Two strains out of 784 held plastic well, and mice given them had about two thirds less left in the gut

Why they ran it

The authors take as given that ingested plastic cannot be metabolised or degraded inside the body, and note that no method exists for clearing what has already been swallowed. Probiotics were the obvious candidate because they were already known to bind other unwanted material such as heavy metals and plasticisers, so the question was whether the same trick would work on plastic in a gut.

Drawn from the paper's introduction.

784 strains isolated from fermented foods were screened against 0.1 micrometre fluorescent polystyrene particles. 87 strains adsorbed more than 60 percent, and the best of them reached 80.5 percent.

Three strains were carried forward: DT66 at 71.4 percent, DT88 at 79.8 percent, and DT22 at 6.2 percent as a deliberately poor binder. Electron microscopy showed the two good binders coated in particles while the poor binder was largely bare.

Mice were given one strain daily for seven days and then a single dose of polystyrene. In the two good binder groups the particles had travelled further down the intestine after twenty minutes: 55.9 and 55.2 percent of intestinal length against 41.0 percent on saline.

In a second experiment mice received plastic and bacteria daily for seven days. Particles remaining in the ileum fell by 61.9 percent with DT66 and 66.8 percent with DT88 against saline. The poor binder DT22 changed residual levels in neither the ileum nor the caecum.

Interleukin 10 fell with plastic exposure. DT88 restored it in both serum and ileum and returned interleukin 6, tumour necrosis factor alpha and interleukin 1 beta to unexposed levels. DT66 restored interleukin 10 in serum only.

Faecal butyrate rose in the treated groups while acetate and propionate did not change, which the authors read as gut motility possibly contributing alongside the binding.

The numbers

Strains screened	784, of which 87 adsorbed more than 60 percent
Highest adsorption in the screen	80.5 percent
Adsorption of the three selected strains	DT88 79.8 percent, DT66 71.4 percent, DT22 6.2 percent
Intestinal transit at 20 minutes	55.9 percent DT66 and 55.2 percent DT88 of intestinal length, against 41.0 percent on saline
Residual polystyrene in the ileum	down 61.9 percent with DT66 and 66.8 percent with DT88; DT22 unchanged
Short chain fatty acids	butyrate rose, acetate and propionate unchanged
Funding and affiliation	all three authors are employees of Bluepha Co., Ltd., which owns the strains

Why this might happen

Shown in this study The study measured this step directly.

The bacteria and the particles clump together into aggregates large enough to see, and the authors' proposal is that something that size is harder to move across the gut wall and easier to sweep along and out.

The strongest evidence for that reading is the strain that was chosen to fail. DT22 adsorbs almost nothing, and it moved neither the transit measure nor the amount left behind, while the two good binders moved both.

The authors also raise a second route they cannot separate from the first. Butyrate rose in the treated animals, and butyrate influences how fast the gut moves things along, so some of the effect may be motility rather than cargo.

Drawn from the paper's discussion, its electron microscopy and its low binding control strain.

What this does not show

This does not show that more plastic left the animals. What the paper calls the excretion rate is the distance the fluorescent particles had travelled along the intestine twenty minutes after a single dose, divided by the total length of the intestine. That is a measure of how far material moved, not a count of what came out. The retention numbers in the second experiment are the stronger evidence and they are measured differently.
This does not show the effect holds for other plastics. The authors state it plainly as a limitation. Adsorption was quantified for polystyrene only. Aggregation with polyethylene, polycarbonate, polypropylene and polyethylene terephthalate was photographed but never measured.
This does not show what happens at real exposure levels. The mice received a milligram of plastic a day for a week, an acute high dose model. The authors name chronic low level exposure as the untested case.
This does not tell you where the aggregates go. The authors say directly that isolating and characterising bacteria and plastic aggregates inside a real gut is difficult and that the fate of those aggregates in other organs remains unclear.

Where this leaves us

Binding measured in a dish now predicts something measurable in an animal, which is the step this field had been missing. It refines rather than confirms the earlier tube work, because the amount retained and the amount bound are not the same quantity.

Male mice given a high daily dose of polystyrene for a week, alongside a daily dose of a single strain.

A study that collects and counts what actually leaves the animal, rather than measuring how far it has travelled, and one that runs at the exposure level a person might plausibly meet.

Caveats worth holding

All three authors are employed by the company that owns the strains tested. The negative control strain performing as a negative control is the internal check that makes the result harder to dismiss.
Male mice only, ten per group, acute high dose exposure over seven days.
The screen used 0.1 micrometre particles and the animal work used 5 micrometre particles, so the size tested in the dish and in the animal were not the same.
Adsorption was quantified for polystyrene only, which the authors list as a limitation.