Introduction

Perfluorinated compounds, a class of organic pollutants characterized by the complete substitution of hydrogen atoms in carbon chains and alkyl groups with fluorine atoms, encompass substances such as perfluoroalkylamines, perfluoroethers, and perfluoroalkanoic acids. These compounds can enter the environment through multiple pathways and are notably resistant to degradation. Recent research has demonstrated a significant correlation between the levels of exposure to these compounds in the human body and the incidence of various diseases1. Due to their high-energy bonds, perfluoroalkanoic acids exhibit excellent physicochemical stability and biological stability, making them of significant industrial value. They are widely used in industries such as textiles, food packaging, cosmetics, coatings, electronics products, and home furnishings2. Among these compounds, Perfluorooctane sulfonic acid (PFOS) consists of fully fluorinated anions in perfluorinated acid sulfate groups. PFOS has half-life of 3.4 years in human serum due to its robust carbon-fluorine bond that imparts thermal-chemical stability; thus it remains uneliminated in both external environments and within the human body3.

As a result of the widespread application of PFOS in industrial production, it permeates the atmosphere, water, soil, and other environmental media through multiple pathways4. At present, human epidemiological studies and experimental animal studies have reported that PFOS can accumulate in liver tissue, kidney tissue, bone, etc. PFOS can affect the functions of liver, thyroid, immune system, nervous system and reproductive system5,6,7,8,9,10. Serum data from 1303 participants recruited in an epidemiological study in China showed that the median concentration of PFOS was 14.85 ng/mL. This figure far exceeds the value of 5 ng PFOS/mL plasma established by the German Committee on Human Biomonitor11,12.

A growing body of research has established a link between perfluorooctane sulfonate (PFOS) exposure and diseases of the female reproductive system. A clinical study conducted on Chinese women demonstrated a negative correlation between serum PFOS concentrations and extended menstrual cycle length, as well as menorrhagia13. In addition, other studies have suggested that PFOS exposure is associated with primary ovarian insufficiency and preterm birth14,15. In female infertility, endometrial function and receptivity status are key factors affecting pregnancy success. Decidualization is a morphological and functional change experienced by endometrial stromal cells to support endometrial receptivity16. At present, it has been reported that PFOS inhibits decidualization of stromal cells, and PFOS disrupts the regeneration of cortisol in decidual tissue by inhibiting the reduction of key proinflammatory cytokines in maternal-fetal immune intolerance, thus impairing decidualization and immune tolerance environment in early pregnancy17. In addition, PFOS exposure can reduce the activity of protein kinases PKA, ROCK and Akt/PKB and prolactin secretion in human endometrial stromal cells, thereby affecting the process of endometrial decidualization18.

Few studies have evaluated the effect of PFOS on endometrial stromal cells in vitro, despite reports of its effects on the reproductive system. The process of PFOS interference with endometrial decidualization, which forms the substrate for embryo implantation, must be further investigated. Therefore, To enhance our understanding of the mechanisms of endometrial regulation after PFOS exposure, our study aimed to assess whether any associations exist between the following: (1) PFOS exposure and endometrial receptivity indexes; (2) As a landmark structure of endometrial implantation window, pinopodes are closely related to the process of embryo implantation. We evaluated the correlation between PFOS concentration and pinopodes morphology.

Materials and methods

Chemicals and reagents

PFOS was obtained from Dr. Ehrenstorfer (Augsburg, Germany) and dissolved in dimethyl sulfoxide (DMSO); The culture medium DMEM was obtained from Biosharp (China); fetal bovine serum was obtained from Animal Blood Ware (China); 0.25% trypsin ethylenediaminetetraacetic acid solution and penicillin-streptomycin double antibody solution were obtained from Basaimedia (China). The collagenase I was obtained from Solarbio (China). The cell counting kit-8(CCK8) and cell tissue lysate for RNA extraction were obtained from Coolaber (China).

Isolation and culture of endometrial stromal cells

As described previously19, primary human endometrial stromal cells were isolated and cultured. Inclusion criteria for endometrial tissue extraction: (1) healthy women of childbearing age between 20 and 30 years old undergoing hysteroscopy in our hospital; (2) regular menstrual cycle; (3) endometrial inflammation, infection and other diseases were excluded (4) infectious diseases, recent medication and other influencing factors were excluded. Briefly, the human endometrium tissue was poured into Petri dishes and rinsed with phosphate-buffered saline (PBS) at least thrice until there was no blood. The chopped endometrial tissue was transferred to a 15 mL centrifuge tube and 3 mL of 1 mg/mL collagenase I was added. After 1 h in a 37 °C water bath, 3 mL DMEM was added to prevent overdigestion. The digested cell supernatant was filtered and collected using a 70 μm screen (Biosharp, Labgic) and then centrifuged to obtain stromal cells. Subsequently, the stromal cells were resuspended in DMEM medium containing 10% fetal bovine serum and antibiotics (1% penicillin-streptomycin solution) and cultured in a cell culture incubator at at 37 °C with 5% CO2. Written informed consent was received from patients before inclusion. The research project received approval from the Ethics Review Committee of the First Hospital of Lanzhou University (Approval number: LDYYSZLLKH2023-04).

Cell viability assay

Cell Counting Kit-8 reagent was used for simple and accurate analysis of cell proliferation and toxicity. Human endometrial stromal cells were seeded into 96-well plates at a density of 3000 cells per well. The cells were treated with PFOS at various concentrations (0.01, 0.1, 1, 10 and 100 µM), and then cultured for 24, 36, 48, 60 and 72 h, respectively. Further, 10 µL of CCK-8 solution was added to each well, and the cells were incubated at 37 °C for 2 h to detect cell viability.

Real-time quantitative PCR (RT-qPCR) analysis

The mRNA expression levels of Bax, Bcl-2, HOXA10, and ITGB3 were measured using RT-qPCR. Total RNAs of endometrial stromal cells were extracted with Trizol reagent (Coolaber, Beijing, China). Nanodrop 2000 spectrophotometer was used to assess the RNA purity. FastKing gDNA Dispelling RT SuperMix kit (Tiangen, Beijing, China) was used to synthesize first-strand complementary DNA. The amplification program was run in QuantStudio™ 3 System (Thermo Fisher Scientific, Rockford, IL). SuperReal PreMix Plus was used to amplify DNA. Amplification was performed for 40 cycles consisting of 5 s at 95 ℃, 30 s at 60 ℃, and then 15 s at 95 ℃. Each experiment was performed in triplicates. The primers used for qPCR were synthesized by Shanghai Sangon Biotechnology Co., Ltd. and are presented in Table 1.

Table 1 Primer sequences for RT-qPCR.
Full size table

Western blot analysis

All proteins of endometrial stromal cells were extracted by cell lysis buffer (Coolaber). Proteins were quantified using a bicinchoninic protein assay kit (Coolaber, Beijing, China). Equal amounts of the extracts were electrophoresed on sodium dodecyl sulfate-polyacrylamide gel and transferred onto a polyvinylidene difluoride membrane (Millipore, Bedford, MA) using the electrophoresis & transmembrane system of Bio-Rad (Hercules, CA). The membrane was incubated with primary antibodies including GAPDH, Bax, Bcl2, HOXA10, ITGB3, FOXO1, cytokeratin 18, and vimentin from Proteintech (China). The membrane was rinsed and incubated with HRP-conjugated secondary antibody (Proteintech, Wuhan, China). Immune-reactive proteins were detected using meilunbio FGSuper sensitive enhanced chemiluminescence (Meilun, Dalian, China). The values of optical densities of the proteins were converted into quantitative data and analyzed using image J. Each experiment was performed in triplicate. In addition, to detect multiple proteins on the samemembrane, we trimmed the membrane according to theprotein molecular weight range in the instructions priorto hybridization with the antibody, and all original blotsas well as replicates are provided in related files.

Flow cytometry assay

Cell apoptosis was detected using an Apoptosis Detection kit (Lianke, China). Collect the suspended cells, add an appropriate amount of pre-cooled 1×Binding Buffer and mix by a pipettor blowing, so that the resuspension is filtered through a screen and transferred to a flow-through tube, in which a single stained tube was added with 5 µL of Annexin V-FITC or 10 µL of PI, respectively, and incubated for 5 min away from light. Detection of annexin V-FITC by FITC using flow cytometry (LSRFortessa) (Ex = 488 nm; Em = 530 nm) and PI (Ex = 535 nm; Em = 615 nm).

Transcriptomics analysis

RNA extraction and cDNA synthesis as described previously (“Real-time quantitative PCR analysis”). RNA sequences were analyzed by Shanghai Majorbio Bio-pharm Biotechnology Co., Ltd. (Shanghai, China). The sequencing library was performed by Majorbio using an NovaSeq 6000 PE150 platform (Illumina). Trimmomatic software was used to process raw reads20. Next, Raw sequencing data (fastq) were aligned using the HISAT2 (version 2.1.0). In order to summarize read counts for each gene, HTSeq (Version 0.6.1) was used to generate gene-level count data. Finally, genes with differential expression (log2FC > 1 and P value < 0.05) were filtered using the DEseq algorithm. Genetic variations was analyzed by KEGG (www.kegg.jp/kegg/kegg1.html).

Cell cycle analysis

Cell cycle stage were detected with a cell cycle assay kit (Lianke, China) using flow cytometry. Human endometrial stromal cells were seeded into 6-well plates and incubated overnight. The cells were treated with PFOS (0.01, 0.1, 1 µM) for 24 h. Next, the cells were collected and washed 1 time with PBS, subsequently, 1 mL of DNA staining solution and 10 µL of permeabilization solution were added and mixed by vortexing, and finally incubated for 30 min in the darkness at room temperature. The samples were subsequently analyzed by flow cytometry.

Molecular docking

First, we downloaded the molecular structure of PFOS from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) and the structures of the target proteins HOXA10, ITGB3 from the PDB database (https://www.rcsb.org/structure), and PyMoL software was used to remove water and original ligands. Secondly, we used AutoDock Vina21 to perform molecular docking and PyMOL to visualize the docking results. Finally, the binding strength and activity of protein targets and PFOS were assessed by molecular docking binding energy.

Immunofluorescence staining

Twenty-four hours following PFOS exposure, endometrial stromal cells were fixed with 4% paraformaldehyde for 30 min and then washed with washing buffer (PBS containing 0.01% Triton X-100, 0.01% Tween, and 2% BSA). Afterward, the endometrial stromal cells were sealed in a washing buffer at 37 ℃ for 2 h and incubated overnight at 4 ℃ with rabbit anti-HOXA10 and anti-ITGB3 polyclonal antibody (Bioss, Beijing, China). The following day, endometrial stromal cells were washed with washing buffer (thrice for 10 min each) and then incubated with goat anti-rabbit immunoglobulin G secondary antibody (Bioss, Beijing, China) at 37 ℃ for 1 h. Endometrial stromal cells were mounted on glass slides and observed under a fluorescence microscope (Nikon, Japan).

Transmission electron microscopy (TEM)

We collected cells with 0.1µM PFOS exposure intervention for 24 h and sent them to electron microscopy for detection. The cells in the culture flasks were collected and centrifuged, and the supernatant was discarded. TEM fixative (Servicebio, China) was added to the tubes, mixed, and then stored and transported at 4℃. They were observed and photographed using a transmission electron microscope. Transmission Electron Microscope using hitachis’ factory production (HT7800).

Mice exposure model

Animal experimentation protocols were approved by the Ethics Committee of the First Hospital of Lanzhou University (Approval number: LDYYSZLLKH2023-05), and were conducted in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals and ARRIVE guidelines. The 6-8-week-old female C57BL/6 mice were purchased from the Experimental Animal Center of Lanzhou University. Mice were housed with controlled temperatures (22 ± 1 ℃), humidity (40–60%), and light/dark cycles (12 h each) under pathogen-free conditions. Mice in the PFOS group were given different concentrations of PFOS (0.5 mg/kg, 1 mg/kg, and 2 mg/kg) by gavage for 21 consecutive days, while the control group was given an equal amount of saline, and after 21 days, sampling was carried out and mice uteri were placed in a 2.5% glutaraldehyde fixative (Servicebio, China) for subsequent electron microscopic observation of pinopodes. Mice were euthanized with a lethal dose of sodium pentobarbital (120 mg/kg).

Estrous cycle determination

Mice vaginal smears were collected at 08:00 a.m. every day for 21 consecutive days, starting from day 1 of modelling. The detached vaginal epithelial cells were fixed and subjected to hematoxylin and eosin (HE) staining, followed by observation under a light microscope (Olympus, model BX51) to determine the stage of the estrous cycle in which the mice were in. Proestrus consists of mostly round nucleated epithelial cells; estrus was characterized by cornified squamous epithelial cells; metestrus consists of cornified epithelial cells and leukocytes; The diestrus stage has a predominance of leucocytes.

Statistical analysis

The statistical analyses were performed using IBM SPSS 22.0 and Graphpad prism 9. Continuous variables were presented as means ± standard deviations, and a t-test was used for comparisons. If normality was not satisfied, the Mann − Whitney U test was used for comparisons. Categorical variables were presented as a percentage. P < 0.05 was considered statistically significant. Western blotting and PCR were repeated thrice.

Results

PFOS inhibited cell viability

The chemical structural formula of PFOS is shown in Fig. 1A. To assess the purity of the primary endometrial stromal cells culture, immunofluorescence staining of the cells with anti-CK18 (a marker of epithelial cells) and anti-vimentin (a specific marker of stromal cells) were performed. As shown in Fig. 1B, vimentin predominantly localized within the cytoplasm of endometrial stromal cells. To assess PFOS cytotoxicity, we used a CCK-8 to evaluate the effect of 0.01µM -100µM PFOS on endometrial stromal cell proliferation for 24 h. The CCK-8 assay revealed that PFOS inhibited human endometrial stromal cells in a concentration-dependent manner, compared to the control group (Fig. 1C). We further treated cells with 0.1µM PFOS for 24–72 h and found that cell viability decreased with increasing duration of action (Fig. 1D). The morphology of the cells changed in response to PFOS, by becoming more wrinkled and sparse (Fig. 1E).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Effect of PFOS on human endometrial stromal cells viability. (A) Molecular formula of PFOS. (B) Characterization of primary human endometrial stromal cells. (C) The effect of different concentrations of PFOS on cell viability was detected by CCK8 assay. (D) Effect of PFOS on cell viability at different times of action. (E) Effect of different concentrations of PFOS on cell morphology. Results are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

PFOS induced apoptosis in endometrial stromal cells

Cell RNA was extracted after 24 h of 0.1µM PFOS exposure and amplified for analysis. Next, we used RT-qPCR and Western Blot technology to detect the expression of apoptosis-related genes BAX and BCL-2 in cells. As shown in Fig. 2A, PFOS exposure significantly increased the mRNA expression of BAX, a pro-apoptotic indicator, and decreased the mRNA expression of BCL-2, an anti-apoptotic indicator, compared with the control group. Endometrial stromal cells were treated with 0.01µM, 0.1µM, 1µM PFOS concentrations for 24 h, followed by WB and flow cytometry apoptosis assays.Consistent with our qRT-PCR results, PFOS treatment significantly increased BAX protein expression and decreased BCL-2 protein expression (Fig. 2B, E). We further examined the effect of PFOS on apoptosis by flow cytometry, as shown in Fig. 2C-D, Q2 and Q3 area indicated late apoptotic or necrotic cells and early apoptotic cells, respectively, and their combination indicated the proportion of apoptotic cells. Our results showed that 1 µM PFOS treatment significantly increased apoptosis in normal cells (3.06% vs. 15.28%).

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Effect of PFOS on endometrial stromal cells apoptosis. (A) The mRNA expression levels of BAX, BCL2 were detected by using RT-qPCR. (B, E) Expression and quantification of BAX and BCL2 proteins in cells. (C, D) Cell apoptosis was measured by flow cytometry. Results are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Transcriptomics analysis

We used 0.1µM PFOS exposure for 24 h as a transcriptomics intervention condition. The effects of PFOS were further analysed by transcriptomics. The PCA showed satisfactory separation between the two groups, the PFOS group all clustered on the left side and the control group all clustered on the right side (Fig. 3A). The Venn diagram shows the number of common and specific genes between the two groups (Fig. 3B). The volcano plots showed the differential genes between the control and PFOS groups, there were a total of 1204 differentially expressed genes between the two groups, of which 441 were up-regulated and 763 were down-regulated (Fig. 3C). As shown in Fig. 4A, the heatmap showed the top 20 genes with the greatest differential expression. We carried out GO and KEGG enrichment analyses to examine the differences in gene function enrichment and pathways between the two groups. The results of the GO annotation analysis showed that the differential genes were mainly enriched in cell cycle process, mitotic cell cycle process, and regulation of biological process, etc. (Fig. 4B). KEGG enrichment analysis revealed that these differential genes were mainly enriched in in cell cycle, cell adhesion molecules, and homologous recombination, etc. (Fig. 4C). We found that the cell cycle is a biological process and pathway in which differential genes are predominantly enriched. Therefore, we performed the cell cycle assay by flow cytometry, and found that PFOS treatment resulted in an increased proportion of the endometrial stromal cells population in the G1 phases (increased cell cycle arrest) (Fig. 4D).

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Transcriptomic analysis after PFOS treatment. (A) The PCA scores plot of control and PFOS group, n = 3 per group. (B) Venn diagram of thecontrol and PFOS group. (C) Volcano plots of the differential genes in control and PFOS group, red denotes the up-regulation of differential genes, blue denotes down-regulation of differential genes, and grey represents non-significant differential genes.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Enrichment analysis of differential genes. (A) Heat map of the top 20 differential genes between the two groups. (B) GO enrichment analysis. (C) KEGG enrichment analysis. (D) Cell cycle analysis.

PFOS down-regulated endometrial receptivity in endometrial stromal cells

Subsequently, we examined the impact of PFOS on endometrial receptivity following a 24-hour exposure period across various concentrations. HOXA10 and ITGB3 are important indicators associated with endometrial receptivity. Molecular docking showed that the binding energies of PFOS with HOXA10 and ITGB3 were − 6.2 kcal/mol and − 6.5 kcal/mol, respectively, suggesting a strong interaction between them (Fig. 5A). RT-qPCR results showed that PFOS treatment significantly reduced the mRNA levels of HOXA10 and ITGB3 compared with the control group (Fig. 5B), similarly, Western Blot results showed that PFOS treatment significantly reduced the protein levels of HOXA10, ITGB3, and FOXO1 (Fig. 5C, D). Consistent with the Western Blot results, immunofluorescence further confirmed that PFOS reduced the expression of HOXA10, ITGB3 and FOXO1 in endometrial stromal cells (Fig. 5E, F, G). Quantitative analysis of immunofluorescence is shown in (Fig. 5H). In addition, transmission electron microscopy showed that the mitochondrial structure was impaired in the endometrial stromal cells after 1 µM PFOS treatment for 24 h, as evidenced by the slight swelling of mitochondria (red arrows), the thinning of the matrix within them, and the breakage of the cristae (Fig. 5I), which provided a more specific explanation for the negative effects of PFOS on the endometrial stromal cells.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Effect of PFOS on endometrial receptivity. (A) Molecular docking of PFOS with HOXA10 and ITGB3. (B) The mRNA expression levels of HOXA10, ITGB3 were detected by using RT-qPCR. (C-D) Expression and quantification of HOXA10, ITGB3 and FOXO1 proteins in cells. Mitochondrial damage (red arrow) in endometrial stromal cells. Immunofluorescence staining reveals expression of HOXA10 (E), ITGB3 (F) and FOXO1(G) in endometrial stromal cells. (H) Immunofluorescence statistical analysis. (I) Transmission electron microscopy showed the normal mitochondrial morphology (green arrow), Scale bar = 200 μm. Results are presented as the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Effects of PFOS exposure in mice

We further explored the effects of PFOS by mice experiments (Fig. 6A). Monitoring of the estrous cycle in all groups of mice revealed that PFOS administered mice exhibited disturbances in the estrous cycle. It was seen that the control group presented a normal estrous cycle, PFOS exposure led to disruption of the estrous cycle in mice, with the low (0.5 mg/kg) and medium (1 mg/kg) PFOS dose groups showing a stay in one period for multiple days, and the high (2 mg/kg) PFOS dose group persisting in a non-estrous state (Fig. 6C, D). In addition, as shown in Fig. 6B, the gross morphology of the mice uterus was significantly thinner and smaller after treatment with 2 mg/kg of PFOS compared with the control group. Scanning electron microscopy results showed severe cytosolic synapse damage with increasing concentrations of PFOS drug exposure in mice, indirectly responding to decreased endometrial receptivity (Fig. 6E).

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Effects of PFOS exposure in mice. (A) Flow chart of mice handling in each group. (B) The change in gross morphology of the uterine tissue after PFOS treatment. (C) Typical images of vaginal smears from mice at different stages of the estrous cycle, Scale bar = 200 μm. (D) Effects of PFOS exposure on the estrous cycle in mice (P: proestrus; E: estrus; M: metestrus; D: diestrus). (E) Pinopodes morphology under scanning electron microscope, Scale bar = 100 μm, n = 5 per group.

Discussions

The issue of declining fertility rates among women has garnered significant attention in recent discourse, with endometrial receptivity being increasingly recognized as a critical factor in successful fertilization. This study integrates both cellular and animal model experiments to investigate the impact of PFOS on the endometrium and to elucidate the potential mechanisms underlying its effects.

PFOS is a part of PFAS, even though its applies have been banned by China since 201922, the risks on human health caused by its high emmision and persistence previously can still be worrying. The distribution of PFOS in human body is wide, so far, it has been detected in blood, urine, and breast milk23,24,25. Beyond that, the existence of PFOS in placenta is also noteworthy for its highest concentration among various PFAS in placenta which reaches to 0.84ng/g26. Reported by a study on primary human decidual stromal cells, PFOS can disrupts cortisol regeneration and then impairs decidualization and immune tolerance environment in early pregnancy17. Anothor investigation on women with endometriosis indicated that their median concentration of total PFOS in serum was 12.6ng/mL, notably higher than normal wmen. Meanwhile, the concentrations of total PFOS and its isomeride were significantly associated with endometriosis27. These studies showed that PFOS exposure may cause negative effects on uterus.

Our study demonstrated that PFOS reduces the viability of primary human endometrial stromal cells in a manner that is both concentration- and time-dependent. Taking this as the starting point, we explored the effect mechnaism of PFOS on uterus subsequently. After treated with 1µM PFOS, the proportion of apoptotic cells obviously increased, from 3.06 to 15.28%. Meanwhile, the apoptosis-related molecular BAX and BCL-2, their mRNA and protien expression were also notably changed. Then, transcriptomics analysis showed the differential genes between PFOS treated group and control group were mainly enriched in cell cycle, the subsequent result also found obvious cell cycle arrest in G1 phase after 1µM PFOS treatment. Apoptosis is a programmed death in cell, strongly associated with cell proliferation28. Cell cycle can regualte apoptosis, manifesting as cell apoptosis needs to effect on moleculars in G1 late phase and the transformation from G1 phase to S phase needs the control of p5329,30, which can lead the decrease of BCL-2 and the increase of BAX to trigger apoptosis31,32. These changes on apoptosis and cell cycle provides possible explanation for uterus toxicity of PFOS.

Endometrial receptivity is modulated by estrogen, progesterone, and various endocrine factors, which collectively facilitate the attachment, invasion, and development of the blastocyst33,34. The impairment of endometrial receptivity is deemed as the major reaseon of embryo implantation failure35. Studies have shown that pinopodes are processes formed by edema and fusion of the apical microvilli of the endometrial epithelium, which are susceptible to exogenous estrogen and progesterone, and participate in the adhesion process of blastocysts to the endometrium during the implantation window36. HOXA10 plays a central role in endometrial development during the menstrual cycle, and the expression of HOXA10 is related to the levels of estrogen and progesterone. Expression increased during the luteal phase and reached peak levels at WOI. HOXA10 improves endometrial receptivity and receives blastocyst signals, modifying the endometrium to decidualize, enabling trophoblast invasion and placentation37. ITGB3 was co-expressed in the whole normal menstrual cycle and increased in the mid-secretory period, which was consistent with the window period of implantation. They may also serve as good endometrial receptivity markers38.Vascular endothelial growth factor (VEGF) plays crucial role in embryo implantation and development in early pregnancy, its increased expression is conducive to improve endometrial receptivity39. Meanwhile, FOXO1 can regualte VEGF-A directly. It plays important role in angiogenesis through down-regualting anti-angiogenic signal CD36, activating pro-angiogenic signal VEGF and building polarity of endothelial cells40. HOXA10, which is also necessary for embryo implantation, its expression in endometrium was lower in women with endometriosis, adenomyosis and recurrent abortion41,42. Besides, the expression of ITGB3 was decreased in women with recurrent abortion and positvely related to the expression of HOXA1043,44,45. Our results found PFOS exposure caused decreased mRNA and protein expression of HOXA10, ITGB3 and FOXO1, confirming the damage of PFOS on endometrial receptivity.

Another study indicated that PFOS exposure can disrupt trophoblast motility including migration, invasion and angiopoiesis by excessive reactive oxygen species (ROS), reduced ATP and declined mitochondrial membrane potential46 while mitochondrial dysfunction will cause polarization of M1-like macrophages in decidua and induce recurrent abortion47. We also found swollen mitochondria and cristae fracture in primary human endometrial stromal cells, these mitochondrial damage characteristics were consistent with previous studies. Our results provided more evidences for confirming the negative effects of PFOS on uterus.

The menstrual cycle may also be influenced by PFOS. In a cross-sectional study, women with higher concentrations of PFOS exhibited increased odds of experiencing irregular and prolonged menstrual cycles13. Its finding was also supported by another study48. Animal experiments showed that took 10 mg/kg PFOS orally exhibited activation of AVPV-kisspeptin neurons, caused decreased gonadotropin-releasing hormone (GnRH), progesterone and luteinizing hormone (LH) and thus induced decreased corpus luteum and prolonged gestation period49. In addition, it was also reported that PFOS exposure induced increased estradiol and irregular estrous state more frequently50. In our study, 2 mg/kg treated group had smaller and thiner uterus and persisting in a non-estrous state. With all of these reults together, the effects of PFOS on neruo-reproductive endocrinology may associated with its damages on uterus.

This stuyd combined in vivo and in vitro experiments, exploring the effect mechnaism of PFOS on uterus from different aspects. We provided more experimental evidence for confirming uterus toxicity of PFOS, but there are several limitations. First, sex hormones are also indispensable for the normal function of endometrium, but we did not detected the changes of sex hormones and aromatase activity in mice after PFOS exposure. In addition, we only detected partial phenotypes, whereas PFOS may affect endometrium by modulating some signal pathway, such as PI3K/AKT/mTOR is a classic pathway which realted to apoptosis51,52. Furthermore, the primary human endometrial stromal cells we used were extracted from different individual, it may have certain individual difference so we expected these results could be verified by more stable cell lines. The threats of PFAS are more and more arrestive and it has been an urgent problem. We hope more studies could figure out their pathogenic mechanism, urging related department take effective measures to reduce their emission and intervene the harmful effects early.