I would like to point out that the main study described in this article is not the "inject aggregated amyloid beta fibrils" that is described first. That study was a preliminary study that prompted the main study which uses genetically altered mice.
The test for alzheimers for the first study (previously reported but summarized again) was to quantify how much the mice deviates from solving a maze that they have been trained to solve.
In the first they injected amyloid beta aggregates into mouse brains and found that EPPS administered orally at 30 mg/kg and 100 mg/kg restores the ability of the mice to efficiently solve the maze.
Next they tested toxicity quantified the amount of EPPS that passes the brain/blood barrier. For toxicity they found no signs of toxicity at 2000 mg / kg (20x dosage). For blood/brain barrier, as you go up in blood concentration you should go up in brain concentration if there is a good penetration from blood to brain. If the barrier is high then you immediately get high blood and low brain concentrations. The point where there is no longer a significant increase in brain concentration when increasing blood concentration is used to determine effective dosage concentrations. They found that at 100 mg/kg they were starting to see increased blood/brain ratios so they targeted 10-100 mg/kg for the next study.
MAIN STUDY (which included identifying the dosage level) used mice that were engineered to "get Alzheimer's" starting around 5 months of age because they produce a human gene (transgenic) that is a precursor to form the AB plaques. This transgenic model is established and the mice showed the expected amyloid beta plaques and had difficulty solving the maze at 10.5 months as expected.
Starting at 10.5 months they gave oral doses of EPPS at 10 mg/kg and 30 mg/kg and monitored maze solving along with several additional tests: likelihood to freeze when presented with negative input (fear conditioning) and ability to find hidden platforms when swimming (water maze). Both tests improved significantly to the wild-type (no Alzheimer's) level when taking EPPS. They also did dose dependency at .1 1 and 10 mg/kg. There was a steady improvement at higher doses.
They also took slices of the mouse brain and tested whether or not the neurons responded differently to electrical stimulation. They found no difference in wild-type (WT, non-genetically altered) or transgenic (TG, altered) response to electrical stimulation with and without EPPS. This hints at no difference in neural activity with or without EPPS. They also gave EPPS to WT for the behavioral tests and did not see a difference (although that was not shown in the behavioral test figures).
They also took slices of the brain and stained them with a fluorescent dye to show the Alzheimer's associated plaques. There is a significant quantifiable reduction in plaques in the treated mice.
They used several other techniques to confirm that they were actually AB plaques and they disaggregated by a specific site of activity. I won't go into those specifics, but to say that this was a VERY well designed and executed study across multiple lines of inquiry and all of the lines of inquiry point to the same conclusion:
EPPS rescues hippocampus-dependent cognitive
deficits in APP/PS1 mice by disaggregation of
amyloid-b oligomers and plaques
And that's why it's a Nature article.