As ADAM10, aka α-secretase, cleaves the amyloid precursor protein (APP) in the non-amyloidogenic fashion, its loss could spell trouble. In the May 8 Journal of Clinical Investigation, scientists led by Monica Di Luca, University of Milano, Italy, report that endocytosis whisks α-secretase away from the cell membrane in a process that seems regulated by neuronal activity. The process relies on ADAM10 binding to AP2, an endocytosis modulator that works similarly to PICALM, an AD GWAS hit (see ARF related news story). If confirmed, Di Luca’s results suggest one way cells balance synaptic α-secretase levels. “This study explains the role played by neuronal activity in Aβ secretion,” wrote Charles Duyckaerts, Hospital de la Salpêtrière, Paris, to Alzforum in an e-mail (see full comment below). “It also sheds new light on the [known] connection between endocytosis and Aβ.”

Researchers know a bit about how ADAM10 gets to the synapse. Di Luca’s lab previously found that ADAM10 binds to synapse-associated protein 97 (SAP97), which ushers the enzyme to the postsynaptic membrane. When the researchers disrupted this interaction in mice, Aβ accumulated in the brain, suggesting that ADAM10’s forward trafficking prevents buildup (see Epis et al., 2010). The researchers later found that this same protein interaction is impaired in Alzheimer’s patients (see Marcello et al., 2012).

In this study, Di Luca and colleagues wanted to research the other side of the coin: How does ADAM10 leave the synapse? One candidate was clathrin-mediated endocytosis. This way of engulfing plasma proteins and lipids relies on formation of clathrin-coated vesicles and is known to regulate levels of synaptic proteins. Given the genetic links between endocytosis and AD, the group investigated whether this process removes ADAM10.

Lead author Elena Marcello first checked to see if endocytotic proteins bound ADAM10. Based on coimmunoprecipitation of the secretase from human hippocampal homogenates, she believes it binds the endocytosis adaptor AP2, which helps form clathrin-coated pits. The researchers found that ADAM10 binds through a previously unknown AP2-binding motif within its C-terminal tail. Mutating or deleting that motif abolished binding and enhanced ADAM10 on the surface of monkey kidney cells. The findings imply that clathrin-mediated endocytosis removes the enzyme from the plasma membrane.

What might regulate endocytosis of ADAM10? Since synaptic activity promotes both insertion and removal of neurotransmitter receptors at synapses, could it do the same for ADAM10? To test this, Marcello and colleagues chemically induced both long-term potentiation (LTP) and long-term depression (LTD) in primary hippocampal neuronal cultures, then analyzed how much ADAM10 bound AP2 and SAP97. LTP raised ADAM10/AP2 binding, depleted ADAM10 from membranes, and suppressed α-cleavage of APP. Conversely, LTD raised ADAM10’s binding with SAP97, promoted ADAM10 insertion into the membrane, and hiked its APP cleavage, implying that the process pushed more ADAM10 to the synapse. Together, the results suggest that LTD raises, while LTP lowers, ADAM10’s activity.

What, if anything, does this mean for AD patients? In hippocampal brain homogenates from six people who died in the early phases of sporadic AD, more ADAM10 and AP2 proteins bound each other. However, in a sample of 200 patients with possible AD, no genetic association turned up to explain that closer association. In early stages of AD, an imbalance of ADAM10 endocytosis and exocytosis could lead to the observed reduction of α-secretase activity, wrote the authors. Di Luca does not know why AD brains might have enhanced AP2/ADAM10 binding, but plans to study that next.

It may have to do with cholesterol levels in the membrane, suggested Duyckaerts. He previously reported that elevated cholesterol favors endocytosis of APP and production of Aβ (Cossec et al., 2010). Di Luca and colleagues are now testing small molecules that interfere specifically with the interaction between AP2 and ADAM10. Since AP2 latches onto many other proteins, they want to be sure these compounds only interfere with binding to ADAM10. These compounds may leave more ADAM10 in the membrane to cleave Aβ, she said, offering a potential new therapeutic target.

The study has clear implications for AD, said Stefan Lichtenthaler, German Center for Neurodegenerative Diseases, Munich. “It may provide clues about the molecular mechanisms that underlie the disease—if the increase in AP2/ADAM10 interaction happens early, this could speed up the disease process,” he said. The study also has broader implications for neuronal plasticity and cell surface proteases, he added. “The mechanisms that control protease activity are relatively poorly understood,” he told Alzforum. This study shows that accurate control of protein trafficking is a way to modulate that activity. LTP and LTD seem to work opposite each other to regulate that trafficking, he added. He wonders what other ADAM10 substrates might be affected by this changing synapse activity.

“There is a long history of literature linking clathrin-mediated endocytosis and Aβ,” John Cirrito, Washington University School of Medicine in St. Louis, Missouri, told Alzforum in an e-mail (see ARF related news story and ARF news story). “However those studies generally revolve around APP internalization and Aβ generation,” he added (see full comment below). The current data suggest an additional [ADAM10-based] mechanism by which endocytosis contributes to Aβ formation, he wrote. The results jibe with his previous research suggesting that elevated synaptic activity leads to Aβ generation (see ARF related news story), though the authors did not directly measure peptide levels, he wrote.—Gwyneth Dickey Zakaib


  1. Marcello and colleagues demonstrate that synaptic activity regulates the levels of ADAM10 on the plasma membrane. Long-term potentiation (LTP) induces ADAM10 internalization by clathrin-mediated endocytosis (CME), whereas LTD induces its insertion into the plasma membrane. There is a long history of literature linking CME and Aβ; however, those studies generally revolve around APP internalization and Aβ generation. Work from our group and others shows that synaptic activity causes CME of APP, which increases Aβ production in endosomes. The data in the Marcello paper look at this from a different angle. Here, synaptic activity increases ADAM10 internalization, which decreases its ability to cleave APP; in theory, this would then increase Aβ generation. So taken together, this suggests that synaptic activity and CME may promote Aβ generation by two parallel pathways: 1) increasing amyloidogenic processing of APP within endosomes, and 2) decreasing non-amyloidogenic processing of APP at the plasma membrane.

    As with any good study, lots of questions remain. Is ADAM10 internalized into separate endosomes than APP? Or can ADAM10 and APP still be within the same endosome but maybe the low pH prevents further cleavage? How is ADAM10 trafficked or segregated after CME? The authors demonstrate that activity modulates APP cleavage by ADAM10, which is consistent with a reduction in Aβ, but one thing lacking in this paper is a direct measurement that Aβ generation is changing as a consequence of the altered ADAM10 subcellular localization.

  2. A disintegrin and metalloproteinase 10 (ADAM10) is apparently one of the most critical membrane-associated proteases in the central nervous system (CNS). Its prominent role in the embryonic and adult CNS has been revealed by a number of studies. Next to APP, an increasing number of transmembrane proteins, including Notch receptors and ligands, are subject to ADAM10-mediated shedding. These shedding events are of critical importance to modulate postsynaptic function and synaptic plasticity.

    Based on their previous work, Monica Di Luca´s group convincingly addressed the post-transcriptional regulation of ADAM10 in neurons. Both its transport to the postsynaptic membrane and its removal are central events to regulate synaptic functions, morphology, and the processing of important substrates, including APP. In the current study, the authors focus on the endocytosis of ADAM10 from the postsynaptic membrane. Using mainly coimmunoprecipitation experiments, they showed that, like other surface molecules, ADAM10 endocytosis depends on binding to the clathrin adaptor AP2. This binding seems to be stronger in samples from AD patients. It was shown that abolishing this binding using mutants or pharmacological approaches also reduced ADAM10 endocytosis. Interestingly, the authors found that long-term potentiation induced this process, whereas long-term depression had an opposite effect by stimulating interaction of ADAM10 and SAP97, thereby promoting ADAM10 delivery to the plasma membrane. In a last set of experiments, they showed that this dynamic, and apparently ADAM10-dependent regulation, also led to a differential processing of APP at the α-secretase site. ADAM10 localization at the synaptic membrane more or less determines whether α-secretase processing occurs.

    What are the consequences for AD? The findings certainly increase our basic understanding of how APP processing, synaptic remodeling, and cellular localization of ADAM10 are linked. I do not feel that immediate new therapeutic targets are apparent, since the factors involved affect a number of other proteins as well; for example, the fine-tuned localization of ADAM10 is necessary for the degree of shedding of most likely more than 10 other synaptic membrane proteins. Also, AP2 mediates endocytosis of a huge number of surface proteins. On the other hand, additional intracellular and extracellular factors are needed to control the activity of ADAM10. We and others showed recently, for example, that the integration of ADAM10 in the tetraspanin web is instrumental for forward trafficking of the protease. It is likely that additional factors directly contribute to the cellular localization of ADAM10.

    As discussed by the authors, the Aβ levels in vivo are in part dependent on the activity of ADAM10. This study additionally provides evidence that the degree of neuronal and synaptic activity alters the function (and localization) of ADAM10 and the degree of Aβ production. Based on their initial observations that in AD brains, the endocytic route of ADAM10 is favored, this would directly explain both the increased Aβ production and the problems in synaptogenes as reported in AD.

  3. I read this paper by Marcello et al. with much interest, and I was impressed both by the number of data and the coherence of the hypothesis. It explains the role played by neuronal activity in Aβ secretion. It also sheds new light on the connection between endocytosis and Aβ.

    In addition, it opens new research perspectives: The alteration of ADAM10/AP2 association in AD is currently not explained and could be related to changes in the cell membrane itself. We have, in this respect, shown that increases in membrane cholesterol favor endocytosis and production of Aβ (see Marquer et al., 2011; Cossec et al., 2010).

    I'd add a word of caution on the neuropathology. Only six cases were examined at Braak stage IV—these were apparently the same cases the authors studied before (see Marcello et al., 2012). Braak stage IV pathology is common in asymptomatic aged persons. The diagnostic probability of Alzheimer's disease is only ranked as intermediate in the current diagnostic criteria (Hyman et al., 2012; Montine et al., 2012). Such cases are, by definition, free from tau pathology in the neocortex. Usually, Aβ deposits are, however, already present. This could be determined in the studied cases, for instance, by identifying the stage of amyloid pathology (see Thal et al., 2002). It would be interesting to compare the association of ADAM10/AP2 in the hippocampus (with tangle pathology) and in the frontal cortex (devoid of tangle pathology but possibly with Aβ deposition). Tau accumulation, which may affect synapses, could play a central role in the alteration of ADAM10/AP2 association, as it probably does for the clathrin adaptor PICALM, which we found to colocalize with tangles (Ando et al., 2013). New studies with more advanced cases would strengthen the human data.


    . Local cholesterol increase triggers amyloid precursor protein-Bace1 clustering in lipid rafts and rapid endocytosis. FASEB J. 2011 Apr;25(4):1295-305. PubMed.

    . Clathrin-dependent APP endocytosis and Abeta secretion are highly sensitive to the level of plasma membrane cholesterol. Biochim Biophys Acta. 2010 Aug;1801(8):846-52. PubMed.

    . SAP97-mediated local trafficking is altered in Alzheimer disease patients' hippocampus. Neurobiol Aging. 2012 Feb;33(2):422.e1-10. PubMed.

    . National Institute on Aging-Alzheimer's Association guidelines for the neuropathologic assessment of Alzheimer's disease. Alzheimers Dement. 2012 Jan;8(1):1-13. PubMed.

    . National Institute on Aging-Alzheimer's Association guidelines for the neuropathologic assessment of Alzheimer's disease: a practical approach. Acta Neuropathol. 2012 Jan;123(1):1-11. PubMed.

    . Phases of A beta-deposition in the human brain and its relevance for the development of AD. Neurology. 2002 Jun 25;58(12):1791-800. PubMed.

    . Clathrin adaptor CALM/PICALM is associated with neurofibrillary tangles and is cleaved in Alzheimer's brains. Acta Neuropathol. 2013 Jun;125(6):861-78. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Vienna: In Genetics, Bigger Is Better—Data Sharing Nets Three New Hits
  2. Link Between Synaptic Activity, Aβ Processing Revealed
  3. Traffic Control—Curb Endocytosis to Curb AD Pathogenesis?
  4. Paper Alert: Synaptic Activity Increases Aβ Release

Paper Citations

  1. . Blocking ADAM10 synaptic trafficking generates a model of sporadic Alzheimer's disease. Brain. 2010 Nov;133(11):3323-35. PubMed.
  2. . SAP97-mediated local trafficking is altered in Alzheimer disease patients' hippocampus. Neurobiol Aging. 2012 Feb;33(2):422.e1-10. PubMed.
  3. . Clathrin-dependent APP endocytosis and Abeta secretion are highly sensitive to the level of plasma membrane cholesterol. Biochim Biophys Acta. 2010 Aug;1801(8):846-52. PubMed.

Further Reading


  1. . Alpha-secretase cleavage of the amyloid precursor protein: proteolysis regulated by signaling pathways and protein trafficking. Curr Alzheimer Res. 2012 Feb;9(2):165-77. PubMed.
  2. . Synapse-associated protein-97 mediates alpha-secretase ADAM10 trafficking and promotes its activity. J Neurosci. 2007 Feb 14;27(7):1682-91. PubMed.
  3. . SAP97-mediated local trafficking is altered in Alzheimer disease patients' hippocampus. Neurobiol Aging. 2012 Feb;33(2):422.e1-10. PubMed.
  4. . Blocking ADAM10 synaptic trafficking generates a model of sporadic Alzheimer's disease. Brain. 2010 Nov;133(11):3323-35. PubMed.
  5. . Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nat Genet. 2009 Oct;41(10):1088-93. PubMed.
  6. . Probing sporadic and familial Alzheimer's disease using induced pluripotent stem cells. Nature. 2012 Jan 25;482(7384):216-20. PubMed.
  7. . Tetraspanin12 regulates ADAM10-dependent cleavage of amyloid precursor protein. FASEB J. 2009 Nov;23(11):3674-81. PubMed.
  8. . Low cholesterol stimulates the nonamyloidogenic pathway by its effect on the alpha -secretase ADAM 10. Proc Natl Acad Sci U S A. 2001 May 8;98(10):5815-20. PubMed.

Primary Papers

  1. . Endocytosis of synaptic ADAM10 in neuronal plasticity and Alzheimer's disease. J Clin Invest. 2013 Jun 3;123(6):2523-38. PubMed.