Evaluation on the Characteristics of Gut Microbiome in Polycystic Ovary Syndrome Rats Induced by Dihydrotestosterone or Letrozole

The etiology of polycystic ovary syndrome (PCOS) is unclear. Recent reports indicated that the gut microbiome of PCOS patients and rodents has changed. In this study, we induced the nonaromatizable androgen dihydrotestosterone (DHT) or the aromatase inhibitor letrozole (LET) to induce PCOS model rat to compare the bacterial diversity distribution within and between the two groups. The molecular ecology of the fecal gut microbiota was analyzed by 16S rDNA high-throughput sequencing. Our study found that DHT can reduce the microbial richness in rats. PCoA plots conrmed that DHT group was statistically signicantly separated from C group and LET group. At phylum level, DHT led to a decrease in Bacteroidetes as well as an increase in Cyanobacteria, Tenericutes, Actinobacteria, Spirochaetae and Saccharibacteria. At genus level, LEfSe analysis showed that genus of Bidobacteriales, Vibro, Peptococcus and Turicibacter played roles in the letrozole induced PCOS rats. And Lachnospiraceae_NK4A136_group, Ruminococcus_1, Ruminiclostridium, Anaerotruncus and Anaeroplasma played vital roles in the intestine of DHT induced PCOS rats.


Introduction
Polycystic ovary syndrome (PCOS) is a multifactorial, heterogeneous and complex genetic, endocrine and metabolic disease, which is characterized by hyperandrogenism, hyperinsulinemia, insulin resistance (IR) and chronic anovulation. [1][2][3]. Hyperandrogenism may play an important role in altering the gut microbiome in women with PCOS [4]. Some studies have reported that compared with healthy women, the overall bacterial species richness (α diversity) of the gut microbial community of PCOS women decreased, and several bacterial taxa changed [5,6]. Zhang et al. [7] con rmed that gut microbiome was involved in the development of IR and menstrual disorders in PCOS patients by affecting intestinal wall permeability. Some researchers believed that dysbiosis of gut microbiota can increase the production of ovarian androgen, and then interfere with the development of normal follicle by triggering a chronic in ammatory reaction and IR [8]. The gut microbiome is closely related to PCOS, but there is no consensus on which phylum or genera in gut microbiome is related to PCOS. The heterogeneity of PCOS is frequently re ected in many animal models. Therefore, if a rat model not only shows the characteristics of ovarian and metabolic syndrome, but also shows the dysbiosis of gut microbiota, it would be valuable for further study of new PCOS therapy. Testosterone has been proved to be related to the composition change of the gut microbiota [9]. Letrozole is a non-steroidal aromatase inhibitor, which can block the conversion of testosterone to estradiol [10], and increase the testosterone level by increasing the production of ovarian androgen [11]. Continuous exposure to Dihydrotestosterone (DHT) can establish a PCOS rat model with typical hormonal disorders and ovarian morphological changes [12]. Excessive androgen secretion in early life may lead to hormone secretion disorder, which may lead to PCOS in adulthood [13][14][15]. In this study, female rats took DHT or letrozole continuously from before puberty, which led to adult PCOS rats.

Animals
Fifteen speci c-pathogen free (SPF) level female Sprague Dawley (SD) weighting 50g± 5g, were aged 21 days at the onset of experiments. They were from the Experimental Animal Science Department of Guangzhou University of Chinese Medicine, Guangzhou, China (License number SCXK-2016-0168). The experimental protocols were performed after approval and in accordance with the guidelines set by the Experimental Committee of Guangzhou Medical University (Protocol number: GY2018-021). All rats were provided with humane care in a temperature-controlled room with a 12-hr light/dark cycle (lights on 07:00-19:00) and ad libitum access to food and water in their cages (22 °C-24 °C and 60% humidity).

Study procedure
Fifteen rats were randomly divided into three groups on average. According to previous studies [16,17], from 3w age, the rats in DHT group were subcutaneously injected daily with 83μg 5α-DHT which dissolved in 0.2ml tea oil for 6 weeks to mimic the hyperandrogenic state in women with PCOS. The rats in other two groups were injected daily with the same volume of tea oil alone. Follow the experiment of Kafali H et al [10], from 6week age, the rats in letrozole (LET) group were oral administration with letrozole (1mg/kg body weight of letrozole dissolved in 1 ml of 1% carboxymethyl cellulose (CMC) for 3 weeks. At the same time, the rats in control (C) group and DHT group were daily oral administration with 1 ml of 1% CMC. All the rats were fed commercial chow (Research Diets GB 14924.  [67% carbohydrate, 21% protein, 12% fat, and total 3.45kcal/g], was provided by Guangdong Medical Laboratory Animal Center.

Vaginal smears
The major cell type in vaginal smears obtained every day from 6 weeks old to the end of the experiment were analyzed by microscope to determine the periodic stage. Rats showing regular cycles of 4-5 days complete with the proestrus, estrus, metestrus and diestrus stages were de ned as normal cyclic rats, whereas animals in which the estrus cycle was found arrested in any one of the stages for 4 consecutive days were termed as acyclic rats [18].
The optical density values were read at 450 nm using a microplate reader. HDL-C, LDL-C, total cholesterol and triglyceride levels were quanti ed by following the protocols provided with kits on UniCel DxC 600800 Synchron Clinical System. IR was appraised with the homeostasis model assessment of insulin resistance (HOMA-IR) method. HOMA-IR was calculated using the following formula: HOMA-IR =FBG (mmol/L) * FINS (mU/L)/22.5 [19].

Ovarian Histology
The left ovary of rats was xed in 4% paraformaldehyde and embedded in para n. 5 μm thick sections were prepared and stained with Hematoxylin-Eosin and histo-anatomical changes were observed and photographed under light microscope (BX-51, Olympus, Tokyo, Japan atX40 magni cation).
16s rDNA sequencing data analysis The fecal microbiome for 15 fecal samples collected from 15 rats in the three groups. The 16S rDNA high through put sequencing (V3-V4 region) was performed using an Illumina MiSeq platform. After assembly, quality ltering and the random extraction of sequences at 97% similarity, the operational taxonomic units (OTUs) for species classi cation were obtained. The Chao1 and Shannon α-diversity indices were used to assess bacterial richness and diversity. In the aspect of β diversity, based on Bray-Curtis dissimilarity, permutational ANOVA (PERMANOVA) was used to assess bacterial community compositional difference. Linear discriminant analysis effect size (LEfSe) analysis coupled with the Kruskal-Wallis rank sum test was performed to identify the microbial differences among all groups. In LEfSe analysis, LDA effect size of >3 was used as the threshold.

Statistical analyses
Most statistical evaluations were performed with SPSS 20.0 for Windows (SPSS Inc., Chicago, IL, United States). All data were presented as mean± SEM. One-way ANOVA was used to determine the signi cance, p 0.05 was considered signi cant. When the ANOVA revealed signi cant differences among three groups, post hoc analysis was performed by a Tukey honest signi cant difference test. Kruskal-Wallis test was used for not normally distributed values.

Animal characteristics
The weight of DHT and LET rats increased more than that of control group (ANOVA, P 0.01) (tab. 1). The concentration of HLD-C and the activity of SOD decreased in DHT group, compared with the C group. And the level of MDA in DHT group was higher than in LET group (p 0.05). Levels of T were similar in DHT rats and controls but were signi cantly elevated in LET rats (p 0.05).

Vaginal smears
The vaginas in all rats were mostly opened at the age of 40 days. The normal oestrus cycle of all control rats is 4~5 days, including proestrus, estrus, metestrus and diestrus. Vaginal smears of DHT and LET groups showed predominantly leukocytes, which indicated that they appeared continuously.
Ovarian morphological changes A cystic follicle was considered to be a large uid-lled cyst with an attenuated granulosa cell layer. Cross-sectional area of ovary was smaller in DHT rats and bigger in LET rats than in the controls. Light microscope observation showed that enlarged ovaries with multiple cysts in LET rats, while the ovaries of rats in DHT group had a tendency of follicle enlargement. The number of cystic follicles was increased with a relatively thin granulosa cell layer in DHT and LET rats compared with controls ( Fig. 1).
The biodiversity in the gut microbiome As shown in Fig.2, compared with the other two groups, DHT group had lowest Chao1 α diversity (p 0.01).
The Shannon index in DHT group was signi cantly higher than the LET group (p 0.01), there was no signi cant difference between DHT and C group. Differences in microbial community structure between groups (β-diversity) were examined with PERMANOVA and visualized with PCoA plots (Fig. 3). At the taxonomic level of OTU, signi cant differences in microbial communities were observed between C and DHT rats (p = 0.006) and between LET and DHT rats (p =0.008). No signi cant difference was observed between C and LET group (p=0.054). (Table.2) The microbial community structures at the phylum and genera levels We evaluate the microbial composition by comparing the relative abundances of taxa at the phylum and genera levels. An overview of gut microbiota composition in each group at the phylum level was illustrated in Fig4. As shown in the Fig 5, compared with C group, a decrease in Bacteroidetes as well as an increase in Cyanobacteria, Tenericutes, Actinobacteria, Spirochaetae and Saccharibacteria in DHT group. No signi cant difference was discovered between C group and LET group.
LEfSe is a biomarker discovery tool for high dimensional data and used to determine which OTUs were differentially abundant between groups. The characteristic LDA discriminant histogram was used to calculate the signi cant gut microbiota between two groups to explain the consistency difference of relative abundance between the two groups. At genus level, the results showed that nineteen genera were obviously representative between DHT group and C group, with two (Vibrio and Catabacter) were more abundant in C group, and seventeen obviously changed genera (Lachnospiraceae_NK4A136_group, Ruminococcus_1, Ruminiclostridium, Anaerotruncus, Anaeroplasma, Acetatifactor, Treponema_2, Tyzzerella, Lachnoclostridium, Ruminococcus_2, Enterorhabdus, Coprococcus_1, Lachnospiraceae_FCS020_group, Streptococcus, Victivallis, Tyzzerella_3 and Marvinbryantia) were from DHT group. And four genera were distinctively represented between LET and C groups, with two (Peptococcus and Turicibacter) were more abundant in C group, and two (Vibrio and Bi dobacterium) were more abundant in LET group. (Fig6A-B and Fig7A-B).

Discussion
PCOS is a complex endocrine and metabolic disorder, which is related to ovulation dysfunction, hyperandrogenism, obesity and IR. Hyperandrogenism is a key feature of PCOS, and androgen may promote early follicular differentiation [20]. In present study, rats letrozole group exhibited endocrine and reproductive characteristics, such as hyperandrogenism, abnormal oestrus cycles or complete acyclic, polycystic ovaries, and obesity. DHT induced rats were showed obesity, irregular oestrus cycles, polycystic ovaries, lower level of HDL-C and lower activity of SOD than controls. Abnormal lipid metabolism is often present in women with PCOS, and its incidence rate is reported as high as 50% [21]. In PCOS patients, hyperandrogenemia is related to higher concentrations of plasma TG and lower HDL-C [22,23]. In our study, the DHT rats had lower HDL-C concentration than controls, while letrozole did not reduce HDL-C.
Oxidative stress (OS) is de ned as an imbalance in normal cellular environment, which is considered to be the cause of the development of many pathological diseases or deterioration of symptoms (such as in ammation) [24][25][26]. OS can also contribute to PCOS [27], the activity of SOD which related to the antioxidant capacity in DHT group was lower than in C group. We suspected that DHT may have a stronger and more direct role in regulating HDL-C and SOD activity than letrozole.
It is universally acknowledged that PCOS is a kind of complex heterogeneous disease [28,29], altered gut microbiome may be associated with the occurrence of PCOS [30]. Sex hormones play an important role on the gut microbiome of the host [31,32]. A study demonstrated that α diversity has a negative correlation with hyperandrogenism, total testosterone, and the other primary reproductive and metabolic characteristics of PCOS [10]. We evaluated two α-diversity indices, including the Chao 1 and Shannon index. The Chao1 index which accounted for richness of a community, DHT group owned the lowest richness of the gut microbiome in three groups. The Shannon index is an index that is commonly used to describe the species diversity in a community. No signi cant difference in Shannon index was observed between C and DHT rats, and between C and LET rats. But the Shannon index was signi cantly higher in DHT group than LET group. Alpha diversity analysis showed that the gut microbiota abundance and diversity in DHT group changed signi cantly. And the results demonstrated that the effects of DHT on gut microbial community may be complicated. With respect to beta diversity, PCoA demonstrated signi cant differences between DHT group and the other two groups on the rst axis, suggesting that the microbial community composition from LET group was more similar to C group and DHT could in uence microbial community composition. Changes in gut microbial composition may be unfavorable, and lead to some diseases. In the eld of ecology, species richness is considered to be related to ecosystem health, because diverse communities may improve the stability and productivity of an ecosystem [33]. According to several recent meta-analyses, compared with healthy individuals, it was also observed that α diversity of the gut microbiome reduced in human with metabolic diseases, which is related to human obesity [34][35][36].
DHT exposure altered the abundance of Bacteroidetes, Cyanobacteria, Tenericutes, Actinobacteria, Spirochaetae and Saccharibacteria. Bacteroides is involved in the metabolic process, which could absorb and metabolize polysaccharides, bile acid and steroids in foods that cannot be metabolized by the host. [37]. Turnbaugh et al. found that obese mice had more Firmicutes and less Bacteroides than lean mice, and that this distribution of gut microbiome can better absorb energy [38]. Bacteroidetes and Firmicutes are signi cantly related to metabolic syndrome, and the abundance of Bacteroides increased while that of Firmicutes decreased in exercise rats [39]. So, it showed that gut microbiome disorders may accelerate the process of DHT induced PCOS by affecting energy absorption, and appear obese. In addition, compared with C groups, DHT signi cantly increased the abundance of Actinobacteria, Cyanobacteria and Saccharibacteria. And the relative abundance of Actinobacteria and Saccharibacteria were signi cantly high in DHT group than LET group. The changes in the gut microbiota are related with speci c metabolic states, such as obesity [40]. Actinobacteria was positively correlated with obesity [41], the high level of Actinobacteria aggravated the disorder of the intestinal ora and contributed to the progression from obesity to the related diseases [42]. Saccharibacteria are prominent in the oral cavity, and increased abundance of Saccharibacteria is correlated with the development of periodontitis [43] and in ammatory bowel disease [44]. However, the role of increased abundance of Saccharibacteria in DHT model rats are still unclear, which may be related to the in ammatory conditions. In order to identify the key bacterial genus in the gut microbiota, we used LEfSe multilevel species discrimination. The larger the LDA score shows the more signi cant the difference between the two groups. The results showed that Lachnospiraceae_NK4A136_group, Ruminococcus_1, Ruminiclostridium, Anaerotruncus and Anaeroplasma were abundant in the intestine of DHT induced PCOS rats. The Lachnospiraceae NK4A136 group is an indicator of gut dysbiosis, because when present in great abundance, gut dysbiosis is more severe [45]. Ruminiclostridium has a significantly positive relationship with hyperglycemia [46], and was richness in DHT group. A study showed that the distribution and structure of the intestinal micro ora (including Anaerotruncus) changed after ovariectomy in rats [47]. Anaerotruncus was signi cantly positively correlated with ovary weight [48], which might be in line with rats' ovaries enlarged in our study. In addition, an increased relative abundance of Anaerotruncus species is related to aging and ageassociated in ammation with elevated pro-in ammatory chemokines in a mouse model [49]. Interestingly, as a taxa of the family Ruminococcaceae, genus Ruminococcus_1 is one of the dominant butyrate-producing bacteria [50]. Anaeroplasma is considered as a potential anti-in ammatory probiotic for treating of chronic intestinal in ammation [51], but an increased relative abundance of Ruminococcus_1 and Anaeroplasma were observed in DHT group. In line with that, the concentrations of TNF-α and IL-22 were similar in C group and DHT group.
According to the score of LDA, compare with LET group, Peptococcus and Turicibacter were more abundant in C group. Peptococcus is a Gram-positive bacterium commonly found in digesters, and was positively correlated with butyrate [52]. Butyrate is a short-chain fatty acid, which is helpful for maintaining cellular health by inhibiting the expression of pro-in ammatory cytokines in the intestine through mechanisms including nuclear factor kappa B signaling [53]. Turicibacter was also reported to be related to anti-in ammatory properties [54][55]. The interesting nding was the higher abundance Bi dobacterium in LET group than in C group. Bi dobacterium a traditional probiotic, which can ameliorate visceral fat accumulation and insulin sensitivity in an experimental model of the metabolic syndrome [56]. Bi dobacterium is also considered as a potential candidate drug for treating obesity [57], and is known to enhance intestinal integrity [58].
In a word, our ndings indicate that DHT affects the composition and diversity of gut microbial community, and leads to the gut dysbiosis. Letrozole may bene t to the gut microbiome in some aspects, but it also may have a trend of inhibiting the expression of some anti-in ammatory genera in intestinal tract.

Declarations
Ethics approval: Animal studies were performed in compliance with the Experimental Committee of Guangzhou Medical University.
No con icts of interest, nancial or otherwise, are declared by the authors.

Consent for publication:
Not Applicable Acknowledgements: We are grateful to Guangzhou Medical University for providing us with a laboratory.

Funding:
This work is supported by the research grants from the National Natural Science Foundation of China (No. 81704107).