Abstract Fullerenols, a water-soluble polyhydroxy derivative of fullerene, hold promise in medical and materials science due to their unique properties. However, concerns about their potential embryotoxicity remain. Using a pregnancy mouse model and metabolomics analysis, our findings reveal that fullerenols exposure during pregnancy not only significantly reduced mice placental weight and villi thickness, but also altered the classes and concentrations of metabolites in the mouse placenta. Furthermore, we found that fullerenols exposure reduced the levels of CYP3A4, ERα and estriol (E3), while increasing the levels of estradiol (E2) and oxidative stress both in mouse placenta and placental trophoblast cells, and exogenous supplementation with E3 and ER agonists was effective in restoring these changes in vitro. Moreover, CYP3A4 inhibition was effective in decreasing intracellular E3 levels, whereas overexpression of CYP3A4 resisted the fullerenols-induced decrease in E3 expression Additionally, we synthesized glutathione-modified fullerenols (C[60]-(OH)[n]-GSH), which demonstrated improved biocompatibility and reduced embryotoxicity by enhancing intracellular glutathione levels and mitigating oxidative stress. In summary, our results demonstrated that fullerenols exposure decreased E3 synthesis by inhibiting CYP3A4 and exacerbated oxidative stress through downregulation of estrogen receptor activation and decreased glutathione levels. These findings highlight the risks of fullerenols exposure during pregnancy and offer strategies for safer nanomaterial development. Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03121-7. Keywords: Fullerenols, Placental development, Estriol, Oxidative stress Graphical Abstract [42]graphic file with name 12951_2025_3121_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s12951-025-03121-7. Introduction Fullerenols (fullerols, hydroxylated fullerene) is a water-soluble polyhydroxy derivative of fullerene, with symmetrically arranged hydroxyl groups based on the structure of the C[60] sphere [[43]1]. Increasing the number of hydroxyl groups in nonpolar fullerene magnifies its water solubility; for instance, fullerenes with 22 or 26 hydroxyl groups exhibit enhanced solubility due to the formation of a hydrophilic layer on the molecular surface, and those with 44 hydroxyl groups (C[60](OH)[44]·8H[2]O) demonstrate exceptional water solubility achieving up to 64.9 mg/mL [[44]2, [45]3]. The unique electronic structure, electronegativity and nanoscale of fullerenols derivatives contribute to their effectiveness as antioxidants [[46]4–[47]6]. Fullerenols are primarily utilized in a variety of fields, including biomedical applications (such as drug delivery systems and antioxidants), and material science (for the development of novel nanomaterials) [[48]7, [49]8]. In addition, it can also inhibit the toxicity of harmful substances or drugs, providing a certain degree of protection for biological bodies [[50]9]. Torres et al. found that hydroxylated fullerenols (C[60](OH)[24]) can counteract the cytotoxicity caused by doxorubicin (DOX) via antioxidant and free radical scavenging activities [[51]10]. Tolga et al. also reported that fullerenols suppress cytotoxic and genotoxic effects on human pulmonary fibroblasts caused by pesticides, thus protecting the cells [[52]11]. Although there are studies showing protective effects of fullerenols, the existing inconsistent and uncertain information on the physicochemical, toxicokinetic and toxicological aspects related to fullerenes makes it impossible to ignore its toxicity [[53]12, [54]13]. Noteworthy, the final safety opinion released in October 2023 by the European Union’s Scientific Committee on Consumer Safety (SCCS), specifically addressing fullerenes, hydroxylated fullerenes and their hydrated nanoscale forms (SCCS/1649/23), stands out for its significance [[55]14]. Currently available information led the SCCS to the conclusion that they could not rule out the possibility of genotoxicity and carcinogenicity in any of these materials under review. Research has shown that fullerenols (such as C[60](OH)[7±2] and C[60](OH)[22−26]) can trigger different “cytotoxicities,” like in mice osteosarcoma, human vascular endothelial cells, lens epithelial cells, RAW2647, HeLa cells, and more [[56]15]. C[60](OH)[44]’s potential toxicity to mitochondria not only affects the respiratory chain but also damages the mitochondrial inner membrane, increasing the permeability to H + and K+, thereby elucidating its unknown functions at the subcellular level [[57]16]. Therefore, assessing the potential adverse health outcomes that fullerenols might cause is also critically important. Furthermore, the widespread use of fullerenols nanoparticles has sparked concerns about its environmental impact, notably contaminating water sources, infiltrating the food chain, and thus impacting aquatic life [[58]17]. Considering that a fullerene’s hydroxylated shell inhibits direct core interaction, thereby allowing fullerenols to remain in a highly stable molecular dispersion and resist degradation [[59]18, [60]19], it is reasonable to hypothesize that fullerenols could potentially be widely distributed across diverse aquatic environments. In addition, the presence of fullerenols in the aquatic food chain, brought about by water contamination and subsequent ingestion by specific marine species, notably magnifies from one trophic level to another [[61]20]. Despite the lack of significant biomagnification in the higher trophic levels, the demonstrated ability of fullerenols to pervade aquatic food chains and accumulate in crucial fish organs reflects a noteworthy environmental concern. The anticipated market growth of fullerenols, coupled with their potential for direct human contact through various applications such as creams for skin or drug delivery, has raised widespread concerns about their potential adverse effects on human health [[62]17]. As it stands, cosmetics are likely the primary route of human exposure to fullerenols. They are used in cosmetics as antioxidants and skin conditioners, typically added in amounts ranging from 0.5 to 1%, although they are generally not explicitly labeled [[63]21]. As a relatively new nanomaterial, regrettably, there are currently no specific population exposure data for fullerenols to reference [[64]17]. However, it is undeniable that with the increasing application of fullerenols, human exposure, including in pregnant populations, is a certainty. As a particularly susceptible subpopulation, pregnant women are more likely to be affected by the toxicity induced by nanoparticles (NPs) [[65]22]. During pregnancy, due to alterations in the neuro-endocrine system, NPs may cross the placental barrier and enter fetal organs [[66]23]. A sequence of events potentially culminating in fetal developmental delay or abnormalities may be triggered by NPs accumulation leading to placental barrier damage, oxidative stress, inflammation and changes in gene expression [[67]24, [68]25]. Taking into account the widespread application of fullerenols and their potential toxicity [[69]26, [70]27], it becomes glaringly evident that there is a significant void in dedicated research scrutinizing the impact of fullerenols on embryonic development. Consequently, this underscores the urgent imperative for further in-depth studies in this area, as understanding this could have profound implications for the safe use of such materials in varied applications along with safeguarding maternal and fetal health. Given the broad application of fullerenols and their potential to cross the placental barrier (as nanoparticles), further investigation is warranted into the safety, toxicity, and particularly the embryotoxicity of fullerenols. This study established a pregnancy mouse exposure model using metabolomics methods to investigate the embryotoxicity of fullerenols and to explore its potential mechanisms. Subsequently, in vitro cell experiments were conducted to further verify its effects on placental and trophoblast cells, systematically elucidating the impact of fullerenols on early embryonic development and its possible mechanisms. This provides a reference for the safety evaluation of fullerenols and offers new insights into the embryotoxic mechanisms of nanomaterials. In addition, based on our research findings, we synthesized glutathione-modified fullerenols particles, which exhibit good biocompatibility and lower toxicity. This provides a new approach for the development of safer nanomaterials. Methods and materials Synthesis and characterization of fullerenols Fullerenols were synthesized using the phase-transfer catalytic method with tetrabutylammonium hydroxide (TBAH) as the catalyst [[71]2]. Firstly, a flask containing a mixture of NaOH (Macklin, Shanghai, China), TBAH, toluene solution with C[60] (SZY, Xi’an, China), and water was stirred. After 2 h, the mixture was allowed to stand until it separated into two layers, followed by filtration to obtain a dark brown solution. Methanol was added to extract a yellowish sediment, which was then dissolved and re-precipitated with methanol several times until the pH was less than or equal to 7.5. The precipitate was then dried under vacuum, dissolved in water, and left to hydrolyze for 24 h. Subsequently, it was precipitated again with methanol, rinsed a few times, and dried to yield C[60](OH)[22−26]. The synthesis products were dissolved to a concentration of 10 mg/L and then characterized. Transmission Electron Microscopy (TEM) images of the products were captured using a JEM-2000EX microscope from JEOL (Japan). The dimensions and zeta potential of the PS-NPs were evaluated with a Zetasizer Nano (ZS90, UK). These measurements were replicated thrice under a steady temperature condition of 25 ± 1 °C. Mice fullerenols exposure experiment The experiment, approved by the Animal Ethics Review Committee of Xuzhou Medical University (Ethical Approval Number: 202312T043, Approval Date: 29/12/2023), involved the use of lab animals, specifically 8-week-old Institute of Cancer Research (ICR) mice from Xuzhou Medical University’s Animal Experiment Center. The mice were separated by gender and raised on a 12-hour light/dark cycle. They were housed in an specific-pathogen-free (SPF) - grade animal room with a temperature maintained at 22 ± 2 °C and a relative humidity between 40 − 60%. Female mice were mated with male mice at a ratio of 2:1. Pregnancy was assumed upon detection of a vaginal plug the next morning, designated as gestation day 0.5 (GD0.5). Pregnant mice were randomly assigned to control and different dose groups: low dose (0.1 mg/kg/d), medium-high dose (1 mg/kg/d), and high dose (10 mg/kg/d), with two mice per cage and six mice per group (n = 6). For substances lacking population exposure data, we can refer to animal studies on the chemical itself or structurally similar chemicals to determine the dosing for animal experiments. Therefore, for our animal dosing, we primarily referred to previous studies on fullerenes and fullerenols, where the majority of doses are concentrated in the range of 1–10 mg/kg [[72]28–[73]30]. This is the basis for our choice of animal experimental dosing. Additionally, some studies have used fullerenols as protective agents, with doses of 10 mg/kg being administered [[74]31, [75]32]. Therefore, selecting dose of 0.1, 1, 10 mg/kg for fullerenols exposure allows us to evaluate the potential risks associated with its therapeutic application. From gestation day 3.5 to 17.5, the pregnant mice were administered fullerenols water solution by gavage, while the control group received drinking water instead. On gestation day 18, the fetuses and placentas were extracted for further experiments. The method of sacrifice used for the mice was cervical dislocation, which is a widely recognized and humane method of euthanasia for small rodents such as mice. Baseline data of pregnant mice were recorded throughout the experimental process. Histopathology Placental pathology examination was referenced to our previous study [[76]33]. The placenta was fixed with 4% paraformaldehyde (PFA) (Servicebio, China), dehydrated, embedded in paraffin. Three placental tissues from each group were taken and were cut longitudinally and transversely, respectively. The maximum area of these cuts was selected, with each slice being 3 micrometers thick and the interval between slices also being 3 micrometers. Histologic sections were dewaxed in xylene, rehydrated, and stained with Hematoxylin and Eosin (H&E) dyes (Servicebio, China). Sections were photographed using a digital slice scanning system (Olympus VS120, Japan). Metabolomic analysis of mice placenta In our study, we implemented the Liquid Chromatography-Mass Spectrometry (LC-MS) technique to characterize the metabolomic profile of mice placenta, based on our previous research methodology [[77]34]. Mice placenta samples were prepared by washing with PBS, then homogenizing and centrifuging to isolate the supernatant. The resulting supernatant underwent LC-MS analysis for metabolomic profiling, utilizing an High Performance Liquid Chromatography (HPLC) system for separation before entering the mass spectrometer for highly precise identification of metabolites. LC-MS metabolite screening involves separating metabolites using an HPLC system, then identifying them using a mass spectrometer. Pathway analysis maps these metabolites to known metabolic pathways via databases KEGG, with statistical analysis highlighting significantly enriched pathways. Samples were analyzed using the UHPLC-Q Exactive HF-X system from Thermo Fisher Scientific. For the chromatographic process, 3 µL of the sample were separated on an HSS T3 column (100 mm × 2.1 mm, 1.8 μm) before mass spectrometry detection. Mobile phase A was 95% water with 5% acetonitrile (containing 0.1% formic acid), and mobile phase B was 7.5% acetonitrile, 47.5% isopropanol, and 5% water (also with 0.1% formic acid). Positive and negative ion modes were adopted respectively, and the run time for a single sample was 8 min. The separation gradient in the positive ion mode was as follows: From 0 to 3 min, mobile phase B increased from 0 to 20%; from 3 to 4.5 min, mobile phase B increased from 20 to 35%; from 4.5 to 5 min, mobile phase B increased from 35 to 100%; from 5 to 6.3 min, mobile phase B remained at 100%; from 6.3 to 6.4 min, mobile phase B decreased from 100 to 0%; from 6.4 to 8 min, mobile phase B remained at 0%. The separation gradient in the negative ion mode was as follows: From 0 to 1.5 min, mobile phase B rose from 0 to 5%; from 1.5 to 2 min, mobile phase B rose from 5 to 10%; from 2 to 4.5 min, mobile phase B rose from 10 to 30%; from 4.5 to 5 min, mobile phase B rose from 30 to 100%; from 5 to 6.3 min, mobile phase B was maintained at 100%; from 6.3 to 6.4 min, mobile phase B decreased from 100 to 0%; from 6.4 to 8 min, mobile phase B was maintained at 0%. The flow rate was 0.40 mL/min, and the column temperature was set at 40 °C. Mass spectrometry parameters included acquisition in both positive and negative ion modes with a mass range of 70–1050 m/z. The sheath gas flow rate was 50 psi, the auxiliary gas flow rate was 13 psi, and the auxiliary gas heater temperature was 425 °C. The positive mode ion spray voltage was 3500 V, and the negative mode was − 3500 V. The ion transfer tube temperature was 325 °C, with a normalized collision energy of 20-40-60 V. The resolution for the first-level mass spectrometry was 60,000, and for the second-level, it was 7,500. Data acquisition was conducted in Data-Dependent Acquisition (DDA) mode. Cell culture HTR8/SVneo (HTR) and JEG-3 cells were used for cell experiments. They are two types of human trophoblast cell lines, generally used for researching placental development and trophoblast function. The HTR and JEG-3 cells were respectively cultured in RPMI-1640 and MEM Medium (P/S) with 10% fetal bovine serum, at a temperature of 37 °C, with 5% CO[2], and were passaged every two days. In the toxicity experiments, the fullerenols aqueous solution was diluted to a certain concentration with the culture medium for cell culture. In the toxicity tests, fullerenols were administered at concentrations of 10, 30, and 50 µg/mL. For the mechanism studies, a concentration of 50 µg/mL was utilized. Each experiment was executed with three biological replicates to ensure the reliability and reproducibility of the results. The ERα agonist (1 µM), ER antagonist (10 µM) and estriol (1 µM) were dissolved in DMSO and then diluted to the desired concentration with the culture medium for the experiments. Dilute the chemicals to appropriate concentrations in the culture medium, co-cultivate, and incubate for 24 h before proceeding with subsequent experiments. The amount of DMSO was controlled to be within 0.01% of the solution. Cell proliferation assays Cell proliferation vitality was evaluated using a CCK-8 assay. The seeding density of HTR-8 was 6 × 10^3 mL^− 1 while the density of JEG-3 was 1 × 10^4 mL^− 1. The fullerenols exposure concentrations were set at 10, 30, and 50 µg/mL, with six replicate wells established for each concentration to ensure measurement accuracy. After 24 h, fullerenols were added to the culture medium and cultured for another 24 h. Following the manufacturer’s instructions, a Cell Counting Kit-8 (CCK8) from Abbkine Scientific (China) was utilized. After a 2-hour incubation period, cell viability was measured at an absorbance of 450 nm using a microplate reader Spark (TECAN, Austria) to determine the cell proliferation rate. Cell migration assay Cell migration ability was assessed using a cell scratch assay. Fully grown cells were evenly seeded into a six-well plate, with each well containing 2 mL of cell medium supplemented with 10% fetal bovine serum and penicillin-streptomycin. After a 24-hour incubation period, a scratch was made using a 10 µL sterile pipette tip. The cells were then washed with warm PBS to remove any detached cells, and images of the scratches were immediately captured at the 0-hour mark. Then, the PBS was removed, and the chemicals were dissolved into the medium containing 2% fetal bovine serum. The medium mixture was added to the six-well plate at a volume of 2 mL/well. For another 24 hours’ incubation, the recovery of the scratch region was taken at the same location. The scratch area was quantified using ImageJ software. Estrogen testing The estrogen testing was conducted by enzyme-linked immunosorbent assay (ELISA) method. Specifically, estradiol (E2) levels were measured using the QuicKey Pro E2 (Estradiol) ELISA Kit (Elabscience, China), while estriol (E3) levels were assessed with the QuicKey Pro E3(Estriol) ELISA Kit (Elabscience, China). Placental assemblies or cell homogenates were processed according to the kit instructions, with E3 or E3 concentrations determined by measuring optical density (OD) values at 450 nm and subsequent calculations. RNA extraction and qRT-PCR Total RNA was extracted from cells and tissues using TRIzol reagent (Vazyme BioTech, China) following the manufacturer’s instructions. The RNA was quantified using the Nanodrop 2000c system (Thermo Fisher Scientific, USA). Subsequently, the RNA was reversely transcribed into cDNA using HiScript II Q RT SuperMix for qPCR (+ gDNA wiper) and amplified with ChamQ SYBR qPCR Master Mix (Vazyme BioTech, China) in a 10 µL reaction volume. Real-time fluorescence quantitative PCR was performed using the 7500 fast real-time PCR System (Applied Biosystems, USA). Data were normalized using GAPDH as the internal reference. The primer sequences are provided in Supplementary Material, Table [78]S1. Western blotting The tissues and cells were lysed with RIPA lysis buffer (Beyotime, China) that included protease and phosphatase inhibitors (Beyotime, China). Protein concentration was determined using a BCA reagent kit (Beyotime, China). Proteins were denatured by adding SDS and PBS, then heating at 100 °C for 5 min. Proteins were separated with 12% SDS-PAGE gels (Vazyme BioTech, China) and transferred to a polyvinylidene fluoride (PVDF) membrane (Merck Millipore, MA, USA). QuickBlock Western Blocking Solution was applied for blocking for 20 min. The membranes were then incubated with the ERα (Abcam, Cambridge, HK), CYP3A4 (Abcam, Cambridge, HK) and GAPDH (Proteintech, wuhan, China) overnight at 4 °C, followed by washing with TBST 10 min for 3 times. The dilution concentration of GAPDH is 1/3000, while the dilution concentration of other antibodies is 1/1000. Subsequently, the membranes were incubated with the secondary antibody for 1 h and washed with TBST 10 min for 3 times. Finally, the PVDF membranes were imaged using a Bio-Rad ChemiDocXRS + system (Bio-Rad, USA). GAPDH was used as the loading control. The grayscale values of protein bands in the images were calculated using ImageJ software. Immunofluorescence (IF) staining Treated cells were rinsed with PBS, then immobilized in 4% paraformaldehyde (Servicebio, China), and permeabilized with a Powerful Permeabilization Solution for Immunostaining (Beyotime, China). After washing, cells were blocked with Immunostaining Blocking Solution (Beyotime, China). Subsequently, the cells were incubated with the primary ERα antibody (Abcam, Cambridge, UK) and secondary antibody (Abbkine, Wuhan, China). After that, nucleus were stained with PhyLight™ Enhanced Antifade Mounting Medium with DAPI (Phygene, China) for 5 min. Fluorescent images were captured using a laser scanning confocal microscope (Leica STELLARIS 5, Germany). Detection of reactive oxygen species and oxidative stress levels The detection of reactive oxygen species (ROS) was performed using the ROS assay kit (cat. S0033S) from Beyotime (China). The DCHF-DA probe, currently the most commonly used method for ROS detection, was utilized. Cells were seeded in a 96-well plate at a density of 1 × 10^4 per well and treated with or without fullerenols and/or corresponding pharmaceutical products after 24 h. After 24 h, DCFH-DA probe was diluted at a ratio of 1 : 5000 and incubated for half an hour. Firstly, tissues and cells were ultrasonically homogenized, and then the DCHF-DA probe was added. After following the instructions, fluorescence was measured using a microplate reader (excitation wavelength 488 nm, emission wavelength 525 nm). Cells extracted from tissues were also incubated for half an hour and then tested. Oxidative stress levels were primarily assessed by measuring catalase (CAT), superoxide dismutase (SOD), malondialdehyde (MDA), and glutathione (GSH). The Nanjing Jiancheng Bioengineering Institute kits used included the catalase (CAT) assay kit (visible light method), total superoxide dismutase (T-SOD) assay kit (hydroxylamine method), malondialdehyde (MDA) assay kit (TBA method), and reduced glutathione (GSH) assay kit (spectrophotometric method). Synthesis of reduced glutathione-modified Fullerenols Composite nanoparticles Fullerene C[60] is dissolved in methanol and ultrasonicated for 5–10 min to achieve a uniform dispersion, after which it is stored in the dark. Glutathione and NaOH are mixed in deionized water, followed by the addition of ethanol and stirring until uniformity is achieved. Under nitrogen protection, the mixture is reacted by stirring for 24–48 h. The liquid is evaporated until the solid is obtained, with toluene, ethanol, and water being removed. The obtained solid is dissolved in ultrapure water, then centrifuged to collect the supernatant at a speed of 500–1000 rpm. The supernatant is filtered through a 220 nm aqueous filter membrane, and the filtrate is dialyzed for purification. Finally, the filtrate is freeze-dried to yield the solid product, which consists of reduced glutathione-modified fullerene composite nanoparticles. Statistical analysis Data from the experiment were analyzed using GraphPad Prism 8.0. All values are expressed as mean ± standard error of mean. Homogeneity of variance was assessed prior to performing t-tests for comparisons. For data that did not meet the criteria for parametric tests, the Mann-Whitney U test was conducted. A p-value of less than 0.05 was considered to indicate a statistically significant difference. Mechanism figure is drawn by Figdraw. Results Characterization analysis of fullerene Transmission electron microscope analysis showed that the original particle size of fullerenols was approximately 100 nm (Supplementary Figure [79]S1A). The particles were characterized by their irregular square shape and exhibited good dispersion. The hydrodynamic diameter of fullerenols was measured to be 247.3 ± 1.62 nm, notably larger than the original particle size, suggesting a slight aggregation of fullerenols in water (Supplementary Figure [80]S1B). The zeta potential of the fullerenols was − 18.3 ± 4.5 mV (Supplementary Figure [81]S1C). In addition, the Polymer Dispersion Index (PDI) of PS-NPs was found to be less than 0.5, indicating a good dispersion. These findings confirmed that the fullerenols particles used in our study possessed typical nanoparticle characteristics and met the experimental requirements. Effects of prenatal fullerenols exposure on placental and fetal development in mice Dams exposed to different concentrations of fullerenols water solution maintained normal conditions with no occurrences of miscarriage or discomfort and no significant changes in water consumption, diet or body weight fluctuation (Supplementary Figures S2A-C). Notably, exposure to 10 mg/kg/d fullerenols during pregnancy led to a decrease in the offspring’s placental weight, but with no discernible differences in the diameter of the placenta, offspring’s weight of offspring (Fig. [82]1A-D). At the same time, no significant effects were observed in the low and medium dose groups. In addition, only one stillbirth occurred in each group exposed to medium and low doses of fullerenols, while the 10 mg/kg/d group had a higher number of stillbirths (no statistical significance) during pregnancy (Fig. [83]1E). HE staining results indicated that prenatal fullerenols exposure caused thinning of the placental villous layer, cell shrinkage, serious cell vacuolization, decreased tight junctions between cells and a slight increase in some local intercellular gaps (Fig. [84]1F). All these findings suggested that exposure to 10 mg/g/d of fullerenols during pregnancy might potentially affect the development of the placenta. However, no insights into how fullerenols affects mouse embryo development are available to date, which warrants further exploration. Fig. 1. [85]Fig. 1 [86]Open in a new tab Effects of prenatal fullerenols exposure on placental and fetal development in mice. (A) Placental weight, (B) placental diameter, (C) embryo weight, and (D) stillbirth rate in GD18 after fullerenols pregnancy exposure. (E) Representative pictures of the whole uterus and fetal placentas of control as well as fullerenols group (10 mg/kg/d) mice. Arrowheads indicate stillbirths seen in the fullerenols group. (F) HE staining of mice placentas was performed to observe the pathology of the lateral section of the placenta and localisation of the trophoblast. Data are expressed as mean ± SEM; * indicates difference from control, * indicates p < 0.05, n = 6, where n represents the number of biological replicates Impact of Fullerenols exposure on metabolic profiles of mice Placenta Given the involvement of metabolic pathways in crucial stages of embryonic development, we conducted metabolomic analysis on the placentas of mice exposed to fullerenols at 10 mg/g/d during pregnancy. The orthogonal partial least squares-discriminant analysis (OPLS-DA, VIP > 1, and P < 0.05) displayed clearly distinguished distribution of metabolites among the control group and fullerenols group with no intersections or overlaps (R^2X = 0.703, R^2Y = 0.980, Q^2 = 0.973) (Fig. [87]2A). The results obtained from PLS-DA and PCA correspondingly demonstrate that the two groups exhibit distinct and non-overlapping clusters (Fig. [88]2B, C), hinting at the possibility that fullerenols exposure during gestation may affect embryonic development by influencing the metabolic state of the mouse placenta. After being annotated by the Human Metabolome Database (HMDB) and statistical analysis by chemical classification, totally 29 metabolites that were significantly different between the two groups were identified and screened out (Fig. [89]2D, Table S2). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was then performed to further understand the potential functions of the above differential metabolites (Fig. [90]2E, Supplementary Figure S3A). Among them, pathways such as Estrogen receptor agonists/antagonists, Ovarian steroidogenesis showed statistical differences. We also mapped the network of differential metabolites to explore the associations between them (Supplementary Figure S3B). Among these differential metabolites, estradiol (E2) and estriol (E3) caught our attention. We found that fullerenols exposure leads to an increase in E2 and a decrease in E3, while estrone (E1) was unaffected (Fig. [91]2F). In the body, E3 is usually derived from E2 [[92]35], suggesting that fullerenols exposure impacts the pathway by which E2 is converted into E3 (Fig. [93]2F). In addition, fullerenols exposure also led to a decrease in glutathione (GSH) levels, suggesting that exposure to fullerenols may exacerbate placental oxidative stress. Fig. 2. [94]Fig. 2 [95]Open in a new tab Impact of fullerenols exposure on metabolic profiles of mice placenta. (A) OPLS-DA analysis of control and fullerenols-exposed groups. (B) PLS-DA analysis of control and fullerenols-exposed groups. (C) PCA analysis of control and fullerenols-exposed groups. (D)Heatmap of 29 differential metabolites in control and fullerenols-exposed groups. (E) KEGG pathway enrichment analysis of differential metabolites. (F) Histograms of relative expression of metabolites of interest (estrone(E1), estradiol(E2), estriol(E3), and glutathione (GSH)). Data are expressed as mean ± SEM; * indicates difference from control, * indicates p < 0.05 and ** p < 0.01, n = 6, where n represents the number of biological replicates Fullerenols exposure affects CYP3A4 and estrogen receptor expression in mice Placenta To confirm the results of metabolic analysis, ELISA methods were applied here to measure the levels of E2 and E3. Consistent with expectations, our results also showed that fullerenols exposure during pregnancy remarkably increased placental E2 and reduced in E3 (Fig. [96]3A, B). It is known that the cytochrome P450 enzyme CYP3A4 can regulate E2 by hydroxylating it at the 16α position to synthesize E3 [[97]36]. Given the dysregulation of placental E3 observed above, we therefore proceeded to examine the expression of placental CYP3A4 caused by fullerenols exposure. It turned out that fullerenols exposure leads to a decreased trend of CYP3A4 mRNA (Fig. [98]3C). More importantly, the physiological effects of estrogen are primarily mediated through estrogen receptors (ERs), including ERα and ERβ, which then transcriptionally and translationally regulate the expression of downstream target genes [[99]37, [100]38]. Notably, our data indicated that fullerenols exposure contributed to a marked reduction in Esr1 (encoding Erα protein) mRNA expression (Fig. [101]3D), while had no effect on Esr2 (encoding Erβ protein) mRNA expression (Supplementary Figure S3C). Western blot analysis further confirmed the significant reduction in placental CYP3A4 and ERα protein expression from mice exposed to fullerenols (Fig. [102]3E, F, G). Our data suggested that fullerenols might inhibit the conversion of E2 to E3 by downregulating CYP3A4 expression, which further reduces ERα expression. Fig. 3. [103]Fig. 3 [104]Open in a new tab Fullerenols exposure affects CYP3A4 and estrogen receptor expression in mice placenta. Effects on mice placental (A) E2, (B) E3, (C) Cyp3a4 mRNA, (D) Esr1 mRNA, and (E) ERα and CYP3A4 protein levels, after exposure to fullerenols during pregnancy. Quantitative analysis of ERα (F) and CYP3A4 (G) WB protein levels. Effects of fullerenols exposure during pregnancy on placental oxidative stress (H) GSH, (I) T-SOD, (J) ROS and (K) MDA. Data are expressed as mean ± SEM; * indicates differences from control, * indicates p < 0.05 and ** p < 0.01, n = 6, where n represents the number of biological replicates Among the differentially expressed metabolites, GSH, a well-studied endogenous antioxidant, was found to be at a low level in the fullerenols group. We then assessed the level of oxidative stress in the placenta and found that fullerenols exposure leads to an increase in placental ROS and MDA, along with a decrease in T-SOD and GSH (Fig. [105]3H-K). But CAT showed no significant difference (Supplementary Figure S3D). Therefore, we speculated that the enhanced oxidative stress induced by fullerenols may be related to the reduction in ERα expression caused by decreased E3 levels. Effects of Fullerenols exposure on trophoblast cells Based on the fullerenols exposure-caused adverse outcomes in placenta and embryos observed above, HTR-8/SVneo and JEG-3 cells, two cell lines routinely utilized in reproductive toxicity studies, were enrolled to further clarify the effect of fullerenols in vitro. Firstly, the results of CCK-8 assay indicated that fullerenols at a concentration of 50 µg/mL significantly inhibits the viability of both cell lines, while 10 and 30 µg/mL of fullerenols exposure show no effect on the cell viability (Fig. [106]4A). The 24-hour scratch assay validated that fullerenols exposure significantly inhibits the migration of HTR-8/SVneo (Fig. [107]4B) and JEG-3 cells (Supplementary Figure S4A) in a dose-dependent manner. The quantitative measurement based on ELISA methods revealed that 30 and 50 µg/mL of fullerenols exposure effectively enhanced the level of E2 (218.3 pg/mL, P < 0.05, 236.3 pg/mL, P < 0.01) and decreased E3 (33.7 pg/mL, P < 0.05, 23.5 pg/mL, P < 0.01) in HTR-8/SVneo cells compared with those (E2: 149.0 pg/mL, E3: 61.5 pg/mL) in cells without fullerenols exposure (Fig. [108]4C, D). Such alteration of E2 and E3 level was also observed in JEG3 cells (Supplementary Figure S4B, C). Beyond that, the mRNA and protein levels of CYP3A4 and ERα in HTR-8/SVneo cells after fullerenols exposure were also measured. Our results showed that exposure to fullerenols at 10, 30 and 50 µg/mL dramatically attenuated CYP3A4 mRNA and protein expression in HTR-8/SVneo cells, and the levels of ERα mRNA and protein were also lowered in cells exposed to fullerenols at 50 µg/mL. (Fig. [109]4E-I). The decrease in CYP3A4 and ERα mRNA and protein expression also occurred in JEG3 cells (Supplementary Figure S4D-H). The Immunofluorescence experiment also showed that fullerenols reduced the relative fluorescence intensity of ERα protein in both cells (Fig. [110]4J, K, Supplementary Figure S4I, J). The oxidative stress results demonstrated that exposure to 10, 30 and 50 µg/mL of fullerenols significantly decreased both GSH and total SOD (T-SOD) concentration and elevated ROS level, and MDA concentration was markedly increased only at 30 µg/mL of fullerenols exposure (Fig. [111]4L-O). A similar pattern of changes in these oxidative stress indicators was shared in fullerenols-exposed JEG3 cells (Supplementary Figure S4K-N). The results from CAT showed no consistent trends in expression levels between the two cells, suggesting that CAT may not be altered through the mechanism by which fullerenols affects ERα (Fig. [112]4P, Supplementary Figure S4O). The above results indicated that the trends in change of estrogen, CYP3A4, ERα and oxidative stress indicators in our cell model exposed to fullerenols were consistent with those detected in our in vivo model. Besides, given the results described above, the dose of 50 µg/mL was used for subsequent functional experiment. Fig. 4. [113]Fig. 4 [114]Open in a new tab Effects of fullerenols exposure on trophoblast cells in vitro. (A) Cell viability was examined by Cell Counting Kit-8 assay in HTR-8/SVneo and JEG-3 cells after 24 h of fullerenols exposure with different concentrations. (B) Relative migratory area of HTR-8/SVneo cells exposed to different concentrations of fullerenols, magnification: 40×, the relative migratory area (%) of the cells at 24 h is shown on the right side of the graph, Scale bar: 500 μm. The levels of E2 (C) and E3 (D) in HTR-8/SVneo cells after fullerenols exposure. Relative mRNA expression levels of CYP3A4 (E) and ESR1 (F) were detected in cells by qRT-PCR. The corresponding CYP3A4 and ERα protein expression was also detected (G), and the relative expression of the proteins is shown on the right of the figure (H), (I). (J),(K) the immunofluorescence images of ERα and relative fluorescence intensity analysis, Scale bar: 50 μm. Changes in oxidative stress levels due to fullerenols exposure in HTR-8/SVneo: (L) GSH, (M) T-SOD, (N) ROS, (O) MDA and (P) CAT. Data are expressed as mean ± SEM; * indicates differences from control, * indicates p < 0.05 and ** p < 0.01, n = 3, where n represents the number of biological replicates The mechanisms underlying Fullerenols-Induced oxidative stress and cytotoxicity in trophoblast cells Our studies revealed that fullerenols exposure inhibits ERα expression, prompting us to explore whether this inhibition leads to oxidative stress and cytotoxicity. To address this, propyl pyrazole triol (PPT), a selective ERα agonist, and the ICI 182780, a potent ER antagonist was obtained to intervene the transportation of E3 signals in cells exposed to fullerenols. The results of scratch assay showed that PPT effectively alleviated the migration ability of cells inhibited by fullerenols, and the inhibition of cell migration by ICI 182780 was similar with fullerenols exposure (Fig. [115]5A, Supplementary Figure S5A). This indicates that fullerenols may exacerbate cytotoxicity by inhibiting ERα expression. Fig. 5. [116]Fig. 5 [117]Open in a new tab Role of ERα in fullerenols cytotoxicity in HTR-8/SVneo cells. There were four groups: control, fullerenols group (50 μg/ml), fullerenols + ERα agonist PPT group, and ERα antagonist ICI 182780 group. (A) Relative migratory area (%) of cells in different treatment groups, magnification: 40×, Scale bar: 500 μm. Relative mRNA expression levels of (B) ESR1 and (C) ESR2 were detected in cells by RT-PCR, (D) protein levels of ERα were examined, and (E) relative expression analysis was performed. The immunofluorescence images of ERα and the analysis of relative fluorescence intensity are shown in (F), (G), Scale bar: 50 μm. Changes in oxidative stress levels in different treatment groups: (H) GSH, (I) T-SOD, (J) ROS, (K) MDA and (L) CAT. Data are expressed as mean ± SEM; * indicates p < 0.05, ** p < 0.01 and *** p < 0.001, n = 3, where n represents the number of biological replicates ESR1 expression was reduced in both the fullerenols group compared to the control group and in the ICI-182780 group compared to the control group, whereas PPT ameliorated the reduction in ESR1 levels caused by fullerenols exposure (Fig. [118]5B, Supplementary Figure S5B). Both PPT and ICI 182780 had no effect on cellular ESR2 expression (Fig. [119]5C, Supplementary Figure S5C). WB analysis also verified that ICI 182780 inhibited ERα expression and that PPT ameliorated the reduction of ERα caused by fullerenols exposure (Fig. [120]5D, E, Supplementary Figure S5D, E). Similarly, the relative fluorescence intensity of ERα is consistent with the results from WB (Fig. [121]5F, G, Supplementary Figure S5F, G). Additionally, the fullerenols + PPT group exhibited increased GSH and T-SOD levels, and decreased ROS levels, compared to the fullerenols and ICI 182780 groups (Fig. [122]5H-K, Supplementary Figure S5H-K). However, CAT does not follow this trend (Fig. [123]5L, Supplementary Figure S5L). This indicates that fullerenols may exacerbate oxidative stress by inhibiting ERα expression except for CAT. Next, we explored the role of CYP3A4 and E3 in cells with fullerenols exposure. Overexpressing CYP3A4 in fullerenols-treated cells increased E3 levels in cell homogenates compared to fullerenols alone, while direct inhibition of CYP3A4 expression decreased E3 levels (Fig. [124]6A, Supplementary Figure S6A). This suggests that fullerenols primarily reduces E3 levels by inhibiting CYP3A4 expression. Subsequently, adding E3 during fullerenols exposure effectively mitigated the cell migration inhibition (Fig. [125]6B, Supplementary Figure S6B) and restored ERα expression (Fig. [126]6C-G, Supplementary Figure S6C-G), along with reducing oxidative stress levels (Fig. [127]6H-K, Supplementary Figure S6H-K), with the exception of CAT (Fig. [128]6L, Supplementary Figure S6L). However, adding E3 with ICI 182780 to fullerenols-treated cells still resulted in inhibited cell migration, decreased ERα expression, and exacerbated oxidative stress. Fig. 6. [129]Fig. 6 [130]Open in a new tab Role of CYP3A4 and E3 in fullerenols cytotoxicity in HTR-8/SVneo cells. (A) Effects of CYP3A4 overexpression and inhibition on cellular E3 levels. A total of four groups were next divided to explore the effects of E3: control group, fullerenols group (50 μg/ml), fullerenols+E3 group, and fullerenols + E3 + ICI 182780 group. (B) Relative migratory area (%) of cells in different treatment groups, magnification: 40×, Scale bar: 500 μm. (C) The mRNA level of ESR1. (D)The protein level of ERα and (E) the relative expression level of the protein. (F) Immunofluorescence images of ERα and (G) the relative fluorescence intensity of ERα, Scale bar: 50 μm. Changes in oxidative stress levels in different treatment groups: (H) GSH, (I) T-SOD, (J) ROS, (K) MDA and (L) CAT. Data are expressed as mean ± SEM; * indicates p < 0.05 and ** p < 0.01, n = 3, where n represents the number of biological replicates Characterization and analysis of GSH-Modified Fullerenols Our above results have demonstrated that fullerenols exposure reduced E3 levels, leading to decreased ERα expression, which further caused elevated oxidative stress indicators such as increased ROS accumulation and decreased glutathione levels. The reduction in intracellular glutathione levels decreases antioxidant capacity, increasing the risk of oxidative damage to cells and tissues, potentially leading to various diseases [[131]39]. Given the application prospects of fullerenols, it is necessary to design a simple preparation process for biocompatible reduced glutathione-modified fullerenols composite nanoparticles. We provide a preparation method for reduced glutathione-modified fullerenols composite nanoparticles, where glutathione is conjugated to the fullerene carbon cage through a nucleophilic addition reaction, resulting in a simple and feasible preparation process. The prepared reduced glutathione-modified fullerenols composite nanoparticles were characterized. Transmission electron microscopy revealed slight aggregation in the aqueous solution, with good dispersibility (Fig. [132]7A). Infrared spectroscopy indicated successful modification of the fullerene carbon cage with reduced glutathione and multiple hydroxyl groups (Fig. [133]7B). UV-Vis absorption spectroscopy showed that the conjugated structure of fullerenes was retained, with the aqueous solution of fullerenes exhibiting a yellow-brown color and a significant absorption peak around 250 nm in the UV region, confirming the intact conjugated structure of the reduced glutathione-modified fullerenols C[60] (Fig. [134]7C). The hydrated particle size of reduced glutathione-modified fullerenols C[60] in water was around 125 nm with a PDI of 0.147, indicating small aggregate size suitable for cellular uptake (Fig. [135]7D). The Zeta potential was approximately − 21 mV, indicating strong electrostatic repulsion between particles, allowing stable existence in aqueous solutions. Fig. 7. [136]Fig. 7 [137]Open in a new tab Characterization and effects of GSH-modified fullerenols in vitro. (A) Transmission electron micrograph of C[60]-(OH)[n]-GSH. (B) Infrared spectrogram of C[60]-(OH)[n]-GSH solid. (C) UV-Vis absorption spectrum of C[60]-(OH)[n]-GSH aqueous solution. (D) Hydrated particle size of C[60]-(OH)[n]-GSH. (E) Relative migratory area (%) of cells in different treatment groups, magnification: 40×, Scale bar: 500 μm. The viability of HTR-8/SVneo cells at (F) 24 h and (G)48 h of intervention was assayed by Cell Counting Kit-8 assay, and the study groups included control, C[60]-(OH)[n], C[60]-(OH)[n]-GSH, H[2]O[2], H[2]O[2]+C[60]-(OH)[n], and H[2]O[2]+C[60]-(OH)[n]-GSH. C[60]-(OH)[n] and C[60]-(OH)[n]-GSH concentrations were 50 μg/mL, and H[2]O[2] concentration was 1 μg/ml. (H) GSH, (I) ROS, (J) T-SOD, (K) MDA, (L) CAT and (M) ESR1 mRNA were assayed in HTR-8/SVneo cells from different treatment groups. Data are expressed as mean ± SEM; * indicates p < 0.05, n = 3, where n represents the number of biological replicates Effects of GSH-Modified fullerenols in Vitro We further investigated the antioxidant capacity of reduced glutathione-modified fullerenols C[60] (C[60] (OH)[n]-GSH) on cells. The study groups included control, C[60]-OH, C[60](OH)[n] -GSH, H[2]O[2], H[2]O[2] + C[60](OH)[n], and H[2]O[2] + C[60](OH)[n]-GSH. The concentrations of C[60](OH)[n] and C[60](OH)[n]-GSH were 50 µg/ml, and H[2]O[2] was 1 ng/ml. The cell scratch test, CCK-8 results at 24 and 48 h showed that C[60](OH)[n]-GSH had a certain reparative effect on H[2]O[2]-induced cell migration and proliferation inhibition, while C[60](OH)[n] did not (Fig. [138]7E-G). ROS results indicated that C[60](OH)[n]-GSH inhibited H[2]O[2]-induced ROS elevation to some extent, whereas C[60](OH)[n] did not (Fig. [139]7H). GSH content analysis revealed that C[60](OH)[n]-GSH increased intracellular GSH levels and partially mitigated H[2]O[2]-induced GSH reduction, whereas C[60](OH)[n] had no such effect (Fig. [140]7I). In addition, C[60](OH)[n]-GSH has a certain reparative effect on the reduction of T-SOD and the increase of MDA caused by fullerenols, but it has no significant effect on CAT. C[60](OH)[n]-GSH also has no effect on the mRNA expression of ESR1. These results suggest that cellular uptake of C[60](OH)[n]-GSH provides some resistance to oxidative stress. Discussion Despite the wide range of applications of fullerenols in medicine and materials science, their safety is still somewhat controversial [[141]40]. Currently, toxicity studies on fullerenols are mainly confined to in vitro experiments, with in vivo studies often involving acute poisoning and short observation periods, lacking mechanism exploration [[142]17]. For instance, fullerenols has been found to affect CHO, HaCaT, Hela, and HEK293 cell lines, showing some degree of cytotoxicity and phototoxicity, base toxicity, and carcinogenic effects [[143]41]. Xu et al. reported that intratracheal instillation of 5 or 10 mg of C60(OH)n in rats resulted in cell damage, oxidative/nitrosative stress, and inflammation, as assessed by bronchoalveolar lavage fluid biomarkers and pathological evaluation of lung tissue after 3 days of exposure [[144]42]. Roursgaard et al. exposed mice to fullerenols ranging from 0.02 to 200 µg/mouse by intratracheal instillations over 24 h [[145]43]. They found that at 200 µg/mouse, fullerenols demonstrated a pro-inflammatory response. In particular, the SCCS opinion indicates that the potential genotoxicity and carcinogenicity of fullerenols cannot be ruled out [[146]14]. Therefore, it is necessary to evaluate the toxicity of fullerenols. Specifically, there is a lack of research on the effects of prenatal fullerenols exposure on embryonic development. Our study indicates that prenatal exposure to 10 mg/kg/day of fullerenols affects placental development, resulting in thinning of the trophoblast layer and severe cellular vacuolization. This suggests that fullerenols exhibits embryonic developmental toxicity. The integration of in vitro and in vivo experiments necessitates a more cautious extrapolation of conclusions, particularly when generalizing to human populations. The experimental conditions and the specific models used may not fully replicate the complexity of human physiology, thus caution must be exercised in applying these results to clinical scenarios. Metabolomics can help study the impact of environmental exposure on embryonic metabolism [[147]44]. By analyzing changes in metabolites, it can assess the potential effects of environmental factors such as chemicals on embryonic development. By exploring metabolic pathways and interaction networks, metabolomics aids in uncovering critical biological mechanisms in development, including gene expression regulation and cellular signal transduction, thereby enhancing our understanding of the mechanisms underlying developmental diseases [[148]45]. Presently, research on the effects of fullerenols exposure on organism metabolism is limited to plants [[149]46] and bacteria [[150]47], our study, however, comprehensively reveals through metabolomics the impact of prenatal fullerenols exposure on embryonic development. Our metabolomic analysis has primarily focused on the estrogen receptor pathway. While this has provided valuable insights, it is acknowledged that other pathways may also be influenced by the treatments and conditions studied. Future research should delve deeper into these additional pathways to gain a more comprehensive understanding of the broader metabolic effects. Through metabolomics, we found that exposure to fullerenols led to an increase in E2 and a decrease in E3, while E2 can be regulated by CYP3A4 to transform into E3 [[151]36]. Fullerenols may be interfering with lipid synthesis or sterol synthesis at a more fundamental level [[152]48], which could potentially explain the observed effects on E3 levels and ERα expression. Further exploration is required to uncover the detailed molecular interactions and signaling pathways involved in this process. We confirmed in cell experiments that CYP3A4 regulates E3 and discovered that supplementing E3 can inhibit the oxidative stress caused by fullerenols exposure. Therefore, we consider the reduced E3 to be a key metabolite leading to abnormal embryonic development. Commonly, E3 are at very low levels in non-pregnant women, and its estrogenic potency is considered much weaker compared to E2 [[153]49]. However, during pregnancy, E3 level increases about 1,000-fold, which accounts for over 90% of the total circulating estrogen under normal physiological conditions [[154]50, [155]51]. In nature, rodents still reproduce successfully without dramatic increases in E3 during pregnancy, while they may experience a high incidence of embryo resorption, premature delivery, post-term labor, and stillbirth [[156]52]. Interestingly, fetal exogenous E3 treatment significantly improved the chance of pregnancy, reduced the incidence of resorption, and eliminated premature/post-term delivery and stillbirth in prenatally exposed female offspring [[157]52]. Abnormally low levels of unconjugated or free E3 in a pregnant woman may signal chromosomal or congenital anomalies [[158]53]. Additionally, E3 may alter the expression levels and DNA methylation patterns of numerous genes in the uterus and placenta [[159]54, [160]55]. Mechanistically, E3 affects the interaction between estrogen receptors (ER) and various DNA/histone modifiers, as well as the binding of these complexes to target genes [[161]56]. E2 and E3 are both agonists of estrogen receptors ERα and ERβ [[162]57]. Although E2 has a stronger binding affinity than E3, unlike E2, E3 can accumulate in significant amounts in target organs to activate ER receptors [[163]58, [164]59]. Additionally, E2 has a plasma protein binding rate of approximately 98%, while E3 has a rate of about 92% [[165]35, [166]60, [167]61]. As a result, a higher proportion of E3 remains unbound in circulation and is thus able to exert its biological activity. This could explain why E3 exhibits higher-than-expected biological activity compared to E2. Through experiments, we confirmed that fullerenols exposure reduces placental accumulation of E3, diminishing ERα activation, while supplementing E3 can restore ERα expression. Additionally, despite an increase in E2 due to fullerenols exposure, its predominantly bound form results in weaker regulation of ER in the placenta. Many studies have shown that estrogen and estrogen receptors are key regulators of oxidative stress [[168]62]. The reduced expression of ERα may further exacerbates the level of oxidative stress. For example, estrogen can restore the expression levels of antioxidant enzymes GPX1 and GPX4, increase the expression of the rate-limiting enzyme γ-glutamylcysteine synthetase for glutathione synthesis, which maintains the stability of GSH and balances oxidative stress [[169]63, [170]64]. Estrogen receptors can also reduce the activity of MAPK, inhibit apoptosis, and decrease the production of ROS [[171]65]. ERα activates eNOS through the PI3/AKT signaling pathway, leading to an increased NO production and reduced oxidative stress levels [[172]66, [173]67]. These studies indicate that downregulation of estrogen receptors results in heightened oxidative stress, which is consistent with our research findings. Oxidative stress is a primary mechanism of toxicity induced by nanomaterials, including fullerenols [[174]68, [175]69]. The toxicity of fullerenols derivatives may be related to the permeability transition (MPT) of mitochondria, leading to mitochondrial membrane depolarization, reduction in glutathione (GSH), protein thiols, and malondialdehyde (MDA) production, and resulting in oxidative damage to mitochondrial lipids, leading to impaired respiratory function. Moreover, fullerenols may induce cell apoptosis or necrosis, including increased cellular ROS, inhibition of MMP, and release of cytochrome c. Fullerene derivatives activate apoptosis-regulating kinase 1 (ASK1) and c-Jun N-terminal kinase (JNK) signaling pathways, activating oxidative stress and influencing the proliferation of hepatocellular carcinoma cells, further leading to cellular JAK2 V617F mutation. Our research has also identified that fullerenols exposure exacerbates oxidative stress in mouse placenta and trophoblast cells. However, we attribute this primarily to the reduction of E3 inhibiting ERα, a deduction supported by our experimental findings. We also observed that fullerenols had no effect on CAT, which may be because CAT is not regulated by the estrogen receptor pathway. Additionally, our metabolomics analysis of fullerenols exposure identified glutathione (GSH), a biomarker of oxidative stress, and experiments confirmed that fullerenols exposure leads to decreased glutathione levels in trophoblast cells. GSH is one of the primary intracellular antioxidants, capable of neutralizing free radicals and peroxides, thus protecting the placenta and trophoblast cells from oxidative stress damage, which is crucial for maintaining cell health and normal function [[176]70]. GSH is also an essential component of sulfur metabolism, playing a role in the activation of various enzymes and protein synthesis [[177]71, [178]72]. Adequate GSH levels are necessary for the metabolic activities of the placenta and trophoblast cells. Moreover, GSH is one of the main cellular reducing agents, maintaining the appropriate intracellular reducing environment, which helps preserve normal cellular function and structure [[179]73]. Given GSH’s critical physiological roles, to reduce the embryotoxicity of fullerenols, we synthesized reduced glutathione-modified fullerenols. The results demonstrated that C[60](OH)[n]-GSH can increase intracellular GSH levels and has certain reparative effects against H[2]O[2]-induced GSH depletion and inhibition of cell proliferation. C[60](OH)[n]-GSH significantly reduces fullerenols’ embryotoxicity, enhances biocompatibility, and boosts antioxidant and cell protective functions. Besides, glutathione modification enhances the water solubility of nanomaterials by introducing strong polar groups such as hydroxyl and thiol groups, making them more stable in aqueous phases and reducing the likelihood of aggregation [[180]74]. The modification with glutathione also protects nanomaterials from attacks by environmental oxidants or other chemicals, thereby enhancing their stability [[181]75, [182]76]. Moreover, glutathione itself is an endogenous strong antioxidant, and its combination with fullerenols can strengthen the antioxidant capacity through covalent or non-covalent interactions [[183]75]. Additionally, studies have shown that glutathione-modified nanoparticles exhibit good stability under neutral and weakly acidic conditions, with the modification reducing the degradation rate of nanoparticles under light exposure [[184]77]. All of the above fully demonstrate that glutathione-modified fullerenols possess good stability. Conclusion In summary, this study demonstrates that fullerenols exposure during pregnancy leads to a reduction in placental weight and thinning of the placental villi layer. Mechanistically, fullerenols exposure inhibits E3 synthesis in placenta by decreasing CYP3A4 level, and then reduces estrogen receptor activation and leads to decreased glutathione and exacerbated oxidative stress. Furthermore, we synthesized C[60](OH)[n]-GSH, which significantly increased intracellular GSH, repaired H[2]O[2]-induced damage, and showed potential in reducing fullerenols’ toxicity and enhancing biocompatibility and antioxidant functions. Our study reveals the potential toxicity of fullerenols during the embryonic development stage, providing important references