Perinatal exposure to nano-polystyrene: deleterious imprinting on inflammatory bowel diseases Authors: Melvin Airaud, Amandine Roldan, Jeanne Le Cléac’h, Lisa Isoard, Marie Carrière, Sandrine Ménard, Frédérick Barreau Published: 2026 ABSTRACT BackgroundThe extent of production and use of plastics leads to release of nanoplastics (NPL) into the environment, where they undergo weathering that could alter their toxicological properties. NPL enter the human food chain via contaminated food, and food packaging, and within the body they are known to pass through epithelial barriers. In parallel to the increasing exposure to plastics, the incidence of non-communicable diseases, including intestinal diseases, is rising worldwide.This study focuses on how oral exposure of gestating and lactating female mice to 50 nm polystyrene beads, pristine (PS50) or weathered (PS50w) affects intestinal function in offspring. Two aspects of intestinal function were studied: oral tolerance and the onset of intestinal disorders.MethodsGestating mice received 1.25 mg of PS50 or PS50w by oral gavage administered daily from gestational day 15. Exposure was continued during lactation until pups were weaned (Postnatal Day, PND21). To study oral tolerance, offspring were exposed to ovalbumin (OVA) both orally and systemically. To study intestinal disorders, colitis was induced by exposure to Dextran Sulfate Sodium (DSS) from PND63.ResultsPerinatal exposure to PS50 or PS50w had no impact on mortality of gestating or lactating dams or sex ratio in litters, but did lead to increased body weight at PND63 in offspring. No impact on tolerance to OVA was observed. However, macroscopic scores for DSS-induced colitis in adult offspring were significantly worsened in both sexes. In this context, in male offspring, PS50w was more deleterious than PS50.ConclusionsEarly-life exposure to PS50 or PS50w has long-lasting consequences on gut physiology, and effects show sexual dimorphism. The results presented raise questions on the kinetics of later-life effects linked to perinatal PS50 and PS50w exposure, and introduce the notion of imprinting.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12989-026-00700-w. Background Today, plastic represents one of the main environmental pollutants, due to the constant increase in its production and use over the last century, combined with the mismanagement of its decycling. According to the Organization for Economic Co-operation and Development (OECD), the global annual production of plastic rose from 0.5 million metric tons (Mt) in 1950, to 465 Mt in 2019. Approximately 82 Mt of this plastic is mismanaged and littered, with 22 Mt of plastic material ending up in the environment every year [1]. Once in the environment, plastics degrade to become microparticles (MPL; between 0.1 and 5000 µm) and nanoparticles (NPL; between 1 and 100 nm) [2]. These particles also undergo weathering due to exposure to UV, temperature variations and oxidation. The omnipresence of plastic in the environment, in the food chain and in food packaging has raised questions about human exposure and the dangers of MPL and NPL. Exposure mainly occurs through ingestion [3], due to the presence of MPL and NPL in many dietary products, including seafood, honey, sugar, salt, drinking water [4] and meat [5]. Previous studies have estimated that humans ingest between 39,000 and 52,000 MPL annually [4]. No completely reliable data are available for NPL, which are more difficult to detect, particularly as assays can be hindered by cross contamination [6]. The physical and chemical properties of MPL, and even more so NPL, allow them to cross intestinal epithelial barrier. As a result, plastic particles have been detected in numerous human tissues [7–14]. So far, most studies seeking to detect plastic particles have focused on MPL, and polyester (PES) has emerged as the main contaminant [15]. A recent article analyzing MPL and NPL in bottled water reported an estimated 105 particles per liter, of which 20 to 40% were Polystyrene (PS) NPL [16]. Soil analysis gives even more striking results, with detection of up to 100 mg of plastic/kg of soil [17]. Once again, PS is the most prevalent type of plastic detected. As extruded and expanded polystyrene have long been widely used in food packaging, other teams have studied PS leakage into packaged foods, and detected contamination of meat in particular [18]. The question of how this contamination affects consumers’ health remains an open one. In parallel to the increasing production and use of plastics, non-communicable diseases, including intestinal disorders, are becoming more prevalent worldwide [19]. Adverse food reactions are associated with a breakdown in tolerance of the intestinal immune system toward food antigens, whereas inflammatory bowel diseases (IBD) are associated with a breakdown in tolerance toward the natural gut microbiota. Both adverse food reactions and IBD are multifactorial, involving genetic and environmental factors. According to the Developmental Origin of Health and Diseases (DOHaD) concept, the two disorders could be triggered during perinatal development [20, 21]. Indeed, the perinatal period represents a highly dynamic transition during which the intestinal microbiota, the epithelium and the immune system all mature to establish a homeostatic state throughout life. Adverse events in early life, such as exposure to environmental pollutants, could disrupt this homeostasis, leading to the development of intestinal diseases in adulthood. As MPL have been detected in the placenta and in breast milk, fetal and early-life exposure are entirely plausible [14, 22]. This study used mouse models to study the effects of perinatal exposure to 50-nm nanoparticles of polystyrene in pristine (PS50) or weathered (PS50w) states. We tested how exposure impacted oral tolerance establishment and IBD severity. Methods Particle preparation and characterization Polystyrene nanoparticles, 50 nm (PS50), were obtained from Polysciences (Polybeads® 15913 -10 PS-COOH 50 nm, 2.5% solids (w/v) in aqueous suspension; Coefficient of Variance (CV): 15%, 3.64 × 1014 particles/mL). Plastics degrade when exposed to environmental conditions, such as UV light and high temperatures. Such weathering can be simulated (for review, see [23]). Devices such as the Q-SUN Xe1 xenon-arc test chamber (Q-Labs, Bolton, UK) have been commercialized to standardize exposure conditions [24]. The conditions for use of test chambers to artificially weather plastics are set out in ISO 4892–2:2013 [25] and were followed here. Briefly, aqueous suspensions of PS-COOH were placed in a quartz reactor and weathered in a Q-SUN Xe1 xenon-arc test chamber (Q-Labs, Bolton, UK) equipped with a daylight filter, set to an irradiance of 1.44 W.m-2.nm−1 (measured at 420 nm), with no humidity control. During weathering, PS-COOH suspensions were continuously stirred using a magnetic stirrer. The test chamber irradiance was calibrated using a dedicated radiometer before each weathering campaign. The air temperature in the chamber was set to 40 °C to mimic a typical summertime temperature (at this setting, the thermometer placed on a black panel gave a reading of 60 °C). Based on preliminary experiments indicating color changes associated with reticulation of the polymer, the weathering time was set to 96 h. Mice Nulliparous female and male C3H/HeNRj mice (Janvier Labs, Le Genest St Isle, France) were housed in specific pathogen-free (SPF) experimental animal facilities (US006/CREFRE), under 12 h-12 h light-dark cycle, at constant temperature (23 °C +/- 1°C). Pathogen-free conditions were monitored every 6 months, in accordance with the FELASA high-standard recommendations. Animals were fed with SAFE® chow (A04 entretien) and had ad libitum access to filtered tap water. Animal ethics declaration All experimental procedures were approved by the French Ministry for higher education, research and innovation (project authorization number: APAFIS #34331-2021121309394360 v4) and were conducted in accordance with European directive 2010/63/UE. Perinatal exposure to 50-nm polystyrene The experimental design is summarized in Fig. 1. For each exposure condition, three females were housed with two males for 4 days, then males were removed. On the 15th day of gestation (G15), gestating females were identified, individually isolated and randomly assigned to an experimental group: 1.25 mg 50-nm polystyrene (PS50), 1.25 mg of 50-nm weathered polystyrene (PS50w), or vehicle (Mili-Q® purified water). The total dose corresponded to 0.35 mg/g of body weight (g bw)/week for a 25 g mouse. Gestating females received oral gavage with test compounds or vehicle every day from G15 to weaning of offspring on Postnatal Day (PND) 21. The sex ratio of offspring was monitored, and litter sizes were standardized at 6 +/- 1 pups to ensure optimal nursing. Undercount or unisex litters were excluded from the protocol, and supernumerary offspring were euthanized at PND1. At weaning (PND21), mothers were euthanized by cervical dislocation, and jejunum, colon and serum were collected for analysis. To avoid major litter effects, weaned pups from litters in the same experimental group were mixed before assigning them to either the IBD model or the adverse food reaction model.Fig. 1Experimental design to study the impact of perinatal exposure to nanoparticles of polystyrene. C3H/HeNRj mice were indirectly perinatally exposed to 50-nm pristine polystyrene (PS50) or weathered polystyrene (PS50w). Their mothers received oral gavage daily with 1.25 mg of PS50, PS50w or water from the 15th day of gestation (G15) to weaning (postnatal day PND21). Litters were homogenized to 6 +/- 1 pups at birth. Offspring from each exposure group were randomly assigned to the adverse food reaction (AFR) or inflammatory bowel diseases (IBD) protocols. For the AFR protocol, on PND15, mice received oral gavage with ovalbumin (OVA) (tolerized group) or water (immunized group). On PND21, they were immunized with OVA with complete Freund’s adjuvant (CFA) and received a booster immunization (boost) at PND35. They were euthanized one week later (PND42). For the IBD protocol, on PND63, colitis was induced by exposing offspring to 3% dextran sulfate sodium (DSS) in drinking water for 5 days, followed by 2 days of clearance with drinking water before euthanasia Experimental design to study the impact of perinatal exposure to nanoparticles of polystyrene. C3H/HeNRj mice were indirectly perinatally exposed to 50-nm pristine polystyrene (PS50) or weathered polystyrene (PS50w). Their mothers received oral gavage daily with 1.25 mg of PS50, PS50w or water from the 15th day of gestation (G15) to weaning (postnatal day PND21). Litters were homogenized to 6 +/- 1 pups at birth. Offspring from each exposure group were randomly assigned to the adverse food reaction (AFR) or inflammatory bowel diseases (IBD) protocols. For the AFR protocol, on PND15, mice received oral gavage with ovalbumin (OVA) (tolerized group) or water (immunized group). On PND21, they were immunized with OVA with complete Freund’s adjuvant (CFA) and received a booster immunization (boost) at PND35. They were euthanized one week later (PND42). For the IBD protocol, on PND63, colitis was induced by exposing offspring to 3% dextran sulfate sodium (DSS) in drinking water for 5 days, followed by 2 days of clearance with drinking water before euthanasia At the end of the experiment period, all mice were euthanized by cervical dislocation. On figures, data for at least 7 pups are shown. Those 7 pups came from at least 3 independent experiments, involving at least 2 litters/condition. Consequently, a minimum of 6 litters were analyzed per condition. Adverse food reaction (AFR) model At PND15, offspring exposed to PS50, PS50w or vehicle (controls), were tested for oral tolerance induction (OVA-tolerized) and immunization (OVA-immunized) as indicated in Table 1. Briefly, pups in the OVA-tolerization group received 20 mg Grade V ovalbumin from chicken egg white (OVA) (Sigma-Aldrich®, A5503-10 G) by oral gavage in 100 µL of drinking water; pups in the OVA-immunization group received water alone. Groups were sex-matched groups and equal in number. The groups were identified by ear punches to allow appropriate mixing between experimental groups at weaning. One week after priming (PND21), all pups received a subcutaneous injection of 25 µg OVA in PBS, mixed 1:1 with complete Freund’s adjuvant (CFA). A booster subcutaneous injection of 10 µg OVA in PBS was administered 2 weeks later (PND35). Body weights were monitored throughout the process. Following exposure, pups were euthanized by cervical dislocation. Sera were collected by intracardiac puncture and feces by anus sampling. Samples were stored at −80 °C for subsequent biochemical analysis.Table 1Treatments and doses used in the adverse food reaction modelAge of micePND15PND21PND35PND42OVA-tolerizedGavage OVA 20 mg in drinking waterOVA in CFA25 µg scOVA boost10 µg scSacrificeHumoral responseOVA-immunizedGavage drinking waterCFA: Complete Freund’s adjuvant; OVA: ovalbumin; sc: Subcutaneous; PND: Postnatal Day Table 1Treatments and doses used in the adverse food reaction modelAge of micePND15PND21PND35PND42OVA-tolerizedGavage OVA 20 mg in drinking waterOVA in CFA25 µg scOVA boost10 µg scSacrificeHumoral responseOVA-immunizedGavage drinking waterCFA: Complete Freund’s adjuvant; OVA: ovalbumin; sc: Subcutaneous; PND: Postnatal Day Treatments and doses used in the adverse food reaction model OVA in CFA 25 µg sc OVA boost 10 µg sc Sacrifice Humoral response CFA: Complete Freund’s adjuvant; OVA: ovalbumin; sc: Subcutaneous; PND: Postnatal Day Colitis model To test for susceptibility to DSS-induced colitis at PND63, offspring were exposed to 3% (w/v) Dextran Sulfate Sodium (DSS) (MP Biomedical®, 160110) in drinking water for 5 days, followed by 2 days of clearance with tap water. Control mice received drinking water without DSS. Colitis severity was scored by daily observation of the Disease Activity Index (DAI), which includes the following parameters (Table 2): % weight loss, consistency of feces, blood in feces/anus, and general appearance. DAI scoring was adapted from Cooper et al. [26, 27], with a maximum possible score of 12. On the 7th day of induction (PND70), feces were collected and stored at −80 °C, and animals were euthanized to collect ileum and colon, which were immediately frozen in liquid nitrogen and stored at −80 °C for future biochemical analysis. At the time of euthanasia, macroscopic observations of intestinal inflammation were scored based on the following parameters (Table 3): colon length, colon thickness, edema size, presence of hyperemia, mesenteric adherences and presence of feces in the colon. Scoring was adapted from Wallace et al. [28], with a maximum score of 17.Table 2Disease activity indexScore DAI% Weight lossConsistency of fecesBlood in feces/anusGeneral appearance00normalnonormal10–5unformedyesbristly25–10light diarrheaprostrate310–15heavy diarrheaprostrate and bristly415–20lethargicmax = 124314DAI: Disease activity index; max: maximumTable 3Macroscopic damage scoreMacroscopic scoreColon length (cm)Colon thickness (mm)Oedema sizeHyperemiaAdherenceFeces in colon0x ⩾ 7.5x < 0.30absenceabsenceyes17.5 > x ⩾6.50.35 > x ⩾0.30.5 ⩾ x > 0presencelowno26.5 > x ⩾60.40 > x ⩾0.351 ⩾ x > 0.5-medium36 > x ⩾5.50.45 > x ⩾0.401.5 ⩾ x > 1-high4x < 5.5x > 0.45x > 1.5--max = 17444131max: maximum Table 2Disease activity indexScore DAI% Weight lossConsistency of fecesBlood in feces/anusGeneral appearance00normalnonormal10–5unformedyesbristly25–10light diarrheaprostrate310–15heavy diarrheaprostrate and bristly415–20lethargicmax = 124314DAI: Disease activity index; max: maximum Disease activity index DAI: Disease activity index; max: maximum Table 3Macroscopic damage scoreMacroscopic scoreColon length (cm)Colon thickness (mm)Oedema sizeHyperemiaAdherenceFeces in colon0x ⩾ 7.5x < 0.30absenceabsenceyes17.5 > x ⩾6.50.35 > x ⩾0.30.5 ⩾ x > 0presencelowno26.5 > x ⩾60.40 > x ⩾0.351 ⩾ x > 0.5-medium36 > x ⩾5.50.45 > x ⩾0.401.5 ⩾ x > 1-high4x < 5.5x > 0.45x > 1.5--max = 17444131max: maximum Macroscopic damage score max: maximum Colon permeability at PND70 Ex vivo permeability assays were performed in Ussing chamber systems as previously described [29]. Colon fragments from animals in the colitis model group were mounted in Ussing chambers (Physiologic Instruments, San Diego, CA, USA) [30]. Tissues were bathed for 1 h at 37 °C in oxygenated Ringer’s solution containing the following (in mmol/L): 140 Na+, 5.2 K+, 120 Cl−, 25 HCO3−, 1.2 Ca2+, 2.4 HPO42−, 1.2 Mg2+. Fluorescein isothiocyanate–dextran 4 kDa (FD4; Sigma-Aldrich®) 0.4 mg/mL was added to the mucosal compartment. Colon permeability to FD4 was determined by measuring the fluorescence intensity (FI) 485 nm/525 nm in the serological compartment, using an automatic EnSight™, Multimode microplate reader. Permeability was calculated as the ratio of flux/concentration, and expressed as nmol/h/cm2, as previously described [31]. Protein in feces Feces were lysed in 1 mL of PBS with protease inhibitors using Lysing Matrix D tubes (Fast prep ref: 6913,500) and a Precellys 24 Touch homogenizer. The homogenate was centrifuged at 1600 g (4 °C, 10 min). Protein concentration was measured in the supernatant using an Uptima BCA kit (Interchim, Montlucon, France). Serum preparation At euthanasia, intracardiac blood was collected in heparin-free tubes. Whole blood was centrifuged for 10 minutes at 1000 g and serum was collected and stored at −80 °C until analysis. Antibody detection in feces and plasma MaxiSorp™ plates were coated with 5 µg/mL of sheep anti-mouse IgA (Sigma) or goat anti-mouse IgG (SouthernBiotech, Cliniscience, Nanterre, France) antibodies. Plasma or feces supernatant was then applied. Binding was detected with 1.5 µg/mL HRP-conjugated goat anti-mouse IgA (Sigma) or goat anti-mouse IgG (SouthernBiotech). HRP was revealed using TMB, and the reaction was stopped by adding H2SO4 before reading at 450 nm in an automatic Varioskan™ microplate reader. ELISA titration of anti-OVA IgG/IgG1/IgG2a Anti-ovalbumin Ig were quantified using an ELISA titration protocol. Ovalbumin antigen was coated onto a 96-well Nunc MaxiSorp™ plate. Plates were blocked with 5% Fetal Bovine Serum (FBS) in PBS before adding serial dilutions of serum samples. Binding was detected with HRP-conjugated anti-mouse IgG, IgG1 or IgG2a (SouthernBiotech, Cliniscience, Nanterre, France). HRP was revealed as above for antibody detection. Calprotectin measurement in feces Calprotectin levels were determined in feces supernatant using Duoset ELISA kits according to the manufacturer’s protocol (S100 Calcium Binding Protein A8 - S100A8). Blood glucose Blood glucose levels were monitored in the tail vein with a glucose meter (Johnson & Johnson, Issy-les-Moulineaux, France) in the morning (fed conditions) and in the evening (fasting conditions). Our intention was to provide exploratory data without adding fasting stress to our conditions. Statistical analysis Outliers were excluded from all datasets using the ROUT test. Remaining data were statistically analyzed using GraphPad Prism 10.00 (GraphPad software, San Diego, CA). Results (interquartile range, median and minimum and maximum values) are summarized in box-and-whiskers plots. Charts expressing kinetics of effects show mean ± (SEM) values. For between-group comparisons, statistical significance was determined using i) a one-way analysis of variance (ANOVA) with post-hoc Tukey’s test for charts with one experimental factor, ii) a two-way analysis of variance (ANOVA) with post-hoc Tukey’s test for charts with two experimental factors, iii) a repeated measure 2-way analysis of variance (ANOVA) with post-hoc Tukey’s test, or Mixed-effects analysis with post-hoc Tukey’s test in case of missing values. The statistical test used is indicated in the legend of each chart. A threshold of p < 0.05 was set for statistical significance. All p values indicated are 2-tailed. Results Physico-chemical characterization of PS particles Pristine and weathered PS particles were imaged by transmission electron microscopy (Fig. 2A–B, respectively). Their size distribution is reported in Fig. 2C–D. Pristine and weathered particles were also characterized in terms of their average hydrodynamic diameter and zeta potential (Fig. 2E). Their mean hydrodynamic diameters were determined. For pristine particles: 49.96 ± 0.61 polydispersity index 0.053 ± 0.010; for weathered particles: 49.9 ± 1.29 nm polydispersity index 0.048 ± 0.004). Zeta potentials of −34.2±2.9 mV (SD) for pristine mV and −30.0±1.2 mV (SD) for weathered particles were determined.Fig. 2Shape and size of PS particles. Pristine (A) and weathered (B) PS50-COOH nanoparticles were imaged by electron microscopy. Size distributions in the two suspensions were measured via dynamic light scattering (pristine: C; weathered: D). The two suspensions were chemically characterized (E) Shape and size of PS particles. Pristine (A) and weathered (B) PS50-COOH nanoparticles were imaged by electron microscopy. Size distributions in the two suspensions were measured via dynamic light scattering (pristine: C; weathered: D). The two suspensions were chemically characterized (E) Oral exposure to 50-nm PS nanoparticles has no effect on dams Gestating and lactating mice were exposed to daily oral gavage with PS50 or PS50w for 4 weeks. Physiological and serological parameters were monitored in these animals at 3 different end points: lactating day 0, 15 and 21 (L0, L15 and L21) (Fig. 3A). At L0 no effects on mortality of dams nor on litter size or sex ratio were observed (Fig. 3B–C). At L15, no effect on fecal IgG and IgA was detected (Fig. 3 D-E). At L21, no effects on weight, intestine length, IgG and IgA levels in feces (Fig. 3F–J) or serum IgG concentrations were detected (Supplementary Figure 1B). Similarly, fasting and fed blood glucose concentrations were not altered in dams over the 4 weeks of daily oral gavage with PS50 or PS50w (Supplementary Figure 1C) at L21.Fig. 3Oral exposure to 50-nm PS nanoparticles has no effect on dams. (A) Dams exposed to 50-nm polystyrene (PS50) or weathered polystyrene (PS50w) were monitored for several physiological characteristics. (A-J) Physiological and serological parameters were monitored in these animals at 3 different end points lactating day 0, 15 and 21 (L0, L15 and L21). The following characteristics were measured: (B) mortality among dams at birth L0, (C) number of live pups per litter and sex ratio, (D) fecal IgG at L15, (E) fecal IgA at L15 (F) mum’s body weight at L21, (G) small intestine length at L21, (H) large intestine length at L21, (I) fecal concentration of (I) IgG and (J) IgA at L21. Data are represented as box and whisker plots (median, min to max). Statistical significance was analyzed by either two-way ANOVA (C) or one-way ANOVA (D, E , F, G, H, I, J) following a Tukey’s multiple comparisons test. *p < 0.05 vs. indicated group. For mothers, there were at least n = 16 mice per group from at least 3 independent experiments. For offspring, there were at least n = 10 mice per group in at least 3 independent experiments; two litters per condition were pooled before analysis Oral exposure to 50-nm PS nanoparticles has no effect on dams. (A) Dams exposed to 50-nm polystyrene (PS50) or weathered polystyrene (PS50w) were monitored for several physiological characteristics. (A-J) Physiological and serological parameters were monitored in these animals at 3 different end points lactating day 0, 15 and 21 (L0, L15 and L21). The following characteristics were measured: (B) mortality among dams at birth L0, (C) number of live pups per litter and sex ratio, (D) fecal IgG at L15, (E) fecal IgA at L15 (F) mum’s body weight at L21, (G) small intestine length at L21, (H) large intestine length at L21, (I) fecal concentration of (I) IgG and (J) IgA at L21. Data are represented as box and whisker plots (median, min to max). Statistical significance was analyzed by either two-way ANOVA (C) or one-way ANOVA (D, E , F, G, H, I, J) following a Tukey’s multiple comparisons test. *p < 0.05 vs. indicated group. For mothers, there were at least n = 16 mice per group from at least 3 independent experiments. For offspring, there were at least n = 10 mice per group in at least 3 independent experiments; two litters per condition were pooled before analysis At weaning, weight of female offspring is increased following perinatal exposure to PS50w Offspring were analyzed at different end points: post-natal day 15 (PND15), corresponding to the beginning of the oral tolerance protocol; PND21, corresponding to weaning; and PND42, corresponding to the end of the oral tolerance protocol (Fig. 4A). At PND15, fecal IgG levels were similar whatever the treatment applied (Fig. 4B). In contrast, fecal IgA levels were higher in female pups following exposure to PS50w (Fig. 4C). At PND21, body weight of female offspring was increased in litters exposed to PS50w, whereas male offspring showed no such weight difference (Fig. 4D). Exposure to PS50 had no detectable effect at this timepoint.Fig. 4Oral tolerance establishment in offspring at postnatal day 15 is not affected by perinatal exposure to PS50 nor PS50w. (A) Mice exposed perinatally to 50-nm polystyrene (PS50) or weathered polystyrene (PS50w) were monitored at PND15 beginning of oral tolerance protocol, PND21 weaning and PND42 end of oral tolerance protocol (A). PND15 fecal concentrations of (B) IgG and (C) IgA. (D) At PND21 body weight were assessed. Then, at PND42, (E) anti-OVA IgG, (F) anti-OVA IgG1 and (G) anti-OVA IgG2 titers were measured in serum. Data were represented as box and whisker (median, min to max) plots. They were analyzed by either one-way ANOVA (fig. B, C, D) or two-way ANOVA (fig. E, F, G) following a Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01; ***p < 0.001 and ****p < 0.0001 vs. indicated group. For offspring, there is at least n = 10 mice per group from at least 3 independent experiment including 2 litters by condition were pooled and analyzed Oral tolerance establishment in offspring at postnatal day 15 is not affected by perinatal exposure to PS50 nor PS50w. (A) Mice exposed perinatally to 50-nm polystyrene (PS50) or weathered polystyrene (PS50w) were monitored at PND15 beginning of oral tolerance protocol, PND21 weaning and PND42 end of oral tolerance protocol (A). PND15 fecal concentrations of (B) IgG and (C) IgA. (D) At PND21 body weight were assessed. Then, at PND42, (E) anti-OVA IgG, (F) anti-OVA IgG1 and (G) anti-OVA IgG2 titers were measured in serum. Data were represented as box and whisker (median, min to max) plots. They were analyzed by either one-way ANOVA (fig. B, C, D) or two-way ANOVA (fig. E, F, G) following a Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01; ***p < 0.001 and ****p < 0.0001 vs. indicated group. For offspring, there is at least n = 10 mice per group from at least 3 independent experiment including 2 litters by condition were pooled and analyzed Oral tolerance at postnatal day 15 is not affected by perinatal exposure to PS50 or PS50w We then investigated whether perinatal exposure to PS50 or PS50w affected the establishment of oral tolerance upon introduction of solid food (PND15). These experiments also allowed us to examine the consequences of perinatal exposure to PS50 and PS50w on the immune response to vaccination/immunization (Fig. 4E–G). Oral tolerance is characterized by an absence of humoral response to an orally administered antigen (OVA in our model). In our model, pups were orally exposed to OVA or vehicle at PND15. At PND21, OVA was administered subcutaneously to mimic vaccination. The immune response against the OVA antigen was monitored. In control offspring (i.e., not exposed to PS50 or PS50w), OVA tolerization decreased serum anti-OVA IgG titers 1 000-fold compared to the non-tolerized group (Fig. 4E). These data validate our novel protocol for inducing oral tolerance to OVA at PND15. In groups exposed to PS50 or PS50w, OVA tolerization also significantly decreased serum anti-OVA IgG titers, with 1 000-fold differences compared to the equivalent non-tolerized groups (Fig. 4E). Thus, the immune response in offspring was not impaired by perinatal exposure to PS50 or PS50w. We extended our analysis to determine any effects on immune response pathway by studying the Th2 (IgG1 - Fig. 4F) and Th1 (IgG2a - Fig. 4G) responses. In immunized male offspring exposed to PS50w, higher anti-OVA IgG1 titers were detected than in control or PS50-exposed groups (Fig. 4F). In contrast, perinatal exposure to PS50 increased anti-OVA IgG2a titers compared to control group (Fig. 4G). Thus, although perinatal exposure to PS50w increased fecal IgA concentration in female offspring at PND15, no effect on oral tolerance induction was observed. However, in immunized males, perinatal exposure to PS50w affected the immune response pathway, leading to increased serological anti-OVA IgG1 levels; perinatal exposure to PS50 increased serological anti-OVA IgG2a levels in animals of both sexes. Long-term consequences of perinatal exposure to PS50 or PS50w We analyzed the long-term consequences of perinatal exposure to PS50 or PS50w on offspring at PND56, one week before inducing colitis (Supplementary Figure 2A male and Figure 4A female). No impact on blood glycemia, whether natural fasting (male offspring: Supplementary Figure 2B; female offspring: Supplementary Figure 4B) or fed (male offspring: Supplementary Figure 2C; female offspring: Supplementary Figure 4C), was observed. Body weight at PND56 was increased among offspring perinatally exposed to PS50w, but not among those exposed to PS50 (males: Supplementary Figure 2D; females: Supplementary Figure 4D). We investigated whether perinatal exposure to PS50 or PS50w could aggravate DSS-induced colitis at PND63. DSS-induced colitis is exacerbated by perinatal exposure to PS50 and PS50w, with a stronger effect of PS50w in male offspring The increased body weight at PND56 was maintained in males exposed to PS50w at PND63 (beginning of DSS-induced colitis protocol Fig. 5A) (Fig. 5B).Fig. 5DSS-induced colitis is exacerbated by both perinatal PS50 and PS50w exposure with stronger effect by PS50w in males offspring. (A) Mice were perinatally exposed to 50 nm polystyrene (PS50) or weathered polystyrene (PS50w) via their mother, receiving daily oral gavage from days 15th of gestation (G15) to weaning (postnatal day 21) with 1.25 mg of PS50, PS50w or water (A). At PND63, offspring has been (B) weighed and colitis has been orally induced by introducing dextran sulfate sodium (DSS) into drinking water at 3% for 5 days followed by 2 days of regular water. During the experimental DSS procedure (7 days), (C) body weight and the (D) disease activity index (DAI) have been monitored daily. (E-I) Then, at the end of the DSS procedure (PND70), mice have been euthanized and (E) macroscopic score, (F) colon length and (G) thickness, (H) edema size and (I) adherence score have been monitored. Data were represented as box and whisker (median, min to max) plots. They were analyzed by either one-way ANOVA (fig. B) or mixed-effects analysis (fig. C) or repeated measures two-way ANOVA (fig. D) or two-way ANOVA (fig. E, F, G, H, I) following a Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01 and ****p < 0.0001 vs. indicated group. For offspring, there is at least n = 10 mice per group in at least 3 independent experiment including 2 litters by condition were pooled and analyzed DSS-induced colitis is exacerbated by both perinatal PS50 and PS50w exposure with stronger effect by PS50w in males offspring. (A) Mice were perinatally exposed to 50 nm polystyrene (PS50) or weathered polystyrene (PS50w) via their mother, receiving daily oral gavage from days 15th of gestation (G15) to weaning (postnatal day 21) with 1.25 mg of PS50, PS50w or water (A). At PND63, offspring has been (B) weighed and colitis has been orally induced by introducing dextran sulfate sodium (DSS) into drinking water at 3% for 5 days followed by 2 days of regular water. During the experimental DSS procedure (7 days), (C) body weight and the (D) disease activity index (DAI) have been monitored daily. (E-I) Then, at the end of the DSS procedure (PND70), mice have been euthanized and (E) macroscopic score, (F) colon length and (G) thickness, (H) edema size and (I) adherence score have been monitored. Data were represented as box and whisker (median, min to max) plots. They were analyzed by either one-way ANOVA (fig. B) or mixed-effects analysis (fig. C) or repeated measures two-way ANOVA (fig. D) or two-way ANOVA (fig. E, F, G, H, I) following a Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01 and ****p < 0.0001 vs. indicated group. For offspring, there is at least n = 10 mice per group in at least 3 independent experiment including 2 litters by condition were pooled and analyzed In male offspring, following DSS-induction of colitis, we observed a greater loss of body weight (Fig. 5C) and higher DAI scores (Fig. 5D), including for general appearance (Supplementary Figure 3B) and feces consistency (Supplementary Figure 3C). These parameters were unaffected by perinatal exposure to PS50 or PS50w (Fig. 5C - D and Supplementary Figure 3B - C). Higher fecal IgG (Supplementary Figure 3D), IgA (Supplementary Figure 3E), and calprotectin concentrations (Supplementary Figure 3F) were measured following DSS-induction, and colonic permeability to FD4 was enhanced (Supplementary Figure 3 G). Once again, perinatal exposure to PS50 or PS50w had no further effect on these parameters. However, the macroscopic score for colitis was increased in both PS50 groups compared to the control group, and the increase was more marked in the PS50w group (Fig. 5E). Among the parameters of macroscopic score most affected were increased colon thickness (Fig. 5G) and edema size (Fig. 5H). Colon length (Fig. 5F) was only exacerbated following perinatal exposure to PS50w. Perinatal exposure to both forms of PS50 worsened DSS-induced mesenteric adherences (Fig. 5I) compared to the control group. Thus, in adult male offspring, perinatal exposure to PS50 or PS50w worsened colitis severity at induction in adulthood, with PS50w eliciting a stronger effect. DSS-induced colitis is exacerbated by perinatal exposure to PS50 and PS50w in female offspring In female offspring, the increase in body weight in the PS50w-exposed group observed at PND15 persisted at PND56, and continued to PND63 (beginning of DSS-induced colitis protocol Fig. 6A) (Fig. 6B).Fig. 6DSS-induced colitis is exacerbated by both perinatal PS50 and PS50w perinatal exposure in female offspring. (A) Mice were perinatally exposed to 50 nm polystyrene (PS50) or weathered polystyrene (PS50w) via their mother, receiving daily oral gavage from days 15th of gestation (G15) to weaning (postnatal day 21; PND21) with 1.25 mg of PS50, PS50w or water. At PND63, offspring has been (B) weighed and colitis has been orally induced by introducing dextran sulfate sodium (DSS) into drinking water at 3% for 5 days followed by 2 days of regular water. During the experimental DSS procedure (7 days), (C) body weight and the (D) disease activity index (DAI) have been monitored daily. (E-I) Then, at the end of the DSS procedure PND70, mice have been euthanized and (E) macroscopic score, (F) colon length and (G) thickness, (H) edema size and (I) adherence score have been monitored. n = 10 mice minimum per group from at least 2 independent litters. Data were represented as box and whisker (median, min to max) plots. They were analyzed by either one-way ANOVA (fig. B) or mixed-effects analysis (fig. C, D) or two-way ANOVA (fig. E, F, G, H, I) following a Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01; ***p < 0.001 and ****p < 0.0001 vs. indicated group. For offspring, there is at least n = 10 mice per group in at least 3 independent experiment including 2 litters by condition were pooled and analyzed DSS-induced colitis is exacerbated by both perinatal PS50 and PS50w perinatal exposure in female offspring. (A) Mice were perinatally exposed to 50 nm polystyrene (PS50) or weathered polystyrene (PS50w) via their mother, receiving daily oral gavage from days 15th of gestation (G15) to weaning (postnatal day 21; PND21) with 1.25 mg of PS50, PS50w or water. At PND63, offspring has been (B) weighed and colitis has been orally induced by introducing dextran sulfate sodium (DSS) into drinking water at 3% for 5 days followed by 2 days of regular water. During the experimental DSS procedure (7 days), (C) body weight and the (D) disease activity index (DAI) have been monitored daily. (E-I) Then, at the end of the DSS procedure PND70, mice have been euthanized and (E) macroscopic score, (F) colon length and (G) thickness, (H) edema size and (I) adherence score have been monitored. n = 10 mice minimum per group from at least 2 independent litters. Data were represented as box and whisker (median, min to max) plots. They were analyzed by either one-way ANOVA (fig. B) or mixed-effects analysis (fig. C, D) or two-way ANOVA (fig. E, F, G, H, I) following a Tukey’s multiple comparisons test. *p < 0.05; **p < 0.01; ***p < 0.001 and ****p < 0.0001 vs. indicated group. For offspring, there is at least n = 10 mice per group in at least 3 independent experiment including 2 litters by condition were pooled and analyzed In all female offspring, exposure to DSS led to a loss of body weight (Fig. 6C) and increases in DAI scores (Fig. 6D), including for general appearance (Supplementary Figure 5B) and feces consistency (Supplementary Figure 5C). On day 5 of DSS treatment, the PS50w group had a higher DAI score than the control or PS50 groups (Fig. 6D). Among the factors of DAI most affected, both general appearance (Supplementary Figure 5B) and feces consistency (Supplementary Figure 5C) were significantly modified, suggesting a multifactorial cumulative effect. In all groups, DSS-induction of colitis led to increased fecal IgG (Supplementary Figure 5D) and IgA concentrations (Supplementary Figure 5E). Intestinal permeability was also increased following DSS-induction of colitis (Supplementary Figure 5G). Unexpectedly, perinatal exposure to PS50w decreased DSS-induced colonic hyperpermeability compared to PS50. Finally, the increased fecal calprotectin concentration detected in mice with induced colitis was further enhanced in females exposed to PS50 (Supplementary Figure 5F). Macroscopic scores for colitis were increased in females perinatally exposed to PS50 and PS50w compared to control animals (Fig. 6E). Among the items contributing to this increase in macroscopic score, we observed a significant decrease in colon length (Fig. 6F) and significant increases in colon thickness (Fig. 6G), edema size (Fig. 6H) and mesenteric adherences (Fig. 6I). These differences were similar in both the PS50 and PS50w groups. Thus, in adult female offspring, perinatal exposure to PS50 or PS50w aggravated colitis severity, with observable effects from day 5 of colitis induction in the group exposed to PS50w. Discussion Given the widespread distribution of nanoplastics in our environment and in the human food chain, there is an urgent need to study their potential impact on health, especially with regard to the onset of non-communicable diseases. Perinatal exposure, during gestation and lactation, might represent a critical timeframe due to fetal and neonate maturation. At steady state, neither pristine nor weathered polystyrene nanoparticles impacted intestinal, metabolic or immunological parameters in gestational and lactating dams. In contrast, among offspring, perinatal exposure had effects, with a certain degree of sexual dimorphism. Thus, exposure to PS50w led to increased body weight in males at adulthood, and in females from PND21 (weaning). Perinatal exposure to PS50 and PS50w exacerbated colitis severity in both males and females at adulthood (PND63); PS50w had a stronger effect, but only in male offspring. Tolerance toward food antigens was not affected at solid food introduction (PND15) nor following systemic immunization. Compared to MPL, the small size of NPL makes them difficult to quantify and identify in the environment [32]. They therefore represent an invisible threat for environmental, animal and human health. With this study, we wanted to compare the consequences of perinatal exposure to pristine and weathered PS50 to mimic the degradation of plastic present in the environment and to explore potential additional toxic effects. Gestating and lactating mothers were orally exposed, as solid food [32] and beverages [4] represent the main sources of human exposure to NPL. The literature contains widely different estimations of the total mass of MP ingested, with weekly adult exposure ranging from the equivalent of 4.1 μg [33] to 5 g [34], i.e., a difference of six orders of magnitude. In 2022, Pletz [35] revisited the calculations, results and methods of these two studies, and concluded that 4.1 µg/week of MP ingested was a more realistic estimate. Dose of exposure to MNP is still under debate due to limited device and technology to detect and quantify MNP and might dramatically vary depending on the region. Indeed, Zhao and You [36] estimated MNP ingestion to 1.8 g/month in France, 2.4 g/month in USA and 12.9 g/month in Indonesia (close to the dose used in this study). Of note, Zhao and You highlight a tremendous increase (6 fold) of ingested and inhaled particles from 1990 to 2018 [36]. First, we analyzed the physiological condition of mothers and offspring of perinatal to determine the consequences of exposure to PS50 or PS50w. Exposure to PS50 or PS50w during gestation did not induce mortality, and had no effect on sex ratio at birth. At PND15, before initiating the oral tolerance protocol, higher fecal IgA concentrations were detected in female offspring exposed to PS50w; IgG levels were not altered. These results suggest that perinatal exposure to PS50w could increase the intestinal immune response in female offspring. At weaning (PND21), no effects of PS50 exposure (pristine or weathered) were observed on metabolism (body weight or glycemia) in mothers receiving oral gavage. In contrast, body weight was increased in female offspring exposed to PS50w. At PND56 and PND63, increased body weight was observed in both male and female offspring. These results corroborate those of Jeong et al. [37], who also observed increased body weight in offspring following early-life exposure to PS NPL. The sexual dimorphism for this increased body weight, with earlier effects in females, was also highlighted by Jeong et al. [37] and Amereh et al. [38], who also reported differences in the secretion of sexual hormones including testosterone, luteinizing hormones and follicle-stimulating hormones. In our study of oral tolerance establishment (tolerance toward food antigens), no difference in establishment of oral tolerance was observed following perinatal exposure to PS50 or PS50w. In contrast, perinatal exposure to both PS50 and PS50w affected the immune response to systemic immunization. Thus, perinatal exposure to PS50w increased anti-OVA IgG1 titers, whereas exposure to PS50 increased anti-OVA IgG2a in immunized male offspring, suggesting stronger immune responses with a possible tendency toward allergic reactions. The lasting consequences of perinatal exposure to PS50 and PS50w were studied here by assessing IBD severity at adulthood (tolerance toward bacterial antigens). Our data showed that, in DSS-induced colitis, only female offspring following perinatal exposure to PS50 had increased fecal calprotectin concentrations. No lasting consequences of perinatal exposure to NPL were observed on intestinal barrier at steady state. These results support those of Schwarzfischer et al. [39], who reported that exposure to PS NPL through drinking water (for 6 months at 0.056 mg/g bw/week) did not modify cytokine expression or the formation of tight junctions in the colon. However, Liang et al. [40] reported that oral exposure to PS50 particles for 28 days, at a daily dose very close to the one used here, increased ROS production by intestinal epithelial cells. At higher doses, they reported increased numbers of apoptotic cells and greater intestinal permeability [40]. These stronger effects are probably due to the dose administered, the duration of exposure and the age at time of exposure. Our data indicated that perinatal exposure to PS50 or PS50w aggravates adult-onset DSS-induced colitis severity. Moreover, in males, perinatal exposure to PS50w had a more deleterious impact on colitis severity than PS50. In an inflammatory context, it was recently shown that exposure for 6 months to PS NPL (0.056 mg/g bw/week for a 25-g mouse) in drinking water did not aggravate DSS-induced colitis [39]. This divergence in results may be due to the fact that Schwarzfischer et al. studied adult animals, whereas we investigated the consequences of exposure during early life, at a critical period for the establishment of intestinal homeostasis according to the DOHaD concept. Our observations support our hypothesis that early-life exposure to PS NPL, through maternal exposure, could trigger deleterious imprinting in the gastrointestinal tract that subsequently persists throughout life. Conclusion Our results provide new insights into the potential hazards of nanoparticles of polystyrene on non-communicable diseases and the consequences of maternal exposure to nanoparticles of polystyrene on offspring health by modifying perinatal imprinting. Together, our data show that early exposure to PS50 or PS50w has a deleterious impact on gut development during the perinatal period, with lasting consequences on intestinal homeostasis leading to aggravation of colitis at adulthood in offspring of both sexes. These effects persist throughout life, and are also expressed as altered intestinal homeostasis, in particular in the form of the intestinal immune response toward intestinal microbiota. Increased body weight gain and enhanced colitis susceptibility were observed in females compared to males following perinatal exposure to polystyrene nanoparticles. This sex dimorphism emphasizes the importance of investigating effects in both sexes. The stronger effects observed with weathered particles underscore the need to distinguish effects based on plastic structure (pristine versus weathered). Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary material 1 Supplementary material 1