Kitome or Placental Microbiome?

Placental dysfunction is known to cause many complications associated with pregnancy and delivery, however, the reason for initial dysfunction often remains a mystery. To investigate these causes, some researchers are investigating the microbiota associated with the placenta. In 2014, a study by Aagaard et al¹ challenged the long-standing idea of a sterile, bacteria-free mammalian womb. Utilizing 16S rRNA sequencing, the study was able to identify several bacterial DNA sequences in placenta samples and discovered a microbiome that closely resembled that of the oral cavity.

In the following years, several more studies were published that supported the findings by Aagaard and further developed the profile of the placenta microbiota. However, a recent study published in Nature by M.C. de Goffau et al.² analyzed over 500 unique placenta samples from healthy and complicated pregnancies and deliveries with multiple analytical methods and concluded that all bacteria signals, with the exception of Streptococcus agalactiae, are a result of laboratory contamination and sample handling.

Data Analysis Points to Contamination

While M.C. de Goffau initially sought to determine whether the placenta microbiota of a healthy pregnancy was distinctly different from that of a complicated pregnancy, they found that the majority of bacterial DNA in the samples was due to contamination. These results are supported by an extensive study by the University of Pennsylvania in 2016 that compared the bacterial sequences in placenta samples to that of the contamination background of a laboratory setting and was unable to identify any distinguishing differences³.

The group identified five main points of contamination that occurred during their study. This included contamination from:

1. Vaginal bacteria during labor and delivery
2. PBS during biopsy tissue wash
3. Trace DNA in extraction kits used for DNA extraction
4. Trace DNA in reagents used during amplification
5. Reagents and equipment used during sequencing

These contamination patterns were identified using principal component analysis (PCA) and any bacteria found to be correlated with processing rather than sampling was assumed to be contamination. Additionally, any bacteria found in the negative controls or buffers were also eliminated as contamination. The only bacteria signal that was sample associated was Streptococcus agalactiae, which was verified by metagenomics and 16S rRNA sequencing, and was independent of other variables. Furthermore, of the 537 collected placenta samples, S. agalactiae was positively identified in 7 samples by metagenomics, 16S sequencing, and qPCR.

Could Bacteria Have Been Missed?

One of the main criticisms voiced by many researchers, including Aagaard, was the concern that bacteria may exist in such low quantities that the extraction and sequencing methods were not able to detect the bacterial DNA. Efficient sample lysis and sensitive downstream applications are integral in addressing these concerns. Unfortunately, bias during sample lysis is extremely prevalent in microbiome studies due to lack of proper controls and validation of cell lysis methods.

The study did not indicate that the sample lysis methods in the Qiagen QIAamp® Miniprep Kit and MP Biomedical FastDNA™ Spin Kit had been optimized to ensure complete lysis was achieved. Although the study did include a single Gram-negative bacterium control in some samples (S. bongori), this control does not sufficiently represent the wide range of bacteria that may be present, including tough-to-lyse Gram-positive bacteria. As a result, the bias in the lysis step cannot be efficiently assessed. To ensure a method is achieving complete lysis and bacterial DNA is not left behind in intact cells, the lysis method would need to be optimized using a mixture of Gram-positive and Gram-negative bacteria with varying hardiness.

The ZymoBIOMICS Microbial Community Standard is composed of 7 bacteria (Gram-negative and Gram-positive) and 2 yeast of varying hardiness to best represent a mixed microbial community. Utilizing this standard to validate lysis and sequencing methods can help reduce instances of missing bacterial DNA due to inefficient lysis and resolve some critics concerns. Additionally, the ZymoBIOMICS Microbial Community Standard II (Logarithmic Distribution) can be used to identify the sensitivity of extraction and identification methods as it includes microbes at logarithmic dilutions down to 100 microbes.

Contamination from Trace DNA in Reagents

The majority of bacterial DNA identified in the samples in this study was found to originate from the DNA extraction kit used to process the sample, specifically the reagents. All sample handling and PCR preparation set up was done in a class 2 biological safety cabinet using single-use sterile equipment in order to reduce contamination risks as much as possible. Despite these precautions, the amplification of trace DNA present in the reagents in the extraction kit was still high enough to be detected at levels on par with potential sample bacteria. The high sensitivity of 16S rRNA gene amplicon sequencing and metagenomic sequencing makes low bioburden reagents and extraction kits essential to microbiome studies, especially when dealing with extremely low biomass samples such as these.

ZymoBIOMICS DNA kits are certified low bioburden and guaranteed to contain less than 3 bacterial genomic copies per µL of eluate. This has been rigorously optimized and tested by quantitative amplification of the 16S rRNA gene, and will help prevent contamination caused by kit reagents.

Unanswered Questions

The only sample-associated pathogen S. agalactiae found in this study was not associated with any common complications in pregnancy such as Preeclampsia, delivery of small for gestational age infants, or spontaneous pre-term birth. Even though the study’s initial question remains unanswered, identifying S. agalactiae as a true bacteria associated with the placenta can still aid doctors tremendously. If the pathogen is present in the womb before labor, the bacteria contamination can lead to fatal sepsis in the infant and routine screening and targeted antibiotics for the pathogen could potentially prevent nearly 200 neonatal deaths per year. However, limitations in diagnostic methods processing time, contamination in workflow reagents, and lack of controls still need to be addressed. Read the full report here.

The measured contamination and uncertainties of this study highlight the sensitivity of sequencing for microbiomics applications, and the need for low bioburden reagents and well-defined standards to ensure accurate results. Utilizing Zymo Research’s bias-free lysis system, microbial standards, and certified low bioburden reagents enables unbiased, ultra-pure DNA for all downstream applications including NGS and metagenomics.

References:
1. Aagaard, K., et al. “The Placenta Harbors a Unique Microbiome.” Science Translational Medicine, vol. 6, no. 237, 2014, doi:10.1126/scitranslmed.3008599.
2. Goffau, Marcus C. De, et al. “Human Placenta Has No Microbiome but Can Contain Potential Pathogens.” Nature, vol. 572, no. 7769, 2019, pp. 329–334., doi:10.1038/s41586-019-1451-5.
3. Lauder, Abigail P., et al. “Comparison of Placenta Samples with Contamination Controls Does Not Provide Evidence for a Distinct Placenta Microbiota.” Microbiome, vol. 4, no. 1, 2016, doi:10.1186/s40168-016-0172-3.