Does β amyloid accumulation drive Alzheimer’s disease? In favor of this idea, numerous known mutations in amyloid precursor protein step up Aβ production and lead to early onset AD. Critics of the theory contend, however, that familial AD may be different, and amyloid merely a marker or side effect in late-onset disease. Now, in a July 11 Nature paper, scientists announce the discovery of a protective APP mutation. Not only does the rare allele reduce sporadic AD risk, it also slows the rate of cognitive decline in an elderly population, report researchers led by Kári Stefánsson at deCODE Genetics, Reykjavik, Iceland, and Ryan Watts at Genentech, South San Francisco, California. The scientists found that β-secretase (BACE1) struggles to cut APP containing the variant, resulting in less production of toxic Aβ fragments. A story in The New York Times hailed the finding’s importance.

Scientists in the field greeted the paper with enthusiasm. “This validates and confirms the β amyloid hypothesis,” said Rudy Tanzi at Massachusetts General Hospital, Boston, adding, “The genetic data look solid.” The findings should encourage companies developing amyloid-targeting drugs, commentators agreed. “It shows that if you reduce Aβ, you can prevent AD, and that is very much a proof of concept for the drug industry,” said Gerard Schellenberg at the University of Pennsylvania, Philadelphia. He pointed out that up until now, the data linking Aβ levels to AD risk come from early onset cases. “This is the first data that says that Aβ levels are important for late-onset AD,” he said. Importantly, the finding strengthens ties between the two forms of the disease, implying that results from planned trials in familial AD populations (see ARF related news story; ARF news story; ARF news story) may apply to sporadic AD, too, Schellenberg added.

Scientists know of about two dozen APP mutations that hasten AD onset, including such famous players as the Swedish and London mutations, but had never turned up a protective allele. To search for one, first author Thorlakur Jonsson catalogued APP coding variations in whole-genome data from almost 1,800 Icelanders. Once the authors determined haplotypes associated with each variation, they predicted the APP sequence in some 370,000 additional Icelanders by matching their haplotypes. The researchers then looked for associations between coding variants and AD cases in this large population. They found a rare allele, present in about 1 percent of Icelanders, that is five times more common in non-demented elderly than in those with AD. The single-nucleotide polymorphism results in the substitution of a threonine for an alanine at position 673 of APP (A673T). The protective allele occurs at a similar frequency in several Scandinavian populations, but is much rarer in North America, found in only about one in 5,000 people, the authors report.

The protective coding variant is located only two amino acids away from the BACE1 cleavage site, suggesting it might defend against cutting. To examine this, the Genentech group measured levels of BACE1 cleavage products in a human cell line transfected with either wild-type or A673T APP. Cells that received the protective variant churned out about 40 percent less Aβ40 and Aβ42 than did cells containing wild-type APP. Likewise, in a purified enzyme assay, the secretase snipped wild-type APP twice as quickly as it did the A673T variant. Intriguingly, several pathogenic mutations also occur near the BACE1 cleavage site. A valine substitution at position 673 increases AD risk (see ARF related news story; Giaccone et al., 2010), and the Swedish mutations lie just on the other side of the cleavage site. All three of these alleles pumped up production of Aβ40 and Aβ42 in the cellular assay, the authors found. Watts is currently examining crystal structures of BACE1 and APP to determine how the amino acid substitutions alter cutting.

In theory, the finding provides a proof of principle for pharmacologic BACE1 inhibitor programs, Stefánsson said. However, Watts cautioned against overinterpreting the data, noting that people who carry one protective allele have produced about 20 percent less Aβ overall for their entire lifespan. “It doesn’t tell you that BACE inhibition is going to work in the elderly,” he noted. Alison Goate at Washington University, St. Louis, Missouri, agreed that timing is the key issue, asking, “How early are we going to have to reduce Aβ for [a therapy] to be effective?” Tanzi pointed out that BACE1 has many other substrates (see, e.g., ARF related news story; ARF related news story; ARF related news story), and suggested an inhibitor might need to be selective for APP in order to be safe enough for such long-term use. John Hardy at University College London, U.K., noted the data imply that a modest lifelong reduction in Aβ production alone is not harmful.

To look at the allele’s effect on cognition, the authors examined cognitive testing data from roughly 3,700 non-demented nursing home residents between 80 and 100 years of age. The 41 residents who carried an A673T allele had better cognitive scores at all ages than did their non-carrier peers, although cognition in both groups declined over time. “I think these data show that Aβ production is a major contributing factor to age-related cognitive decline,” Watts said, pointing out that amyloid pathology is widespread at these ages, even in the absence of overt clinical symptoms, and may overshadow other sources of cognitive decline. Goate agreed, suggesting that the steeper cognitive decline in the non-carriers is probably due to preclinical AD being more common in this group.

In future work, Stefánsson will tackle the question of how early researchers should start lowering Aβ levels, he told Alzforum. He plans to test the cognitive functions of carriers and non-carriers at younger ages, looking for the point where the two groups begin to separate. Treatment should be started no later than that, Stefánsson suggested. One intriguing implication of this work is that most elderly people might benefit from Aβ reduction, given how common AD pathology is in older brains, Stefánsson said. He added, “I think it is possible that all of us are destined to develop AD if we live long enough.”—Madolyn Bowman Rogers


  1. This elegant and important study finds an association between a rare APP variant and a lower risk of AD, and it provides strong evidence to suggest that the protective effects of this variant may be attributable to reductions in BACE1-mediated APP cleavage. This study provides additional support for the amyloid hypothesis and the potential role of BACE1 inhibitors in the preclinical treatment of AD.
    It also illustrates the potential value of whole-genome sequencing studies, when used in conjunction with relevant basic scientific research, to advance the understanding of AD and the discovery of promising investigational treatments.

    This study only adds to the interest that my Alzheimer's Prevention Initiative (API) colleagues, other researchers, and I have in the possibility of evaluating suitable BACE1 inhibitors (and related
    agents) in persons at increased risk of developing AD, and our interest in learning more about safety, tolerability, and certain other effects in clinically affected patients. We see particular value in the possibility of evaluating anti-amyloid production treatments like these in ApoE4 carriers and young adult PS1 mutation carriers, since these agents might have the best chance to exert a profound benefit if started before the accumulation of significant fibrillar amyloid.

  2. I agree with the comments underlying the importance of this paper, which supports the pathogenic role of APP processing by BACE in sporadic AD. As noted in many comments, these findings also suggest that inhibition of BACE processing of APP will be a beneficial therapeutic approach for both familial and sporadic AD. Perhaps, this treatment may also improve performance in elderly with no obvious cognitive deficiency. BACE1 inhibitors may, however, have toxic effects related to the important biological functions of BACE processing of other substrates. Therefore, as noted by Dr. Tanzi, "inhibitors might need to be selective for APP in order to be safe enough for such long-term use." Interestingly, a molecule with these features has been recently described—MoBA, Modulator of Β-processing of APP (Tamayev et al., 2012), and may represent a leading compound to develop drugs that interfere with BACE1 processing of APP without inhibiting the proteolytic activity of BACE1 on the other substrates.

    However, contrary to what is said by many, I do not think that the finding validates or confirms the amyloid cascade hypothesis. Reduction of β-processing of APP will first result in a reduction of sAPPβ and β-CTFs. Therefore, this finding is also consistent with alternative hypotheses of AD pathogenesis which point to sAPPβ and/or β-CTFs, and not Aβ, as the main pathogenic APP-derived metabolites (Tamayev et al., 2012; Nikolaev et al., 2009; Tamayev and D'Adamio, 2012).


    . β- but not γ-secretase proteolysis of APP causes synaptic and memory deficits in a mouse model of dementia. EMBO Mol Med. 2012 Mar;4(3):171-9. PubMed.

    . APP binds DR6 to trigger axon pruning and neuron death via distinct caspases. Nature. 2009 Feb 19;457(7232):981-9. PubMed. RETRACTED

    . Inhibition of gamma-secretase worsens memory deficits in a genetically congruous mouse model of Danish dementia. Mol Neurodegener. 2012 Apr 26;7(1):19. PubMed.

  3. New APP Mutation Does Not Advance the Amyloid Hypothesis Debate
    In response to the question, Does β amyloid accumulation drive Alzheimer’s disease?, the answer is “most likely no,” and the new data presented by Jonsson et al. does not change that answer.

    That the amyloid hypothesis inadequately accounts for the current data has been known for some time. It lacks a theoretical foundation from which the physiological generation of Aβ can be understood, and therapeutic approaches based on its premises have all failed. Furthermore, there is no significant correlation between Aβ accumulation and cognitive deterioration in either humans or in mouse models, and Aβ-containing senile plaques have been found in the brains of approximately 30 percent of individuals with no signs of dementia (Crystal et al., 1988; Price et al., 2009).

    This does not necessarily indicate that Aβ is not a key factor in AD. The weakness of the amyloid hypothesis is not that it links Aβ and AD, but rather that it places Aβ as the key pathogenic trigger of the disease and, accordingly, ought to be the main target for therapeutic purposes. To date, all the available evidence—including clinical drug trials—indicates that this is an increasingly unlikely scenario.

    Rather, the hypothesis that best fits the current body of evidence in the AD field, including the work by Jonsson et al., is that Aβ is a key element of the brain’s adaptive response to stress. The idea of an adaptive response to stress in the brain, in the context of both aging and AD pathogenesis, has been put forward in various forms by scientists including George Perry, Mark Smith, Karl Herrup, and others (Stranahan et al., 2011; Stranahan and Mattson, 2012; Nunomura et al., 2001; Pappolla et al., 2002; Castellani et al., 2009; Herrup, 2010; Castello and Soriano, 2012). The idea is that failure of this adaptive response, or its chronic activation, is what leads to sporadic AD, instead of overproduction of Aβ itself. The nature of that stress in the brain is wide ranging, encompassing microglia activation, accumulation of reactive oxygen species, and cholesterol dysregulation, and the adaptive response it elicits involves regulation of APP through its cleavage products, including Aβ and sAPPα (Castello and Soriano, 2012; Stranahan and Mattson, 2012; Castellani et al., 2009).

    In the case of the A673T mutation, all other factors being comparable, the combination of lesser Aβ levels, changes in APP-driven cholesterol regulation, and/or higher levels of sAPPα would all result in a more efficient adaptive response, accounting for the delay in AD age of onset described by Jonsson et al. (Castello and Soriano, 2012). Note also that the A673T mutation, creating a stronger adaptive response, is also consistent with the superior performance in cognitive tests of patients with the mutation. Mutation carriers retain cognitive ability for a longer period of time even when they do develop AD, as is visually demonstrated by the supplementary figures in Jonsson et al.

    Overall, the message is that we should not rush back into the race to find the best BACE inhibitors to fight AD, as many commentators and scientists have suggested here and in the media. The line of inquiry involving BACE biology in AD has been in progress for over a decade, with negative outcomes, recent examples of which appear in the sidebar next to this very article (ARF related news story; ARF news story).

    We need to reassess the way we approach research in the AD field. Instead of continuing a never-ending search for the elusive final proof that the amyloid hypothesis is correct, we should begin first by restructuring AD diagnosis to focus on symptomatic changes rather than the neuropathological identification of amyloid plaques, as has been suggested (Dubois et al., 2010). The pathological process of AD suggests variations of the same stress response. Therefore, our research should focus on the primary issue—the sources of stress that create a need for the adaptive response involving APP and Aβ.


    . Oxidative damage is the earliest event in Alzheimer disease. J Neuropathol Exp Neurol. 2001 Aug;60(8):759-67. PubMed.

    . Revising the definition of Alzheimer's disease: a new lexicon. Lancet Neurol. 2010 Nov;9(11):1118-27. PubMed.

    . Reimagining Alzheimer's disease--an age-based hypothesis. J Neurosci. 2010 Dec 15;30(50):16755-62. PubMed.

    . Neuropathology of nondemented aging: presumptive evidence for preclinical Alzheimer disease. Neurobiol Aging. 2009 Jul;30(7):1026-36. PubMed.

    . Diet-induced elevations in serum cholesterol are associated with alterations in hippocampal lipid metabolism and increased oxidative stress. J Neurochem. 2011 Aug;118(4):611-5. PubMed.

    . Recruiting adaptive cellular stress responses for successful brain ageing. Nat Rev Neurosci. 2012 Mar;13(3):209-16. PubMed.

    . Reexamining Alzheimer's disease: evidence for a protective role for amyloid-beta protein precursor and amyloid-beta. J Alzheimers Dis. 2009;18(2):447-52. PubMed.

    . Cholesterol, oxidative stress, and Alzheimer's disease: expanding the horizons of pathogenesis. Free Radic Biol Med. 2002 Jul 15;33(2):173-81. PubMed.

    . Clinico-pathologic studies in dementia: nondemented subjects with pathologically confirmed Alzheimer's disease. Neurology. 1988 Nov;38(11):1682-7. PubMed.

  4. This is a nice finding by Jonsson et al., and adds more evidence to the idea published in 2009 by an Italian group that changes to this amino acid position in APP can prevent or facilitate aggregation of human Aβ.

    It seems that amino acids at position 2 of hAβ with larger side chains than alanine may prevent aggregation. Di Fede showed that a transition, A2V, in Aβ was protective in heterozygotes, whereas homozygotic individuals had a higher risk in the studied family tree.

    They also presented aggregation profiles of pure and mixed Aβ species that nicely supported the aggregation idea. Thus, the question remains whether it is indeed a BACE-related mechanism or just an aggregation problem of these Aβ species due to stereology. In our hands, cerebral injection and chronic infusion of A2V Aβ1-6 peptides reduced plaque formation in mice. That could be exploited as a treatment option (unpublished data).

    As we know, mouse Aβ is different in three amino acids at the N-terminal, and these differences completely prevent plaque formation and fibril generation in mice.


    . A recessive mutation in the APP gene with dominant-negative effect on amyloidogenesis. Science. 2009 Mar 13;323(5920):1473-7. PubMed.

    . Neuropathology of the recessive A673V APP mutation: Alzheimer disease with distinctive features. Acta Neuropathol. 2010 Dec;120(6):803-12. PubMed.

  5. This study provides elegant genetic evidence that reducing the BACE processing of APP protects against late-onset AD. The finding complements the long-standing observations that increased BACE processing (APPswe mutation or elevated BACE levels [1]) enhances the risk for AD.

    What gets ignored in ensuing discussions, however, is the fact that BACE processing of APP generates not only Aβ peptides, but also β-CTFs and AICD peptides (studies from three groups show that BACE processing enhances AICD generation and signaling [2-4]). Since both β-CTF and AICD cause AD-like pathological features in vivo in mouse models (5,6), the present study is consistent with, but cannot be claimed to support, the amyloid hypothesis (the causal role of Aβ in AD).

    Nonetheless, this study is important because it identifies a unique human population (A673T carriers) that can be used to further validate neuroimaging (PIB imaging) and CSF biomarkers. Indeed, it will be very informative to see whether biomarker changes in this population appear at the same time as in the non-carriers or are significantly delayed.


    . Beta-site amyloid precursor protein cleaving enzyme 1 levels become elevated in neurons around amyloid plaques: implications for Alzheimer's disease pathogenesis. J Neurosci. 2007 Apr 4;27(14):3639-49. PubMed.

    . Nuclear signaling by the APP intracellular domain occurs predominantly through the amyloidogenic processing pathway. J Cell Sci. 2009 Oct 15;122(Pt 20):3703-14. PubMed.

    . The transcriptionally active amyloid precursor protein (APP) intracellular domain is preferentially produced from the 695 isoform of APP in a {beta}-secretase-dependent pathway. J Biol Chem. 2010 Dec 31;285(53):41443-54. PubMed.

    . Dual role of alpha-secretase cleavage in the regulation of gamma-secretase activity for amyloid production. J Biol Chem. 2010 Oct 15;285(42):32549-56. PubMed.

    . Impairments in learning and memory accompanied by neurodegeneration in mice transgenic for the carboxyl-terminus of the amyloid precursor protein. Brain Res Mol Brain Res. 1999 Mar 20;66(1-2):150-62. PubMed.

    . Alzheimer's disease-like pathological features in transgenic mice expressing the APP intracellular domain. Proc Natl Acad Sci U S A. 2009 Oct 27;106(43):18367-72. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. NIH Director Announces $100M Prevention Trial of Genentech Antibody
  2. DIAN Grows, Gets Ready for Therapeutic Trials
  3. Anti-Amyloid Treatment in Asymptomatic AD Trial
  4. Good Gene, Bad Gene?—New APP Variant May Be Both
  5. BACE Secrets: Newly Identified Substrates May Regulate Plasticity
  6. BACE Knockouts Show More Synaptic Troubles
  7. BACE Regulates Sodium Channels, Neuronal Excitability

Paper Citations

  1. . Neuropathology of the recessive A673V APP mutation: Alzheimer disease with distinctive features. Acta Neuropathol. 2010 Dec;120(6):803-12. PubMed.

External Citations

  1. The New York Times

Further Reading

Primary Papers

  1. . A mutation in APP protects against Alzheimer's disease and age-related cognitive decline. Nature. 2012 Aug 2;488(7409):96-9. PubMed.