Unfortunately, it has proven very difficult to develop drugs capable of ameliorating the disease. After a tremendous burst of progress in the 1990s, the pace of discoveries has slowed. Dr. Saido believes that part of the difficulty is the inadequacy of current mouse models to replicate the real conditions of Alzheimer's disease and allow an understanding of the underlying mechanisms that lead to neurodegeneration. In fact, much of the research in Alzheimer's disease over the past decade may be flawed, as it was based on unrealistic models.
The problem with older mouse models is that they overexpress a protein called amyloid precursor protein, or APP, which gives rise to the amyloid-beta (Abeta) peptides that accumulate in the brain, eventually leading to the neurodegeneration that characterizes Alzheimer's disease. However, in mice the overexpression of APP gives rise to effects which are not seen in human Alzheimer's disease.
For example, the APP mutant mice often die of unknown causes at a young age, and the group believes this may be related to the generation of toxic fragments of APP, such as CTF-beta. In addition, some of the fragments of APP could be neuroprotective, making it difficult to judge whether drugs are being effective due to their effect on Abeta peptides, which are known to be involved in human AD, or whether it is due to other effects that would not be seen in human disease. In addition, the gene for expressing APP is inserted in different places in the genome, and may knock out other genes, creating artifacts that are not seen in humans.