Epigallocatechin‐3‐O‐gallate modulates the diversity of gut microbiota in ovariectomized rats

Abstract Epigallocatechin‐3‐O‐gallate (EGCG) exists as one of the major active components of green tea and has been studied extensively; however, the relationship between EGCG and the changes in the gut microflora of ovariectomized (OVX) rats as a model of menopause women have not yet been studied. Female Wistar rats were fed on a maintenance material diet and underwent either ovariectomy or SHAM surgery. The ovariectomized rats were divided into OVX group with the treatment of placebo or EGCG group which was treated with EGCG by oral gavage. After 8 weeks of treatment, anxiety‐like behaviors were assessed using elevated plus maze test (EMP) and open field test (OFT). The serum estradiol concentration was assayed through ELISA. High‐throughput V3–V4 16S rDNA sequencing was conducted to assess the microbial diversity in fecal samples collected from all rats. EGCG, at a concentration of 10 mg/kg, caused behavioral changes in rats similar to anxiety. In EPM, OVX rats spent less time in open arms than SHAM group rats and EGCG group rats (F = 16.043, p < .01). In OFT, the total travelled distance and the number of entries for EGCG group were higher compared with OVX group (F = 30.939, H = 13.107, respectively; p < .01). In addition, the distribution and composition of intestinal microflora in rats changed after ovariectomy. EGCG modulated the diversity of gut microbiota in OVX group at the phylum and the genus levels. Our results suggested that the composition of gut microbiota and anxiety in OVX rats were simultaneously affected by EGCG, and therefore, the two conditions might be strongly related, yet the deeper mechanistic links need further exploration.

ovariectomized rats were divided into OVX group with the treatment of placebo or EGCG group which was treated with EGCG by oral gavage. After 8 weeks of treatment, anxiety-like behaviors were assessed using elevated plus maze test (EMP) and open field test (OFT). The serum estradiol concentration was assayed through ELISA.
High-throughput V3-V4 16S rDNA sequencing was conducted to assess the microbial diversity in fecal samples collected from all rats. EGCG, at a concentration of 10 mg/ kg, caused behavioral changes in rats similar to anxiety. In EPM, OVX rats spent less time in open arms than SHAM group rats and EGCG group rats (F = 16.043, p < .01).
In OFT, the total travelled distance and the number of entries for EGCG group were higher compared with OVX group (F = 30.939, H = 13.107, respectively; p < .01). In addition, the distribution and composition of intestinal microflora in rats changed after ovariectomy. EGCG modulated the diversity of gut microbiota in OVX group at the phylum and the genus levels. Our results suggested that the composition of gut microbiota and anxiety in OVX rats were simultaneously affected by EGCG, and therefore, the two conditions might be strongly related, yet the deeper mechanistic links need further exploration.

K E Y W O R D S
antianxiety effect, EGCG, intestinal microbiota, ovariectomized rat undergo such changes is directly related to the low level of endogenous estrogens. Furthermore, 65% of postmenopausal women have metabolic dysfunction, which is attributed to the association of the gut microbiota with lack of estrogen (Leeners, Geary, Tobler, & Asarian, 2017). It is reported that the gut microbiota modulates estrogen homeostasis via enterohepatic circulation (Flores et al., 2012). Thus, the effect of menopause on gut microbiota has attracted a lot of attention.
Accumulating evidence from previous studies suggests that the gut microbiota regulates mood disorders, including the two most prominent neuropsychiatric disorders that affect millions of people worldwide: anxiety and depression, via a microbiotagut-brain axis (Rieder, Wisniewski, Alderman, & Campbell, 2017).
Different laboratories and studies have reported that germ-free (GF) mice exhibit consistently reduced anxiety symptoms (Clarke et al., 2013;Neufeld, Kang, Bienenstock, & Foster, 2011). It was also displayed by another study that the anxiety-like and depression-like behaviors were reduced in healthy male Balb/C mice after feeding L. rhamnosus (Bravo et al., 2011), which resulted in altered gut microbiota. Furthermore, the related human study also supports the relationship between microbiota and anxiety (Stevens et al., 2018).
Estrogen replacement therapy was tested as the predominant means to induce antianxiety effects in several studies (Hiroi & Neumaier, 2011;Weiser, Foradori, & Handa, 2008). However, its side effects are inevitable, including increased thrombosis risk, inducement of breast cancer, and serious mental disorder (Rozenberg, Vandromme, & Antoine, 2013). It is very necessary to develop alternative approaches that are more effective and safer. In recent years, the search for effective natural antidepressant and antianxiety agents from plants has gained much interest in the field of psychotropic drug research and development. Tea contains a lot of polyphenols, which exhibit strong antioxidant capacities. The main component of tea polyphenols is catechin. Around 50%-60% catechins found in tea are (-)-Epigallocatechin-3-gallate (EGCG). EGCG exhibits antioxidative, anti-inflammatory, and anticarcinogenic properties and other biological activities (Higdon & Frei, 2003;Pan, Lin-Shiau, Ho, Lin, & Lin, 2000). Furthermore, it has been reported that EGCG could improve bone microarchitecture and reduce stress of oxidation in bladder for ovariectomized (OVX) rats, whose bladder cells were excessively active due to surgical menopause (Chen et al., 2013;Juan et al., 2012). Park et al. (2010) suggested that EGCG could reverse caffeine-induced anxiogenic-like effects in rats. Therefore, EGCG may exhibit antianxiety effects in ovariectomized rats. The changes caused by estrogen depletion in OVX rodents have many characteristics similar to changes in menopausal women, such as weight gain, adipose tissue inflammation, anxiety, depression, and hormone changes. Therefore, OVX rodents may be a suitable model for studying postmenopause. However, how EGCG influences microbiome environments in the gastrointestinal tract of OVX rats is still unrevealed. The current study was planned for verifying the effects of EGCG on the diversity of gut microbiota in addition to anxiety in OVX rats.

| Animal preparation
All animal usage in this study was approved and supervised by the Institutional Animal Care and Use Committee of Jilin University. Female Wistar rats (250 ± 20.0 g, 10 weeks) were acquired from the Changchun Yi Si Laboratory Animal Technology Co., Ltd. (SCXK-(JI) 2016-003). All rats were fed with standard lab chow in a constant light/dark cycle (12h/12h) at room temperature and 70% relative humidity during the course of the experiments. This study was subject to the Institutional Guidelines for Animal Research; and the authors obeyed the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). A total of 30 rats were randomly selected and sorted into three groups: the group of SHAM surgery (SHAM group, n = 10), bilateral OVX group (n = 10), and EGCG-treated group (EGCG group, n = 10). Individual of EGCG group was treated with EGCG (Shanghai Yuan Ye Biotechnology Co., Ltd.; dissolved in saline) at 10 mg/kg through oral gavage once a day. SHAM and OVX individuals were treated with the same amount of saline by the same means of administration. The ovaries of OVX and EGCG rats were surgically removed, and the procedures including cutting two minimal incisions on the back of each rat under 10% chloral hydrate (0.35 ml/kg) induced anesthesia. The SHAM surgery was carried out by exposing the ovaries and sewing them back in. Starting from one week after the surgery, all rats accepted either EGCG or the saline vehicle every day for 8 weeks.

| Behavioral test
Anxiety-related activities were analyzed in open field test (OFT) and elevated plus maze (EPM) test. All behavioral analyses were carried out between 09:00 and 14:00, recorded using a visual tracking system (EthoVision XT 8.5 for video tracking) and monitored by a trained observer blinded to the precedent procedures.

| Open field test (OFT)
A 5-min OFT was carried out to investigate exploratory behavior and anxiety-related behavior in rodents. All rats were subjected to OFT setup according to previous literature (Carrey, McFadyen, & Brown, 2000). After 24-hr interval since the last dose of drugs was administered, animals were released onto the floor of the open field squares (100 cm × 100 cm × 40 cm) which were electronically divided into 9 equal squares, surrounded by base and walls made of nonreflecting gray material. The central square was defined as the ''center'' region.
The initial placement of each rat was in a corner square, facing the wall, and the observation lasted for 5 min. The total travelled distance was associated with activity; and the frequency of entries into the center region was associated with emotionality; these data were automatically recorded. As a result of OFT, the relief of anxiety-like behavior was proved by more time spent in the center region (Hiroi & Neumaier, 2011). The OFT apparatus was carefully cleaned according to laboratory protocols after each session in all behavioral experiments.

| Elevated plus maze (EPM) test
All groups of female rats were subjected to EPM which is widely used to test anxiety-like behavior and responses (Pellow, Chopin, File, & Briley, 1985). The maze was raised 80 cm above the floor, made of dark gray Plexiglas walls, and includes four arms (50 cm in length and 10 cm in width); two of the arms were closed arms with ledges of 40 cm in height, and two arms were open arms with ledges of 0.5 cm in height. A central square connecting the arms measured 10 cm × 10 cm. Individual animal was placed in the central square facing an open arm at the start.
Then, the observation lasted for 5 min using a video camera set above the cage and was analyzed with tracking software. The time spent in the open arms and the frequency of entries into the center of the maze were noted. The EPM apparatus was carefully wiped clean after each test. It was presumed that the animals were more affected by anxiety or similar emotions if they spent more time in the closed arms.

| Preparation of specimens
After all behavior tests, the rats were anesthetized by overdose of chloral hydrate and sacrificed by exsanguination. Before being anesthetized, fecal samples were collected in tubes. Blood samples for estradiol concentration assay were collected by puncturing the abdominal aorta before death and bleeding into tubes. Then, serum was extracted with centrifuge (1,000 g for 10 min) and stored at −80℃.

| Determination of estradiol concentration
The plasma level of estradiol was measured by ELISA assay performed at automated instrument made by Beijing Science and Technology Co., Ltd. (ARB12673). All measurements were conducted and controlled up to the standard manual.

| PCR and sequencing
Fragments in hypervariable regions of 16S V4 rDNA were amplified with specific primer set (515F: 5ʹ-GTGCCAGCMGCCGCGGTAA-3ʹ, 806R: 5ʹ-GGACTACHVGGGTWTCTAAT-3ʹ). The conditions for PCR steps were stated as the following: incubation at 98°C for 3 min; 30 cycles at 98°C for 45 s, 55°C for 45 s, and 72°C for 45 s; 72°C for 7 min; and holding at 4°C. The products of PCR were purified with Ampure XP beads (AGENCOURT) to remove the unspecific products. The sequencing of the product was performed by Illumina HiSeq 2500 PE250 (Illumina) following the instrumental manual.
The library was established based on the MiSeq (Illumina) platform.

| Statistical analysis
One-way ANOVA with LSD was used for statistical analysis among groups to assess the homogeneity of variance, and the rank test using the Kruskal-Wallis method was used to analyze missing variance data. Calculations were done by the SPSS version 11.5 software (SPSS Inc.). p < .05 indicated statistically significant. Figure 1 showed the levels of serum estradiol in different groups after 8-week EGCG administration. The levels of serum estradiol in OVX group were lower compared with SHAM group (F = 267.058, p < .01). However, the level of serum estradiol for EGCG group rats showed no significant difference compared with OVX group.

| The effect of EGCG on anxiety-like behaviors of OVX rats in EPM and OFT
As observed in Figure 2a Table 1). In addition, center zone duration in EPM apparatus was less for OVX rats compared with SHAM group rats and EGCG group rats (F = 41.066, p < .01, Figure 2b and Table 1). Moreover, frequency of open arm duration was less for OVX group rats compared with SHAM group rats (F = 13.605, p < .01, Figure 2c and Table 1). Therefore, OVX rats showed increase in anxiety-like behavior. EGCG administration attenuated the anxiety-like alterations in OVX rats.
During OFT, the total travelled distance and the number of entries for OVX group were significantly reduced compared with SHAM F I G U R E 1 The effect of EGCG on serum estradiol level as the result of 8-week treatment. *p < .01, compared with SHAM group group, indicating an increase of anxiety-like behavior in OVX group (F = 30.939, p < .05, Table 2 and Figure 3). Meanwhile, both the two measurements for EGCG group were higher compared with OVX group indicating the antianxiety effect of EGCG (F = 30.939, H = 13.107, respectively; p < .01). Velocity of OVX group rats was less compared with SHAM group rats, but EGCG group showed no difference with respect to velocity compared with OVX group (F = 30.939, p < .01).  group. The differences revealed that the samples of EGCG group showed enhanced biodiversity than those of OVX group (Table 3).

| β-diversity analysis
The β-diversity of each group was assessed with principal component analysis (PCA) referring to the differentiated OTUs for SHAM, F I G U R E 2 Anxiety-like behaviors in EPM. 2A shows open arm time duration(s) for the three groups. EGCG treatment ameliorated the anxiety-like symptom compared with OVX group (one-way ANOVA with LSD). 2B shows EGCG treatment decreased center zone stay of OVX animals compared with SHAM group and EGCG group (one-way ANOVA with LSD). 2C shows that the frequency of entries in open arms in EPM apparatus was significantly increased in EGCG group compared with OVX group (Rank test). *p < .05, **p < .01, compared with SHAM group; #p < .05, compared with OVX group

| Several phylum and genus differed between OVX and EGCG
At the phylum level, EGCG group had higher relative abundance levels of Actinobacteria, Verrucomicrobia, and Firmicutes compared with OVX group (Table 4). At the genus level, it was also found that the abundance levels of several genera in EGCG group including Bacteroidetes were increased compared with OVX group (Table 5).

| D ISCUSS I ON
In this study, after ovariectomy, serum estradiol level for SHAM group was increased compared with both the OVX and EGCG groups, which confirmed the findings in precedent literature (Zhang et al., 2017). Behavior tests showed that treatment of ovariectomized rats with EGCG greatly enhanced their exploratory behavior and activity during OFT and EMP tests. This suggested that EGCG treatment had an antianxiety-like effect. Furthermore, our current findings showed that, compared with SHAM group, the species richness decreased and the species evenness increased significantly in OVX rats, and EGCG treatment enhanced the richness but reduced the evenness, consequently increased diversity of the fecal microbiota.
More than 80% of women would suffer various degrees of psychological and/or physical abnormalities during menopause (Gracia &

F I G U R E 3
The effects of EGCG on the observations during OFT. 3A showed that the total travelled distance in EGCG group increased compared with OVX (p < .05, one-way ANOVA with LSD). 3B showed that velocity of OVX group rats decreased compared with SHAM group (p < .01). 3C showed that frequency in center zone was less for OVX group rats compared with SHAM (p < .01, one-way ANOVA with LSD) and EGCG group (p < .05, one-way ANOVA with LSD). *p < .05, **p < .01, compared with SHAM group; #p < .05, compared with OVX group F I G U R E 4 The distribution of OTUs for the three groups after 8-week treatment of EGCG Freeman, 2018). The persistent absence of ovarian hormones following ovariectomy in rats was proposed to be the animal model for early postmenopause stage (Bosee & Di Paolo, 1995). Anxiety is a common phenomenon in menopausal female of both human and animals (Dalal & Agarwal, 2015;Dornellas et al., 2018). In this study, the ovariectomized group rats exhibited anxiety-like behavior as observed through OFT and EPM. This corroborated the findings of Fedotova et al. that the anxiety-like state is enhanced in OVX rats compared with normal rats (Fedotova, Pivina, & Sushko, 2017). Meanwhile, OVX rats which were treated with EGCG showed a decreased anxiety-like state compared with untreated OVX rats. The results indicated that EGCG was able to decrease anxiety-like activities which is in agreement with many previous studies (Park et al., 2010;Vignes et al., 2006).
Increasing number of studies have reported that anxiety and depression may be influenced by microbiota-gut-brain axis (Rieder et al., 2017). The underlying associations between microbiota and psychopathological conditions had previously been recognized, for example, microbiota alteration was involved in hypersensitivity of the hypothalamus-pituitary-adrenal (HPA) axis in response to stressors (Sudo et al., 2004), and psychological stress could cause microbial translocation resulting increased gut inflammation (Rieder et al., 2017). Therefore, the millions of bacteria in the gut may substantially influence on mental disorders, including anxiety and depression. Note: Data are expressed as the means ± SD (n = 6 in SHAM group, n = 6 in OVX group, n = 4 in EGCG group). Note: Data are expressed as the means ± SD (n = 6 in SHAM group, n = 6 in OVX group n = 4 in EGCG group).

TA B L E 5
Differential gut microbiota at the genus level mice was increased (Jin et al., 2018). The present study also found the OUT in OVX rats was different from SHAM rats. EGCG exhibits a wide range of physiological functions, including the anticancer, anti-hyperlipidemia, anti-inflammation, and anti-hypertension effects (Legeay, Rodier, Fillon, Faure, & Clere, 2015).
EGCG also exhibits the effect of preventing bone loss in ovariectomized rats. Recent studies have found that EGCG also exhibits antianxiety effects (Park et al., 2010). Moreover, Most, Penders, Lucchesi, Goossens, & Blaak (2017) found that supplementation of EGCG + RES for 12 weeks could affect the gut microbiota composition in men but not in women. It indicated that EGCG may affect the gut microbiota composition in menopausal women who showed lower estradiol concentration. Our study found that treatment with EGCG for 8 weeks could affect the gut microbiota composition, and at the same time, EGCG also affected the anxiety-like behavior in OVX rats which is a famous model for menopausal women. Thus, these findings illuminated the effects of EGCG on neuropsychic diseases, and it revealed that EGCG could be a drug of great potential for menopausal women in the future. However, more evidence was still required to verify that EGCG exhibits antianxiety effects by modulating intestinal microbiota.

| CON CLUS IONS
The present study is the first to reveal that dietary EGCG not only affects the diversity of gut microbiota but also improves anxiety symptoms in ovariectomized rats. At the phylum level, Actinobacteria, Verrucomicrobia, and Firmicutes showed higher abundance in SHAM and EGCG groups compared with OVX group.
At the genus level, 11 genera in EGCG group or SHAM group were different from those in OVX group, including Bacteroidetes. Based on the findings, it was further assumed that the diversity of gut microbiota and anxiety in OVX rats are strongly related supported by the simultaneous effects of EGCG. Although the deeper mechanistic links need further exploration, such as the explanation of relevant molecular signal pathway, it still would be expected that EGCG had a great potential in ameliorating the symptoms of anxiety in postmenopausal women.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

E TH I C A L A PPROVA L
The study was carried out at Department of Physiology, College of