If proteins could guest on TV game shows, then amyloid-β would make a good candidate for What’s My Line?—the object of which is to guess someone’s invariably obscure profession. The many incarnations of Aβ would complicate the task, since it is unclear which is the most physiologically relevant. When they decided to study Aβ, researchers led by Inna Slutsky at Tel Aviv University, Israel, chose to focus on the profession rather than the professional. As reported in the November 22 Nature Neuroscience online, the researchers offer evidence that endogenous Aβ, true identity unclear, can modulate short-term synaptic plasticity, which appears to suffer when levels of the peptide are too high, or too low. “This means you need an optimal amount of Aβ to maintain facilitation in normal systems,” Slutsky told ARF. “If you don’t have it at all, you have as much trouble as if you have too much.” If true, the findings could have repercussions for treatments aimed at reducing Aβ production.

Joint first authors Efrat Abramov and Iftach Dolev blocked degradation of the peptide using inhibitors of neprilysin, a peri-synaptic protease that degrades Aβ. The idea was that neprilysin inhibitors would acutely raise the concentration of endogenously released Aβ in and around the synapse. To get the opposite effect, the researchers used antibodies to deplete Aβ. They used FM dyes, which fluoresce only after being incorporated into recycling synaptic vesicles, to measure the effects of their manipulations on synaptic activity in cultured rat hippocampal cells and slices. These dyes allow researchers to measure synaptic activity at high resolution, down to the single bouton level (for a review, see Gaffield and Betz, 2006).

The researchers found that either thiorphan, a neprilysin inhibitor, or AF1126, a polyclonal antibody that blocks neprilysin, caused a significant increase in levels of Aβ40 and Aβ42 in rat hippocampal cell culture medium (as judged by ELISA). This came with increased release probability—in other words, the chance that synaptic vesicles would fuse with the cell membrane and dump their contents into the cleft. The researchers calculated that this heightened release probability increased synaptic strength by two- to threefold. To test if the neprilysin effect was specific for Aβ, the researchers repeated the experiments with hippocampal cultures taken from amyloid-β precursor protein (APP)-negative mice. In those cells, thiorphan did not affect release probability or synaptic strength; nor did it affect synapse turnover if cultures were pretreated with either β- and γ-secretase inhibitors to block Aβ production, or with anti-Aβ antibodies. All told, the experiments support the idea that Aβ modulates synaptic activity by stimulating release of synaptic vesicles, the scientists believe. One potential drawback is that the researchers do not know exactly which species of Aβ might be responsible for the observed effects. In addition, neprilysin and Aβ antibodies may have multiple substrates, causing some researchers to question whether the observations can truly be attributed to Aβ.

To investigate whether this Aβ-induced synaptic release mattered functionally, the researchers measured conductance between pairs of rat neurons and also among neurons in a hippocampal network, i.e., the CA3-CA1 Schaffer collateral pathway in hippocampal slices. Using whole cell recordings, the researchers found that blocking Aβ degradation caused an increase in basal miniature excitatory postsynaptic currents (mEPSCs). To test how this relates to firing in a more physiologically relevant setting, Abramov and colleagues tested the effect of thiorphan when cells were undergoing bursts of activity. They found that while the first EPSC was increased in the presence of thiorphan, subsequent ones were decreased, indicating that the inhibitor, presumably through increasing synaptic Aβ, caused a decrease in short-term synaptic facilitation and hence an increase in short-term depression. At the network level, Abramov and colleagues also found an increase in excitability when neprilysin activity was blocked. Acute application of thiorphan to hippocampal slices increased basal excitatory postsynaptic potentials (EPSPs), and during bursts of activity the inhibitor reduced amplitudes of the third and subsequent action potentials, again indicative of short-term synaptic depression.

The opposite experiment—that is, decreasing Aβ by using the monoclonal antibody HJ5.1—surprisingly generated the same result, i.e., an increase in short-term depression. “This was unexpected,” said Slutsky. “It doesn’t matter if you have too much peptide or too little peptide. [Either way,] you get strong synaptic depression and very little synaptic facilitation.”

The findings fit in with some previous observations. Roberto Malinow and colleagues reported that Aβ can temper glutamate receptor-drive synaptic transmission, for example, (see ARF related news story on Kamenetz et al., 2003 id=27086). But in this case, Slutsky and colleagues propose that Aβ effects are entirely presynaptic. How the peptide influences presynaptic release is unclear. Slutsky said she is working on the mechanism.

“I view this as top-notch work providing exciting new clues to the potential role β amyloid plays at synapses,” Gunnar Gouras, Cornell University, New York, told ARF (see also comment below).

The work may have implications for AD. Abramov and colleagues went on to test long-term effects of inhibiting Aβ degradation using hippocampal cultures from wild-type mice. Mouse Aβ is less amyloidogenic than human Aβ, and mice normally don’t form Aβ aggregates or plaques. Blocking neprilysin in the hippocampal cultures led to a reduction in the number of synapses and a growth in synapse size in as little as 48 hours. “This is extremely important because it suggests that you don’t need aggregates or big oligomers,” said Slutsky. “I feel that the toxicity that comes at very late stages has nothing to do with the physiology or pathology of the protein at the initial stages,” she said.—Tom Fagan


  1. Taken together, the findings of this intriguing study suggest a novel role for endogenous Aβ as a positive modulator of excitatory transmission. Acutely increasing levels of Aβ rapidly and reversibly increased the number of active synapses and the amount of neurotransmitter released by each active synapse. Acutely reducing levels of Aβ reduced the number of active synapses by about half, with more modest effects on neurotransmitter release. While the data presented are compelling, they are hard to reconcile with previous reports that have suggested an inhibitory role for Aβ. In particular, they would seem to conflict with the lack of effect of acute inhibition of Aβ production on basal transmission in hippocampal slices (Kamenetz et al., 2003). One can, however, imagine scenarios where small differences in the concentration of Aβ, oligomeric state of the peptide, and/or the duration of synaptic exposure, could shift the balance between positive and negative feedback regulation. Indeed, Abramov et al. report that longer-term (48 hours) inhibition of Aβ degradation decreased the number of active synapses. Future experiments will be needed to tease apart the specific roles of endogenous Aβ under a variety of normal conditions and at different stages of Alzheimer disease.


    . APP processing and synaptic function. Neuron. 2003 Mar 27;37(6):925-37. PubMed.

  2. This is a detailed and elegant study reporting on acute changes in pre-synaptic release probability related to β amyloid (although it seems that the authors cannot fully rule out a role also of βC-terminal fragments of APP). Intriguingly, they provide evidence that Aβ plays a role in history-dependent enhancement of synaptic transmission. This paper builds on other recent studies providing evidence for a physiological role of Aβ at synapses. This group has a lot of expertise in these methods, and it is exciting to get such an excellent new group into the field of AD research. At the same time, one can speculate that if another group with expertise in post-synaptic mechanisms did a similarly detailed study, they might find some direct effects there as well.

    It is difficult to isolate pre- and post-synapses since, of course, they relate to each other. APP certainly traffics to both ends of the neuron and Aβ also localizes to both ends. What APP and Aβ are doing at synapses remains unclear, although this study provides new insights on acute pre-synaptic effects. They interpret that extracellular Aβ is the driving force behind the synaptic enhancement that they observed, although in their only experiments directly applying extracellular Aβ (Supplementary Fig. 5), they apparently required several orders of magnitude higher levels of the peptide than they find released endogenously. Our recent study showed that neprilysin also regulates the intraneuronal pool of Aβ (Tampellini et al., 2009).

    At the end of their discussion, the authors then turn to how their work might relate to AD. Specifically, they look at the potential deleterious effect of having too much Aβ in the setting of more prolonged (48 hours) inhibition of neprilysin. Interestingly, and in agreement with our previous publication on transgenic compared to wild-type neurons (Almeida et al., 2005), they observed (in Supplementary Fig. 10) reduced but enlarged FM1-43 puncta/pre-synapses.


    . Synaptic activity reduces intraneuronal Abeta, promotes APP transport to synapses, and protects against Abeta-related synaptic alterations. J Neurosci. 2009 Aug 5;29(31):9704-13. PubMed.

    . Beta-amyloid accumulation in APP mutant neurons reduces PSD-95 and GluR1 in synapses. Neurobiol Dis. 2005 Nov;20(2):187-98. PubMed.

  3. This is quite an interesting paper. Abramov and colleagues carefully investigated the role of Aβ in presynaptic function, and found that Aβ works as a positive regulator of transmitter release. This is the first report on a physiological role of Aβ as far as I know. Because increased Aβ enhances basal transmitter release, GABA signals may be increased in APP Tg mice to keep a balance between excitation and inhibition of neurons as Mucke reported (see Palop et al., 2007).

    In their figure 7d, Abramov and colleagues showed that presynaptic facilitation by bursts was diminished by both increased and reduced levels of Aβ. Thus, an appropriate level of Aβ may be required for maintaining healthy synaptic activity. This result may explain why Aβ immunization does not clearly halt a progression of dementia.

    I would like to hear Dr. Mucke's opinion.


    . Aberrant excitatory neuronal activity and compensatory remodeling of inhibitory hippocampal circuits in mouse models of Alzheimer's disease. Neuron. 2007 Sep 6;55(5):697-711. PubMed.

    View all comments by Akihiko Takashima

Make a Comment

To make a comment you must login or register.


News Citations

  1. Synapses Sizzle in Limelight of Symposium Preceding Neuroscience Conference, Orlando: Day 2

Paper Citations

  1. . Imaging synaptic vesicle exocytosis and endocytosis with FM dyes. Nat Protoc. 2006;1(6):2916-21. PubMed.

Further Reading

Primary Papers

  1. . Amyloid-beta as a positive endogenous regulator of release probability at hippocampal synapses. Nat Neurosci. 2009 Dec;12(12):1567-76. PubMed.