Midtrimester microbial DNA variations in maternal serum of women who experience spontaneous preterm birth

Abstract

Objectives: To evaluate if midtrimester maternal serum contains microbial DNA and whether it differs between women with spontaneous preterm birth (SPTB) and those delivering at term.Study design: In this retrospective case-control study, we identified 20 healthy nulliparas with SPTB at 24-33 weeks of a nonanomalous singleton in 2014. Each case was matched by race/ethnicity to a control delivering at 39-40 weeks. Serum samples, collected at 15-20 weeks and stored at -80 C, were thawed and DNA extracted. PCR with primers targeting the 16S rDNA V4 region were used to prepare an amplicon library, sequenced using Illumina MiSeq, and analyzed using quantitative insight into microbial ecology (QIIME). Taxonomy was assigned using Ribosomal Database program (RDP) Classifier (threshold 0.8) against a modified Greengenes database. Differences in number of observed species, microbial alpha-diversity and beta-diversity, and taxa level analyses were undertaken.Results: All 40 samples were included. Women with SPTB had more unique observed species (p = .046) and higher mean alpha-diversity by Shannon index (but not Chao1 or Simpson) (p = .024). Microbial composition was different between groups by Bray-Curtis clustering (p = .03) but not by weighted (p = .13) or unweighted Unifrac (p = .11). Numerous taxa in the Firmicutes, Proteobacteria, and Actinobacteria phyla differed between groups (p < .05).Conclusions: SPTB is associated with distinct microbial DNA changes detected in midtrimester maternal serum.

Keywords: Maternal serum; microbial DNA; midtrimester; spontaneous preterm birth.

Figure 1.

Representations of differences in microbial composition between women who experienced spontaneous preterm birth (SPTB) and term birth by three analytic methods. A) Rarefaction curve demonstrating the number of observed species seen by sequences per samples for the SPTB (blue) and term (red) groups. Greater number of unique species were detected in the preterm as compared to term samples (p=0.046). B) Box plot of Shannon indices for the term birth group (Left bar; mean Shannon index = 5.39) and SPTB group (Right bar; mean = 6.44). The groups were significantly different (p = 0.024). C) Comparison of the overall microbial diversity between serum samples from patients with term versus SPTB. Principal coordinate analysis was used to generate a plot using Bray-Curtis metrics. Clustering of samples in the two groups (red = term birth, blue = SPTB) was significantly different (p<0.03).

Figure 2.

Phylum level relative abundance (y-axis) in samples comparing between women having spontaneous preterm birth (SPTB) and term birth