Colin L Masters1 and Konrad Beyreuther1
Our contributions in the 1980s (Kang et al. 1987; Martins et al. 1986; Masters et al. 1985a,b) to the purification and N-terminal sequencing of the amyloid plaque cores (APC) of Alzheimer's disease (AD) and the discovery of its biogenesis from a neuronal precursor (the amyloid protein precursor - APP) by proteolytic cleavages (the P- and y- secretases) need to be seen against the background of many years of prior research activity from a diverse range of individuals and groups.
Prior to the mid-1960s, very little progress had been made in understanding the pathological significance and biochemical nature of the amyloid depositions either in the brain (the concept of amyloid dates from Virchow's description of cerebral corpora amylacea) or systemically. Von Braunmuhl was among the leaders of the German school of pathologists who attempted to understand the "colloidal" nature of amyloid. The same school had developed the use of the cotton dyes, such as Congo Red, for the differentiation of amyloid from other proteinaceous deposits. The fact that abnormal degenerative and regenerative changes occurred in response to cerebral amyloid deposition was fully appreciated but the origin of the cerebral deposits remained enigmatic, particularly since some forms were clearly associated with small blood vessels, the amyloid congophilic angiopathy (ACA) of Pantelakis. The concept of a vascular or hematogenous origin of the cerebral amyloid plaque clearly arose during this period, and was promoted by the general (non-neuropathologically trained) pathologists who were used to evaluating the systemic forms of amyloidosis.
Three major intellectual streams emerged during the latter half of the 1960s. First, Friede described the histochemical reactivity of the AD "senile" plaque and observed
1 Department of Pathology, The University of Melbourne and the Mental Health Research Institute of Victoria
2 Centre for Molecular Biology, The University of Heidelberg
Address for correspondence: Colin L Masters, Department of Pathology, The University of Melbourne, Victoria 3010, Australia * [This is an abbreviated version of a longer article written for the Alzheimer 100 Special Issue of the Journal of Alzheimer's Disease edited by G. Perry, J. Avila, J. Kinoshita and M. Smith.]
the enzymatic activity specific for acetylcholinesterase. This observation eventually developed into the cholinergic theory of AD and the current class of cholinesterase inhibitors that are useful in the symptomatic treatment of AD. Second, the emergence of the technology behind electron microscopy led to the "great tangle debate" between the schools of Bob Terry and Michael Kidd: was the Alzheimer neurofibrillary tangle (NFT) a paired helical filament or a twisted tubule? Third, and more productively, the unlocking of the structural basis of the systemic amyloid filament had begun.
Early studies on the structural and biochemical nature of the Alzheimer plaque - the 1970s
In the early 1970s, little progress was made in understanding the amyloid plaques in both AD and the "slow virus" diseases. Remarkably, the first attempts at purifying the APC came from James Austin's group in Denver, Colorado, in 1971-1972, but they had used formalin-fixed tissues. Nevertheless, their studies did reveal the presence of non-proteinaceous elements such as silicon.
In the first half of the 1970s, George Glenner had made spectacular progress in the biochemical elucidation of the AL types of systemic amyloid. Glenner, who died in 1995 at the age of 67 from the complications of cardiac amyloidosis, was among the first to obtain N-terminal sequences on the amyloid AL light chains. He had conducted his seminal studies at the NIH in Bethesda, having moved there in 1958.
Masters' first post-doctoral position was with EP Richardson in Boston in 19761977, continuing earlier studies on the AD/CJD amyloid plaques at a morphological level. In late 1977, Masters moved to the NIH laboratories of Carleton Gajdusek and Joe Gibbs, to continue the collaboration that had started in 1968. After some discussion with Joe Gibbs, it was agreed that Masters should start a project on purifying and characterizing the amyloid plaques from the human transmissible diseases (kuru, CJD and the Gerstmann-Straussler Syndrome). Joe Gibbs, of course, knew of George Glenner's scientific reputation and ofhis presence on the NIH campus, and he suggested that Masters visit Glenner's laboratory. By this time, Glenner had begun to think about the AD-amyloid connection and had decided that the best approach would be to isolate the AD amyloid from the leptomeningeal vessels, but he had not commenced work on this subject. Masters and Glenner met on two or three occasions, and Glenner then drafted a research proposal, which he sent to Joe Gibbs. As a result of this, an intramural NIH conference was convened in 1978, at which Glenner, Masters, Gibbs and other intramural scientists presented ideas on how to approach the general methods involved in purifying the AD/CJD amyloids. From the very start, Glenner assumed that the AD amyloid was derived from the vascular compartment, in the same manner as had been identified for the AA and AL proteins.
Over the ensuing two years (1978-1980), Glenner and Masters met only on one or two occasions in Glenner's laboratory to discuss progress. At this stage, samples had not been exchanged between their respective laboratories at the NIH; neither was it clear that Glenner had actually begun dissecting any human AD brain tissues.
Masters' approach at the NIH was to try and adapt the known detergent-high salt extraction methods previously used for the purification of intermediate filaments, relying on the relative insolubility of the APC of both kuru and GSS brains. Because of the scarcity of tissue samples and the low numbers of APC in these conditions, Masters also began using AD brains to establish the methodologies.
While these studies were going on at the NIH from 1978 to mid-1980, other connections and collaborations were being established. In the AD field, several groups were actively engaged in the purification of the NFT, including Dennis Selkoe in Boston and Henryk Wisniewski and his team at Staten Island (including Khalid Iqbal and Patricia Merz). On one of their many visits to Salem and Boston, Gajdusek and Masters dropped in unannounced to see Selkoe at his McLean laboratory, to check on progress with his NFT preparations. This would have been in 1979, and it was apparent at that time that Selkoe was not directly working on an APC purification strategy. In contrast, Pat Merz and Steve Bobin at Staten Island were very interested in the amyloid purifications in both scrapie/kuru/GSS and AD. Masters, Bobin and Merz set up a collaboration in which they shared protocols, samples, and techniques. This collaboration eventually resulted in a publication in 1983 (submitted in November 1982), which was the first to describe in detail some of the methods that had been jointly developed for the purification of amyloid from AD, GSS and scrapie frozen brains. This paper concentrated on the electron microscopic appearances of the different types of amyloid filaments, which at the time was a very controversial area because of the studies emerging from the Prusiner laboratory in San Francisco. In retrospect, it was evident that we had relatively "pure" preparations of scrapie/GSS amyloid in our laboratories at a time well before Prusiner had "pure" preparations of the prion protein (PrP). If we had been able to solubilize and characterize the scrapie/GSS amyloid protein at that time, it would have led us directly to the PrP protein, pre-empting Prusiner's later discoveries by several years.
Dramatic discoveries on many fronts in the 1980's
Masters left the NIH labs in the latter half of 1980 for a year in Heidelberg with Melitta Schachner before returning to Australia in 1981, where he recommenced studies on the sporadic and genetic cerebral AD/CJD amyloids. George Glenner moved to San Diego in 1982, having published a major review on the "P-fibrilloses" in 1980. His attempts to dominate the amyloid nomenclature debate with his emphasis on their P-pleated sheet structure were to be carried forward in his AD studies in California, where he found more ready access to AD brain tissues. During 1982-1983, Masters was in communication with Glenner, and samples of pure AD-APC were sent to him on the understanding that it was a collaboration in which he would perform X-ray diffraction studies. It was never clear what became of those samples, as results were never forthcoming from his laboratory. Also, in 1982, Prusiner visited Masters' Australian laboratory while attending the International Congress of Biochemistry in Perth, at which he set up the collaboration with Charles Weissman that was to lead to the eventual cloning of the PrP gene. During Prusiner's visit, Masters discussed progress in isolating the AD and GSS amyloid. Although Prusiner, at that time and for many years thereafter, maintained that his "prion rods" were distinct from amyloid fibrils and Merz's "scrapie associated fibrils," it came as a great surprise that he subsequently consulted with Glenner and published observations on the Congo Red negative birefringence of the aggregated prion rods.
In 1983, another surprising paper appeared from Michael Kidd, Mike Landon and David Allsop in the Nottingham Medical School, disclosing their method of AD-APC purification (discontinuous sucrose gradient with subtilisin pre-digestion) and showing that their total amino acid composition was different from the known AA/AL amyloid proteins. We had not been aware of competitors other than Glenner. Much later, we also learned that Alex Roher had also been working on purified APC. In subsequent discussions with Allsop and Landon, it was clear that they had made plans to determine the N-terminal sequence, but their chosen collaborator failed to deliver. Moreover, they apparently had not discovered a method to solubilize the APC, a precondition for determining the N-terminal sequence.
In retrospect, it was clear that Glenner had been very busy and productive during 1983 and 1984, as his two papers on the N-terminal sequence of the AD amyloid protein appeared in May and August 1984. As expected, he had confined himself to the amyloid extractable from the leptomeningeal vasculature, and his method required predigestion with collagenase and (partial) solubilization in 6M guanidine-HCl, followed by Sephadex G100 chromatography. He also found the amyloid was soluble in 88% formic acid for HPLC. In his first paper, he obtained an N-terminal sequence as far as residue 24, with a mistake at residue 11 (identified Gln instead of Glu). He named contents of the two G100 peaks "p1, p2peptide" after the "p-pleated sheet" configuration determined by X-ray crystallography (in contravention to the International Amyloidosis Nomenclature Committee rules, which required the A-"x" system). He predicted that the p1/p2 peptides would be derived from a unique serum protein precursor.
Masters first saw this paper when travelling to an EMBO-sponsored meeting on the Transmissible Spongiform Encephalopathies being organized by Alan Dickinson in Edinburgh. At this meeting, Konrad Beyreuther and Stan Prusiner were present: Beyreuther attended because of his association with Heino Diringer who had interested him in some of the properties of scrapie fibrils isolated in Diringer's Berlin laboratories; Prusiner was there with some important unpublished information on the N-terminal sequence of PrP. Masters approached Beyreuther, known for his expertise in amino acid sequencing, to help with his studies on the AD amyloid plaque cores. By that time, Masters had also determined their solubility in strong chaotropes, such as guanidine, and had discovered formic acid to be the most effective solvent (a tip derived from the previous generation of Australian wool protein chemists). Beyreuther readily agreed to collaborate, and Masters sent purified AD-APC to him and Gerd Multhaup for sequencing at the Institute for Genetics, Cologne. Our method for the APC purification now consisted of a pepsin digestion, Triton X100/high salt extraction followed by separation on a discontinuous sucrose gradient. Beyreuther and Multhaup were able to solubilize the AD-APC in formic acid and obtain a very ragged N-terminal sequence as far as residue 28 (four more than Glenner!). On SDS gels, dimers and higher order aggregates were readily observed of the 4kD monomer, which at the time, in conformance with the International Nomenclature rules, we referred to as "A4" (and the oligomers as A8,A16,A64, etc, the "A" standing for either Amyloid or Alzheimer); see Fig. 1). We noted the pH-dependence of this aggregation process as being typical of protonation of histidines.
Glenner's next paper appeared while we were making rapid progress with our own analyses. He now referred to the "p1/p2peptide" as "the p protein," corrected
Fig. 1. Image of gel of native APC Ap from Masters et al. (1985a) showing oligomers of Ap (lane 2) of mass 8kd (As, dimer) and 16kd (A16, tetramer?). These oligomeric species of Ap have subsequently become the prime suspect in the quest for the "toxic species"
his sequencing error at residue 11, and predicted that, since the amyloid N-terminal sequence from a vascular preparation from a case of Down's syndrome was the same as from AD, there would be a gene defect on chromosome 21 responsible for AD.
By late 1984, we had assembled enough data from our APC studies to draft a manuscript that was submitted to PNAS, and accepted in January 1985. In the acknowledgments, we thanked Steve Bobin, Michael Landon and George Glenner for "helpful discussions." This statement was certainly true for Bobin and Landon, with whom we had developed cordial relationships. Glenner, however, maintained a very "stand-offish" attitude and even had the presumption to request further supplies of our purified APC (see Fig. 2 - "1 mg would be fine"!). Our PNAS paper was published in June 1985. Subsequent discussions with Glenner showed that he believed that we could never have obtained our results without reference to his 1-24 sequence. For many years thereafter, he maintained that the basic amyloid subunit was 28 residues in length. Initially, we ourselves were uncertain whether the N-terminal raggedness
I diversity of California, San Diego (M-012), La .lolla. CA 921M3
BIQCHEM1CAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
fi^Z Pages 1131-1135
ALZHEIMERN DISEASE AM) DOWN'S SYNDROME: \jJm\RJNG OF A l'NIQUK CEREBROVASCULAR AMYLOID FIBRIL PROTEIN
I diversity of California, San Diego (M-012), La .lolla. CA 921M3
Received Jurte 26, 1984
Hl MM A Ii V: Tin- cerebrovascular dinyitdd protein from a '•'/■.;>[ ,iI.!i Down \s iyndtonic was isolated and purified, Amino acid sequence analysis showed il. to be homologous to that of the 6 protein of Alzheimer's disease. This is the first chemical evidence of a relationship between Down's syndrome anil Alzheimer's diseasct. It suggests that Down's syndrome may be a predict,, able model fi>r Alzheimer's disease. Assuming the 'J protein is a human gene product, il als« suggests that the genetic defect in Alzheimer's disease is localized on chromosome 21.
Fig. 2. Image of a reprint sent to CLM from George Glenner, at some time in late 1984, with the inscription "1 mg would be fine," referring to the collaboration in which Masters had previously sent him samples of purified AD-APC for X-ray diffraction studies. Glenner was now requesting further supplies. No results ever came from this collaboration was an artefact of the preparative method or caused by non-specific degradation of material remaining in situ for extended periods.
The most important question that needed to be addressed in early 1985 was the origin of the cerebral amyloid. We immediately set out to raise antisera to the purified and fractionated APC and to a variety of synthetic peptides of A4 (the Ap peptide, as it subsequently became known). Using these antisera on AD brain sections, we were privileged to be the first to see the full extent of amyloid deposition in the human AD brain - a major revelation to the eyes of a classically trained neuropathologist! The Nottingham group had drawn attention to the similarity in amino acid composition between APC and NFT preparations, and we were very surprised to find similar (but more ragged) N-terminal sequences from our own NFT preparations. Even more surprising, some antisera raised to both native and synthetic APC/A4 reacted with a subpopulation of NFT in situ. All of the antisera that reacted with APC also strongly reacted with the vascular amyloid. We also observed that the APC might have a non-proteinaceous component. From these observations we made several bold predictions, including that the A4(Ap ) subunit would be of neuronal origin, would consist of about 40 residues, and would be derived from a precursor protein. The concept of a neuronal origin of an intracerebral amyloidogenic protein (diametrically opposed to the prevailing views of Glenner, Frangione and Wisniewski) received further support from the studies of Ghiso and Frangione, who showed that a neuronally derived protein, cystatin C, was the cause of a rare Icelandic congophilic angiopathy. But the more compelling evidence for the neuronal origin of the AD-APC/ACA was to come eventually from the cloning of the Ap precursor protein (APP) itself (Kang et al. 1987). Once we had determined the N-terminal sequence of the AD amyloid, it was clear that the major challenge ahead was to use this information to derive a cDNA clone to uncover the precursor protein. This came to fruition in the second half of 1986 (Kang et al. 1987), when the APP gene was sequenced, disclosing the proteolytic origin of Ap. Our studies at that time indicated that the C-terminus of Ap was around positions 42/43. The neuronal origin of the Ap was also confirmed through studies showing high levels of APP mRNA expression in the brain.
Our observations in 1985-1986 that the AD brain is under severe oxidative stress (Martins et al. 1986) were the first to suggest that the accumulation of Ap in the AD brain might cause damage through some redox-active chemistry. This is currently one of our major strategies directed at therapeutic interventions in AD, in which we have increasing evidence that the Ap fragment itself is driving the oxidative stress through metal-catalyzed oxidation. In some sense this closes a loop of investigation that has occupied us over the past several decades.
Was this article helpful?
Curb Sugar Cravings Once And For All With These Powerful Techniques. Sugar sensitive people might be low in specific neurochemicals that help us feel calm, centered, confident, and optimistic. Sugar is a drug that temporarily makes the sugar sensitive feel better, but with damaging consequences.