The discovery and mapping to chromosome 21 of the Alzheimer's amyloid gene: history revised.

I was invited by the editor of the Journal of Alzheimer’s Disease to give a short account of our efforts twenty years ago to clone the gene that encodes the amyloid-β peptide, or beta (β)-protein as it was called. At that time, I was at the New York State Institute for Basic Research at Staten Island and a member of the Department of Molecular Biology headed by David Miller. Director of the Institute was the late Henry Wisniewski, a distinguished neuropathologist and expert on Alzheimer disease (AD) whose research group was working on the pathological consequences and molecular nature of insoluble aggregates found in AD including neurofibrillary tangles (NFTs) and amyloid plaques. At the time, the primary structure of amyloid-β was not known, but in 1984 using guanidine hydrochloride, George Glenner was able to solubilize amyloid isolated from leptomeningeal vessels of Down syndrome and AD patients. Using HPLC, he determined that a main component in his solution was a small peptide of an apparent molecular mass of 4.2 kDa. He called this peptide “β-protein” because of its ability to form noncovalent polymers of β-pleated sheet structures that aggregated to form fibers identified histologically by Congo red staining and optical polarization properties. He published this data including the sequence of the first 24 residues of β-protein in two BBRC reports [1, 2]. At the time, I was working mostly on the genetics of prokaryotes and of scrapie prion (PrP) proteins [3,4], but being at the Institute of Basic Research, I developed an interest in AD. The technology of gene cloning at that time was rather primitive compared to today’s standards and was mostly based on reverse genetics using oligonucleotide probes to mimic nucleotide sequence encoding amino acid residues and/or antibodies against specific antigens. Publication of the β-protein (hereafter referred to as amyloid-β) sequence stimulated my interest to clone the encoding gene, and with the encouragement of both Henry and David, my group, consisting of myself, Ram Ramakrishna and Gloria Wolfe, went to work on the problem. We prepared oligonucleotide probes based on Glenner’s amino acid sequence and started by screening genomic libraries because the origin of amyloid-β was unknown, and it was not clear whether or not the protein was expressed in the brain. Actually, at the time, the predominant concept was Glenner’s suggestion that amyloid-β was an abnormal derivative of a serum precursor protein that was endocytosed and processed in the lysosomal compartment of cerebrovascular endothelial cells. It was thought that following processing, amyloid-β was aggregated to form amyloid fibrils that were then exocytosed and deposited in the neuropil or in cerebral vessels (for a detailed account of this concept, see [5]). After a while, it became clear that screening genomic clones was not a very productive route of investigation, as we encountered difficulties stemming