Fertility factors affect the vaginal microbiome in women of reproductive age

Abstract Problem For women of reproductive age, achieving a successful pregnancy requires both the normal functioning of reproductive endocrine and the health of the reproductive tract environment. We aimed to study how these fertility factors, such as female age, baseline sexual hormone levels, tubal patency, and vaginal pH, affect the composition of vaginal microbiome. Method of study The 16S rRNA sequencing was carried on vaginal microbiome samples from 85 women of reproductive age without vaginal infections or reproductive endocrine diseases. The detailed correlations between fertility factors and vaginal microbiome were quantified by Spearman's rank tests. A linear discriminant analysis was carried out to explore the effects of fertility factors on the relative abundances of vaginal bacterial species. Results The vaginal pH, levels of basal E2, LH, and FSH all had significant effects on the distribution of vaginal microbiome. The relative abundances of vaginal bacterial species, including Escherichia coli, Streptococcus agalactiae, and Prevotella intermedia, were significantly different due to the host's state of reproductive endocrine and tubal patency. It was worth noting that women with tubal obstruction, or prolonged menstrual cycle, or antral follicle count >15, or vaginal pH > 4.5 all had a higher abundance of Escherichia coli in vagina. Conclusion The fertility factors associated with the reproductive endocrine and the genital tract environment affected vaginal microbiome in women of reproductive age. The species Escherichia coli, Streptococcus agalactiae, Prevotella intermedia, etc could be used as biomarkers to reflect the pathological state of reproductive endocrine and genital tract.


| INTRODUC TI ON
For women of reproductive age, achieving a successful pregnancy requires both the normal functioning of reproductive endocrine and the health of the reproductive tract environment.
Reproductive endocrine plays an essential role in promoting gonadal development, maintaining fertility, and ensuring reproductive health. Meanwhile, a healthy reproductive tract environment also helps support successful embryo implantation and maintains pregnancy.
As the age in women grows, the levels of sex hormones change and the vaginal microbiome is in constant flux. 1 During a woman's reproductive years, the fluctuating levels of hormones that regulate the menstrual cycle are an important influence on the vaginal microbiome, 2 especially the estradiol (E2) and progesterone (P). 3,4 The levels of sex hormones would affect the components of the female genital tract defensive barriers, including the mucous viscosity, epithelial barrier thickness, immune cell frequency, and resident vaginal microbes. 3,5 Previous studies have explored the physiological interaction between sex hormones and vaginal flora. Estrogen promotes hyperplasia and thickening of vaginal epithelia and the increase in glycogen. 6 As the dominant bacteria in the vagina, Lactobacillus can change the glycogen into lactic acid, maintain the acidic environment of the vagina, inhibit the growth of other pathogens, 7,8 and strengthen the immune system. [9][10][11] When a woman enters menopause, the level of estrogen will significantly reduce, while the level of follicle-stimulating hormone (FSH) increases, and the vaginal pH environment changes from acidic to weakly acidic, which leads to the colonization of a large number of mixed bacteria. [12][13][14] Meanwhile, the estradiol-based hormone replacement therapy can maintain Lactobacillus dominance in post-menopausal women, supporting a link between estradiol and lactobacilli. 12,15 The levels of sex hormones in the reproductive endocrine system are closely related to female fertility. Clinically, the levels of Therefore, in this study, we examined the vaginal microbiome of women of reproductive age and tried to study how these fertility factors associated with the reproductive endocrine and reproductive tract environment affect the composition of the vaginal microbiome.

| Ethics statement
This study was approved by the Ethical Committee of Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine (Approval No. 2016012904). All participants have provided written informed consent prior to their inclusion in the study. Women who had taken any antibiotic drugs in the past 3 months were also excluded.

| Sample collection and clinical measurements
All participants reported their conditions of menstruation and fallopian tube patency. We recorded their age, height, and weight.
The vaginal pH values were tested from the posterior fornix during the menstrual interval. We detected serum AMH levels using an ultrasensitive enzyme-linked immunosorbent assay (Kangrun Biotech, catalog number KR-AMH-001). Since the levels of basic sexual hormones were regarded as stable at the baseline during the early follicular phase of the menstrual cycle, 17

| PCR amplification and 16S rRNA gene sequencing
Amplification and sequencing were carried out as protocols de-

| Sequence processing and data analysis
Raw Illumina FASTQ files were demultiplexed, quality filtered, and analyzed by Quantitative Insights Into Microbial Ecology (QIIME, version 1.8.0). 19 The operational taxonomic units (OTUs) were clustered with a 97% similarity cutoff using UPARSE. 20 Taxonomy was assigned to representative sequences using the Ribosomal Database Project (RDP) classifier in QIIME 21 with a confidence value of 0.8 against the Silva bacterial 16S rRNA gene dataset (version 119) and Greengenes database for species-level taxonomy information.

| Gene function prediction
We used Canoco 5.0 software to perform redundancy analysis (RDA) at the genus level. 22 For the multiple comparisons, the false discovery rate (FDR) adjusting method was employed to avoid type I errors.
Provided by Benjamini and Hochberg, it advocated control of the expected proportion of falsely rejected hypotheses. 23 The correlation coefficients between the relative abundance of vaginal bacteria and clinical factors were calculated by Spearman's rank tests. Linear discriminant analysis (LDA) effect size (LEfSe) method was used to discover the high-dimensional biomarker and explain the genomic feature identification that characterizes the differences between various biological conditions. 24

| Vaginal microbial community composition of the women of reproductive age
The clinical results of fertility factors of 85 women of reproductive age were shown in frequency bar charts ( Figure S1). 16S rRNA gene sequencing generated a total of 3 218 048 high-quality reads, which

| Analyses of correlations between the fertility factors and vaginal microbiome
We used the redundancy analysis (RDA) to visually examine the effect of fertility factors on the distribution of samples ( Figure 2).
The results showed that vaginal pH, levels of serum E2, LH, and FSH all had significant effects on the distribution of vaginal microbiota (P < .05). Among them, vaginal pH and the levels of serum E2 had the most significant effects (contribution of 29.5% and 21.7%, respectively).
Next, we quantified the correlations between the relative abundance of vaginal flora and fertility factors by Spearman's rank tests.
The results were presented in a heatmap ( Figure 3). The vaginal bacteria were arranged according to the value of the correlation coefficient between the relative abundance of vaginal bacteria and the vaginal pH.
As can be seen from the results, the relative abundance of Lactobacillus was negatively correlated with vaginal pH, suggesting that Lactobacillus was more likely to exist in a weakly acidic environment, which is a normal vaginal environment. Meanwhile, the relative abundances of Escherichia/Shigella, Gardnerella, Prevotella, and Enterococcus were positively correlated with vaginal pH. That means when the vaginal pH rises, the colonization of these bacteria would increase.
The relative abundance of Paraprevotella was negatively correlated with age and FSH. It was suggested that this bacterium was more commonly found in the vagina of women with normal ovarian function or younger. Meanwhile, the relative abundances of genera Varibaculum, Streptococcus, and Veillonella were positively related to age, indicating that the colonization of these bacteria in the vagina may increase with female age.
In addition, the relative abundances of genera Aerococcus and Atopobium were negatively correlated with LH. The relative abundance of Gemella was positively correlated with FSH/LH and that of Streptococcus was positively correlated with E2.

The vaginal genera negatively correlated with AFC included
Varibaculum, Porphyromonas, Granulicatella, and Anaerococcus, while the relative abundance of Escherichia/Shigella was positively correlated with AFC.

| Specific vaginal species in each subgroup of fertility factors
We then divided the women of reproductive age into multiple subgroups based on each fertility factor. It was found that there were significant differences in the abundances of some vaginal species between/among the subgroups of several fertility factors,  We also found out that when the number of AFC was at different levels, the abundances of vaginal bacterial species in women of reproductive age also had specific differences. For women whose AFC was between 0-6 and 6-15, the relative abundances of unclassified Varibaculum, unclassified Granulicatella, and unclassified Anaerococcus in the vagina were higher than that in AFC > 15 group.

| D ISCUSS I ON
The results above suggested that the fertility factors associated with reproductive endocrine and the reproductive tract environment could affect the vaginal microbiome in women of reproductive age.
The relative abundances of several vaginal bacterial species were significantly different due to the host's age, levels of basic sexual hormones, antral follicle count, fallopian tube patency, menstrual cycle length, and vaginal pH.
We found that in older women, or women with higher levels of basal E2, the abundance of Streptococcus agalactiae in the vagina was higher than the normal group. Streptococcus agalactiae, also known as Group B Streptococcus (GBS), is an important pathogen in aerobic vaginitis. It rarely causes infections in healthy adults; however, occasionally it may cause morbidity in older women, pregnant women, or the patients with underlying predisposing conditions. 25 Elderly adults account for >40% of persons with invasive GBS disease in the United States, colonized at vaginal and rectal sites.
It causes skin and soft-tissue infections, and bacteremia with no identified focus. 26,27 In pregnant and postpartum women, GBS is a frequent cause of asymptomatic bacteriuria, urinary tract infection, and upper genital tract infection et al. 28,29 It is noted that invasive maternal infection with GBS is associated with pregnancy loss and preterm delivery. 30,31 In the early follicular phase of women of reproductive age, high levels of basal E2 (>60-80 pg/mL) with normal FSH levels may indicate a poor ovarian response. Early follicular E2 levels are better used with basal FSH to assess ovarian reserve. 32 However, the relationship between GBS and the ovarian response has not been studied, and our study was the first to find that GBS infection might be associated with the decline of ovarian function. Lactobacillus iners dominated vaginal community type seems to be less stable or more in transition than the other community types and more associated with vaginal dysbiosis. 36 Our results suggested that high level of basal E2 may have an impact on the colonization of Liners.
Lactobacillus plantarum strains produce anti-infective agents, including hydrogen peroxide, and are able to co-aggregate efficiently with vaginal pathogens. 8 It helps treat vulvovaginal candidiasis and bacterial vaginosis, and restore a normal vaginal microbiota. 37,38 Moreover, it was first found that the abundances of Atopobium vaginae and unclassified Aerococcus in the vagina were correlated with the basal LH level. LH activity had previously been deemed potentially detrimental to reproductive function. 39 However, Sun et al 40 found that a high basal LH level did not have any adverse effects on oocyte quality and embryo quality and did not lead to poor outcomes of IVF/ICSI treatment. Moreover, more embryos and more top-quality embryos developed in the group with LH > 10 mIU/mL than in the other groups.
The genera Prevotella, Atopobium, and Gardnerella were associated with bacterial vaginosis (BV), which may be linked to an increased risk of preterm labor. 11 Abnormal vaginal microbiota, such as a high abundance of Gardnerella vaginalis or Atopobium vaginae, may negatively affect the clinical pregnancy rate in IVF patients. 41 Based on our results, it was also first discovered that the number of antral follicles was also associated with the colonization of genera Granulicatella, Anaerococcus, and Varibaculum in the vagina. Previous researches have shown that the relative abundances of genera Granulicatella and Anaerococcus were increased in bacterial vaginosis patients. 42 Varibaculum cambriense has been isolated from intrauterine devices and human vagina and abscess specimens. 43 It was interesting to notice that women with AFC > 15 or pro- The results above suggest that vaginal bacteria may affect the patency of fallopian tubes by causing the female genital infection.
In addition, we found that women with normal menstrual cycles have more unclassified Ochrobactrum in the vagina than women with the prolonged menstrual cycle, which is an emerging opportunistic pathogen in immunocompromised patients. 53 It should be noted that although subjects with mycoplasma or chlamydia infection were excluded according to the clinical laboratory tests, some subjects were found to had Mycoplasma or Ureaplasma colonized in the vagina with a relative abundance of <5%. This may be due to the higher sensitivity and precision of 16S rRNA sequencing.

| CON CLUS IONS
The fertility factors associated with reproductive endocrine and the genital tract environment had effects on the vaginal microbiome in women of reproductive age. The bacterial species Escherichia coli, Lactobacillus iners, Streptococcus agalactiae, Prevotella intermedia, etc could be used as biomarkers to reflect the pathological state of reproductive endocrine and genital tract.

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

PU B LI C S H A R I N G O F DATA
Sequence data were archived in the NCBI Sequence Read Archive (SRA) under accession number SRP136384.