(175bf) Assessing the Metabolic Responses of 3D Heparg Cells to Dibutyl Phthalate and DEHP Exposure: Insights from Untargeted Metabolomics

Phthalates, widely found in various consumer products including PVC plastics, medical devices, and food packaging, play a significant role in modern manufacturing. Among these, di-n-butyl phthalate (DBP) and di(2-ethylhexyl) phthalate (DEHP) are prominently utilized. There is growing evidence to suggest that exposure to these phthalates may be linked to metabolic disorderss such as obesity, insulin resistance, and nonalcoholic fatty liver disease (NAFLD). However, the precise impact of these chemicals on the human metabolome remains unclear. This study aims to dissect the metabolomic alterations in 3D hepatocyte cultures following exposure to DBP and DEHP, shedding light on how such exposures may influence metabolic disease pathways.

HepaRG spheroids, also known as Hepoid-HepaRG, were cultured within a collagen matrix, following the method outlined by Rose et al. (2022). The procedure began by embedding Hepoid-HepaRG cells in type I collagen extracted from bovine skin (provided by Capricorn Scientific, Germany) and mixed with William's E medium to attain a collagen solution concentration of 1.5 mg/ml. The medium was enhanced with 5% Fetal Bovine Serum (FBS), antibiotics (100 U/ml penicillin and streptomycin), 2 mM Glutamine, 105 M hydrocortisone hemisuccinate, and 5 μg/ml insulin, with its pH adjusted to 7.4 using 0.1 N NaOH. Cells were introduced at a density of 6 × 105 cells/ml into this mixture, which was subsequently aliquoted into 96-well plates for WST-1 assays and 384-well plates for ATP assays. Following an one hour incubation at 37 °C in a 5% CO2 environment, an equal volume of medium was supplemented to the gels that had solidified. Five days post embedding in collagen, the HepaRG spheroids were exposed to three DEHP and DBP concentrations ranging from 100 pM to 10 uM, with a final DMSO concentration of 0.1% in the media. Media was refreshed every 48 or 72 hours throughout the duration of the study.

Briefly, 250 μl aliquot of each polar sample was transferred to a new Eppendorf tube after vortexing at 2000 rpm for 5 sec. Aliquots were dried under a gentle nitrogen flow using a Techne Sample concentrator (Time: 180 min, Temperature: 45 °C). Then, the samples were resuspended using 120 μl of solvent water/methanol (70:30), followed by vigorous vortex for 5 sec at 2000 rpm. Samples were centrifuged at 15000 rpm for 10 min, and the supernatants were transferred to autosampler vials with inserts. A mixture of multiple internal standards was added to the solvent to assess system stability for each sample analysed. The sample preparation procedure was the same for the nonpolar samples except for the solvent used for metabolite extraction; 75 µl of a methanol/chloroform (3:1) solution containing a mixture of internal standards.

Global untargeted metabolomics analysis on the Hepoid-HepaRG cells exposed to the two plastisizers was performed using an Agilent 1290 infinity UHPLC System coupled to an Agilent 6540 HRMS-QTOF in both positive and negative ionization modes. Additionally, polar and nonpolar samples were analysed with a RP and HILIC column, respectively, in order to increase the coverage of the detected metabolites. The data were acquired between 50 and 1700 m/z at a scan rate of 1.5 spectra/sec in centroid mode at a resolution of 40,000 FWHM. The source conditions were as follows: gas temperature 300oC, drying gas 7 L/min, nebulizer 50 psig, fragmentor 250 V, skimmer 65 V, and capillary voltage 3500V or -3500V in positive and negative modes, respectively. Agilent MassHunter Software v.B.06.01 was used to collect the data, followed by data pre-processing based on R packages.

Initially, raw data from the analyses were converted to the standardized .mzML format using the msConvertGUI tool from the ProteoWizard suite. The centWave algorithm was then applied for peak detection, effectively identifying chromatographic peaks from the complex data set. Subsequent alignment of samples was performed using the obiwarp method with a binSize of 0.6 to ensure consistency across samples. The correspondence matching phase grouped chromatographic peaks across samples into coherent features, using a density-based approach along the retention time axis with optimized parameters (binSize of 0.25, bandwidth (bw) of 20, and a minimum fraction (minFraction) of 0) facilitated by the IPO R package. In the data cleaning phase, features underwent rigorous filtering to ensure reliability: variables with over 80% missing values in QC samples or 0% missing in solvents (to avoid contamination) were discarded. Additionally, features exhibiting a Relative Standard Deviation (RSD) greater than 30% in QC samples were excluded to maintain data quality. To further refine the dataset, peak intensities were log-transformed, scaled, and normalized, minimizing systematic and experimental variations. This meticulous data preprocessing and quality control approach provides a solid foundation for reliable and insightful metabolomic analysis.

Significantly differential features (DEFs) were identified using the ANOVA unequal variance test in R, with Benjamini-Hochberg FDR correction for multiple comparisons to reduce false positives, setting a p-value cut-off at 0.05. Fold change, calculated as the ratio between control and DEHP or DBP exposed cells, was considered significant at a cut-off of 2.0. These DEFs then underwent metabolite annotation using the annotation R package xMSAnnotator developed by Uppal et al. (2017) to perform accurate mass mapping with online compound databases (HMDB, LipidMaps, and KEGG). Queries of accurate mass values in compound databases provided several matches with a mass error below 25. The confirmation of the identified selected biomarkers was performed by comparison with the RT, and fragmentation pattern of authentic analytical standards from the EnvE-X in-house library, or MS/MS spectra that are available in databases like HMDB, Metlin, and LipidMaps. Pathway analysis was conducted using Fisher's Exact method, with a focus on pathways from KEGG database enriched with specific metabolites (p-value<0.05) and excluding predicted metabolites to minimize false positives.

Metabolomic analysis of Hepoid-HepaRG cells exposed to DEHP at concentrations of 100 pM, 100 nM, and 10 μM identified 65, 72, and 67 statistically significant annotated metabolites, respectively. These metabolites were associated with six significant metabolic pathways, with the pathways of phenylalanine, tyrosine, and tryptophan biosynthesis (p-value = 0.0017), tryptophan metabolism (p-value = 0.004), sphingolipid metabolism (p-value = 0.005), aminoacyl-tRNA biosynthesis (p-value = 0.008), phenylalanine metabolism (p-value = 0.01), and biosynthesis of unsaturated fatty acids (p-value = 0.02) being the most significant.

Similarly, exposure to DBP at concentrations of 100 pM, 100 nM, and 10 μM resulted in the identification of 66, 70, and 74 statistically significant annotated metabolites, respectively. The metabolic shifts were particularly notable in glycerolipid, glycerophospholipid, nucleoside, and organoheterocyclic compound metabolisms. The analysis revealed five statistically significant pathways influenced by DBP exposure, including phenylalanine, tyrosine, and tryptophan biosynthesis (p-value = 0.001), tryptophan metabolism (p-value = 0.003), sphingolipid metabolism (p-value = 0.036), aminoacyl-tRNA biosynthesis (p-value = 0.005), and phenylalanine metabolism (p-value = 0.009).

The untargeted metabolomic analyses revealed 88 unique metabolites that were statistically significant following exposure to both DEHP and DBP. Among these, 56 metabolites were commonly altered across all DBP treatments, and 51 were common across all DEHP treatments. These metabolites spanned several classes, including fatty acyls, ceramides, sterol lipids, glycerolipids, and carboxylic acids. Notably, several detected ceramides and Glycerolipids, such as Cer(d18:0/20:0), DG(15:0/0:0/24:1n9), and TG(20:1(11Z)/20:1(11Z)/18:2(9Z,12Z)) are implicated in insulin resistance and metabolic syndrome (Chavez et al., 2014, Finck and Hall, 2015, Grundy, 1999), highlighting potential mechanisms by which phthalate exposure contributes to these conditions. Furthermore, amino acids such as L-tryptophan, L-phenylalanine, and L-serine were significantly perturbed by both DBP and DEHP exposures, linking these environmental chemicals to obesity risk (Wahl et al., 2013). Pathway analysis underscored the disruption of metabolic pathways related to lipid and amino acid metabolism. Common pathways affected by both phthalates included phenylalanine, tyrosine, and tryptophan biosynthesis; tryptophan metabolism; sphingolipid metabolism; aminoacyl-tRNA biosynthesis; and phenylalanine metabolism.

This study's untargeted metabolomics analysis offers a thorough characterisation of the broad metabolic disruptions caused by DBP and DEHP exposure in HepaRG cells. The identification of significant metabolites and pathways, particularly those associated with lipid and amino acid metabolism, highlights the extensive impact these phthalates can have on cellular processes. This research not only sheds light on the toxicological effects of DBP and DEHP but, also, underscores the utility of metabolomics in uncovering potential biomarkers for exposure and toxicity, paving the way for future studies aimed at mitigating the health risks associated with phthalate exposure.