Consistent with the in vitro results memantine

Consistent with the in vitro results, memantine ameliorated Aβ-induced cytotoxicity in rat primary cortical cultures. Studies reported that memantine protected neurons against the damage caused by aggregated Aβ40 or Aβ42 oligomers in vivo[26], [27]. These results suggest that, in addition to its neuroprotective effect, memantine reduces the toxicity of oligomeric Aβs by inhibiting molecular aggregation. Furthermore, memantine inhibited the aggregation not only of human Aβ(1–42) but also of other Aβ peptides, including mouse Aβ, [Pyr3]- Aβ(3–42) (found in AD patients), Aβs carrying amino ac 723 substitutions linked to early-onset FAD, and Aβs with different N- or C-terminal lengths. These results suggest that memantine inhibits Aβ aggregation regardless of the amino acid sequence around the mutation site in Aβ. Of note, the effect of memantine on [Pyr3]-Aβ(3–42) and the Aβ mutant D23 N was weak, indicating that the N-terminus and residue D23 of Aβ may be involved in the interaction and/or effect of memantine on Aβ oligomerization. Additional mutational studies are needed to clarify the mode of action of memantine on Aβ aggregation. Our results suggest that memantine may benefit a wide range of patients, including those with SAD and FAD. Memantine inhibited the aggregation of recombinant human Aβ(1–42) and disassembled preformed Aβ(1–42) fibrils in a concentration-dependent manner in vitro. However, amantadine, a structurally similar NMDA antagonist, did not affect the formation and disassembly of Aβ aggregates at the same concentrations. Therefore, the effects of memantine on Aβ aggregation are mediated by a NMDA receptor-independent mechanism. In this study, we used Aβ peptides at the concentrations of 1–5 μM and memantine at 0.1–3 mM. In patients treated with memantine, its concentration in the CSF ranged from 0.05 to 0.3 μM [36]. In 128 participants from the DIAN cohort, the Aβ42 concentration in the CSF was 218 pg/mL in mutation carriers (in one of three genes, APP, PSEN1 or PSEN2) and 398 pg/mL in noncarriers [37]. In the ADNI cohort, the mean concentration of Aβ(1–42) in the CSF was 229.0, 141.0, and 134.0 pg/mL in cognitively normal subjects, mild cognitive impairment (MCI) patients, and AD patients, respectively [38]. These values indicate that the concentration of memantine and Aβ42 overlaps with the concentration used in this study, suggesting that the effect of memantine found here would be observed in MCI or AD patients treated with memantine.
Acknowledgements We are grateful to C. Nishimoto for the assistance in conducting the study and to S. Kumakura, R. Murakami, and K. Saito for their support and suggestions. This study was supported by Daiichi-Sankyo Co., Ltd.
Introduction Alzheimer's disease (AD) is a complex neurological disorder characterized by a progressive dementia as a consequence of synaptic failure [1], [2]. The amyloid hypothesis maintains that the pivotal event in AD is the production of toxic amyloid-β (Aβ) peptides following the proteolytic cleavage of the amyloid precursor protein (APP) [3]. In animal models intracerebral injections of Aβ peptides caused synapse damage and impaired memory formation [4], [5]. Neurodegeneration is not directly proportional to concentrations of Aβ; rather that it is dependent upon numerous factors including the state of Aβ aggregation and specific Aβ conformations. Perhaps the key to understanding the amyloid hypothesis is the realization that there exist conformational forms of disease-relevant Aβ, while other conformations are less toxic or even biologically inert. Therefore, we sought to identify factors that specifically affected the release of toxic forms of Aβ. Chinese hamster ovary (CHO) cells stably transfected with cDNA encoding the human (APP)751, known as 7PA2 cells, have been extensively used as they release soluble Aβ [6], [7] that have similar properties to the soluble Aβ species found within the brains of AD patients [8], [9], [10], [11].