In large quantities, Aβ poisons synapses; in small amounts, it can enhance plasticity. What explains this dual action? In the October 22 Journal of Neuroscience, researchers led by Robert Nichols at the University of Hawaii at Manoa, Honolulu, offer a clue. They found that the beneficial synaptic activity of Aβ localizes to a small N-terminal fragment, Aβ15, which activates nicotinic acetylcholine receptors and enhances long-term potentiation (LTP). Injection of this fragment into wild-type mice improved learning. Moreover, treating hippocampal slices with Aβ15 before adding Aβ42 preserved synaptic plasticity. The fragment exists in the cerebrospinal fluid of healthy adults, suggesting that it could play a role in normal cognition.

“The results are exciting,” said Jerrel Yakel at the National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina. “This paper seems to indicate that Aβ15 can block the deleterious effects of full-length Aβ. That gives us hope for designing small-molecule therapeutics.” Yakel has collaborated with Nichols in the past, but was not involved in the current study.

Secretases in the brain chop up amyloid precursor protein (APP) in a number of ways, producing a bewildering variety of Aβ fragments. Sequential cuts by β-secretase (BACE1) and γ-secretase yield amyloidogenic peptides such as Aβ42, while α-secretase followed by γ-secretase cleavage generates fragments that are believed to be less harmful. Recently, researchers uncovered a third pathway, in which a snip by BACE1 precedes cleavage by α-secretase (see Portelius et al., 2011). This produces Aβ14, Aβ15, and Aβ16, which span the N-terminal portion of Aβ42.

These fragments exist in normal CSF (see Portelius et al., 2007), are elevated in Alzheimer’s disease (see Portelius et al., 2010), and build up after γ-secretase inhibition (see Apr 2010 news story). However, researchers do not know what function, if any, these small peptides serve in the brain. Nichols and colleagues had previously observed that full-length Aβ42 activated the α7 nicotinic acetylcholine receptor (see Dougherty et al., 2003Tong et al., 2011), leading them to wonder if the smaller fragments might share this activity.

To investigate this, joint first authors James Lawrence and Mei Tong treated cultured neuroblastoma cells with Aβ15. At picomolar concentrations, the peptide activated both α7 and α4β2 nicotinic acetylcholine receptors, and did so more strongly than did Aβ42. Next, the authors looked at the effect on synaptic plasticity in wild-type mouse hippocampal slices. At minuscule, femtomolar concentrations, Aβ15 roughly doubled LTP after electrical stimulation. Curiously, larger picomolar and nanomolar concentrations had no effect. The reason for the loss of activity at higher doses remains unexplained, Nichols wrote to Alzforum. However, a similar phenomenon has been observed for Aβ42. There, picomolar concentrations enhance LTP in slices, while nanomolar and higher levels inhibit it (see Jan 2009 news story).

Commentators agreed this mystery deserves further exploration. “That low and high concentrations of Aβ (or its fragments) can have opposing effects has been underappreciated. Moving forward, it will be important to keep in mind the effect of a very wide range of Aβ concentrations on plasticity, as well as have a renewed interest in all APP fragments,” John Cirrito at Washington University in St. Louis wrote to Alzforum (see full comment below).

Could different segments of Aβ42 be responsible for enhancing and inhibiting neurotransmission? To pin down the portion that activates acetylcholine receptors, Nichols and colleagues mutated various residues of Aβ15. They localized the agonist activity to a segment six amino acids long, representing residues 10-15. The finding may help explain Aβ42’s opposing effects, Yakel noted. “The 10-15 hexapeptide might be responsible for the enhancing effect, and another part of Aβ42 might be the inhibiting part,” he told Alzforum.

What happens when both Aβ15 and Aβ42 are present? The shorter peptide seems to block the effects of the larger one, the authors found. Pretreating slices with 500 nM Aβ15 prevented the drop in LTP normally seen after adding 500 nM Aβ42. In slices from APPswe AD model mice, Aβ15 likewise rescued LTP.

Does the effect of Aβ15 on LTP translate into better memory in living animals? The authors injected picomolar concentrations into the hippocampi of wild-type mice. Because the peptide diffuses from the point of injection, the effective concentration in the hippocampus probably ended up in the femtomolar range, similar to the in vitro experiments, Nichols noted by email. The treatment boosted memory in a fear-conditioning trial, leading the animals to freeze nearly twice as often when they were put in a box where they had previously felt an electrical shock.

Many questions remain, including how much of the peptide a healthy brain produces, and how that changes in Alzheimer’s disease or in aging. “In particular, we wonder whether Aβ15 levels at the synapse fall in AD,” Nichols wrote to Alzforum.

Other questions revolve around the effects of secretase inhibition. Several BACE inhibitors are currently in clinical trials (see Dec 2013 news storyOct 2014 news story), and γ-secretase modulators are being developed (see Apr 2011 conference story; Aug 2013 news story). Knowing how these compounds alter Aβ15 levels might help predict the clinical impact of treatment, said Nichols. He plans to investigate this further.

Erik Portelius at the University of Gothenburg, Sweden, who first identified Aβ15, wrote to Alzforum, “From when we first discovered that Aβ15 increased in response to γ-secretase inhibitor treatment, we have wondered whether inducing the α-secretase pathway is ‘good or bad.’ Lawrence and colleagues clearly show that inducing this pathway is good.” (See full comment below.)—Madolyn Bowman Rogers


  1. From when we first discovered that the β/α generated fragment Aβ1-15 increased in response to γ-secretase inhibitor (GSI) treatment, we have wondered whether inducing the α-secretase pathway is “good or bad.” Lawrence and colleges clearly show that inducing this pathway is good, and their speculation on optimizing the production of Aβ1-15 is intriguing. With the knowledge that several GSIs have failed in clinical trials, some due to worsening in cognition, it is important to follow up the data presented and investigate how it translates into humans.

    View all comments by Erik Portelius
  2. This study by Lawrence et al. on the activity of the N-terminal fragment of Aβ raises some important questions regarding the physiological roles of APP cleavage products. As APP is cleaved, it makes sense that the products could serve as activity-dependent feedback with downstream effects that regulate the system. Whether this feedback is functioning properly (physiologically normal) or impaired (pathological) could influence disease conditions. There are well-documented cases of redundancy in many systems, and the set of experiments identifying the two histidines (his-13 and his-14) and the YEVHHQ region as the key component of the functional effects of the N-terminal domain of Aβ are particularly compelling. Although there are many more combinations of Aβ fragments to examine in this way, this is an important mechanistic finding.

    These results also raise many questions about modulatory effects of the full-length Aβ peptide and specific cleavage products that will be important to study in further depth. Is the improved conditioned fear performance seen on administration of low concentration of Aβ42 due to processing of the peptide to Aβ1-15? (Could Aβ1-15 be measured in tissue to see if Aβ1-42 is cleaved after injection?) Or, could the neuromodulatory activity of Aβ1-15 be retained in the full-length peptide under specific conformations (i.e., is the 1-15 region within Aβ1-42 accessible)? Do the positive effects seen on pretreatment with Aβ1-15 represent competition of the peptide with Aβ1-42, rather than Aβ1-15 signaling through a receptor that protects the system?  Would the same effects be seen by pretreating with an α7-nicotinic agonist? 

    The initial reports that low concentrations of Aβ could augment synaptic signaling seemed contrary to a large literature that high concentrations of Aβ potently inhibit synaptic plasticity. However, several independent groups, including nice studies by Dr. Nichols’ group, have now shown picomolar concentrations of Aβ do, in fact, augment synaptic plasticity and behavior. That low and high concentrations of Aβ (or its fragments) can have opposing effects has been underappreciated. Moving forward, it will be important to keep in mind the effect of a very wide range of Aβ concentrations on plasticity, as well as have a renewed interest in all APP fragments.

    View all comments by John Cirrito

Make a Comment

To make a comment you must login or register.


News Citations

  1. Sweet 16: Novel APP Processing Pathway and a New Biomarker?
  2. Aβ Boosts Memory; Mint/X11 Proteins Boost Aβ?
  3. Merck BACE Inhibitor Clears a Safety Hurdle, Gets New Trial
  4. Lilly Teams Up With AstraZeneca for BACE Inhibitor Phase 2/3 Trial
  5. Barcelona: Allosteric γ Modulation Moves Toward Clinic
  6. More Evidence that γ-Secretase Modulators Spare Essential Substrates

Research Models Citations

  1. APPSwe

Paper Citations

  1. . A novel pathway for amyloid precursor protein processing. Neurobiol Aging. 2011 Jun;32(6):1090-8. PubMed.
  2. . Characterization of amyloid beta peptides in cerebrospinal fluid by an automated immunoprecipitation procedure followed by mass spectrometry. J Proteome Res. 2007 Nov;6(11):4433-9. PubMed.
  3. . Distinct cerebrospinal fluid amyloid beta peptide signatures in sporadic and PSEN1 A431E-associated familial Alzheimer's disease. Mol Neurodegener. 2010;5:2. PubMed.
  4. . Beta-amyloid regulation of presynaptic nicotinic receptors in rat hippocampus and neocortex. J Neurosci. 2003 Jul 30;23(17):6740-7. PubMed.
  5. . Role of key aromatic residues in the ligand-binding domain of alpha7 nicotinic receptors in the agonist action of beta-amyloid. J Biol Chem. 2011 Sep 30;286(39):34373-81. PubMed.

Further Reading

Primary Papers

  1. . Regulation of presynaptic Ca2+, synaptic plasticity and contextual fear conditioning by a N-terminal β-amyloid fragment. J Neurosci. 2014 Oct 22;34(43):14210-8. PubMed.