Is there a placental microbiota? A critical review and re-analysis of published placental microbiota datasets

Abstract

The existence of a placental microbiota is debated. The human placenta has historically been considered sterile and microbial colonization was associated with adverse pregnancy outcomes. Yet, recent DNA sequencing investigations reported a microbiota in typical human term placentas. However, this detected microbiota could represent background DNA or delivery-associated contamination. Using fifteen publicly available 16S rRNA gene datasets, existing data were uniformly re-analyzed with DADA2 to maximize comparability. While Amplicon Sequence Variants (ASVs) identified as Lactobacillus, a typical vaginal bacterium, were highly abundant and prevalent across studies, this prevalence disappeared after applying likely DNA contaminant removal to placentas from term cesarean deliveries. A six-study sub-analysis targeting the 16S rRNA gene V4 hypervariable region demonstrated that bacterial profiles of placental samples and technical controls share principal bacterial ASVs and that placental samples clustered primarily by study origin and mode of delivery. Contemporary DNA-based evidence does not support the existence of a placental microbiota.ImportanceEarly-gestational microbial influences on human development are unclear. By applying DNA sequencing technologies to placental tissue, bacterial DNA signals were observed, leading some to conclude that a live bacterial placental microbiome exists in typical term pregnancy. However, the low-biomass nature of the proposed microbiome and high sensitivity of current DNA sequencing technologies indicate that the signal may alternatively derive from environmental or delivery-associated bacterial DNA contamination. Here we address these alternatives with a re-analysis of 16S rRNA gene sequencing data from 15 publicly available placental datasets. After identical DADA2 pipeline processing of the raw data, subanalyses were performed to control for mode of delivery and environmental DNA contamination. Both environment and mode of delivery profoundly influenced the bacterial DNA signal from term-delivered placentas. Aside from these contamination-associated signals, consistency was lacking across studies. Thus, placentas delivered at term are unlikely to be the original source of observed bacterial DNA signals.

Keywords: 16S rRNA gene sequencing; Low microbial biomass sampling; Microbiome; Placenta; Sterile womb.

Fig. 1

Study inclusion flowchart Four searches were performed on PubMed to identify studies for inclusion in the re-analysis. Filtering criteria were: primary research article, 16S rRNA gene sequencing data, placentas obtained from term deliveries, sequencing data accessible with published accession number, and sufficient metadata available to parse sequencing data into individual samples.

Fig. 2

Conclusions of thirteen studies evaluating the existence of a placental microbiota, which included data from multiple placentas delivered via cesarean section at term The studies are principally separated and contrasted depending upon whether they included technical controls to account for potential background DNA contamination.

Fig. 3

Principal Coordinates Analyses of the beta diversity of bacterial DNA profiles between placental samples and technical controls in published placental microbiota studies Studies were included if technical controls were sequenced and made publicly available to account for background DNA contamination. Beta diversity between placental (red open circles) and technical control (black open circles) samples is illustrated by study in PCoA plots based on the Bray-Curtis dissimilarity index. Genus level classifications of the top ten ASVs in placental samples and technical controls by total reads are plotted at their weighted average positions (grey diamonds). Asterisks indicate ASV sequence genus level classifications which were assigned by NCBI BLAST with the highest percent identity in excess of 95%.

Fig. 4

Principal Coordinates Analyses of the beta diversity of the bacterial DNA profiles of placental and vaginal/vaginal-rectal samples in placental microbiota studies Prior published studies were included if vaginal or vaginal-rectal samples were sequenced and made publicly available alongside placental samples. The top ten ASVs shared between placental samples and technical controls, and the top ten ASVs in vaginal samples are plotted at their weighted average positions in the ordination space (grey diamonds) and their genus level classifications are noted. Agglomerated genus level classifications were plotted for the Liu dataset instead of ASVs since no ASVs were greater than 1% mean relative abundance across placental samples. Asterisks indicate ASV sequence genus level classifications which were assigned by NCBI BLAST with the highest percent identity in excess of 95%.

Fig. 5

Placental and technical control samples cluster by study origin, mode of delivery, and gestational age at delivery A) Beta diversity between placental (open circles) and technical control samples (open triangles) from studies which sequenced the V4 hypervariable region of the 16S rRNA gene is visualized through principal coordinate analysis (PCoA) based on the Bray-Curtis dissimilarity index. B) Beta diversity of placental samples without technical control samples from each study. C) Placental samples from the same six studies were characterized by mode of delivery and gestational age at delivery. Weighted average positions of ASVs greater than 1% were plotted as grey diamonds and labelled with genus level classifications. Asterisks indicate ASV sequence genus level classifications which were assigned by NCBI BLAST with the highest percent identity in excess of 95%.

Fig. 6

Heatmaps of the bacterial DNA profiles of placental and technical control samples from studies which sequenced the V4 hypervariable region of the 16S rRNA gene demonstrating a high degree of overlap between these two sample types ASVs are listed by study if they had a mean relative abundance greater than 1% across placental samples (green bar). Red asterisks indicate ASVs which had a mean relative abundance greater than 2% across all technical control samples (purple bar) from that study. Regular asterisks indicate ASV sequence genus level classifications which were assigned by NCBI BLAST with the highest percent identity in excess of 95%.