Lack of detection of a human placenta microbiome in samples from preterm and term deliveries

Abstract

Background: Historically, the human womb has been thought to be sterile in healthy pregnancies, but this idea has been challenged by recent studies using DNA sequence-based methods, which have suggested that the womb is colonized with bacteria. For example, analysis of DNA from placenta samples yielded small proportions of microbial sequences which were proposed to represent normal bacterial colonization. However, an analysis by our group showed no distinction between background negative controls and placenta samples. Also supporting the idea that the womb is sterile is the observation that germ-free mammals can be generated by sterile delivery of neonates into a sterile isolator, after which neonates remain germ-free, which would seem to provide strong data in support of sterility of the womb.

Results: To probe this further and to investigate possible placental colonization associated with spontaneous preterm birth, we carried out another study comparing microbiota in placenta samples from 20 term and 20 spontaneous preterm deliveries. Both 16S rRNA marker gene sequencing and shotgun metagenomic sequencing were used to characterize placenta and control samples. We first quantified absolute amounts of bacterial 16S rRNA gene sequences using 16S rRNA gene quantitative PCR (qPCR). As in our previous study, levels were found to be low in the placenta samples and indistinguishable from negative controls. Analysis by DNA sequencing did not yield a placenta microbiome distinct from negative controls, either using marker gene sequencing as in our previous work, or with shotgun metagenomic sequencing. Several types of artifacts, including erroneous read classifications and barcode misattribution, needed to be identified and removed from the data to clarify this point.

Conclusions: Our findings do not support the existence of a consistent placental microbiome, in either placenta from term deliveries or spontaneous preterm births.

Keywords: 16S rRNA gene; Microbiome; Placenta; Preterm birth; Shotgun metagenomics.

Fig. 1

Quantitative PCR (qPCR) analysis of the 16S rRNA gene abundance in the samples studied. Values shown are the cycle of threshold (CT) of each sample. The limit of detection is a CT level of 40 (horizontal line). Samples with no detectable signal are shown above the line. Statistical comparisons of between sample types are in Additional file 1: Table S4. Data was pooled for samples generated using two DNA different extraction kits

Fig. 2

Overview of bacterial abundances inferred using 16S rRNA marker gene sequencing. a Heatmap showing relative abundances of bacterial taxa per sample (columns). Reads were aggregated at the genus level, which are shown as the different rows. Denser sample labeling is available in Additional file 2: Figure S2. Samples are grouped by sample type and further grouped by term or preterm delivery (more detailed sample labeling in Additional file 2: Figure S2). b PCoA of unweighted UniFrac distances, for all samples, colored by sample type. p values for group comparisons are in Additional file 1: Table S12. c PCoA of unweighted UniFrac comparing blanks to placenta samples

Fig. 3

Reads counts from shotgun metagenomic sequence analysis separated by sample type. “Total Reads” are all reads from the HiSeq sequencing run; “Nonhost Reads” are the reads remaining after Sunbeam human filtering, “Komplexity Filtered Reads” are those remaining after filtering low-complexity reads from the Nonhost reads, “Kraken Classified Reads” are those that remained after Komplexity filtering and were classified with Kraken, after removing Chordata, Arthropoda, and Apicomplexa

Fig. 4

Heatmap showing the relative abundance of classified reads in each sample from shotgun metagenomic sequencing. Columns show individual samples, rows show genuses, grouped by phylum. The lineages shown are the top three most abundant classifications per sample. Denser sample labeling is available in Additional file 2: Figure S5