Maternal Bisphenol B Exposure and Mammary Gland Development of Offspring: A Time-Series Analysis.

The results showed that BPB exposure did not alter the maternal weight at any time point during gestation ( Figure 1 A), with an unchanged overall food intake, except at GD2 and GD14 ( Figure 1 B). In this study, the total time of off-nest behaviors and eating/drinking was suppressed by BPB exposure; however, the time spent on these behaviors every day was not changed significantly except the licking/grooming time on lactation day 2 ( Figure 1 G–L). As for nursing, BPB exposure did not alter arched-back and passive nursing behaviors at any time point ( Figure 1 C–F). In addition, maternal BPB exposure did not alter the number, sex ratio, and weight of female and male offspring at any time point after birth ( Figure 1 M and N and Figure S2 ). Maternal weight and behavior of mice exposed to BPB. (A) Effects of BPB exposure on dam’s weight ( n = 9/group). (B) Effects of BPB exposure on food intake of pregnant mice ( n = 9/group). Time spent by dams displaying nursing (C), arched-back nursing (D), passive nursing (E), licking/grooming (G), eating/drinking (I), and off-nest behaviors (K) from LD2 to LD8 ( n = 6/group). Pooled time spent by dams displaying nursing (F), licking/grooming (H), eating/drinking (J), and off-nest behaviors (L) from LD2 to LD8 ( n = 6/group). (M) Number of offspring of dams after BPB exposure ( n = 9/group). (N) Female offspring’s weight of dams exposed to BPB ( n = 44–47 from 9 dams at PND1, n = 25–40/group from 9 dams at PND7, n = 13–25/group from 9 dams at PND14, n = 12–25/group from 9 dams at PND21, n = 9–15/group from 9 dams at PND90). * p < 0.05, compared with the corresponding vehicle group. Next, the vaginal opening (VO) and vaginal cytology during the regular cycling of female offspring were analyzed. The average age of VO in the BPB exposure group was PND26.19, while that of the vehicle control group was PND28.00, suggesting that sexual maturity was slightly accelerated by maternal BPB exposure significantly ( p = 0.07) ( Figure 2 A). The mice begin sexual cycling immediately after their vaginal orifice opens. Next, observation of vaginal epithelial cell structure based on the vaginal smear assay was employed as a relatively noninvasive method to monitor reproductive cycles for four consecutive weeks. Our results indicated that the total and average duration of normal diestrus, estrous, metestrus and proestrus did not significantly change between groups ( Figure 2 B and Figure S3 ). In addition, serum estradiol and progesterone levels of female offspring were measured at PND21. Prior to VO at PND21, the mean serum estradiol concentration of female offspring in the BPB group decreased significantly when compared with vehicle control (Veh: 2.5 ng/mL; BPB: 1.12 ng/mL; p < 0.01) ( Figure 2 C). Serum progesterone concentrations were also significantly lower than vehicle controls in BPB groups at PND21 ( p < 0.01) ( Figure 2 D). Next, ex vivo hypothalamic explant incubation was carried out to compare the BPB effects on GnRH secretion at PND21. As shown in Figure 2 E–G, the GnRH interpulse interval was modestly but significantly higher in BPB groups when compared with the vehicle control group ( p < 0.01). Pubertal timing (VO) and estrous cycle of female offspring after maternal exposure to BPB or vehicle. (A) Effects of maternal BPB exposure on the age of VO in female offspring ( n = 10–16/group). (B) Pooled days of four stages of the female offspring estrous cycle ( n = 6/group from different litters) after BPB exposure. (C) Estradiol and (D) progesterone serum levels of female offspring at PND21 ( n = 4–6/group from different litters). Representative image of graphically expressed GnRH interpulse interval of Veh (E) and BPB-exposed (F) mice using connected dots at PND21. (G) Mean time of GnRH interpulse interval of hypothalamic explants of vehicle control and BPB-exposed female offspring at PND21 ( n = 3/group from different litters). ** p < 0.01, compared with the corresponding vehicle group. Endocrine disorder may be related to the development of the mammary gland, but our results did not show any morphological changes in the mammary glands of female offspring in the maternal BPB exposure group at different time points relative to the vehicle control using whole-mount staining ( Figure 3 ). Briefly, during the juvenile development periods (PND1 and PND10), the number of total branches, the area of the branch, and the longitudinal length did not change significantly. The pubertal mammary gland development index after weaning (PND21) such as branch area, length and TEB area was reduced slightly without significance. At PND90, which represents the adult age of female mice whereby ductal extension and branching equalize, BPB treatment did not change mammary gland development represented by unchanged length, width, area, and density. The organ ratio of the mammary glands to body weight in the BPB group remained unchanged for PND21 and PND90 ( Figure 3 H). Characteristics of mammary gland from BPB-exposed females on different time points. (A) Representative whole-mount staining images of mammary gland at PND1, PND10, PND21 and PND90 (control: n = 4, BPB: n = 5 at PND1; n = 3/group at PND10; control: n = 3, BPB: n = 4 at PND21; n = 7/group at PND90, from different litters). (B) Total branch, (C) area and (D) length of mammary gland at PND1, PND10 and PND21. (E) TEB area of mammary gland at PND21. (F) Length, width and area of mammary gland at PND90. (G) Mammary gland branches density at PND90. (H) Organ to body ratio of mammary gland at PND21 and PND90 (control: n = 5, BPB: n = 7 at PND21; control: n = 9, BPB: n = 12 at PND90, from different litters). Although maternal BPB exposure did not alter the branch development by whole-mount staining, whether its potential indicators such as cell differentiation, histopathological, and genetic changes have changed needs further attention. Fully formed adult mammary epithelium consists of bilayered ducts, formed by luminal cells surrounded by a basal layer composed mainly of myoepithelial cells. As shown in Figure 4 A, maternal BPB exposure did not alter the luminal (CD29 low /CD24 + ) to basal (CD29 hi /CD24 + ) cell ratio at PND90 ( Figure 4 A and B). In addition, maternal BPB exposure did not induce a significant reduction in organoid branch formation and elongation ( Figure 4 G and Figure S4 ). Luminal cells can be subdivided into hormone receptor-expressing and nonhormone receptor-expressing luminal cells (HR + and HR – , respectively) and Sca1 marking both HR + cells and their progenitors. 37 , 39 BPB exposure increased the proportion of differentiated HR + cells relative to the vehicle control group ( Figure 4 C and D). We hypothesized that the increased HR + cells are related to serum hormone levels. Serum estradiol and progesterone levels were measured at PND90, and serum estradiol levels were restored to normal level when compared with PND21, while progesterone concentration remained significantly lower in the BPB group ( p < 0.01) ( Figure 4 E and F). Luminal, basal MECs and HR + cell differentiation of female offspring at PND90. (A) Representative CD24/CD29 FACS plot showing relative amounts of lineage marker negative luminal and basal MECs from PND90 female offspring of vehicle and BPB treatment group ( n = 5/group from different litters). (B) Quantification of the percentages of basal and luminal MECs. (C) Representative Sca1-BV711 histograms (HR + cells) of individual female offspring at PND90 after maternal vehicle and BPB exposure. (D) Quantification of the percentages of luminal HR + subpopulations. (E) Estradiol and (F) progesterone serum levels of female offspring at PND90 (control: n = 5, BPB: n = 6, from different litters). (G) Organoid morphologies of vehicle control and BPB groups grown in matrigel for 7 days (control: n = 3, BPB: n = 4, from different litters) at PND90. * p < 0.05, ** p < 0.01, compared with the corresponding vehicle group. Next, to determine the mechanism by which maternal BPB exposure possibly altered hormone concentration and HR + luminal cells, RNA-seq combined with a time-series analysis of mammary glands of female offspring at four time points was performed. First, the possible GO and KEGG enrichment of DEGs at different time points was analyzed ( Figures S5–S8 ). At PND1, BPB exposure mainly altered epidermis development related genes. At PND10, cell cycle phase transition and phosphorus metabolic process were enriched by DEGs after BPB exposure. At PND21, DEGs mainly took part in some fatty acid metabolic process. At PND90, oxidative phosphorylation and epidermis development process were enriched after BPB exposure. In transcriptome data analysis, time-series analysis is mostly used to analyze the dynamic changes of different genes with time points, especially when samples (belonging to the same processing method) may show different biological characteristics due to different time points. Time-series analysis is also a method to analyze the development process, direction, and trend of time series and predict the possible goals in the time domain in the future. Therefore, the time-series analysis for RNA-sequence results of genes is not a simple difference analysis between different time points but focuses on the expression trend. In this study, 891 DEGs were identified in the vehicle control group, and 2533 were identified in the maternal BPB exposure group; 527 DEGs were commonly shared by the vehicle control and BPB groups ( Figure 5 A and B). These results strongly suggested that maternal BPB exposure could cause a widespread change in the transcriptomic profile of the mammary glands of female offspring over time. DEGs in the vehicle control group were indicative of normal development. Thus, evaluating genes that are alternatively regulated in the BPB-exposed group would be insightful. Effects of maternal BPB exposure on the whole transcriptome of the mammary glands of female offspring at different time points. (A) Heat maps showing the time-series expressed DEGs in mammary glands of the vehicle and BPB groups ( n = 3/group from different litters). (B) Venn diagram representing the number of time-series expressed DEGs in vehicle and BPB-treated offspring. GO enrichment showing the biological processes enriched with the DEGs with a time series in (C) vehicle group and (D) BPB group. (E) Disease analysis (up) and mammary gland related disease analysis (down) of time-series expressed DEGs only in BPB group. (F) GO enrichment of the time-series expressed DEGs of mammary gland related diseases. (G) Sankey plot of five GO enrichment of time-series expressed DEGs of mammary gland related diseases. (H) Mammary gland development related gene expression at different time points in vehicle and BPB groups. Blue line: BPB group, Grey line: Vehicle control group. * p < 0.05, ** p < 0.01, *** p < 0.001, compared with the corresponding vehicle group. Next, GO analysis was conducted to examine the potential biological pathways involved in BPB-exposure-induced changes in the mammary glands. The results showed that 364 DEGs with a time-series expression pattern in the control group were mainly enriched in the metabolic processes of biological macromolecules including lipids, carbohydrates, and nucleosides ( Figure 5 C). Among the 2006 DEGs in the BPB group only, the morphogenesis of a branching structure, the morphogenesis of a branching epithelium, and the branching morphogenesis of an epithelial tube were overrepresented, which may be related to disturbed mammary gland development. Some pathways involved in leukocyte cell–cell adhesion activation and system development were also enriched ( Figure 5 D). Of the BPB group, 2006 DEGs were imported into GeneCards for disease enrichment analysis. The top five enriched diseases were those of mental health, breast cancer, colorectal cancer, prostate cancer, and lung cancer; the top five diseases mostly related to the breast and mammary gland, including 220 of 2006 DEGs, were breast cancer, breast adenocarcinoma, breast disease, breast ductal carcinoma, and estrogen-receptor positive breast cancer ( Figure 5 E). Thus, 220 genes associated with breast disease were subjected to GO analysis to determine the biological processes involved in mammary gland development induced by maternal BPB exposure. The results of GO analysis showed that maternal BPB exposure altered epithelial cell proliferation, gland morphogenesis, and gland development ( Figure 5 F and G). Furthermore, the expression of genes participating in these biological processes was analyzed at four time points. As shown in Figure 5 H, 12 genes showed a time-series pattern in the BPB group and were altered significantly relative to their expression in the control group at one or more time points. Briefly, the expressions of the progesterone receptor ( Pgr ), epidermal growth factor receptor ( Egfr ), GATA binding protein 3 ( Gata3 ), TNF superfamily member 11 ( Tnfsf11 ), and E74 like ETS transcription factor 5 ( Elf5 ) were upregulated on PND90 after maternal BPB exposure. Considering that Pg levels in the BPB group were significantly suppressed at PND90, the toxicity of BPB but not Pg enhanced or activated the expression of Pgr . These results further implied an endocrine-disrupting effect of BPB. Moreover, maternal BPB exposure elevated the expression of fibroblast growth factor receptor 2 ( Fgfr2 ), vitamin D receptor ( Vdr ), and Gli1 on PND1 only. The expression of chemokine (C–C motif) ligand 5 ( Ccl5 ) in the BPB group was downregulated at birth, prepubertal, and pubertal stages and returned to normal levels in adulthood. Kinase insert domain protein receptor ( Kdr ) was upregulated at PND21, nerve growth factor receptor ( Ngfr ) was upregulated at PND10, and leptin ( Lep ) was elevated at PND21 and PND90 in the BPB group. Endocrine pathways, including PPAR, estrogen, progesterone, growth hormone, and thyroid hormone, were also analyzed at different time points. At PND1, the most obviously disturbed pathways were estrogen and peroxisome proliferator activated receptor (PPAR), and PND10 were thyroid hormones. At PND21, PPAR pathway-related genes such as peroxisome proliferator activated receptor alpha ( Pparα ), growth hormone-related genes, insulin receptor substrate 1 ( Irs1 ) and Irs3 , and the thyroid hormone-related gene, thyroid hormone receptor beta ( Thrb ), were significantly altered. 44 DEGs participated in the endocrine signaling pathway at PND90, and Pgr was significantly elevated at this point ( Figures S9 and S10 ). As mentioned in the RNA-seq results, the Pgr expression was upregulated at PND90 after maternal BPB exposure. In addition, the progesterone levels in serum of female offspring at PND21 and PND90 both showed a downward trend. To validate the Pgr expression, RT-PCR and IHC staining were performed. The results showed that maternal BPB exposure indeed elevated the level of mRNA expression of Pgr in female offspring mammary glands at PND90 ( Figure 6 A). IGV browser showed Pgr region in the mouse mm10 genome, displaying results for all three mice from vehicle control and BPB group, and aligned reads in the BPB group were richer than in the vehicle control group ( Figure 6 B). Furthermore, IHC staining results confirmed that the expression of PR was indeed elevated in nuclear mammary gland luminal cells at PND90 after BPB exposure ( Figure 6 D). In addition, IHC results of PR in human tumor tissue showed a strong expression in cytoplasmic, membranous, and nuclear of mammary gland when compared with normal tissue ( Figure 6 E). Then, the mRNA expression of breast cancer related biomarkers such as Gata3 , Egfr , Tnfsf11 , and Elf5 showed similar trends to Pgr at PND90 ( Figures S11–S13 ). In order to detect whether the alternation of these breast cancer related genes can cause pathological changes, the H&E stained sections were examined by a professional pathologist. The results did not indicate significant foci of the tumor, but a few ducts showed dilatations, lobular hyperplasia, and inflammatory cell infiltration in the BPB-treated groups at PND90 ( Figure 6 C and Table 1 ). Progesterone receptor (PR) expression in mammary gland of offspring at PND90 after BPB exposure. (A) Pgr mRNA expression of mammary gland of female offspring at PND90 using RT-PCR ( n = 5/group from different litters). (B) IGV genome browser visualization of RNA-seq reads alignment of Pgr for all three mice at PND90 in vehicle and BPB group. (C) Some abnormal structures of mammary gland of female offspring at PND90 ( n = 3/group from different litters). (D) Representative IHC images of PND90 mammary glands immunostained with PR ( n = 3/group from different litters). (E) Representative IHC images of human normal and tumor tissue immunostained with PR collected from online database. ** p < 0.01, compared with the corresponding vehicle group. Note: the numbers in brackets represent the degree of histopathological damage, 0 represents normal, 1 represents mild, 2 represents moderate, and 3 represents severe.

BPB (CAS: 77-40-7, purity >98%) was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan) and was dissolved in corn oil (HY-Y1888, MCE, NJ, U.S.A.) without tocopherol. Carnoy’s fixative was prepared using 60% absolute ethanol, 30% chloroform, and 10% glacial acetic acid. The whole-mount staining solution was prepared using 1 g of carmine alum and 2.5 g of potassium aluminum sulfate in 500 mL of H 2 O. The primary antibodies (CD29-FITC, CD24-APC, CD31-BV421, CD45- BV421, Ter119-BV421, Sca1-BV711 and CD49b-PE) for flow cytometry were purchased from Biolegend. Fifteen male and 30 female CD-1 (ICR) mice (7–8-week-old) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All mice were housed in light-, humidity-, and temperature-controlled rooms. Following a week’s habituation, male and female mice (quantity ratio = 1:2) were mated in a separate wire cage for a night. The next morning, plugs were checked; 23 plug-positive mice were considered to be pregnant and housed separately with free access to food and water. The day the plug was detected was recorded as embryonic day 0.5 (E0.5). At E0.5, pregnant mice were randomly divided into the vehicle control and BPB exposure groups, with each group containing 11 and 12 animals, respectively. In the BPB exposure group, plug-positive mice were treated with 300 μg/kg bw BPB by gavage every other day throughout the gestation period. Control mice were treated with corn oil following the same procedure. Finally, there were 102 offspring from 9 litters in the control group, with 46 female and 56 male offspring; there were 101 offspring from 9 litters in the BPB group, with 47 female and 54 male offspring. On PND1, PND10, PND21, and PND90, female offspring were euthanized, sacrificed, and disinfected with 70% alcohol; the mammary glands were harvested for RNA-seq, whole mount staining, histopathological analysis, flow cytometry analysis, and organoid culture. For further examination, serum was used for the detection of hormone contents. All animal procedures were approved by the Institutional Animal Care and Use Committee of Shanxi University (SXULL2019028). An infrared camera (360AP5C-J) was used to record maternal behavior for 1 h during the dark phase from lactational days 2 to 8 directly in the home cage under undisturbed conditions. Briefly, every video was visualized, and the conditions and duration of each behavior were manually recorded. 34 The feed weight of each pregnant mouse was weighed at 20:00 (W1) and 7:00 (W2) the next day. The food intake of each mouse was calculated as W1 minus W2. Pubertal maturation was determined as previously described. 35 Briefly, female offspring were randomly chosen (1–2 per litter), and the vaginal opening period was examined to monitor female puberty. From PND40 to PND70, a vaginal smear was conducted to estimate the estrous cycle at 7:30 am every morning. At PND1, PND10, PND21, and PND90, the fourth pairs of mammary glands were harvested. For each female offspring, the left side of the fourth pair of mammary glands was collected for whole-mount staining; a piece of the right side was used for hematoxylin and eosin (H&E) staining, and the remaining tissues were collected for flow cytometry analysis, organoid culture, and RNA-seq. The mammary gland was placed on a charged slide for 30 min and immersed overnight in Carnoy’s fixative. The next day, the mammary glands on the slides were rinsed in 70%, 50%, and 25% alcohol for 15 min. The mammary glands were stained with a carmine alum solution overnight. After rinsing in 70%, 95%, and 100% alcohol for 15 min, the mammary glands were defatted with xylene and observed under an optical microscope. Total branches, duct length, mammary gland area and terminal end bud (TEB) area and branching density were determined using the ImageJ software as previously described. 36 A part of the fourth mammary gland was collected and incubated in formaldehyde fixative (Wuhan Servicebio Technology Co., Ltd.) for 24 h at 4 °C. The biological samples were then paraffin-embedded and cut into 5–6 μm segments. Finally, the sections were stained with H&E. All histopathology samples were evaluated by a pathologist who was blind to treatment. Sections were dewaxed with xylene, rehydrated with gradient alcohol, and processed for antigen retrieval with citric acid retrieval buffer (Wuhan Service Biotechnology Co., Ltd.). Then, sections were incubated with 3% H 2 O 2 for 18 min to eliminate endogenous peroxidase and with 0.3% triton-X100 for 30 min to lyse the cell membrane and release nuclear protein. After blocking with goat serum, the sections were incubated overnight with progesterone receptor (PR) (Wuhan Service Biotechnology Co., Ltd.). The next day, sections were rinsed with PBS for 5 min three times and incubated with biotin-labeled secondary antibody working fluid for 18 min. After being washed with PBS, sections were incubated with horseradish peroxidase-labeled streptavidin working solution that was added on the sections and incubated at 37 °C for 18 min. Then, sections were stained with DAB (Beijing Zhongshan Jinqiao Biotechnology Co., Ltd., Beijing, China) for 5 min and hematoxylin for 1 min. Finally, the segments were dehydrated, cleared, sealed with neutral resin, and observed under a light microscope. Immunostaining images of human tissues were obtained from the Human Protein Atlas ( http://www.proteinatlas.org/ ). One of the fourth mammary glands without lymph node of each female offspring at PND90 were dissected into pieces and digested with 10 mL DMEM/F12 (Procell Life Science & Technology Co., Ltd.) containing 2.5% fetal calf serum (FCS) (Dcell Biologics), 5 μg/mL insulin (Beijing Solarbio Science & Technology Co., Ltd.), 50 μg/mL gentamicin (Beijing Solarbio Science & Technology Co., Ltd.), 2 mmol/L glutamine (Beijing Solarbio Science & Technology Co., Ltd.), and 0.01 mg collagenase A (Beijing Solarbio Science & Technology Co., Ltd.) for 3 h at 37 °C. Next, tissues were trypsinized with 0.05% Trypsin-EDTA for 5–10 min. The digested tissue was then filtered with a 70 μm cell strainer to prepare a single-cell suspension. The single-cell suspension was resuspended in 0.2% BSA in PBS, and the number of cells was adjusted to 2 × 10 6 /mL. Then, these cells were incubated with the following primary antibodies for analysis (4 °C for 30 min): CD29-FITC (Biolegend), CD24-APC (Biolegend), CD31-BV421 (Biolegend), CD45-BV421 (Biolegend), Ter119-BV421 (Biolegend), Sca1-BV711 (Biolegend), and CD49b-PE (Biolegend). Following washing and resuspension in PBS, the proportion of primary mouse mammary epithelial cells was analyzed using BD flow cytometry (LSRFortessa X-20). The gating strategy is shown in Figure S1 in the Supporting Information . Postanalysis of sorted cells and generation of typical images were conducted using the FlowJo software. Parts of the single-cell suspension of the fourth mammary gland without lymph node of female offspring at PND90 were suspended with 100 μL DMEM/F12 medium containing 5 μg/mL insulin (Beijing Solarbio Science & Technology Co., Ltd.), 1 μg/mL hydrocortisone (Topscience Co., Ltd.), 10 ng/mL mouse EGF (PeproTech, Inc.), 2 mmol/L glutamine (Beijing Solarbio Science & Technology Co., Ltd.), 50 μg/mL gentamicin (Beijing Solarbio Science & Technology Co., Ltd.), penicillin and streptomycin (Beijing Solarbio Science & Technology Co., Ltd.), and 10% FCS (Dcell Biologics) and maintained in low adhesion 8-well plates for 3 days. Then, the single-cell suspension grown on low adhesion 8-well plates was trypsinized with 0.05% Trypsin-EDTA (Beijing Solarbio Science & Technology Co., Ltd.) for 10 min, suspended with 200 μL liquid Matrigel (Corning), and plated onto 8-well chamber slides. Organoids were grown in 200 μL DMEM/F12 supplemented with Insulin-Transferrin-Selenium (ITS) medium supplement, penicillin, streptomycin and 2.5 nmol/L FGF-2 (PeproTech, Inc.). The medium was refreshed every 3 days. After 7 days of culture, the morphology of the organoids was observed by photographing. Detailed information can be seen in a previously published study. 37 Serum estrogen and progesterone (Pg) levels were determined using ELISA kits. 50 μL of mouse serum was used for each sample. Instructions detailed in the ELISA kit were followed (Shanghai Aimeng Youning Biotechnology Co., Ltd.). The detection limit for the progesterone assay was 1.56–100 ng/mL, and that for estrogen was 0.31–20 ng/mL. Three female offspring from different litters were sacrificed on PND21, and the hypothalamus was harvested. Then, the explants were transferred into individual chambers, submerged in MEM, and cultured in a static incubator containing a water-saturated atmosphere of 5% CO 2 at 37 °C. The incubation medium was collected and renewed every 7.5 min for 4 h. The GnRH concentration was determined following the ELISA kit instructions (Shanghai Aimeng Youning Biotechnology Co., Ltd.). The detection limit was 15.6–1000 pg/mL. Total RNA was extracted from the mammary glands dissected at different time points, and the purity and integrity of RNA were validated; RNA levels were quantified by a relative quantification method (the ΔΔCT method) as previously described in Text S1 . Primer sequences are listed in Table S1 . RNA-seq was performed by Shanghai Sinomics Corporation (Shanghai, China). Differentially expressed genes (DEGs) in the time-series patterns throughout different time points were screened using the maSigPro R package (1.64.0, R version 4.1.0, threshold: adjusted p -value <0.05). The edgeR package (3.30.3) was used to analyze DEGs ( p 1.5) between the BPB and vehicle control groups at different time points. Gene ontology (GO) enrichment analysis for DEGs were conducted using the clusterProfiler package (version: 4.2.0; threshold: adjusted p -value <0.05) in R (4.1.0). The MalaCards human disease database ( http://www.malacards.org/ ), an integrated compendium of annotated diseases mined from 68 data sources, was used for disease enrichment analysis. Set enrichment analysis via GeneAnalytics was performed by probing the overlap between genes associated with an entity in GeneCards and disease-related genes. 38 If the data are following a normal distribution, comparisons between two groups were performed by using a two-tailed Student’s t test. If not, Kruskal–Wallis nonparametric tests were used for comparing the differences between BPB and control. Differences were considered statistically significant at p < 0.05. RNA-seq were analyzed by using the maSigPro R package and the edgeR package. The data were analyzed using SPSS 23.0, and figures were generated using R (version 4.1.0), FlowJo 10, and GraphPad Prism 8.

Previous studies on the effects of perinatal EDC exposure on the mammary glands of mice have focused on estradiol, atrazine, dioxin, oxybenzone, ethinyl estradiol, and BPA. However, the doses used in these studies were relatively higher than environmental exposure concentration. For BPA, the reference dose concentration (RDC) is 0.05 mg/kg of bw/day, the no observed adverse effect level (NOAEL) is 5 mg/kg of bw/day, and the low observed adverse effect level (LOAEL) is 50 mg/kg bw/day. No data showed RDC, NOAEL, or LOAEL for BPB at present. Importantly, the maternal serum content of BPB was 0.1 ppb (unpublished data not shown), which was similar to those in humans, as previously indicated. The present study focused on the effects of BPB (300 μg/kg bw) at relatively lower doses on mammary gland development across critical time windows. During pregnancy, females are especially sensitive to estrogens, which control the development of the reproductive system and modulate neural circuits of the CNS, and consequently neuroendocrine, behavioral, and cognitive functions. The most direct effects of BPB exposure were on maternal weight and nursing behaviors. Both the present study and that of López-Rodríguez showed no differences in maternal weight between the treatment and control groups. 34 Previous studies have shown that exposure to BPA (40 μg/kg/day, orally) during pregnancy and lactation can affect maternal care, especially licking and grooming behaviors. 40 However, in the present study, BPB exposure did not affect maternal nursing behavior, and neither licking/grooming nor arched-back/passive nursing was affected. A similar study suggested that exposure to EDC mixture did not alter maternal behavior directly but rather through generations via a multigenerational nongenomic mechanism, affecting the hypothalamic response to stress. 34 Although maternal and nursing behaviors were not disturbed, the effects of BPB exposure on offspring cannot be ignored. Interest in pubertal alterations has been boosted by recent epidemiological reports of disturbed puberty onset in humans, which may be related to EDC exposure or food nutrition. 41 − 43 As early as 1998, Ashby and Tinwell suggested that subcutaneous injections of 600 and 800 mg/kg BPA could induce premature VO in rats. 44 Later, other groups also reported early VO age in rats treated with BPA perinatally 45 or postnatally. 46 , 47 The dosage chosen in the above studies was higher than that in environmental exposure. A recent study under environmentally relevant conditions of female mice perinatally exposed to BPA also showed an early vaginal opening. 48 In the present study, BPB did not induce premature VO in offspring mice significantly. In addition, the estrous cycle of female offspring also showed no obvious changes upon BPB exposure, although the duration of estrus lasted for 3 days in mouse BPB1 and BPB6. However, estrogen and progesterone, which play important roles in the regulation of VO, were downregulated at PND21, and progesterone levels were consistently suppressed at PND90. The decrease in serum estradiol levels after perinatal exposure to BPA has also been observed in rats. 49 In addition, recent experiments have shown that BPA disrupts granulose cell function in vitro , lowering the progesterone levels, 50 inhibits follicle growth and decreases hormone production in adult ovarian antral follicles. 51 Therefore, a reduction in serum estradiol and progesterone levels in female offspring after BPB exposure may be due to the inhibition of follicle growth of female offspring induced by BPB treatment. However, in Tucker’s study, estrogen and progesterone levels were elevated at PND20, and progesterone concentrations also tended to be depressed in bisphenol-exposed animals at 3 months after maternal bisphenols exposure. 31 In addition, estrogen and progesterone production were stimulated by gonadotropins from the anterior pituitary, which was triggered by GnRH secretion. 52 Interestingly, the GnRH interpulse interval was prolonged in offspring at puberty after maternal BPB exposure, which may result in decreased levels of estrogen and progesterone. In the pollutant-exposed group, maternal BPB exposure disrupted the hormonal homeostasis in female offspring at puberty; however, these changes are not enough to cause premature vaginal opening and changes in the estrus cycle. Hormonal homeostasis can influence mammary gland development in offspring, which may account for the many alterations and manifestations that arise in the gland later in life. We further examined whether maternal BPB exposure affected mammary gland development in offspring. As indicated, mammary gland branches in the BPB group showed no significant changes at birth or prepuberty. During the pubertal period, the number and area of TEBs did not change in the BPB group. In adult life, in utero BPB exposure also did not alter the length, width, or area of the mammary glands. However, studies on BPA, BPS, and BPAF have shown that early life exposure to these EDCs alters pubertal mammary gland development in CD-1 mice, which can increase susceptibility to tumors later in life. 31 One of the possible reasons for this difference between BPB and other BPA analogues is the different molecular structures, exposure dosages, and durations. The dosage in this study was 300 μg/kg, similar to that in Tucker’s study (500 μg/kg); BPA and BPAF elevated branching density but not BPS; BPS increased the number of TEBs, while none of them altered the mammary epithelial area at PND20. Further, 500 μg/kg BPA exposure did not change prenatal pubertal mammary gland development scores of female offspring at PND20, PND28, PND35, and PND56; the same dosage of BPA analogues only changed the scores at PND20 for BPAF and at PND35 and PND56 for BPS. In addition, the scores changed more significantly at PND20 and PND35 when the dosage was increased to 5000 μg/kg. Therefore, compared with BPA, BPAF, and BPS, BPB has a weaker impact on the development of mammary gland branches. Rodents are considered a low estrogen level model. Another study on monkeys, a high estrogen level model, closely resembling the human condition, suggested that 400 μg/kg BPA exposure increased the density of mammary buds and the overall development of mammary glands in monkeys of the BPA group as compared to the unexposed monkeys. 53 Therefore, although there was no obvious effect on the morphological development of the mammary glands of offspring mice, our current BPB exposure dose may cause damage to the mammary glands of high-hormone level model humans. Thus, it is necessary to examine the possible cellular and molecular changes during this developmental process. Ductal elongation during mammary gland development requires an accurate interplay between luminal and basal epithelial cells. In the present study, BPB exposure did not alter the luminal-to-basal epithelial cell ratio in female offspring. HR + luminal epithelial cells are considered essential mediators of the interplay between luminal and basal cells by sensing hormones and producing paracrine signals for neighboring HR – luminal cells. 54 Interestingly, maternal BPB exposure increased the number of HR + luminal cells at PND90. As early as the 21st century, Iguchi and colleagues intended to screen estrogen-responsive genes in the mouse reproductive system; these data are useful for elucidating the mechanisms of persistent changes in the reproductive system induced by perinatal estrogen exposure. 55 Previous in vivo and in vitro studies have indicated the important roles of PR and ER in mammary gland injuries and breast cancer induced by BP exposure. 27 , 56 , 57 RNA-seq results also suggested that Pgr expression remained after birth and was elevated in mammary glands under stimulation of the estrous cycle at PND90 in the BPB group. The reason Pgr changed only at PND90 was that periodic stimulus of the estrous cycle and in utero exposure to BPB increased luminal cell sensitivity in the mammary gland in offspring to Pg, although the expression of Pg was relatively lower than that of the vehicle control. However, according to the RNA-seq results, the expression of Esr1 did not change at any time point. In addition, combined time-series analysis and pairwise comparison based on every single sensitive window screened out other genes related to mammary gland development and hormone homeostasis such as Egfr (PND90) and Vdr (PND1). EGFR signaling is mediated by amphiregulin (Areg), a downstream mediator of estrogen and progesterone, and is both necessary and sufficient for sustaining mammary epithelial cell proliferation. 58 Areg in the current study showed a significant time-dependent increase in the BPB group relative to the vehicle control group by time-series analysis, although the expression remained unchanged relative to a single time point ( Figure S14 ). Mammary glands from VDR-KO mice exhibit accelerated growth and branching morphogenesis compared to glands from age- and weight-matched WT mice. 59 However, our findings suggest that in utero BPB exposure increased the expression of Vdr at PND1, which did not change significantly at the following developmental time points. In addition, endocrine pathway analysis suggested that mammary glands of offspring were sensitive to estrogen and PPAR at PND1, thyroid hormone at PND10, growth hormone and thyroid hormone at PND21, and endocrine signaling pathway at PND90 after in utero BPB exposure; PND90 was the most sensitive time point to the endocrine pathway, which may be a cumulative result. In addition, the present study pays more attention to the health of adult offspring about whether these changes in endocrine signals induced by maternal BPB exposure will affect the health of adults, such as the occurrence of breast cancer. Although no cancer foci were observed through histopathological changes, maternal BPB exposure indeed induces inflammatory cell infiltration and lobular hyperplasia in the mammary gland of offspring at PND90 with no significant changes at PND10 and PND21 ( Figure S15 ), which may affect the long-term health of the breast. Further, breast cancer related genes such as Gata3 , Egfr , and Fgfr2 , which regulate epithelial–mesenchymal transition, migration, and tumor invasion, also showed elevated trends at PND90 after BPB exposure, which further confirmed the long-term adverse effects of maternal BPB exposure on the breast health of female offspring. In conclusion, maternal exposure to 300 μg/kg of BPB disrupted endocrine homeostasis of female offspring and endocrine pathways related to nuclear receptors, especially PR , which played an important role in mammary gland dilatations, lobular hyperplasia, and inflammatory cell infiltration in adult life. Future studies, including single-cell approaches, should be used to characterize whether the transgenerational effects of EDC mixture target specific PR-positive cellular populations.

By the end of 2020, 7.8 million women were diagnosed with breast cancer in the past 5 years; among them, those in 2020 alone accounted for nearly 30% of the total, making it the world’s most prevalent cancer. 1 The risk of breast cancer is mostly influenced by nongenetics, and over 70% of these cases are noninherited or sporadic. 2 Recent studies have suggested that breast cancer risk is related to abnormal breast development and is thought to be modified by the environment and lifestyle. There are three critical windows for mammary gland growth: first is the prenatal period of mammary epithelial sprout development, second is the peripubertal period of expansive proliferation, and third is gland maturation and alveologenesis during pregnancy and lactation. As a female reproductive organ, many hormones can influence the development of mammary glands by regulating ductal growth, elongation, branching, and differentiation. For endocrine disruptive compounds (EDCs), a class of exogenous hormones, their exposure during these developmental windows of mammary gland is suspected to increase the risk of breast cancer. Bisphenol A (BPA), a common EDC, is widely used in certain food-contact materials, including polycarbonate plastics, epoxy resins, and thermal papers and was first approved by the U.S. Food and Drug Administration (FDA) during past years. However, representative research on BPA exposure levels in populations in the United States, the 2003–2004 National Health and Nutrition Examination Survey (NHANES III), found a relatively high detection rate of BPA (93%) in 2517 human urine samples. The adverse effects of BPA on reproductive, metabolic, and child development have led regulatory bodies worldwide to ban BPA from baby bottles over the last decades. 3 − 6 Therefore, regulatory policies and public concern have stimulated the production of alternative substances to replace BPA, and various bisphenol analogues have been used for industrial applications. BPB differs from BPA only by a single methyl group on the central carbon and is currently registered as an indirect food additive by the FDA for use in food-contact resins and polymer coatings. Recently, BPB has been detected in food samples, including beverages, dairy products, seafood, cereals, meat, and fruits. 7 Monitoring survey on common population indicated that daily dietary intakes of BPB for adults in the United States reached 0.35 ng/kg. 8 BPB is also detected in thermal paper samples and personal care products. 9 Among environmental residuals, BPB is one of the most poorly investigated and detected bisphenols, likely suggesting its specialized applications and localized emissions. Limited studies indicate that BPB is present in surface water of the Pearl River (7.6 ng/L), 10 10.9% of indoor dust samples in China, 11 sedimentary microplastics in Hong Kong, 12 and sewage sludge samples collected in the United States. 13 Due to the low concentration of BPB in different environmental media, few studies have focused on the internal exposure of BPB in humans. The mean concentrations of serum BPB were 0.765 and 0.255 ng/mL in people working in dense industrial areas and pregnant women, respectively. 14 The detection rates of BPB and conjugated BPB were 89.0% and 10% in the serum of pregnant women and human urine samples, respectively. 15 , 16 However, BPB has not been detected in infant serum in China. 17 Although the available data of BPB are limited, BPB meets the definition of an endocrine disrupting chemical by WHO. On May 5, 2011, BPB was added to the The Endocrine Disruption Exchange (TEDX) list of potential “Endocrine Disruptors”. Recent studies have also revealed that BPB exposure is related to disturbed male and female reproductive functions, 18 − 20 hepatic lipid accumulation, 21 and sex-specific neurotoxic effects. 22 The most hazardous effects of BPB are its interaction with the endocrine system by modulating estrogenic, androgenic, steroidogenesis-related, thyroid hormone, and other endocrine activities. 23 Some studies have suggested that the cytotoxicity and binding affinity of BPB with the hormone receptor such as estrogen receptor alpha are stronger than those of BPA. The IC50 at 48 h on Ctenopharyngodon idella kidney cells was 72.44 μmol/L for BPB, and the rank order of cytotoxicity represented by the half inhibition concentration is BPAF > BPB > BPA > BPF > BPS. 24 The results of membrane affinity data also confirm that BPB results in more toxic effects than BPA 25 and displays a 9-fold binding affinity with G protein-coupled estrogen receptor relative to BPA. 26 Importantly, BPB strongly activates estrogen receptor alpha in human breast cancer cells. 27 In vivo studies show that BPB exposure during pregnancy is related to the endometriosis 28 and enhances the risk of anemia during pregnancy. 15 In mammary glands, low-dose (25 μg/kg/day) gestational or perinatal BPA exposure induces a stimulation of mammary growth, whereas higher doses (250 μg/kg/day) exerts an opposite effect. 29 , 30 Female mice prenatally exposed to BPAF and BPS develop proliferative epithelial lesions by midlife, concomitant with a significant inflammatory response that may predispose them to tumor formation later in life. 31 A previous study suggested that prenatal exposure to BPAF or BPS induced precocious development of the mammary gland, with an incidence greater than BPA-exposed mice. 31 An increased prevalence of intraductal hyperplasia was observed in BPA females exposed to 250 μg/kg at postnatal day (PND) 400 and not at higher or lower doses. 32 Females exposed to 200 μg BPS/kg/day had significantly more alveolar buds at adulthood compared to controls. 33 Although these results may lack direct comparability, they provide sufficient evidence to suggest that BPA alternatives have similar or potentially even stronger detrimental effects on mammary gland development. Considering the endocrine disruption effects of BPB, it is reasonable to conduct further research on whether fetal exposure to BPB can disturb mammary gland development at different stages in female offspring and further examine its long-term effects on breast health. In this study, in order to examine whether maternal exposure to BPB at relatively low concentrations could alter postnatal and adult mammary gland development of offspring, pregnant mice were orally exposed to corn oil or 300 μg/kg body weight (bw) BPB throughout gestation, and the mammary glands of female offspring at different time points were harvested to evaluate morphological, pathological, and molecular time-series changes under critical developmental stages.