolution mass spectrometry^c^LOD, limit of detection (ng/g wet weight), was calculated based on matrix blanks for each tissue using the formula LOD = mean_blank_ + t_0.01, n-1_ *SD_blank_, where mean_blank_ is the mean of blank measures, t_0.01, n-1_ is Student’s t-value for n – 1 degrees of freedom at the 99% confidence level, and SD_blank_ is the standard deviation of the blank measures; n = 6; S/N = signal to noise ratio^d^Mean ± standard deviation; NA = data not available
LC-HRMS analysis revealed three OH-PCB 11 peaks, X_1_, X_2_, and X_3_ ([M-H]^−^, m/z 236.98767; Fig. 1, A1 and A2) in serum at retention times of 8.11, 8.21, and 8.34 min, respectively. The detection frequencies of X_1_ and X_2_ OH-PCB 11 peaks were 33% and 83% in both the low-dose and high-dose groups, respectively. The X_3_ metabolite was only detected in serum from the high-dose group, with a detection frequency of 33%. The detection frequencies of the OH-PCB 11 metabolites in the LC-HRMS analysis differ from those observed in the GC–MS/MS analysis, most likely due to differences in the limits of detection between the two analytical methods or a limitation in the liquid–liquid extraction protocol used in the GC–MS/MS analysis workflow. Additionally, two peaks corresponding to PCB 11 sulfates, Y_1_ and Y_2_ ([M-H-SO_3_]^−^, m/z 236.98800, [M-H]^−^, m/z 316.94434; Fig. 1, B1 and B2), were observed at 6.66 and 6.78 min, respectively. The PCB 11 sulfate Y_1_ peak was detected in 50% of low-dose samples, whereas both Y_1_ and Y_2_ metabolites were observed in 100% of the high-dose group. Two OH-PCB 11 sulfate peaks, Z_1_ and Z_2_ ([M-H-SO_3_]^−^, m/z 252.98254; [M-H]^−^, m/z 332.93951; Fig. 1, C1 and C2), were detected at 6.68 and 6.89 min, respectively. Both OH-PCB 11 sulfate peaks were detected in 50% and 100% of low- and high-dose exposure samples, respectively. These findings suggest that PCB 11 and its metabolites do not accumulate at detectable levels in the dams at the time point investigated, likely due to the rapid elimination of PCB 11 in laboratory animals (Hu et al. 2013; Zhang et al. 2021).Fig. 1. Identification of putative PCB 11 metabolites in serum of exposed postpartum dams. Representative ion chromatograms and mass spectrometric data support the presence of (A) OH-PCB 11 and (B) PCB 11 sulfate in a high-dose sample and (C) OH-PCB 11 sulfate metabolite in a low-dose sample. (A1) OH-PCB 11 metabolites detected at retention times 8.11 min, 8.21 min, and 8.34 min (m/z 236.98767); (A2) accurate masses of isotope ions corresponding to OH-PCB 11 eluting at 8.21 min; (B1) PCB 11 sulfate detected at 6.66 min and 6.78 min (m/z 316.94434); (B2) accurate masses of PCB 11 sulfate isotope ions and a fragment ion ([M-H-SO_3_]^−^, m/z 236.98800) eluting at 6.66 min; (C1) OH-PCB 11 sulfate detected at 6.68 min and 6.89 min (m/z 332.93951); (C2) accurate masses of isotope ions and a fragment ([M-H-SO_3_]^−^, m/z 252.98254) of OH-PCB 11 sulfate eluting at 6.89 min. (D) Proposed metabolic pathway of PCB 11 in dam serum samples. CYP: cytochrome; SULT: sulfotransferaseFig. 2Hepatic proteome alterations induced by low and high doses of PCB 11 in postpartum day 21 dams. Volcano plots show differentially expressed hepatic proteins following (A) low-dose (1.0 mg/kg; n = 5) and (B) high-dose (6.0 mg/kg; n = 5) PCB 11 exposure, with significance defined by FDR ≤ 0.05. The top ten KEGG pathway enrichments identified via STRING analysis are presented for (C) low-dose and (D) high-dose exposure groups
Impact of PCB 11 exposure during pregnancy on the maternal hepatic proteome
While HC-PCBs are known to alter the hepatic proteome (Rignall et al. 2013), the effects of LC-PCBs, such as PCB 11, on the hepatic proteome remain poorly investigated. Furthermore, the impact of exposure to PCBs during pregnancy and lactation on the maternal liver proteome has not been characterized. To address these knowledge gaps, we first performed global proteomic profiling of liver tissue, followed by an analysis of drug-metabolizing enzymes in dams exposed to PCB 11 throughout pregnancy and lactation, to determine whether PCB 11 exposure elicits proteomic changes similar to those induced by other PCB congeners.
Global alterations of the liver proteome following PCB 11 exposure during pregnancy and lactation
Principal component analysis (PCA) revealed distinct separation between vehicle control and both PCB 11 exposure groups (Online Resource Figure S1). Relative to the vehicle controls, 123 hepatic proteins were significantly altered (FDR ≤ 0.05; Fig. 2A and Online Resource Table S9) in the low-dose group, with 3 increased and 120 decreased. The most decreased proteins included dihydrolipoamide S-acetyltransferase (DLAT), ubiquitin carboxyl-terminal hydrolase 26 (USP26), malic enzyme 1 (ME1), cytochrome P450 2D10 (CYP2D10), and SEC23 homolog A, COPII coat complex component (SEC23A). The most increased proteins were pre-mRNA processing factor 38A (PRPF38A), solute carrier family 6 member 13 (SLC6A13), and elastin microfibril interfacer 1 (EMILIN1).
In the high-dose group, 234 proteins were significantly altered (FDR ≤ 0.05; Fig. 2B, Online Resource Table S12), with 56 increased and 178 decreased. The top five decreased proteins were serine protease 1 (PRSS1) and 2 (PRSS2), chymotrypsin-like (CTRL), Ras-related C3 botulinum toxin substrate 1 (RAC1), and DLAT. The top five increased proteins were solute carrier family 6 member 13 (SLC6A13), CCR4-NOT transcription complex subunit 4 (CNOT4), tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein epsilon (YWHAE), asparaginyl-tRNA synthetase 1 (NARS1), and chromodomain helicase DNA binding protein 1 (CHD1).
KEGG pathway analysis revealed enrichment of differentially abundant proteins (FDR ≤ 0.05) in the low- (Fig. 2C; Online Resource Table S10) and high-dose (Fig. 2D; Online Resource Table S13) exposure groups across several hepatic metabolic pathways. Key pathways included “metabolic pathways”, “fatty acid degradation”, “carbon metabolism”, “tryptophan metabolism”, “beta-alanine metabolism”, “valine/leucine/isoleucine degradation”, “histidine metabolism”, “biosynthesis of amino acids”, “pyruvate metabolism”. “Fatty acid metabolism” was enriched only in the low-dose group, while the “TCA cycle” was specific to the high-dose group. Protein–protein network analyses for low- and high-group exposure are depicted in the Online Resource Figures S2 and S3, respectively. Comparison of maternal hepatic proteins between exposure doses (Online Resource Figure S4, Table S15) identified 119 common proteins (FDR ≤ 0.05) enriched in pathways related to amino acid, energy, and xenobiotic metabolism, including drug-processing enzymes such as acyl-CoA’s, epoxide hydrolases, carboxylesterases, P450s, and flavin-containing monooxygenases.
Disruption of hepatic phase I biotransformation in the maternal proteome
Effect of PCB 11 exposure on liver P450 enzymes
PCBs are known to induce hepatic P450 enzymes (Robertson et al. 1984), but the effects of PCB 11 during pregnancy and lactation remain poorly understood. Using global proteomics, we identified eleven hepatic P450 enzymes altered by maternal PCB 11 exposure, irrespective of the dose (Fig. 3). Compared to vehicle controls, both low- and high-dose PCB 11 exposure significantly reduced (FDR ≤ 0.05) the protein abundance of CYP1A2, CYP2C23, CYP2C29, CYP2C40, CYP2C70, CYP2D10, CYP2E1, and CYP3A41, enzymes critical for phase I biotransformation and steroid biosynthesis, suggesting a broad suppression of hepatic biotransformation capacity by PCB 11 in mouse dams exposed via the diet.Fig. 3. Protein abundance of cytochrome P450 enzymes in female adult wildtype mice exposed to PCB 11. Total protein abundance of P450 enzymes was measured in vehicle (sterile peanut butter; 0 mg/kg; n = 5; green), low (1.0 mg/kg; n = 5; yellow) and high (6.0 mg/kg; n = 5; red) exposure groups. P450 enzyme families detected include CYP1A (A), CYP2C subtypes (B-F), CYP2D (G), CYP2E (H), CYP3A subtypes (I-K). Outliers were identified with Grubbs’ test, resulting in the removal of one sample from CYP2C23 protein analyses. Data were analyzed using ordinary one-way ANOVA followed by Dunnett’s multiple comparison test with GraphPad Prism v10.1
We did not detect CYP2B enzymes in the liver, irrespective of the experimental group. Given that many NDL-PCBs induce CYP2B enzymes (Uwimana et al. 2019), CYP2B10 expression was specifically examined but was absent in the proteomics analysis. PCB 11 exposure did not significantly affect (P > 0.05) CYP2B10 mRNA expression, assessed by RT-PCR (Online Resource Figure S5), or protein levels, assessed by western blotting (Online Resource Figure S6).
STRING and KEGG pathway analysis of drug-metabolizing enzymes
To further examine PCB 11 effects on hepatic drug metabolism, STRING analysis was performed at a confidence threshold score ≥ 0.900. Eight enzymes were significantly altered in the low-dose group (CYP1A2, CYP2C23, CYP2C29, CYP2C40, CYP2C70, CYP2E1, EPHX1, EPHX2; FDR ≤ 0.05; Fig. 4A), and nine in the high-dose group, including the eight drug metabolizing enzymes plus CES1D (FDR ≤ 0.05; Fig. 4C).Fig. 4. Characterization of cytochrome P450 and steroid hormone biosynthesis pathway in female adult mice exposed to PCB 11. Protein–protein interaction (PPI) network and KEGG pathway enrichment analyses were used to assess hepatic drug metabolism pathways following PCB 11 exposure. Panels (A, B) depict the PPI network (interaction score ≥ 0.900) and significantly enriched KEGG pathways (FDR ≤ 0.05) in the low-dose group (1.0 mg/kg; n = 5; yellow), while panels (C, D) show high-dose exposure (6.0 mg/kg; n = 5; red). Panel (E) highlights the steroid biosynthesis pathway and enzymes identified between both exposures
KEGG pathway enrichment of these proteins subsequently identified overlapping pathways in both PCB 11 exposure groups, including “steroid hormone biosynthesis,” “chemical carcinogenesis,” “arachidonic acid metabolism,” “retinol metabolism,” “serotonergic synapse,” “metabolic pathways,” “inflammatory mediator regulation of TRP channels,” “metabolism of xenobiotics by P450s,” and “drug metabolism – cytochrome P450” (FDR ≤ 0.05; Fig. 4B and D; Online Resource Tables S11 and S14).
“Steroid hormone biosynthesis” was enriched in both groups, suggesting potential endocrine-disrupting effects of PCB 11. While PCBs are established endocrine disruptors (Streifer et al. 2024), evidence for PCB 11 and/or its metabolites remains limited. In the low-dose group, seven hepatic enzymes (CYP1A2, CYP2C23, CYP2C29, CYP2C40, CYP2C70, CYP2D10, and CYP2E1; Fig. 4E) are involved in converting dehydroepiandrosterone (DHEA) to 16α-hydroxydehydroepiandrosterone (16α-OH-DHEA). In the high-dose group, these enzymes plus hydroxysteroid 11-beta dehydrogenase 1 (HSD11B1; Fig. 4E) were altered, supporting disruption of steroid hormone metabolism following PCB 11 exposure during pregnancy and lactation.
Discussion
This study reports the first maternal hepatic proteomic characterization following PCB 11 exposure during pregnancy and lactation, revealing disruptions in mitochondrial and fatty acid metabolism (e.g., DLAT, ACSL family), suppression of P450 enzymes, and perturbation of steroidogenic pathways. These observations provide novel mechanistic insights into how non-legacy LC-PCBs, such as PCB 11, can alter maternal detoxification and endocrine homeostasis, advancing our understanding of PCB effects on the mammalian liver proteome.
PCB 11 is an LC-PCB frequently detected in environmental matrices and consumer products (Liu et al. 2022). Human exposure to PCB 11 can occur through ingestion and inhalation (Ampleman et al. 2015; Grimm et al. 2015). PCB 11 and its hydroxylated metabolites were present at higher concentrations in the serum from exposed dams than those previously reported in human samples. In the current study, PCB 11 concentrations were mostly below the limit of detection, except for one sample with a PCB 11 level of 5.9 ng/g ww. PCB 11 was identified as a major congener in the serum of pregnant women from the Markers of Autism Risk in Babies – Learning Early Signs (MARBLES) cohort, with a mean concentration of 0.490 ng/mL (range: 0.005 to 1.717 ng/mL) (Sethi et al. 2017). Hydroxylated PCB 11 metabolites were identified in this study, including 4-OH-PCB 11 at concentrations of up to 189 ng/g wet weight, 5-OH-PCB 11 at 8 ng/g wet weight, and 6-OH-PCB 11 at 1.9 ± 1.2 ng/g wet weight. In contrast, a published study of human serum reported only trace levels of PCB 11 metabolites (0.001 to 0.006 ng/g ww); however, interpretation of those findings is limited by the small sample size (three donors) (Zhu et al. 2013).
Furthermore, hydroxylated and sulfated PCB 11 metabolites have been detected in serum from populations in the United States (Grimm et al. 2015; Zhang et al. 2022). Unlike HC-PCBs, which bioaccumulate in human serum and adipose tissue (Mussalo-Rauhamaa 1991), PCB 11 undergoes rapid biotransformation into hydroxylated and sulfated metabolites (Hu et al. 2013; Zhang et al. 2021). P450 enzymes efficiently oxidize PCB 11 to OH-PCBs (Kennedy et al. 1981), which are subsequently sulfated by SULT1A1 and SULT2A1 to form PCB sulfates (Duffel and and Lehmler 2024). In HepG2 cells, PCB 11 was metabolized to complex metabolite mixtures (Zhang et al. 2020), and in vivo studies detected an OH-PCB 11 sulfate in mouse serum and liver after oral exposure (Zhang et al. 2021). Similarly, our study observed low detection frequencies of PCB 11 and its metabolites in the dam serum and tissues. As only a single time point was analyzed in this and earlier studies, the toxicokinetics of PCB 11 and its metabolites in pregnant or lactating rodents remains unknown; future time-series experiments will be essential to fully characterize the relationship between PCB 11 exposure and downstream proteomic responses.
Global proteomic profiling revealed consistent downregulation of DLAT, a core component of the pyruvate dehydrogenase complex involved in cellular energy metabolism (Sutendra et al. 2014), and upregulation of solute carrier SLC6A13, a GABA transporter (Schlessinger et al. 2012), following PCB 11 exposure. DLAT also mediates cuproptosis (Tsvetkov et al. 2022), a copper-dependent form of cell death, though its relationship with PCB exposure remains unexplored. Interestingly, DLAT was downregulated in the intestinal proteome of Takifugu rubripes exposed to increasing copper in a dose-dependent manner (Xia et al. 2024), suggesting that PCB 11 may similarly disrupt copper homeostasis, potentially inducing proteotoxic stress and cell death (Solmonson and DeBerardinis 2018). Future research should examine how the copper-binding activity of DLAT may be influenced by PCB-induced oxidative stress or mitochondrial dysfunction. Additionally, SLC family genes, including SLC6A13, regulate the uptake of endogenous compounds and xenobiotics (Girardi et al. 2020). Prior studies in female mice exposed to PCB mixtures reported dose-dependent downregulation of hepatic SLC gene transcripts (Lim et al. 2020), supporting the broader impact of PCBs on solute carrier expression, an area that merits further investigation.
To investigate the metabolic consequences of PCB 11, we conducted network analyses of differentially expressed hepatic proteins, revealing enrichment in mitochondrial and lipid metabolism pathways. Acyl-CoA synthetases activate fatty acids for beta-oxidation and lipid synthesis (Yan et al. 2015). PCB 11 exposure reduced hepatic expression of long-chain (ACSL1, ACSL5) and medium-chain (ACSM1) acyl-CoA synthetases. Knockout studies of Acsl1 (Li et al. 2009) in mice and Acsl5 in rat primary hepatocytes (Bu and Mashek 2010) have shown diminished synthesis of triglycerides, suggesting PCB 11 may disrupt lipid metabolism and mitochondrial function in postpartum dams. ACSM1 reduction may also impair glycine conjugation of medium-chain fatty acids (Fouché et al. 2025). In a Leigh syndrome mouse model, a human disorder that exhibits infantile neurodevelopmental deficits, hepatic dysfunction, and liver failure, hepatic Acsm1 transcript levels were downregulated (Fouché et al. 2025). Thus, reduced Acsm1 abundance following PCB 11 exposure may also suggest an impairment to glycine conjugation and elimination of medium-chain fatty acids in postpartum dams, potentially impacting offspring development. Interestingly, this contrast with in vitro studies in SIRT3-deficient fibroblasts, where 3,3'-dichlorobiphenyl-4-ol, a PCB 11 metabolite, increased expression of fatty acid metabolism genes (e.g., Acsbg2, Acsm2, Acsl1, Slc27a5, etc*.*) (Alam et al. 2018). Combined with decreased DLAT expression, our findings suggest PCB 11 impairs hepatic fatty acid metabolism, promoting metabolic dysfunction through mechanisms distinct from those observed in cell-based models.
The liver is the primary site for PCB metabolism and detoxification (Wahlang et al. 2017). Isoform-dependent changes in hepatic P450 enzyme activity during pregnancy and lactation can influence PCB toxicokinetics (Anderson 2005). HC-PCBs affect P450 expression in adult rodents in a congener-specific way, but the effects of PCB 11 on hepatic P450 levels are unknown. Here, we investigated drug-metabolizing enzymes observed in the hepatic proteome of mouse dams exposed to PCB 11 throughout pregnancy and lactation. Several P450 enzymes (CYP1A2, CYP2C23, CYP2C29, CYP2C40, CYP2C70, CYP2D10, CYP2E1, and CYP3A41) were downregulated following PCB 11 exposure. CYP2B enzymes were found to be below the detection limit of the proteomics analysis; however, CYP2B10 expression was detected in the liver at both the mRNA and protein levels. Notably, exposure to PCB 11 did not have a significant effect on CYP2B10 expression. These findings indicate that PCB 11 exposure does not activate the AhR, which is typically linked to increased CYP1A enzyme expression in the rodent liver after exposure to DL-PCBs (Jin et al. 2021). Additionally, it does not activate the canonical CAR/PXR pathway, which is responsible for inducing CYP2B in the mouse liver following exposure to NDL-PCBs (Al-Salman and Plant 2012; Wahlang et al. 2015).
PCB 11 is oxidized by CYP1A and CYP2B enzymes (Kennedy et al. 1981); however, the roles of other P450 subfamilies in metabolizing LC-PCBs remain unknown. Experimental evidence suggests that CYP1A2 and CYP2E1 participate in LC-PCB metabolism. For example, PCB 28 undergoes hydroxylation by these enzymes in Drosophila melanogaster (Idda et al. 2020), and PCB 20 and PCB 22 induce cytotoxicity via CYP2E1- and CYP1A2-mediated bioactivation in human L-02 hepatocytes (Liu et al. 2017). Given the potential roles of CYP1A2 and CYP2E1 in PCB metabolism, their decreased abundance suggests that PCB 11 exposure may alter its own metabolism capacity during sensitive physiological windows, such as pregnancy and lactation. However, despite reduced P450 expression, PCB 11 was still rapidly eliminated from mouse dams.
PCB 11 exposure also affected hepatic proteins involved in steroid biosynthesis, suggesting an endocrine-disrupting potential of PCB 11. Within the steroid biosynthesis pathway, PCB 11 exposure was associated with a reduced abundance of seven enzymes involved in the conversion of dehydroepiandrosterone (DHEA) to 16α-hydroxy-DHEA, a key intermediate in endocrine homeostasis. This reduction in P450 enzyme levels may result in a reduction in the production of 16α-hydroxy-DHEA and altered downstream steroid hormone levels. For example, the 16α-hydroxy-DHEA serves as a precursor for estriol, which plays a critical role in pregnancy and fetal development (Frederiksen et al. 2020). Additionally, a positive correlation was observed between PCB and DHEA levels in human breast milk, indicating that PCBs may disrupt endocrine function and affect DHEA in fetal offspring (Rennert et al. 2012). Lactational metabolic adaptations suppress P450 expression (He et al. 2005); accordingly, HSD11B1 was downregulated in high-dose PCB 11 exposure, indicating broader disruption of hepatic steroid metabolism.
Our findings indicate that, in contrast to HC-PCBs, PCB 11 exposure during pregnancy and lactation may impair maternal detoxification mechanisms, potentially increasing fetal and neonatal susceptibility. While the systemic accumulation of PCB 11 and its metabolites was not observed, proteomic alterations in the liver indicate disruptions in energy homeostasis and endocrine function. Limitations of this study include the focus on a single PCB congener and the lack of cell-type resolution in the proteomic analysis. To build on these results, future studies should incorporate integrated omics approaches, combining proteomics with metabolomics and transcriptomics, to elucidate the mechanistic underpinnings of maternal PCB 11 toxicity following exposure to this LC-PCB during pregnancy and lactation. Moreover, investigating sex-specific effects and the functional consequences of enzyme dysregulation in the liver following maternal exposure to PCB 11 will also be crucial to fully characterize the risks to the dam and offspring associated with this and other LC-PCBs.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1