A largely overlooked amyloid-β (Aβ) peptide could play a greater role in the pathogenesis of Alzheimer disease (AD) than previously envisaged, scientists claim. Mice engineered to overexpress Aβ43 were found to display impairment of short-term memory and acceleration of amyloid-β pathology. This was accompanied by pronounced accumulation of Aβ43 in plaque cores that were similar in biochemical composition to those observed in the brains of affected human patients.
In fact, the researchers state, Aβ43 showed a higher propensity to aggregate and was more neurotoxic than Aβ42, which has to date been thought of as the main amyloidogenic and pathogenic agent in AD.
Takaomi C. Saido, Ph.D., and colleagues at Japan’s Riken Brain Science Institute, describes this research in Nature Neuroscience in a paper titled “Potent amyloidogenicity and pathogenicity of Aβ43.”
Building on independent research suggesting that Aβ43 appears to be as prone to aggregate in vitro as Aβ42, the team set out to investigate the possibility that Aβ43 may also contribute to AD pathology.
The researchers exploited a presenilin-1 (PS1) mutation known as R2781, which results in selective overproduction of Aβ43 in vitro and has also been found in patients with atypical forms of Alzheimer-like disease.
Unfortunately, mice engineered to be homozygous for PS1-R2781 died at the embryonic stage, so a heterozygous R2781 knock-in model was generated for initial in vivo evaluation. PS1-R2781 heterozygous knock-in mice were normal in terms of development and anatomy at both three and 24 months of age.
ELISA assays developed to assess levels of Aβ40, Aβ42, and Aβ43 in mouse brains at different time points, though, showed that Aβ43 increased on aging. Additionally, there were increases in the Aβ42/Aβ40 and Aβ43/Aβ40 ratios in older heterozygous mice, which appeared to be caused primarily by a decrease in Aβ40. The R278I mutation in addition led to an elevation in the Aβ43/Aβ42 ratio in aged mice.
“Taken together, these findings indicate that the PS1-R278I mutation gives rise to a modest in vivo effect in terms of the levels of endogenous Aβ species under heterozygous conditions,” the researchers state.
To test the effects of homozygocity on Aβ ratios, the team generated mouse embryonic fibroblasts that were homozygous for the mutated presenilin gene. Aβ43 markedly increased in the homozygous knock-in MEFs, whereas Aβ42 levels remained unchanged. The combined findings hinted that the R2781 mutation acts to inhibit Aβ43 to Aβ40 conversion, leading to increased Aβ43 levels and concomitant decrease of Aβ40, but without altering Aβ42 levels.
The researchers moved on to evaluate Aβ proteins in a mouse model of disease generated by crossing animals heterozygous for the R2781 mutation, with mice carrying a human amyloid precursor protein (APP) isoform. These animals exhibited short-term memory loss before progressing to develop brain plaques and accelerated amyloid pathology. Pathological Aβ deposits started to accumulate in the double-mutants at around six months of age, whereas it took about 12 months for APP transgenic mice to begin to show signs of amyloid deposits. Behaviorally, 3–4 month old APP × PS1-R278I mice also exhibited short-term memory impairment.
The researchers then quantified the steady-state levels of Aβ40, Aβ42, and Aβ43 in the brains of APP and APP × PS1-R278I mice at three and nine months. Notably, only the double-mutant mice exhibited selective elevation of Aβ43 at three months, which is before the pathological deposition of Aβ but at about the same time that the animals started to show short-term memory impairment.
This suggests that Aβ43 may be the initial seeding species and the trigger for memory deficits in the mouse model, the authors comment. In contrast, Aβ40 and Aβ42 levels started to increase at around nine months, they note.
To compare the effects of Aβ43 with those of Aβ42, the team developed a separate mouse model by crossing APP transgenic animals with those carrying a PS1 mutation (PS1-M146V) that leads to overproduction of Aβ42. As expected, these animals developed selective accumulation of Aβ42, but although the respective steady-state levels of Aβ42 in the APP × PS1-M146V mice was about 10-fold greater than that of Aβ43 in APP × PS1-R287I mice at nine months, the total plaque areas were actually similar in the two animals.
Interestingly, analysis of plaque composition in the two double mutants suggested that Aβ43 was even more prone to seed cores in plaque formation than Aβ42 and that a relatively small amount of Aβ43 is sufficient to accelerate Aβ amyloidosis and induce plaque core formation in vivo.
Neurotoxicity studies separately indicated that Aβ43 demonstrates a higher level of neural toxicity than Aβ40 or Aβ42. “These results indicate that Aβ43 directly affects neural toxicity and induces synaptic dysfunction, which would contribute to short-term memory impairments before the amyloidogenesis,” the Riken team writes.
The overall findings led the researchers to conclude that Aβ43 should be separately analyzed from Aβ42 in terms of its role in AD. The observation that basal Aβ43 levels substantially increased with aging in wild-type mice could in addition make Aβ43 a potentially valuable cerebrospinal fluid biomarker for presymptomatic diagnosis of AD as well as a putative therapeutic target.