Feeling good on the outside often helps us feel good on the inside. Could the same be true for neurons? In today’s issue of the journal Neuron, Frank LaFerla and colleagues at the University of California, Irvine, report that antibodies to amyloid-β not only clear up extracellular amyloid plaques, but they promote clearance of intracellular neurofibrillary tangles too. The results should be music to the ears of vaccine developers, and may boost the flagging amyloid cascade hypothesis.

First author Salvatore Oddo and colleagues discovered the effect when they passively immunized triple transgenic (3xTg-AD) mice with antibodies to Aβ. Developed by LaFerla and colleagues, 3xTg-AD mice harbor presenilin, APP, and tau mutants, and recapitulate the two classic hallmarks of Alzheimer’s disease (AD), amyloid plaques and neurofibrillary tangles (see Oddo et al., 2003 and ARF related news story for descriptions of this mouse model).

Oddo immunized the animals by injecting antibodies into one side of the hippocampus (antibodies used were commercially available monoclonals to human Aβ). Examining the mice seven days later, the authors found that the ipsilateral side showed a marked reduction in the number of plaques, as judged by immunoreactivity and thioflavin S staining. In contrast, there was no improvement on the side of the hippocampus that received no antibody.

Plaque clearance by passive immunization is not big news in and of itself (see, for example, a recent roundup of various vaccination strategies and their outcomes), but what is remarkable about this study is that the authors found that intracellular aggregates of tau, the major component of neurofibrillary tangles, had also disappeared. This loss was also restricted to the treated side of the hippocampus.

The reduction of both Aβ and tau aggregates suggests a “direct relationship between the development and expression of these two neuropathological lesions,” write the authors. And to confirm this relationship, Oddo and colleagues looked at the time course of events following immunization. If their hypothesized relationship held true, then Aβ clearance should precede tau clearance, and indeed, this is what the authors found. Three days after immunization, only the amyloid plaques had cleared; the tau aggregates did not begin to disappear until two days later.

How does passive immunization with anti-Aβ lead to clearance of intracellular tau? Some clues may come from the clearance of Aβ itself. Oddo and colleagues found that it is not only extracellular Aβ but also the intracellular form that is cleared—perhaps because it diffuses out of neurons to maintain an intra/extracellular equilibrium. Could reducing the amount of intracellular Aβ help prevent tau aggregates, and how?

LaFerla and colleagues explored the possibility that the proteasome may be involved because impairment of this crucial protein disposal has been implicated in AD (see the ARF milestone paper by Keller et al., 2000) and it has been reported that Aβ may inhibit subunits of the complex (see Gregori et al., 1997). When the authors injected mice with Aβ antibody and the proteasome inhibitor epoxomicin, amyloid plaque clearance occurred as before, but clearance of tau was almost nonexistent. This supports the theory that intracellular Aβ may inhibit the proteasome and prevent degradation of unwanted tau.

These results give a big boost to the vaccine development camp. One of the major criticisms of Aβ vaccination therapy is that it just tackles one of the effects of AD (extracellular plaques) but not the cause (internal neuronal dysfunction). But this paper suggests that a vaccination approach may do both. It also lends support to the amyloid cascade hypothesis. “These findings…provide strong supporting evidence for the amyloid cascade hypothesis, which stipulates the Aβ accumulation triggers the onset of AD and that tau hyperphosphorylation, subsequent neurofibrillary tangle formation, and cell death are downstream consequences of Aβ aggregation,” write the authors.

Whether passive immunization against Aβ would have such dramatic effects in humans remains to be seen. But a few caveats are worth keeping in mind. The authors found that tau clearance only occurs if the protein hasn’t become hyperphosphorylated. “Because the clearance of the tau pathology is dependent on its phosphorylation state, it indicates that Aβ immunotherapy late during the disease course may still effectively clear amyloid plaques, although it will be insufficient to impact the neurofibrillary pathology,” La Ferla and colleagues write. For this reason they suggest that it will be essential to test tau antibodies, particularly those recognizing phosphorylated forms, for the ability to clear tau.

Also, as with most passive immunizations, the effects are not permanent (see also comment below by Dave Morgan). About 30 days after vaccination, aggregates reemerge. Interestingly enough, it is the Aβ plaques that form first, followed by tau aggregates. Now there’s some good evidence for the cascade hypothesis.—Tom Fagan.

Oddo S, Billings L, Kesslak JP, Cribbs DH, LaFerla FM. Abeta immunotherapy leads to clearance of early, but not late, hyperphosphorylated tau aggregates via the proteasome. Neuron. 2004 August 5;43:321-332. Abstract


Make a Comment

To make a comment you must login or register.

Comments on this content

  1. There are two key features to this manuscript. The first is the obvious linkage between the presence of excess Aβ and the accumulation of early-stage phospho-tau variants. This is the first observation that reducing Aβ may slow the rate of tau filament formation (assuming early forms progress to late forms of hyperphosphorylated tau). It is consistent with the earlier reports that elevating Aβ in tau transgenic mice increased phospho-tau levels. If the same holds true in AD brain, it suggests that amyloid lowering therapies are likely have benefit in slowing AD progression.

    A second outcome of the study is it suggests that not only can Aβ be rapidly removed from the brain, but that it returns very rapidly, as well. What took 12 months to accumulate originally, returns within 45 days. This implies that in transgenic mice, the accumulated Aβ and early forms of phospho-tau pathology are likely in an equilibrium state, with the rate of production exceeding the rate of removal by some amount, and age contributing to an even greater disequilibrium between these processes. Modifying the steady-state conditions with antibodies (to increase removal) or secretase inhibitors (to reduce production) leads to shifts in the steady-state pool of these substances over days. Once the equilibrium conditions are restored after clearance of the antibody, the amyloid deposits reemerge at levels comparable to that at the time of the antibody injection. We observe a similar rapid recovery of Aβ deposits after removal by activating microglia with LPS injections into hippocampus (Herber et al., Experimental Neurology, in press).

    While the structural modifications of amyloid deposits in AD brain make such rapid removal unlikely to be complete, the autopsy reports from the suspended vaccine trial argue that some of the deposited Aβ is removable. These are exciting results linking the βAPtist and tauist theologies and argue that disrupting the pathway leading to neurodegeneration at any point is likely to be useful in treating AD.

  2. These are a series of beautifully done studies, which strongly support the Aβ cascade hypothesis. However, to fully understand the potential of Aβ reduction to alter the clinical course of Alzheimer’s disease, it would be important to know whether lowering Aβ improves cognitive function in APP/tau mice. Moreover, if it does, is the improvement in cognitive function transient or is it sustained? Finally, it would be interesting to identify the specific molecular form of Aβ responsible for inducing the accumulation of abnormally phosphorylated tau.

  3. While the data presented using antibodies is consistent with results obtained in other studies, I find the data using the gamma-secretase inhibitor DAPT doubtful.

    Oddo, et al., write "Using an alternative approach, we show that administration of the gamma-secretase inhibitor, DAPT, leads to similar results. Therefore, the finding that Abeta-based interventions successfully clear the tau pathology in these 3xTg-AD mice provides compelling evidence in support of the amyloid cascade hypothesis."

    This result is very hard to believe. How can an acute injection of DAPT cause Abeta plaque disappearance (figure 10 in the paper)?

    Linking the plaque disappearance from the use of antibodies or DAPT is stretching it. Chronic use of DAPT should prevent plaque formation and certainly not dissolve, as quickly as the authors report, Abeta plaques.

    There was a poster at the AD conference which would suggest the DAPT data is also probably dodgy, or not as simple as it is presented (see "Inhibition of beta-amyloid production and clearance of senile plaques in transgenic mice" by Joanna L. Jankowsky1, Jason Wen2, Hilda H. Slunt2, Victoria Gonzales2, Nancy A. Jenkins3, Neal G. Copeland3, David R. Borchelt2,
    Presentation Number: P2-038.)

    In that paper, the authors concluded that amyloid plaque formation can be quickly arrested by inhibiting A-beta production. However, mature amyloid plaques are not rapidly cleared, and more time (as much as several months), or additional manipulations, may be required to reverse this pathology.

    Otherwise, it is an excellent paper, as it clearly demonstrates that targeting Abeta seems the right thing to do (as opposed to going after tau) to find a cure for AD.

  4. The paper from the LaFerla lab is most interesting, and extremely well done, which I have mentioned in recent interviews to other news outlets. However, I do not think the data in this paper resolve controversies about the validity of the amyloid cascade hypothesis, although there are important lessons to be learned from these elegant studies. For example, the paper indicates that it is possible to remove Aβ deposits with anti-Aβ antibodies, as also shown by others, but that this has no effect on established or fully formed AD-like NFTs composed of hyperphosphorylated tau proteins. On the other hand somatodendritic tau is reduced by clearing Aβ deposits with anti-Aβ antibodies, presumably through the proteasome, according to this study, but the significance of somatodendritic tau is not yet understood since it is not a marker of or diagnostic lesion for any brain disorder, and it is uncertain if somatodendritic tau has any deleterious effects or behavioral consequences in the mice studied here or would have any such effects in human beings. Finally, these studies appear to confirm what has been learned from the four postmortem studies of Elan vaccine patients to date which is that Aβ immune therapy can clear Aβ deposits in AD patients, but, as in the mice here, fully formed NFTs do not appear to be affected by this immune therapy. Thus, it may be necessary to have multiple treatment approaches to AD including those that are Aβ-centric and tau-centric.

  5. The paper by Oddo et al published in Neuron 2004 is most interesting. Together with the earlier reports (Oddo et al., 2003a; Oddo et al., 2003b) the authors describe a “missing link” between Aβ-amyloid and Tau pathology in an animal model for Alzheimer's disease. It is particularly interesting that intraneuronal Aβ precedes Tau pathology and injection of anti-Aβ antibodies reverses intraneuronal Aβ “accumulation”. Upon single injection of anti-Aβ antibodies, intracellular Aβ was first cleared followed by somatodendritic Tau, however, accumulated (Gallyas stained) Tau pathology was not reversed. Such an animal model offers the possibility to answer fundamental questions in Alzheimer`s disease pathology.

    It would be important to know for instance whether intraneuronal Aβ, somatodendritic and hyperphosphorylated Tau impairs neuron function in this mouse model, and whether age-dependent synapse and neuron loss occurs. We and others found that intraneuronal Aβ per se is not sufficient for neuron loss. All APP transgenic models we have studied so far, elicit intraneuronal Aβ staining. A striking neuron loss was only seen in two different APP/PS1 transgenic mouse models with an enhanced Aβ42 to Aβ ratio (Blanchard et al., 2003; Casas et al., 2004; Schmitz et al., 2004). Interestingly, the neuron loss was not correlated with any obvious Tau pathology.

    Together with the observations from Oddo et al. it is important to note that intraneuronal Aβ triggers both neuron loss, Tau trafficking and phosphorylation. Clearance of plaques can be successfully achieved by immunotherapeutic approaches as shown by a number of reports. However, since there is no correlation between plaque load and neuron loss in APP/PS1 transgenic mice (Casas et al., 2004; Schmitz et al., 2004), the question is still open whether neuron loss can be prevented by active or passive immunization. Oddo et al (2004) have clearly shown that clearing the extracellular Aβ pool also reduces the intracellular Aβ pool. We and others assume that there is an equilibrium between these pools (and maybe between different intracellular pools). However, there is a lack of information on the nature of “aggregated” Aβ species (dimers, protofibrills…?) within neurons in transgenic mouse brain and the precise molecular events leading to neuron death in Alzheimer's disease. While immunotherapy against Aβ removes intra- and extracellular Aβ, the next step needs to be shown whether it helps to stop neuron death.

    Blanchard et al. (2003). Time sequence of maturation of dystrophic neurites associated with Abeta deposits in APP/PS1 transgenic mice. Exp Neurol 184, 247-263. Abstract

    Casas et al. (2004). Massive CA1/2 neuronal loss with intraneuronal and N-terminal truncated Aβ42 accumulation in a novel Alzheimer transgenic model. Am. J. Pathol. in press.

    Oddo et al. (2004). Abeta Immunotherapy Leads to Clearance of Early, but Not Late, Hyperphosphorylated Tau Aggregates via the Proteasome. Neuron 43, 321-332. Abstract

    Oddo et al. (2003a). Amyloid deposition precedes tangle formation in a triple transgenic model of Alzheimer's disease. Neurobiol Aging 24, 1063-1070. Abstract

    Oddo et al. (2003b). Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron 39, 409-421. Abstract

    Schmitz et al. (2004). Hippocampal neuron loss exceeds amyloid plaque load in a transgenic mouse model of Alzheimer's disease. Am J Pathol 164, 1495-1502. Abstract

  6. The paper by Oddo et al, published in Neuron 2004, reaffirmed my belief that Frank LaFerla and his group at UC Irvine will be (if they are not already) at the forefront when it comes to advancing our knowledge of AD. I also heard Dr. LaFerla's talk on this data at the recent Philadelphia meeting, and thought it was one of the best presentations I heard. In regard to the comments by Thomas Bayer, who correctly inferred that "while immunotherapy against Aβ removes intra- and extracellular Aβ, the next step needs to be shown whether it helps to stop neuron death," in a recent report published in JCI (Caspase-cleavage of tau is an early event in AD tangle pathology, 114: 121-130, 2004), it was reported that in 12-month 3xTg-AD mice, caspase-cleavage of tau is evident and colocalizes with both an early tangle marker (MC1) and intraneuronal Aβ in CA1 pyramidal neurons. However, in their most recent report in Neuron, Oddo et al. did not examine whether removal of intraneuronal Aβ by immunotherapy abrogated caspase cleavage of tau. Although the demonstration of caspase-cleaved tau in this mouse model is not definitive for apoptosis or neuronal death occurring, it can be inferred that the activation of caspases and cleavage of cellular proteins (such as tau) most likely contributes to neuronal demise. In this regard, it would have been interesting to examine if immunotherapy prevents caspase activation and cleavage of tau in 3xTg-AD mice.

    Rissman, R.A., Poon, W.W., Blurton-Jones, M., Oddo, S., Torp, R., LaFerla, F.M., Rohn, T.T. and Cotman, C.W. (2004). Caspase-cleavage of tau is an early event in Alzheimer disease tangle pathology. J Clin Invest. 2004 Jul;114(1):121-30. (See commentary, page 23.) Abstract

Comments on Primary Papers for this Article

  1. Important and interesting work, but one question. What kind of mice would be a relevant control for 3XTg mice? Sarcastically speaking, transgenic expression of BSA may accelerate tau pathology in tau transgenic mice. Thus, there remains the danger that using 3XTg mice for drug screening might produce artifacts.

    View all comments by Takaomi Saido


News Citations

  1. Synapses Sizzle in Limelight of Symposium Preceding Neuroscience Conference, Orlando: Day 2
  2. New Orleans: Immunotherapy—The Game Is Still in Town

Paper Citations

  1. . Triple-transgenic model of Alzheimer's disease with plaques and tangles: intracellular Abeta and synaptic dysfunction. Neuron. 2003 Jul 31;39(3):409-21. PubMed.
  2. . Impaired proteasome function in Alzheimer's disease. J Neurochem. 2000 Jul;75(1):436-9. PubMed.
  3. . Binding of amyloid beta protein to the 20 S proteasome. J Biol Chem. 1997 Jan 3;272(1):58-62. PubMed.
  4. . Abeta immunotherapy leads to clearance of early, but not late, hyperphosphorylated tau aggregates via the proteasome. Neuron. 2004 Aug 5;43(3):321-32. PubMed.

Further Reading


  1. . Abeta immunotherapy leads to clearance of early, but not late, hyperphosphorylated tau aggregates via the proteasome. Neuron. 2004 Aug 5;43(3):321-32. PubMed.


  1. Philadelphia: The Enemy Within—Neurodegeneration From Intraneuronal Aβ

Primary Papers

  1. . Abeta immunotherapy leads to clearance of early, but not late, hyperphosphorylated tau aggregates via the proteasome. Neuron. 2004 Aug 5;43(3):321-32. PubMed.