Term placentas were obtained from a cohort of 40 anonymous healthy pregnant women (average age of 25.9 ± 5.2 years old), carefully matched for gestational delivery week (37.8 ± 2.6 weeks), who provided consent following uncomplicated pregnancies and underwent caesarean section deliveries. Noteworthy, these placentas were not analyzed for MP they eventually could already contain. Ethical approval for research involving human subjects was obtained from the Ethics Committee (CAAE 58129422.3.0000.5013). Sampling took place at Hospital Universitario Professor Alberto Antunes (HUPAA) of the Federal University of Alagoas (UFAL) between February 2023 and September 2023. Exclusion criteria encompassed gestational diabetes, gestational hypertensive diseases, chorioamnionitis, membrane rupture, and any other maternal medical conditions, as well as known fetal aneuploidy, fetal growth restriction, and macrosomia.

The placental explants were obtained following a protocol based on the methods initially developed by Trowell (42), with current modifications implemented and extensively published (43). Cotyledon fragments were thoroughly rinsed in glass-filtered phosphate buffer solution (PBS, 1.6 μm Whatman GF/A, Sigma-Aldrich, St. Louis, MO, USA) to remove maternal blood and to avoid baseline plastic exposure or any contamination by MP. Only the villous tissue was minced with a number 21 scalpel, generating approximately 5 mg of tissue per well. These placental explants were cultured in individual glass Petri dishes in DMEM/F12 culture medium (Sigma-Aldrich) with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin and 1% gentamicin (Sigma-Aldrich), and incubated at 37° C and 5% CO2. To prevent contamination from laboratory-derived plastics, all laboratory materials, typically constituted by plastic, were substituted for glass materials. Whenever a plastic substitute could not be employed, this limitation was addressed in the text.

Polystyrene microplastics (PS-MPs) of 5 μm functionalized with carboxylic surface groups were used (Magsphere, Los Angeles, CA, USA). Thus, the structure and composition of polystyrene were corroborated using Raman spectroscopy (XploRA Raman spectrometry; spectral range 300–2200 cm−1, 532 and 785 nm laser diodes, 600 lines per mm grating, Horiba, Japan) and analyzed with KnowItAll software (Wiley Science Solutions, Hoboken, NJ, USA), and the free microplastic libraries SLOPP and SLOPPe ( Supplementary Figure 1 ). No pre- or post-treatments were performed in the PS-MPs. Exposure time and concentration were defined in subsequent experiments, also based on the results reported in literature (44). Bisphenol A (BPA) (Sigma-Aldritch) was employed as positive control for oxidative stress experiments at the concentration of 1 μM according to the literature (45).

Placental explants were assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-ditetrazolium bromide (MTT) assay, following the manufacturer´s protocol (INLAB, SP, Brazil). Explants were treated with PS-MP concentrations of 0.1, 1, 10, and 100 µg/mL (w/v) for 24 h to establish an experimental non-lethal concentration of PS-MP. Briefly, placental explants were transferred to a 96-well plastic plate, and incubated for 4 h at 37°C with 5 mg/mL of MTT in DMEM/F12 (Sigma-Aldrich). Following PS incubation, the absorbance was measured at 540 nm using a Bio-Rad Multiwell Plate reader (Bio-Rad, Hercules, CA, USA). All experiments were performed in triplicates (n=3).

PS-MP cytotoxicity was measured using the Pierce LDH Cytotoxicity Assay Kit, according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). Briefly, placental explants were treated or not with PS-MP for 24, 48 and 72 h for a time-dependent cytotoxicity assessment. Following PS-MP exposure, 50 μL of conditioned medium from each sample were transferred to a 96-well plastic plate, and 50 μL of Reaction Mixture were added. After 30 minutes, protected from light, 50 μL of Stop Solution were added to each well, and the absorbance was measured at 490 and 680 nm on a Bio-Rad Multiwell Plate reader (Bio-Rad). To calculate the percentage of cytotoxicity, the LDH activity of the spontaneous LDH Release Control (water-treated) was subtracted from the PS-treated sample LDH activity and divided by the total LDH activity [(Maximum LDH Release Control activity) – (Spontaneous LDH Release Control activity)], and multiplied by 100, according to the manufacturer. The experiments were made in triplicates, and the cytotoxicity for each n was calculated (n=3). Noteworthy, we have accompanied the control group dynamics of LDH release through time, to make sure that the placental explants were not naturally dying over time, and up to 72 h of culture they were considered viable (Average O.D. of 1.22 at 24 h, 1.3 at 48 h, and 1.529 at 72 h, while the maximun LDH release positive control had an average O.D. of 2.23 at 24 h, 2.3 at 48 h, and 2.46 at 72 h).

The O2•– production was evaluated by two methods. The first method focused on the overall production of O2•– by nitroblue tetrazolium (NBT) (Sigma-Aldritch) (46). After 1 h of incubation, a solution of 1 M KOH and DMSO (v/v) was added, and the placental explants were transferred to a 96-well plastic plate. The absorbance was measured at 630 nm using a spectrophotometer (AJX-6100PC, Micronal). The second method was performed to access only the mitochondrial O2•– production, through mitoSOX staining (47) at 5 µM at an excitation wavelength (λex) of 510 nm, emission wavelength (λem) of 580 nm, and a 10 nm slit width for both excitation and emission. (Invitrogen, Waltham, MA, USA), and according to the manufacturer’s instructions. The fluorescence was measured with two probes (Invitrogen) and a spectrofluorimeter (Shimadzu RF5301, Shimadzu, Columbia, MD, USA). All experiments were conducted in triplicates (n=3).

The production of H₂O₂ was monitored using the fluorescent probe Amplex Red (10 μM, Thermo Fisher Scientific) (48). This experiment was conducted with peroxidase (1 U/mL) at an excitation wavelength (λex) of 563 nm, emission wavelength (λem) of 587 nm, and a 3 nm slit width for both excitation and emission. The H₂O₂ was expressed in µmol/mg of protein-1 and rendering to a calibration curve. All experiments were conducted in triplicates (n=8).

The placental explants homogenate (80 μg protein) were incubated in 50 mM sodium carbonate buffer (NaCO₃/NaHCO₃ at pH = 10.2 with 0.1 mM EDTA) at 37°C. The reaction was initiated by adding 20 μL epinephrine (150 mM) in acetic acid (0.05% v/v). The absorbance was measured at 480 nm with a dual-beam scanning spectrophotometer (AJX-6100PC, Micronal) for 5 min. One SOD unit was defined as the amount of protein to inhibit the autoxidation of epinephrine (1 μM) per minute. The results were expressed as U mg-1 protein (49). All experiments were conducted in triplicates (n=5).

To determine the decomposition of H₂O₂ in water and oxygen, 40 μg equivalent of protein (from the placental explants homogenate) was added to 50 mM phosphate buffer (KH₂PO₄ + Na₂HPO₄ with pH = 7.0) at 25°C. The reaction was started by adding 10 mM of H2O2. The analytical signal decrease was monitored at 240 nm with a dual-beam scanning spectrophotometer (AJX-6100PC, Micronal) for 2 min. One unit of CAT was defined as the amount of protein required to convert 1 μM of H2O2 to H₂O and O₂ per minute (50). The results were expressed as K mg-1 protein and all experiments were conducted in triplicates (n=5).

The GPx activity was monitored by determining the decrease in nicotinamide adenine dinucleotide phosphate (NADPH) analytical signal at 340 nm using a scanning spectrophotometer (AJX-6100PC, Micronal) for 3 min at 20°C (51). The mixture contained 80 μg protein (from the placental explants homogenate), 0.05 M Na2HPO4 (pH = 7.0), 5 mM EDTA, 84 μM NADPH, 1.1 mM sodium azide, 1.5 mM of GSH, 0.1 U of GR, and 90 μM of H2O2. One enzyme unit was defined as the amount required for the oxidation of 1 μM of NADPH min-1 mg-1 of protein. The results were expressed as U mg-1 protein and all experiments were conducted in triplicates (n=5).

The GSH/GSSG ratio was assessed by measuring the levels of GSH and GSSG (52). The homogenates containing 50 μg and 100 μg of protein and o-phthalaldehyde (1 mg mL-1) were utilized. For the GSH assay, 0.1 M NaH2PO4 (pH 8.0) and 0.005 M EDTA were used as buffer, whereas 0.1 M NaOH was the buffer used for the GSSG reaction. Fluorescence was monitored using a spectrofluorometer (Flex Station 3, Molecular Devices LLC, San Jose, CA, USA) with an excitation wavelength (λex) of 350 nm and an emission wavelength (λem) of 420 nm. An analytical curve based on known concentrations of GSH and GSSG was used to determine the concentrations of these compounds in the samples. All experiments were conducted in triplicates (n=5).

Total protein quantification of the placental explants was conducted using the Bradford method, and a BSA solution (3 mg mL-1) was used as a standard for total protein determination. The total SH content in the samples was determined based on the reaction with 5,5’ Dithiobis (2 Nitrobenzoic Acid) (DTNB, Sigma-Aldritch). The homogenate (100 μg protein) was incubated in the dark with DTNB (500 μM) for 30 min. Afterwards, the final volume was adjusted with extraction buffer (50 mM Tris-HCl with 1.0 mM EDTA, pH 7.4), and the absorbance was measured at 412 nm with a dual-beam scanning spectrophotometer (AJX-6100PC, Micronal, Sao Paulo, Brazil) (53). All experiments were conducted in triplicates (n=5).

A total of 450 μg of tissue homogenate protein was combined with 200 μL of 30% (v/v) trichloroacetic acid (TCA). Subsequently, Tris-HCl (10 mM) was added and centrifuged at 1180 g for 10 min at 4°C. Then, the collected supernatant was mixed with 0.73% (m/v) thiobarbituric acid, which reacted with lipid peroxidation products to form a pink-colored compound. Following this, the mixture was incubated for 15 min at 100°C and cooled down. The absorbance was measured using a spectrophotometer (AJX-6100PC, Micronal) at 535 nm. The results were expressed in mmol mg-1 of protein (54). All experiments were conducted in triplicates (n=5).

The quantification of carbonyl group formation in 100 µL of tissue homogenate was conducted using the reaction with 2,4-dinitrophenylhydrazine (DNPH) (55). The reaction control was treated with 2.5 M HCl, whereas the samples groups were treated with 10 mM DNPH (diluted in 2.5 M HCl). Subsequent to room temperature incubation, 2 M trichloroacetic acid (TCA) was added, followed by centrifugation at 10,000 g for 10 min at 4°C. After sequential steps of centrifugation and washing, an ethanol and ethyl acetate mixture (1:1 v/v) was added. The groups were finally suspended in 6 M guanidine and centrifuged again. The collected supernatant was used for absorbance measurement at 370 nm using a plate reader, normalized to the protein content (Tecan Infinite 200 pro, Männerdorf, Switzerland). The results were expressed in pmol mg-1 protein. All experiments were conducted in triplicates (n=5).

The supernatants of 72 h explant cultures were prepared by adding 350 μl of 0.1 M phosphate buffer containing 3-[Trimethylsilyl] propionic-2,2,3,3-d4 acid sodium salt at 0.1 mM in D2O. The prepared samples were then transferred to 5 mm NMR tubes. All 1H NMR spectra were acquired on a Bruker 600 MHz spectrometer Avance III (Bruker BioSpin, Billerica, MA, USA) equipped with a PA BBO 600 S3 BBF-H-D-05 Z SP probe. Before the experiments, the NMR spectrometer was calibrated daily following strict standard operating procedures to ensure the highest spectral quality and reproducibility. A sealed tube containing a 99.8% MeOD standard sample was used to ensure that supernatants were run at 298 K. The ‘edte’ command controlled the temperature’s stability over time. The second calibration with sucrose aimed to optimize water suppression, and spectrometer sensitivity. An automatic tune and matching procedure, deuterium signal lock, and automatic shimming routine were performed before each experiment. The NMR spectra were recorded using the 1D NOESY pulse sequence with water suppression (pulse program: noesygppr1d). For all experiments, 128 scans were recorded with 32 k data points, a spectral width of 20.029 ppm, and a relaxation delay of 4 s. The spectral processing was realized automatically using TopSpin 3.5 (Bruker BioSpin, Billerica, MS, United States), performing phase and baseline correction, and calibrating the TSP reference signal at δ0.00 ppm. The peaks were identified using the Human Metabolome Database (HMDB) (The Metabolomics Innovation Centre, Canada) and the ChenomX NMR Suite (Chenomx Inc., Canada). Criteria such as chemical shifts and multiplicity were employed for peak identification. The dataset underwent further processing using the R statistical software, version 4.1, along with the R package PepsNMR (R Core Team, Vienna, Austria). The spectral region spanning from 4.7 to 5.1 ppm, which includes the residual water signal, was selectively suppressed. Biomarker discovery and classification were performed using the online platform MetaboAnalyst version 6 (www.metaboanalyst.ca).

The supervised model Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) was executed to evaluate the capacity of metabolomics data in distinguishing PS-MP exposed samples from control samples, considering the complete set of identified metabolites. The defined confidence region was set to 95%, and one hundred permutations were employed for model evaluation. Significant metabolites (biomarkers) were selected by analyzing loadings and variable importance in the projection (VIP) plot from the OPLS-DA model, indicating the variables crucial for group discrimination. Additionally, a heatmap analysis was applied to reveal metabolic patterns between the groups and a t-test with false discovery rate (FDR)-corrected p values was employed to assess the significance of metabolites. The metabolomics biomarkers data were correlated with oxidative stress parameters. Finally, to estimate the association between the two datasets, it was applied the Pearson correlation coefficient and a multiblock Partial Least Squares Discriminant Analysis (PLS-DA) model, also known as Data Integration Analysis for Biomarker Discovery using Latent Components (DIABLO) (56). All experiments were conducted in triplicates (n=16).

The 100 μg/mL PS-MP in DMEM/F12 culture media had their zeta potential measured by a nano/zeta sizer ZS90 (Malvern Instruments Ltd) with 250 scans made in five different samples.

With the exception of metabolomics already described, the statistical analyses was performed using the one-way ANOVA followed by Bonferroni multiple comparisons. All the statistical analyses were performed using GraphPad (version 7.0; GraphPadSoftware, Inc.), with p < 0.05 considered statistically significant.

Exposure to different concentrations of PS-MP for 24 h did not affect explant viability as determined by the screening MTT test ( Figure 1A ). Based on these results of viability, the concentration of PS-MP to be tested in all further experiments were 100 μg/mL, as it was not only inducing viability reduction, but the increased tendency observed caught our attention.

To assess if such concentration would present time-dependent cytotoxicity, the LDH release assay was performed after exposure of the placental explants for 24, 48 and 72 h. The cytotoxicity after 24 h was 10.37 ± 3.79%, after 48 h was 23.19 ± 3.09%, and after 72 h was 40.32 ± 7.02%. The 30% increase in cytotoxicity from 24 to 72 h was statistically significant (p < 0.01) ( Figure 1B ).

As MP-PS exposure showed time-dependent cytotoxicity to placental explants, we evaluated if they would trigger an oxidative stress response. We assessed the production of mitochondrial O₂•– and overall O₂•– levels in placental explants using MitoSOX staining and the NBT assay, respectively. Both methods demonstrated a significant increase in O₂•– levels, with mitochondrial O₂•– increasing by 3.5-fold and total O₂•– showing a similar elevation upon exposure to PS-MP (p < 0.01 and p < 0.05, respectively) ( Figures 2A, B ). The similar pattern was observed in the analysis of H₂O₂ generation, which greatly increased from 6.29 ± 0.43 A.U. to 30.10 ± 0.43 A.U. (p < 0.001) ( Figure 2C ). The activity of antioxidant enzymes was also assessed. SOD catalyzes O₂•– in O₂ and H₂O₂, helping to regulate mitochondrial function as it scavenges O₂•–, produces H₂O₂ and inhibits ONOO- production by protecting NO, while CAT and GPx are responsible of breaking H₂O₂ in O₂ and H₂O. We observed a great reduction of SOD and CAT activity when explants were exposed to PS, as CAT was reduced from 0.056 ± 0.007 K mg-1 protein in the control group to 0.014 ± 0.008 K mg-1 protein in the PS group (p < 0.05) ( Figure 2D ). Similarly, SOD was reduced from 428.5 ± 34.58 U mg-1 protein in the control group to 150.4 ± 49.84 U mg-1 protein in the PS group (p < 0.01) ( Figure 2E ). In contrast, GPx activity was increased from 0.036 ± 0.002 U mg-1 protein in the control group to 0.061 ± 0.008 U mg-1 protein (p < 0.05) ( Figure 2F ).

Since GPx activity was increased, we evaluated GSH and GSSG levels. While GSH had a tendency of reduction in the PS group ( Figure 2G ), and GSSG was unchanged ( Figure 2H ), their ratio was reduced in the PS group, from 3.84 ± 0.43 to 2.91 ± 0.38 (p < 0.05) in the control group ( Figure 2I ). Altogether, these results indicate the presence of oxidative stress due to reduced SOD and CAT capabilities, and a reduced GSH/GSSG ratio coupled with increased GPx activity indicates this antioxidant defense system is actively engaged in counteracting the oxidative stress.

Another well-known marker of oxidative stress is the total SH content, which prevent uncontrolled oxidation of proteins by Cystein aminoacids. We observed a clear reduction of the total SH content, from 93.93 ± 10.84 mmol mg-1 protein in the control group to 56.05 ± 2.96 mmol mg-1 protein in the PS exposed group (p < 0.05) ( Figure 2J ).

Due to the observation that exposure to PS-MP results in an increase in all reactive oxygen species (ROS), it is crucial to comprehend the extensive oxidative stress experienced by placental explants. Therefore, we conducted further analyses on ROS production indicators, including measurements of malondialdehyde (MDA) and carbonylated protein levels. MDA is one of the final products of polyunsaturated fatty acids peroxidation, while carbonylation of proteins is defined as the covalent, nonreversible modification of the side chains of cysteine, histidine and lysine residues by lipid peroxidation end products, resulting in their loss of function. Our results showed that MDA level substantially increased, from 0.66 ± 0.52 μM mg-1 protein in the control group to 4.26 ± 0.60 μM mg-1 protein in the PS group (p < 0.001) ( Figure 2K ). This elevation was concomitant with a rise in protein carbonylation, escalating from 9.18 ± 1.20 nmol mg-1 protein in the control group to 29.16 ± 4.06 nmol mg-1 protein in the PS group (p < 0.01) ( Figure 2L ).

All supernatant samples underwent analysis using 1H NMR. Figures 3A, B depict the metabolite identification and overlapped spectra of PS-exposed and control groups. We employed orthogonal projections to latent structures discriminant analysis (OPLS-DA) for its ability to distinguish predictive from non-predictive (orthogonal) variations. As such, OPLS-DA revealed substantial metabolite differences, separating both clusters ( Figure 3C ). The model exhibited a high degree of separation between groups and good predictive performance through cross-validation (Q2 0.747, R2Y 0.855, and p < 0.01) ( Figure 3D ). The variable importance in the projection (VIP) score highlighted multiple significant alterations, with formate, valine, tyrosine, isoleucine, isobutyrate, hypoxanthine, and alanine being the most important variables ( Figure 3E ). Furthermore, the heatmap visualization of both groups illustrated the similarities among samples within the control or PS groups and the main differences between the analyzed groups ( Figure 3F ), corroborating findings from the correlation plots and OPLS-DA.

The relative concentrations, calculated by normalizing the molar concentration of each metabolite to the total molar concentration of all metabolites for each sample in the two groups, were compared using box-and-whisker plots. The data obtained demonstrated that the PS-MP exposed group exhibited lower levels of alanine (p = 0.03, Figure 4 ), formate (p < 0.001, Figure 4 ), glutaric acid (p = 0.02, Figure 4 ), and maltotriose (p = 0.03, Figure 4 ), whereas showing increase in valine (p = 0.03, Figure 4 ) and tyrosine (p < 0.001, Figure 4 ), confirming the results of previous analyses. These findings suggest potential impairments in folate, and amino acid metabolisms, and alterations in the TCA cycle ( Figure 5 ).

Additionally, the observed metabolic changes correlated with oxidative stress alterations found in placental explants ( Supplementary Figure 2 ). The heatmaps highlight the Pearson correlation among metabolites and between metabolites and oxidative stress ( Supplementary Figures 2A, B ). Pairwise variable associations were also extracted from a similarity matrix derived from the DIABLO model ( Supplementary Figures 2C, D ). This measure, analogous to a correlation coefficient, estimates the variables’ association relevant to group discrimination. Consequently, the reduced antioxidant system directly correlated with reduced glucose, maltotriose, carnitine, and glutaric acid metabolites, while CAT-reduced levels correlated negatively with lactate levels. The increased GPx, negatively correlated with almost all metabolites (isoleucine, isobutyrate, tyrosine, phenylalanine, lactate, histidine, alanine, carnitine, and formate). Additionally, the reduced GSH/GSSG ratio positively correlated with valine, ethanol, isoleucine, isobutyrate, tyrosine, phenylalanine, lactate, histidine, and alanine reduction. Also noteworthy is the negative correlation between the increase in carbonylated proteins with glutaric acid and hypoxanthine ( Supplementary Figure 2 ).