Like good cholesterol and bad cholesterol, the Alzheimer's community has realized that there may also be good Aβ and bad Aβ. What if an antibody was developed to target only the latter? A tall order, yet scientists from the University of California, Irvine, may have done just that. Reporting in yesterday’s Science, Charles Glabe and colleagues describe a remarkable antibody that only recognizes and binds to soluble oligomers of Aβ. This is significant because not only are these oligomers the building blocks of fibrillar Aβ, but they are also more neurotoxic (see ARF related news story).

First author Rakez Kayed and coworkers raised the antibody by vaccinating rabbits with synthetic molecules designed to mimic the structure of soluble Aβ oligomers. When the authors tested the antibody, they found it reacted with the oligomers but failed to recognize soluble monomers or insoluble fibrillar forms of the peptide. In an astonishing twist, Kayed found that the antibody also recognized soluble oligomers of other amyloid proteins, including α-synuclein, human insulin, islet amyloid polypeptide (IAPP), and prion peptide 106-126. This suggests that oligomers of all these unrelated proteins must have a common structural surface based on their peptide bond backbone and quite irrespective of their particular amino acid side chains.

Kayed and colleagues next asked if the antibody could prevent the toxic effects of amyloid on cultured neurons. When the latter were incubated with soluble oligomers of Aβ40 or Aβ42, only about 20 percent of cells remained viable, yet including the oligomer antibody in the cell culture increased this number to 80 percent. The antibody was similarly able to rescue cells from the toxic effects of other soluble oligomers, including insulin, IAPP, lysozyme, and polyQ proteins. One implication is that one antibody might be useful against a variety of amyloid disorders.

In AD brain samples, the antibody stained clusters of Aβ deposits distinct from those fibrillar deposits detected by thioflavin-S. Furthermore, the antibody failed to detect any deposits in samples from dementia-free individuals, despite the presence of occasional Aβ plaques in these people’s brains. The antibody also reacts with brain lysates from people who have exhibited early signs of AD. Together, this leads the authors to suggest that soluble Aβ oligomers, also called Aβ-derived diffusible ligands (ADDLs), "may represent the initial stage of amyloid deposition."—Tom Fagan


  1. An initial read of the paper by Kayed et al. suggests that an antibody was developed that can see oligomeric forms of Aβ in vitro. What seems most interesting is that the antibody also appears to recognize oligomeric forms of many other proteins that aggregrate. This suggests a common structure to oligomers.

    This antibody should be very useful to test many questions in the future, and it will be interesting to see what effects it has in vivo. Data in Figure 3 of the paper suggests that the antibody stains areas around plaques, but presumably not fibrillar Aβ, in human brain. It appears that it only stains areas that are in some way in the vicinity of plaques.

    An important future question to address is whether oligomeric forms of Aβ only occur in association with aggregated forms such as fibrils. Some have speculated that they occur in brains well before or independently of the process of fibril formation. The reagent in this paper should allow this issue to be addressed. This study appears to suggest that the process of oligomer and fibril formation are linked, as they occur in the same regions in brain.

    Also, use of this antibody should be useful in the future to see if "oligomeric" forms of Aβ are present in physiological solutions (e.g., blood or CSF).

  2. Over the past decade, credible evidence has gradually accrued in support of the idea that small, prefibrillar forms of amyloidogenic proteins (oligomers) may be key cytotoxic agents in Alzheimer's disease and other proteopathies. Several suggestions have been made as to which oligomeric species is most culpable, and how oligomers produce their toxic effects, but we lacked a ready means of directly demonstrating the presence of these elusive molecules in afflicted organs. Kayed, Glabe, and colleagues have developed a remarkably specific antibody, produced in rabbits by immunization with a molecular mimic of oligomeric Aβ, that recognizes oligomers within a specific size range. This oligomer-specific antibody is able to detect accumulations of these molecules even in tissue sections from the AD brain. Notably, the antibody binds (and blocks the toxicity of) not only oligomers of Aβ (and not that of fibrillar Aβ), but also those formed from amyloidogenic proteins with diverse amino acid sequences.

    This study demonstrates once again the exquisitely fine-tuned ability of antibodies to discriminate among determinants formed of primary, secondary, tertiary, and quaternary structural elements of proteins. The narrowness or breadth of the specificity can be adjusted with judicious screening.

    This new means of scrutinizing (and impeding) pathogenic forms of protein has several potential applications. The therapeutic implications are clear. If oligomeric toxicity is central to a variety of age-associated proteopathies (and this appears to be so), the presence of the right antibodies could reduce the risk of developing these disorders and enhance the overall quality of old age.

    Experimentally, the oligomer-specific antibody represents a new type of tool for illuminating the pathogenicity of disease-related proteins. For example, it is now possible to get a preliminary purchase on the comparative pathobiology of oligomerization. Both aged nonhuman primates and AβPP-transgenic mice are surprisingly refractory to Aβ-induced neuronal loss, despite the accumulation of large quantities of Aβ in brain; is this because oligomers fail to reach toxic levels in the brains of these species, and if so, why? It would also be informative to know if this antibody reacts with any normal proteins, or more intriguingly, are there formerly unrecognized, misfolded oligomeric proteins that appear with aging or insult?

    Kayed et al. have taken an important step toward demystifying the link between abnormal protein assembly and Alzheimer's disease. The findings also strengthen the concept that many degenerative diseases share important pathogenic features that are not necessarily tied to amyloid itself. Finally, the data add weight to the argument that scientists and clinicians seeking new treatments for Alzheimer's disease should be conversing, in earnest, with those studying seemingly disparate disorders such as Parkinson's disease, mad cow disease and type 2 diabetes, to name only a few.—Lary Walker, Pfizer, Ann Arbor, Michigan; Harry LeVine, University of Kentucky, Lexington, Kentucky.

  3. The paper by Kayed et al. describes the development, characterization, and use of a reagent that promises to be of great utility in deciphering the role of soluble amyloid oligomers in Alzheimer’s disease and a host of other diseases involving protein aggregation. Based on prior experimental evidence suggesting that soluble oligomeric Aβ exist as protein micelles (Soreghan et al., 1994), the authors generated a molecular mimic in which the C-terminus of Aβ1-40 was covalently linked to colloidal gold particles. The mimic displayed many of the physical properties of synthetic Aβ oligomers, but was significantly more stable and therefore useful as an antigen. Antibodies (referred to as anti-oligo) raised to this antigen specifically recognized Aβ oligomers, but not fibrils or monomer.

    Temporal analysis of in-vitro aggregation of Aβ1-40 and 1-42 by EM and dot blot with anti-oligo revealed that the appearance of ADDLs (Lambert et al., 1998) and protofibrils (Harper et al., 1997; Walsh et al., 1997) were coincident with anti-oligo immunoreactivity, and indicate that both ADDLs and protofibrils share a common structural epitope.

    Amazingly, anti-oligo also specifically detects soluble oligomeric aggregates formed by α-synuclein, islet amyloid polypeptide, polyglutamine, lysozyme, insulin, and prion protein 106-126. As with Aβ, anti-oligo did not detect the monomeric or fibrillar versions of these proteins. Thus, anti-oligo recognizes a common structural epitope independent of primary sequence.

    Importantly, anti-oligo prevented oligomer-mediated toxicity of all the proteins tested, whereas antibodies not specific for oligomers had no effect. However, the finding that the monoclonal antibody 6E10 can bind oligomers (although not specifically, Fig. 1A), but cannot attenuate oligomer-mediated toxicity (Fig. 2D) is unexpected, particularly since monoclonal antibodies generated using conventional means can rapidly reverse memory deficits in AβPP-transgenic mice by the presumed targeting of nonfibrillar soluble Aβ oligomers, e.g., Dodart et al., 2002 (see ARF related news story). This also raises a question about the role of low n-oligomers. Apparently anti-oligo apparently does not detect oligomers smaller than octamers (Fig. 1D). Are these the Aβ species that underlie reversible effects on memory and learning (Dodart et al., 2002, Walsh et al., 2002, see ARF related news story), whereas larger oligomers may be responsible for the irreversible loss of cells?

    Whatever the effects of low n-oligomers, the observation that toxicity mediated by ADDLs and/or protofibrils can be ameliorated by anti-oligo suggests that their common structure may mediate toxicity by a common mechanism, and it offers a unique target for therapeutic intervention. This may enable further optimization of vaccination strategies, but in addition it suggests the possibility of designing small molecules that specifically bind to and disrupt the oligomer-specific conformation. Development of small-molecule disrupters would be particularly pertinent for use in diseases involving intracellular aggregates that are inaccessible to antibodies.

    As with many significant advances, the work reported by Charlie Glabe’s group raises as many questions as it answers. For example, will the anti-oligo ELISA prove a useful diagnostic tool? The anti-oligo ELISA could lend itself to the inexpensive screening of thousands of samples and, in addition, might prove useful for the detection of other amyloid-related diseases. In this regard, the detection of oligomers in brain homogenates from patients with AD and mild Braak changes, but not in nondemented controls, is encouraging (see supporting online material), and provides the first evidence for the in-vivo relevance of protofibrils. Clearly, the availability of this unique antibody will facilitate a fuller understanding of the role of soluble oligomers in AD and other protein aggregation diseases. Exciting times!


    . Surfactant properties of Alzheimer's A beta peptides and the mechanism of amyloid aggregation. J Biol Chem. 1994 Nov 18;269(46):28551-4. PubMed.

    . Diffusible, nonfibrillar ligands derived from Abeta1-42 are potent central nervous system neurotoxins. Proc Natl Acad Sci U S A. 1998 May 26;95(11):6448-53. PubMed.

    . Atomic force microscopic imaging of seeded fibril formation and fibril branching by the Alzheimer's disease amyloid-beta protein. Chem Biol. 1997 Dec;4(12):951-9. PubMed.

    . Amyloid beta-protein fibrillogenesis. Detection of a protofibrillar intermediate. J Biol Chem. 1997 Aug 29;272(35):22364-72. PubMed.

    . Immunization reverses memory deficits without reducing brain Abeta burden in Alzheimer's disease model. Nat Neurosci. 2002 May;5(5):452-7. PubMed.

    . Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature. 2002 Apr 4;416(6880):535-9. PubMed.


    Please see my commentary on this article at Science SAGE KE (1 May 2003) [ Full Text ].

Make a Comment

To make a comment you must login or register.


Other Citations

  1. ARF related news story

Further Reading

No Available Further Reading

Primary Papers

  1. . Common structure of soluble amyloid oligomers implies common mechanism of pathogenesis. Science. 2003 Apr 18;300(5618):486-9. PubMed.