Early Life Polychlorinated Biphenyl 126 Exposure Disrupts Gut Microbiota and Metabolic Homeostasis in Mice Fed with High-Fat Diet in Adulthood

Evidence supports the potential influence of persistent organic pollutants (POPs) on the pathogenesis and progression of obesity and diabetes. Diet-toxicant interactions appear to be important in diet-induced obesity/diabetes; however, the factors influencing this interaction, especially the early life environmental exposure, are unclear. Herein, we investigated the metabolic effects following early life five-day exposure (24 μg/kg body weight per day) to 3,3′,4,4′,5-pentacholorobiphenyl (PCB 126) at four months after exposure in mice fed with control (CTRL) or high-fat diet (HFD). Activation of aryl hydrocarbon receptor (AHR) signaling as well as higher levels of liver nucleotides were observed at 4 months after PCB 126 exposure in mice, independent of diet status. Inflammatory responses including higher levels of serum cytokines and adipose inflammatory gene expression caused by early life PCB 126 were observed only in HFD-fed mice in adulthood. Notably, early life PCB 126 exposure worsened HFD-induced impaired glucose homeostasis characterized by glucose intolerance and elevated gluconeogenesis and tricarboxylic acid (TCA) cycle flux without worsening the effects of HFD related to adiposity in adulthood. Furthermore, early life PCB 126 exposure resulted in diet-dependent changes in bacterial community structure and function later in life, as indicated by metagenomic and metabolomic analyses. These data contribute to a more comprehensive understanding of the interactions between diet and early life environmental chemical exposure.


Introduction
Persistent organic pollutants (POPs) are toxic chemicals that are resistant to environmental degradation [1]. Most POPs are hydrophobic and poorly metabolized, thus making such compounds susceptible to bioaccumulation in animal and human fat [2]. POPs are toxic to human health, causing developmental defects, metabolic diseases, cancer, and, in some cases, death [3][4][5].

Histopathology
Formalin-fixed liver tissues were embedded in paraffin wax, sectioned, and stained with hematoxylin and eosin (H&E). Oil red O staining of liver sections was performed on optimal cutting temperature (OCT)-embedded frozen liver sections by Histoserv, Inc. (Germantown, MD, USA).

Blood Clinical Biochemistry and Cytokine Analysis
Liver injury markers, including serum alanine transaminase (ALT) and alkaline phosphatase (ALP), were assessed using the VetScan VS2 Chemistry Analyzer and the Mammalian Liver Profile (Abaxis Inc., Union City, CA, USA). Serum cytokine levels were assessed with a BioPlex 200 mouse cytokine array/chemokine array 32-Plex by Eve Technologies (Calgary, AB, Canada).

Histopathology
Formalin-fixed liver tissues were embedded in paraffin wax, sectioned, and stained with hematoxylin and eosin (H&E). Oil red O staining of liver sections was performed on optimal cutting temperature (OCT)-embedded frozen liver sections by Histoserv, Inc. (Germantown, MD, USA).

Blood Clinical Biochemistry and Cytokine Analysis
Liver injury markers, including serum alanine transaminase (ALT) and alkaline phosphatase (ALP), were assessed using the VetScan VS2 Chemistry Analyzer and the Mammalian Liver Profile (Abaxis Inc., Union City, CA, USA). Serum cytokine levels were assessed with a BioPlex 200 mouse cytokine array/chemokine array 32-Plex by Eve Technologies (Calgary, AB, Canada).

Glucose Tolerance Test
The glucose tolerance test was performed as previously described [22] and followed a 6 h fast. In brief, blood glucose was measured following intraperitoneal injection of 40% glucose (2.0 g/kg body weight). The blood glucose concentrations were measured using OneTouch Ultra 2 Meter (LifeScan, Malvern, PA, USA).

Liver Triglyceride and Glutathione Quantification
Liver triglyceride was measured using the Triglyceride Colorimetric Assay Kit according to the manufacturer's instructions (Cayman Chemical, Ann Arbor, MI, USA). The ratios of reduced glutathione (GSH) and oxidized glutathione (GSSG) were measured with the GSH/GSSG Ratio Detection Assay Kit according to the manufacturer's recommendations (Abcam, Cambridge, UK).

Tissue RNA Isolation and qPCR
Total RNA was isolated from liver, adipose, and intestine tissues by TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. cDNA was synthesized from 1 µg of RNA using qScript cDNA SuperMix (Quanta Biosciences, Gaithersburg, MD, USA). Quantitative PCR (qPCR) reactions were performed using SYBR green QPCR master mix with QuantStudio 3 Real-Time PCR system (Thermo Fisher Scientific, Waltham, MA, USA). The primers used were listed in Supplementary Table S2. The data were normalized to Gapdh mRNA levels using the ∆∆C T method.

1 H NMR Based Metabolomics Experiments
Sample preparations for cecal content, liver, and urine samples were performed as previously described [23]. All 1 H NMR spectra were recorded using a Bruker Avance NEO 600 MHz spectrometer equipped with an inverse cryogenic probe (Bruker Biospin, Germany) at 298 K. NMR spectra were acquired with a typical 1D NMR spectrum named NOESYPR1D. The metabolites were confirmed with a set of 2D NMR spectra and published results [23,24]. 1 H NMR spectra for liver lipids were corrected for phase and baseline distortions, referenced to TMS (δ = 0.0), and integrated using TopSpin 3.6 and AMIX 3.9 (Bruker Biospin, Ettlingen, Germany). The absolute quantification for lipid classes in liver were calculated as previously reported [24]. The spectra process and quantification for liver hydrophilic and cecal bacterial metabolites were performed using Chenomx NMR Suite (Chenomx Inc., Edmonton, AB, Canada).

Bile Acid Quantitation by UPLC-MS/MS
Quantitative analysis of bile acids in feces was performed with an Acquity UPLC system coupled to a Waters Xevo TQS MS with an ACQUITY C8 BEH (2.1 × 100 mm, particle size 1.7 µm) UPLC column (all from Waters, Milford, MA, USA) [25]. The analytes were detected by multiple reaction monitoring (MRM) or selected ion monitoring (SIM) and normalized by their respective internal deuterated standard. The results were quantified by comparing the integrated peaks against a standard curve.

Metagenomic Analysis
DNA was extracted from cecal contents using the E.Z.N.A. Stool DNA Kit (Omega Bio-Tek Inc., Norcross, GA, USA). Cecal DNA samples were submitted to the Pennsylvania State University Genomics Core Facility for NextSeq Mid-Output 150 × 150 paired end sequencing. The obtained demultiplexed reads were checked for quality using FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (accessed on 5 July 2022)). To remove low-quality reads and the "contaminant" host reads from the metagenomic sequences, reads were trimmed using trimmomatic [26] and then aligned with the genome of C57BL/6J mice and filtered using Kneaddata [27]. Clean metagenomic sequence reads were analyzed using the Kraken2 taxonomic sequence classification approach on a standard Kraken database comprising all complete bacterial, viral, and archeal genomes in RefSeq [28]. The abundance of the various species was estimated using Bracken [29]. For functional classification, the reads were concatenated and then processed with default settings using HUMAnN3 [30]. Briefly, the reads were mapped to uniref90 database using DIAMOND [31], and the relative abundance of each gene family was quantified. The uniref-based gene family abundances were regrouped to KEGG orthologous groups, which were further used to make inferences on metabolic pathways.

Statistics
All data values are expressed as mean ± standard deviation (SD) or median and interquartile range. Graphical illustrations and two-tailed unpaired t-test analyses were performed using GraphPad Prism 6.0 (GraphPad, San Diego, CA, USA).

Early Life PCB 126 Exposure Persistently Affects AHR Signaling and Liver Nucleotide Metabolism in Both CTRL and HFD-Fed Mice in Adulthood
Early life PCB 126 exposure (24 µg/kg body weight per day for five days) had no persistent effect on body weight or liver histological analysis in mice fed a CTRL or HFD in adulthood (Supplementary Figure S2). As expected, and confirming the considerably long half-life of PCB 126 in the rodents [32], increased mRNA expression of AHR target genes was observed in liver and ileal samples from CTRL and HFD-fed mice at four months after PCB 126 exposure ( Figures 1B and S3). 1 H NMR analysis showed significantly higher levels of liver purines and pyrimidines from both CTRL and HFD-fed mice with PCB 126 exposure in adulthood ( Figure 1C,D). Early life PCB 126 exposure resulted in the significantly higher levels of adenosine monophosphate (AMP), adenosine diphosphate (ADP), adenosine triphosphate (ATP), xanthine, and cytidine 5 -monophosphate (CMP) in the liver of mice fed with HFD and higher levels of xanthine and uridine in the liver of mice fed with CTRL in adulthood ( Figure 1C,D).

Early Life PCB 126 Exposure Causes Inflammatory Response in HFD-Fed Mice in Adulthood
No significant changes in blood biochemical markers were observed in CTRL or HFDfed mice with early life PCB 126 exposure ( Figure 2A). Early life PCB 126 exposure resulted in the significantly higher ratios of oxidized glutathione (GSSG) to reduced glutathione (GSH) in the liver of mice with HFD but not in CTRL-fed mice in adulthood ( Figure 2B). Early life PCB 126 resulted in significant changes in two serum cytokines, including a lower level of leukemia inhibitory factor (LIF) and a higher level of lipopolysaccharide-induced CXC chemokine (LIX) in mice with HFD later in life but no changes in mice with CTRL ( Figure 2C and Supplementary Table S3). The gene expression of the proinflammatory cytokines in adipose tissues, including tumor necrosis factor alpha (Tnfα) and NAD(P)H quinone dehydrogenase 1 (Nqo1), was significantly increased in early life PCB 126 exposure mice with HFD but not in CTRL-fed mice in adulthood ( Figures 2D and S4).

Early Life PCB 126 Exposure Does Not Worsen the Effects of HFD Related to Adiposity in Adulthood
Oil red O staining of the liver showed no obvious changes, such as increased lipid accumulation and droplet size, in both CTRL and HFD-fed mice in adulthood with PCB 126 exposure ( Figure 3A). Consistently, the quantitative 1 H NMR analysis and triglyceride assay kit showed no significant changes in hepatic lipids, including fatty acids, in CTRL or HFD-fed mice in adulthood with early life PCB 126 exposure ( Figure 3B-D). Consistent with the metabolomics data, no significant changes in the mRNA expression of genes involved in de novo fatty acid biosynthesis were observed in the liver from CTRL or HFD-fed mice with PCB 126 exposure ( Figure 3E).

Early Life PCB 126 Exposure Aggravates Impaired Glucose Homeostasis in HFD-Fed Mice in Adulthood
POPs have recently been linked with the development of type 2 diabetes and other metabolic diseases [33]. Quantitative 1 H NMR analysis revealed that early life PCB 126

Early Life PCB 126 Exposure Aggravates Impaired Glucose Homeostasis in HFD-Fed Mice in Adulthood
POPs have recently been linked with the development of type 2 diabetes and other metabolic diseases [33]. Quantitative 1 H NMR analysis revealed that early life PCB 126 exposure resulted in significantly higher levels of metabolites involved in the tricarboxylic acid (TCA) cycle, including pyruvate, aspartate, fumarate, succinate, citric acid, glutamate, and α-ketoglutarate, in the liver from HFD-fed mice in adulthood ( Figure 4A). Notably, the glucose tolerance test showed impaired glucose tolerance in HFD-fed mice later in life with early life PCB 126 exposure ( Figure 4B). No significant changes were observed in metabolites involved in the TCA cycle or glucose tolerance test in CTRL-fed mice with early life PCB 126 exposure ( Figure 4A,B). The mRNA expression of genes involved in gluconeogenesis was also significantly higher in the liver of HFD-fed mice with early life PCB 126 exposure ( Figure 4C). Together, these results indicated that early life exposure to PCB 126 aggravates impaired glucose homeostasis in HFD-fed mice in adulthood.

Early Life PCB 126 Exposure Results in Diet-Dependent Changes in Bacteria Community, Gene Levels, and Metabolism in Adulthood
In order to investigate the influence of early life PCB 126 exposure on the gut microbiota with two diets, metagenomics and 1 H NMR-and mass-spectrometry-based metabolomics analysis was performed. Metagenomic data showed that early life PCB 126 exposure resulted in significant changes in bacteria community and gene levels that were specific in two different diets in adulthood ( Figure

Early Life PCB 126 Exposure Results in Diet-Dependent Changes in Bacteria Community, Gene Levels, and Metabolism in Adulthood
In order to investigate the influence of early life PCB 126 exposure on the gut microbiota with two diets, metagenomics and 1 H NMR-and mass-spectrometry-based metabolomics analysis was performed. Metagenomic data showed that early life PCB 126 exposure resulted in significant changes in bacteria community and gene levels that were specific in two different diets in adulthood ( Figures 5A-C and S5). Early life PCB 126 exposure resulted in a significant decrease in the relative abundance of Bacteroidetes phyla and a higher ratio of Firmicutes/Bacteroidetes in cecal contents from mice with HFD in adulthood (Figures 5B and S5A). No pronounced effect in the ratio of Firmicutes to Bacteroidetes but significant changes in Firmicutes and Verrucomicrobia were observed in CTRL-fed mice with early life PCB 126 exposure (Figures 5B and S5A). Early life PCB 126 exposure resulted in significant decreases in the relative abundances of genera Lacrimispora, Anaerocolumna, Anaerobutyricum, Anaerostipes, Mediterraneibacter, Blautia, Butyrivibrio, Bacillus, Clostridioides, Hungatella, Eubacterium, Lachnoclostridium, Ruthenibacterium, Roseburia, Streptococcus, and Caproiciproducens but increases in the relative abundance of genus Akkermansia in CTRL-fed mice; and significant decreases in the relative abundances of genera Muribaculum, Duncaniella, Bacteroides, Parabacteroides, and Prevotella but increases in the relative abundances of genera Romboutsia and Adlercreutzia in HFD-fed mice ( Figure 5A and Supplementary Tables S4 and S5). We also analyzed the microbial metabolic pathway based on the KEGG database. Early life PCB 126 resulted in the induced expression of genes involved in L-histidine degradation, L-lysine biosynthesis, inosine-5 -phosphate biosynthesis, ADP-L-glycero-beta-D-manno-heptose biosynthesis, Calvin-Benson-Bassham cycle, TCA cycle, superpathway of pyridoxal 5'-phosphate biosynthesis and salvage, pyridoxal 5'-phosphate biosynthesis, tetrapyrrole biosynthesis, flavin biosynthesis, and phosphopantothenate biosynthesis but reduced expression of genes involved in stachyose degradation and purine ribonucleosides degradation in CTRL-fed mice ( Figure 5C). Early life PCB 126 exposure reduced the expression of modules for the superpathway of L-aspartate and L-asparagine biosynthesis, L-histidine biosynthesis, L-arginine biosynthesis, phosphopantothenate biosynthesis, and the superpathway of coenzyme A biosynthesis in HFD-fed mice ( Figure 5C). beta-D-manno-heptose biosynthesis, Calvin-Benson-Bassham cycle, TCA cycle, superpathway of pyridoxal 5'-phosphate biosynthesis and salvage, pyridoxal 5'-phosphate biosynthesis, tetrapyrrole biosynthesis, flavin biosynthesis, and phosphopantothenate biosynthesis but reduced expression of genes involved in stachyose degradation and purine ribonucleosides degradation in CTRL-fed mice ( Figure 5C). Early life PCB 126 exposure reduced the expression of modules for the superpathway of L-aspartate and L-asparagine biosynthesis, L-histidine biosynthesis, L-arginine biosynthesis, phosphopantothenate biosynthesis, and the superpathway of coenzyme A biosynthesis in HFD-fed mice ( Figure  5C).
The influence of early life PCB 126 exposure on bacterial metabolites from mice fed with two diets was also investigated. We quantified the levels of common bacterial metabolites in the urine, feces, and cecal content samples using 1 H NMR and UPLC-MS/MS analysis ( Figure 5D-F and Supplementary Figures S5 and S6). Early life PCB 126 exposure resulted in significantly lower levels of urine microbial metabolites, such as hippurate and phenylacetylglycine (PAG), in the mice with two diets in adulthood ( Figure 5D). Early life PCB 126 resulted in a significantly higher level of cecal glycerol in HFD-fed mice later in life but not in CTRL-fed mice ( Figure 5E). Early life PCB 126 did not exhibit a pronounced effect on cecal short-chain fatty acids (SCFAs) in both CTRL and HFD-fed mice, which was not consistent with what we observed in mice fed with standard chow diet [15] (Supplementary Figure S5B). The variation could be due to the different composition of gut microbiota and reduced formation of SCFAs with HFD and purified diet (CTRL) compared to the chow diet, which is rich in fermentable fiber [34]. The UPLC-MS/MS analysis of bile acid composition showed significantly increased unconjugated and conjugated bile acids in feces from both CTRL and HFD-fed mice in adulthood with PCB 126 exposure ( Figure 5F and Supplementary Figure S6).  The influence of early life PCB 126 exposure on bacterial metabolites from mice fed with two diets was also investigated. We quantified the levels of common bacterial metabolites in the urine, feces, and cecal content samples using 1 H NMR and UPLC-MS/MS analysis ( Figures 5D-F, S5 and S6). Early life PCB 126 exposure resulted in significantly lower levels of urine microbial metabolites, such as hippurate and phenylacetylglycine (PAG), in the mice with two diets in adulthood ( Figure 5D). Early life PCB 126 resulted in a significantly higher level of cecal glycerol in HFD-fed mice later in life but not in CTRL-fed mice ( Figure 5E). Early life PCB 126 did not exhibit a pronounced effect on cecal short-chain fatty acids (SCFAs) in both CTRL and HFD-fed mice, which was not consistent with what we observed in mice fed with standard chow diet [15] (Supplementary Figure S5B). The variation could be due to the different composition of gut microbiota and reduced formation of SCFAs with HFD and purified diet (CTRL) compared to the chow diet, which is rich in fermentable fiber [34]. The UPLC-MS/MS analysis of bile acid composition showed significantly increased unconjugated and conjugated bile acids in feces from both CTRL and HFD-fed mice in adulthood with PCB 126 exposure (Figures 5F and S6).

Discussion
Early life exposure to environmental pollutants is an increasing global problem, adversely and continuously affecting human health [14]. Over the last 50 years, the diet pattern of humans has shifted due to a rapid transition to saturated fat and refined sugars [35]. Accumulating evidence suggests that the connections between diet and environmental exposures are complex and important [19][20][21]; however, the interactions between diet and early life environmental exposure at doses not associated with overt toxicity have not been fully investigated. In this study, we systematically investigated the interaction between early life PCB 126 exposure and mice eating a CTRL or HFD. We determined diet-independent and diet-dependent metabolic and bacterial changes caused by early life exposure to PCB 126. We also demonstrated that early life PCB 126 exposure worsens HFD-induced impaired glucose homeostasis without worsening the effects of HFD related to adiposity in adulthood.
The diet-independent metabolic changes induced by early life PCB 126 exposure were observed in mice later in life. PCB 126, the most potent dioxin-like PCB, shows persistent AHR activation at four months after early life exposure in mice with CTRL and HFD due to its long half-life [32,36]. The significantly higher levels of liver nucleotides were observed with PCB 126 exposure in both CTRL and HFD-fed mice, which has also been reported with air pollution and PCB 126 exposure mice fed with standard chow diet [15,37]. To ensure the accuracy of DNA and RNA synthesis for replication, transcription, and translation, nucleotide metabolism is tightly regulated at all levels to maintain constant pools of pyrimidines and purines [38]. The disruption in nucleotide synthesis by PCB 126 exposure might support the unrestrained proliferation, regardless of diet status, which is usually observed in cancer cells and virus-infected cells [38].
Early life PCB 126 exposure worsens HFD-induced impaired glucose homeostasis without worsening the effects of HFD related to adiposity in adulthood. Emerging evidence suggests that POPs are involved in key mechanisms linking obesity and diabetes [39]. Our results show that early life PCB 126 worsens HFD-induced impaired glucose homeostasis in adulthood characterized by glucose intolerance and elevated in gluconeogenesis and the TCA cycle. A similar observation has been reported in a long-term PCB 126 exposure to adult mice that showed systemic effects associated with early end point of diabetes, including insulin resistance and impaired glucose metabolism induced by PCB 126 [10]. Future studies, including hepatic and systemic insulin sensitivity, are indeed warranted to fully capture the impact of changes in glucose homeostasis with early life PCB 126 exposure and HFD. It is interesting to note that PCB 126 and PCB 153 [19] both worsen diet-induced obesity/metabolic syndrome, suggesting the PCB effects on obesity/diabetes may depend more on diet interactions than on the congeners themselves. More work is needed to determine the role of ligand-activated transcription factors, such as the AHR, in relationship with environmental pollutants and the metabolic syndrome. Notably, the reduction in serum LIF and increase in serum LIX were observed in mice with HFD induced by early life PCB 126 exposure, suggesting an inflammatory response caused by early life PCB 126 that can contribute to the dysfunctional regulation of caloric intake and energy expenditure commonly present in obesity [40,41]. This notion is supported by the observation of up-regulated gene expression of several inflammatory mediators in adipose from HFD-fed mice with early life PCB 126 exposure. These are further accompanied by the significantly higher ratio of liver GSSG to GSH in HFD-fed mice with early life PCB 126 exposure, suggesting a marker of oxidative stress that is considered to be involved in various diseases, including obesity and diabetes [42]. Further evidence, such as H&E and immunohistochemical staining, is warranted to determine inflammation severity in the adipose tissue. Surprisingly, we did not observe increases in hepatic lipogenesis in mice fed a CTRL or HFD induced by early life PCB 126 exposure, which was reported in the experiments performed on human HepaRG liver cells, C57BL/6J mice fed with standard chow diet, and male Sprague Dawley rats fed with AIN-93G diet with PCB 126 exposure [15,43,44]. Emerging evidence demonstrated the direct role of AHR in liver lipid oxidation and the lipogenic pathway [45,46]. Our observation could be partially explained by the reduced availability of AHR ligands in the purified diets as compared to the standard chow diet [47] and the complexity of interactions between dietary nutrients and early life environmental pollutants. More studies are needed to rigorously test these variables using more variable animal models for the metabolic syndrome.
Early life PCB 126 exposure resulted in diet-dependent changes in bacterial community structure and function in adulthood. It is increasingly clear that HFD alters the gut microbiota community structure, such as a decrease in Bacteroidetes and an increase in Firmicutes, which have been associated with obesity and chronic diseases [48,49]. We believe that diet not only plays a significant role in shaping the microbiome but also modifies the interactions between gut microbiota and the pollutants [50]. We observed a higher ratio of Firmicutes to Bacteroidetes in HFD-fed mice later in life with early life PCB 126 exposure, considering it a relevant marker of adverse changes in the gut microbiome community in obese subjects [49,51]. The decreased levels of genera Blautia and Butyrivibrio, which are butyrate-producing bacteria with anti-inflammatory properties [35,52], were observed in early life PCB 126 exposure mice with CTRL diet but not with HFD. The difference could be due to the significantly lower abundance of these bacteria with HFD [52]. Notably, early life PCB 126 exposure not only altered the bacteria community but also significantly changed microbial metabolism in adulthood. PCB 126 exposure resulted in reduced levels of urine hippurate in mice fed with CTRL and HFD, indicating decreased gut microbiome diversity, which is associated with multiple disorders, including the metabolic syndrome [53]. Microbial biosynthesis of amino acids in the human gastrointestinal tract plays a potential role in the host's amino acid homeostasis [54]. The significant changes in microbial amino acid metabolism seen in both CTRL and HFD-fed mice with early life PCB 126 exposure suggest a disruption between microbial activity and host amino acid and energy homeostasis, which may be associated with the development of obesity [55]. Interestingly, a reduction in coenzyme A biosynthesis was also observed in HFD-fed mice with early life PCB 126 exposure, possibly indicating a disruption in the synthesis and degradation of fatty acids and phospholipids [56]. Consistently, we observed significantly higher levels of cecal glycerol with early life PCB 126 exposure in HFD-fed mice, suggesting the disturbance of fat absorption in the small intestine [57]. Similar observation has been reported with higher glycerol in fecal samples of patients suffering from Crohn's disease, suggesting gut inflammation caused by the destruction of microbiota that is commonly associated with obesity [58]. More research is needed focusing on microbial fatty acid and lipid metabolism pathways to better understand the interaction with pollutant exposure and HFD. Furthermore, early life PCB 126 exposure resulted in elevated bile acid biosynthesis in mice later in life in both CTRL and HFD-fed mice, which could be due to the remodeling of gut microbiota and persistent AHR activation by PCB 126. Similar observations were seen in our previous studies showing significantly higher levels of bile acids in early life at five days of PCB 126 exposure in mice fed with standard chow diet [15].

Conclusions
Collectively, these data demonstrate that early life PCB 126 exposure persistently and significantly affects AHR signaling and liver nucleotide metabolism in mice fed with a CTRL or HFD in adulthood. Notably, we determined that early life PCB 126 worsened HFD-induced impaired glucose homeostasis characterized by glucose intolerance and elevated gluconeogenesis and TCA cycle flux, as well as changes in serum cytokines and adipose inflammatory genes, without worsening the effects of HFD related to adiposity in adulthood. Early life PCB 126 exposure resulted in diet-dependent changes in bacterial community structure and function later in life. These results provide a better understanding of the interactions between diet and early life environmental exposures and demonstrate that diet is a critical factor in the biochemical consequences of early life POP exposure involving the development of metabolic diseases.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/metabo12100894/s1, Figure S1: Experimental schedule of the present study; Figure S2: (A) Body weight of mice recorded every week from CTRL or HFD-fed mice with vehicle or PCB 126 exposure. (B) Light microscopic examination of H&E-stained liver sections from CTRL or HFD-fed mice with vehicle or PCB 126 exposure; Figure S3: qPCR analysis of mRNA levels of AHR targeted gene in the ileum from CTRL or HFD-fed mice with vehicle or PCB 126 exposure; Figure S4: qPCR analysis of mRNA levels of inflammatory cytokines in the adipose from CTRL or HFD-fed mice with vehicle or PCB 126 exposure; Figure S5: (A) Relative abundance of cecal bacteria from phylum in cecal content from CTRL or HFD-fed mice with vehicle or PCB 126 exposure. (B) NMR analysis of cecal short-chain fatty acids (SCFAs) from CTRL or HFD-fed mice with vehicle or PCB 126 exposure; Figure S6: Quantitative UPLC-MS analysis of bile acids in the feces from CTRL or HFD-fed mice with vehicle or PCB 126 exposure; Table S1: Composition of control diet (CTRL) F4031 and high fat diet (HFD) F3282.; Table S2: mRNA gene-targeted primers used in this study; Table S3: Serum cytokine (32-Plex) assay results; Table S4: Relative abundances of significantly changed genus in mice fed with CTRL; Table S5: Relative abundances of significantly changed genus in mice fed with HFD.

Conflicts of Interest:
The authors declare no conflict of interest.