Di‐(2‐ethylhexyl) phthalate exposure induces female reproductive toxicity and alters the intestinal microbiota community structure and fecal metabolite profile in mice

Abstract Di‐(2‐ethylhexyl) phthalate (DEHP) is one of the most commonly used plasticizers, and it is widely applied in various plastic products. DEHP is an endocrine‐disrupting chemical (EDC) that has been shown to disrupt the function of reproductive system in females. Although many studies have shown that DEHP potentially causes female reproductive toxicity, including depletion of the primordial follicle and decreased sex hormone production, the specific mechanisms by which DEHP affects female reproduction remain unknown. In recent years, research focused on the intestinal flora has provided an idea to eliminate our confusion, and gut bacterial dysbiosis may contribute to female reproductive toxicity. In the present study, the feces of DEHP‐exposed mice were collected and analyzed using 16S rRNA amplicon sequencing and untargeted global metabolite profiling of metabolomics. DEHP obviously causes reproductive toxicity, including the ovarian organ coefficient, estradiol level, histological features of the ovary and estrus. Furthermore, DEHP exposure alters the structure of the intestinal microbiota community and fecal metabolite profile in mice, suggesting that the reproductive toxicity may be caused by gut bacterial dysbiosis and altered metabolites, such as changes in the levels of short‐chain fatty acid (SCFA). Additionally, it is well known that changes in gut microbiota and fecal metabolites cause inflammation and tissue oxidative stress, expectedly, we found oxidative stress in the ovary and systemic inflammation in DEHP exposed mice. Thus, based on our findings, DEHP exposure may cause gut bacterial dysbiosis and altered metabolite profiles, particularly SCFA profiles, leading to oxidative stress in the ovary and systemic inflammation to ultimately induce female reproductive toxicity.


| INTRODUCTION
Di-(2-ethylhexyl) phthalate (DEHP) is one of the most commonly used plasticizers, and it is widely used in the production of various cosmetics, personal care products, food storage containers, pharmaceuticals, building materials, children's toys, medical tubing, and other polyvinyl chloride products. 1-3 DEHP is continuously released from plastic products and directly infiltrates food, water and air; thus, humans are exposed to DEHP daily via ingestion, inhalation, and dermal absorption. 2,[4][5][6] In recent years, DEHP and its active metabolic product, mono (2-ethylhexyl) phthalate (MEHP), have been successively detected in many tissues, including the liver, blood, cord blood, breast milk, placenta, amniotic fluid, and early gestation villi. [7][8][9] According to previous studies, DEHP is associated with hepatic, renal and neural injures or diseases. [10][11][12] In addition, DEHP is an endocrine-disrupting chemical (EDC) that has been shown to disrupt the function of reproductive system in both females and males. 13,14 Several studies have confirmed that the obvious toxicity of DEHP is mediated by its effects on gonadal steroidogenesis and an accompanying decrease in reproductive function and fertility. 13,15,16 The average exposure of DEHP in humans ranges from 3 to 30 mg/kg (body weight)/d, and occupational exposure levels have been calculated to even reach up to 300 to 600 mg/kg/d. 17,18 Based on epidemiological data, decreased conception rates, increased miscarriage rates, decreased estrogen levels, and abnormal ovulation are associated with the occupational exposure of female workers to DEHP. 19,20 Thus, the health risks of DEHP exposure in humans have attracted increasing attention.
The ovary is the primary reproductive organ in the female, and it regulates female endocrinology and provides a microenvironment for follicle development. 21 DEHP and MEHP have been shown to induce a depletion of the primordial follicles and a decrease in sex hormone production. [22][23][24] However, the precise mechanisms by which DEHP affects female reproduction remain unclear. The human gastrointestinal tract contains 10 13 -10 14 microbiota that consist of 1500 microbial species and are characterized by more than 3 million genotypes, and numerous studies have shown that the intestinal microbiota is inextricably linked to human health. 25,26 Approximately 1000 bacterial species and 7000 bacterial strains have been identified using the currently available advanced sequencing technology. 27 Recent studies have revealed the bidirectional relationship between the estrogen level and gut microbiota in females with diseases induced by abnormal estrogen levels; estrogens are potentially regulated by the gut microbiota through secretion of β-glucuronidase, which deconjugates estrogens into active forms, and estrogens regulate the gut microbiota through immunoregulation. [28][29][30] In addition, gut microbiota dysbiosis may increase intestinal permeability and alter the levels of host gut metabolites, allowing bacterial endotoxins, such as lipopolysaccharide, to be transported into the circulation and activate the inflammatory response to promote disease development in females. 31,32 Therefore, we hypothesized that women of child-bearing age who are exposed to DEHP would exhibit abnormal follicular development and infertility that may be mediated by alterations in the gut bacterial composition and metabolite profiles.
In the present study, we aimed to determine the effects of DEHP exposure on ovarian damage and homeostasis of the gut microbiome and fecal metabolome, and to explore the relationship between DEHP-induced reproductive disease and the gut microbiome or fecal metabolome in an effort to provide deeper insights into the etiology of reproductive toxicity. For the precise analysis, mice were exposed to different concentrations of DEHP for 30 days, and then the effects of DEHP exposure on reproductive toxicity, including ovarian pathology and serum hormone levels, were detected, and fecal microbiomemetabolome responses were evaluated using 16S rRNA pyrosequencing and nontargeted metabolomics. Together with the dual omics approach, we attempted to identify differences in strains and metabolites between DEHP-exposed and normal mice, as the differences in strains and metabolites can be used as biomarkers for the diagnosis or auxiliary diagnosis of DEHP-induced reproductive toxicity. More importantly, our study will more accurately describe the intestinal microbial and fecal metabolic states associated with reproductive toxicity and enable a complete understanding of the mechanisms of the gut microbiota and host health that may facilitate further studies and provide a valuable reference for human fecal microbiota transplantation to treat metabolic reproductive diseases in the future.

| DEHP exposure and sample collection
DEHP (99% purity) was purchased from Sigma-Aldrich, and diluted in edible corn oil (Jinlong, China) to prepare stock and working solutions.
The doses administered in this study were 500 and 1500 mg/kg body weight per day. The doses were primarily selected based on the equivalent dose ratio calculated using the surface areas of humans and mice. Thirty female mice were randomly and divided into three groups, and animals in the control group were administered an equal volume of corn oil by oral gavage. All animals were intragastrically administered DEHP (mixed with corn oil) for 30 days and doses were adjusted daily according to changes in body weight. The doses of DEHP were based on previous studies. 14,33,34 30 days later, mice in each group was anesthetized by intraperitoneally injecting 0.3 mL/100 g of body weight of a 10% chloral hydrate (Sigma, Missouri) solution, and the ovaries and venous blood were collected. The ovarian organ coefficient was evaluated by the ratio of two ovary weight to body weight.

| Analysis of estrous cyclicity
During DEHP exposure, vaginal smears were collected from the mice each day to monitor estrous cyclicity. The stage of estrus was determined by performing vaginal cytology using a light microscope and was recorded based on previously defined and well-documented criteria. 35 The percentage of days in estrus was calculated by dividing the number of days in estrus by the number of days in the study and multiplying that value by 100. The percentage of days in metestrus/ diestrus was calculated by dividing the number of days in either metestrus or diestrus by the number of days in the study and multiplying that value by 100. The evaluation of estrous cyclicity was repeated 3 times.

| Hematoxylin and eosin staining of ovarian tissues
Ovarian tissue samples were collected from the euthanized mice, fixed with 4% paraformaldehyde for 24 h, and subsequently processed for histological staining. Tissues were then dehydrated using increasing concentrations of alcohol ranging from 60% to absolute alcohol, cleared and infiltrated with xylene and embedded in paraffin. The paraffin blocks of ovaries were sectioned into 5-mm slices and sequentially stained with a hematoxylin solution for 5 min and an eosin solution for 5 min. All reagents were purchased from Solarbio Science and Technology Ltd. (Beijing, China). The slides were viewed and images were captured under an Olympus microscope and camera system. The whole ovary was continuously sliced into numerous sections that were placed on glass slides in order, and the numbers of whole follicles and atretic follicles were observed under a microscope and recorded from every five sections to compare the numbers of atretic follicles between groups.

| Measurement of estradiol levels
Blood was collected from euthanized mice and plasma was separated by centrifugation at 3000 rpm for 15 min within 1 h after blood collection and stored at −80 C until analysis. Estradiol levels were measured in plasma samples using an enzyme-linked immunosorbent assay (ELISA) kit (Beyond Biotechnology, Shanghai, China) according to the manufacturer's instructions. Briefly, the estradiol standard and the plasma samples were added to the 96-well plate provided with the kit, horseradish peroxidase-labeled estradiol was added to the wells containing the samples to competitively bind the estradiol antibody labeled on the plate and then the free estradiol was removed by washing the plate. After coloration and termination, the absorbance of each well in the plate was detected using a multimode microplate reader (Thermo Scientific) at 450 nm.
Finally, the estradiol levels in the samples were calculated from the standard curve.

| Fecal sample collection and DNA extraction
Fresh fecal samples were collected from control, 500 and 1500 mg/kg (fecal samples from random six animals in each group) DEHP-exposed groups, placed in sterile tubes in an ice bath, and immediately transferred to the laboratory. All fecal samples were stored at −80 C.
Microbial genomic DNA was isolated and purified from each fecal sample using a DNA Extraction Kit (Omega Bio-tek, Norcross) according to the manufacturer's instruction. The DNA integrity and concentrations were determined using a Nanodrop2000 spectrophotometer (Thermo Fisher Scientific) and Qubit R 2.0 Fluorometer (Life Technologies, CA).

| 16S rDNA sequencing and bioinformatics analysis
Polymerase chain reaction (PCR) was used to amplify the V3-V4 region of the 16S rDNA with specific primers. Equimolar concentrations of PCR products were purified, quantified and sequenced using the Illumina Hiseq2500 platform according to the standard protocols provided by Novogene Technology Co. Ltd. (Beijing, China). Pairedend raw reads were merged from the original DNA fragments with FLASH (version 1.2.7). After qualified filtering with Trimmomatic (version 0.33), clean reads were extracted and effective tags were produced by removing the chimeric sequences using UCHIME (version 4.2). Finally, high-quality effective tags were evaluated according to Q20 and Q30 bases and used for bioinformatics analysis.
Using Uparse, all effective reads were clustered into operational taxonomic units (OTUs) based on 97% sequence similarity using QIIME (version 1.

| Measurement of SOD and MDA levels
The levels of superoxide dismutase (SOD) in the ovaries of the 500 and 1500 mg/kg DEHP-exposed and the control groups were quantified using a Total Superoxide Dismutase Assay Kit with NBT (Beyotime, Shanghai, China) according to the manufacturer's instructions. Superoxide anion radicals that are not scavenged by SOD reduce NBT to a blue formazan product. The absorbance of the blue formazan product in each sample was detected with a multimode microplate reader at 450 nm. Malondialdehyde (MDA) levels in ovarian tissue extract were measured using a Lipid Peroxidation MDA Assay Kit (Beyotime, Shanghai, China), which is based on the ability of MDA to react with thiobarbituric acid to generate a red product. The absorbance of each sample was determined using a multimode microplate reader (Thermo Scientific) at 535 nm.

| Measurement of proinflammatory factors
IL-1β and TNF-α were measured in plasma samples using an enzymelinked immunosorbent assay (ELISA) kit (Jonln Biotech., Shanghai, China) according to the manufacturer's instructions. Briefly, the standard of IL-1β, TNF-α and the plasma samples were added to the 96-well plate provided with the kits, horseradish peroxidase-labeled IL-1β and TNF-α were added to the wells containing the samples to competitively bind the IL-1β and TNF-α antibody labeled on the plate and then the free IL-1β and TNF-α were removed by washing the plate. After coloration and termination, the absorbance of each well in the plate was detected using a multimode microplate reader (Thermo Scientific) at 450 nm. Finally, the levels of IL-1β and TNF-α in the samples were calculated from the standard curve.

| Statistical Analysis
The differences among groups are presented as means ± SD and were analyzed using one-way ANOVA with GraphPad Prism 5.0 software. If the variance was equal, the Student-Newman-Keuls (SNK) test was used for the pairwise comparison; otherwise, the Games-Howell test was used. p < .05 was considered a significant difference in all analyses. ANOVA was performed to identify significant differences in the abundances of intestinal bacteria and metabolites.
Pearson's correlation coefficients were calculated to recognize the correlations between perturbed intestinal microbiota and changes in fecal metabolites.

| Reproductive toxicity was induced by DEHP exposure
During the 30 days of DEHP exposure, the control, 500 and 1500 mg/kg DEHP groups exhibited similar levels of body weight gain, and thus continuous DEHP exposure for 30 days did not alter the body weight of mice ( Figure 1(A)). Significantly decreased ovary organ coefficient (Figure 1(B)) and estradiol levels ( Figure 1(C)) were also observed in all DEHP-treated groups. Significant differences in follicle morphology were observed in histological examinations of the ovarian tissues from mice exposed to DEHP under a light microscope.
The numbers of primary, secondary and antral follicles were decreased, and the numbers of oocytes in the primary follicles were obviously decreased by the 500 and 1500 mg/kg DEHP treatments.
In the groups treated with DEHP, oocyte loss occurred in the primary follicles, along with a loose structure and detachment of granular cells, wide intercellular spaces between granulosa cells and theca cells  3.2 | Gut microbiota dysbiosis was induced by DEHP exposure The mouse fecal microbiota was characterized by performing 16S rDNA sequencing. Alpha diversity was used as a measure of the complexity of the species diversity and richness by calculating several indices, including Chao1, and Shannon and Simpson indices. The Chao1 histogram showed the species number determined in each sample.
The flattening of the rarefaction curve based on the values of OTUs and Shannon indices indicated that our data volume covered all species of the community in the samples. Moreover, we revealed both increases and decreases in different gut bacterial species in the control, 500 and 1500 mg/kg DEHP-exposed groups ( Figure S1), respectively.
By examining the unweighted UniFrac distance, 500 and 1500 mg/kg DEHP-exposed mice and controls were separated on the PCoA plot ( Figure 2(A)) and NMDS (Figure 2(B)). PCoA and NMDS analyses of the relative abundance of different bacterial taxa indicated a considerable separation among these three groups, suggesting a change in the structure of the bacterial community in DEHP-exposed mice. The gut microbiota in the control, 500 and 1500 mg/kg DEHPexposed groups were dominated by five phyla: Firmicutes, Bacteroidetes, Verrucomicrobia, Actinobacteria, and Epsilonbacteraeota ( Figure 2(C)). The relative abundance of Firmicutes was increased, and the relative abundances of Bacteroidetes, Actinobacteria, and Epsilonbacteraeota were decreased in the DEHP-exposed groups compared to the control group. Interestingly, a higher abundance of Verrucomicrobia was detected and this phylum was an important component of the bacterial community in 1500 mg/kg DEHP-exposed group (Figure 2(C)). To further identify the specific bacterial species as the biomarkers after DEHP exposing, the abundances Akkermansia and Bacteroidaceae (Figure 2(I)) were decreased in mice exposed to 500 and 1500 mg/kg DEHP. The taxonomic distributions of the fecal microbiota showed significant differences at the class, family, and genus levels in mice exposed to 500 and 1500 mg/kg DEHP, as shown in Figure S2. in the 500 mg/kg DEHP-exposed group were observed ( Figure S3C and S3D). These results revealed a significant difference in the gut microbiota between the control and DEHP-exposed groups. Here, we particularly considered obvious differences in the gut microbiota after DEHP exposure.

| Effects of DEHP exposure on fecal metabolite profiles in mice
Using an untargeted strategy, the fecal metabolome associated with functional characteristics of the gut microbiome was studied. 1500 mg/kg DEHP-exposed group was valid, and comparisons between the control and 500 mg/kg DEHP group ( Figure S4A) and between the 500 and 1500 mg/kg DEHP groups were also valid ( Figure S4B). Significantly altered metabolites were identified based F I G U R E 2 Comparison of the fecal microbiota structures and distributions among control, 500 and 1500 mg/kg DEHP exposed groups. 1500 mg/kg DEHP-exposed groups, 165 metabolites were downregulated and 65 were upregulated, which were visualized in a volcano plot in Figure 3(D). In the comparison of the control and 500 mg/kg DEHP-exposed groups, 103 metabolites were downregulated and 99 were upregulated ( Figure S4C). In the comparison of the 500 and 1500 mg/kg DEHP-exposed groups, 179 metabolites were downregulated and 30 were upregulated ( Figure S4D). The numbers of differentially altered metabolites in the control, 500 and 1500 mg/kg DEHP groups are presented in Figure 3E.
A heat map of three groups was constructed to visualize the results of the major abundant metabolites and summarize the distributions of the most significantly differentially altered metabolites distinguishing the three groups ( Figure 4A). In addition, the differentially altered metabolites are listed in Table 1, and the levels of those metabolites were significantly different in the 500 and 1500 mg/kg DEHP groups compared with the control groups. These metabolites were products of carbohydrate metabolism, protein digestion and absorption, and fatty acid metabolism. The enrichment of the differentially abundant metabolites between the control and 1500 mg/kg DEHP-exposed groups was determined by analyzing Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and the differentially abundant metabolites were related to some metabolic pathways. In the comparison of the control and 500 mg/kg DEHP-exposed groups, tyrosine metabolism, ubiquinone and other terpenoid-quinone biosynthesis, amino sugar and nucleotide sugar metabolism, histidine metabolism, phenylalanine, tyrosine and tryptophan biosynthesis, and synthesis and degradation of ketone bodies were the affected pathways ( Figure 4(B)). In the comparison of the control and 1500 mg/kg DEHPexposed groups, phenylalanine metabolism, steroid biosynthesis, amino sugar and nucleotide sugar metabolism, purine metabolism, pyrimidine metabolism, and riboflavin metabolism were affected (Figure 4(C)). In the comparison of the 500 and 1500 mg/kg DEHP-exposed groups, pyrimidine metabolism, riboflavin metabolism, pyruvate metabolism, porphyrin and chlorophyll metabolism, and amino acid (valine, leucine and isoleucine) biosynthesis were affected ( Figure S5). In addition, the differentially altered metabolites enriched in these signaling pathways are listed in Table S1.

| Correlation between the Fecal Microbiota and Metabolites
Spearman's correlation analysis was conducted in this study to further explore the functional correlations between the composition of the

| Oxidative stress in the ovary and systemic inflammation
We detected the indicators of oxidative stress MDA and SOD in ovarian tissues and proinflammatory factors in the blood of DEHP-exposed mice to elucidate the mechanism by which the DEHP-induced changes in the fecal microbiota and metabolites cause female reproductive toxicity. The MDA concentration was increased ( Figure 6(A)) and the SOD concentration was decreased (Figure 6(B)) in the DEHP-exposed groups compared with the control group. The analysis of systemic proinflammatory factors showed that IL-1β and TNF-α levels were increased in mice after DEHP exposure (Figure 6(C) and (D)).

| DISCUSSION
DEHP is one of the most environmentally abundant endocrinedisrupting chemicals that cause reproductive abnormalities in humans, T A B L E 1 The differential metabolites in groups of 500 and 1500 mg/kg DEHP compared with the control Denotes p < .01 between DEHP exposed and control, and, b Denotes p < .05 between DEHP exposed and control.
and DEHP can leach out of plastic beverage or food containers and readily enter the blood following oral ingestion. DEHP disrupts normal reproductive and ovarian function, 22 alters follicular development during weaning and maturity, 38 impairs the steroidogenesis of ovarian follicular cells, 39 and induces premature ovarian failure. 40 In addition, prenatal exposure to DEHP exerts multigenerational and transgenerational effects on female reproduction. 17 In the present study, DEHP expo- Atm genes, and leads to abnormal follicular growth and cell division. 43 Additionally, DEHP induces oxidative stress by promoting ROS generation and inhibits steroid synthesis by modulating the expression of steroidogenic responsive genes in granulosa cells; it also activates the Bax/Bcl-2 and the caspase-3-mediated mitochondrial apoptotic pathway to induce apoptosis. 44 Thus, DEHP induces oxidative stress in ovary and alters ovarian function to induce female reproductive toxicity. However, the specific mechanisms underlying the toxic effects of DEHP are an area that is largely unexplored.
Based on accumulating evidence, the gut microbiota plays an important role in maintaining animal health, and an imbalance in the intestinal flora is known to be associated with different metabolic diseases. The gut microbiome modulates the estrogen level in the host by secreting β-glucuronidase, an enzyme that suppresses the binding of estrogen to its receptors and its subsequent physiological downstream effects. 29,45 The composition of the gut microbiota was recently shown to play an important role in sexual dimorphism or sex-specific differences mediated by estrogen, and estradiol and estrogenlike compounds induce changes in the gut microbiome to improve the sexual dimorphism of subjects with metabolic syndrome. 46 Females are more resistant to gut injury compared to their male counterparts, and gut injury is decreased in male mice after androgen inhibition; therefore, the authors postulated that the gut epithelial barrier integrity was modified by estrogen. 47 Baker et al. published a detailed review showing that the gut microbiome is the principal regulator of circulating estrogen levels and described the relationship between gut microbes and estrogen-modulated diseases. 28 According to these studies, homeostasis of the gut microbiome is closely related to normal estrogen levels; however, the causal relationship between these factors requires further study. Meanwhile, the role of estrogens in female reproductive system development and maintenance is well defined. 48 In the present study, DEHP did not change the body weight that similar with a study of DEHP (750 mg/kg/d for 30 days) affected female reproduction, while DEHP (300 mg/kg/d for 4 weeks) could increase obesity-induced damage to the male reproductive system in mice, 22,49 in addition, a study showed males and females respond differently to DEHP not only in age but also in non-monotonic doseresponse curve, 50  The human body is reported to be exposed to DEHP at concentrations of up to 30 mg/kg/day, and occupational exposure levels of up to 600 mg/kg/d have been reported. 15,18 The doses to which the mice in present study were exposed were primarily selected based on the equivalent dose ratio calculated using the surface areas of humans and mice, and this exposure dose is consistent with human exposure in toxicological studies and might enable us to better assess the toxicological health effects. In the present study, the concentrations of DEHP to which mice were exposed were 500 and 1500 mg/kg.
Although the a 3-fold difference existed between the these two concentrations, the toxic effects, including estrogen levels, organ ratio is specifically increased in obese individuals. 53,54 In addition, obesity correlated with a reduction in estrogen levels in a previous study. 55 A decrease in the abundance of beneficial bacteria and an increase in the abundance of pathogenic populations induce inflammation and are closely relate to the absence of estrogen. 28,56 A study showed that the F/B ratio and Escherichia coli abundance were higher in ovariectomized animals than normal females. 57 Estrogen not only could increased abundance of Akkermansia and Bifidobacterium but also could inhibit the overgrowth of Proteobacteria and Escherichia coli. 58 These studies similar with our studies indicated that the bidirectionality effects between the reproductive system and gut microbiota, however, the chronological relationship between estrogen and gut microbiota caused by DEHP still needs to be further proved.
In the present study, the abundances of Verrucomicrobia and Akkermansia were significantly increased in 1500 mg/kg DEHPexposed mice. Akkermansia is a member of the phylum Verrucomicrobia that is involved in the process of lipid metabolism. In previous studies, the abundance of Verrucomicrobia was significantly increased in monkeys with alcoholic fatty liver and nonobese mice fed a high-fat diet. 59,60 In contrast, the abundance of Verrucomicrobia was significantly decreased in obese mice and patients with non-alcoholic fatty liver cirrhosis. 61 In the present study, the abundances of Helicobacter, Campylobacter, Lachnospiraceae and Erysipelotrichia were increased in DEHP-exposed groups, according to the LEfSe analysis and LDA score. Helicobacter is a genus of gram-negative bacteria and its major cell wall component lipopolysaccharide (LPS) induces systemic effects and even modifies lipid profiles, stimulates vascular inflammation, and exacerbates atherogenesis. 71 Additionally, the most widely known species Helicobacter pylori is strongly associated with gastric cancer in Asia. 72 Campylobacter is a gram-negative bacteria that induces diarrhea. 73 Members of the family Lachnospiraceae, such as Ruminococcus gnavus, express superantigen and activate the IgA response, which play a critical role in intestinal homeostasis. 74 In addition, Ruminococcaceae and Lachnospiraceae participate in protecting against Clostridium difficile infection. 75 The loss of protective gut commensal strains of the family Lachnospiraceae and an increase in the abundance of colitogenic strains of the family Erysipelotrichaceae in Nlrp12-deficient mice causes colonic inflammation. 76 Here, alterations in the bacterial composition of the gut represent one possible explanation for DEHP-induced female reproductive toxicity and systemic inflammation due to increased LPS levels; however, in-depth research must be con-  of Ruminiclostridium is closely related to the production of SCFAs and inhibition of the endotoxin LPS, and Ruminiclostridium is one of the dominant genera responsible for volatile fatty acid production. 33,79 The SCFA-producing bacteria of the gut microbiota, such as Bacteroides, Lactobacillus, and Lachnospiraceae, show negative correlations with fecal LPS concentrations. 80 An increase in the abundance of SCFA-producing bacteria might reduce inflammation. 43 The concen-  86,87 In addition, a change in the F/B ratio is directly related to SCFA metabolism. 88 Butyrate plays an important role in improving colonic defense barriers by increasing the expression of tight junction proteins, and the DEHP-induced increase in the levels of inflammatory factors may be partially due to the reduction in butyrate and subsequent increase in colonocyte permeability. 89 A study showed the changes of fecal microbiota composition and metabolites are related to the plasma reproductive hormones during pregnant and lactating stages in Bama mini pigs model. 90 The SCFA not only regulates the estradiol secretion but also participates in endogenous estrogen receptor-alpha-mediated signaling. 91,92 These result implied that the SCFAs potentially play an important role in DEHP-induced female reproductive toxicology. Overall, the reduction in the SCFA levels after DEHP exposure may be due to gut bacterial dysbiosis and alterations in gut permeability. In a study reporting that perinatal bisphenol A exposure induces chronic inflammation in rabbit offspring, the authors found that females exhibited more severe inflammation than males due to the overexpression of estrogen receptor-2. 93 In addition, both bisphenol A and DEHP have estrogen interference effect in reproductive toxicity.
Many fecal metabolites generated by the gut microbiota, particularly benzene homologs and their derivatives, induce a systemic inflammatory response and oxidative stress. 94 Phenylacetic acid is related to a more pro-oxidant and immune-stimulated status, which are both negatively associated with fecal propionate levels, whereas phenylpropionic acid is directly related to the fecal acetate level. 95 Leucine, isoleucine and valine are essential amino acids termed branched-chain amino acids due to the presence of an aliphatic sidechain, and high concentrations of branched-chain amino acids potentially exert deleterious effects and induce a pro-inflammatory and oxidative stress status. 96 We detected oxidative stress and systemic inflammation in ovarian tissues to clarify the mechanism of DEHP and involving the intestinal microbiota and fecal metabolites in female reproductive toxicity. As expected, oxidative stress and inflammation were increased after DEHP exposure. Thus, oxidative stress in the ovary and systemic inflammation may be the main factors contributing to female reproductive toxicity, which may be caused by changes in the intestinal flora and fecal metabolites following DEHP exposure.
Therefore, the levels of the amino acids leucine, isoleucine and valine were significantly altered by DEHP exposure, according to feces metabolomic profile, and these amino acids may be involved in systemic inflammation and oxidative stress. Moreover, DEHP-induced female reproductive toxicity and ovarian damage might be due to increased levels of oxidative stress and inflammation caused by metabolites from feces. Although the levels of many potential metabolic markers were increased after DEHP exposure at the doses used in this study, further studies are required to determine whether these metabolites are directly related to the inflammatory response and oxidative stress.

CONFLICT OF INTEREST
The authors declare no potential conflict of interest.