A possible link between the vaginal microbiome, bacterial vaginosis and preterm birth

Bacterial vaginosis (BV), previously called nonspecific vaginitis or Gardnerella-associated vaginitis, is a condition that is estimated to affect 10 to15% of women of reproductive age. The condition is caused by an imbalance of the naturally occurring vaginal bacterial flora and is associated with genital tract infections, resulting in an increased risk of ectopic pregnancy, pelvic inflammatory disease, chronic pelvic pain, and tubal factor infertility. BV is diagnosed based on positive results of three of four clinical criteria commonly known as “Amsel’s criteria”: presence of a homogeneous discharge; vaginal pH greater than 4.5; potassium hydroxide test for volatile amines, and the microscopic presence of clue cells, which are vaginal epithelial cells having a stippled appearance due to being covered with bacteria. This condition is thought to stem from a shift in the vaginal microbial community from mainly Lactobacillus species to a more diverse polymicrobial community although the same microbial species can be found in the vaginal tract of asymptomatic women as well. This paradox illustrates a clear gap between the clinical assessment of this condition and our understanding of the pathophysiology of the female genital tract.

BV clue cells

CLUE cells are key to diagnosis and are squamous epithelial cells appearing stippled or granualetd due to bacteria adherence to the epithelial cell surface.

This gap in our knowledge is widened when we consider the impact of BV on pregnancy. For pregnant women, BV increases the risk of late miscarriage, preterm labour, and preterm premature rupture of membranes, resulting in preterm delivery. In addition, the risk of developing chorioamnionitis and endometritis is increased with BV. In the United States alone, preterm delivery is associated with an annual cost to society of $26.2 billion and is the single most significant factor contributing to neonatal morbidity and mortality. The World Health Organization reported that about 13 million premature births occur globally each year directly causing nearly 28% of all neonatal deaths (1 million each year) and overall the world’s preterm birth rate has continued to rise. For example, the rate of singleton preterm births has increased by 14% since 1990, and now accounts for 11.1% of deliveries worldwide. In 2006, preterm births constituted 12.8% of live births in the United States, an increase of 20% since 1990.

A recent study demonstrated a role for BV in preterm birth by showing that BV more than doubles the risk of preterm delivery in asymptomatic patients and of preterm labour in patients with symptoms, as well as significantly increasing the risk of late miscarriages and maternal infection. Microbial infection may contribute to approximately 25% of preterm births, with bacterial colonization rates as high as 79% for birth at 23 weeks of gestation, declining to 11% at 31 to 34 weeks. Moreover, there are specific associations between various vaginal microbes linked to BV (Mycoplasma hominis, Ureaplasma urealyticum, Gardnerella vaginalis and Atopobium vaginae, alone or together) and preterm birth. Abnormal vaginal flora has also been associated with shortening of the cervix in pregnancy, a known risk factor for preterm delivery. It is puzzling that although treatment for BV with metronidazole was shown to clear symptoms and signs of BV, metronidazole treatment was unable to decrease rates of preterm delivery. The question remains, then, “What is the relationship between the vaginal tract microbiome and preterm birth?”

Culture-based identification of vaginal microbes has provided an incomplete and biased understanding of the vaginal microbiome. This approach resulted in a common belief that the vaginal microbiome included only a few microbial species, primarily the “beneficial” bacteria from the genus Lactobacillus. It was thought that an overgrowth of bacterial species such as G. vaginalis in addition to a reduction in the lactobacilli population resulted in BV. The advances in modern molecular culture-independent techniques and next-generation sequencing approaches allow us to characterize microbial communities in detail which was previously not possible. These techniques have been applied to survey the human vaginal microbial community and amniotic fluid during preterm labour. The diversity of species present has exceeded that of classic culture-based methods by several orders of magnitude. Not only is there incredible diversity of microorganisms present in individual women, but also the vaginal populations in different healthy women are highly varied. A recently study by co-principle investigator (PI) Wilson supports this observation.

While we still know very little about how the vaginal microbiome is impacted by age, menstrual cycle, genetic background or other factors in the daily lives of women, demonstration projects and consortium sequencing through the NIH funded Human Microbiome Project (HMP; http://nihroadmap.nih.gov/hmp/fundedresearch.asp) will provide important microbiome and metagenome baseline data. In addition, the PIs (White and Wilson) of the NIH award at the UIUC are supplying vaginal isolates to the JCVI for reference genome sequencing under the HMP U54 mechanism, and this has resulted in the release and publication of the complete genome of G. vaginalis 409-05. Reference genome sequencing is in progress at the JCVI for at least nine other vaginal isolates from genera other than Gardnerella. As such additional biological, technological, and reference genome resources are now available, enabling the application of translational biomedical approaches based on genomics, metagenomics and computational biology to improve women’s reproductive health.

All previous culture-independent studies of the vaginal microbiome focused only on the prokaryotic community by sequencing the 16S rRNA gene (rDNA), neglecting the other potential key players: bacteriophage, viruses and fungi. In addition, none of these investigations looked at the metabolic potential (functional gene content) of the vaginal microbiome, which we will investigate in our study, as has been done in other environmental systems. We will use high throughput genomic technologies (454 and Illumina) to survey the fungal, bacterial, phage and viral microbiomes as they relate to BV and preterm birth. Microbial and gene predictions (annotation and construction of genomes based on microbial reference genomes from the HMP) will be used to test the hypothesis that a defined set of microbial taxa or gene(s) can be predictors (quantitative trait loci or QTLs) of increased risk and incidence of preterm birth. These microbial genetic markers will lead to new and rapid diagnostics that will ultimately provide medical professionals with predictors of risk for these clinical outcomes that have profound impacts on women’s health.

Specifically, we will use deep, tag-based 454 sequencing of the V1-V3 hypervariable regions of the bacterial 16S rDNA and the internal transcribed spacer (ITS) region of fungal rDNA, and metagenomic sequencing of viral/bacteriophage genomes to assess the structure of the vaginal microbiome and associations of specific taxanomic groups with BV and preterm birth. In addition, we will use a combination of 454 and Illumina metagenomic sequencing, whole genome assemblies from metagenome sequencing and comparative analyses to determine the metabolic potential of non-BV versus BV-associated vaginal microbiomes and of vaginal microbiomes associated with preterm birth. This will allow us to define the identifiable microbiome taxa or gene(s) that are linked to an increased risk and incidence of BV and preterm birth.

Our overall approach will provide a tremendous amount of new information about the factors that have the greatest impacts on vaginal microbiome dynamics and provide critical data on risk factors and how changes in the vaginal ecosystem accompany or predict disease development and progression. Ultimately, we envision these studies leading to personalized (genomic) medical diagnostics that would be capable of predicting the risk for these clinical outcomes, which impact women’s health.

Derrick E. Fouts, Bryan A. White, and Karen E. Nelson (The J. Craig Venter Institute)

Categories: Feature Articles

Tags: , , , ,

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: