Deciphering the effect of reproductive tract microbiota on human reproduction.

Reproductive tract microbiota at the embryo‐maternal interface is receiving increasing attention in human reproduction because it may impact not only the chances of achieving a pregnancy, but also the health status of the mother and the child before and after delivery. The vaginal microbiome of pregnant women who deliver at term is abundant in Lactobacilli , with a significant dominance of CST‐I, ‐II, ‐III, and ‐V, whereas high levels of Gardnerella , Ureaplasma , or other bacteria belonging to the CST‐IV are often associated with preterm birth and obstetrical complications. 64 , 65 , 66 , 67 , 68 , 69 On the other hand, the postpartum vaginal microbiome is characterized by a shift to CST‐IV even in those pregnancies in which a high Lactobacillus abundance was maintained throughout pregnancy. 69 The dominance of Lactobacillus in the vagina during pregnancy could be explained by the increase in estrogen levels, but it is also hypothesized that this protective microbiome serves as a barrier to prevent the growth of potential pathogens in the fetal membranes and placenta. For this reason, dysbiotic deviations from the LD vaginal profile during pregnancy result in increased risk of miscarriage, premature rupture of membranes, and preterm birth. 70 , 71 , 72 However, these associations have been mainly reported in White women and could be related to the host background, based on the lack of significant association between BV and preterm birth in Black women. 73 Another upper genital tract infection, deciduitis, is the inflammation of the maternal tissue in the basal plate of the placenta. This infection is proposed to be the consequence of untreated preconception chronic endometritis that produces immune and inflammatory responses at the maternal‐fetal interface, leading to preterm labor. 74 , 75 From the fetal side, infection of the fetal membranes, known as chorioamnionitis, is also an important obstetrical complication caused by bacteria ascending from the vagina and colonizing the choriodecidual space and the amniotic fluid. The most commonly isolated pathogens from amniotic fluid are Bacteroides species, E. coli , G. vaginalis , M. hominis , Peptostreptococci , Streptococci , and Ureaplasma urealyticum , but in some cases, the amniotic inflammation is accompanied by negative microbial cultures, highlighting the importance of high‐sensitivity molecular diagnosis of the microbiome in low biomass environments to detect and prevent severe obstetrical complications. 76 , 77 Evidence from several groups indicates that infertile patients harbor a differential reproductive tract microbiota (lower and/or upper) compared to healthy and fertile women. 4 , 78 , 79 , 80 Thus, it is worth considering whether IVF outcomes could be influenced by the microbial taxa present in the reproductive tract during infertility treatment. Indeed, deviations from the low‐diversity vaginal microbiome, including BV, have been significantly correlated with decreased pregnancy rates after IVF. 81 , 82 The association between endometrial dysbiosis and ART failure has long been studied using classical bacterial culture of the tip of the transfer catheter used for embryo transfer. Since the mid‐1990s, the isolation of Enterococci , Enterobacteriaceae ( E. coli and Klebsiella pneumoniae ), Streptococci , Staphylococci , and/or Gram‐negative bacteria has been correlated with lower implantation and pregnancy rates and increased miscarriage rates, while isolation of Lactobacilli or samples with negative cultures for the aforementioned pathogens has been correlated with better reproductive results. 83 , 84 , 85 , 86 , 87 Using the same strategy, the isolation of Streptococcus viridans was significantly correlated to decreased live birth rate, with only 7% of successful pregnancies compared to 88% when Lactobacillus was isolated from the catheter tip at the time of embryo transfer. 85 These studies have now been confirmed using NGS, which is able to characterize the endometrial microbiota at the molecular level in infertile patients with different ART indications: unexplained infertility, repeated implantation failure (RIF), and recurrent pregnancy loss (RPL). 10 , 11 , 43 , 44 The percent composition of Lactobacillus in the endometrium differs between healthy volunteers (85.7%), non‐IVF patients (73.9%), and IVF patients (38%), showing that a high percentage of infertile patients subjected to ART present an abnormal endometrial bacterial profile. 11 This is consistent with previous findings showing that 46% (15/32) of IVF patients with a receptive endometrium presented an NLD microbiota. 10 The first report on the endometrial microbiome's influence on ART using 16S rRNA gene sequencing found that Lactobacillus and Flavobacterium were the most represented genera in the tip of the embryo transfer catheter; however, there was no statistically significant association between the assessed microbiome and the reproductive outcomes of those patients. 44 Subsequently, the uterine microbiome was investigated by 16S targeted sequencing of endometrial fluid collected from 35 RIF patients with a receptive endometrium. Here, the percentage of Lactobacilli was indeed associated with reproductive outcomes (Figure 4 ). NLD microbiota strongly associated with poor reproductive outcomes, compared to patients with LD microbiota, in terms of implantation rate ( P  = 0.02), pregnancy rate ( P  = 0.03), ongoing pregnancy ( P  = 0.02), and live birth rates ( P  = 0.002). 10 Interestingly, these results have been partially confirmed by a group evaluating the endometrial fluid of 79 IVF patients and its impact on reproductive outcomes. In that study, 62% of infertile patients receiving ART presented an NLD microbiome, and interestingly, patients achieving a pregnancy after IVF presented an average content of Lactobacillus of 96.5% ± 33.6%, suggesting that an LD endometrium might favor implantation. 11 Together, these results support the relevance of uterine health, even at the microbiological level, for a successful implantation and pregnancy, and point to endometrial infections (ie, CE) as causes of infertility. The estimated prevalence of CE in infertile patients ranges from 0.2% to 46%. 88 CE is also an additional risk factor for infertility in RIF patients; while RIF patients present an implantation rate of 46% after embryo transfer, this rate falls to 15% in RIF patients with concomitant CE. 89 The abundance of Lactobacilli in endometrial fluid samples is associated with reproductive outcomes in ART patients. Reprinted from Moreno et al, 10 with permission from Elsevier At the end of the 20th century, Egbase et al unequivocally demonstrated the relevance of the reproductive tract microbiota for reproductive success. In this study, the authors compared the bacteria isolated from the transfer catheter tip of a mock transfer performed at oocyte retrieval—when prophylactic antibiotics were administered—and at the time of embryo transfer—performed 48 hours later—with their pregnancy results (Egbase et al 83 ). The results of this study showed that isolation of endometrial pathogens ( Enterococcus , Staphylococcus , E. coli , and other mixed cultures) at embryo transfer correlated with decreased clinical pregnancy rate per transfer compared to those with negative cultures or those that have responded to prophylactic antibiotics after a positive culture at egg retrieval (18.7% vs 41.3% and 38.1%, P  < 0.01, respectively). 90 Following this study, it is logical to hypothesize that decreasing the numbers of bacterial pathogens in the reproductive tract and increasing the proportion of beneficial Lactobacillus could improve reproductive outcomes in those patients with an abnormal microbiota. To address this, two different approaches are under study: the use of antibiotics and the use of probiotics. The use of antibiotics has been widely studied to treat BV and prevent preterm birth (reviewed in [ 30 , 37 ]). But the usefulness of antibiotic use before embryo transfer remains controversial; while it is efficient in reducing upper genital tract contamination, no beneficial role has been observed in pregnancy outcome. 91 This can be due to the lack of specificity of broad‐spectrum antibiotics that could impair not only the growth of dysbiotic bacteria, but also the protective Lactobacilli . Other studies describing the impact of prescribed antibiotics to treat CE in infertile patients have yielded dissimilar results. A recent systematic review and meta‐analysis evaluated the efficacy of CE antimicrobial therapy to improve pregnancy outcomes in RIF patients. 92 The results show that antibiotic therapy is not only effective at eliminating the cause of infection by targeting specific pathogens, but also is useful to ameliorate infertility in RIF patients with CE. Specifically, effective antibiotic therapy against CE improved implantation, clinical pregnancy, ongoing pregnancy, and live birth rates in a subsequent IVF cycle after CE was resolved. Moreover, there were no statistically significant differences in these outcomes between patients with resolved CE and those without CE (Figure 5 ). These findings suggest that diagnosing and treating CE in RIF patients before embryo transfer could be a useful intervention to eliminate the source of infection, improve the endometrial microbial health, and increase the live birth rates in these patients. 92 Importantly, antibiotic therapy has been also effective in patients with CE and unexplained infertility to increase their chances to conceive spontaneously and maintain a safe pregnancy to term after CE resolution. 93 Effect of antimicrobial therapy on reproductive outcomes in ART patients. A, Ongoing pregnancy rate and/or live birth rate. B, Clinical pregnancy rate. C, Implantation rates. Reprinted from Vitagliano et al, 92 with permission from Elsevier Another possible strategy to modulate the reproductive tract microbiome is the use of probiotics. These are live biotherapeutic products containing one or more strains of the desired bacteria, in this case Lactobacillus , that are administered to colonize the niche while displacing potential dysbiotic bacteria. Several oral and vaginal probiotics are currently commercially available, the majority of them including L. crispatus , L. gasseri , Lactobacillus plantarum , Lactobacillus reuteri , and Lactobacillus rhamnosus . This could offer an interesting approach to restore a healthy LD microbiota while overcoming the disadvantages of antibiotic treatment such as antibiotic resistance, high rate of recurrent infections after treatment, and side effects derived from the clearance of endogenous off‐target flora in other body sites. 94 , 95 However, the efficacy of a probiotic therapy alone in reverting BV and other reproductive tract infections is not certain. For example, treatment with vaginal L. crispatus for one cycle results in colonization with this strain in up to 60% of patients. 96 A two‐step therapy with vaginal probiotics following antibiotic treatment could be useful to first fight the fastidious bacteria and then repopulate the reproductive tract with Lactobacillus strains. 97 , 98 , 99

The majority of studies on the female reproductive tract microbiome are focused on the vagina because of the ease of sampling; yet, several studies have demonstrated the existence of bacterial colonization beyond the vagina, showing that the upper reproductive tract is not sterile. For example, an active uterine microbiome has been characterized in healthy reproductive‐age women, 9 , 10 , 11 but bacteria have also been found to inhabit the fallopian tubes and the ovaries, 12 with Lactobacillus the most abundant genus throughout the female reproductive tract. 13 Recently, a study surveying the female reproductive tract confirmed the existence of a microbiota continuum starting in the vagina and progressing to the deepest organs in the tract—cervix, uterus, tubes, ovaries, and even colonizing the pouch of Douglas—in women with non‐infectious conditions. The microbiome of the vagina in healthy reproductive‐age women presents a biomass of approximately one billion bacteria per gram of vaginal fluid with low diversity, mainly composed of one or few Lactobacillus species, representing 90%‐95% of the total bacteria in the reproductive tract. 14 , 15 However, the vaginal microbiota is not dominated by Lactobacillus throughout a woman's lifetime. Indeed, in childhood, anaerobes and Escherichia coli predominate. 16 , 17 After puberty, the estrogen rise leads to the production and accumulation of glycogen, which is essential for Lactobacillus growth and the colonization of the vaginal epithelium; the dominance of Lactobacillus is maintained during the reproductive years. Finally, after menopause, the proportion of Lactobacillus species decreases again due to the drop in endogenous estrogen. Interestingly, the Lactobacillus content, as well as a low vaginal pH, is maintained in women receiving hormonal replacement therapy during menopause. 18 , 19 , 20 The first description of the vaginal microbiome in a set of 396 reproductive‐age women using NGS for the 16S rRNA bacterial gene revealed the existence of five distinct community state types (CSTs) depending on the abundance of the bacteria identified. CST‐I, ‐II, ‐III, and ‐V are dominated by Lactobacillus crispatus , Lactobacillus gasserii , Lactobacillus iners , and Lactobacillus jensenii , respectively. 21 The role of Lactobacillus is to maintain vaginal homeostasis by producing lactic acid to lower the vaginal pH. 22 , 23 This feature, together with the production of hydrogen peroxide, bacteriocins, and other antimicrobial compounds, facilitates the adhesion of Lactobacilli to the vaginal epithelial cells and competition for the nutrients in the niche to deter the growth of pathogenic bacteria. 24 , 25 , 26 , 27 , 28 In contrast, CST‐IV is characterized by increased diversity due to polymicrobial colonization with facultative anaerobic bacteria such as Gardnerella , Prevotella , Megasphaera , Atopobium , and Dialister to the detriment of Lactobacilli . 21 Shifts between different CSTs occur in women, sometimes involving acquisition of the CST‐IV microbiome. In fact, sexual activity and the transition through CST‐IV are risk factors for bacterial vaginosis (BV). 29 However, the CST‐IV profile is also common in asymptomatic women depending on their racial background. The percentage of women with a lower genital tract dominated by Lactobacillus is 90%, 80%, 60%, and 60%‐37% in White, Asian, Hispanic, and Black populations, respectively. 30 The variation of microbiota profiles in these populations may reflect not only racial or genetic predisposition to one or other types of bacteria, but also geographic, social, and/or economic factors. 30 Among the endogenous factors known to contribute to microbiome changes are hormonal changes during the menstrual cycle. These changes are associated with shifts in vaginal bacterial content, with menses representing the phase in which the microbiome is more diverse, while the oestradiol and progesterone peaks are more bacterially stable times. 31 Some external factors can also modulate the vaginal microbiota, such as hygiene habits, sexual exposure, change of sexual partners, and use and type contraceptives. 31 , 32 , 33 , 34 , 35 , 36 The factors influencing the structure and composition of the cervicovaginal microbiota were recently reviewed by Kroon and colleagues (Figure 1 ). 37 Factors influencing the composition of the cervicovaginal microbiota. Reprinted from Kroon et al, 37 with permission from Elsevier Our knowledge about the upper reproductive tract is not as wide, but it is now accepted that the internal organs also harbor a low but active biomass microbiota (Figure 2 ). The first evidence of bacterial colonization of the human uterine cavity was reported more than 30 years ago following cultivation of endometrial samples obtained either transcervically or after hysterectomy. 38 , 39 , 40 , 41 Bacteria (most often Lactobacillus spp, Mycoplasma hominis , Gardnerella vaginalis , and Enterobacter spp) were recovered in at least 25%‐30% of samples cultured. Later, molecular techniques such as polymerase chain reaction (PCR) were used to demonstrate that the upper genital tract, particularly the uterus, presents bacterial taxa other than those found in paired vaginal samples in both healthy women and women with BV. 9 In this study, 95% of the endometrial samples analyzed tested positive for bacterial DNA, although the overall number of recovered bacteria was significantly lower in endometrial than in vaginal samples. This observation was recently supported by the results of Chen and co‐workers, who quantified the bacterial load of samples collected along the reproductive tract using quantitative PCR and 16S rRNA NGS to show that the upper reproductive tract (peritoneal fluid and endometrium) contains 10 000 times less bacteria than the vagina. 42 The quantitative differences in bacterial load observed between the lower tract and the upper tract could be due to the cervical barrier, which may partially inhibit the ascension of bacteria from the vagina. Other hypotheses suggest a specific immune response in the internal organs or differential environmental conditions can lead to differential bacteria growth in both ends of the reproductive tract. 9 Despite bearing an ultralow biomass, this is an active microbiota, as demonstrated by the isolation of Lactobacillus —the most abundant bacteria in the reproductive tract— Actinomyces and Staphylococcus , among other bacterial genera, upon cultivation of fresh peritoneal fluid samples. 42 History of uterine microbiota investigation. Reprinted from Bakeret al, 100 under the terms of the Creative Commons Attribution License (CC BY) Despite some inconsistencies owing to differences in experimental design, several reports to date agree that the most abundant phyla in the uterine cavity are Firmicutes, Bacteroidetes, Proteobacteria, and Actinobacteria. 10 , 42 , 43 , 44 Moreover, the genus Lactobacillus has been consistently identified as the most represented taxa in the endometrium, 9 , 10 , 44 , 45 while Gardnerella , Streptococcus , Staphylococcus , Bifidobacterium , Prevotella , Atopobium , and Sneathia are present in smaller proportions. 10 , 11 , 45 Because it is difficult to obtain samples from the upper genital tract of healthy women, few studies have reported on the “normal” upper reproductive tract microbiome. Using 16S rRNA gene sequencing, the endometrial microbiome of fertile and healthy patients was investigated using minimally invasive methods for the collection of endometrial fluid, and this was compared to vaginal aspirates of the same subjects. This study confirmed that Lactobacillus is the most abundant genus in endometrial samples; Gardnerella , Bifidobacterium , Streptococcus , and Prevotella were also detected, all of which were previously identified in the lower reproductive tract. 10 Further, this study helped to classify the endometrial microbiota profile into Lactobacillus ‐dominated (LD) and non‐ Lactobacillus ‐dominated (NLD) according to the abundance of this bacterial genus (cutoff 90%) and its value to predict reproductive success (see below). Consistent with previous findings, women usually present similar microbiome profiles in the upper and lower genital tract. However, 20% of the women in whom the bacterial taxa were identified in endometrial samples showed significant differences in their paired vaginal aspirates. 10 Moreover, this study showed in 22 fertile and asymptomatic donors that the endometrial microbiota is stable during the 5‐day period in which endometrial receptivity—the ability to enable embryo implantation—is achieved, between the pre‐receptive state (2 days after the LH surge) and the receptive state (7 days after LH peak). 10 These results were corroborated in a study assessing vaginal and endometrial samples of healthy volunteers, non‐infertile patients, and patients receiving in vitro fertilization (IVF) treatment in Japan. This study showed that the healthy volunteers presented LD microbiota (≥90% Lactobacillus species), with 25% of the women presenting different taxonomic profiles in endometrial and vaginal samples. 11 When healthy volunteers were sampled again in different phases of the cycle or in the consecutive cycle, the microbial profiles were similar, suggesting that the endometrial microbiome may be stable over time. Importantly, however, the sample size was small, and other studies are needed to support this conclusion. 11 The results of these studies support the idea of a microbiota continuum along the reproductive tract and ascension from the vagina as the most plausible mode of upper genital tract colonization, as proposed after G. vaginalis biofilms were found in the fallopian tubes of patients diagnosed with BV. 46 Migration of microorganisms from other organs/tracts to the reproductive tract via hematogenous spreading has also been suggested based on the similarities between the cervical and gastrointestinal tract microbial features in obese patients, or the correspondence between oral and placental microbiota in pregnant women. 47 , 48

Mounting evidence demonstrates the importance of reproductive tract microbiota in women's health and reproductive function. This knowledge supports the practice of testing and treating, if required, the microbiota of the reproductive tract as a part of the personalized treatment in IVF settings. This practice is currently possible thanks to technical improvements that enable the study of bacterial communities at the molecular level, offering a new microbiological perspective at an unprecedented resolution. However, a deeper understanding of the microbiome will be useful to improve the gynecological and obstetrical health of every woman, regardless of her age, fertility status, or plans to conceive. In this regard, studies combining the most advanced technologies are required to decipher the interactions between host and microbiome to better define what a normal microbiome is, and the consequences of potential deviations from this normal microbiome.

Inmaculada Moreno is research manager at Igenomix SL, Spain. Carlos Simon is Chief Scientific Officer and own stocks in Igenomix SL. Human/animal rights : This article does not contain any studies with human and animal subjects performed by the any of the authors. Approval by ethics committee : This review does not contain original research, so approval by a suitable Ethics Committee is not applicable.

The female urogenital tract microbiota has long been interrogated using culture to assess the identity of the different microorganisms isolated from this body niche and their impact in reproductive pathophysiology. However, a catalogue of the microbial diversity inhabiting the female reproductive tract has only recently been generated, with the advent of sensitive molecular techniques such as next‐generation sequencing (NGS) of bacterial DNA. In the last decade, the Human Microbiome Project has enabled the study of the structure and composition of the microbiome at different body sites, revealing that the female reproductive tract microbiota accounts for approximately 9% of the total bacterial load in humans. This community predominantly comprises Lactobacilli in healthy women, although other genera have been identified, namely, Prevotella , Gardnerella , Atopobium , Sneathia , Bifidobacterium , Megasphaera , and Anaerococcus . 1 , 2 , 3 , 4 This is an interesting finding unique to the human reproductive tract microbiome, as other mammals present a vaginal microbiota not dominated by Lactobacillus . 5 Similarly, the male reproductive tract presents an active microbiome, as revealed by the presence of bacteria in seminal fluid samples from infertile men and healthy sperm donors. Interestingly, the bacterial communities found in the seminal samples are associated with semen health. In this regard, Lactobacillus may play a protective role; bacteria such as Anaerococcus , Pseudomonas , or Prevotella are mainly found in low‐quality sperm. 6 , 7 These data together suggest that the bacterial communities in the reproductive tract play important roles at different stages of the reproductive process, starting with gamete formation, fertilization, pregnancy establishment and maintenance, and even the microbial colonization of the newborn. 8 Based on this concept, various studies have sought to define the features and composition of a “normal/healthy” microbiome and establishing the potential shifts leading to a “dysbiotic/abnormal” microbiota. The aim of this review was to synthesize the current knowledge on the female reproductive tract microbiome and the role of bacteria in human reproduction.