After their production in endosomes, intraneuronal membrane-bound amyloid-β (Aβ) accumulates in multivesicular bodies (MVBs), but the route they take from there, to end up as plaque-forming extracellular peptides, remains unknown. MVBs are bags of bags, organelles that contain vesicles destined for delivery to lysosomes, where their contents are degraded, or to the plasma membrane, where the vesicles are released intact as exosomes into the extracellular space. Those exosomes can be Aβ’s ticket to ride out of the cell, according to a paper in the PNAS online edition this week.

The work, from Kai Simons and colleagues at the Max Planck Institute in Dresden, Germany, traces the footsteps of APP and its fragments in cultured cells as it is cleaved by β-secretase (BACE) in early endosomes. After that, a small fraction of Aβ (less than 1 percent) is routed to multivesicular bodies, and released into the media in association with exosomes, the researchers report. The results suggest that exosomes provide an escape route to the extracellular space for at least a portion of intracellular Aβ. In support of this idea, the researchers show that the amyloid plaques in human Alzheimer disease brain contain the exosome protein Alix.

In a related paper, researchers from the lab of Shoichi Ishiura at the University of Tokyo show that BACE can associate with the exosomal marker and lipid raft protein flotillin-1. Together with the demonstration by Simons and colleagues that flotillin is present in Aβ-containing exosomes, the work raises the possibility that that exosomal Aβ is a product of BACE and γ-secretase that travel together with APP on lipid rafts.

Aβ has been spotted in MVBs before: A few years ago, researchers from Gunnar Gouras’s lab reported that the major site of intraneuronal accumulation of Aβ42 in mouse brain is in these organelles (Takahashi et al., 2002). Similar results were later reported in mouse brain (Langui et al., 2004). More recent work says that Aβ accumulation impairs the MVB sorting pathway by slowing passage of proteins and interfering with ubiquitin-dependent protein degradation (see ARF related news story).

In the new study, first author Lawrence Rajendran and colleagues localize β-secretase (BACE) cleavage of APP to early endosomes using an antibody (ANJJ) that recognizes sAPPβ, the soluble BACE cleavage product of APP. sAPPβ co-localized with early and late endosome markers, but not exocytic markers in Hela or Neuro2a cells expressing the Swedish mutant of APP (swAPP). Their results support the idea that APP and BACE are co-internalized from the cell surface, and that BACE cleavage occurs in early endosomes. Work from others has shown that γ-secretase is also present in endosomes, where it cleaves sAPPβ to produce membrane-bound Aβ peptides.

From early endosomes, Aβ has the potential to recycle to the membrane for immediate exocytosis, but instead it appears to travel on down the endosomal pathway into MVBs. The vesicles that fill MVBs are formed by invagination of the endosome membrane to create internal vesicles, which are then sorted for degradation or release at the plasma membrane. Using immunoelectron microscopy, the researchers detected Aβ peptides in MVBs in Neuro2a cells. To ask if Aβ might be exocytosed via this route, the researchers looked for membrane-associated Aβ in culture supernatants. Differential centrifugation of culture supernatants revealed a very small fraction (about one tenth of 1 percent) of extracellular Aβ in the membrane-containing high-speed pellet. The same pellet contained a significant fraction (about 20 percent) of the extracellular sAPPβ, but little sAPPα or full-length APP. The membrane-associated Aβ was in endosomes, as shown by its cosedimentation on sucrose gradients with the endosomal markers flotillin and Alix, and immunoelectron microscopy that detected the Aβ on Alix- and flotillin-positive 60-100 nm-diameter vesicles.

The results show that Aβ can be released from cells in exosomes, but the question remains if this pathway contributes significantly to the accumulation of extracellular Aβ in AD. To directly address this question, Rajendran and colleagues looked for Alix in brains of three AD patients, as well as two control subjects and two with Parkinson disease. They detected Alix immunostaining around small neuritic plaques and in large diffuse plaques from all three AD brains tested but no trace of the protein in the PD or control brains. Other investigators have found flotillin in amyloid plaques in mouse brain (Kokubo et al., 2005). Both results are consistent with the release and incorporation of exosomal Aβ into extracellular amyloid deposits.

While only a small amount of secreted Aβ was found in association with exosomes, the authors speculate that the vesicles, which are enriched for the amyloid-promoting ganglioside GM1, could act as nucleation centers for amyloid plaques. They cite the example of the amyloid-forming protein pMel1 (see ARF related news story), which is known to travel in MVBs and be released via exosomes. They propose that the internal environment in MVB vesicles or exosomes might promote amyloid fibril formation (reviewed in Theos et al., 2005). Alternatively, exosomal release could provide a slow drip of Aβ over the long course of AD. It is also possible that MVBs could carry and release soluble Aβ, they note.

The exosomal marker flotillin is also known as an organizer of lipid rafts—membrane microdomains that have been implicated in signal transduction and APP processing. In a related paper, the Japanese group report that FLOT-1 directly interacts with BACE in BACE1-expressing HEK cells. Overexpression of FLOT-1 or another raft protein, CAV-1, results in increased recruitment of BACE into lipid rafts, but decreased BACE activity and Aβ secretion. The results provide a physical link between BACE and lipid raft proteins, and raise the possibility that BACE and Aβ might transit from endosomes to exosomes in lipid rafts. Because rafts are enriched in the amyloid-promoting GM1, and provide a platform for γ-secretase (see ARF related news story), exosomes might contain a perfect mix of ingredients for formation of amyloid fibrils.—Pat McCaffrey


  1. Rajendran and colleagues provide further insights into the biology of Alzheimer disease (AD). They present more evidence for endosomes as important sites of Aβ generation using elegant cell biology experiments. By crosslinking experiments, they show that there was increased colocalization of APP with BACE in early endosomes. Aβ secretion was dramatically reduced by inhibiting endocytosis or reducing recycling. They provide further confirmation for Aβ localization to multivesicular bodies (MVBs) by immunoelectron microscopy and present exciting new data on Aβ in secreted exosomes. It has been described that in some cells MVBs can fuse with the plasma membrane and secrete their inner (luminal) vesicles (exosomes). Interestingly, the authors localize the exosome component Alix to plaques in AD brain, and note that flottilin-1, known to be contained in exosomes, was previously shown to associate with plaques. Yet, they point out that, at least in APP transfected neuroblastoma cells, exosome-associated Aβ appears to account for only about 1 percent of Aβ secreted. They hypothesize that in the slowly progressive disease process, secreted exosome-associated Aβ may be important in the progression of the disease. Despite evidence in support of intraneuronal MVB Aβ involvement in plaque formation, the authors did not consider the alternative—that plaque-associated Alix and flotillin-1, which localize to exosomes and MVB inner vesicles, could originate from MVBs without being secreted.

  2. I strongly recommend this article. But what percent of amyloid-β is bound to exosomes in AD is still shrouded in mystery. Much more experimental data are required to understand this better.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Paper Alert: Intraneuronal Aβ Impairs Multivesicular Body Sorting
  2. Is It Good for You? Amyloid Shows New Side in Mammalian Cells
  3. γ-Secretase Activity is Mostly Intracellular—Needs PEN-2 C-terminus

Paper Citations

  1. . Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am J Pathol. 2002 Nov;161(5):1869-79. PubMed.
  2. . Subcellular topography of neuronal Abeta peptide in APPxPS1 transgenic mice. Am J Pathol. 2004 Nov;165(5):1465-77. PubMed.
  3. . Part of membrane-bound Abeta exists in rafts within senile plaques in Tg2576 mouse brain. Neurobiol Aging. 2005 Apr;26(4):409-18. PubMed.
  4. . The Silver locus product Pmel17/gp100/Silv/ME20: controversial in name and in function. Pigment Cell Res. 2005 Oct;18(5):322-36. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Alzheimer's disease beta-amyloid peptides are released in association with exosomes. Proc Natl Acad Sci U S A. 2006 Jul 25;103(30):11172-7. PubMed.
  2. . BACE1 interacts with lipid raft proteins. J Neurosci Res. 2006 Sep;84(4):912-7. PubMed.