Immunotherapy, through either passive or active immunization, remains a viable and aggressively pursued strategy for ridding the brain of amyloid β, the major component found in the amyloid plaques that are the hallmark of Alzheimer disease (see ARF related news story). But in this week’s online edition of PNAS, researchers present a slight variation on the theme—antibodies not against Aβ, but against the β-secretase (BACE) site in the Aβ precursor protein (AβPP). Because these antibodies could prevent BACE from proteolytically cleaving AβPP, a prerequisite for Aβ production, they have the potential to prevent the Aβ peptide from forming in the first place.

Beka Solomon and colleagues at Tel Aviv University, Israel, used what they call multiple antigenic peptide (MAP)—an antigen comprising eight copies of the BACE cleavage site—to generate BACE site antibodies in mice. First author Michal Arbel and colleagues then made monoclonal antibodies by fusing hybridoma cells with splenocytes from mice that had the highest MAP antibody titers. Then, selecting the BACE site monoclonal antibody with the highest affinity for MAP, they tested its ability to interfere with Aβ production.

Arbel and colleagues found that when the monoclonal antibody, which they called BSS1, was incubated with Chinese hamster ovary cells overexpressing AβPP, secretion of Aβ into the culture medium was reduced by up to 22 percent. In contrast, antibodies to the N-terminal of AβPP had no effect on Aβ levels. The authors also found that BSS1 had an even greater impact on intracellular Aβ, reducing it by about 50 percent, with a concomitant reduction in the amount of membrane-bound AβPP C-terminal fragments (C99), indicating that BACE cleavage had indeed been affected. The authors also found that the antibody was rapidly internalized by the cells, appearing in early endosomes within about 60 minutes of incubation. The antibody had no effect on α-secretase-mediated cleavage, an indication of its specificity for the BACE site on the protein.

When the authors immunized AβPP transgenic mice with MAP, they found that the animals did better than control littermates in the Morris water maze, a standard test of learning and memory function, though the numbers of animals used in these experiments were small. Also, in these animals the amyloid burden in the brain dropped by over half.

BACE itself has come in for intense scrutiny as a possible therapeutic target, but as Solomon and colleagues point out, BACE activity might be required for cleavage of other proteins, or other putative proteases might be capable of carrying out cleavage of AβPP at the BACE site. The use of anti-BACE site antibodies would circumvent both these potential pitfalls. The fact that the antibodies can be internalized is also good news, given the increasing evidence that intracellular cleavage and toxicity of Aβ may play a role in AD (see ARF live discussion and ARF related news story). In fact, because upregulation of endocytosis might lead to enhanced cleavage of AβPP and increased plaque formation in non-familial cases of AD (see, for example, Nixon et al., 2003), the authors suggest that their immunologic approach may be “a therapeutic strategy in AD treatment that is worthy of further investigation.”—Tom Fagan


  1. The creative approach reported by Arbel and colleagues to reduce Aβ burden using site-directed APP antibodies that inhibit β-cleavage offers a possible alternative to specific BACE inhibitors, which have been challenging to design and could interfere with the processing of BACE substrates likely to be uncovered in the future. Another attractive feature of the approach is that it strategically targets APP processing within the endocytic pathway, the site of the earliest appearing cellular disturbances related to AD and a likely source for Aβ overproduction in sporadic AD. The approach has a possible advantage over immunization strategies directed squarely at β-amyloid removal: Blocking Aβ production may allow for a more gradual process of amyloid/Aβ clearance better tolerated by the cerebral vasculature. One possible disadvantage, however, is that the β-site-directed antibodies also cross-react with APP holoprotein and may, therefore, affect its function adversely.

    Judgments about the promise of this approach as a possible therapy rest on confirming the provocative, though very preliminary, findings in the mouse amyloidopathy model and establishing that the mechanism of action demonstrated in the cell culture experiments applies in all respects to the Aβ lowering seen in brain. The much greater effect seen in vivo than in vitro needs to be understood. The 8-month age of the Tg2576 mice analyzed in the study is a possible factor to consider in this regard because the amyloid phenotype is quite variable at this young age.

  2. Recent exciting work from Beka Solomon’s group demonstrates proof-of-concept for a novel way to block β-secretase activity using active immunization in mice to generate antibodies directed against the β-secretase cleavage site of APP. In a paper by Arbel et al., WT mice were immunized with a multiple antigen peptide containing eight copies of an eight-residue peptide representing the site at which β-secretase cleaves APP. To enhance immunogenicity, one of the two mutant residues found in the APP Swedish mutation was used in the immunogen. Antibodies were generated, purified, and used for in vitro studies of APP processing and Aβ generation in hAPPWT-transfected CHO cells. The antibodies recognized full-length APP but not Aβ. When co-incubated with APP-CHO cells, the antibodies were internalized, colocalized with APP in early endosomes, and resulted in reduced intracellular Aβ (50 percent by 5 days) and extracellular Aβ (22 percent at 9 hours, but effect wore off later). APP CTF C99 was also diminished. Preliminary data from a pilot study in 8-month-old APP Tg2576 mice showed improved cognition and reduced cerebral Aβ burden in immunized mice compared to nonimmunized controls. The authors suggest that the antibodies cross the blood-brain barrier (BBB) and inhibit β-secretase activity in the brain, although evidence for this is not yet provided. Unfortunately, the authors confused our previously published work as evidence that Aβ antibodies cross the BBB and bind plaques following passive immunization in APP Tg mice. In our paper by Seabrook et al., we used active immunization with human or rodent Aβ peptides and showed that mouse Ig could be immunohistochemically detected in plaques in both immunized and nonimmunized mouse brain cryosections, similar to findings of human IgG in cerebral plaques of brain tissue from Aβ-immunized and nonimmunized AD patients as reported by Nicoll et al. Aβ-specific immunoglobulins were undetectable in brain by ELISA or Western blot. Thus, our results do not provide evidence that Aβ antibodies cross the BBB and bind plaques.

    The novel method of blocking β-secretase activity described by Arbel et al. using immunotherapy is creative and exciting. It will be interesting to learn more about the effects of such an immunogen in vivo, including a full analysis of both the humoral and cellular immune responses. A robust T cell response to APP via active immunization may wreak havoc on the body, indicating that passive transfer of antibodies against the β-secretase cleavage site of APP may be a safer alternative.


    . Inhibition of amyloid precursor protein processing by beta-secretase through site-directed antibodies. Proc Natl Acad Sci U S A. 2005 May 24;102(21):7718-23. PubMed.

    . Species-specific immune response to immunization with human versus rodent A beta peptide. Neurobiol Aging. 2004 Oct;25(9):1141-51. PubMed.

    . Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: a case report. Nat Med. 2003 Apr;9(4):448-52. PubMed.

  3. Our article is related to the very important accelerated endocytosis process as presented firstly by the Nixon group. The question of what leads to this process is key for understanding the early process of AD. I believe that a clue can be attained from Down syndrome (DS) cases. As the Nixon group has previously shown, the accelerated endocytosis is clearly present in DS as well as sporadic forms of AD at the very early stages.

    DS is caused by triplication of chromosome 21. By brief examination of this chromosome, I have seen some interesting genes that are related to the endocytosis process. Two of them, coding for the proteins synaptojanin and synaptogamin, might be related to the accelerated endocytosis. In Down syndrome and in the sporadic form of AD, overexpression of these genes may be a result of a defect in cholesterol homeostasis, for example, malfunctions of cholesterol uptake from the bloodstream (related to ApoE activity?). The cells may accelerate endocytosis as a mechanism to increase cholesterol uptake to compensate for the faulty cholesterol endocytosis.

    Overexpression of these two proteins in combination with APP may shed more light on this very early stage of the disease. It is interesting to note that these proteins are missing from the mouse chromosome 17, which is the mouse homolog to chromosome 21. Addition of APP and synaptojanin and synaptogamin to a mouse model of DS or vice versa, addition of synaptojanin and synaptogamin to hAPP mouse models, would create a much more realistic model of sporadic cases.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Sorrento: Immunotherapy Update Hot Off Lectern of AD/PD Conference
  2. Tackling Alzheimer’s from the Outside In

Paper Citations

  1. . The neuronal endosomal-lysosomal system in Alzheimer's disease. J Alzheimers Dis. 2001 Feb;3(1):97-107. PubMed.

Other Citations

  1. ARF live discussion

Further Reading

No Available Further Reading

Primary Papers

  1. . Inhibition of amyloid precursor protein processing by beta-secretase through site-directed antibodies. Proc Natl Acad Sci U S A. 2005 May 24;102(21):7718-23. PubMed.