The differential oral microbiome in preterm birth
MetadataShow full item record
Prior studies have shown that the low-level microbiome in the placenta is most similar to the non-pregnant oral microbiome, suggesting a hematogenous route of bacterial transmission. Based on these studies, we theorized that a disruption of the normal balance of pathogenic and commensal microorganisms in the oral cavity will lead to conditions that favor colonization of the placenta and amniotic cavity, leading to inflammation and preterm birth. We hypothesized that an altered oral microbiome profile will promote preterm birth. Our study aimed to compare metagenomic profiles of saliva and tongue in women delivering preterm to those of women delivering at term. Unstimulated saliva and tongue brushings were collected according to an IRB-approved protocol from patients who delivered preterm, and from age and race-matched patients who delivered at term. Exclusion criteria included obvious risk factors for preterm birth or other major complications (multiple gestation; history of or current cervical cerclage; history of hypertension or diabetes; prior history of preterm birth or preeclampsia; age less than 16 or greater than 45) as well as immunologic problems (HIV, organ transplant, etc). Oral samples were collected within 24-48 hours after delivery. Samples were analyzed using 16S rDNA-based sequencing, where the DNA was extracted and then amplified by PCR using 16S rDNA primers. Data was processed using the UPARSE/SINTAX pipeline, and differentially abundant taxa were determined using the LEfSe method and MaAsLin2. The parent study enrolled 100 patients who delivered preterm and 205 patients who delivered at term. A subset of patients had oral samples collected, and 95 saliva and 70 tongue samples were analyzed using 16S rDNA-based sequencing. Communities from tongue and saliva were significantly different between women who delivered preterm and those who delivered at term, although not between those delivering low birthweight versus normal birthweight infants. When assessing beta diversity using the unweighted unifrac metric, patients who delivered very preterm (VPT, < 32 weeks) had a tongue microbiome that was consistently and statistically different from the tongue microbiome of patients who delivered both term (T, > 37 weeks) and late preterm (LPT, 32-<37). Differences remained significant after controlling for tobacco use. In a three-way comparison of saliva samples between the groups using MaAsLin2, the genus Lachnoanaerobaculum was significantly less abundant in the VPT group. Our studies suggest a tongue and salivary microbiome that differs between women delivering term and late preterm versus very preterm. Future studies are required to prospectively confirm differences and may yield data to design noninvasive tests to predict preterm birth risk.