Aβ42 is known to be a primary amyloidogenic and pathogenic agent in Alzheimer's disease. However, the role of Aβ43, found just as frequently in patient brains, remains unresolved. We generated knockin mice containing a pathogenic presenilin-1 R278I mutation that causes overproduction of Aβ43. Homozygous mice exhibited embryonic lethality, indicating that the mutation involves loss of function. Crossing amyloid precursor protein transgenic mice with heterozygous mutant mice resulted in elevation of Aβ43 levels, impairment of short-term memory, and acceleration of Aβ pathology, accompanying pronounced accumulation of Aβ43 in plaque cores similar to the biochemical composition observed in patient brains. Consistently, Aβ43 showed a higher propensity to aggregate and was more neurotoxic than Aȕ42. Other pathogenic presenilin mutations also caused overproduction of Aβ43 in a manner correlating with Aβ42 and with age of disease onset. These findings indicate that Aβ43, an overlooked species, is potently amyloidogenic, neurotoxic, and abundant in vivo. 3 Alzheimer's disease, the most common form of dementia, is characterized by two pathological features in the brain, extracellular senile plaques and intracellular neurofibrillary tangles. Senile plaques consist of amyloid-β peptide (Aβ) generated from amyloid precursor protein (APP) through sequential proteolytic processing by β-secretase and γ-secretase. Two major forms of Aβ exist, Aβ40 and Aβ42, with Aβ42 being more neurotoxic due to its higher hydrophobicity, which results in faster oligomerization and aggregation 1 . A number of mutations associated with early-onset familial Alzheimer's disease (FAD) have been identified in the APP, PSEN1 and PSEN2 genes, and these mutations lead to accelerated production of Aβ42 or an increase in the Aβ42/Aβ40 ratio. Together these findings indicate that Aβ42 plays an essential role in the initiation of pathogenesis. However, the possible involvement of longer Aβ species that also exist in Alzheimer's disease brains has not yet been fully investigated.Thus far, various longer Aβ species, such as Aβ43, Aβ45, Aβ48, Aβ49 and Aβ50, have been qualitatively described in Alzheimer's disease brains 2 . Similar Aβ species have also been found in transgenic mice that overexpress APP carrying FAD-linked mutations 3 . Further quantitative studies have revealed that Aβ43 is deposited more frequently than Aβ40 in both sporadic Alzheimer's disease (SAD) and FAD [4][5][6][7] .How these Aβ species with different C-terminal ends are generated from the precursor has mainly been investigated by cell biological and biochemical methods. A number of studies 8,9 demonstrated that γ/ε-cleavage by γ-secretase activity controls the fate of the C-terminal end. Aβ43, generated from Aβ49 via Aβ46, is subsequently converted to Aβ40 by γ-secretase whereas Aβ42 is independently generated from Aβ48 via Aβ45. It has also been reported that the FAD-associated I213T mutation in the PSEN1 gene increases the generation of longer Aβ species, such as Aβ43, Aβ45 a...