In Mice, Plastic Particles Harmed Pregnancy Through Vaginal Microbes
By the 14th day of pregnancy, the particles were almost impossible to find. Whatever the plastic had done to these mice, it had done before the pregnancy began, and something else was carrying the damage forward.
Wan Liu, Huidong Zhang and colleagues point to the community of microbes that lives in the vagina. Their study, published Sept. 25, 2026, in ACS Nano, exposed female mice to high doses of aged polystyrene nanoparticles before the animals were pregnant. Miscarriage followed, and the embryos that kept developing were held back. The work was led from Central South University in Changsha and the Eighth Affiliated Hospital of Sun Yat-sen University in Shenzhen, with co-authors at several other Chinese universities and hospitals.
Their starting point was a gap. Nanoplastic toxicity has been reported in one organ after another, while its effect on the vaginal microbes that healthy reproduction depends on was largely unknown.
The paper's own title calls the doses high, which is the first thing to hold on to. This is a mechanism traced in an animal at a deliberately heavy exposure, not a measured risk to people.
The chain the researchers describe has several links, and all of them persist. Exposure before pregnancy tipped the balance of the vaginal community, and one bacterium, Aerococcus urinae, became more abundant. A transporter called Slc1a5, which pulls the amino acid glutamine into cells, was suppressed, and glutamine in the vaginal tissue ran down. Genes that drive inflammation were turned up. All of it carried over from before pregnancy into pregnancy, which is how the particles could be almost undetectable in the vagina on day 14 while the pregnancies still failed.
Two sets of experiments test that chain rather than describe it. The authors report that transplanting A. urinae into the vaginal tract made both the sequence and the pregnancy outcomes worse. That casts the bacterium as something that aggravates the damage, not something that produces it by itself. Three interventions then pushed the other way: raising Slc1a5 again, supplying glutamine or treating the inflammation. Each recovered the processes and eased the outcomes.
The vaginal route is the new part, and it is narrower than it sounds. That plastic particles can reach a mouse pregnancy through microbes was already on record. A 2025 paper in the Journal of Nanobiotechnology traced nanoplastic-induced fetal growth restriction to the gut community and to the placenta's handling of nicotinamide, a form of vitamin B3. A 2024 paper in the same journal as this one went at the placenta directly, reporting that polystyrene nanoplastics interfere with the cells that build it. What the authors present as new is the vaginal community working as a lasting route from exposure to pregnancy loss, and they describe it as a previously unrecognized long-term nanoplastic reproductive toxicity.
Whether any of this describes women is a separate question, and the dose is not the only reason for care. The vaginal microbiome of a laboratory mouse is not a small version of a woman's. Healthy human communities are usually dominated by Lactobacillus, bacteria that keep the tissue acidic. Conventional mice carry a different mix and are poorly colonized by the human species, which is why researchers build mice with humanized microbial communities to study the human system. A route that runs through microbes in a mouse vagina cannot be assumed to run the same way in a woman.
The human end of the question has support of its own, short of cause and effect. Marwa Saadaoui and colleagues reviewed that literature in Frontiers in Cellular and Infection Microbiology in 2023: a disturbed vaginal community accompanies several adverse pregnancy outcomes, and its part in miscarriage is still not well understood.
The particles used here were aged, which is the state plastic tends to be in by the time it is out in the world, where nanoplastics have been measured in Arctic sea ice and snow.
The team's own conclusion is that this gives an experimental basis for treatments aimed at the chain. What the work more plainly gives is a short list of things to look for: one bacterium, one amino acid, one transporter and a window that opens before pregnancy instead of during it.