The emergence of the concept of the microbiome, together with the development of molecular-based techniques, particularly polymerase chain reaction (PCR) amplification using the 16S ribosomal RNA (rRNA) gene, has dramatically increased the detection of microorganisms, the number of known species, and the understanding of bacterial communities that are relevant to maternal–fetal medicine in health and disease. Culture-independent methods enable characterization of the microbiomes of the reproductive tract of pregnant and nonpregnant women, and have increased our understanding of the role of the uterine microbiome in adverse obstetric outcomes. While bacterial ascent from the vaginal tract is recognized as the primary cause of intrauterine infection, the microbiomes of the gastrointestinal, oral, and respiratory tracts are shown to be involved by means of hematogenous spread. The transmission of maternal microbiomes to the neonate, by vaginal delivery or cesarean section, is shown to affect health from birth to adulthood.
Microbiota and obstetric outcomes
Accumulating evidence links microorganisms to the etiology of various maternal–fetal conditions and to some of the “great obstetrical syndromes,” among them preterm delivery (PTD), premature preterm rupture of membranes (PPROM), premature labor, intrauterine growth restriction, gestational diabetes, late abortions, stillbirth, and hyperemesis gravidarum . PTD is the leading cause of neonatal morbidity and mortality worldwide . Every year, an estimated 15 million babies throughout the world are born before 37 weeks of gestation; about one million of them die from complications of PTD ( http://www.who.int/mediacentre/factsheets/fs363/en , updated Nov. 2013, accessed: March 17, 2014). Intrauterine infection is estimated to account for as much as 25–40% of PTD . However, cultivating bacteria has not revealed the precise etiologies of PTD or of other adverse obstetric conditions.
The emergence of the concept of the microbiome, as the genomes of microorganisms and their hosts of a particular environment, and the relationships between them , has transformed research on human health and disease. Of the microorganisms comprising human microbiomes, 90% are estimated as uncultivable . The development of DNA-sequencing technology has led to the establishment of metagenomics, which enables identification of the host genome, together with that of microorganisms. The 16S ribosomal RNA (rRNA) gene is generally used for polymerase chain reaction (PCR) amplification, due to its presence in all bacteria and its inclusion of both highly conserved and heterogenic sequences. A number of collaborations have emerged to implement the revolutionary technologies, among them MetaHIT financed by the European Commission and the Human Microbiome Project, an initiative of the United States National Institute of Health. The latter set out in 2007 to characterize the genomic sequences of microbial communities at five sites of the human body, one of them being the urogenital tract .
Over the past two decades, implementation of molecular-based (culture-independent) techniques for the characterization of bacteria has revealed the great diversity of the microbiome of the reproductive tract, enabled identification of many bacteria on the species and not only genus level, challenged the concept of “normal microbiota,” and changed the understanding of such pathologies as bacterial vaginosis (BV). We discuss the contribution of metagenomics to major issues of maternal–fetal medicine. Starting with the lower reproductive tract, we present the current state of knowledge regarding the vaginal microbiome in women of reproductive age and in pregnancy. Next, we discuss the intrauterine microbiota, which receives bacteria from both the vagina and hematogenous sources. We focus on the implications of bacterial transmission on obstetric outcomes and briefly discuss possible long-term implications on the developing life of the newborn.
The vaginal microbiome of healthy reproductive-age women
Normal vaginal flora has traditionally been defined as Lactobacillus predominant . However, culture-independent techniques have revealed the diverse and variable composition of the healthy vaginal microbiome. In the landmark study of asymptomatic reproductive-age North American women by Ravel et al., Lactobacillus species dominated in four of five groups of bacterial clusters . The most common species, L. iners , L. crispatus , L. gasseri , and L. jensenii , have been shown to predominate in the vaginal microbiota of other populations, though in differed proportions . Groups of bacterial clusters were found to be associated according to ethnic groups .
The observation that strictly anaerobic organisms, rather than Lactobacillus , dominated in one of Ravel et al.’s five groups of bacterial clusters indicates that the dominance of Lactobacillus is not a necessary characteristic of healthy vaginal flora. Nevertheless, even in that group, Lactobacillus species were detected in 104 of the 108 specimens. Interestingly, all community groups identified by Ravel et al., including the one not dominated by Lactobacillus , contained lactic-acid-producing bacteria. In addition to Lactobacillus , Megasphaera , Streptococcus , and Atopobium also presented. The detection of bacterial genera that have been associated with abnormal vaginal microbiota, such as Prevotella , Atopobium , Gardnerella , Megasphaera , and Mobiluncus , in asymptomatic reproductive-age women challenge the traditional distinction between healthy and abnormal flora. In addition to differences observed between ethnic groups and interindividual differences, considerable intraindividual differences have been documented. A study of 32 healthy reproductive-age women over a 16-week period found bacterial community composition, time in the menstrual cycle, and sexual activity to be associated with deviations from the stability of bacterial vaginal communities . Despite the variability observed, the vaginal microbiome is shown to have a higher stability than other bodily habitats, namely the oral region, skin, and distal gut (stool) .
The vaginal microbiome of healthy reproductive-age women
Normal vaginal flora has traditionally been defined as Lactobacillus predominant . However, culture-independent techniques have revealed the diverse and variable composition of the healthy vaginal microbiome. In the landmark study of asymptomatic reproductive-age North American women by Ravel et al., Lactobacillus species dominated in four of five groups of bacterial clusters . The most common species, L. iners , L. crispatus , L. gasseri , and L. jensenii , have been shown to predominate in the vaginal microbiota of other populations, though in differed proportions . Groups of bacterial clusters were found to be associated according to ethnic groups .
The observation that strictly anaerobic organisms, rather than Lactobacillus , dominated in one of Ravel et al.’s five groups of bacterial clusters indicates that the dominance of Lactobacillus is not a necessary characteristic of healthy vaginal flora. Nevertheless, even in that group, Lactobacillus species were detected in 104 of the 108 specimens. Interestingly, all community groups identified by Ravel et al., including the one not dominated by Lactobacillus , contained lactic-acid-producing bacteria. In addition to Lactobacillus , Megasphaera , Streptococcus , and Atopobium also presented. The detection of bacterial genera that have been associated with abnormal vaginal microbiota, such as Prevotella , Atopobium , Gardnerella , Megasphaera , and Mobiluncus , in asymptomatic reproductive-age women challenge the traditional distinction between healthy and abnormal flora. In addition to differences observed between ethnic groups and interindividual differences, considerable intraindividual differences have been documented. A study of 32 healthy reproductive-age women over a 16-week period found bacterial community composition, time in the menstrual cycle, and sexual activity to be associated with deviations from the stability of bacterial vaginal communities . Despite the variability observed, the vaginal microbiome is shown to have a higher stability than other bodily habitats, namely the oral region, skin, and distal gut (stool) .
Microbiomes of the reproductive tract of pregnant women
The vaginal microbiome
In two studies using 16S rRNA sequencing, less bacterial diversity was observed in the vaginal microbiota of healthy pregnant women than in that of nonpregnant women . In the pregnant women, Lactobacillus species were more abundant and other phylotypes less so, and the bacterial community group that is not dominated by Lactobacillus was less common . Despite the relative stability of the vaginal microbiota of pregnant women, communities tended to shift from dominance by one Lactobacillus species to another . The authors suggested that the enhanced stability may increase resilience and thus protect against ascending infection and its sequela, though the mechanism of such protection remains to be elucidated .
The uterine microbiome
The uterus was traditionally considered sterile. The documentation in 1927 of positive bacterial cultures in the amniotic fluid of women undergoing cesarean sections who had been in labor for >6 h, compared with negative cultures in those who were in labor for <6 h (Harris 1927), supported the proposition of a sterile uterine cavity. The isolation of microbes from amniotic fluid was long considered an indication of infection or disease. In one study that refuted the paradigm of the sterile uterus, only six cultures isolated from 55 endometria of asymptomatic women without a history of previous pelvic infection were sterile; 231 bacterial species were detected from the positive cultures (Hemsell 1989). Investigations of the uterine microbiome of healthy, term pregnancies remain scanty, even with the availability of culture-independent techniques. Han et al. identified nine bacterial species that presented in almost half of amniotic fluid samples that tested positive with PCR, but were not cultivated. Compared to culture-dependent methods, culture-independent methods are estimated to have increased bacterial detection in the amniotic cavity by about 30–50% and increased the number of detected species by up to fivefold . Among them, the cultivation-resistant anaerobic bacteria belonging to the family Fusobacteriaceae is now recognized as a common inhabitant of amniotic fluid. The detection of bacteria in the intrauterine cavity by PCR, in the absence of signs of infection, confirms the proposition of a non-sterile uterus. Further, evidence accumulated from the animal kingdom indicates the universality of microbial transmission from mothers to newborns, of commensal bacteria, and not only pathogens . More investigation is needed to characterize the members of the healthy uterine microbiome, interactions among these members, and diversity among populations.
Microbiomes that affect obstetric outcomes
The sites of maternal infection that may come in direct contact with the fetus are amniotic fluid, the placenta, fetal membranes, and the umbilical cord. In addition, hematogenous transmission from the mother to the fetus may involve microbiomes that are not in direct contact with the fetus, such as those of the gastrointestinal, respiratory, and oronasal tracts.
The uterine microbiome and adverse obstetric outcomes
Bacterial invasion of the amniotic cavity has been indicted as the main cause of neonatal mortality worldwide . Cultivated bacterial pathogens from the amniotic cavity have been associated with miscarriage , chorioamnionitis , PROM , and PTD . Culture-independent techniques have contributed greatly to our understanding of the associations between infection, inflammation, and adverse obstetric outcomes. Positive bacterial cultures from the amniotic fluid of women who delivered preterm were found to be inversely related to gestational age, with different species more commonly isolated at earlier or later gestation . However, culture-independent technologies have detected bacterial sequences in the fetal membranes of preterm deliveries and term deliveries, both with and without labor , as well as in up to 70% of women undergoing elective cesarean sections at term , thus indicating that the presence of bacteria does not in itself cause preterm labor.
The detection of elevated levels of proinflammatory cytokines in the amniotic fluid of women who delivered before 34 weeks of gestation supported the etiology of intra-amniotic infection in PTD . However, proinflammatory cytokines were also detected in culture-negative amniotic fluid . Though such a finding may result from treatment with antibiotics, culture-independent techniques have revealed bacterial presence and species that were not identified by cultivation, and that may induce inflammatory responses. Gardella et al. detected 16S ribosomal DNA (rDNA) sequences in 36% of culture-negative amniotic fluid samples of women in preterm labor who had elevated levels of the proinflammatory cytokine interleukin-6 (IL-6). The culture-negative amniotic fluid specimens that contained 16s rDNA of Leptotrichia or Fusobacterium had particularly high IL-6 levels. Han et al. reported that about two-thirds of the bacterial species detected in amniotic fluid by culture-independent methods were not isolated by culture. Further, PCR-positive bacterial identification has itself been associated with elevated IL-6 levels in amniotic fluid , as well as histological chorioamnionitis and funisitis , and early-onset neonatal sepsis (EONS) . The observations of a shorter amniocentesis-to-delivery interval among women who were PCR positive and a dose–response association between bacterial rDNA abundance and gestational age at delivery support the possibility of a causal relationship . In a recently published systematic review of women delivering before term, culture-independent techniques revealed intrauterine bacterial infections in 349/761 (46%) .
The capability of detecting greater numbers and species of bacteria in amniotic fluid raises the expectations of deciphering associations between microorganisms in amniotic fluid and adverse perinatal outcomes. Microbial invasion of the amniotic cavity has been assessed as polymicrobial in 24–67% of the cases . The clinical significance of a positive PCR of amniotic fluid appears to be equivalent to that of a positive culture . Still, the role of bacteria in PTD is not yet understood. A recent investigation of bacteria from fetal membranes of women who delivered preterm or term, with or without labor, reported an association of increased bacterial presence with a thinner chorion, regardless of infection, gestational age, or labor status. This suggests that bacterial presence may lead to chorion thinning and ultimately to PPROM .
The vaginal microbiome and adverse obstetric outcomes
Bacterial ascent from the vaginal tract to the uterus, possibly crossing the placental barrier, is recognized by both culture-dependent and culture-independent studies as the primary source of intrauterine infection. The finding that bacterial species common to the genital tract were among the most common bacteria detected in the uterine cavity supports this pathogenesis .
Bacterial vaginosis
BV, the most common vaginal disease, has been found to be associated with intra-amniotic infection and to be a strong risk factor for PTD and spontaneous abortion . The prevalence rates of BV vary greatly among ethnic groups and geographical regions, reaching as high as 60% . The Nugent scoring system, which is based on morphologic bacterial identification by Gram staining , is considered the gold standard for diagnosing BV . High Nugent scores (indicating abnormal flora) have been associated with PTD , delivery of low-birth-weight infants , and early and late miscarriage .
The relationship between BV and PTD is not yet understood. Spontaneous recovery of BV during pregnancy was reported not to be associated with a reduced risk of PTD . While antibiotic treatment has demonstrated effectiveness in reducing Nugent scores, especially high scores , its effect on reducing rates of PTD has been shown to be either negative or limited to certain circumstances, according to medical history or characteristics of the BV infection . As a high Nugent score was shown to be associated with a better response to antibiotic treatment, and as a high Nugent score is allocated in the presence of Mobiluncus , antibiotics were proposed to be particularly effective against Mobiluncus . Recently, using culture-independent techniques, Mobiluncus was in fact found to be associated with an increased risk of PTD . By contrast, BV-associated bacteria (BVAB-3), which has a high specificity (>96%) for BV , was shown to decrease the risk of PTD . This unexpected finding could be due to interactions observed between BVAB-3 and cohabiting bacteria, as well as with race/ethnicity . Future culture-independent studies may be expected to reveal particular bacterial species and communities that may affect the risk of PTD and the specific impact of antibiotics on them.
Based on Nugent criteria, the absence of Lactobacillus bacteria (score of 4) is a sufficient criterion for the classification of abnormal vaginal flora, and thus may lead to overdiagnosis of BV . On the other hand, elevated pH and high Nugent scores have been detected in some bacterial communities that have high proportions of Lactobacillus species . Moreover, culture-independent techniques reveal considerable diversity in Lactobacillus species. For example, L. vaginalis was found to induce an immune response that is similar to that elicited by BVAB . Interestingly, L. iners , whose detection only became possible with the development of 16S rRNA gene sequencing , persists in the BV environment , as well as comprising one of the most common bacterial species of the healthy microflora . This Lactobacillus species was shown to express different genes in the healthy and BV environments .
BV appears to be a syndrome, with varying characteristics, symptoms, outcomes, and responses to antibiotic treatment . Culture-independent techniques have revealed greater bacterial diversity in the vaginal microbiota of women with BV than in those without , as well as several bacterial species that were not detected through cultivation . On the other hand, some bacterial species, such as G. vaginalis and A. vaginae , which have been considered as indications of a disease state , are not uncommon inhabitants of healthy vaginas . The diversity of the vaginal microbiome has been suggested as a factor that increases the risk of PTD . The ambiguity in the characterization of BV may explain the notion of asymptomatic BV among women considered healthy , as well as the high recurrence to BV after antibiotic treatment – reversion to a state that is not dominated by Lactobacillus , and which is not necessarily pathological according to the current and broader understanding of normal vaginal flora. The recent documentation of a period of conversion that precedes symptomatic acute BV elucidates changes occurring in the transition from the healthy to the BV state .
Other types of vaginitis
Vulvovaginal candidosis (VVC) is the most common cause of vaginal disease after BV. The number of cases of diagnosis is estimated to be as high as 40% of women with vaginal complaints in the primary-care setting . Using 16S rRNA sequencing, no distinct bacterial profile was discerned for VVC in Chinese women . The diversity of the vaginal microbe was greater for those with VVS than for normal control specimens, though less than for those diagnosed with BV. Specimens of mixed BV and VVC infection showed a unique vaginal microbiota community structure, different from that of BV alone, with all BV/VVC microbiomes showing abundant Lactobacillus . The upshot on obstetric outcomes of the presence of candidiasis in conjunction with BV warrants investigation, as does the impact of increased bacterial diversity.
Though its investigation is scanty compared to that of BV, aerobic vaginitis (AV) has been found to be associated with late miscarriage, chorioamnionitis, and PTD . The application of 16S rRNA sequencing revealed that bacteria associated with aerobic vaginitis, specifically members of the genera Streptococcus and Staphylococcus , as well as Escherichia coli , present frequently in healthy vaginal microbiota .
The hematogenous route to intra-amniotic infection
Though bacteria ascending from the vagina appear to be the primary source of intra-amniotic infections, the hematogenous route, from the placenta, is also recognized.
The gastrointestinal microbiome and adverse obstetric outcomes
The evidence that some microorganisms colonize both the gastrointestinal and the reproductive tracts demonstrates the interaction between these microbiomes. As an example, the intestinal tract is recognized as the primary reservoir of group B streptococci (GBS) and thus the predominant source of vaginal colonization in pregnant women . However, women with hydrogen peroxide-producing lactobacilli in the vagina were found to be more likely to have GBS only in the rectum, and not in the vagina . Further, the prevalence of BV was found to be lower (5%), among women with hydrogen peroxide-producing lactobacilli in both the vagina and rectum than in those with such bacteria in only one or neither of the regions . Maternal–fetal infection by Listeria monocytogenes is another example of migration of bacteria from the gastrointestinal to the reproductive tract. Of 37 culture-confirmed cases of L. monocytogenes recorded in Denmark during the period 1994–2005, 12 (32%) ended in abortion or stillbirth . Using culture-independent techniques, 44.4% of bacterial species that were identified from either the vaginal or rectal microflora of pregnant women were detected in both sites . Ingestion of lactobacillus probiotics has been shown to induce changes in the vaginal bacterial community, as well as an inflammatory response . This provides evidence of interaction between the vaginal and gastrointestinal microbiomes, and supports the proposition that oral probiotics may reduce candidal infection and BV .
The oral microbiome and adverse obstetric outcomes
A review of the literature and a meta-analysis of 22 studies including 12047 pregnant women showed that women with periodontitis had an increased risk of PTD and of delivering a low-birth-weight infant . PCR using 16S and 23S rRNA supports the possibility of microbial transmission from the oral to the uterine cavity. Porphyromonas gingivalis was detected in both the amniotic fluid and the subgingival plaque of eight women in premature labor . Further, the same uncultivated strain of Bergeyella sp. was isolated from both the amniotic fluid and the subgingival plaque of a pregnant woman with an intrauterine infection, and not detected in her vaginal tract . Investigations are needed to determine the extent of transmission of bacterial species from the oral to the intrauterine cavity and the involvement of such bacteria in intrauterine infection.
The respiratory microbiome and adverse obstetric outcomes
The respiratory tract is recognized as a potential source of hematogenous infection , though the evidence for a causal relationship is currently scanty. In support, bacterial species common to the respiratory tract, such as Haemophilus influenzae , H. parainfluenzae , and S. pneumoniae , have been identified, though in low frequency in the intra-amniotic cavity of women delivering preterm . However, as these and other bacteria also present in the vaginal microbiota , more studies are needed to investigate the role of respiratory tract infection in intra-amniotic infection and in ensuing adverse pregnancy outcomes. A recent report showed a 17-fold increased risk of invasive unencapsulated H. influenzae disease among pregnant women in England and Wales . Contracting this infection during the first 24 weeks of pregnancy was associated with fetal loss and extremely premature birth; contact during the second half of pregnancy was associated with premature birth and stillbirth.
The placental microbiome and adverse obstetric outcomes
The high rate of recurrence of PTD suggests a sustained risk factor. This prompted investigation of a placental microbiome and, specifically, the investigation of the microflora in the basal plate of the placenta, which is the tissue layer directly at the maternal–fetal interface . Bacteria of diverse morphologies were detected in the basal plates of 27% of placentas examined: in 54% of women who delivered spontaneously preterm, before 28 weeks; and in 26% of those who delivered at term.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
