How uterine microbiota might be responsible for a receptive, fertile endometrium

How uterine microbiota might be responsible for a receptive, fertile endometrium Abstract BACKGROUND Fertility depends on a receptive state of the endometrium, influenced by hormonal and anatomical adaptations, as well as the immune system. Local and systemic immunity is greatly influenced by microbiota. Recent discoveries of 16S rRNA in the endometrium and the ability to detect low-biomass microbiota fueled the notion that the uterus may be indeed a non-sterile compartment. To date, the concept of the ‘sterile womb’ focuses on in utero effects of microbiota on offspring and neonatal immunity. However, little awareness has been raised regarding the importance of uterine microbiota for endometrial physiology in reproductive health; manifested in fertility and placentation. OBJECTIVE AND RATIONALE Commensal colonization of the uterus has been widely discussed in the literature. The objective of this review is to outline the possible importance of this uterine colonization for a healthy, fertile uterus. We present the available evidence regarding uterine microbiota, focusing on recent findings based on 16S rRNA, and depict the possible importance of uterine colonization for a receptive endometrium. We highlight a possible role of uterine microbiota for host immunity and tissue adaptation, as well as conferring protection against pathogens. Based on knowledge of the interaction of the mucosal immune cells of the gut with the local microbiome, we want to investigate the potential implications of commensal colonization for uterine health. SEARCH METHODS PubMed and Google Scholar were searched for articles in English indexed from 1 January 2008 to 1 March 2018 for ‘16S rRNA’, ‘uterus’ and related search terms to assess available evidence on uterine microbiome analysis. A manual search of the references within the resulting articles was performed. To investigate possible functional contributions of uterine microbiota to health, studies on microbiota of other body sites were additionally assessed. OUTCOMES Challenging the view of a sterile uterus is in its infancy and, to date, no conclusions on a ‘core uterine microbiome’ can be drawn. Nevertheless, evidence for certain microbiota and/or associated compounds in the uterus accumulates. The presence of microbiota or their constituent molecules, such as polysaccharide A of the Bacteroides fragilis capsule, go together with healthy physiological function. Lessons learned from the gut microbiome suggest that the microbiota of the uterus may potentially modulate immune cell subsets needed for implantation and have implications for tissue morphology. Microbiota can also be crucial in protection against uterine infections by defending their niche and competing with pathogens. Our review highlights the need for well-designed studies on a ‘baseline’ microbial state of the uterus representing the optimal starting point for implantation and subsequent placenta formation. WIDER IMPLICATIONS The complex interplay of processes and cells involved in healthy pregnancy is still poorly understood. The correct receptive endometrial state, including the local immune environment, is crucial not only for fertility but also placenta formation since initiation of placentation highly depends on interaction with immune cells. Implantation failure, recurrent pregnancy loss, and other pathologies of endometrium and placenta, such as pre-eclampsia, represent an increasing societal burden. More robust studies are needed to investigate uterine colonization. Based on current data, future research needs to include the uterine microbiome as a relevant factor in order to understand the players needed for healthy pregnancy. immunology, cytokines, endometrium, implantation, uterus, pregnancy, menstrual cycle, recurrent miscarriage Introduction Maternal microbial colonization contributes to development of the unborn child by micronutrient provision, xenobiotic metabolism and enhancing maternal energy conversion (Macpherson et al., 2017). Currently, the concept of the ‘sterile womb’, the paradigm that the fetus grows up in a sterile environment, is highly debated (Perez-Munoz et al., 2017). If bacteria (or their compounds) were naturally present in the uterus, their role even before pregnancy, in maintenance of the uterus deserves attention. Various recent reviews focused on correlations between commensal uterine colonization, fertility problems and pregnancy complications (Franasiak and Scott, 2017; Moreno and Franasiak, 2017; Power et al., 2017; Prince et al., 2014). Bacteria are known to affect immunity (Hooper et al., 2012; Littman Dan and Pamer Eric, 2011). If they already impact the immune environment of the uterus before pregnancy, this would greatly impact the receptive potential of the endometrium, as well as the ability to correctly initialize placenta formation. Successful embryo implantation requires both a synchronous development and an intricate interplay between the hatched blastocyst and endometrium. This depends on a receptive state of the endometrium. In the first week after fertilization, the blastocyst makes initial contact with the highly specialized endometrium. During the female cycle, if the endometrium fails to undergo the proper adaptations to acquire this receptive state, infertility and impaired placentation may be the consequence. Since bacteria can play a role in morphological changes of, for example, mucosal cells, implications of microbiota in decidualization can be envisioned. At the same time, commensal microbiota may convey protection towards pathogenic species contributing to uterine health. Studies evaluating the endometrial microbiome and its role in fertility are limited and therefore we depend on extrapolating knowledge from other colonized body sites. A vast number of studies focusing on the gut microbiota have pointed out the contribution of microbes to immunity and development, and the necessity of commensal colonization to achieve a basal, healthy immune state (Littman Dan and Pamer Eric, 2011; Cho and Blaser, 2012; Geva-Zatorsky et al., 2017). We hypothesize that, if indeed present, the endometrial microbiome or its compounds can have implications for (pre-) decidualization, and therefore fertility. The objective of this review is to outline the available evidence regarding uterine microbiota, focusing on recent findings based on 16S rRNA, and to depict the possible importance of uterine colonization for a receptive endometrium. We aim to highlight a possible role of uterine microbiota in host immunity and tissue adaptation, as well as conferring protection against pathogens. Methods PubMed and Google Scholar were searched for articles in English indexed from 1 January 2008 to 1 March 2018 for ‘16S rRNA’, ‘uterus’ and related search terms (Supplementary Fig. SI) to assess available evidence on uterine microbiome analysis. Studies on the upper female genital tract were investigated regarding data specifically on the uterine compartment. A manual search of the references within the resulting articles was performed. Studies on microbiota of other body sites were additionally assessed. The non-sterile uterus? Microbiota and mammals depend on their symbiotic relationship. While microbes receive a steady nutrient supply through the host, the host benefits from vital contributions of the microbe to physiological processes such as epithelial homeostasis, and is supplied with a natural barrier against colonization by pathogenic species (Buchon et al., 2009; Eberl, 2010). Bacterial colonization also plays a crucial role in modulation of host immunity. Metagenomic analysis has propelled research on natural colonization of the human body forward, revealing microbiota at body sites that were previously considered to be sterile. This includes the upper reproductive tract and placenta, challenging the classic dogma of a ‘sterile womb’ as coined by Henry Tissier more than a century ago (Tissier, 1900). Due to the earlier limitations in microbial characterization and challenges in sample acquisition, the significance of endometrial bacteria may have been missed or overlooked (Viniker, 1999). Now, increasing evidence in favor of an endometrial microbiome (Franasiak and Scott, 2015; Fang et al., 2016; Khan et al., 2016; Moreno et al., 2016; Verstraelen et al., 2016; Walther-Antonio et al., 2016; Chen et al., 2017; Miles et al., 2017; Tao et al., 2017) and a placental commensal colonization (Aagaard et al., 2014; Prince et al., 2016) indicates the need for a paradigm shift. The endometrium: an immunologically suited niche for microbiota In the gut, the intestinal immune system needs to be highly adapted to withstand the continuous threat posed by colonization of the large mucosal surface, separated from host tissue by only a single layer of epithelial cells (Hooper and Macpherson, 2010). In a symbiotic colonization, bacterial growth is safe for the host as long as it is contained within an assigned compartment. Therefore, tissue invasion must be limited, thus preventing potentially harmful inflammation or disruption of the niche for the necessary symbiotic relationship. Hooper and Macpherson (2010) described three types of immunological barriers needed for intestinal microbial homeostasis: anatomically limiting exposure of resident bacteria to the systemic immune system; immune mediators restricting direct contact between epithelia and microbes; and rapid detection and killing of bacteria upon barrier breach. All three pre-requisites are met by the endometrial mucosa. In the uterus, the single layer of columnar epithelial cells that proliferate to form glandular cells in the secretory phase of the menstrual cycle forms a strong barrier through tight junctions (Wira et al., 2015). The uterine mucosal surface and the endometrial fluid (EF) contain infection-controlling molecules, known as antimicrobial peptides (AMPs), with fluctuating levels during the menstrual cycle (Wira et al., 2014). AMPs are known to contribute to the health of the female reproductive tract with implications for fertility and pregnancy (Frew and Stock, 2011). An example of an AMP found in the uterus is the secretory leukocyte protease inhibitor, which has antiviral and antifungal properties, and also acts as a bactericidal against gram-negative as well as gram-positive bacteria such as Escherichia coli and Staphylococcus aureus (King et al., 2000). Furthermore, the endometrial lymphocytes in the mucosal layer are present throughout all stages of the menstrual cycle, ready to act upon pathogen invasion (Givan et al., 1997). Therefore, according to the required properties, the uterus could offer a safe niche for symbiotic colonization. Assessing commensal uterine colonization Not only theoretically does the endometrium offer an environment suited for coexistence with bacteria but also, in accordance with earlier hypotheses, such as Viniker’s theory of the ‘bacteria endometrialis’ inhabiting the uterine cavity (Viniker, 1999), various approaches have now provided evidence for uterine colonization, as outlined below. Initial assessment of endometrial colonization was based on observations of cells in culture. It has long been thought that the few observed growing species from endometrial cultures after hysterectomy resulted from the sterile character of the uterus, and the bacteria that were found were ascribed to a pathogenic condition (Ansbacher et al., 1967; Moller et al., 1995). However, women with no sign of infection also showed some bacterial growth in samples maintained in vitro. Species such as Lactobacillus (Duff et al., 1983; Cowling et al., 1992), Gardnerella vaginalis, Enterobacter, and Streptococcus agalactiae (Moller, et al., 1995), E. coli and Enterococcus faecalis were typically found in patients with uterine pathologies, such as endometritis (Cicinelli et al., 2008). Other studies used catheter tips that were used in embryo transfer (ET) for microbial assessment based on culturing. Bacterial growth was associated with outcomes of IVF treatment, with varying results. If bacterial growth was observed upon culturing the catheter tips used for ET during IVF, live birth rates were found to be either decreased (Egbase et al., 1996, 1999; Fanchin et al., 1998; Salim et al., 2002), increased or not affected at all, depending on the study or bacterial species cultured (Moore et al., 2000; Selman et al., 2007). This underlines the need to carefully assess the outcomes and limitations of such an approach, as discussed below. However, even if smears led to growing cultures, pregnancy was possible (e.g. clinical pregnancy rate 28 versus 17% in the positive culture group, n = 279) (Fanchin et al., 1998). Also, using isolates other than from catheter tips, endometrial colonization could be shown for asymptomatic as well as symptomatic women (Koren and Spigland, 1978; Naessens et al., 1987; Stray-Pedersen et al., 1982). Taken together, microbial presence could not be ascribed to a certain pregnancy outcome based on culture-dependent methods. As we now know, culture-dependent characterization of microbial communities is associated with limitations (Supplementary Table SI). In culture-based approaches, rapidly growing, aerobic species dominate, leaving rare species that demand specific culture conditions undetected (Relman, 2002; Verstraelen et al., 2004). Molecular approaches allow detection of species that will not be revealed by culture-dependent techniques. Mitchell et al. (2015) examined uterine swabs and EF from hysterectomies by quantitative PCR (qPCR) for 12 bacterial species, including Atopobium vaginae, Prevotella spp., Lactobacillus crispatus, Lactobacillus iners, G. vaginalis and bacterial vaginosis-associated bacterium 1 (BVAB1). All of the selected species could be detected in vaginal samples and to a varying extent in the endometrium. Clear differences could be found between vaginal and endometrial samples. While A. vaginae was more commonly detected in vagina, L. iners and BVAB1 were more likely to be detected in endometrial samples. Of note, 95% of hysterectomy samples showed the presence of bacterial DNA. Also based on selected targets, Swidsinski et al. (2013) used fluorescent in-situ hybridization probes for G. vaginalis, A. vaginae, Lactobacillus, Bacteroides, Prevotella, Enterobacteriaceae and Eubacteria. Again, the endometrial microbiotic environment was shown to be different than that of the vagina. However, the need of these studies to select specific probes targeting a certain species is intrinsically biased. Rather than evaluating the presence and/or abundance of a certain species, amplicon sequencing of the hypervariable 16 S region of the ribosomal RNA makes it possible to identify species present in a sample (Clarridge, 2004). The metagenomic molecular approach of high-throughput 16 S rRNA sequencing allows for a more complete view, reflecting the diversity and relative abundance of microbiota. However, also this technique comes with its limitations as outlined below. While shotgun whole genome sequencing (WGS) offers advantages to amplicon 16S sequencing (Ranjan et al., 2016; Tessler et al., 2017), as yet this approach has not been employed to study endometrial microbiota. Evidence of endometrial microbiota by 16S sequencing In the Human Microbiome Project (HMP), efforts are made to understand more about the natural colonization of various body sites, its physiological importance and implications for disease (Peterson et al., 2009). In the context of the HMP, data on gut and vaginal colonization increase, but the upper reproductive tract has not been characterized extensively yet (Gevers et al., 2012). However, the importance of microbiota for human reproduction is increasingly acknowledged (Younes et al., 2017). Recently, several studies targeted a putative endometrial microbiome through 16S rRNA sequencing, and each of the studies documented the presence of uterine microbiota (Fang, et al., 2016; Franasiak et al., 2016; Khan, et al., 2016; Moreno, et al., 2016; Verstraelen, et al., 2016; Walther-Antonio, et al., 2016; Chen, et al., 2017; Miles, et al., 2017; Tao, et al., 2017). An overview of these recent studies assessing human uterine microbial composition based on 16S rRNA sequencing is given in Table I. Tips used for ET, swabs, biopsies or aspirates were employed to obtain endometrial samples. The studies used next-generation sequencing and assigned the sequence reads to different operational taxonomic units (OTUs). Phylotypes or, depending on the fit against the database used for taxonomic annotation, genotype or species could be assigned to the 16S rRNA gene amplicon sequence. Depending on the study, varying 16S rRNA regions were targeted simultaneously, since depending on which hypervariable V1–V9 region is targeted, more differentiation at genus or species level is possible (Chakravorty et al., 2007). Alpha (within a subject) and beta (between different subjects) diversity were assessed using different diversity indices (Hugerth and Andersson, 2017). One of the classifications often used to express alpha diversity and species richness (i.e. how different species are sequenced in total) is the Shannon Diversity Index. Another often-used index is Chao1, using rare classes of OTUs (Franasiak et al., 2016; Moreno et al., 2016). Beta diversity is often expressed through the Bray–Curtis dissimilarity (Walther-Antonio, et al., 2016). Taxonomic distribution profiles of identified genera or species are markedly different between the studies. Table I Studies presenting uterine microbiome assessment based on 16S rRNA. Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces CT, cycle threshold; GnRHa, GnRH agonist; OTU, operational taxonomic unit; PGM, personal genome machine; qPCR, quantitative PCR; RT, reproductive tract. Table I Studies presenting uterine microbiome assessment based on 16S rRNA. Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces CT, cycle threshold; GnRHa, GnRH agonist; OTU, operational taxonomic unit; PGM, personal genome machine; qPCR, quantitative PCR; RT, reproductive tract. Franasiak et al. (2016) characterized the endometrial microbiome of 33 women by collecting material from the tip of the catheter used for ET during ART. Both in the group of non-ongoing and ongoing pregnancies, Lactobacillus and Flavobacterium were identified as the most abundant genera. However, in this initial study only women undergoing ART for unspecified reasons were included. Therefore, it is unclear to what extent these findings represent the colonization of a healthy, fertile endometrium. The same holds true for the study by Tao et al. (2017) assessing the microbiota obtained from catheter tips during ET of 70 women. Lactobacillus was abundantly detected (>90% of OTUs in 33 women, >70% in 50 women). Samples also frequently contained Corynebacterium (40 women), Staphylococcus (38 women), Streptococcus (38 women) and Bifidobacterium (15 women). To achieve reliable sequencing results, particular attention was given to sample preparation, as sampling using remaining tissue from catheter tips only yields ultra-low bacterial cell counts. The authors were able to reliably determine bacterial abundance when more than 60 bacterial cells were present per sample, although no estimation was given of how many cells the catheter tips likely contain. Verstraelen et al. (2016) identified Bacteroides as the dominant genus, present in >90% of the women that were included. All women underwent hysteroscopy in the absence of uterine anomalies. Again, the cohort of women employed sought medical help due to recurrent implantation failure (RIF) or recurrent pregnancy loss. The 12 of the 19 women included showed 75% similarity in bacterial composition (Bray–Curtis dissimilarity of 24.6%, range 13.2–34.3%) due to general high abundance of Bacteroidetes and Proteobacteria taxa. Next to this core of Bacteroidetes and Proteobacteria, five women also presented with L. iners (n = 2), P. amnii (n = 1) or L. crispatus (n = 1) as the most abundant species. Large dissimilarities were observed in two women (79.0 and 90.7% Bray–Curtis dissimilarity) showing Prevotella spp. or L. crispatus dominance, which the authors ascribe to possible contamination from the vagina. Control swabs from the vagina could have given an indication if the distinctive microbial profile of these women indeed reflected an artifact from the transcervical sampling approach but no such data are available. In a study focusing on the relation of endometrial polyps to local microbiota, Fang et al. (2016) included 10 fertile women as a control population in their 16 S analysis of transcervical uterine swabs. Proteobacteria, Firmicutes and Actinobacteria were consistently found at phylum level; Enterobacter, Pseudomonas, and Lactobacillus at genus level. Moreno et al. (2016) were also able to include fertile women (n = 13 for comparison of vaginal aspirates and EF, n = 22 for compositional differences between pre-receptive and receptive phase) of reproductive age (Moreno, et al., 2016). Lactobacillus (71.1%), Gardnerella (12.6%), Bifidobacterium (3.7%), Streptococcus (3.2%) and Prevotella (0.9%) were the most identified genera. Endometrial microbial composition was categorized into either Lactobacillus-dominated (>90% Lactobacillus spp.) versus non-Lactobacillus dominated (>10% bacteria other than Lactobacillus, such as A. vaginae, G. vaginalis, species of the genera Clostridium, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia). The 18 of the 22 women presented with stable microbiota profiles when comparing pre-receptive and receptive phase, of which 12 women were continuously categorized in the Lactobacillus-dominated group, and six women categorized in the non-Lactobacillus group independent from sampling time point. This indicates that bacterial community composition was relatively stable in most women. Contradictory to the results by Franasiak et al. (2016); Moreno et al. (2016) found an association between Lactobacillus abundance and pregnancy outcome. Implantation was decreased (23.1 versus 60.7%), and pregnancy rates declined (13.3 versus 58.8%) when the women showed a non-Lactobacillus dominated endometrial phenotype at the time of ET. An especially negative impact on reproductive outcome was observed when G. vaginalis and Streptococcus species were present in abundance. These results were independent of pH of the sample, known to be affected by Lactobacillus species (Tachedjian et al., 2017). Unlike the other transcervical sampling approaches, Walther-Antonio et al. (2016) studied samples taken from the uterus, Fallopian tubes, and ovaries removed during hysterectomy (n = 31), in addition to pre-operative vaginal and cervical swabs and scrapes. Their results showed Shigella and Barnesiella as dominant species of the endometrial microbiome in accordance with earlier culture-based observations (Cicinelli, et al., 2008). High abundance of Staphylococcus, Blautia and Parabacteroides was found in benign uterine conditions (n = 10), and Bacteroides, and Faecalibacterium were associated with women who presented with cancer as reason for hysterectomy (n = 17). Miles et al. (2017) also took samples from various tissue types obtained during hysterectomy. Their data on 10 women, comprising a wide range of uterine pathologies, shows high variability. Species such as Lactobacillus, Acinetobacter, Blautia, Corynebacterium and Staphylococcus were abundantly found in some of the women. The authors also note that even after various attempts, no sequencing data could be obtained for endometrium from a patient with atrophic endometrium. Similarly, Chen et al. (2017) investigated the microbiota of different sites of the female reproductive tract accessed through surgery. Tissue was obtained during laparoscopy or laparotomy, therefore, not reflecting healthy, fertile women of reproductive age but the patients enrolled had conditions known not to involve infection. In these conditions, Lactobacillus, Pseudomonas, Acinetobacter, Vagococcus and Sphingobium were frequently detected. When evaluating the results of the recent 16 S endometrial studies, certain species were found in more than one study (Table I). Some of the described findings are addressed in two recent reviews (Franasiak and Scott, 2017; Moreno and Franasiak, 2017), neither including a concluding summary of the species found in the different studies, nor a meta-analysis. Microbiota in the uterus—real or artifact? Until now, available 16S data on uterine microbiota do not allow conclusions on a ‘core uterine microbiome’ due to a number of limitations. When investigating the questioned low-biomass microbiome, studies need to be designed with a special focus on possible contamination. In the following paragraph, we outline the pitfalls of the presented studies. Contamination acquired during sample acquisition and processing It is currently highly debated whether microbiota found to be involved in reproduction, such as in the placenta, are in fact merely the result of contamination and an artifact of the study design (Kliman, 2014). Contamination always has to be a concern when studying 16S data. Especially when investigating a suspected low-biomass microbiome, as found within the uterus, the impact of various (contaminating) handling steps makes it hard to detect low-abundance microbes originating only from the sampling site (van der Horst et al., 2013; Laurence et al., 2014; Salter et al., 2014). To overcome the hurdles associated with the need for highly sensitive detection, the 16S studies on endometrial colonization employed different controls (Table I). To examine the possible contribution of contamination when studying the placenta, Lauder et al. (2016) examined the results from ‘air swabs’ waved in the laboratory space, as well as unused sterile swabs, in comparison to placental tissue samples. They could not distinguish between contamination controls and placental samples. Compositions of microbiota in negative controls are known to be associated with the DNA extraction kits (Kim et al., 2017). Misrepresentation of data needs to be minimized by using special DNA isolation kits for low microbial biomass samples and committing to one kit type for all samples (Kim, et al., 2017). Low-biomass microbial communities are sensitive to contamination and misinterpretation Lauder et al. (2016) suggest the presence of a placental microbiome below their level of detection as a reason for not being able to distinguish between air swabs and sampled material. Detection of a low-biomass microbiome (where for 1 g of placental tissue only 20–2000 ng of bacterial DNA is estimated to be extracted) is a major issue when assessing endometrial microbiota (Antony et al., 2015). Chen et al. included an assessment of the bacterial biomass in their approach. Copy numbers, as calculated from qPCR on vaginal typical Lactobacillus species, were related to the relative abundance of the species as found in 16 S rRNA sequencing, showing that microbial numbers decrease from vagina (1010–1020 copies/sample) through the cervix (108–1010 copies/sample) to the uterus (102–103 copies/sample) (Chen, et al., 2017). As the authors point out, this still represents higher numbers than those that can potentially be detected in background noise (Salter, et al., 2014) or in their controls of sterile PBS, sterile saline, and ultrapure water. Accordingly, even though the detected biomass is small, the detected species are not an artifact or contamination. Another source of misinterpretation that is gaining attention is the approach in which species distribution is related to the obtained sequencing library. Disease-associated changes in Prevotella and Bacteroides abundances found in Crohn’s disease were found to be rooted in proportional profiling and had to be ascribed to a general decrease in microbiota richness in Crohn’s disease (Vandeputte et al., 2017). If abundance is interpreted as a relative proportion, a change in the presence of a certain species presents as increase or decrease of a possibly unrelated species. Low-biomass microbiota, as detected in the uterus, are highly sensitive to this misrepresentation of data. A quantitative assessment of genus abundance proportional to cell counts and microbial load of the sample would allow a more precise representation of data (Vandeputte, et al., 2017). Vaginal microbiota as a source of contamination Due to its high biomass relative to the uterine microbiome and the high abundance of Lactobacillus in the vagina, misrepresentation of species distribution cannot be excluded. Especially when using a transcervical approach, contamination by vaginal bacteria needs to be controlled for. Moreno et al. (2016) compared endometrial and vaginal samples, and only 2 of the 26 sample pairs (two pairs per woman) of EF and vaginal aspirate showed the same microbial profile. While in some samples only minor differences were observed, six paired samples exhibited totally different profiles when women belonged to the group of non-Lactobacilli dominated subjects. In nine of the 13 women, Lactobacillus consistently reflected the dominant genus with more than 90% colonization in the pre-receptive and receptive phase of endometrium and vagina. Notably, some common species, such as A. vaginae, G. vaginalis, Prevotella and Sneathia, of both endometrium and vagina were identified. Also, the study that circumvented this risk, by sampling from uteri obtained during hysterectomy, showed a correlation between the vaginal and uterine microbiomes (Walther-Antonio, et al., 2016). However, it is unclear how the uterine microenvironment is contaminated by transfer of cervicovaginal bacteria into the excised uterus. As manipulation of the uterus during surgery might enable passage of bacteria from the lower reproductive tract, it cannot be excluded that the correlation of vaginal and uterine microbiota is just an artifact. This correlation was also seen by Verstraelen et al. (2016), who detected endometrial phylotypes known to be associated with the vagina. Lactobacillus crispatus was present in 12 of the 19 women, and most abundant in 5 of 19 individuals (range from 17.1 to 79.1% of the total number of sequence reads). Also, species associated with vaginal microbiota imbalance, such as Prevotella (P. amnii, Prevotella sp., P. timonensis, P. disiens), were abundantly expressed in two subjects (18.4 and 55.4% of total sequence reads) (Verstraelen, et al., 2004; Martin and Marrazzo, 2016; Onderdonk et al., 2016). Likewise, G. vaginalis was present in six women, but at an abundance of <1%. While trying to avoid vaginal contamination through the use of a sheath-protected sampling surface and trying to prevent cervicovaginal contact of the sampling tool, it cannot be excluded that insertion or extraction of the device facilitates bacterial carry-over. Verstraelen et al. (2016) pointed out that, even though a certain degree of species overlap exists, it is not to be assumed that endometrial samples reflect vaginal colonization, since the diversity found in endometrial samples is higher than in the vagina. While we believe that it is unlikely that the distribution profiles from endometrial swabs are merely a representation of vaginal colonization, the possibility of carrying over traces from vaginal microbiota onto the sample cannot be neglected. Likewise, insertion of a sampling device or speculum could carry over bacteria from the lower reproductive tract into the uterus. It is also unclear how surgical access of the uterus coincided with disruption of the cervical barrier allowing passage of bacteria that would normally not reside in the uterus as a result of the pre-operative preparation. Live bacteria? It cannot be ignored that, without further measures, sequencing of 16S rRNA does not differentiate between living bacteria or (dead) bacterial fragments. Chen et al. studied cultures of freshly collected peritoneal fluid samples from the Pouch of Douglas (the lowest part of the abdominal cavity, between uterus and rectum), a site shown to be similar in microbial diversity, and even lower in terms of microbial abundance, as compared to the endometrium (Chen et al., 2017). Eight different isolates were cultured belonging to seven genera, such as Lactobacillus, Staphylococcus and Actinomyces, in 5 out of 15 samples, whereas various negative controls did not lead to bacterial growth when cultured. These results indicate that the detected 16S reads do not solely reflect dead bacterial fragments. Conclusions on the validity of the current evidence Based on the current study designs, no information on bacterial detection and distributions of a ‘core uterine microbiome’ can be extracted. Contamination during sampling by surrounding microbiota and reagent kits have not been addressed systematically on a large cohort of healthy, fertile women yet, in order to allow any conclusions on the continuous baseline colonization of the uterus independent of pathologies. Future studies also need to address the question of whether detected 16S results from bacterial fragments rather than live organisms by using a more robust approach than culture in vitro since (even though proving that culture is not impossible) often no growth can be established in cases of these ultra-low biomass samples. The variability in culturing success probably also results from variation between women in species composition and biomass of live bacteria. Aside from this, even 16S rRNA data, reflecting a proportion of the overall community that is dead or fragmented, represent ligands for the host cells to recognize and act upon. Thereby, these inactive bacterial fragments can still contribute to a physiologic interaction with host cells. We conclude that well-setup large cohort studies are needed to define a healthy uterine microbiome. We want to stress that even though no uterus-typical bacterial profile can be established yet, existing data are not to be underrated as evidence towards a natural microbial presence within the uterus. A number of the presented studies do employ negative controls and/or additional vaginal swabs. Since species abundance of these controls was not identical to the results from endometrial material, we should consider a physiological importance of local microbiota. Endometrium—starting point for decidualization, implantation and placentation To understand how microbiota could play a role in the uterine interplay of cells involved in the per-implantation period, we first need to understand how the endometrium forms the basis for successful implantation and placentation. Despite the growing interest from both clinicians and scientists in the process of implantation, the mechanisms underlying human implantation remain poorly understood. Each month, before the presence of a developing embryo, the stage for formation of the placenta is set. During the mid-to-late luteal phase spontaneous decidualization occurs; i.e. the transformation of the endometrium into a receptive state, independent of the presence or absence of a conceptus. And 6–10 days after ovulation (during Days 19–24 of the menstrual cycle) the receptive state of the endometrium is shaped. This is referred to as ‘the window of implantation’ (WOI), lasting for 2–4 days (Navot et al., 1991; Wilcox et al., 1999). The substantial physiological adaptations needed for correct cyclical changes of the endometrium are orchestrated by fluctuations of progesterone and oestrogen (Jabbour et al., 2006; Henriet et al., 2012), as well as by immune cells and their products (Salamonsen et al., 2002; Singh et al., 2011). Endometrial stromal fibroblasts differentiate to become larger, rounded fibroblast-like stromal cells. To achieve this, the cytoskeleton and plasma membrane undergo modifications (Thie et al., 1995; Martin et al., 2000; Murphy, 2000). The luminal epithelium must be able to interact with the blastocyst. Interactions also involve adhesion molecules, such as integrins, L-selectin ligands (e.g. L-selectin ligands expressed by the luminal epithelium and L-selectin receptors of the blastocyst) and oligosaccharides (Cross et al., 1994; Prakobphol et al., 2006; Cha et al., 2012; Kang et al., 2014). We summarized key endometrial adaptations in the peri-implantation period (Fig. 1) to show the importance of enabling the correct interaction of endometrium and blastocyst. Figure 1 View largeDownload slide Illustration of key elements in blastocyst–endometrium interaction needed for a receptive endometrium. Due to its important contributions during early placentation, the endometrium is key to healthy pregnancy. During the window of opportunity (Days 19–24 of the menstrual cycle), the endometrium resides in a receptive state that allows selection, apposition and attachment of a healthy blastocyst. Implantation is marked by invasive growth and differentiation of trophoblast cells. Proper interaction of invading trophoblast and local immunity is needed to achieve correct villi development and a healthy placenta. All these highly regulated properties of the endometrium needed in the initial phase of pregnancy are possibly affected by uterine microbiota. IVS, intervillous space; uNK, uterine natural killer cell. Figure 1 View largeDownload slide Illustration of key elements in blastocyst–endometrium interaction needed for a receptive endometrium. Due to its important contributions during early placentation, the endometrium is key to healthy pregnancy. During the window of opportunity (Days 19–24 of the menstrual cycle), the endometrium resides in a receptive state that allows selection, apposition and attachment of a healthy blastocyst. Implantation is marked by invasive growth and differentiation of trophoblast cells. Proper interaction of invading trophoblast and local immunity is needed to achieve correct villi development and a healthy placenta. All these highly regulated properties of the endometrium needed in the initial phase of pregnancy are possibly affected by uterine microbiota. IVS, intervillous space; uNK, uterine natural killer cell. Already at this stage, irreversible shortcomings in placental vascularization pose an origin for diseases such as pre-eclampsia or intra-uterine growth restriction (Conrad et al., 2017). On the maternal side, the underlying mechanisms regulating correct migration of the trophoblast and vascular remodeling may be rooted in the endometrium. Due to its importance for correct placentation, the endometrium is crucial for developing a healthy placenta, and therefore a healthy pregnancy (Burton et al., 2010). If microbial compounds constitute another physiologic player in the complex uterine environment, its natural impact will likely also contribute to implantation and placentation. Uterine immunity during implantation and placentation as a possible interaction partner of local microbiota Local immune cells belong to the factors known to influence extravillous trophoblast migration, and are therefore instrumental to maternal spiral artery remodeling (Moffett et al., 2015). The immune system is intimately involved in all aspects of the reproductive process, particularly around the time of conception, and in the peri-implantation period (Robertson et al., 2011). Different stages during the initiation of pregnancy hold specific immunologic challenges: to enable a healthy pregnancy, the maternal immune system must support the introduction of semen to allow fertilization, the initial contact of blastocyst and endometrium, and correct formation of the placenta (Robertson and Moldenhauer, 2014). All of these processes are subjects of extensive study, and a comprehensive overview is beyond the scope of a single review. Natural killer (NK) cells, T cells and antigen presenting cells (APCs) are a common focus when investigating the immunology of pregnancy (Emmer et al., 2000; Moffett and Colucci, 2014; Moffett et al., 2004; Plaks et al., 2008; Tilburgs et al., 2006, 2009; Trundley and Moffett, 2004; van der Meer et al., 2004). We wish to highlight the importance of a tightly regulated uterine immune environment that could potentially be affected by microbiota-induced factors. Recent discoveries on the interplay of the intestinal microbiome, their metabolites, and host immunity, give an impression of the vast impact that the endometrial microbiome could have (Hooper et al., 2012; Postler and Ghosh, 2017; Round and Mazmanian, 2009). We outline here the implications of gut microbiota affecting cell types known to be important within the uterus. Lymphocytes involved in the peri-implantation period NK cells are the most abundant of immune cells in the endometrium representing ~70% of all hematopoietic cells present (Bulmer et al., 1991; Moffett-King, 2002). The term endometrial NK cells (eNK; or uterine NK cells; uNK cells) refers to NK cells in the non-pregnant endometrium, and decidual NK (dNK) cells residing in the placental membrane are phenotypically and functionally different from eNK cells or peripheral blood NK (pbNK) cells (Moffett-King, 2002; Male et al., 2011). In the endometrium, the non-cytolytic and potent cytokine secretors, CD56brightCD16−NK cells, dominate. Only a small fraction of eNK cells belongs to the CD56+CD16+NK population specialized in killing infected cells (van der Meer, et al., 2004; Hanna and Mandelboim, 2007), contrary to the NK cell division in peripheral blood where the majority are CD56+CD16+. Throughout the female cycle, eNK cells continue to accumulate until the first half of gestation. Abnormal levels of NK cells, measured 20 days after ET, were shown to be associated with RIF (Ledee-Bataille et al., 2004). Direct involvement of eNK cells in implantation remains unclear, as successful implantation is possible in NK cell deficient mice (Le Bouteiller and Piccinni, 2008). More evidence points towards an involvement of dNK cells in placenta-formation processes after initial implantation. However, an indirect role of the human CD56brightCD16−eNK cells already during implantation cannot be neglected due to their high abundance and ability to secrete large amounts of cytokines, such as IFN-γ, tumor necrosis factor alpha (TNFα), granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin 10 (IL-10) (Hanna and Mandelboim, 2007). Just as eNK and dNK cells are distinct from pbNK cells, the mucosal NK cells of the gut that are continuously exposed to the gut microbiome are distinct from pbNK cells. This distinction is characterized by limited IFNγ production and an absence of cytotoxic effector perforin, granzymes, FasL and TNF-related apoptosis-inducing ligand (TRAIL) in mucosal NK cells of the gut (Satoh-Takayama et al., 2008). Of note, germ-free mice lacked these IL-22+NKp46+ cells (Sanos et al., 2009). Hence, the specialized intestinal NK cells might be the result of an adaptation to the microbiota-rich environment. It needs to be evaluated whether the reduced cytotoxic capacity of uNK cells (Manaster and Mandelboim, 2010) has any relation to the uterine microbiota. APCs, such as macrophages (Mɸ) and dendritic cells (DCs), constitute 10-20% of uterine leukocytes, expressing major histocompatibility complex class II molecules (Dekel et al., 2010; Mori et al., 2016). Through integrating microbial, environmental, and self-derived stimuli, APCs are key to the successful initiation of an appropriate immune response (Medzhitov and Janeway, 1997; Swiatczak and Rescigno, 2012). Throughout the menstrual cycle, mature CD83+ DCs and CD68+ Mɸ increase, peaking in the late secretory phase (Rieger et al., 2004). Mɸ are, together with NK cells, the main cytokine producers in the human endometrium (Kämmerer et al., 2004). Through secreting leukemia inhibitory factor (LIF) and IL1B, Mɸ play a known role in the establishment of endometrial receptivity by increasing cell surface fucosylated structures, allowing trophectoderm attachment (Robertson and Moldenhauer, 2014). Even more pronounced than seen for NK cells, monocytes of the heavily colonized gut mucosa are markedly different in phenotype and function compared to blood-derived monocytes. Even though they reside in close proximity to bacteria colonizing the intestinal lumen, the ability of the local macrophages to mount a pro-inflammatory response seems to be tuned down. Their expression of, for example, lipopolysaccharide (LPS)-binding protein CD14 receptor (Smith et al., 1997), CD89, implicated in IgA-enhanced phagocytosis is decreased (Smith et al., 2001). Also, various FCγ receptors (CD16, CD32, CD64), integrins (CD11a, CD11b, CD11c, CD18) and the IL-2 receptor CD25 are, amongst others, downregulated on intestinal macrophages. Their ability to produce cytokines (IL-1, IL-6, TNFα) is diminished (Smythies et al., 2005). It needs to be investigated whether this Smad-induced IκBα expression and NF-κB inactivation-mediated phenotype, termed ‘inflammation anergy’, is paralleled at the fetal–maternal interface (Smythies et al., 2010). T cells represent the third largest fraction of immune cells found in human endometrium (van der Molen et al., 2014; Shanmugasundaram et al., 2016; Feyaerts et al., 2017). Since, in the absence of pregnancy, T cells mostly reside in the deeper layers of the endometrium, their relation to fertility disorders more likely reflects a role in early placenta formation after implantation, rather than a contribution to the pre-implantation period. Experimental evidence on the functional importance of T cells in the WOI is scarce. It has been suggested that a switch in the ratio of Th1 and Th2 cells (Th1/Th2 paradigm) within the endometrium is needed to prepare for implantation, as a pronounced decrease in Th1 was accompanied by an increase in Th2 cells seen from the secretory phase towards early pregnancy (Saito et al., 1999). More recently, regulatory T (Treg) cells and Th17 cells have been included in the discussion (Figueiredo and Schumacher, 2016; Saito et al., 2010). Increasing evidence on the shaping of T cell subsets by microbiota based on studies of the gut is emerging (Omenetti and Pizarro, 2015). Like in the endometrium, mucosal immunity of the intestines depends on an interplay of IFN-γ producing Th1 cells, Th17 cells that secrete IL-17 and IL-22, and innate lymphoid cells with Th2-cytokine secreting functions, and suppressive Foxp3+ Treg cells. Differentiation of various T cell subsets is impaired in mice with disrupted gut microbiota (Ivanov et al., 2008; Gaboriau-Routhiau et al., 2009). Commensal segmented filamentous bacteria were shown to be important for Th17-mediated mucosal protection (Ivanov et al., 2009). A detailed understanding of the mechanisms underlying bacteria-mediated T cell induction is not yet available. One of the few examples in which the link between microbe and host is known, is polysaccharide A (PSA) originating from the capsule of Bacteroides fragilis (Ivanov, et al., 2009). Bacteroides, found in a number of the endometrial studies analyzing 16S rRNA, tunes the Th17 response of intestinal T cells, and causes a systemic increase of circulating CD4+ T and Th1 cells (Johnson et al., 2015; Mazmanian et al., 2005, 2008). Other microbiota are known to induce and cause accumulation of Treg cells, essential in maintaining tolerance and known to be important for pregnancy (Saito et al., 2005). In the gut, Treg mediated IL-10 production is thought to be critical to maintain intestinal microbial homeostasis by limiting continuous activation of Th1 and Th17 cells. Commensal microbiota promote de novo generation and activation of mucosal Treg cells (Geuking et al., 2011) as shown by mice colonized with altered Schaedler flora species (Lactobaillus, Bacteroides, species of Flexistipes phylum and gram-positive bacteria of the Firmicutes, Bacillus-Clostridium group) (Dewhirst et al., 1999; Geuking, et al., 2011). Focusing on specific strains, colonization of mice with Clostridia strains was shown to induce expansion of Treg cells locally in the lamina propria and systemically, resulting in increased resistance to colitis and IgE responses in adult mice (Atarashi et al., 2011). These intestinally induced Treg were shown to express a T-cell receptor repertoire specific for bacteria of the luminal content of colonized mice, allowing tolerance towards individual species (Lathrop et al., 2011). Similar generation and activation of immune cell types by endometrial microbiota could be essential for correct homeostasis regarding maintenance of a microbiome and programming the local immunity within the uterus. The majority of work on microbiota and immunity focuses on T cell homeostasis, but especially cells of the innate arm of the immune system are involved in the peri-implantation period. Cytokines and chemokines belong to the soluble factors known to be associated with reproductive health. Induced by either endometrial cells or immune cells recruited upon blastocyst encounter, the implantation period is characterized by an increase in pro-inflammatory Th1 cytokines, such as IL-6, IL-8, LIF and TNF-α, with implications for immune cell recruitment and activation of the endometrium, as reviewed previously (van Mourik et al., 2009; Dominguez et al., 2010). Throughout the female cycle, healthy endometrial stromal cells constitutively secrete the chemokine CCL2, also known as monocyte chemoattractant protein 1 involved in directing monocytes, T cells and DCs to inflammatory or tumor tissue (Carr et al., 1994) (Shi and Pamer, 2011; Li et al., 2012). Also, decidual stromal cells obtained during first trimester (~8 weeks in both studies) secrete CCL2, which was shown to attract Th2 and Th17 cells, thereby influencing the local T cell balance (He et al., 2012; Wu et al., 2014). In addition, macrophage recruitment and polarization (Sierra-Filardi et al., 2014), and plasmacytoid DCs were affected by CCL2. It was recently shown that microbiota are important in establishing baseline CCL2 secretion controlling homeostatic trafficking of plasmacytoid DCs (Swiecki et al., 2017). As no functional contribution of endometrial microbiota has been shown yet, the physiologically closest possible microbiome, as indicated by Chen et al. (2017), is probably the cervicovaginal microbiome. Cytokine concentrations were shown to be affected depending on the microbial profiles of 16S rRNA analysis of cervicovaginal lavage. This was reflected in lower IL-1beta, IL-8 and IL-10 levels in bacterial vaginosis (BV) (Santos-Greatti et al., 2016). Also other cytokines, such as LIF, GM-CSF, colony-stimulating factor-1, heparin binding epidermal growth factor-like growth factor, insulin-like growth factor Iand II, contribute to fertility through their role in supporting blastocyst development in the pre-implantation phase (Robertson and Moldenhauer, 2014). Decreased LIF expression of endometrial epithelial cells was shown to be associated with infertility (Margioula-Siarkou et al., 2016). Cervicovaginal levels of cytokines were affected by the bacterial flora found (Marconi et al., 2014) and a probiotic supplement was shown to affect vaginal interleukin changes (Bisanz et al., 2014). It remains to be seen if different uterine microbial profiles could be associated with altered chemokine/cytokine profiles and effects on reproductive success. Lessons learned from the gut microbiome: possible interactions between local microbiome and endometrium Besides influence on systemic and tissue resident lymphocytes, the microbiota and host interact on various levels. The relation between microbes and host is not a black and white phenomenon; and thus surpasses the simplified view of commensal versus pathogen. The immune system shapes a homeostatic coexistence, i.e. by eliminating species that might not only be harmful but also disturb the balance (Eberl, 2010). Two different types of battles are continuously fought when hosting a microbial community: one that is responsible for facing the daily challenge posed by containment of symbiotic microbes, and one for defense against pathogenic microbes that breach containment (Sansonetti, 2004; Eberl, 2010). Although our understanding of the microbiome–host interaction is still rudimentary, its functional importance, especially on various arms of innate and adaptive immunity, is evident (Cénit et al., 2014; Littman Dan and Pamer Eric, 2011). Below, we give an overview of the different mechanisms of how endometrial colonization might have implications for the uterine immune-environment, and thus for fertility. Pattern recognition receptors To detect viral, fungal and bacterial pathogens, the innate immune system senses pathogen-associated molecular patterns (PAMPs) through pattern recognition receptors (PRRs), such as toll-like receptors (TLRs), RIG-I-like receptors, NOD-like receptors (NLRs) and C-type lectin receptors, such as collectins, selectins, endocytic receptors and proteoglycans (Tschopp et al., 2003; Oger et al., 2009; Takeuchi and Akira, 2010; Botos et al., 2011). PAMPs include, for example, cell wall mannans (yeast), formylated or lipo-peptides, peptidoglycans, teichoic acids and bacterial cell-wall components such as LPS (Janeway and Medzhitov, 1998). Activation of TLR induces nuclear factor-kappa B, which initiates an inflammatory cascade including upregulation of markers to mount an adaptive immune response (Medzhitov et al., 1997). PRRs form the first line of defense against sexually transmitted disease or any other pathogens that can access the female reproductive tract through the vagina (Wira et al., 2005). For example, decidual and uterine stromal and epithelial cells express various TLRs and other PRRs, and the cells mount a potent inflammatory response upon receptor recognition (Young et al., 2004; Schaefer et al., 2005; Aflatoonian and Fazeli, 2008; Ghosh et al., 2013; D’Ippolito et al., 2016; Anders et al., 2017). An overview of TLRs of the endometrium is given in Table II. The expression of TLR mRNA and protein changes throughout the menstrual cycle, as shown by different groups (Jorgenson et al., 2005; Aflatoonian et al., 2007; Hirata et al., 2007; King et al., 2009; Nikzad et al., 2013). Receptor expression on endometrial cells was found to be low in the proliferative phase and increased in the secretory phase for TLRs 1–10. Together with intracellular receptors, there is a constitutively high expression of pathogenicity sensors lining the female reproductive tract (Ghosh, et al., 2013). All of these receptors can sense their specific PAMP and react by inducing signaling cascades. Table II Pattern recognition receptor expression on endometrial cells. Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) CpG, Cytosin-phosphatidyc-guanin; dsRNA/DNA, double-stranded RNA/DNA; EEC, endometrial epithelial cells; ESC, endometrial stromal cells; LPS, lipopolysaccharide; LTA, lipoteichoic acid; MDA-5, Melanoma Differentiation-Associated protein 5; nd, not defined; (x), PAMPs, pathogen-associated molecular pattern; presence shown for whole endometrial sample; RIG-1,retinoic acid-inducible gene 1; ssRNA, single-stranded RNA; TLR, toll-like receptor. Table II Pattern recognition receptor expression on endometrial cells. Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) CpG, Cytosin-phosphatidyc-guanin; dsRNA/DNA, double-stranded RNA/DNA; EEC, endometrial epithelial cells; ESC, endometrial stromal cells; LPS, lipopolysaccharide; LTA, lipoteichoic acid; MDA-5, Melanoma Differentiation-Associated protein 5; nd, not defined; (x), PAMPs, pathogen-associated molecular pattern; presence shown for whole endometrial sample; RIG-1,retinoic acid-inducible gene 1; ssRNA, single-stranded RNA; TLR, toll-like receptor. Evidence accumulates that PRRs are one way for microbiota and host to communicate. Round et al. (2011) examined how B. fragilis induces symbiosis via TLR2 of CD4+ T cells. PSA of B. fragilis induces Foxp3+ regulatory cells that are needed to counteract the otherwise Th17-mediated killing of Bacteroides (Round et al., 2011). Nod2-deficient mice were more susceptible to pathogenic infection (Petnicki-Ocwieja et al., 2009). Nod2 was shown to control dysregulated commensal colonization, and commensal resident bacteria are in turn needed to induce Nod2 expression. TLRs and NLRs are suspected to play a role in the periconceptual regulation as they are major players in the cascade of cytokine induction (Robertson, et al., 2011; Sirota et al., 2014). Protection against pathogens Maintaining colonization of commensal bacteria protects the host against pathogens. Residential microbiota successfully compete for the niche by exhibiting better adaptation than an occasionally invading pathogenic species, termed the ‘colonization resistance’ concept (Costello et al., 2012; Buffie and Pamer, 2013). Commensal microbiota are specialized in nutrient usage of their habitat, and deplete the environment of nutrients needed for pathogenic species. Consumption of a limited resource can starve the pathogenic invader (Kamada et al., 2013). E. coli, for example, competes in consumption of sugars, amino acids and other nutrients with the pathogenic enterohaemorrhagic E. coli (EHEC) (Fabich et al., 2008; Leatham et al., 2009). Symbionts also defend their niche (Ubeda et al., 2017). An example is the commensal Clostridium scindens whose colonization protects against C. difficile infection by production of secondary bile acids (Buffie et al., 2015). Another contribution by commensals is continuous receptor stimulation leading to TLR upregulation; needed to sense potentially dangerous bacteria (Vaishnava et al., 2008). Mice and humans treated with antibiotics, diminishing the intestinal microbiota, are highly susceptible to antibiotic-resistant strains because of a reduced expression of antimicrobial defense mechanisms (Brandl et al., 2008). In the female reproductive tract, inhibition of gonococci (Neisseria gonorrhoeae) adherence is observed in the presence of Lactobacilli when using an in vitro model of endometrial epithelial cells (Spurbeck and Arvidson, 2008). In an endocervical epithelial cell model, TLR agonists, as a surrogate for microbial products, were shown to stimulate antimicrobial products and mucins (Radtke et al., 2012). Maintaining commensal colonization within the endometrium may likely offer a similar means of effective protection against uterine infections. Tissue adaptation In addition to lymphocytes, epithelial cells play a crucial role in coexistence with commensal colonization. An intact epithelium allows for safe colonization without extensive (pathogenic) barrier breach, through a physical barrier and reaction of the epithelial cells towards bacterial ligands. On the other hand, bacteria also contribute to a healthy barrier development. Increasing evidence highlights the importance of microbiota for gut development and morphogenesis. Known adaptations of microbiota-implicated tissue adaptations include intestinal epithelial cell differentiation (Yu et al., 2016), support of epithelial cell regeneration (Banasaz et al., 2002), modulation of epithelial cell permeability (Cario et al., 2007), and vascularization (Stappenbeck et al., 2002). In the uterus, commensal bacteria might contribute to the remodeling needed for a receptive state of the endometrium. The genus Bacteroides, constituting 30% of the endometrial community in 90% of women, is known to be tightly involved in intestinal tissue development by influencing epithelial cell maturation and maintenance (Wexler, 2007; Maier et al., 2015). Hooper et al. (2001) showed that B. thetaiotaomicron induced gene expression of proteins involved in nutrient absorption, angiogenesis, intestinal maturation and mucosal barrier reinforcement. Bacteroides was shown to be essential for intestinal vascularization through signaling mediated by Paneth cells, the main type of epithelial cell in the intestine secreting antimicrobial compounds (Bevins and Salzman, 2011). Further studies using endometrial epithelial cells are warranted to ascribe similar physiological functions to the uterine microbiome during endometrial remodeling. The continuous reparative processes following the secretory phase might benefit from microbial support mechanisms. Dysbiosis in this critical period of preparing for blastocyst invasion might be one of the unrevealed causes leading to RIF due to failed remodeling. Another interesting tissue adaptation involved in the peri-implantation period is the change in barrier function of the endometrial epithelium. Features of tight junctions, such as area of whole junction and the area enclosed by junctional strands, decrease from Days 13 to 22 of the menstrual cycle, which could be beneficial for implantation (Murphy et al., 1992). Redistribution of adherens junctions and desmosomes during the WOI may also prepare for trophoblast invasion (Buck et al., 2012). The tight junctional barrier function is regulated by nutrients and cytokines (Turner, 2009). As a side effect of increased barrier permeability, more barrier breach of the uterine microbiota can occur. This in turn can lead to a more pro-inflammatory immune environment with mucosal immune cells stimulated to secrete the cytokines that are beneficial for the peri-implantation period. Discussion Conclusions on the possible presence of uterine colonization Even before the era of 16S RNA analysis, it was put forward that ‘it is difficult to envision that a mucosa that is continuously exposed to microorganisms present in the lower genital tract and that is regularly invaded by sperm that can carry microorganisms into the endometrial cavity, may be free of bacteria and its products such as endotoxin’ (Romero et al., 2004). We also agree that, based on the various lines of evidence for the existence of an endometrial microbiota, the era of the sterile uterus, free of any microbial compounds, has to come to an end. Taken together, this variety of immunological adaptations due to microbial presence contributes to establishing the right balance of tolerance versus reactivity towards bacterial ligands, and likely has a physiological impact on pregnancy (Fig. 2). Following the suggestion by Espinoza et al. (2006), it can be envisioned that as long as no pathologic pro-inflammatory response against the microorganisms of the endometrial mucosa is started, the presence of a certain microbial coexistence does not harm the conceptus, and might well contribute to vital processes in the peri-implantation period and pregnancy. Figure 2 View largeDownload slide Uterine microbiota may contribute to healthy endometrium physiology. Upon pattern recognition receptor stimulation, epithelial cells release soluble factors, such as cytokines, affecting local lymphocyte populations. If bacteria are naturally present in the uterus, similar interaction with host cells as seen in mucosa of the gut can be envisioned. Local lymphocytes, e.g. antigen presenting cells, could sense microbes by probing through the epithelia, or upon barrier breach, thereby initiating a signaling cascade. Presence of commensals might alter the mucosal T cell balance and the involved cytokines can have an effect on the local immune environment. Cells important for healthy implantation and placentation, such as uNK cells, are potentially affected. Therefore colonization might have an impact on the important initial steps of endometrial physiology. APC, antigen presenting cell; IL-10, interleukin 10; PSA, polysaccharide A; PRR, pattern recognition receptor; Th17, T helper 17 cells; Treg, regulatory T cell. Figure 2 View largeDownload slide Uterine microbiota may contribute to healthy endometrium physiology. Upon pattern recognition receptor stimulation, epithelial cells release soluble factors, such as cytokines, affecting local lymphocyte populations. If bacteria are naturally present in the uterus, similar interaction with host cells as seen in mucosa of the gut can be envisioned. Local lymphocytes, e.g. antigen presenting cells, could sense microbes by probing through the epithelia, or upon barrier breach, thereby initiating a signaling cascade. Presence of commensals might alter the mucosal T cell balance and the involved cytokines can have an effect on the local immune environment. Cells important for healthy implantation and placentation, such as uNK cells, are potentially affected. Therefore colonization might have an impact on the important initial steps of endometrial physiology. APC, antigen presenting cell; IL-10, interleukin 10; PSA, polysaccharide A; PRR, pattern recognition receptor; Th17, T helper 17 cells; Treg, regulatory T cell. Additional studies required to define the core microbiome Sample cohort Even though several studies have evaluated the microbiome of the endometrium, a ‘baseline’ or ‘core’ microbiome of this healthy endometrium has not been defined. The studies presented here on 16S rRNA evaluation of endometrial microbiota vary to a large extent in experimental approach. Differences in sample population, sample acquisition method, 16S region, controls and bacterial genotyping make it impossible to combine the data in a more powerful meta-analysis. Of note, all studies enrolled small and highly selected cohorts. Except for the studies by Fang et al. (2016) assessing 10 healthy women, and Moreno et al. (2016) including 22 healthy women, all studies enrolled subjects seeking treatment because of reproductive complications or uterine abnormalities (Fang, et al., 2016; Moreno, et al., 2016). Therefore, the majority of data does not reflect a putative steady-state healthy endometrial microbiome. Only recruitment of healthy, fertile women would provide insight into the microbiome of a fertile, receptive endometrium. However, without underlying medical reasons, sampling in healthy women is unethical and therefore a major limitation to investigating uterine microbiota. Controlling for vaginal and environmental contamination A transvaginal approach carries the risk of contamination by the cervicovaginal microbiome. Due to the high biomass of the vaginal relative to the uterine microbiome and the high abundance of Lactobacillus, misrepresentation of species distribution cannot be excluded. To achieve more conclusive results in the near future, transvaginal studies on endometrial microbiota need to strictly incorporate control samples from vagina and cervix. Additionally, contamination by reagents needs to be controlled for by including controls of all extraction kits used, a common source for misleading data (Glassing et al., 2016). Also, air swabs and swabs of other environmental sources of microbial exposure need to be taken into account when trying to establish conclusive results on the low-biomass microbiome of the uterus. Standardized sample processing In the presented studies, major differences in the processing of material, data collection and analyses are all sources of variation (Gohl et al., 2016). The choice of hypervariable region (V1–V9) of the 16S rRNA, containing species-specific sequences used for assessing microbial diversity, influences the detection results. V1 for example was found to be well suited to distinguish between S. aureus species, whereas V2 was more successful in Mycobacterium species differentiation (Chakravorty, et al., 2007). Depending on the variable region of the 16 S rRNA, different species are preferentially detected. A standardized approach in which study design, sampling method, DNA extraction with generation of 16 S amplicons and sequencing, assigning of OTUs to genus/species, and reporting of the findings is needed (Parnell et al., 2017) to achieve conclusive results for endometrial and placental microbiome studies. Without these standardized, well-controlled study designs the uterine microbiome will still not come into its own in the future. Points of interest Timing of intervention Owing to the dynamic fluctuations of microbiota in the uterus, the timing of microbial influence(s) on a healthy pregnancy needs attention. Depending on when the pregnant (or soon-to-be pregnant) woman undergoes any external microbial (pro-/pre or antibiotic) intervention, this might alter the physiologic processes of implantation, placenta formation or placenta maintenance if microbiota are involved. There is increasing evidence that BV is associated with infertility and pregnancy complications (Campisciano et al., 2017). Interestingly, the risk of delivering preterm because of BV is not reduced after clearance of infection by antibiotics (Pelzer et al., 2016). This might be explained by the fact that the effects of BV have already taken place and the initiated cascade of changes induced by the bacterial imbalance cannot be reversed. Accordingly, treatment of BV before 20 weeks of gestation might decrease the risk of BV-associated preterm birth, suggesting that the early phase of pregnancy in particular is sensitive to microbial impact (McDonald et al., 2007). Systematic reviews on other conditions of bacterial imbalance associated with pregnancy and their treatment might narrow down the window in which a certain microbe-associated effect on pregnancy takes place. In the future, robust evidence for a uterine microbiome might also fuel novel options for intervention concerning fertility and prevention of preterm birth risk in women with a history of BV. Chicken or egg It is possible that the state of inflammation is a sign of dysregulated immunity, rather than only a result of the presence of pathogenic, bacterial colonization. Chronic endometritis, the persistent inflammation of the endometrial lining, is associated with RIF (Johnston-MacAnanny et al., 2010) and successful antibiotic treatment of endometritis could clear the condition (as characterized by accumulation of plasma cells in biopsy samples), but clearance did not improve the chance of implantation. In this case, a certain microbial profile might not represent the cause but, instead, the effect of an immune state that is unfavorable for pregnancy. Interestingly, bacteria found in the endometrium causing the inflammation were not found in the vagina, suggesting that they are commensals in a dysbiotic state, rather than pathogenic intruders (Cicinelli, et al., 2008). However, at present, we cannot exclude an active effect of microbiota on the physiological changes needed for a receptive endometrium. Functional studies are needed to understand the host–microbe interaction in the uterus. More robust information on the species involved in uterine colonization will allow further functional assessment using in vitro endometrial models (Laniewski et al., 2017). Next to this, the association of bacterial metabolites as a cause or consequence of certain microbial colonization and the state of fertility needs to be examined. Yeoman et al. (2013) evaluated metabolomic profiling as a fast and cost-effective diagnostic tool for BV (Yeoman et al., 2013). It remains to be elucidated whether uterine metabolites can be a proxy for the infertility related to dysbiosis. Also, since naturally occurring microbes continuously breach the endothelial barrier and are a part of homeostasis, microbes will continuously shape the immune environment. How this impacts the tightly regulated immunological processes involved from endometrium preparation towards a healthy pregnancy is yet to be discovered. Dynamic bugs Recently, the importance of the microbiome for regulation of rhythmic biological changes has been stated (Thaiss et al., 2016). Fluctuating microbial community structures might direct hormonal changes (Kundu et al., 2017). It is tempting to speculate that similar dynamics are involved in the female cycle. Moreno et al. (2016) observed differences in the microbiota detected in the pre-receptive versus receptive phase in the endometrium. The extent of this effect varied. However, also for the vaginal microbiota it was shown that while in some women low constancy and high species turnover is observed, the extent of these changes can be very small for others (Gajer et al., 2012). Until now, sample size prohibits conclusions on putative characteristic changes of the endometrial microbiota associated with the induction of the WOI. More in-depth studies on changes in the endometrial microbiome are needed to assess whether microbiota play a role in orchestrating the cyclic endometrial changes and establishing a receptive endometrium. Studies examining 16S rRNA from endometrium alongside first trimester and term placental samples are needed to show if, and how far, the microbes of the fetal–maternal interface relate to the pre-pregnancy state. Origin The origin of the endometrial, and especially the placental, microbiome is highly discussed. Different routes of colonization have been suggested, including the gut, vagina or oral cavity. We believe that with placental development originating from decidualized endometrium, it would be highly unlikely that the endometrial and placental microbiome are completely independent. In the intimate relationship of endometrium/decidua and placenta, the trophoblast cells invading the uterine mucosal tissue have to be in contact with the residential microbiota. Systematically comparing the healthy, pre-pregnancy endometrium to placental samples will help to elucidate whether the placental colonization results from the bacteria already present in the uterus. A vaginal origin is one of the suggested routes of uterine colonization, as a number of species seem to be residents of both body sites. Transfer of labeled macrospheres by a ‘uterine peristaltic pump’ activity from vagina to uterus has been shown (Kunz and Beil, 1997; Zervomanolakis et al., 2007). Evidence from mouse studies points out that a change in TLRs and AMP expression and function allows the passage of vaginal bacteria through the cervix into the uterus upon cervical infection. Remarkably, in non-pregnant mice, bacteria could travel from the lower reproductive tract into the uterus, but this was prohibited during pregnancy. These findings suggest that a healthy, uninfected cervix, regulates the ascents of vaginal bacteria (Racicot et al., 2013). This theory is supported by earlier findings of differences in biochemical and chemical properties of the cervix depending on the pregnancy state (Hassan et al., 2010; Hein et al., 2002). In line with theories on oral transmission of bacteria into the uterus, an association of periodontal disease and pregnancy complications is highly discussed (Ren and Du, 2017). Indeed, bacteria such as Lactobacillus or A. vaginae that were found in the 16S assessment of the endometrial microbiome, are associated with caries, but not healthy plaques (Kumar et al., 2003; Aas et al., 2005). Interestingly, the placental microbiome resembles that of the oral cavity more than that of gut or even vagina (Aagaard, et al., 2014). Several studies also show an association of cytokines and bacteria from amniotic fluid with oral microbiota (Dortbudak et al., 2005; Han et al., 2006). It remains to be seen whether bacteria of the oral cavity can indeed travel via the circulatory system to selectively colonize the uterus before or during pregnancy (Zaura et al., 2014). Jimenez et al. (2008) elegantly showed that in utero bacterial colonization of the fetus takes place, which can be influenced by maternal oral uptake of microbiota. Pregnant mice were fed with a genetically labeled Enterococcus faecium strain isolated from human breast milk, and the bacteria could be detected in pups delivered by C-section. The authors point out that next to the possible oral route, APCs of the gut might sense and actively harvest intestinal bacteria, facilitating their spread to other body sites. This sampling of luminal bacteria by DCs has been shown for the gut (Rescigno et al., 2001). A similar way of harvesting might already take place before pregnancy to prepare the uterus for placentation. If and how microbiota of gut and uterus might interact remains to be established. The study by Jimenez et al. (2008) fits with the ‘in utero colonization hypothesis’, which in itself is heavily debated. Recently, Perez-Muños et al. (2017) have pointed out the shortcomings of published data surrounding the theory that newborns acquire first microbial colonization in utero. Studies on in utero colonization and microbial colonization of the non-pregnant uterus share the same methodological and technical limitations that have not been addressed sufficiently to allow any conclusions to be drawn about a core uterine microbiome. We wish to point that even though possibly not determining colonization of the pup in utero, as highlighted by the authors, maternal fertility or uterine health might have still have been affected by the microbial status. Importantly, a role in establishing protection against pathogenic invasion can be envisioned, a threat that mice in protected breeding facilities are a lot less exposed to than humans. Therefore, the fact that germ-free mice are fertile does not rule out a possible impact of uterine microbiota on normal physiology of the human endometrium. Conclusion Microbiota might be another piece in the complex mechanism contributing to the cogwheels of hormones, immune cells and physiological adaptations that are needed for successful pregnancy. As described in this review, a vast number of possible physiological contributions of the uterine microbiota can be inferred from other body sites. We highlight that the assessment of uterine microbiota still suffers from many limitations. Nevertheless, the available evidence shows that the presence of uterine 16S rRNA is not solely the result of sampling or analysis errors and deserves to be acknowledged. We conclude that the concept of the sterile endometrium, and the uterine compartment in general, is outworn, although the true core uterine microbiome still needs to be assessed. Functional studies are needed to elucidate the physiological importance of the microbiome in fertility. The challenge of studying reproductive immunology and the microbiota involved is that research on all of the different aspects is still in its infancy; microbiome, immunity, endocrinology in pregnancy, and placental and fetal development need to be studied together to obtain a more comprehensive overview. Thus, experts in various fields, such as microbiology and immunology, need to co-operate. Without this, such a basal process as pregnancy, needed for healthy offspring, will continue to be a mystery. Supplementary data Supplementary data are available at Human Reproduction Update online. Authors’ roles M.B. wrote the article. R.G.M., I.J. and G.F. reviewed and edited the article. All authors approved the final version of the article. Funding RIMLS (Radboud Institute for Molecular Life Sciences) grant to M.B. Conflict of interest All authors declare that they have no conflict of interest. References Aagaard K , Ma J , Antony KM , Ganu R , Petrosino J , Versalovic J . The placenta harbors a unique microbiome . Sci Transl Med 2014 ; 6 : 237ra265 . Aas JA , Paster BJ , Stokes LN , Olsen I , Dewhirst FE . Defining the normal bacterial flora of the oral cavity . J Clin Microbiol 2005 ; 43 : 5721 – 5732 . Google Scholar CrossRef Search ADS PubMed Aflatoonian R , Fazeli A . Toll-like receptors in female reproductive tract and their menstrual cycle dependent expression . J Reprod Immunol 2008 ; 77 : 7 – 13 . Google Scholar CrossRef Search ADS PubMed Aflatoonian R , Tuckerman E , Elliott SL , Bruce C , Aflatoonian A , Li TC , Fazeli A . Menstrual cycle-dependent changes of Toll-like receptors in endometrium . Hum Reprod 2007 ; 22 : 586 – 593 . Google Scholar CrossRef Search ADS PubMed Anders AP , Gaddy JA , Doster RS , Aronoff DM . Current concepts in maternal-fetal immunology: recognition and response to microbial pathogens by decidual stromal cells . Am J Reprod Immunol 2017 ; 77 . doi:10.1111/aji.12623 . Ansbacher R , Boyson WA , Morris JA . Sterility of the uterine cavity . Am J Obstet Gynecol 1967 ; 99 : 394 – 396 . Google Scholar CrossRef Search ADS PubMed Antony KM , Ma J , Mitchell KB , Racusin DA , Versalovic J , Aagaard K . The preterm placental microbiome varies in association with excess maternal gestational weight gain . Am J Obstet Gynecol 2015 ; 212 : 653.e651 – 616 . Atarashi K , Tanoue T , Shima T , Imaoka A , Kuwahara T , Momose Y , Cheng G , Yamasaki S , Saito T , Ohba Y et al. . Induction of colonic regulatory T cells by indigenous clostridium species . Science 2011 ; 331 : 337 – 341 . Google Scholar CrossRef Search ADS PubMed Banasaz M , Norin E , Holma R , Midtvedt T . Increased enterocyte production in gnotobiotic rats mono-associated with Lactobacillus rhamnosus GG . Appl Environ Microbiol 2002 ; 68 : 3031 – 3034 . Google Scholar CrossRef Search ADS PubMed Bevins CL , Salzman NH . Paneth cells, antimicrobial peptides and maintenance of intestinal homeostasis . Nat Rev Microbiol 2011 ; 9 : 356 – 368 . Google Scholar CrossRef Search ADS PubMed Bisanz JE , Seney S , McMillan A , Vongsa R , Koenig D , Wong L , Dvoracek B , Gloor GB , Sumarah M , Ford B et al. . A systems biology approach investigating the effect of probiotics on the vaginal microbiome and host responses in a double blind, placebo-controlled clinical trial of post-menopausal women . PLoS One 2014 ; 9 : e104511 . Google Scholar CrossRef Search ADS PubMed Botos I , Segal DM , Davies DR . The structural biology of toll-like receptors . Structure 2011 ; 19 : 447 – 459 . Google Scholar CrossRef Search ADS PubMed Brandl K , Plitas G , Mihu CN , Ubeda C , Jia T , Fleisher M , Schnabl B , DeMatteo RP , Pamer EG . Vancomycin-resistant enterococci exploit antibiotic-induced innate immune deficits . Nature 2008 ; 455 : 804 – 807 . Google Scholar CrossRef Search ADS PubMed Buchon N , Broderick NA , Chakrabarti S , Lemaitre B . Invasive and indigenous microbiota impact intestinal stem cell activity through multiple pathways in Drosophila . Genes Dev 2009 ; 23 : 2333 – 2344 . Google Scholar CrossRef Search ADS PubMed Buck VU , Windoffer R , Leube RE , Classen-Linke I . Redistribution of adhering junctions in human endometrial epithelial cells during the implantation window of the menstrual cycle . Histochem Cell Biol 2012 ; 137 : 777 – 790 . Google Scholar CrossRef Search ADS PubMed Buffie CG , Bucci V , Stein RR , McKenney PT , Ling L , Gobourne A , No D , Liu H , Kinnebrew M , Viale A et al. . Precision microbiome reconstitution restores bile acid mediated resistance to Clostridium difficile . Nature 2015 ; 517 : 205 – 208 . Google Scholar CrossRef Search ADS PubMed Buffie CG , Pamer EG . Microbiota-mediated colonization resistance against intestinal pathogens . Nat Rev Immunol 2013 ; 13 : 790 – 801 . Google Scholar CrossRef Search ADS PubMed Bulmer JN , Morrison L , Longfellow M , Ritson A , Pace D . Granulated lymphocytes in human endometrium: histochemical and immunohistochemical studies . Hum Reprod 1991 ; 6 : 791 – 798 . Google Scholar CrossRef Search ADS PubMed Burton GJ , Jauniaux E , Charnock-Jones DS . The influence of the intrauterine environment on human placental development . Int J Dev Biol 2010 ; 54 : 303 – 312 . Google Scholar CrossRef Search ADS PubMed Campisciano G , Florian F , D’Eustacchio A , Stankovic D , Ricci G , De Seta F , Comar M . Subclinical alteration of the cervical-vaginal microbiome in women with idiopathic infertility . J Cell Physiol 2017 ; 232 : 1681 – 1688 . Google Scholar CrossRef Search ADS PubMed Cario E , Gerken G , Podolsky DK . Toll-like receptor 2 controls mucosal inflammation by regulating epithelial barrier function . Gastroenterology 2007 ; 132 : 1359 – 1374 . Google Scholar CrossRef Search ADS PubMed Carr MW , Roth SJ , Luther E , Rose SS , Springer TA . Monocyte chemoattractant protein 1 acts as a T-lymphocyte chemoattractant . Proc Natl Acad Sci USA 1994 ; 91 : 3652 – 3656 . Google Scholar CrossRef Search ADS PubMed Cha J , Sun X , Dey SK . Mechanisms of implantation: strategies for successful pregnancy . Nat Med 2012 ; 18 : 1754 – 1767 . Google Scholar CrossRef Search ADS PubMed Chakravorty S , Helb D , Burday M , Connell N , Alland D . A detailed analysis of 16 S ribosomal RNA gene segments for the diagnosis of pathogenic bacteria . J Microbiol Methods 2007 ; 69 : 330 – 339 . Google Scholar CrossRef Search ADS PubMed Chen C , Song X , Wei W , Zhong H , Dai J , Lan Z , Li F , Yu X , Feng Q , Wang Z et al. . The microbiota continuum along the female reproductive tract and its relation to uterine-related diseases . Nat Commun 2017 ; 8 : 875 . Google Scholar CrossRef Search ADS PubMed Cho I , Blaser MJ . The human microbiome: at the interface of health and disease . Nat Rev Genet 2012 ; 13 : 260 – 270 . Google Scholar CrossRef Search ADS PubMed Cicinelli E , De Ziegler D , Nicoletti R , Colafiglio G , Saliani N , Resta L , Rizzi D , De Vito D . Chronic endometritis: correlation among hysteroscopic, histologic, and bacteriologic findings in a prospective trial with 2190 consecutive office hysteroscopies . Fertil Steril 2008 ; 89 : 677 – 684 . Google Scholar CrossRef Search ADS PubMed Clarridge JE . Impact of 16 S rRNA gene sequence analysis for identification of bacteria on clinical microbiology and infectious diseases . Clin Microbiol Rev 2004 ; 17 : 840 – 862 . Google Scholar CrossRef Search ADS PubMed Conrad KP , Rabaglino MB , Post Uiterweer ED . Emerging role for dysregulated decidualization in the genesis of preeclampsia . Placenta 2017 ; 60 : 119 – 129 . Google Scholar CrossRef Search ADS PubMed Costello EK , Stagaman K , Dethlefsen L , Bohannan BJ , Relman DA . The application of ecological theory toward an understanding of the human microbiome . Science 2012 ; 336 : 1255 – 1262 . Google Scholar CrossRef Search ADS PubMed Cowling P , McCoy DR , Marshall RJ , Padfield CJ , Reeves DS . Bacterial colonization of the non-pregnant uterus: a study of pre-menopausal abdominal hysterectomy specimens . Eur J Clin Microbiol Infect Dis 1992 ; 11 : 204 – 205 . Google Scholar CrossRef Search ADS PubMed Cross JC , Werb Z , Fisher SJ . Implantation and the placenta: key pieces of the development puzzle . Science 1994 ; 266 : 1508 – 1518 . Google Scholar CrossRef Search ADS PubMed Cénit MC , Matzaraki V , Tigchelaar EF , Zhernakova A . Rapidly expanding knowledge on the role of the gut microbiome in health and disease . Biochim Biophys Acta 2014 ; 1842 : 1981 – 1992 . Google Scholar CrossRef Search ADS PubMed Dekel N , Gnainsky Y , Granot I , Mor G . Inflammation and Implantation . Am J Reprod Immunol 2010 ; 63 : 17 – 21 . Google Scholar CrossRef Search ADS PubMed Dewhirst FE , Chien CC , Paster BJ , Ericson RL , Orcutt RP , Schauer DB , Fox JG . Phylogeny of the defined murine microbiota: altered Schaedler flora . Appl Environ Microbiol 1999 ; 65 : 3287 – 3292 . Google Scholar PubMed Dominguez F , Gadea B , Mercader A , Esteban FJ , Pellicer A , Simon C . Embryologic outcome and secretome profile of implanted blastocysts obtained after coculture in human endometrial epithelial cells versus the sequential system . Fertil Steril 2010 ; 93 : 774 – 782.e771 . Google Scholar CrossRef Search ADS PubMed Dortbudak O , Eberhardt R , Ulm M , Persson GR . Periodontitis, a marker of risk in pregnancy for preterm birth . J Clin Periodontol 2005 ; 32 : 45 – 52 . Google Scholar CrossRef Search ADS PubMed Duff P , Gibbs RS , Blanco JD , Clair PJS . Endometrial culture techniques in puerperal patients . Obstet Gynecol 1983 ; 61 : 217 – 222 . Google Scholar PubMed D’Ippolito S , Tersigni C , Marana R , Di Nicuolo F , Gaglione R , Rossi ED , Castellani R , Scambia G , Di Simone N . Inflammosome in the human endometrium: further step in the evaluation of the ‘maternal side’ . Fertil Steril 2016 ; 105 : 111-118.e111 – 114 . Eberl G . A new vision of immunity: homeostasis of the superorganism . Mucosal Immunol 2010 ; 3 : 450 – 460 . Google Scholar CrossRef Search ADS PubMed Egbase PE , al-Sharhan M , al-Othman S , al-Mutawa M , Udo EE , Grudzinskas JG . Incidence of microbial growth from the tip of the embryo transfer catheter after embryo transfer in relation to clinical pregnancy rate following in-vitro fertilization and embryo transfer . Hum Reprod 1996 ; 11 : 1687 – 1689 . Google Scholar CrossRef Search ADS PubMed Egbase PE , Udo EE , Al-Sharhan M , Grudzinskas JG . Prophylactic antibiotics and endocervical microbial inoculation of the endometrium at embryo transfer . Lancet 1999 ; 354 : 651 – 652 . Google Scholar CrossRef Search ADS PubMed Emmer PM , Nelen WL , Steegers EA , Hendriks JC , Veerhoek M , Joosten I . Peripheral natural killer cytotoxicity and CD56(pos)CD16(pos) cells increase during early pregnancy in women with a history of recurrent spontaneous abortion . Hum Reprod 2000 ; 15 : 1163 – 1169 . Google Scholar CrossRef Search ADS PubMed Espinoza J , Erez O , Romero R . Preconceptional antibiotic treatment to prevent preterm birth in women with a previous preterm delivery . Am J Obstet Gynecol 2006 ; 194 : 630 – 637 . Google Scholar CrossRef Search ADS PubMed Fabich AJ , Jones SA , Chowdhury FZ , Cernosek A , Anderson A , Smalley D , McHargue JW , Hightower GA , Smith JT , Autieri SM et al. . Comparison of carbon nutrition for pathogenic and commensal Escherichia coli strains in the mouse intestine . Infect Immun 2008 ; 76 : 1143 – 1152 . Google Scholar CrossRef Search ADS PubMed Fanchin R , Harmas A , Benaoudia F , Lundkvist U , Olivennes F , Frydman R . Microbial flora of the cervix assessed at the time of embryo transfer adversely affects in vitro fertilization outcome . Fertil Steril 1998 ; 70 : 866 – 870 . Google Scholar CrossRef Search ADS PubMed Fang R-L , Chen L-X , Shu W-S , Yao S-Z , Wang S-W , Chen Y-Q . Barcoded sequencing reveals diverse intrauterine microbiomes in patients suffering with endometrial polyps . Am J Transl Res 2016 ; 8 : 1581 – 1592 . Google Scholar PubMed Feyaerts D , Benner M , van Cranenbroek B , van der Heijden OWH , Joosten I , van der Molen RG . Human uterine lymphocytes acquire a more experienced and tolerogenic phenotype during pregnancy . Sci Rep 2017 ; 7 : 2884 . Google Scholar CrossRef Search ADS PubMed Figueiredo AS , Schumacher A . The T helper type 17/regulatory T cell paradigm in pregnancy . Immunology 2016 ; 148 : 13 – 21 . Google Scholar CrossRef Search ADS PubMed Franasiak JM , Scott RT Jr . Reproductive tract microbiome in assisted reproductive technologies . Fertil Steril 2015 ; 104 : 1364 – 1371 . Google Scholar CrossRef Search ADS PubMed Franasiak JM , Scott RT . Endometrial microbiome . Curr Opin Obstet Gynecol 2017 ; 29 : 146 – 152 . Google Scholar CrossRef Search ADS PubMed Franasiak JM , Werner MD , Juneau CR , Tao X , Landis J , Zhan Y , Treff NR , Scott RT . Endometrial microbiome at the time of embryo transfer: next-generation sequencing of the 16 S ribosomal subunit . J Assist Reprod Genet 2016 ; 33 : 129 – 136 . Google Scholar CrossRef Search ADS PubMed Frew L , Stock SJ . Antimicrobial peptides and pregnancy . Reproduction 2011 ; 141 : 725 – 735 . Google Scholar CrossRef Search ADS PubMed Gaboriau-Routhiau V , Rakotobe S , Lecuyer E , Mulder I , Lan A , Bridonneau C , Rochet V , Pisi A , De Paepe M , Brandi G et al. . The key role of segmented filamentous bacteria in the coordinated maturation of gut helper T cell responses . Immunity 2009 ; 31 : 677 – 689 . Google Scholar CrossRef Search ADS PubMed Gajer P , Brotman RM , Bai G , Sakamoto J , Schutte UM , Zhong X , Koenig SS , Fu L , Ma ZS , Zhou X et al. . Temporal dynamics of the human vaginal microbiota . Sci Transl Med 2012 ; 4 : 132ra152 . Google Scholar CrossRef Search ADS Geuking MB , Cahenzli J , Lawson MA , Ng DC , Slack E , Hapfelmeier S , McCoy KD , Macpherson AJ . Intestinal bacterial colonization induces mutualistic regulatory T cell responses . Immunity 2011 ; 34 : 794 – 806 . Google Scholar CrossRef Search ADS PubMed Geva-Zatorsky N , Sefik E , Kua L , Pasman L , Tan TG , Ortiz-Lopez A , Yanortsang TB , Yang L , Jupp R , Mathis D et al. . Mining the human gut microbiota for immunomodulatory organisms . Cell 2017 ; 168 : 928 – 943.e11 . Google Scholar CrossRef Search ADS PubMed Gevers D , Knight R , Petrosino JF , Huang K , McGuire AL , Birren BW , Nelson KE , White O , Methé BA , Huttenhower C . The Human Microbiome Project: a community resource for the healthy human microbiome . PLoS Biol 2012 ; 10 : e1001377 . Google Scholar CrossRef Search ADS PubMed Ghosh M , Shen Z , Fahey JV , Crist SG , Patel M , Smith JM , Wira CR . Pathogen recognition in the human female reproductive tract: expression of intracellular cytosolic sensors NOD1, NOD2, RIG-1, and MDA5 and response to HIV-1 and Neisseria gonorrhea . Am J Reprod Immunol 2013 ; 69 : 41 – 51 . Google Scholar CrossRef Search ADS PubMed Givan AL , White HD , Stern JE , Colby E , Gosselin EJ , Guyre PM , Wira CR . Flow cytometric analysis of leukocytes in the human female reproductive tract: comparison of fallopian tube, uterus, cervix, and vagina . Am J Reprod Immunol 1997 ; 38 : 350 – 359 . Google Scholar CrossRef Search ADS PubMed Glassing A , Dowd SE , Galandiuk S , Davis B , Chiodini RJ . Inherent bacterial DNA contamination of extraction and sequencing reagents may affect interpretation of microbiota in low bacterial biomass samples . Gut Pathog 2016 ; 8 : 24 . Google Scholar CrossRef Search ADS PubMed Gohl DM , Vangay P , Garbe J , MacLean A , Hauge A , Becker A , Gould TJ , Clayton JB , Johnson TJ , Hunter R et al. . Systematic improvement of amplicon marker gene methods for increased accuracy in microbiome studies . Nat Biotechnol 2016 ; 34 : 942 – 949 . Google Scholar CrossRef Search ADS PubMed Han YW , Ikegami A , Bissada NF , Herbst M , Redline RW , Ashmead GG . Transmission of an uncultivated Bergeyella strain from the oral cavity to amniotic fluid in a case of preterm birth . J Clin Microbiol 2006 ; 44 : 1475 – 1483 . Google Scholar CrossRef Search ADS PubMed Hanna J , Mandelboim O . When killers become helpers . Trends Immunol 2007 ; 28 : 201 – 206 . Google Scholar CrossRef Search ADS PubMed Hassan SS , Romero R , Pineles B , Tarca AL , Montenegro D , Erez O , Mittal P , Kusanovic JP , Mazaki-Tovi S , Espinoza J et al. . MicroRNA expression profiling of the human uterine cervix after term labor and delivery . Am J Obstet Gynecol 2010 ; 202 : 80.e81 – 88 . Google Scholar CrossRef Search ADS He YY , He XJ , Guo PF , Du MR , Shao J , Li MQ , Li DJ . The decidual stromal cells-secreted CCL2 induces and maintains decidual leukocytes into Th2 bias in human early pregnancy . Clin Immunol 2012 ; 145 : 161 – 173 . Google Scholar CrossRef Search ADS PubMed Hein M , Valore EV , Helmig RB , Uldbjerg N , Ganz T . Antimicrobial factors in the cervical mucus plug . Am J Obstet Gynecol 2002 ; 187 : 137 – 144 . Google Scholar CrossRef Search ADS PubMed Henriet P , Gaide Chevronnay HP , Marbaix E . The endocrine and paracrine control of menstruation . Mol Cell Endocrinol 2012 ; 358 : 197 – 207 . Google Scholar CrossRef Search ADS PubMed Hirata T , Osuga Y , Hamasaki K , Hirota Y , Nose E , Morimoto C , Harada M , Takemura Y , Koga K , Yoshino O et al. . Expression of toll-like receptors 2, 3, 4, and 9 genes in the human endometrium during the menstrual cycle . J Reprod Immunol 2007 ; 74 : 53 – 60 . Google Scholar CrossRef Search ADS PubMed Hooper LV , Littman DR , Macpherson AJ . Interactions between the microbiota and the immune system . Science 2012 ; 336 : 1268 – 1273 . Google Scholar CrossRef Search ADS PubMed Hooper LV , Macpherson AJ . Immune adaptations that maintain homeostasis with the intestinal microbiota . Nat Rev Immunol 2010 ; 10 : 159 – 169 . Google Scholar CrossRef Search ADS PubMed Hooper LV , Wong MH , Thelin A , Hansson L , Falk PG , Gordon JI . Molecular analysis of commensal host-microbial relationships in the intestine . Science 2001 ; 291 : 881 – 884 . Google Scholar CrossRef Search ADS PubMed Hugerth LW , Andersson AF . Analysing microbial community composition through amplicon sequencing: from sampling to hypothesis testing . Front Microbiol 2017 ; 8 : 1561 . Google Scholar CrossRef Search ADS PubMed Ivanov II , Atarashi K , Manel N , Brodie EL , Shima T , Karaoz U , Wei D , Goldfarb KC , Santee CA , Lynch SV et al. . Induction of intestinal Th17 cells by segmented filamentous bacteria . Cell 2009 ; 139 : 485 – 498 . Google Scholar CrossRef Search ADS PubMed Ivanov II , Frutos Rde L , Manel N , Yoshinaga K , Rifkin DB , Sartor RB , Finlay BB , Littman DR . Specific microbiota direct the differentiation of IL-17-producing T-helper cells in the mucosa of the small intestine . Cell Host Microbe 2008 ; 4 : 337 – 349 . Google Scholar CrossRef Search ADS PubMed Jabbour HN , Kelly RW , Fraser HM , Critchley HO . Endocrine regulation of menstruation . Endocr Rev 2006 ; 27 : 17 – 46 . Google Scholar CrossRef Search ADS PubMed Janeway CA Jr. , Medzhitov R . Introduction: the role of innate immunity in the adaptive immune response . Semin Immunol 1998 ; 10 : 349 – 350 . Google Scholar CrossRef Search ADS PubMed Jimenez E , Marin ML , Martin R , Odriozola JM , Olivares M , Xaus J , Fernandez L , Rodriguez JM . Is meconium from healthy newborns actually sterile? Res Microbiol 2008 ; 159 : 187 – 193 . Google Scholar CrossRef Search ADS PubMed Johnson JL , Jones MB , Cobb BA . Polysaccharide A from the capsule of Bacteroides fragilis induces clonal CD4+ T cell expansion . J Biol Chem 2015 ; 290 : 5007 – 5014 . Google Scholar CrossRef Search ADS PubMed Johnston-MacAnanny EB , Hartnett J , Engmann LL , Nulsen JC , Sanders MM , Benadiva CA . Chronic endometritis is a frequent finding in women with recurrent implantation failure after in vitro fertilization . Fertil Steril 2010 ; 93 : 437 – 441 . Google Scholar CrossRef Search ADS PubMed Jorgenson RL , Young SL , Lesmeister MJ , Lyddon TD , Misfeldt ML . Human endometrial epithelial cells cyclically express Toll-like receptor 3 (TLR3) and exhibit TLR3-dependent responses to dsRNA . Hum Immunol 2005 ; 66 : 469 – 482 . Google Scholar CrossRef Search ADS PubMed Kamada N , Chen GY , Inohara N , Nunez G . Control of pathogens and pathobionts by the gut microbiota . Nat Immunol 2013 ; 14 : 685 – 690 . Google Scholar CrossRef Search ADS PubMed Kang YJ , Forbes K , Carver J , Aplin JD . The role of the osteopontin-integrin alphavbeta3 interaction at implantation: functional analysis using three different in vitro models . Hum Reprod 2014 ; 29 : 739 – 749 . Google Scholar CrossRef Search ADS PubMed Khan KN , Fujishita A , Masumoto H , Muto H , Kitajima M , Masuzaki H , Kitawaki J . Molecular detection of intrauterine microbial colonization in women with endometriosis . Eur J Obstet Gynecol Reprod Biol 2016 ; 199 : 69 – 75 . Google Scholar CrossRef Search ADS PubMed Kim D , Hofstaedter CE , Zhao C , Mattei L , Tanes C , Clarke E , Lauder A , Sherrill-Mix S , Chehoud C , Kelsen J et al. . Optimizing methods and dodging pitfalls in microbiome research . Microbiome 2017 ; 5 : 52 . Google Scholar CrossRef Search ADS PubMed King AE , Critchley HO , Kelly RW . Presence of secretory leukocyte protease inhibitor in human endometrium and first trimester decidua suggests an antibacterial protective role . Mol Hum Reprod 2000 ; 6 : 191 – 196 . Google Scholar CrossRef Search ADS PubMed King AE , Horne AW , Hombach-Klonisch S , Mason JI , Critchley HO . Differential expression and regulation of nuclear oligomerization domain proteins NOD1 and NOD2 in human endometrium: a potential role in innate immune protection and menstruation . Mol Hum Reprod 2009 ; 15 : 311 – 319 . Google Scholar CrossRef Search ADS PubMed Kliman HJ . Comment on ‘The placenta harbors a unique microbiome’ . Sci Transl Med 2014 ; 6 : 254le254 . Google Scholar CrossRef Search ADS Koren Z , Spigland I . Irrigation technique for detection of Mycoplasma intrauterine infection in infertile patients . Obstet Gynecol 1978 ; 52 : 588 – 590 . Google Scholar PubMed Kumar PS , Griffen AL , Barton JA , Paster BJ , Moeschberger ML , Leys EJ . New bacterial species associated with chronic periodontitis . J Dent Res 2003 ; 82 : 338 – 344 . Google Scholar CrossRef Search ADS PubMed Kundu P , Blacher E , Elinav E , Pettersson S . Our gut microbiome: the evolving inner self . Cell 2017 ; 171 : 1481 – 1493 . Google Scholar CrossRef Search ADS PubMed Kunz G , Beil D , Deiniger H , Einspanier A , Mall G , Leyendecker G . The uterine peristaltic pump . In: Ivell R , Holstein A-F (Eds). The Fate of the Male Germ Cell. US, Boston, MA : Springer , 1997 , 267 – 277 . Google Scholar CrossRef Search ADS Kämmerer U , von Wolff M , Markert UR . Immunology of human endometrium . Immunobiology 2004 ; 209 : 569 – 574 . Google Scholar CrossRef Search ADS PubMed Laniewski P , Gomez A , Hire G , So M , Herbst-Kralovetz MM . Human three-dimensional endometrial epithelial cell model to study host interactions with vaginal bacteria and Neisseria gonorrhoeae . Infect Immun 2017 ; 85 :pii: e01049-16 . Google Scholar CrossRef Search ADS PubMed Lathrop SK , Bloom SM , Rao SM , Nutsch K , Lio CW , Santacruz N , Peterson DA , Stappenbeck TS , Hsieh CS . Peripheral education of the immune system by colonic commensal microbiota . Nature 2011 ; 478 : 250 – 254 . Google Scholar CrossRef Search ADS PubMed Lauder AP , Roche AM , Sherrill-Mix S , Bailey A , Laughlin AL , Bittinger K , Leite R , Elovitz MA , Parry S , Bushman FD . Comparison of placenta samples with contamination controls does not provide evidence for a distinct placenta microbiota . Microbiome 2016 ; 4 : 29 . Google Scholar CrossRef Search ADS PubMed Laurence M , Hatzis C , Brash DE . Common contaminants in next-generation sequencing that hinder discovery of low-abundance microbes . PLoS One 2014 ; 9 : e97876 . Google Scholar CrossRef Search ADS PubMed Le Bouteiller P , Piccinni MP . Human NK cells in pregnant uterus: why there? Am J Reprod Immunol 2008 ; 59 : 401 – 406 . Google Scholar CrossRef Search ADS PubMed Leatham MP , Banerjee S , Autieri SM , Mercado-Lubo R , Conway T , Cohen PS . Precolonized human commensal Escherichia coli strains serve as a barrier to E. coli O157:H7 growth in the streptomycin-treated mouse intestine . Infect Immun 2009 ; 77 : 2876 – 2886 . Google Scholar CrossRef Search ADS PubMed Ledee-Bataille N , Dubanchet S , Coulomb-L’hermine A , Durand-Gasselin I , Frydman R , Chaouat G . A new role for natural killer cells, interleukin (IL)−12, and IL-18 in repeated implantation failure after in vitro fertilization . Fertil Steril 2004 ; 81 : 59 – 65 . Google Scholar CrossRef Search ADS PubMed Li M-Q , Li H-P , Meng Y-H , Wang X-Q , Zhu X-Y , Mei J , Li D-J . Chemokine CCL2 enhances survival and invasiveness of endometrial stromal cells in an autocrine manner by activating Akt and MAPK/Erk1/2 signal pathway . Fertil Steril 2012 ; 97 : 919 – 929.e911 . Google Scholar CrossRef Search ADS PubMed Littman Dan R , Pamer Eric G . Role of the commensal microbiota in normal and pathogenic host immune responses . Cell Host Microbe 2011 ; 10 : 311 – 323 . Google Scholar CrossRef Search ADS PubMed Macpherson AJ , de Aguero MG , Ganal-Vonarburg SC . How nutrition and the maternal microbiota shape the neonatal immune system . Nat Rev Immunol 2017 ; 10 : 508 – 517 . Google Scholar CrossRef Search ADS Maier E , Anderson R , Roy N . Understanding how commensal obligate anaerobic bacteria regulate immune functions in the large intestine . Nutrients 2015 ; 7 : 45 . Google Scholar CrossRef Search ADS Male V , Sharkey A , Masters L , Kennedy PR , Farrell LE , Moffett A . The effect of pregnancy on the uterine NK cell KIR repertoire . Eur J Immunol 2011 ; 41 : 3017 – 3027 . Google Scholar CrossRef Search ADS PubMed Manaster I , Mandelboim O . The unique properties of uterine NK cells . Am J Reprod Immunol 2010 ; 63 : 434 – 444 . Google Scholar CrossRef Search ADS PubMed Marconi C , Santos-Greatti MM , Parada CM , Pontes A , Pontes AG , Giraldo PC , Donders GG , da Silva MG . Cervicovaginal levels of proinflammatory cytokines are increased during chlamydial infection in bacterial vaginosis but not in lactobacilli-dominated flora . J Low Genit Tract Dis 2014 ; 18 : 261 – 265 . Google Scholar CrossRef Search ADS PubMed Margioula-Siarkou C , Prapas Y , Petousis S , Milias S , Ravanos K , Kalogiannidis I , Mavromatidis G , Haitoglou C , Prapas N , Rousso D . LIF and LIFR expression in the endometrium of fertile and infertile women: a prospective observational casecontrol study . Mol Med Rep 2016 ; 13 : 4721 – 4728 . Google Scholar CrossRef Search ADS PubMed Martin JC , Jasper MJ , Valbuena D , Meseguer M , Remohi J , Pellicer A , Simon C . Increased adhesiveness in cultured endometrial-derived cells is related to the absence of moesin expression . Biol Reprod 2000 ; 63 : 1370 – 1376 . Google Scholar CrossRef Search ADS PubMed Martin DH , Marrazzo JM . The vaginal microbiome: current understanding and future directions . J Infect Dis 2016 ; 214 : S36 – S41 . Google Scholar CrossRef Search ADS PubMed Mazmanian SK , Liu CH , Tzianabos AO , Kasper DL . An immunomodulatory molecule of symbiotic bacteria directs maturation of the host immune system . Cell 2005 ; 122 : 107 – 118 . Google Scholar CrossRef Search ADS PubMed Mazmanian SK , Round JL , Kasper DL . A microbial symbiosis factor prevents intestinal inflammatory disease . Nature 2008 ; 453 : 620 – 625 . Google Scholar CrossRef Search ADS PubMed McDonald HM , Brocklehurst P , Gordon A . Antibiotics for treating bacterial vaginosis in pregnancy . Cochrane Database Syst Rev 2007 ; 24 : CD000262 . Medzhitov R , Janeway CA . Innate immunity: impact on the adaptive immune response . Curr Opin Immunol 1997 ; 9 : 4 – 9 . Google Scholar CrossRef Search ADS PubMed Medzhitov R , Preston-Hurlburt P , Janeway CA Jr. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity . Nature 1997 ; 388 : 394 – 397 . Google Scholar CrossRef Search ADS PubMed Miles SM , Hardy BL , Merrell DS . Investigation of the microbiota of the reproductive tract in women undergoing a total hysterectomy and bilateral salpingo-oopherectomy . Fertil Steril 2017 ; 107 : 813 – 820.e811 . Google Scholar CrossRef Search ADS PubMed Mitchell CM , Haick A , Nkwopara E , Garcia R , Rendi M , Agnew K , Fredricks DN , Eschenbach D . Colonization of the upper genital tract by vaginal bacterial species in nonpregnant women . Am J Obstet Gynecol 2015 ; 212 : 611.e611 – 619 . Google Scholar CrossRef Search ADS Moffett A , Colucci F . Uterine NK cells: active regulators at the maternal-fetal interface . J Clin Invest 2014 ; 124 : 1872 – 1879 . Google Scholar CrossRef Search ADS PubMed Moffett A , Hiby SE , Sharkey AM . The role of the maternal immune system in the regulation of human birthweight . Philos Trans R Soc Lond B Biol Sci 2015 ; 370 : 20140071 . Google Scholar CrossRef Search ADS PubMed Moffett A , Regan L , Braude P . Natural killer cells, miscarriage, and infertility . Br Med J 2004 ; 329 : 1283 – 1285 . Google Scholar CrossRef Search ADS Moffett-King A . Natural killer cells and pregnancy . Nat Rev Immunol 2002 ; 2 : 656 – 663 . Google Scholar CrossRef Search ADS PubMed Moller BR , Kristiansen FV , Thorsen P , Frost L , Mogensen SC . Sterility of the uterine cavity . Acta Obstet Gynecol Scand 1995 ; 74 : 216 – 219 . Google Scholar CrossRef Search ADS PubMed Moore DE , Soules MR , Klein NA , Fujimoto VY , Agnew KJ , Eschenbach DA . Bacteria in the transfer catheter tip influence the live-birth rate after in vitro fertilization . Fertil Steril 2000 ; 74 : 1118 – 1124 . Google Scholar CrossRef Search ADS PubMed Moreno I , Codoner FM , Vilella F , Valbuena D , Martinez-Blanch JF , Jimenez-Almazan J , Alonso R , Alama P , Remohi J , Pellicer A et al. . Evidence that the endometrial microbiota has an effect on implantation success or failure . Am J Obstet Gynecol 2016 ; 215 : 684 – 703 . Google Scholar CrossRef Search ADS PubMed Moreno I , Franasiak JM . Endometrial microbiota—new player in town . Fertil Steril 2017 ; 108 : 32 – 39 . Google Scholar CrossRef Search ADS PubMed Mori M , Bogdan A , Balassa T , Csabai T , Szekeres-Bartho J . The decidua-the maternal bed embracing the embryo-maintains the pregnancy . Semin Immunopathol 2016 ; 38 : 635 – 649 . Google Scholar CrossRef Search ADS PubMed Murphy CR . The plasma membrane transformation of uterine epithelial cells during pregnancy . J Reprod Fertil Suppl 2000 ; 55 : 23 – 28 . Google Scholar PubMed Murphy CR , Rogers PA , Hosie MJ , Leeton J , Beaton L . Tight junctions of human uterine epithelial cells change during the menstrual cycle: a morphometric study . Acta Anat (Basel) 1992 ; 144 : 36 – 38 . Google Scholar CrossRef Search ADS PubMed Naessens A , Foulon W , Cammu H , Goossens A , Lauwers S . Epidemiology and pathogenesis of ureaplasma urealyticum in spontaneous abortion and early preterm labor . Acta Obstet Gynecol Scand 1987 ; 66 : 513 – 516 . Google Scholar CrossRef Search ADS PubMed Navot D , Bergh PA , Williams M , Garrisi GJ , Guzman I , Sandler B , Fox J , Schreiner-Engel P , Hofmann GE , Grunfeld L . An insight into early reproductive processes through the in vivo model of ovum donation . J Clin Endocrinol Metab 1991 ; 72 : 408 – 414 . Google Scholar CrossRef Search ADS PubMed Nikzad H , Haddad Kashani H , Kabir-Salmani M , Akimoto Y , Iwashita M . Expression of galectin-8 on human endometrium: molecular and cellular aspects . Iran J Reprod Med 2013 ; 11 : 65 – 70 . Google Scholar PubMed Oger P , Bulla R , Tedesco F , Portier A , Dubanchet S , Bailly M , Wainer R , Chaouat G , Ledee N . Higher interleukin-18 and mannose-binding lectin are present in uterine lumen of patients with unexplained infertility . Reprod Biomed Online 2009 ; 19 : 591 – 598 . Google Scholar CrossRef Search ADS PubMed Omenetti S , Pizarro TT . The Treg/Th17 axis: a dynamic balance regulated by the gut microbiome . Front Immunol 2015 ; 6 : 639 . Google Scholar CrossRef Search ADS PubMed Onderdonk AB , Delaney ML , Fichorova RN . The human microbiome during bacterial vaginosis . Clin Microbiol Rev 2016 ; 29 : 223 – 238 . Google Scholar CrossRef Search ADS PubMed Parnell LA , Briggs CM , Mysorekar IU . Maternal microbiomes in preterm birth: recent progress and analytical pipelines . Semin Perinatol 2017 ; 41 : 392 – 400 . Google Scholar CrossRef Search ADS PubMed Pelzer E , Gomez-Arango LF , Barrett HL , Nitert MD . Maternal health and the placental microbiome . Placenta 2016 ; 54 : 30 – 37 . Google Scholar CrossRef Search ADS PubMed Perez-Munoz ME , Arrieta MC , Ramer-Tait AE , Walter J . A critical assessment of the ‘sterile womb’ and ‘in utero colonization’ hypotheses: implications for research on the pioneer infant microbiome . Microbiome 2017 ; 5 : 48 . Google Scholar CrossRef Search ADS PubMed Peterson J , Garges S , Giovanni M , McInnes P , Wang L , Schloss JA , Bonazzi V , McEwen JE , Wetterstrand KA , Deal C et al. . The NIH human microbiome project . Genome Res 2009 ; 19 : 2317 – 2323 . Google Scholar CrossRef Search ADS PubMed Petnicki-Ocwieja T , Hrncir T , Liu Y-J , Biswas A , Hudcovic T , Tlaskalova-Hogenova H , Kobayashi KS . Nod2 is required for the regulation of commensal microbiota in the intestine . Proc Natl Acad Sci USA 2009 ; 106 : 15813 – 15818 . Google Scholar CrossRef Search ADS PubMed Plaks V , Birnberg T , Berkutzki T , Sela S , BenYashar A , Kalchenko V , Mor G , Keshet E , Dekel N , Neeman M et al. . Uterine DCs are crucial for decidua formation during embryo implantation in mice . J Clin Invest 2008 ; 118 : 3954 – 3965 . Google Scholar PubMed Postler TS , Ghosh S . Understanding the holobiont: how microbial metabolites affect human health and shape the immune system . Cell Metab 2017 ; 26 : 110 – 130 . Google Scholar CrossRef Search ADS PubMed Power ML , Quaglieri C , Schulkin J . Reproductive microbiomes . Reprod Sci 2017 ; 24 : 1482 – 1492 . Google Scholar CrossRef Search ADS PubMed Prakobphol A , Genbacev O , Gormley M , Kapidzic M , Fisher SJ . A role for the L-selectin adhesion system in mediating cytotrophoblast emigration from the placenta . Dev Biol 2006 ; 298 : 107 – 117 . Google Scholar CrossRef Search ADS PubMed Prince AL , Antony KM , Chu DM , Aagaard KM . The microbiome, parturition, and timing of birth: more questions than answers . J Reprod Immunol 2014 ;104–105: 12 – 19 . Prince AL , Ma J , Kannan PS , Alvarez M , Gisslen T , Harris RA , Sweeney EL , Knox CL , Lambers DS , Jobe AH et al. . The placental membrane microbiome is altered among subjects with spontaneous preterm birth with and without chorioamnionitis . Am J Obstet Gynecol 2016 ; 214 : 627.e621 – 627.e616 . Racicot K , Cardenas I , Wunsche V , Aldo P , Guller S , Means RE , Romero R , Mor G . Viral infection of the pregnant cervix predisposes to ascending bacterial infection . J Immunol 2013 ; 191 : 934 – 941 . Google Scholar CrossRef Search ADS PubMed Radtke AL , Quayle AJ , Herbst-Kralovetz MM . Microbial products alter the expression of membrane-associated mucin and antimicrobial peptides in a three-dimensional human endocervical epithelial cell model . Biol Reprod 2012 ; 87 : 132 . Google Scholar CrossRef Search ADS PubMed Ranjan R , Rani A , Metwally A , McGee HS , Perkins DL . Analysis of the microbiome: advantages of whole genome shotgun versus 16S amplicon sequencing . Biochem Biophys Res Commun 2016 ; 469 : 967 – 977 . Google Scholar CrossRef Search ADS PubMed Relman DA . New technologies, human-microbe interactions, and the search for previously unrecognized pathogens . J Infect Dis 2002 ; 186 : S254 – S258 . Google Scholar CrossRef Search ADS PubMed Ren H , Du M . Role of maternal periodontitis in preterm birth . Front Immunol 2017 ; 8 : 139 . Google Scholar PubMed Rescigno M , Urbano M , Valzasina B , Francolini M , Rotta G , Bonasio R , Granucci F , Kraehenbuhl JP , Ricciardi-Castagnoli P . Dendritic cells express tight junction proteins and penetrate gut epithelial monolayers to sample bacteria . Nat Immunol 2001 ; 2 : 361 – 367 . Google Scholar CrossRef Search ADS PubMed Rieger L , Honig A , Sutterlin M , Kapp M , Dietl J , Ruck P , Kammerer U . Antigen-presenting cells in human endometrium during the menstrual cycle compared to early pregnancy . J Soc Gynecol Investig 2004 ; 11 : 488 – 493 . Google Scholar CrossRef Search ADS PubMed Robertson SA , Chin PY , Glynn DJ , Thompson JG . Peri-conceptual cytokines—setting the trajectory for embryo implantation, pregnancy and beyond . Am J Reprod Immunol 2011 ; 66 : 2 – 10 . Google Scholar CrossRef Search ADS PubMed Robertson SA , Moldenhauer LM . Immunological determinants of implantation success . Int J Dev Biol 2014 ; 58 : 205 – 217 . Google Scholar CrossRef Search ADS PubMed Romero R , Espinoza J , Mazor M . Can endometrial infection/inflammation explain implantation failure, spontaneous abortion, and preterm birth after in vitro fertilization? Fertil Steril 2004 ; 82 : 799 – 804 . Google Scholar CrossRef Search ADS PubMed Round JL , Lee SM , Li J , Tran G , Jabri B , Chatila TA , Mazmanian SK . The Toll-like receptor 2 pathway establishes colonization by a commensal of the human microbiota . Science 2011 ; 332 : 974 – 977 . Google Scholar CrossRef Search ADS PubMed Round JL , Mazmanian SK . The gut microbiome shapes intestinal immune responses during health and disease . Nat Rev Immunol 2009 ; 9 : 313 – 323 . Google Scholar CrossRef Search ADS PubMed Saito S , Nakashima A , Shima T , Ito M . Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy . Am J Reprod Immunol 2010 ; 63 : 601 – 610 . Google Scholar CrossRef Search ADS PubMed Saito S , Sasaki Y , Sakai M . CD4+CD25high regulatory T cells in human pregnancy . J Reprod Immunol 2005 ; 65 : 111 – 120 . Google Scholar CrossRef Search ADS PubMed Saito S , Tsukaguchi N , Hasegawa T , Michimata T , Tsuda H , Narita N . Distribution of Th1, Th2, and Th0 and the Th1/Th2 cell ratios in human peripheral and endometrial T cells . Am J Reprod Immunol 1999 ; 42 : 240 – 245 . Google Scholar CrossRef Search ADS PubMed Salamonsen LA , Zhang J , Brasted M . Leukocyte networks and human endometrial remodelling . J Reprod Immunol 2002 ; 57 : 95 – 108 . Google Scholar CrossRef Search ADS PubMed Salim R , Ben-Shlomo I , Colodner R , Keness Y , Shalev E . Bacterial colonization of the uterine cervix and success rate in assisted reproduction: results of a prospective survey . Hum Reprod 2002 ; 17 : 337 – 340 . Google Scholar CrossRef Search ADS PubMed Salter SJ , Cox MJ , Turek EM , Calus ST , Cookson WO , Moffatt MF , Turner P , Parkhill J , Loman NJ , Walker AW . Reagent and laboratory contamination can critically impact sequence-based microbiome analyses . BMC Biol 2014 ; 12 : 87 . Google Scholar CrossRef Search ADS PubMed Sanos SL , Bui VL , Mortha A , Oberle K , Heners C , Johner C , Diefenbach A . RORgammat and commensal microflora are required for the differentiation of mucosal interleukin 22-producing NKp46+ cells . Nat Immunol 2009 ; 10 : 83 – 91 . Google Scholar CrossRef Search ADS PubMed Sansonetti PJ . War and peace at mucosal surfaces . Nat Rev Immunol 2004 ; 4 : 953 – 964 . Google Scholar CrossRef Search ADS PubMed Santos-Greatti MMV , da Silva MG , Ferreira CST , Marconi C . Cervicovaginal cytokines, sialidase activity and bacterial load in reproductive-aged women with intermediate vaginal flora . J Reprod Immunol 2016 ; 118 : 36 – 41 . Google Scholar CrossRef Search ADS PubMed Satoh-Takayama N , Vosshenrich CAJ , Lesjean-Pottier S , Sawa S , Lochner M , Rattis F , Mention J-J , Thiam K , Cerf-Bensussan N , Mandelboim O et al. . Microbial flora drives interleukin 22 production in intestinal NKp46+ Cells that provide innate mucosal immune defense . Immunity 2008 ; 29 : 958 – 970 . Google Scholar CrossRef Search ADS PubMed Schaefer TM , Fahey JV , Wright JA , Wira CR . Innate immunity in the human female reproductive tract: antiviral response of uterine epithelial cells to the TLR3 agonist Poly(I:C) . J Immunol 2005 ; 174 : 992 . Google Scholar CrossRef Search ADS PubMed Selman H , Mariani M , Barnocchi N , Mencacci A , Bistoni F , Arena S , Pizzasegale S , Brusco GF , Angelini A . Examination of bacterial contamination at the time of embryo transfer, and its impact on the IVF/pregnancy outcome . J Assist Reprod Genet 2007 ; 24 : 395 – 399 . Google Scholar CrossRef Search ADS PubMed Shanmugasundaram U , Hilton JF , Critchfield JW , Greenblatt RM , Giudice LC , Averbach S , Seidman D , Shacklett BL , Smith-McCune K . Effects of the levonorgestrel-releasing intrauterine device on the immune microenvironment of the human cervix and endometrium . Am J Reprod Immunol 2016 ; 76 : 137 – 148 . Google Scholar CrossRef Search ADS PubMed Shi C , Pamer EG . Monocyte recruitment during infection and inflammation . Nat Rev Immunol 2011 ; 11 : 762 – 774 . Google Scholar CrossRef Search ADS PubMed Sierra-Filardi E , Nieto C , Dominguez-Soto A , Barroso R , Sanchez-Mateos P , Puig-Kroger A , Lopez-Bravo M , Joven J , Ardavin C , Rodriguez-Fernandez JL et al. . CCL2 shapes macrophage polarization by GM-CSF and M-CSF: identification of CCL2/CCR2-dependent gene expression profile . J Immunol 2014 ; 192 : 3858 – 3867 . Google Scholar CrossRef Search ADS PubMed Singh M , Chaudhry P , Asselin E . Bridging endometrial receptivity and implantation: network of hormones, cytokines, and growth factors . J Endocrinol 2011 ; 210 : 5 – 14 . Google Scholar CrossRef Search ADS PubMed Sirota I , Zarek SM , Segars JH . Potential influence of the microbiome on infertility and assisted reproductive technology . Semin Reprod Med 2014 ; 32 : 35 – 42 . Google Scholar CrossRef Search ADS PubMed Smith PD , Janoff EN , Mosteller-Barnum M , Merger M , Orenstein JM , Kearney JF , Graham MF . Isolation and purification of CD14-negative mucosal macrophages from normal human small intestine . J Immunol Methods 1997 ; 202 : 1 – 11 . Google Scholar CrossRef Search ADS PubMed Smith PD , Smythies LE , Mosteller-Barnum M , Sibley DA , Russell MW , Merger M , Sellers MT , Orenstein JM , Shimada T , Graham MF et al. . Intestinal macrophages lack CD14 and CD89 and consequently are down-regulated for LPS- and IgA-mediated activities . J Immunol 2001 ; 167 : 2651 – 2656 . Google Scholar CrossRef Search ADS PubMed Smythies LE , Sellers M , Clements RH , Mosteller-Barnum M , Meng G , Benjamin WH , Orenstein JM , Smith PD . Human intestinal macrophages display profound inflammatory anergy despite avid phagocytic and bacteriocidal activity . J Clin Invest 2005 ; 115 : 66 – 75 . Google Scholar CrossRef Search ADS PubMed Smythies LE , Shen R , Bimczok D , Novak L , Clements RH , Eckhoff DE , Bouchard P , George MD , Hu WK , Dandekar S et al. . Inflammation anergy in human intestinal macrophages is due to Smad-induced IκBα expression and NF-κB inactivation . J Biol Chem 2010 ; 285 : 19593 – 19604 . Google Scholar CrossRef Search ADS PubMed Spurbeck RR , Arvidson CG . Inhibition of Neisseria gonorrhoeae epithelial cell interactions by vaginal Lactobacillus species . Infect Immun 2008 ; 76 : 3124 – 3130 . Google Scholar CrossRef Search ADS PubMed Stappenbeck TS , Hooper LV , Gordon JI . Developmental regulation of intestinal angiogenesis by indigenous microbes via Paneth cells . Proc Natl Acad Sci USA 2002 ; 99 : 15451 – 15455 . Google Scholar CrossRef Search ADS PubMed Stray-Pedersen B , Bruu AL , Molne K . Infertility and uterine colonization with ureaplasma urealyticum . Acta Obstet Gynecol Scand 1982 ; 61 : 21 – 24 . Google Scholar CrossRef Search ADS PubMed Swiatczak B , Rescigno M . How the interplay between antigen presenting cells and microbiota tunes host immune responses in the gut . Semin Immunol 2012 ; 24 : 43 – 49 . Google Scholar CrossRef Search ADS PubMed Swidsinski A , Verstraelen H , Loening-Baucke V , Swidsinski S , Mendling W , Halwani Z . Presence of a polymicrobial endometrial biofilm in patients with bacterial vaginosis . PLoS One 2013 ; 8 : e53997 . Google Scholar CrossRef Search ADS PubMed Swiecki M , Miller HL , Sesti-Costa R , Cella M , Gilfillan S , Colonna M . Microbiota induces tonic CCL2 systemic levels that control pDC trafficking in steady state . Mucosal Immunol 2017 ; 10 : 936 – 945 . Google Scholar CrossRef Search ADS PubMed Tachedjian G , Aldunate M , Bradshaw CS , Cone RA . The role of lactic acid production by probiotic Lactobacillus species in vaginal health . Res Microbiol 2017 ; 168 : 782 – 792 . Google Scholar CrossRef Search ADS PubMed Takeuchi O , Akira S . Pattern recognition receptors and inflammation . Cell 2010 ; 140 : 805 – 820 . Google Scholar CrossRef Search ADS PubMed Tao X , Franasiak JM , Zhan Y , Scott RT , Rajchel J , Bedard J , Newby R , Scott RT , Treff NR , Chu T . Characterizing the endometrial microbiome by analyzing the ultra-low bacteria from embryo transfer catheter tips in IVF cycles: next generation sequencing (NGS) analysis of the 16 S ribosomal gene . Hum Microbiome J 2017 ; 3 : 15 – 21 . Google Scholar CrossRef Search ADS Tessler M , Neumann JS , Afshinnekoo E , Pineda M , Hersch R , Velho LFM , Segovia BT , Lansac-Toha FA , Lemke M , DeSalle R et al. . Large-scale differences in microbial biodiversity discovery between 16 S amplicon and shotgun sequencing . Sci Rep 2017 ; 7 : 6589 . Google Scholar CrossRef Search ADS PubMed Thaiss CA , Levy M , Korem T , Dohnalova L , Shapiro H , Jaitin DA , David E , Winter DR , Gury-BenAri M , Tatirovsky E et al. . Microbiota diurnal rhythmicity programs host transcriptome oscillations . Cell 2016 ; 167 : 1495 – 1510.e1412 . Google Scholar CrossRef Search ADS PubMed Thie M , Harrach-Ruprecht B , Sauer H , Fuchs P , Albers A , Denker HW . Cell adhesion to the apical pole of epithelium: a function of cell polarity . Eur J Cell Biol 1995 ; 66 : 180 – 191 . Google Scholar PubMed Tilburgs T , Roelen DL , van der Mast BJ , van Schip JJ , Kleijburg C , de Groot-Swings GM , Kanhai HH , Claas FH , Scherjon SA . Differential distribution of CD4(+)CD25(bright) and CD8(+)CD28(-) T-cells in decidua and maternal blood during human pregnancy . Placenta 2006 ; 27 : S47 – S53 . Google Scholar CrossRef Search ADS PubMed Tilburgs T , Scherjon SA , van der Mast BJ , Haasnoot GW , Voort-Maarschalk MV-v , Roelen DL , van Rood JJ , Claas FH . Fetal–maternal HLA-C mismatch is associated with decidual T cell activation and induction of functional T regulatory cells . J Reprod Immunol 2009 ; 82 : 148 – 157 . Google Scholar CrossRef Search ADS PubMed Tissier H Recherches sur la flore intestinale normale et pathologique du nourrisson. Doctoral Dissertation. University of Paris 1900 . Trundley A , Moffett A . Human uterine leukocytes and pregnancy . Tissue Antigens 2004 ; 63 : 1 – 12 . Google Scholar CrossRef Search ADS PubMed Tschopp J , Martinon F , Burns K . NALPs: a novel protein family involved in inflammation . Nat Rev Mol Cell Biol 2003 ; 4 : 95 – 104 . Google Scholar CrossRef Search ADS PubMed Turner JR . Intestinal mucosal barrier function in health and disease . Nat Rev Immunol 2009 ; 9 : 799 – 809 . Google Scholar CrossRef Search ADS PubMed Ubeda C , Djukovic A , Isaac S . Roles of the intestinal microbiota in pathogen protection . Clin Transl Immunology 2017 ; 6 : e128 . Google Scholar CrossRef Search ADS PubMed Vaishnava S , Behrendt CL , Ismail AS , Eckmann L , Hooper LV . Paneth cells directly sense gut commensals and maintain homeostasis at the intestinal host-microbial interface . Proc Natl Acad Sci USA 2008 ; 105 : 20858 – 20863 . Google Scholar CrossRef Search ADS PubMed van der Horst J , Buijs MJ , Laine ML , Wismeijer D , Loos BG , Crielaard W , Zaura E . Sterile paper points as a bacterial DNA-contamination source in microbiome profiles of clinical samples . J Dent 2013 ; 41 : 1297 – 1301 . Google Scholar CrossRef Search ADS PubMed van der Meer A , Lukassen HG , van Lierop MJ , Wijnands F , Mosselman S , Braat DD , Joosten I . Membrane-bound HLA-G activates proliferation and interferon-gamma production by uterine natural killer cells . Mol Hum Reprod 2004 ; 10 : 189 – 195 . Google Scholar CrossRef Search ADS PubMed van der Molen RG , Schutten JH , van Cranenbroek B , ter Meer M , Donckers J , Scholten RR , van der Heijden OW , Spaanderman ME , Joosten I . Menstrual blood closely resembles the uterine immune micro-environment and is clearly distinct from peripheral blood . Hum Reprod 2014 ; 29 : 303 – 314 . Google Scholar CrossRef Search ADS PubMed van Mourik MS , Macklon NS , Heijnen CJ . Embryonic implantation: cytokines, adhesion molecules, and immune cells in establishing an implantation environment . J Leukoc Biol 2009 ; 85 : 4 – 19 . Google Scholar CrossRef Search ADS PubMed Vandeputte D , Kathagen G , D’Hoe K , Vieira-Silva S , Valles-Colomer M , Sabino J , Wang J , Tito RY , De Commer L , Darzi Y et al. . Quantitative microbiome profiling links gut community variation to microbial load . Nature 2017 ; 551 : 507 – 511 . Google Scholar PubMed Verstraelen H , Verhelst R , Claeys G , Temmerman M , Vaneechoutte M . Culture-independent analysis of vaginal microflora: the unrecognized association of Atopobium vaginae with bacterial vaginosis . Am J Obstet Gynecol 2004 ; 191 : 1130 – 1132 . Google Scholar CrossRef Search ADS PubMed Verstraelen H , Vilchez-Vargas R , Desimpel F , Jauregui R , Vankeirsbilck N , Weyers S , Verhelst R , De Sutter P , Pieper DH , Van De Wiele T . Characterisation of the human uterine microbiome in non-pregnant women through deep sequencing of the V1-2 region of the 16 S rRNA gene . PeerJ 2016 ; 4 : e1602 . Google Scholar CrossRef Search ADS PubMed Viniker DA . Hypothesis on the role of sub-clinical bacteria of the endometrium (bacteria endometrialis) in gynaecological and obstetric enigmas . Hum Reprod Update 1999 ; 5 : 373 – 385 . Google Scholar CrossRef Search ADS PubMed Walther-Antonio MR , Chen J , Multinu F , Hokenstad A , Distad TJ , Cheek EH , Keeney GL , Creedon DJ , Nelson H , Mariani A et al. . Potential contribution of the uterine microbiome in the development of endometrial cancer . Genome Med 2016 ; 8 : 122 . Google Scholar CrossRef Search ADS PubMed Wexler HM . Bacteroides: the good, the bad, and the nitty-gritty . Clin Microbiol Rev 2007 ; 20 : 593 – 621 . Google Scholar CrossRef Search ADS PubMed Wilcox AJ , Baird DD , Weinberg CR . Time of implantation of the conceptus and loss of pregnancy . N Engl J Med 1999 ; 340 : 1796 – 1799 . Google Scholar CrossRef Search ADS PubMed Wira CR , Fahey JV , Rodriguez-Garcia M , Shen Z , Patel MV . Regulation of mucosal immunity in the female reproductive tract: the role of sex hormones in immune protection against sexually transmitted pathogens . Am J Reprod Immunol 2014 ; 72 : 236 – 258 . Google Scholar CrossRef Search ADS PubMed Wira CR , Grant-Tschudy KS , Crane-Godreau MA . Epithelial cells in the female reproductive tract: a central role as sentinels of immune protection . Am J Reprod Immunol 2005 ; 53 : 65 – 76 . Google Scholar CrossRef Search ADS PubMed Wira CR , Rodriguez-Garcia M , Patel MV . The role of sex hormones in immune protection of the female reproductive tract . Nat Rev Immunol 2015 ; 15 : 217 – 230 . Google Scholar CrossRef Search ADS PubMed Wu HX , Jin LP , Xu B , Liang SS , Li DJ . Decidual stromal cells recruit Th17 cells into decidua to promote proliferation and invasion of human trophoblast cells by secreting IL-17 . Cell Mol Immunol 2014 ; 11 : 253 – 262 . Google Scholar CrossRef Search ADS PubMed Yeoman CJ , Thomas SM , Miller ME , Ulanov AV , Torralba M , Lucas S , Gillis M , Cregger M , Gomez A , Ho M et al. . A multi-omic systems-based approach reveals metabolic markers of bacterial vaginosis and insight into the disease . PLoS One 2013 ; 8 : e56111 . Google Scholar CrossRef Search ADS PubMed Younes JA , Lievens E , Hummelen R , van der Westen R , Reid G , Petrova MI . Women and their microbes: the unexpected friendship . Trends Microbiol 2017 ; 26 : 16 – 32 . Google Scholar CrossRef Search ADS PubMed Young SL , Lyddon TD , Jorgenson RL , Misfeldt ML . Expression of Toll-like receptors in human endometrial epithelial cells and cell lines . Am J Reprod Immunol 2004 ; 52 : 67 – 73 . Google Scholar CrossRef Search ADS PubMed Yu Y , Lu L , Sun J , Petrof EO , Claud EC . Preterm infant gut microbiota affects intestinal epithelial development in a humanized microbiome gnotobiotic mouse model . Am J Physiol Gastrointest Liver Physiol 2016 ; 311 : G521 – G532 . Google Scholar CrossRef Search ADS PubMed Zaura E , Nicu EA , Krom BP , Keijser BJ . Acquiring and maintaining a normal oral microbiome: current perspective . Front Cell Infect Microbiol 2014 ; 4 : 85 . Google Scholar CrossRef Search ADS PubMed Zervomanolakis I , Ott HW , Hadziomerovic D , Mattle V , Seeber BE , Virgolini I , Heute D , Kissler S , Leyendecker G , Wildt L . Physiology of upward transport in the human female genital tract . Ann N Y Acad Sci 2007 ; 1101 : 1 – 20 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Human Reproduction Update Oxford University Press

How uterine microbiota might be responsible for a receptive, fertile endometrium

Loading next page...
 
/lp/ou_press/how-uterine-microbiota-might-be-responsible-for-a-receptive-fertile-TGT2hpI2H3
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1355-4786
eISSN
1460-2369
D.O.I.
10.1093/humupd/dmy012
Publisher site
See Article on Publisher Site

Abstract

Abstract BACKGROUND Fertility depends on a receptive state of the endometrium, influenced by hormonal and anatomical adaptations, as well as the immune system. Local and systemic immunity is greatly influenced by microbiota. Recent discoveries of 16S rRNA in the endometrium and the ability to detect low-biomass microbiota fueled the notion that the uterus may be indeed a non-sterile compartment. To date, the concept of the ‘sterile womb’ focuses on in utero effects of microbiota on offspring and neonatal immunity. However, little awareness has been raised regarding the importance of uterine microbiota for endometrial physiology in reproductive health; manifested in fertility and placentation. OBJECTIVE AND RATIONALE Commensal colonization of the uterus has been widely discussed in the literature. The objective of this review is to outline the possible importance of this uterine colonization for a healthy, fertile uterus. We present the available evidence regarding uterine microbiota, focusing on recent findings based on 16S rRNA, and depict the possible importance of uterine colonization for a receptive endometrium. We highlight a possible role of uterine microbiota for host immunity and tissue adaptation, as well as conferring protection against pathogens. Based on knowledge of the interaction of the mucosal immune cells of the gut with the local microbiome, we want to investigate the potential implications of commensal colonization for uterine health. SEARCH METHODS PubMed and Google Scholar were searched for articles in English indexed from 1 January 2008 to 1 March 2018 for ‘16S rRNA’, ‘uterus’ and related search terms to assess available evidence on uterine microbiome analysis. A manual search of the references within the resulting articles was performed. To investigate possible functional contributions of uterine microbiota to health, studies on microbiota of other body sites were additionally assessed. OUTCOMES Challenging the view of a sterile uterus is in its infancy and, to date, no conclusions on a ‘core uterine microbiome’ can be drawn. Nevertheless, evidence for certain microbiota and/or associated compounds in the uterus accumulates. The presence of microbiota or their constituent molecules, such as polysaccharide A of the Bacteroides fragilis capsule, go together with healthy physiological function. Lessons learned from the gut microbiome suggest that the microbiota of the uterus may potentially modulate immune cell subsets needed for implantation and have implications for tissue morphology. Microbiota can also be crucial in protection against uterine infections by defending their niche and competing with pathogens. Our review highlights the need for well-designed studies on a ‘baseline’ microbial state of the uterus representing the optimal starting point for implantation and subsequent placenta formation. WIDER IMPLICATIONS The complex interplay of processes and cells involved in healthy pregnancy is still poorly understood. The correct receptive endometrial state, including the local immune environment, is crucial not only for fertility but also placenta formation since initiation of placentation highly depends on interaction with immune cells. Implantation failure, recurrent pregnancy loss, and other pathologies of endometrium and placenta, such as pre-eclampsia, represent an increasing societal burden. More robust studies are needed to investigate uterine colonization. Based on current data, future research needs to include the uterine microbiome as a relevant factor in order to understand the players needed for healthy pregnancy. immunology, cytokines, endometrium, implantation, uterus, pregnancy, menstrual cycle, recurrent miscarriage Introduction Maternal microbial colonization contributes to development of the unborn child by micronutrient provision, xenobiotic metabolism and enhancing maternal energy conversion (Macpherson et al., 2017). Currently, the concept of the ‘sterile womb’, the paradigm that the fetus grows up in a sterile environment, is highly debated (Perez-Munoz et al., 2017). If bacteria (or their compounds) were naturally present in the uterus, their role even before pregnancy, in maintenance of the uterus deserves attention. Various recent reviews focused on correlations between commensal uterine colonization, fertility problems and pregnancy complications (Franasiak and Scott, 2017; Moreno and Franasiak, 2017; Power et al., 2017; Prince et al., 2014). Bacteria are known to affect immunity (Hooper et al., 2012; Littman Dan and Pamer Eric, 2011). If they already impact the immune environment of the uterus before pregnancy, this would greatly impact the receptive potential of the endometrium, as well as the ability to correctly initialize placenta formation. Successful embryo implantation requires both a synchronous development and an intricate interplay between the hatched blastocyst and endometrium. This depends on a receptive state of the endometrium. In the first week after fertilization, the blastocyst makes initial contact with the highly specialized endometrium. During the female cycle, if the endometrium fails to undergo the proper adaptations to acquire this receptive state, infertility and impaired placentation may be the consequence. Since bacteria can play a role in morphological changes of, for example, mucosal cells, implications of microbiota in decidualization can be envisioned. At the same time, commensal microbiota may convey protection towards pathogenic species contributing to uterine health. Studies evaluating the endometrial microbiome and its role in fertility are limited and therefore we depend on extrapolating knowledge from other colonized body sites. A vast number of studies focusing on the gut microbiota have pointed out the contribution of microbes to immunity and development, and the necessity of commensal colonization to achieve a basal, healthy immune state (Littman Dan and Pamer Eric, 2011; Cho and Blaser, 2012; Geva-Zatorsky et al., 2017). We hypothesize that, if indeed present, the endometrial microbiome or its compounds can have implications for (pre-) decidualization, and therefore fertility. The objective of this review is to outline the available evidence regarding uterine microbiota, focusing on recent findings based on 16S rRNA, and to depict the possible importance of uterine colonization for a receptive endometrium. We aim to highlight a possible role of uterine microbiota in host immunity and tissue adaptation, as well as conferring protection against pathogens. Methods PubMed and Google Scholar were searched for articles in English indexed from 1 January 2008 to 1 March 2018 for ‘16S rRNA’, ‘uterus’ and related search terms (Supplementary Fig. SI) to assess available evidence on uterine microbiome analysis. Studies on the upper female genital tract were investigated regarding data specifically on the uterine compartment. A manual search of the references within the resulting articles was performed. Studies on microbiota of other body sites were additionally assessed. The non-sterile uterus? Microbiota and mammals depend on their symbiotic relationship. While microbes receive a steady nutrient supply through the host, the host benefits from vital contributions of the microbe to physiological processes such as epithelial homeostasis, and is supplied with a natural barrier against colonization by pathogenic species (Buchon et al., 2009; Eberl, 2010). Bacterial colonization also plays a crucial role in modulation of host immunity. Metagenomic analysis has propelled research on natural colonization of the human body forward, revealing microbiota at body sites that were previously considered to be sterile. This includes the upper reproductive tract and placenta, challenging the classic dogma of a ‘sterile womb’ as coined by Henry Tissier more than a century ago (Tissier, 1900). Due to the earlier limitations in microbial characterization and challenges in sample acquisition, the significance of endometrial bacteria may have been missed or overlooked (Viniker, 1999). Now, increasing evidence in favor of an endometrial microbiome (Franasiak and Scott, 2015; Fang et al., 2016; Khan et al., 2016; Moreno et al., 2016; Verstraelen et al., 2016; Walther-Antonio et al., 2016; Chen et al., 2017; Miles et al., 2017; Tao et al., 2017) and a placental commensal colonization (Aagaard et al., 2014; Prince et al., 2016) indicates the need for a paradigm shift. The endometrium: an immunologically suited niche for microbiota In the gut, the intestinal immune system needs to be highly adapted to withstand the continuous threat posed by colonization of the large mucosal surface, separated from host tissue by only a single layer of epithelial cells (Hooper and Macpherson, 2010). In a symbiotic colonization, bacterial growth is safe for the host as long as it is contained within an assigned compartment. Therefore, tissue invasion must be limited, thus preventing potentially harmful inflammation or disruption of the niche for the necessary symbiotic relationship. Hooper and Macpherson (2010) described three types of immunological barriers needed for intestinal microbial homeostasis: anatomically limiting exposure of resident bacteria to the systemic immune system; immune mediators restricting direct contact between epithelia and microbes; and rapid detection and killing of bacteria upon barrier breach. All three pre-requisites are met by the endometrial mucosa. In the uterus, the single layer of columnar epithelial cells that proliferate to form glandular cells in the secretory phase of the menstrual cycle forms a strong barrier through tight junctions (Wira et al., 2015). The uterine mucosal surface and the endometrial fluid (EF) contain infection-controlling molecules, known as antimicrobial peptides (AMPs), with fluctuating levels during the menstrual cycle (Wira et al., 2014). AMPs are known to contribute to the health of the female reproductive tract with implications for fertility and pregnancy (Frew and Stock, 2011). An example of an AMP found in the uterus is the secretory leukocyte protease inhibitor, which has antiviral and antifungal properties, and also acts as a bactericidal against gram-negative as well as gram-positive bacteria such as Escherichia coli and Staphylococcus aureus (King et al., 2000). Furthermore, the endometrial lymphocytes in the mucosal layer are present throughout all stages of the menstrual cycle, ready to act upon pathogen invasion (Givan et al., 1997). Therefore, according to the required properties, the uterus could offer a safe niche for symbiotic colonization. Assessing commensal uterine colonization Not only theoretically does the endometrium offer an environment suited for coexistence with bacteria but also, in accordance with earlier hypotheses, such as Viniker’s theory of the ‘bacteria endometrialis’ inhabiting the uterine cavity (Viniker, 1999), various approaches have now provided evidence for uterine colonization, as outlined below. Initial assessment of endometrial colonization was based on observations of cells in culture. It has long been thought that the few observed growing species from endometrial cultures after hysterectomy resulted from the sterile character of the uterus, and the bacteria that were found were ascribed to a pathogenic condition (Ansbacher et al., 1967; Moller et al., 1995). However, women with no sign of infection also showed some bacterial growth in samples maintained in vitro. Species such as Lactobacillus (Duff et al., 1983; Cowling et al., 1992), Gardnerella vaginalis, Enterobacter, and Streptococcus agalactiae (Moller, et al., 1995), E. coli and Enterococcus faecalis were typically found in patients with uterine pathologies, such as endometritis (Cicinelli et al., 2008). Other studies used catheter tips that were used in embryo transfer (ET) for microbial assessment based on culturing. Bacterial growth was associated with outcomes of IVF treatment, with varying results. If bacterial growth was observed upon culturing the catheter tips used for ET during IVF, live birth rates were found to be either decreased (Egbase et al., 1996, 1999; Fanchin et al., 1998; Salim et al., 2002), increased or not affected at all, depending on the study or bacterial species cultured (Moore et al., 2000; Selman et al., 2007). This underlines the need to carefully assess the outcomes and limitations of such an approach, as discussed below. However, even if smears led to growing cultures, pregnancy was possible (e.g. clinical pregnancy rate 28 versus 17% in the positive culture group, n = 279) (Fanchin et al., 1998). Also, using isolates other than from catheter tips, endometrial colonization could be shown for asymptomatic as well as symptomatic women (Koren and Spigland, 1978; Naessens et al., 1987; Stray-Pedersen et al., 1982). Taken together, microbial presence could not be ascribed to a certain pregnancy outcome based on culture-dependent methods. As we now know, culture-dependent characterization of microbial communities is associated with limitations (Supplementary Table SI). In culture-based approaches, rapidly growing, aerobic species dominate, leaving rare species that demand specific culture conditions undetected (Relman, 2002; Verstraelen et al., 2004). Molecular approaches allow detection of species that will not be revealed by culture-dependent techniques. Mitchell et al. (2015) examined uterine swabs and EF from hysterectomies by quantitative PCR (qPCR) for 12 bacterial species, including Atopobium vaginae, Prevotella spp., Lactobacillus crispatus, Lactobacillus iners, G. vaginalis and bacterial vaginosis-associated bacterium 1 (BVAB1). All of the selected species could be detected in vaginal samples and to a varying extent in the endometrium. Clear differences could be found between vaginal and endometrial samples. While A. vaginae was more commonly detected in vagina, L. iners and BVAB1 were more likely to be detected in endometrial samples. Of note, 95% of hysterectomy samples showed the presence of bacterial DNA. Also based on selected targets, Swidsinski et al. (2013) used fluorescent in-situ hybridization probes for G. vaginalis, A. vaginae, Lactobacillus, Bacteroides, Prevotella, Enterobacteriaceae and Eubacteria. Again, the endometrial microbiotic environment was shown to be different than that of the vagina. However, the need of these studies to select specific probes targeting a certain species is intrinsically biased. Rather than evaluating the presence and/or abundance of a certain species, amplicon sequencing of the hypervariable 16 S region of the ribosomal RNA makes it possible to identify species present in a sample (Clarridge, 2004). The metagenomic molecular approach of high-throughput 16 S rRNA sequencing allows for a more complete view, reflecting the diversity and relative abundance of microbiota. However, also this technique comes with its limitations as outlined below. While shotgun whole genome sequencing (WGS) offers advantages to amplicon 16S sequencing (Ranjan et al., 2016; Tessler et al., 2017), as yet this approach has not been employed to study endometrial microbiota. Evidence of endometrial microbiota by 16S sequencing In the Human Microbiome Project (HMP), efforts are made to understand more about the natural colonization of various body sites, its physiological importance and implications for disease (Peterson et al., 2009). In the context of the HMP, data on gut and vaginal colonization increase, but the upper reproductive tract has not been characterized extensively yet (Gevers et al., 2012). However, the importance of microbiota for human reproduction is increasingly acknowledged (Younes et al., 2017). Recently, several studies targeted a putative endometrial microbiome through 16S rRNA sequencing, and each of the studies documented the presence of uterine microbiota (Fang, et al., 2016; Franasiak et al., 2016; Khan, et al., 2016; Moreno, et al., 2016; Verstraelen, et al., 2016; Walther-Antonio, et al., 2016; Chen, et al., 2017; Miles, et al., 2017; Tao, et al., 2017). An overview of these recent studies assessing human uterine microbial composition based on 16S rRNA sequencing is given in Table I. Tips used for ET, swabs, biopsies or aspirates were employed to obtain endometrial samples. The studies used next-generation sequencing and assigned the sequence reads to different operational taxonomic units (OTUs). Phylotypes or, depending on the fit against the database used for taxonomic annotation, genotype or species could be assigned to the 16S rRNA gene amplicon sequence. Depending on the study, varying 16S rRNA regions were targeted simultaneously, since depending on which hypervariable V1–V9 region is targeted, more differentiation at genus or species level is possible (Chakravorty et al., 2007). Alpha (within a subject) and beta (between different subjects) diversity were assessed using different diversity indices (Hugerth and Andersson, 2017). One of the classifications often used to express alpha diversity and species richness (i.e. how different species are sequenced in total) is the Shannon Diversity Index. Another often-used index is Chao1, using rare classes of OTUs (Franasiak et al., 2016; Moreno et al., 2016). Beta diversity is often expressed through the Bray–Curtis dissimilarity (Walther-Antonio, et al., 2016). Taxonomic distribution profiles of identified genera or species are markedly different between the studies. Table I Studies presenting uterine microbiome assessment based on 16S rRNA. Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces CT, cycle threshold; GnRHa, GnRH agonist; OTU, operational taxonomic unit; PGM, personal genome machine; qPCR, quantitative PCR; RT, reproductive tract. Table I Studies presenting uterine microbiome assessment based on 16S rRNA. Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces Cohort Sampling Results Ref. Aim Subjects (no. of women) Cohort specification (Inclusion criteria) Age (years) Technique Number of endometrial samples taken Controls Sequencing platform Variable regions Consistently found species Additional findings Franasiak et al. (2016) Characterization endometrial microbiome at the time of embryo transfer by reproductive outcome •33 Undergoing ART Average 35.9 (range 22.5–3.0) Transcervical; distal portion of transfer catheter used for embryo transfer 1 2 Escherichia coli controls, negative controls from reagents. Ion PGM™ system sequencing (Thermo Fisher) V2,3,4,6,7,8,9 Lactobacillus and Flavobacterium No association in Lactobacillus content and pregnancy outcome Verstraelen et al. (2016) Investigation of the presence of a uterine microbiome •11 Recurrent implantation failure Median 32 (range 25–39) Transcervical; Tao Brush™ Endometrial Sampler 1 Not defined MiSeq® (Illumina) V1–V2 Bacteroidetes phylum, making up one-third of overall population, second most abundant: Proteobacteria (incl. Pelomonas, Beta-and Gammaproteobacteria related to Escherichia/Shigella) High similarity in 90% of women of 75% (25% Bray–Curtis dissimilarity). Additional high abundance of Lactobacillus iners, Prevotellaamnii or L.crispatus in five women. •7 Recurrent pregnancy loss Khan et al. (2016) Investigation of endometrial microbial colonization related to endometriosis •32 Endometriosis, undergoing laparoscopy (16 of which undergoing GnRH treatment) Range 21–47 Transcervical swabs 1 Not defined MiSeq® (Illumina) Custom primers, see Mori et al. (2016) Lactobacillacae, Streptococcaceae, Staphylococaceae, Enterobacteriaceae, and Moraxellaceae as predominant families of 58 bacterial candidates. Increase in microbial colonization during menstrual phase. GnRH reatment impacts bacterial proportions. Microbial accumulation in endometriosis patients compared to control. •32 Fertile (16 of which undergoing GnRH treatment, uterine myoma) Range 21–52 Fang et al. (2016) Investigation of endometrial microbial colonization related to endometrial polyps •10 Fertile Average 30.9 (±1.56) Transcervical swabs 1 Vaginal swabs MiSeq® (Illumina) V4 Proteobacteria (73%), Firmicutes (14%) and Actinobacteria (5%) on phylum level; Enterobacter (33%), Pseudomonas (24%) and Lactobacillus (6%) on genus level in healthy cohort Differences in detected phyla/genera in patient versus control cohort less Enterobacter and Pseudomonas whereas more Lactobacillus than in diseased. Higher Shannon diversity in patient cohort. •10 Endometrial polyps 34.4 ± 2.44 •10 Endometrial polyps and chronic endometritis Average 35.2 ± 1.3 Moreno et al. (2016) Investigation of the presence of a uterine microbiome •13 Fertile Range 18–35 Transcervical; endometrial fluid aspirated through a catheter Two samples endometrial fluid per woman, obtained in pre-receptive and receptive phase within the same menstrual cycle (n = 26) Vaginal aspirates 454 pyro-sequencing on 454 Life Sciences GS FLX+ instrument (Roche) V3–V5 71.7% Lactobacillus, 12.6% Gardnerella, 3.7% Bifidobacterium, 3.2% Streptococcus, 0.9% Prevotella; if samples non-Lactobacillus dominated Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera abundant Stratification into Lactobacillus versus non-Lactobacillus dominated group (containing high proportion of Atopodium, Clostridium, Gardnerella, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia genera). Hormonal regulation of microbiome •22 Fertile Range 18–35 One endometrial sample taken in menstrual cycle before embryo transfer No variation in bacterial community composition in pre-receptive versus receptive phase in most subjects (n = 18 of 22) Impact on reproductive outcome •35 Infertile, undergoing IVF, receptive endometrium Range 25–40 Negative association of non-Lactobacillus dominated subjects with pregnancy outcome (decreased implantation of 23.1% versus 60.7%; pregnancy rates 33.3% versus 70.6%; ongoing pregnancy rates 13.3% versus 58.8%, and live birth rates 6.7% versus 58.8%). Walther-Antonio et al. (2016) Composition of the uterine microbiome, its role in endometrial cancer •10 Benign uterine conditions (pelvic pain, abnormal bleeding, fibroids, prolapse) Median 44.5 (range 43.5–2.5) Hysterectomy; Uterus, Fallopian tubes and ovaries excised, biopsy and scrapes taken. 1 Lysogeny broth kept open during sample acquisition swabbed, controls of DNA extraction/ microbiome enrichment process. Pre-operative vaginal/ cervical swabs MiSeq (Illumina) V3–V5 Shigella, Barnesiella, Staphylococcus, Blautia, Parabacteroides Bacteroides and Faecalibacterium dominant in cancer group, enrichment of Firmicutes (Anaerostipes, ph2, Dialister, Peptoniphilus, 1-68, Ruminococcus, Anaerotruncus), Spirochaetes (Treponema), Actinobacteria (Atopobium vaginae), Bacteroidetes (Bacteroides and Porphyromonas), and Proteobacteria (Arthrospira). Higher diversity in cancer cohort, especially within the uterus. Presence of Atopobium vaginae and spp. 99% matching P. somerae as predictor of disease status. •4 Endometrial hyperplasia (cancer precursor) Median 54 (range 50.75–2.5) •17 Endometrial cancer Median 64 (range 58–71) Miles et al. (2017) Investigation of microbiota composition within female reproductive tract of women undergoing hysterectomy/ bilateral salpingo-oopherectomy (pilot) •10 Undergoing hysterectomy and bilateral salpingo-oopherectomy Median 49 (range 41–59) Varying route of hysterectomy, transvaginal (1/10), abdominal (1/10), robot-assisted (1/10), and laparoscopic (7/10) 1 Vaginal swabs (preceding surgery) 454 pyro-sequencing on Life Sciences instrument (Roche) V1–V3 Lactobacillus, Acinetobacter, Blautia, Corynebacterium Staphylococcus Various attempts to sequence endometrial samples of patient with atrophic endometrium were negative. Tao et al. (2017) Investigation of endometrial microbial colonization at the time of embryo transfer during IVF using ulta-low bacteria counts •70 Undergoing IVF Average 36.2 (range 22.3–46) Transcervical; distal portion of transfer catheter used for embryo transfer 1 MiSeq (Illumina) V4 Lactobacillus spp 33 samples contained over 90% of Lactobacillus abundance, 50 samples over 70%, Corynebacterium (40 patients), Staphylococcus (38 patients) Streptococcus (38 patients), Bifidobacterium (15 patients) 16 S analysis tested on diluted single- and poly-microbial samples. Sample processing allows reliable results on species classification and abundance with low biomass sampling of 60 bacterial cells. Chen et al. (2017) Investigation of microbiota composition within female reproductive tract •80 Surgery for conditions not known to involve infection (hysteromyoma, adenomyosis, endometriosis, salpingemphraxis) Median 31 (range 22–48) Laparoscopy or laparotomy swabs, transcervical swabs 2 (one obtained transcervically, one surgically) Dry swabs, pre-operative skin area, swabs of gloves used by surgeons. Swabs taken through cervical os to test influence of contamination compared to operative sampling, PBS/physiological saline as diluent negative controls controls for sample processing, DNA extraction, and real-time qPCR (the latter included additional ultrapure water control) Ion PGM™ system sequencing (Thermo Fisher) V4 - V5 Lactobacillus (30.6%), Pseudomonas (9.1%), Acinetobacter (9.1%), Vagococcus (7.3%) and Sphingobium (5%) Higher diversity in uterus compared to vaginal and cervical samples. High similarity between samples taken by entrance through cervical os versus surgical access through abdomen. Microbial profiles different than controls. High reproducibility. Determination of signature OTUs increasing in abundance from vagina to peritoneal fluid. High intra-individual correlation versus clear distinction between upper and lower RT also in inter-individual analysis Estimation of bacterial biomass of female reproductive tract Copy number obtained by qPCR species-specific for four major vaginal Lactobacillus species, divided by corresponding relative abundance based on 16 S rRNA sequencing 1 Lowest biomass compared to other sites of reproductive tract (decreasing from 1010 to 1020 copies/sample at lower third of vagina towards posterior fornix and cervical canal and endometrium with 102–103 copies/sample). Much lower CT value of endometrial samples compared to negative controls. Higher diversity corresponds to lower biomass. Validation if sequencing live bacteria or debris •15 Median 33 (range 24–41) Culturing of live bacteria (additional to 16 S rRNA sequencing) 1 Positive cultures for 5 out of 15 samples; eight different isolates belonging to seven genera such as Lactobacillus, Staphylococcus and Actinomyces CT, cycle threshold; GnRHa, GnRH agonist; OTU, operational taxonomic unit; PGM, personal genome machine; qPCR, quantitative PCR; RT, reproductive tract. Franasiak et al. (2016) characterized the endometrial microbiome of 33 women by collecting material from the tip of the catheter used for ET during ART. Both in the group of non-ongoing and ongoing pregnancies, Lactobacillus and Flavobacterium were identified as the most abundant genera. However, in this initial study only women undergoing ART for unspecified reasons were included. Therefore, it is unclear to what extent these findings represent the colonization of a healthy, fertile endometrium. The same holds true for the study by Tao et al. (2017) assessing the microbiota obtained from catheter tips during ET of 70 women. Lactobacillus was abundantly detected (>90% of OTUs in 33 women, >70% in 50 women). Samples also frequently contained Corynebacterium (40 women), Staphylococcus (38 women), Streptococcus (38 women) and Bifidobacterium (15 women). To achieve reliable sequencing results, particular attention was given to sample preparation, as sampling using remaining tissue from catheter tips only yields ultra-low bacterial cell counts. The authors were able to reliably determine bacterial abundance when more than 60 bacterial cells were present per sample, although no estimation was given of how many cells the catheter tips likely contain. Verstraelen et al. (2016) identified Bacteroides as the dominant genus, present in >90% of the women that were included. All women underwent hysteroscopy in the absence of uterine anomalies. Again, the cohort of women employed sought medical help due to recurrent implantation failure (RIF) or recurrent pregnancy loss. The 12 of the 19 women included showed 75% similarity in bacterial composition (Bray–Curtis dissimilarity of 24.6%, range 13.2–34.3%) due to general high abundance of Bacteroidetes and Proteobacteria taxa. Next to this core of Bacteroidetes and Proteobacteria, five women also presented with L. iners (n = 2), P. amnii (n = 1) or L. crispatus (n = 1) as the most abundant species. Large dissimilarities were observed in two women (79.0 and 90.7% Bray–Curtis dissimilarity) showing Prevotella spp. or L. crispatus dominance, which the authors ascribe to possible contamination from the vagina. Control swabs from the vagina could have given an indication if the distinctive microbial profile of these women indeed reflected an artifact from the transcervical sampling approach but no such data are available. In a study focusing on the relation of endometrial polyps to local microbiota, Fang et al. (2016) included 10 fertile women as a control population in their 16 S analysis of transcervical uterine swabs. Proteobacteria, Firmicutes and Actinobacteria were consistently found at phylum level; Enterobacter, Pseudomonas, and Lactobacillus at genus level. Moreno et al. (2016) were also able to include fertile women (n = 13 for comparison of vaginal aspirates and EF, n = 22 for compositional differences between pre-receptive and receptive phase) of reproductive age (Moreno, et al., 2016). Lactobacillus (71.1%), Gardnerella (12.6%), Bifidobacterium (3.7%), Streptococcus (3.2%) and Prevotella (0.9%) were the most identified genera. Endometrial microbial composition was categorized into either Lactobacillus-dominated (>90% Lactobacillus spp.) versus non-Lactobacillus dominated (>10% bacteria other than Lactobacillus, such as A. vaginae, G. vaginalis, species of the genera Clostridium, Megasphaera, Parvimonas, Prevotella, Sphingomonas or Sneathia). The 18 of the 22 women presented with stable microbiota profiles when comparing pre-receptive and receptive phase, of which 12 women were continuously categorized in the Lactobacillus-dominated group, and six women categorized in the non-Lactobacillus group independent from sampling time point. This indicates that bacterial community composition was relatively stable in most women. Contradictory to the results by Franasiak et al. (2016); Moreno et al. (2016) found an association between Lactobacillus abundance and pregnancy outcome. Implantation was decreased (23.1 versus 60.7%), and pregnancy rates declined (13.3 versus 58.8%) when the women showed a non-Lactobacillus dominated endometrial phenotype at the time of ET. An especially negative impact on reproductive outcome was observed when G. vaginalis and Streptococcus species were present in abundance. These results were independent of pH of the sample, known to be affected by Lactobacillus species (Tachedjian et al., 2017). Unlike the other transcervical sampling approaches, Walther-Antonio et al. (2016) studied samples taken from the uterus, Fallopian tubes, and ovaries removed during hysterectomy (n = 31), in addition to pre-operative vaginal and cervical swabs and scrapes. Their results showed Shigella and Barnesiella as dominant species of the endometrial microbiome in accordance with earlier culture-based observations (Cicinelli, et al., 2008). High abundance of Staphylococcus, Blautia and Parabacteroides was found in benign uterine conditions (n = 10), and Bacteroides, and Faecalibacterium were associated with women who presented with cancer as reason for hysterectomy (n = 17). Miles et al. (2017) also took samples from various tissue types obtained during hysterectomy. Their data on 10 women, comprising a wide range of uterine pathologies, shows high variability. Species such as Lactobacillus, Acinetobacter, Blautia, Corynebacterium and Staphylococcus were abundantly found in some of the women. The authors also note that even after various attempts, no sequencing data could be obtained for endometrium from a patient with atrophic endometrium. Similarly, Chen et al. (2017) investigated the microbiota of different sites of the female reproductive tract accessed through surgery. Tissue was obtained during laparoscopy or laparotomy, therefore, not reflecting healthy, fertile women of reproductive age but the patients enrolled had conditions known not to involve infection. In these conditions, Lactobacillus, Pseudomonas, Acinetobacter, Vagococcus and Sphingobium were frequently detected. When evaluating the results of the recent 16 S endometrial studies, certain species were found in more than one study (Table I). Some of the described findings are addressed in two recent reviews (Franasiak and Scott, 2017; Moreno and Franasiak, 2017), neither including a concluding summary of the species found in the different studies, nor a meta-analysis. Microbiota in the uterus—real or artifact? Until now, available 16S data on uterine microbiota do not allow conclusions on a ‘core uterine microbiome’ due to a number of limitations. When investigating the questioned low-biomass microbiome, studies need to be designed with a special focus on possible contamination. In the following paragraph, we outline the pitfalls of the presented studies. Contamination acquired during sample acquisition and processing It is currently highly debated whether microbiota found to be involved in reproduction, such as in the placenta, are in fact merely the result of contamination and an artifact of the study design (Kliman, 2014). Contamination always has to be a concern when studying 16S data. Especially when investigating a suspected low-biomass microbiome, as found within the uterus, the impact of various (contaminating) handling steps makes it hard to detect low-abundance microbes originating only from the sampling site (van der Horst et al., 2013; Laurence et al., 2014; Salter et al., 2014). To overcome the hurdles associated with the need for highly sensitive detection, the 16S studies on endometrial colonization employed different controls (Table I). To examine the possible contribution of contamination when studying the placenta, Lauder et al. (2016) examined the results from ‘air swabs’ waved in the laboratory space, as well as unused sterile swabs, in comparison to placental tissue samples. They could not distinguish between contamination controls and placental samples. Compositions of microbiota in negative controls are known to be associated with the DNA extraction kits (Kim et al., 2017). Misrepresentation of data needs to be minimized by using special DNA isolation kits for low microbial biomass samples and committing to one kit type for all samples (Kim, et al., 2017). Low-biomass microbial communities are sensitive to contamination and misinterpretation Lauder et al. (2016) suggest the presence of a placental microbiome below their level of detection as a reason for not being able to distinguish between air swabs and sampled material. Detection of a low-biomass microbiome (where for 1 g of placental tissue only 20–2000 ng of bacterial DNA is estimated to be extracted) is a major issue when assessing endometrial microbiota (Antony et al., 2015). Chen et al. included an assessment of the bacterial biomass in their approach. Copy numbers, as calculated from qPCR on vaginal typical Lactobacillus species, were related to the relative abundance of the species as found in 16 S rRNA sequencing, showing that microbial numbers decrease from vagina (1010–1020 copies/sample) through the cervix (108–1010 copies/sample) to the uterus (102–103 copies/sample) (Chen, et al., 2017). As the authors point out, this still represents higher numbers than those that can potentially be detected in background noise (Salter, et al., 2014) or in their controls of sterile PBS, sterile saline, and ultrapure water. Accordingly, even though the detected biomass is small, the detected species are not an artifact or contamination. Another source of misinterpretation that is gaining attention is the approach in which species distribution is related to the obtained sequencing library. Disease-associated changes in Prevotella and Bacteroides abundances found in Crohn’s disease were found to be rooted in proportional profiling and had to be ascribed to a general decrease in microbiota richness in Crohn’s disease (Vandeputte et al., 2017). If abundance is interpreted as a relative proportion, a change in the presence of a certain species presents as increase or decrease of a possibly unrelated species. Low-biomass microbiota, as detected in the uterus, are highly sensitive to this misrepresentation of data. A quantitative assessment of genus abundance proportional to cell counts and microbial load of the sample would allow a more precise representation of data (Vandeputte, et al., 2017). Vaginal microbiota as a source of contamination Due to its high biomass relative to the uterine microbiome and the high abundance of Lactobacillus in the vagina, misrepresentation of species distribution cannot be excluded. Especially when using a transcervical approach, contamination by vaginal bacteria needs to be controlled for. Moreno et al. (2016) compared endometrial and vaginal samples, and only 2 of the 26 sample pairs (two pairs per woman) of EF and vaginal aspirate showed the same microbial profile. While in some samples only minor differences were observed, six paired samples exhibited totally different profiles when women belonged to the group of non-Lactobacilli dominated subjects. In nine of the 13 women, Lactobacillus consistently reflected the dominant genus with more than 90% colonization in the pre-receptive and receptive phase of endometrium and vagina. Notably, some common species, such as A. vaginae, G. vaginalis, Prevotella and Sneathia, of both endometrium and vagina were identified. Also, the study that circumvented this risk, by sampling from uteri obtained during hysterectomy, showed a correlation between the vaginal and uterine microbiomes (Walther-Antonio, et al., 2016). However, it is unclear how the uterine microenvironment is contaminated by transfer of cervicovaginal bacteria into the excised uterus. As manipulation of the uterus during surgery might enable passage of bacteria from the lower reproductive tract, it cannot be excluded that the correlation of vaginal and uterine microbiota is just an artifact. This correlation was also seen by Verstraelen et al. (2016), who detected endometrial phylotypes known to be associated with the vagina. Lactobacillus crispatus was present in 12 of the 19 women, and most abundant in 5 of 19 individuals (range from 17.1 to 79.1% of the total number of sequence reads). Also, species associated with vaginal microbiota imbalance, such as Prevotella (P. amnii, Prevotella sp., P. timonensis, P. disiens), were abundantly expressed in two subjects (18.4 and 55.4% of total sequence reads) (Verstraelen, et al., 2004; Martin and Marrazzo, 2016; Onderdonk et al., 2016). Likewise, G. vaginalis was present in six women, but at an abundance of <1%. While trying to avoid vaginal contamination through the use of a sheath-protected sampling surface and trying to prevent cervicovaginal contact of the sampling tool, it cannot be excluded that insertion or extraction of the device facilitates bacterial carry-over. Verstraelen et al. (2016) pointed out that, even though a certain degree of species overlap exists, it is not to be assumed that endometrial samples reflect vaginal colonization, since the diversity found in endometrial samples is higher than in the vagina. While we believe that it is unlikely that the distribution profiles from endometrial swabs are merely a representation of vaginal colonization, the possibility of carrying over traces from vaginal microbiota onto the sample cannot be neglected. Likewise, insertion of a sampling device or speculum could carry over bacteria from the lower reproductive tract into the uterus. It is also unclear how surgical access of the uterus coincided with disruption of the cervical barrier allowing passage of bacteria that would normally not reside in the uterus as a result of the pre-operative preparation. Live bacteria? It cannot be ignored that, without further measures, sequencing of 16S rRNA does not differentiate between living bacteria or (dead) bacterial fragments. Chen et al. studied cultures of freshly collected peritoneal fluid samples from the Pouch of Douglas (the lowest part of the abdominal cavity, between uterus and rectum), a site shown to be similar in microbial diversity, and even lower in terms of microbial abundance, as compared to the endometrium (Chen et al., 2017). Eight different isolates were cultured belonging to seven genera, such as Lactobacillus, Staphylococcus and Actinomyces, in 5 out of 15 samples, whereas various negative controls did not lead to bacterial growth when cultured. These results indicate that the detected 16S reads do not solely reflect dead bacterial fragments. Conclusions on the validity of the current evidence Based on the current study designs, no information on bacterial detection and distributions of a ‘core uterine microbiome’ can be extracted. Contamination during sampling by surrounding microbiota and reagent kits have not been addressed systematically on a large cohort of healthy, fertile women yet, in order to allow any conclusions on the continuous baseline colonization of the uterus independent of pathologies. Future studies also need to address the question of whether detected 16S results from bacterial fragments rather than live organisms by using a more robust approach than culture in vitro since (even though proving that culture is not impossible) often no growth can be established in cases of these ultra-low biomass samples. The variability in culturing success probably also results from variation between women in species composition and biomass of live bacteria. Aside from this, even 16S rRNA data, reflecting a proportion of the overall community that is dead or fragmented, represent ligands for the host cells to recognize and act upon. Thereby, these inactive bacterial fragments can still contribute to a physiologic interaction with host cells. We conclude that well-setup large cohort studies are needed to define a healthy uterine microbiome. We want to stress that even though no uterus-typical bacterial profile can be established yet, existing data are not to be underrated as evidence towards a natural microbial presence within the uterus. A number of the presented studies do employ negative controls and/or additional vaginal swabs. Since species abundance of these controls was not identical to the results from endometrial material, we should consider a physiological importance of local microbiota. Endometrium—starting point for decidualization, implantation and placentation To understand how microbiota could play a role in the uterine interplay of cells involved in the per-implantation period, we first need to understand how the endometrium forms the basis for successful implantation and placentation. Despite the growing interest from both clinicians and scientists in the process of implantation, the mechanisms underlying human implantation remain poorly understood. Each month, before the presence of a developing embryo, the stage for formation of the placenta is set. During the mid-to-late luteal phase spontaneous decidualization occurs; i.e. the transformation of the endometrium into a receptive state, independent of the presence or absence of a conceptus. And 6–10 days after ovulation (during Days 19–24 of the menstrual cycle) the receptive state of the endometrium is shaped. This is referred to as ‘the window of implantation’ (WOI), lasting for 2–4 days (Navot et al., 1991; Wilcox et al., 1999). The substantial physiological adaptations needed for correct cyclical changes of the endometrium are orchestrated by fluctuations of progesterone and oestrogen (Jabbour et al., 2006; Henriet et al., 2012), as well as by immune cells and their products (Salamonsen et al., 2002; Singh et al., 2011). Endometrial stromal fibroblasts differentiate to become larger, rounded fibroblast-like stromal cells. To achieve this, the cytoskeleton and plasma membrane undergo modifications (Thie et al., 1995; Martin et al., 2000; Murphy, 2000). The luminal epithelium must be able to interact with the blastocyst. Interactions also involve adhesion molecules, such as integrins, L-selectin ligands (e.g. L-selectin ligands expressed by the luminal epithelium and L-selectin receptors of the blastocyst) and oligosaccharides (Cross et al., 1994; Prakobphol et al., 2006; Cha et al., 2012; Kang et al., 2014). We summarized key endometrial adaptations in the peri-implantation period (Fig. 1) to show the importance of enabling the correct interaction of endometrium and blastocyst. Figure 1 View largeDownload slide Illustration of key elements in blastocyst–endometrium interaction needed for a receptive endometrium. Due to its important contributions during early placentation, the endometrium is key to healthy pregnancy. During the window of opportunity (Days 19–24 of the menstrual cycle), the endometrium resides in a receptive state that allows selection, apposition and attachment of a healthy blastocyst. Implantation is marked by invasive growth and differentiation of trophoblast cells. Proper interaction of invading trophoblast and local immunity is needed to achieve correct villi development and a healthy placenta. All these highly regulated properties of the endometrium needed in the initial phase of pregnancy are possibly affected by uterine microbiota. IVS, intervillous space; uNK, uterine natural killer cell. Figure 1 View largeDownload slide Illustration of key elements in blastocyst–endometrium interaction needed for a receptive endometrium. Due to its important contributions during early placentation, the endometrium is key to healthy pregnancy. During the window of opportunity (Days 19–24 of the menstrual cycle), the endometrium resides in a receptive state that allows selection, apposition and attachment of a healthy blastocyst. Implantation is marked by invasive growth and differentiation of trophoblast cells. Proper interaction of invading trophoblast and local immunity is needed to achieve correct villi development and a healthy placenta. All these highly regulated properties of the endometrium needed in the initial phase of pregnancy are possibly affected by uterine microbiota. IVS, intervillous space; uNK, uterine natural killer cell. Already at this stage, irreversible shortcomings in placental vascularization pose an origin for diseases such as pre-eclampsia or intra-uterine growth restriction (Conrad et al., 2017). On the maternal side, the underlying mechanisms regulating correct migration of the trophoblast and vascular remodeling may be rooted in the endometrium. Due to its importance for correct placentation, the endometrium is crucial for developing a healthy placenta, and therefore a healthy pregnancy (Burton et al., 2010). If microbial compounds constitute another physiologic player in the complex uterine environment, its natural impact will likely also contribute to implantation and placentation. Uterine immunity during implantation and placentation as a possible interaction partner of local microbiota Local immune cells belong to the factors known to influence extravillous trophoblast migration, and are therefore instrumental to maternal spiral artery remodeling (Moffett et al., 2015). The immune system is intimately involved in all aspects of the reproductive process, particularly around the time of conception, and in the peri-implantation period (Robertson et al., 2011). Different stages during the initiation of pregnancy hold specific immunologic challenges: to enable a healthy pregnancy, the maternal immune system must support the introduction of semen to allow fertilization, the initial contact of blastocyst and endometrium, and correct formation of the placenta (Robertson and Moldenhauer, 2014). All of these processes are subjects of extensive study, and a comprehensive overview is beyond the scope of a single review. Natural killer (NK) cells, T cells and antigen presenting cells (APCs) are a common focus when investigating the immunology of pregnancy (Emmer et al., 2000; Moffett and Colucci, 2014; Moffett et al., 2004; Plaks et al., 2008; Tilburgs et al., 2006, 2009; Trundley and Moffett, 2004; van der Meer et al., 2004). We wish to highlight the importance of a tightly regulated uterine immune environment that could potentially be affected by microbiota-induced factors. Recent discoveries on the interplay of the intestinal microbiome, their metabolites, and host immunity, give an impression of the vast impact that the endometrial microbiome could have (Hooper et al., 2012; Postler and Ghosh, 2017; Round and Mazmanian, 2009). We outline here the implications of gut microbiota affecting cell types known to be important within the uterus. Lymphocytes involved in the peri-implantation period NK cells are the most abundant of immune cells in the endometrium representing ~70% of all hematopoietic cells present (Bulmer et al., 1991; Moffett-King, 2002). The term endometrial NK cells (eNK; or uterine NK cells; uNK cells) refers to NK cells in the non-pregnant endometrium, and decidual NK (dNK) cells residing in the placental membrane are phenotypically and functionally different from eNK cells or peripheral blood NK (pbNK) cells (Moffett-King, 2002; Male et al., 2011). In the endometrium, the non-cytolytic and potent cytokine secretors, CD56brightCD16−NK cells, dominate. Only a small fraction of eNK cells belongs to the CD56+CD16+NK population specialized in killing infected cells (van der Meer, et al., 2004; Hanna and Mandelboim, 2007), contrary to the NK cell division in peripheral blood where the majority are CD56+CD16+. Throughout the female cycle, eNK cells continue to accumulate until the first half of gestation. Abnormal levels of NK cells, measured 20 days after ET, were shown to be associated with RIF (Ledee-Bataille et al., 2004). Direct involvement of eNK cells in implantation remains unclear, as successful implantation is possible in NK cell deficient mice (Le Bouteiller and Piccinni, 2008). More evidence points towards an involvement of dNK cells in placenta-formation processes after initial implantation. However, an indirect role of the human CD56brightCD16−eNK cells already during implantation cannot be neglected due to their high abundance and ability to secrete large amounts of cytokines, such as IFN-γ, tumor necrosis factor alpha (TNFα), granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin 10 (IL-10) (Hanna and Mandelboim, 2007). Just as eNK and dNK cells are distinct from pbNK cells, the mucosal NK cells of the gut that are continuously exposed to the gut microbiome are distinct from pbNK cells. This distinction is characterized by limited IFNγ production and an absence of cytotoxic effector perforin, granzymes, FasL and TNF-related apoptosis-inducing ligand (TRAIL) in mucosal NK cells of the gut (Satoh-Takayama et al., 2008). Of note, germ-free mice lacked these IL-22+NKp46+ cells (Sanos et al., 2009). Hence, the specialized intestinal NK cells might be the result of an adaptation to the microbiota-rich environment. It needs to be evaluated whether the reduced cytotoxic capacity of uNK cells (Manaster and Mandelboim, 2010) has any relation to the uterine microbiota. APCs, such as macrophages (Mɸ) and dendritic cells (DCs), constitute 10-20% of uterine leukocytes, expressing major histocompatibility complex class II molecules (Dekel et al., 2010; Mori et al., 2016). Through integrating microbial, environmental, and self-derived stimuli, APCs are key to the successful initiation of an appropriate immune response (Medzhitov and Janeway, 1997; Swiatczak and Rescigno, 2012). Throughout the menstrual cycle, mature CD83+ DCs and CD68+ Mɸ increase, peaking in the late secretory phase (Rieger et al., 2004). Mɸ are, together with NK cells, the main cytokine producers in the human endometrium (Kämmerer et al., 2004). Through secreting leukemia inhibitory factor (LIF) and IL1B, Mɸ play a known role in the establishment of endometrial receptivity by increasing cell surface fucosylated structures, allowing trophectoderm attachment (Robertson and Moldenhauer, 2014). Even more pronounced than seen for NK cells, monocytes of the heavily colonized gut mucosa are markedly different in phenotype and function compared to blood-derived monocytes. Even though they reside in close proximity to bacteria colonizing the intestinal lumen, the ability of the local macrophages to mount a pro-inflammatory response seems to be tuned down. Their expression of, for example, lipopolysaccharide (LPS)-binding protein CD14 receptor (Smith et al., 1997), CD89, implicated in IgA-enhanced phagocytosis is decreased (Smith et al., 2001). Also, various FCγ receptors (CD16, CD32, CD64), integrins (CD11a, CD11b, CD11c, CD18) and the IL-2 receptor CD25 are, amongst others, downregulated on intestinal macrophages. Their ability to produce cytokines (IL-1, IL-6, TNFα) is diminished (Smythies et al., 2005). It needs to be investigated whether this Smad-induced IκBα expression and NF-κB inactivation-mediated phenotype, termed ‘inflammation anergy’, is paralleled at the fetal–maternal interface (Smythies et al., 2010). T cells represent the third largest fraction of immune cells found in human endometrium (van der Molen et al., 2014; Shanmugasundaram et al., 2016; Feyaerts et al., 2017). Since, in the absence of pregnancy, T cells mostly reside in the deeper layers of the endometrium, their relation to fertility disorders more likely reflects a role in early placenta formation after implantation, rather than a contribution to the pre-implantation period. Experimental evidence on the functional importance of T cells in the WOI is scarce. It has been suggested that a switch in the ratio of Th1 and Th2 cells (Th1/Th2 paradigm) within the endometrium is needed to prepare for implantation, as a pronounced decrease in Th1 was accompanied by an increase in Th2 cells seen from the secretory phase towards early pregnancy (Saito et al., 1999). More recently, regulatory T (Treg) cells and Th17 cells have been included in the discussion (Figueiredo and Schumacher, 2016; Saito et al., 2010). Increasing evidence on the shaping of T cell subsets by microbiota based on studies of the gut is emerging (Omenetti and Pizarro, 2015). Like in the endometrium, mucosal immunity of the intestines depends on an interplay of IFN-γ producing Th1 cells, Th17 cells that secrete IL-17 and IL-22, and innate lymphoid cells with Th2-cytokine secreting functions, and suppressive Foxp3+ Treg cells. Differentiation of various T cell subsets is impaired in mice with disrupted gut microbiota (Ivanov et al., 2008; Gaboriau-Routhiau et al., 2009). Commensal segmented filamentous bacteria were shown to be important for Th17-mediated mucosal protection (Ivanov et al., 2009). A detailed understanding of the mechanisms underlying bacteria-mediated T cell induction is not yet available. One of the few examples in which the link between microbe and host is known, is polysaccharide A (PSA) originating from the capsule of Bacteroides fragilis (Ivanov, et al., 2009). Bacteroides, found in a number of the endometrial studies analyzing 16S rRNA, tunes the Th17 response of intestinal T cells, and causes a systemic increase of circulating CD4+ T and Th1 cells (Johnson et al., 2015; Mazmanian et al., 2005, 2008). Other microbiota are known to induce and cause accumulation of Treg cells, essential in maintaining tolerance and known to be important for pregnancy (Saito et al., 2005). In the gut, Treg mediated IL-10 production is thought to be critical to maintain intestinal microbial homeostasis by limiting continuous activation of Th1 and Th17 cells. Commensal microbiota promote de novo generation and activation of mucosal Treg cells (Geuking et al., 2011) as shown by mice colonized with altered Schaedler flora species (Lactobaillus, Bacteroides, species of Flexistipes phylum and gram-positive bacteria of the Firmicutes, Bacillus-Clostridium group) (Dewhirst et al., 1999; Geuking, et al., 2011). Focusing on specific strains, colonization of mice with Clostridia strains was shown to induce expansion of Treg cells locally in the lamina propria and systemically, resulting in increased resistance to colitis and IgE responses in adult mice (Atarashi et al., 2011). These intestinally induced Treg were shown to express a T-cell receptor repertoire specific for bacteria of the luminal content of colonized mice, allowing tolerance towards individual species (Lathrop et al., 2011). Similar generation and activation of immune cell types by endometrial microbiota could be essential for correct homeostasis regarding maintenance of a microbiome and programming the local immunity within the uterus. The majority of work on microbiota and immunity focuses on T cell homeostasis, but especially cells of the innate arm of the immune system are involved in the peri-implantation period. Cytokines and chemokines belong to the soluble factors known to be associated with reproductive health. Induced by either endometrial cells or immune cells recruited upon blastocyst encounter, the implantation period is characterized by an increase in pro-inflammatory Th1 cytokines, such as IL-6, IL-8, LIF and TNF-α, with implications for immune cell recruitment and activation of the endometrium, as reviewed previously (van Mourik et al., 2009; Dominguez et al., 2010). Throughout the female cycle, healthy endometrial stromal cells constitutively secrete the chemokine CCL2, also known as monocyte chemoattractant protein 1 involved in directing monocytes, T cells and DCs to inflammatory or tumor tissue (Carr et al., 1994) (Shi and Pamer, 2011; Li et al., 2012). Also, decidual stromal cells obtained during first trimester (~8 weeks in both studies) secrete CCL2, which was shown to attract Th2 and Th17 cells, thereby influencing the local T cell balance (He et al., 2012; Wu et al., 2014). In addition, macrophage recruitment and polarization (Sierra-Filardi et al., 2014), and plasmacytoid DCs were affected by CCL2. It was recently shown that microbiota are important in establishing baseline CCL2 secretion controlling homeostatic trafficking of plasmacytoid DCs (Swiecki et al., 2017). As no functional contribution of endometrial microbiota has been shown yet, the physiologically closest possible microbiome, as indicated by Chen et al. (2017), is probably the cervicovaginal microbiome. Cytokine concentrations were shown to be affected depending on the microbial profiles of 16S rRNA analysis of cervicovaginal lavage. This was reflected in lower IL-1beta, IL-8 and IL-10 levels in bacterial vaginosis (BV) (Santos-Greatti et al., 2016). Also other cytokines, such as LIF, GM-CSF, colony-stimulating factor-1, heparin binding epidermal growth factor-like growth factor, insulin-like growth factor Iand II, contribute to fertility through their role in supporting blastocyst development in the pre-implantation phase (Robertson and Moldenhauer, 2014). Decreased LIF expression of endometrial epithelial cells was shown to be associated with infertility (Margioula-Siarkou et al., 2016). Cervicovaginal levels of cytokines were affected by the bacterial flora found (Marconi et al., 2014) and a probiotic supplement was shown to affect vaginal interleukin changes (Bisanz et al., 2014). It remains to be seen if different uterine microbial profiles could be associated with altered chemokine/cytokine profiles and effects on reproductive success. Lessons learned from the gut microbiome: possible interactions between local microbiome and endometrium Besides influence on systemic and tissue resident lymphocytes, the microbiota and host interact on various levels. The relation between microbes and host is not a black and white phenomenon; and thus surpasses the simplified view of commensal versus pathogen. The immune system shapes a homeostatic coexistence, i.e. by eliminating species that might not only be harmful but also disturb the balance (Eberl, 2010). Two different types of battles are continuously fought when hosting a microbial community: one that is responsible for facing the daily challenge posed by containment of symbiotic microbes, and one for defense against pathogenic microbes that breach containment (Sansonetti, 2004; Eberl, 2010). Although our understanding of the microbiome–host interaction is still rudimentary, its functional importance, especially on various arms of innate and adaptive immunity, is evident (Cénit et al., 2014; Littman Dan and Pamer Eric, 2011). Below, we give an overview of the different mechanisms of how endometrial colonization might have implications for the uterine immune-environment, and thus for fertility. Pattern recognition receptors To detect viral, fungal and bacterial pathogens, the innate immune system senses pathogen-associated molecular patterns (PAMPs) through pattern recognition receptors (PRRs), such as toll-like receptors (TLRs), RIG-I-like receptors, NOD-like receptors (NLRs) and C-type lectin receptors, such as collectins, selectins, endocytic receptors and proteoglycans (Tschopp et al., 2003; Oger et al., 2009; Takeuchi and Akira, 2010; Botos et al., 2011). PAMPs include, for example, cell wall mannans (yeast), formylated or lipo-peptides, peptidoglycans, teichoic acids and bacterial cell-wall components such as LPS (Janeway and Medzhitov, 1998). Activation of TLR induces nuclear factor-kappa B, which initiates an inflammatory cascade including upregulation of markers to mount an adaptive immune response (Medzhitov et al., 1997). PRRs form the first line of defense against sexually transmitted disease or any other pathogens that can access the female reproductive tract through the vagina (Wira et al., 2005). For example, decidual and uterine stromal and epithelial cells express various TLRs and other PRRs, and the cells mount a potent inflammatory response upon receptor recognition (Young et al., 2004; Schaefer et al., 2005; Aflatoonian and Fazeli, 2008; Ghosh et al., 2013; D’Ippolito et al., 2016; Anders et al., 2017). An overview of TLRs of the endometrium is given in Table II. The expression of TLR mRNA and protein changes throughout the menstrual cycle, as shown by different groups (Jorgenson et al., 2005; Aflatoonian et al., 2007; Hirata et al., 2007; King et al., 2009; Nikzad et al., 2013). Receptor expression on endometrial cells was found to be low in the proliferative phase and increased in the secretory phase for TLRs 1–10. Together with intracellular receptors, there is a constitutively high expression of pathogenicity sensors lining the female reproductive tract (Ghosh, et al., 2013). All of these receptors can sense their specific PAMP and react by inducing signaling cascades. Table II Pattern recognition receptor expression on endometrial cells. Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) CpG, Cytosin-phosphatidyc-guanin; dsRNA/DNA, double-stranded RNA/DNA; EEC, endometrial epithelial cells; ESC, endometrial stromal cells; LPS, lipopolysaccharide; LTA, lipoteichoic acid; MDA-5, Melanoma Differentiation-Associated protein 5; nd, not defined; (x), PAMPs, pathogen-associated molecular pattern; presence shown for whole endometrial sample; RIG-1,retinoic acid-inducible gene 1; ssRNA, single-stranded RNA; TLR, toll-like receptor. Table II Pattern recognition receptor expression on endometrial cells. Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) Pattern recognition receptor Expressed on EEC ESC Cyclical changes Pathogen-associated molecular pattern Pathogen Reference Toll-like receptor  TLR1 Plasma membrane x (x) – Lipid-containing PAMPs, e.g. LTA Gram pos. bacteria, fungi Aflatoonian et al. (2007); Young et al. (2004) (together with TLR2/6)  TLR2 Plasma membrane x x x Lipid-containing PAMPs, e.g. LTA (together with TLR2/6), zymosan Gram pos. bacteria, fungi Aflatoonian et al. (2007, Hirata et al. (2007); Young et al. (2004)  TLR6 Plasma membrane x (x) x See TLR1/2 Aflatoonian et al. (2007)  TLR4 Plasma membrane x x x LPS, LipidA Gram neg. bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR5 Plasma membrane x (x) x Flagellin Flagellated bacteria Aflatoonian et al. (2007); Young et al. (2004)  TLR3 Endosome x x x dsRNA Virus Aflatoonian et al. (2007); Hirata et al. (2007), Jorgenson et al. (2005); Schaefer et al. (2005); Young et al. (2004)  TLR7 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR8 Endosome x x – ssRNA Virus, bacteria Aflatoonian et al. (2007)  TLR9 Endosome x x x CpG-rich unmethylated ssDNA Virus, bacteria Aflatoonian et al. (2007); Hirata et al. (2007); Young et al. (2004)  TLR10 Endosome x x x Unknown Aflatoonian et al. (2007) Nod-like receptor  Nod1 Cytoplasm x Low – Peptidoglycan (gram pos. bacteria cell wall component) Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Nod2 Cytoplasm x x x Muramyl dipeptide Gram pos./neg. Bacteria Ghosh et al. (2013); King et al. (2009)  Inflammasome Cytoplasm x Low nd Muramyl dipeptide Gram pos./neg. Bacteria D’Ippolito et al. (2016) RIG-1-like receptor  RIG-1 Cytoplasm x Low nd Short/5′triphosphate dsRNA Virus Ghosh et al. (2013)  MDA-5 Cytoplasm x Low nd dsRNA, preference for long segments Virus Ghosh et al. (2013) C-type lectin  E.g. Dectins, mannose-binding lectin Plasma membrane/endosome/secreted (x) (x) nd β-Glucan, zymosan (yeast cell wall), carbohydrates Virus, fungi Oger et al. (2009) CpG, Cytosin-phosphatidyc-guanin; dsRNA/DNA, double-stranded RNA/DNA; EEC, endometrial epithelial cells; ESC, endometrial stromal cells; LPS, lipopolysaccharide; LTA, lipoteichoic acid; MDA-5, Melanoma Differentiation-Associated protein 5; nd, not defined; (x), PAMPs, pathogen-associated molecular pattern; presence shown for whole endometrial sample; RIG-1,retinoic acid-inducible gene 1; ssRNA, single-stranded RNA; TLR, toll-like receptor. Evidence accumulates that PRRs are one way for microbiota and host to communicate. Round et al. (2011) examined how B. fragilis induces symbiosis via TLR2 of CD4+ T cells. PSA of B. fragilis induces Foxp3+ regulatory cells that are needed to counteract the otherwise Th17-mediated killing of Bacteroides (Round et al., 2011). Nod2-deficient mice were more susceptible to pathogenic infection (Petnicki-Ocwieja et al., 2009). Nod2 was shown to control dysregulated commensal colonization, and commensal resident bacteria are in turn needed to induce Nod2 expression. TLRs and NLRs are suspected to play a role in the periconceptual regulation as they are major players in the cascade of cytokine induction (Robertson, et al., 2011; Sirota et al., 2014). Protection against pathogens Maintaining colonization of commensal bacteria protects the host against pathogens. Residential microbiota successfully compete for the niche by exhibiting better adaptation than an occasionally invading pathogenic species, termed the ‘colonization resistance’ concept (Costello et al., 2012; Buffie and Pamer, 2013). Commensal microbiota are specialized in nutrient usage of their habitat, and deplete the environment of nutrients needed for pathogenic species. Consumption of a limited resource can starve the pathogenic invader (Kamada et al., 2013). E. coli, for example, competes in consumption of sugars, amino acids and other nutrients with the pathogenic enterohaemorrhagic E. coli (EHEC) (Fabich et al., 2008; Leatham et al., 2009). Symbionts also defend their niche (Ubeda et al., 2017). An example is the commensal Clostridium scindens whose colonization protects against C. difficile infection by production of secondary bile acids (Buffie et al., 2015). Another contribution by commensals is continuous receptor stimulation leading to TLR upregulation; needed to sense potentially dangerous bacteria (Vaishnava et al., 2008). Mice and humans treated with antibiotics, diminishing the intestinal microbiota, are highly susceptible to antibiotic-resistant strains because of a reduced expression of antimicrobial defense mechanisms (Brandl et al., 2008). In the female reproductive tract, inhibition of gonococci (Neisseria gonorrhoeae) adherence is observed in the presence of Lactobacilli when using an in vitro model of endometrial epithelial cells (Spurbeck and Arvidson, 2008). In an endocervical epithelial cell model, TLR agonists, as a surrogate for microbial products, were shown to stimulate antimicrobial products and mucins (Radtke et al., 2012). Maintaining commensal colonization within the endometrium may likely offer a similar means of effective protection against uterine infections. Tissue adaptation In addition to lymphocytes, epithelial cells play a crucial role in coexistence with commensal colonization. An intact epithelium allows for safe colonization without extensive (pathogenic) barrier breach, through a physical barrier and reaction of the epithelial cells towards bacterial ligands. On the other hand, bacteria also contribute to a healthy barrier development. Increasing evidence highlights the importance of microbiota for gut development and morphogenesis. Known adaptations of microbiota-implicated tissue adaptations include intestinal epithelial cell differentiation (Yu et al., 2016), support of epithelial cell regeneration (Banasaz et al., 2002), modulation of epithelial cell permeability (Cario et al., 2007), and vascularization (Stappenbeck et al., 2002). In the uterus, commensal bacteria might contribute to the remodeling needed for a receptive state of the endometrium. The genus Bacteroides, constituting 30% of the endometrial community in 90% of women, is known to be tightly involved in intestinal tissue development by influencing epithelial cell maturation and maintenance (Wexler, 2007; Maier et al., 2015). Hooper et al. (2001) showed that B. thetaiotaomicron induced gene expression of proteins involved in nutrient absorption, angiogenesis, intestinal maturation and mucosal barrier reinforcement. Bacteroides was shown to be essential for intestinal vascularization through signaling mediated by Paneth cells, the main type of epithelial cell in the intestine secreting antimicrobial compounds (Bevins and Salzman, 2011). Further studies using endometrial epithelial cells are warranted to ascribe similar physiological functions to the uterine microbiome during endometrial remodeling. The continuous reparative processes following the secretory phase might benefit from microbial support mechanisms. Dysbiosis in this critical period of preparing for blastocyst invasion might be one of the unrevealed causes leading to RIF due to failed remodeling. Another interesting tissue adaptation involved in the peri-implantation period is the change in barrier function of the endometrial epithelium. Features of tight junctions, such as area of whole junction and the area enclosed by junctional strands, decrease from Days 13 to 22 of the menstrual cycle, which could be beneficial for implantation (Murphy et al., 1992). Redistribution of adherens junctions and desmosomes during the WOI may also prepare for trophoblast invasion (Buck et al., 2012). The tight junctional barrier function is regulated by nutrients and cytokines (Turner, 2009). As a side effect of increased barrier permeability, more barrier breach of the uterine microbiota can occur. This in turn can lead to a more pro-inflammatory immune environment with mucosal immune cells stimulated to secrete the cytokines that are beneficial for the peri-implantation period. Discussion Conclusions on the possible presence of uterine colonization Even before the era of 16S RNA analysis, it was put forward that ‘it is difficult to envision that a mucosa that is continuously exposed to microorganisms present in the lower genital tract and that is regularly invaded by sperm that can carry microorganisms into the endometrial cavity, may be free of bacteria and its products such as endotoxin’ (Romero et al., 2004). We also agree that, based on the various lines of evidence for the existence of an endometrial microbiota, the era of the sterile uterus, free of any microbial compounds, has to come to an end. Taken together, this variety of immunological adaptations due to microbial presence contributes to establishing the right balance of tolerance versus reactivity towards bacterial ligands, and likely has a physiological impact on pregnancy (Fig. 2). Following the suggestion by Espinoza et al. (2006), it can be envisioned that as long as no pathologic pro-inflammatory response against the microorganisms of the endometrial mucosa is started, the presence of a certain microbial coexistence does not harm the conceptus, and might well contribute to vital processes in the peri-implantation period and pregnancy. Figure 2 View largeDownload slide Uterine microbiota may contribute to healthy endometrium physiology. Upon pattern recognition receptor stimulation, epithelial cells release soluble factors, such as cytokines, affecting local lymphocyte populations. If bacteria are naturally present in the uterus, similar interaction with host cells as seen in mucosa of the gut can be envisioned. Local lymphocytes, e.g. antigen presenting cells, could sense microbes by probing through the epithelia, or upon barrier breach, thereby initiating a signaling cascade. Presence of commensals might alter the mucosal T cell balance and the involved cytokines can have an effect on the local immune environment. Cells important for healthy implantation and placentation, such as uNK cells, are potentially affected. Therefore colonization might have an impact on the important initial steps of endometrial physiology. APC, antigen presenting cell; IL-10, interleukin 10; PSA, polysaccharide A; PRR, pattern recognition receptor; Th17, T helper 17 cells; Treg, regulatory T cell. Figure 2 View largeDownload slide Uterine microbiota may contribute to healthy endometrium physiology. Upon pattern recognition receptor stimulation, epithelial cells release soluble factors, such as cytokines, affecting local lymphocyte populations. If bacteria are naturally present in the uterus, similar interaction with host cells as seen in mucosa of the gut can be envisioned. Local lymphocytes, e.g. antigen presenting cells, could sense microbes by probing through the epithelia, or upon barrier breach, thereby initiating a signaling cascade. Presence of commensals might alter the mucosal T cell balance and the involved cytokines can have an effect on the local immune environment. Cells important for healthy implantation and placentation, such as uNK cells, are potentially affected. Therefore colonization might have an impact on the important initial steps of endometrial physiology. APC, antigen presenting cell; IL-10, interleukin 10; PSA, polysaccharide A; PRR, pattern recognition receptor; Th17, T helper 17 cells; Treg, regulatory T cell. Additional studies required to define the core microbiome Sample cohort Even though several studies have evaluated the microbiome of the endometrium, a ‘baseline’ or ‘core’ microbiome of this healthy endometrium has not been defined. The studies presented here on 16S rRNA evaluation of endometrial microbiota vary to a large extent in experimental approach. Differences in sample population, sample acquisition method, 16S region, controls and bacterial genotyping make it impossible to combine the data in a more powerful meta-analysis. Of note, all studies enrolled small and highly selected cohorts. Except for the studies by Fang et al. (2016) assessing 10 healthy women, and Moreno et al. (2016) including 22 healthy women, all studies enrolled subjects seeking treatment because of reproductive complications or uterine abnormalities (Fang, et al., 2016; Moreno, et al., 2016). Therefore, the majority of data does not reflect a putative steady-state healthy endometrial microbiome. Only recruitment of healthy, fertile women would provide insight into the microbiome of a fertile, receptive endometrium. However, without underlying medical reasons, sampling in healthy women is unethical and therefore a major limitation to investigating uterine microbiota. Controlling for vaginal and environmental contamination A transvaginal approach carries the risk of contamination by the cervicovaginal microbiome. Due to the high biomass of the vaginal relative to the uterine microbiome and the high abundance of Lactobacillus, misrepresentation of species distribution cannot be excluded. To achieve more conclusive results in the near future, transvaginal studies on endometrial microbiota need to strictly incorporate control samples from vagina and cervix. Additionally, contamination by reagents needs to be controlled for by including controls of all extraction kits used, a common source for misleading data (Glassing et al., 2016). Also, air swabs and swabs of other environmental sources of microbial exposure need to be taken into account when trying to establish conclusive results on the low-biomass microbiome of the uterus. Standardized sample processing In the presented studies, major differences in the processing of material, data collection and analyses are all sources of variation (Gohl et al., 2016). The choice of hypervariable region (V1–V9) of the 16S rRNA, containing species-specific sequences used for assessing microbial diversity, influences the detection results. V1 for example was found to be well suited to distinguish between S. aureus species, whereas V2 was more successful in Mycobacterium species differentiation (Chakravorty, et al., 2007). Depending on the variable region of the 16 S rRNA, different species are preferentially detected. A standardized approach in which study design, sampling method, DNA extraction with generation of 16 S amplicons and sequencing, assigning of OTUs to genus/species, and reporting of the findings is needed (Parnell et al., 2017) to achieve conclusive results for endometrial and placental microbiome studies. Without these standardized, well-controlled study designs the uterine microbiome will still not come into its own in the future. Points of interest Timing of intervention Owing to the dynamic fluctuations of microbiota in the uterus, the timing of microbial influence(s) on a healthy pregnancy needs attention. Depending on when the pregnant (or soon-to-be pregnant) woman undergoes any external microbial (pro-/pre or antibiotic) intervention, this might alter the physiologic processes of implantation, placenta formation or placenta maintenance if microbiota are involved. There is increasing evidence that BV is associated with infertility and pregnancy complications (Campisciano et al., 2017). Interestingly, the risk of delivering preterm because of BV is not reduced after clearance of infection by antibiotics (Pelzer et al., 2016). This might be explained by the fact that the effects of BV have already taken place and the initiated cascade of changes induced by the bacterial imbalance cannot be reversed. Accordingly, treatment of BV before 20 weeks of gestation might decrease the risk of BV-associated preterm birth, suggesting that the early phase of pregnancy in particular is sensitive to microbial impact (McDonald et al., 2007). Systematic reviews on other conditions of bacterial imbalance associated with pregnancy and their treatment might narrow down the window in which a certain microbe-associated effect on pregnancy takes place. In the future, robust evidence for a uterine microbiome might also fuel novel options for intervention concerning fertility and prevention of preterm birth risk in women with a history of BV. Chicken or egg It is possible that the state of inflammation is a sign of dysregulated immunity, rather than only a result of the presence of pathogenic, bacterial colonization. Chronic endometritis, the persistent inflammation of the endometrial lining, is associated with RIF (Johnston-MacAnanny et al., 2010) and successful antibiotic treatment of endometritis could clear the condition (as characterized by accumulation of plasma cells in biopsy samples), but clearance did not improve the chance of implantation. In this case, a certain microbial profile might not represent the cause but, instead, the effect of an immune state that is unfavorable for pregnancy. Interestingly, bacteria found in the endometrium causing the inflammation were not found in the vagina, suggesting that they are commensals in a dysbiotic state, rather than pathogenic intruders (Cicinelli, et al., 2008). However, at present, we cannot exclude an active effect of microbiota on the physiological changes needed for a receptive endometrium. Functional studies are needed to understand the host–microbe interaction in the uterus. More robust information on the species involved in uterine colonization will allow further functional assessment using in vitro endometrial models (Laniewski et al., 2017). Next to this, the association of bacterial metabolites as a cause or consequence of certain microbial colonization and the state of fertility needs to be examined. Yeoman et al. (2013) evaluated metabolomic profiling as a fast and cost-effective diagnostic tool for BV (Yeoman et al., 2013). It remains to be elucidated whether uterine metabolites can be a proxy for the infertility related to dysbiosis. Also, since naturally occurring microbes continuously breach the endothelial barrier and are a part of homeostasis, microbes will continuously shape the immune environment. How this impacts the tightly regulated immunological processes involved from endometrium preparation towards a healthy pregnancy is yet to be discovered. Dynamic bugs Recently, the importance of the microbiome for regulation of rhythmic biological changes has been stated (Thaiss et al., 2016). Fluctuating microbial community structures might direct hormonal changes (Kundu et al., 2017). It is tempting to speculate that similar dynamics are involved in the female cycle. Moreno et al. (2016) observed differences in the microbiota detected in the pre-receptive versus receptive phase in the endometrium. The extent of this effect varied. However, also for the vaginal microbiota it was shown that while in some women low constancy and high species turnover is observed, the extent of these changes can be very small for others (Gajer et al., 2012). Until now, sample size prohibits conclusions on putative characteristic changes of the endometrial microbiota associated with the induction of the WOI. More in-depth studies on changes in the endometrial microbiome are needed to assess whether microbiota play a role in orchestrating the cyclic endometrial changes and establishing a receptive endometrium. Studies examining 16S rRNA from endometrium alongside first trimester and term placental samples are needed to show if, and how far, the microbes of the fetal–maternal interface relate to the pre-pregnancy state. Origin The origin of the endometrial, and especially the placental, microbiome is highly discussed. Different routes of colonization have been suggested, including the gut, vagina or oral cavity. We believe that with placental development originating from decidualized endometrium, it would be highly unlikely that the endometrial and placental microbiome are completely independent. In the intimate relationship of endometrium/decidua and placenta, the trophoblast cells invading the uterine mucosal tissue have to be in contact with the residential microbiota. Systematically comparing the healthy, pre-pregnancy endometrium to placental samples will help to elucidate whether the placental colonization results from the bacteria already present in the uterus. A vaginal origin is one of the suggested routes of uterine colonization, as a number of species seem to be residents of both body sites. Transfer of labeled macrospheres by a ‘uterine peristaltic pump’ activity from vagina to uterus has been shown (Kunz and Beil, 1997; Zervomanolakis et al., 2007). Evidence from mouse studies points out that a change in TLRs and AMP expression and function allows the passage of vaginal bacteria through the cervix into the uterus upon cervical infection. Remarkably, in non-pregnant mice, bacteria could travel from the lower reproductive tract into the uterus, but this was prohibited during pregnancy. These findings suggest that a healthy, uninfected cervix, regulates the ascents of vaginal bacteria (Racicot et al., 2013). This theory is supported by earlier findings of differences in biochemical and chemical properties of the cervix depending on the pregnancy state (Hassan et al., 2010; Hein et al., 2002). In line with theories on oral transmission of bacteria into the uterus, an association of periodontal disease and pregnancy complications is highly discussed (Ren and Du, 2017). Indeed, bacteria such as Lactobacillus or A. vaginae that were found in the 16S assessment of the endometrial microbiome, are associated with caries, but not healthy plaques (Kumar et al., 2003; Aas et al., 2005). Interestingly, the placental microbiome resembles that of the oral cavity more than that of gut or even vagina (Aagaard, et al., 2014). Several studies also show an association of cytokines and bacteria from amniotic fluid with oral microbiota (Dortbudak et al., 2005; Han et al., 2006). It remains to be seen whether bacteria of the oral cavity can indeed travel via the circulatory system to selectively colonize the uterus before or during pregnancy (Zaura et al., 2014). Jimenez et al. (2008) elegantly showed that in utero bacterial colonization of the fetus takes place, which can be influenced by maternal oral uptake of microbiota. Pregnant mice were fed with a genetically labeled Enterococcus faecium strain isolated from human breast milk, and the bacteria could be detected in pups delivered by C-section. The authors point out that next to the possible oral route, APCs of the gut might sense and actively harvest intestinal bacteria, facilitating their spread to other body sites. This sampling of luminal bacteria by DCs has been shown for the gut (Rescigno et al., 2001). A similar way of harvesting might already take place before pregnancy to prepare the uterus for placentation. If and how microbiota of gut and uterus might interact remains to be established. The study by Jimenez et al. (2008) fits with the ‘in utero colonization hypothesis’, which in itself is heavily debated. Recently, Perez-Muños et al. (2017) have pointed out the shortcomings of published data surrounding the theory that newborns acquire first microbial colonization in utero. Studies on in utero colonization and microbial colonization of the non-pregnant uterus share the same methodological and technical limitations that have not been addressed sufficiently to allow any conclusions to be drawn about a core uterine microbiome. We wish to point that even though possibly not determining colonization of the pup in utero, as highlighted by the authors, maternal fertility or uterine health might have still have been affected by the microbial status. Importantly, a role in establishing protection against pathogenic invasion can be envisioned, a threat that mice in protected breeding facilities are a lot less exposed to than humans. Therefore, the fact that germ-free mice are fertile does not rule out a possible impact of uterine microbiota on normal physiology of the human endometrium. Conclusion Microbiota might be another piece in the complex mechanism contributing to the cogwheels of hormones, immune cells and physiological adaptations that are needed for successful pregnancy. As described in this review, a vast number of possible physiological contributions of the uterine microbiota can be inferred from other body sites. We highlight that the assessment of uterine microbiota still suffers from many limitations. Nevertheless, the available evidence shows that the presence of uterine 16S rRNA is not solely the result of sampling or analysis errors and deserves to be acknowledged. We conclude that the concept of the sterile endometrium, and the uterine compartment in general, is outworn, although the true core uterine microbiome still needs to be assessed. Functional studies are needed to elucidate the physiological importance of the microbiome in fertility. The challenge of studying reproductive immunology and the microbiota involved is that research on all of the different aspects is still in its infancy; microbiome, immunity, endocrinology in pregnancy, and placental and fetal development need to be studied together to obtain a more comprehensive overview. Thus, experts in various fields, such as microbiology and immunology, need to co-operate. Without this, such a basal process as pregnancy, needed for healthy offspring, will continue to be a mystery. Supplementary data Supplementary data are available at Human Re