Culture-independent next-generation sequencing technologies have given us a far deeper understanding of the microbiome composition of various important health-related niches, most notably the gut microbiome. 

The oral, vaginal and skin microbiomes have similarly been analysed successfully but things have gotten out of hand with people looking for microbiomes everywhere, even in places previously considered sterile. Outlandish reports (from an ecological point of view) of the supposed brain, blood, breast tissue and, not least, the placental microbiome have been published in various high-impact journals and are creating a perfect storm of confusion. 

The cause of all this misinformation, as shown by Salter et al., is bacterial contamination of reagents. When sites of limited (or zero) microbial biomass are analysed, the bacterial reagent contamination (DNA remains of various dead environmental bacterial species) is instead amplified during the PCR process and is subsequently sequenced and detected. A lack of properly adapted methodology for these extremely low microbial biomass samples and a lack of basic microbiological ecological understanding have led to the current situation.

As a result, the sterile womb paradigm has been challenged by the in utero colonization hypothesis, as reviewed by Perez-Muñoz et al. There are two main lines of evidence for the non-existence of a placental/uterine microbiome. (1) The creation of mice completely devoid of bacteria (axenic animals) via cesarean section would be impossible if there were always bacteria present within the amniotic fluid or if there is an active transfer of bacteria from the mother to the infant. (2) Species identified using culture-independent techniques cannot be found using normal culturing techniques, even though these specific bacteria are normally very easy to culture. Furthermore, if there was an active directed colonization effort in utero, why do we not primarily detect the most important beneficial bacterial groups (e.g., Bifidobacteria)? A directed colonization effort does indeed exist; it is however postnatal, in the form of breastfeeding. Not only does breastfeeding select for the right bacteria, but it does actually provide these specific bacteria, which can be seen both by culturing and sequencing.

All of this is not to say that the womb is always sterile; things do sometimes go wrong. Bacteria are frequently found in the amniotic fluid of mothers who deliver prematurely, particularly in extremely preterm cases. In such cases these bacteria are subsequently also detected in the meconium, as amniotic fluid is swallowed by the foetus. Proposed mechanisms for amniotic colonization include the ascension and translocation of vaginal microbiota, or that they are obtained via the bloodstream (with an oral or faecal origin). Meconium research by itself, while potentially indicative of an in utero colonization event, should however not be seen as reflective of an in utero microbial environment. Rapid dissemination of bacteria through the meconium either via the oral or rectal route, or simply contamination of the meconium during defecation, will allow low numbers of vaginal, faecal, and skin bacteria that are picked up during delivery to end up in the meconium.

Infection of placental tissues (chorioamnionitis) by, for example, Streptococcus agalactiae (GBS) or Listeria monocytogenes (both competent intracellular pathogens) represents another prime example of a non-sterile womb. An article by Urushiyama et al. on the microbiome profile of amniotic fluid in patients with chorioamnionitis (with different levels of placental inflammation) provides an excellent example of real signals of opportunistic pathogens in most of the stage three and some of the stage two cases (severe inflammation). A reagent contamination profile (primarily a combination of Proteobacteria, Firmicutes and Propionibacterium acnes in this study) is seen in most of the stage one and stage zero cases, healthy controls and blanks. It is furthermore known that the risk of chorioamnionitis increases with each vaginal examination; it thus, does not take that much for bacteria to end up in the uterine environment. Microscopic examination typically does not find any evidence for the presence of intracellular bacteria in the large majority of spontaneously delivered placentas (basal plate location), but it is found in about half of the samples from spontaneous deliveries from extremely preterm cases.

In conclusion, the womb is indeed not always a sterile environment as bacteria (opportunistic pathogens) are found in a number of (adverse) circumstances. There is however no such thing as a healthy in utero microbiome.