Mothers and microbes, Part 1: The vaginal microbiome in health and disease

December 20th, 2015 by Amy Proal

“Like mother, like daughter.” The phrase is often invoked to describe how children resemble their parents. While we know that human genes are passed from generation to generation, an expanding body of research now shows that many microbiome populations are also inherited. The microbes a child inherits are acquired from both parents and even siblings. However, microbial populations inherited from the mother have a particularly strong impact on a child’s development and health.

The impact of inherited microbes cannot be underestimated. As described here, humans are superorganisms, whose human cells are vastly outnumbered by those of their microbial inhabitants. So while a child develops thanks to its 20,500 or so human genes, it is tremendously impacted by the billions of genes expressed by its microbiome. Research now indicates that an infant’s microbial populations are inherited from those present in its mother’s placenta, amniotic fluid, vagina, breast tissue, breast milk, and other sources. These populations are often derived from microbes in the mother’s gut, oral cavity, and other body sites. As with most microbiome populations, maternal microbial ecosystems have been identified in areas of the body previously considered to be sterile.

I first became interested in this “maternal microbiome” after reading a 2008 paper by Dave Relman and team at Stanford. The study used molecular tools to detect microbes in the amniotic fluid of women who gave birth prematurely. Many of the bacteria identified by the team had not been previously cultivated or characterized. There was a strong, direct correlation between the number of microbes in the amniotic fluid and an increased risk of premature birth.

2010-Proal-Slovenia.001At the time, I started most of my lectures with the slide to the right, which depicts data from the Relman paper. I wanted my audience to note that even the most traditionally “privileged” body sites are not sterile. I also hoped viewers would entertain the possibility that microbes can be passed from mother to infant. In 2008, however, the traditional view that babies are born sterile was very prevalent. Certain audience members would inevitably struggle with Relman’s data, and suggest that his samples might have been contaminated.

These concerns about contamination have since been proven unfounded. Over the last few years, additional research teams have detected microbes in the amniotic fluid. However, as Relman writes in a letter in the medical journal PNAS, fetal testing has become less invasive over the last few years. This has made it much harder for researchers to obtain samples of amniotic fluid that can be used in research studies.

Nevertheless, around 2010, several metagenomic analyses clarified the presence of a robust vaginal microbiome, members of which coat an infant as it passes through the birth canal. The new data dispelled the notion that babies are born sterile and, in my opinion, made the research community much more open to the study of other possible maternal microbiome populations.

The vaginal microbiome

The vaginal microbiome has now been characterized by multiple teams, with some studies tracking how vaginal microbial populations fluctuate over time. Vaginal microbial populations vary among individuals due to factors such as ethnicity, diet, stress, sexual partners, and obesity. However, some general trends have emerged. In healthy women, several kinds of vaginal microbial communities exist. Most of these are dominated by species of Lactobacillus, while others contain a diverse array of anaerobic microorganisms. Some of the Lactobacillus-dominated communities, such as those with high levels of L. crispatus, are more stable than others. Yet composition of even the most stable vaginal microbiome communities are punctuated by occasional daily fluctuations. For example, menses can also disrupt vaginal microbiome composition, albeit transiently.

The healthy vaginal microbiome produces lactic acid and hydrogen peroxide.

The healthy vaginal microbiome produces lactic acid and hydrogen peroxide. This creates an acidic environment that is less hospitable to pathogens. According to Spanish researcher Juan Álvarez-Calatayud, “the healthy vagina emits organic compounds that keep the pH around 4-4.5, which seems optimal for the Lactobacillus dominated vagina.” In healthy, premenopasal women, vaginal mucous is also rich in nutrients such as glucose and amino acids that facilitate colonization of the vaginal cavity. Levels of this mucous decrease substantially after menopause, causing vaginal populations to decrease significantly. These postmenopausal vaginal populations have been shown to contain fewer Lactobacillus than those of reproductive-aged women, with the exception of women on hormone-replacement therapies.

Composition of the vaginal microbiome also changes during pregnancy. Collectively, the vaginal communities of pregnant women are very stable, and dominated by Lactobacillus. However, they tend to be characterized by lower richness and species diversity than the vaginal microbiomes of non-pregnant women. Relman and team recently performed yet another excellent analysis, this time of the vaginal microbiome during pregnancy. After identifying vaginal microbiome populations using molecular tools, they were able to assign each female study participant into one of five community state types, or CSTs. Four of these CSTs were highly uneven communities dominated by different species of Lactobacillus. The fifth CST was characterized by much greater evenness and taxonomic diversity. While some subjects maintained a single stable CST throughout gestation, the vaginal microbiome composition of other subjects frequently transitioned among different CSTs.

F4.largeInterestingly, women who harbored the Lactobacillus-poor CST were much more likely to give birth prematurely than those with Lactobacillus-dominated CSTs. Preterm birth, or delivery before 37 weeks of gestation, occurs in 11% of pregnancies and is the leading cause of neonatal death. The figures to the right show that, in the Relman study, the Lactobacillus-poor CST vaginal community exhibited both dose–response and temporal relationships with preterm birth.

Preterm birth also is associated with bacterial vaginosis, a community-wide disruption of the vaginal microbiome that approximately doubles the risk of preterm birth. Further analysis of the Lactobacillus-poor CST suggested that a high abundance of the bacterium Gardnerella or the mycoplasma Ureaplasma, combined with low abundance of Lactobacillus, was particularly predictive of preterm birth. Although the Relman study was based on a relatively small number of subjects, a large body of literature implicates both Gardnerella and Ureaplasma as having potential roles in the pathogenesis of preterm delivery. For example, researchers at the University of Alabama found that in 23% of neonates born between 23 to 32 weeks, umbilical blood cultures tested positive for Ureaplasma urealyticum and Mycoplasma hominis.

Vaginosis, which affects over 30% of women, has also been associated with preterm birth in a number of other studies. For example, researchers at Temple University in Philadelphia found that vaginosis in the first trimester more than doubled the risk of spontaneous pregnancy loss in the second. A Belgian research team reported that women with balanced vaginal flora in the first trimester had a 75% lower risk of delivery before 35 weeks compared to women with abnormal vaginal flora (AVF). Furthermore, absence of lactobacilli in the AVF group was associated with increased risks of preterm birth and miscarriage.

Returning to the Relman study, the team also found that, in most cases, maternal vaginal microbiome populations shifted significantly after subjects gave birth. These altered vaginal communities persisted for up to a year, and were characterized by a sharp increase in community diversity. The most prevalent Lactobacillus species, Lactobacilli, were largely replaced by a diverse mixture of anaerobic bacteria including Peptoniphilus, Prevotella, and Anaerococcus. Relman and team point out that a short interval between pregnancies (twelve months or less) is associated with an increased risk of preterm birth. This caused the team to speculate that an “altered vaginal community might affect the outcome of a subsequent pregnancy if conception occurs too soon after delivery.”

Maternal vaginal microbiome populations shifted significantly after subjects gave birth.

Why might vaginosis be associated with preterm birth? In addition to helping the body attack germs, the immune system plays a large role in the birthing process. During healthy birth, a generalized inflammatory response aids in moving an infant out of the womb. Some researchers believe that, in women who suffer from vaginosis, a similar inflammatory response towards pathogens in the vaginal community may incorrectly stimulate labor before an infant has fully matured. Indeed, studies show that any stimulus that prompts inflammation during pregnancy – be it from acute infection, microbiome imbalance, or even gum disease – can trigger labor prematurely.

Vaginosis is also associated with preterm birth because, on the whole, mothers with dysregulated vaginal microbiomes tend to be more generally ill. While vaginosis may develop on its own, it is often diagnosed in women who suffer from other system-wide inflammatory diseases. For example, women who suffer from an autoimmune disease are more likely to suffer from vaginosis. In such cases, the inflammation and dysfunction associated with these additional diagnoses may also induce delivery before an infant has fully matured in the womb.

In the United States, the standard of care for treating bacterial vaginosis is antibiotics. These include metronidazole and clindamycin. Unfortunately, these antibiotics cause side effects and relapse is common. However, some studies have found that specific probiotics may help stabilize vaginal microbiome populations in women with vaginosis. In some cases, probiotics can be administered intravaginally or directly into the vagina. This eliminates the possibility that they might be destroyed by stomach acid or other components of the digestive process. For example, Ling and team studied the vaginal microbiome populations of women administered either the antibiotic metronidazole or intravaginal probiotics. In most subjects, treatment led to the recovery of a Lactobacillus-dominated vaginal microbiome, and levels of pathogenic vaginal bacteria decreased. However, a greater vaginosis cure rate was observed in the probiotic-treated subjects. The probiotics also re-established vaginal homeostasis in a gradual and steady fashion.

In certain cases, probiotics administered orally have also been shown to benefit women with vaginosis, although success rates vary from study to study. Usually these oral probiotics contain one or more strains of Lactobacillus, and are taken in conjunction with an antibiotic. One study of 196 women found that treatment with the antibiotic tinidazole plus Lactobacillus reuteri RC-14+L. rhamnosus GR-1 significantly increased the relative abundance of indigenous vaginal Lactobacilli in women with bacterial vaginosis.

In a 2015 paper on the maternal microbiome and probiotics, Álvarez-Calatayud and team agree that certain species of Lactobacillus, isolated from human milk and administered orally, may help manage vaginosis. They recommend such probiotics be used in lieu of antibiotics, since many antibiotics can cause yeast overgrowth and digestive problems. However, they emphasize that only very specific strains of Lactobacillus have been shown to help women with vaginosis, and many over-the-counter probiotics do not contain such species. Also, many of the probiotic strains used to treat vaginosis are not currently available for purchase in the United States.

Mode of delivery impacts the infant microbiome

Attempts to stabilize the vaginal communities of women with vaginosis are important, since naturally born infants are coated in their mother’s vaginal microbes during and after passage through the birth canal. These vaginal microbes serve as a direct source of protective or pathogenic bacteria very early in life and greatly shape an infant’s initial microbial communities. In 2010, Rob Knight and team at the University of Colorado Boulder studied the impact of the vaginal microbiome on infant microbial development. They compared the early microbial populations of vaginally delivered infants to those present in infants delivered by Cesarian section (C-section).

The team sequenced the microbial communities of mothers at various body sites. As expected, each maternal bacterial community was structured primarily by body habitat, with distinct populations detected in the mouth, skin, and vagina. In contrast to their mothers however, newborns harbored bacterial communities that were essentially undifferentiated across skin, oral, nasal, and gut habitats regardless of delivery mode. In other words, the human microbiome is distributed uniformly across the body during the earliest stages of life.

Relative abundance of the 20 most abundant taxa in mothers' vaginal communities and in the babies they delivered vaginally. Knight et al.

Relative abundance of the 20 most abundant taxa in mothers’ vaginal communities and in the babies they delivered vaginally. Knight et al.

However, microbes forming these early communities vary widely depending on mode of delivery. As expected, C-section babies lacked bacteria from the vaginal community. On the other hand, infants delivered via C-section harbored bacterial communities (across all body habitats) that were most similar to the skin communities of their mother. As with adult skin communities, these infants harbored an abundance of the bacterium Staphylococcus. Knight and team hypothesized that this finding might explain why infants delivered by C-section are often more susceptible to certain pathogens. For example, 64 to 82% of reported methicillin-resistant Staphylococcus aureus (MRSA) skin infections in newborns have occurred in C-section delivered infants.

In addition, each mother’s vaginal bacterial community was significantly more similar to her own infant’s microbiome than to that of other vaginally delivered babies. In contrast, skin bacterial communities of C-section mothers were no more similar to their own babies than to the other infants born via C-section. This led Knight and team to speculate that exposure to skin bacteria in the hospital environment could contribute to the microbiomes of C-section delivered babies.

Bacterial diversity in operating rooms. Source: Dominguez-Bello et al.

Bacterial diversity in operating rooms. Source: Dominguez-Bello et al.

Indeed, this past year Dominguez-Bello and team at New York University searched for bacteria present in four different operating rooms. Humans shed up to 37 million bacteria into the environment per hour. In addition, operating rooms tend to lack natural ventilation and, regardless of cleaning efforts, harbor skin microbes from occupants (surgeons, nurses, cleaning personnel). The team characterized bacteria present in the dust, floors, walls, ventilation grids, armrests, and lamps of each operating room immediately after C-section births. Bacteria identified at these locations corresponded greatly to common human skin microbes, including Staphylococcus and Corynebacterium. All rooms sampled had airborne skin bacteria that accumulate on surfaces. Bacteria in human skin flakes were found in ventilation grids, and live skin bacteria were prevalent on top of operating room lamps. Consequently, a major source of the C-section infant microbiome does appear to be derived from the skin microbes of operating room personnel.

It’s important to note that the skin-derived microbiome associated with C-section birth has not yet been significantly tied to disease or diminished microbiome health. More studies are needed to better understand how early microbiome composition impacts development. However, a number of studies show that infants born via C-section have a higher likelihood of developing conditions such as asthma, obesity and type 1 diabetes. While mode of delivery may be one variable contributing to these higher rates of disease, mothers whose infants are born via C-section are often more ill than those that give birth vaginally. They also tend to suffer from complicated pregnancies associated with higher rates of infection and system-wide dysfunction. Clearly, these other variables impacting the health of the mother must also be taken into consideration when assessing negative outcomes associated with C-section birth.

Nevertheless, because vaginal birth is more natural than C-section delivery, some people have gone so far as to coat infants delivered by C-section with vaginal microbes after birth. As mentioned in a 2014 TED talk, Rob Knight and his partner did this after their own daughter was born via emergency C-section. Knight does not mention the specifics of how vaginal microbes were acquired or administered to his infant. He is also unsure if coating his daughter in this fashion has benefited her health, but mentions that she has not suffered any ear infections to date. Most importantly, he states that larger clinical trials are underway to study any protective effects of this “coating” in other babies born via C-section.

While these further clinical trials are certainly warranted, it seems prudent that the vaginal microbiome of any “donor” involved in coating infants with vaginal microbes be carefully sequenced. Unless a mother’s or donor’s vaginal microbiome is in a state of balance, the infant might acquire undesirable microbes in the process. This is especially true since many mothers who deliver via C-section may also suffer from vaginosis. Also, as discussed above, the vaginal microbiome appears to become significantly altered after birth. So the microbes an infant acquires while passing through the vaginal canal might be different from those the same mother would harbor even shortly after delivery.

Stay tuned next week for more on the placental microbiome, breast milk microbiome, and more…

2 thoughts on “Mothers and microbes, Part 1: The vaginal microbiome in health and disease

Leave a Reply

Your email address will not be published. Required fields are marked *