This paper offers an interesting novel mechanism by which anti-Aβ antibodies mediate intraneuronal Aβ reduction. Accumulating data support the role of intraneuronal Aβ in the early stages of AD. Indeed, intracerebral injection of anti-Aβ antibody to the triple-transgenic AD mice model reduced intraneuronal Aβ, and this correlated with improvement in cognitive function (1). Many mechanisms were suggested to account for extracellular Aβ reduction; however, the mechanism by which the antibodies affect intracellular Aβ is not clear.

To address this issue, Tampellini and colleagues performed an extensive work employing both a neuronal cell line and primary neurons of the Tg2576 AD transgenic model. In their work, the authors show that anti-Aβ antibodies administered in the cells' growing media are internalized into the cells by binding to the Aβ sequence of the full-length APP, and that the antibodies accelerate Aβ degradation through the endosomal-lysosomal pathway. Importantly, Aβ carboxy terminal-specific antibodies that preclude APP binding fail to enter the cells and affect...  Read more

This paper offers an interesting novel mechanism by which anti-Aβ antibodies mediate intraneuronal Aβ reduction. Accumulating data support the role of intraneuronal Aβ in the early stages of AD. Indeed, intracerebral injection of anti-Aβ antibody to the triple-transgenic AD mice model reduced intraneuronal Aβ, and this correlated with improvement in cognitive function (1). Many mechanisms were suggested to account for extracellular Aβ reduction; however, the mechanism by which the antibodies affect intracellular Aβ is not clear.

To address this issue, Tampellini and colleagues performed an extensive work employing both a neuronal cell line and primary neurons of the Tg2576 AD transgenic model. In their work, the authors show that anti-Aβ antibodies administered in the cells' growing media are internalized into the cells by binding to the Aβ sequence of the full-length APP, and that the antibodies accelerate Aβ degradation through the endosomal-lysosomal pathway. Importantly, Aβ carboxy terminal-specific antibodies that preclude APP binding fail to enter the cells and affect intracellular Aβ, while antibodies directed to the ectodomain of APP follow the same internalization route; however, they did not alter Aβ levels. The authors preclude decreased processing of APP by both β- and γ-secretase as a possible mechanism for Aβ reduction and, interestingly, β-cleavage products are increased as a result of antibody treatment.

These results support our previously reported data regarding antibodies directed to the β-secretase cleavage site of APP (2). Anti-β-site antibody trafficking is quite similar to that described by Tampellini et al., being co-localized with early endosomal markers starting at 2 minutes after incubation and evident at the lysosomes at later time points (from 45 minutes and later; unpublished data). Unlike their results, intracellular Aβ reduction mediated by anti-β-site antibody in our hands was associated with a decrease in C99 levels, suggesting an inhibition of β-secretase cleavage (2). But similar to the results presented here, despite antibody internalization, the APP amino terminus antibody failed to affect Aβ levels in our hands.

Unlike mAb 6E10 presented in this paper, in our hands mAb 10D5 (Elan Pharmaceuticals) and antibody 196, both targeted to the amino terminus of Aβ, fail to bind cell surface APP and showed no internalization, probably due to their cryptic epitope (unpublished data).

The results presented in this paper by the Gouras research group present a mechanism by which passive administration of antibodies against Aβ peptide reduces intracellular Aβ levels. This mechanism comes in addition to previously proposed mechanisms that mainly account for extracellular Aβ clearance. This paper provides new insights regarding the mechanism underlying the beneficial effect of Aβ immunotherapy in terms of reduced cerebral Aβ and increased cognitive function.

References:
1. 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 Aug 5;43(3):321-32. Abstract

2. Arbel M, Yacoby I, Solomon B. 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. Epub 2005 May 13. Abstract

View all comments by Michal Arbel-Ornath

In this manuscript, Gouras and colleagues showed that anti-Aβ antibodies generated against the extracellular domain of Aβ were able to reduce intracellular Aβ accumulation in a neuronal cell line and primary neurons. In contrast, they showed that treatment with the G2-11 antibody, which is raised against the intramembranous region of Aβ, did not reduce intraneuronal Aβ accumulation. This is consistent with our previous work, where we showed that both active and passive Aβ immunization was able to clear intraneuronal Aβ deposits in vivo (Oddo et al., 2004; Billings et al., 2005; Oddo et al., 2006).

To determine the mechanism of intraneuronal Aβ clearance, Gouras and colleagues nicely showed that, in their experimental setup, the antibodies 6E10 and 4G8 are internalized by endocytosis and this internalization process is required for intraneuronal Aβ clearance. We have shown that in the 3xTg-AD mice, after a single intrahippocampal injection...  Read more

In this manuscript, Gouras and colleagues showed that anti-Aβ antibodies generated against the extracellular domain of Aβ were able to reduce intracellular Aβ accumulation in a neuronal cell line and primary neurons. In contrast, they showed that treatment with the G2-11 antibody, which is raised against the intramembranous region of Aβ, did not reduce intraneuronal Aβ accumulation. This is consistent with our previous work, where we showed that both active and passive Aβ immunization was able to clear intraneuronal Aβ deposits in vivo (Oddo et al., 2004; Billings et al., 2005; Oddo et al., 2006).

To determine the mechanism of intraneuronal Aβ clearance, Gouras and colleagues nicely showed that, in their experimental setup, the antibodies 6E10 and 4G8 are internalized by endocytosis and this internalization process is required for intraneuronal Aβ clearance. We have shown that in the 3xTg-AD mice, after a single intrahippocampal injection of 6E10, extracellular Aβ deposits are removed first, followed by the clearance of intraneuronal Aβ deposits. Moreover, we showed that once the antibody diffuses away, intraneuronal Aβ accumulates first and subsequently extracellular plaques are also detected. We concluded that there is a dynamic relationship between these two pools of Aβ. Our hypothesis, however, is not mutually exclusive with the results presented here by Gouras's group. The two mechanisms of intraneuronal Aβ clearance can be independent of each other or maybe have a synergistic effect on the clearance of intraneuronal Aβ in vivo.

View all comments by Salvatore Oddo

Riddle me this, Batman: If you took a cell lysate and incubated it with antibody to a particular cellular protein, would it then be possible to detect that protein in an ELISA? Or would the (first) antibody mask the protein? And... isn't that possibly what Tampellini et al. have done? Granted, this would still require internalization of the antibody - something that has been well-documented in other systems. And, it would probably still indicate a helpful effect, as the antibody would probably inhibit aggregation. But, it would mean that the title of the paper is technically wrong. In other words, I did not see direct, quantitative evidence that A-beta was "reduced." I say this not to be a naysayer - just to point out a technical difficulty.

View all comments by Steve Barger

Reply by Davide Tampellini, Michael Lin, Gunnar Gouras
Regarding the technical question by Dr. Barger, we used two other methods besides ELISA. In the Western blot in Figure 1B, neurons were treated with Aβ antibody, washed, lysed in 6 percent SDS, and direct-loaded onto the gel, run under denaturing conditions. There was no antibody-based capture process that would be confounded by previous antibody treatment. Moreover, the denaturing conditions, which we also use for immunoprecipitation experiments, separate antibody-antigen complexes. Thus, the decreased Aβ band intensities in antibody treated cells are not due to obscuration by the antibody treatment. Additionally, using immunofluorescence (Figure 2E), the Aβ42-specific detecting antibody recognizes an entirely different epitope from the treating antibody.

View all comments by Gunnar K. Gouras
View all comments by Michael Lin

We are investigating clearance of intraneuronal amyloid-β in Alzheimer brain sections using incubation (two or four days) with monocytes of control or Alzheimer subjects. Control monocytes intrude into neurons by their processes and upload oligomeric amyloid-β; Alzheimer patients' monocytes upload less and suffer apoptosis; apoptotic monocytes spill amyloid-β which becomes fibrillar and leads to congophilic angiopathy.

We agree with the comments by Alex Roher that vascular amyloidosis is a critical problem and is not easily solved by vaccination, which does not increase degradation. The issue is that amyloid-β needs to be degraded inside the brain, the modes of transport out of the brain are restricted. Macrophages of control subjects are much more effective in degrading amyloid-β in vitro compared to macrophages of Alzheimer patients.

References:
Fiala M, Philip Liu, Araceli Espinosa-Jeffrey et al. Innate immunity and transcription of MGAT-III and Toll-like receptors in Alzheimer's disease patients are improved by bisdemethoxycurcumin. Proc Natl Acad Sci USA. 2007;104:12849-12854. Abstract

View all comments by Milan Fiala