The discovery that synaptic activity releases amyloid-β (Aβ) into the interstitial fluid between neurons raised a bit of a conundrum. How can brain activity simultaneously protect against Alzheimer’s and elevate Aβ? Is the “use it or lose it” mantra flawed? Not necessarily. In the August 3 Journal of Neuroscience, researchers led by John Cirrito at Washington University, St. Louis, Missouri, report that the relationship between neuron activity and Aβ is not simply a linear one. If neuron activity is high enough, it suppresses production of the peptide via a complex signaling pathway, according to the research. The findings may help explain why areas of the brain that are most vulnerable to Aβ accumulation might be protected by increased brain activity.

“The findings were a complete surprise,” Cirrito told ARF. First author Deborah Verges and colleagues were looking at the effect of infusing NMDA, a glutamate receptor agonist, into the brains of transgenic APP/PS1 mice. Instead of the expected increase in interstitial fluid (ISF) Aβ, they saw the opposite. “What was supposed to be a three-day experiment then turned into months,” said Cirrito. During that time, the researchers found that the amount of NMDA was crucial to the Aβ response. At lower doses of the agonist (~0.5 μM) ISF Aβ jumped 1.5- to twofold, but higher doses of NMDA (20 μM and above) quelled ISF Aβ to well below baseline levels.

The researchers used a microdialysis probe to deliver the agonist. The tiny probe is inserted into the brain of a living mouse, where it takes up solutes, such as Aβ, for analysis. Chemicals added to the probe can be delivered to the brain in a reverse of the process. The method was perfected by Cirrito when working with David Holtzman, also at Washington University. Though the researchers do not know exactly how much NMDA reached the receptors in the brain using this reverse dialysis, they found that it was sufficient to induce neuronal activity in the vicinity of the probe, as measured by electroencephalography. Even with only 1 μM NMDA in the probe, they recorded more electrical spikes than normal. Spike frequency and intensity grew as the researchers raised the amount of NMDA. At 250 μM, the mice exhibited brief periods of epileptiform activity over three hours. To check if reductions in Aβ might be due to neuronal damage, the authors washed out the NMDA and followed what happens in the ISF. Aβ levels recovered to near normal within a day, suggesting that neurons in the vicinity were functioning normally.

Flooding the areas surrounding the microdialysis probe with NMDA may not be the best mimic of physiological glutamate receptor activation, Cirrito admitted. But when the researchers added an NMDA receptor antagonist to the probe, they found it boosted ISF Aβ, indicating that endogenous activation of these receptors normally tones down Aβ release.

How does NMDA receptor activation reduce Aβ? To answer that question, Verges and colleagues first looked to electrical activity. When they blocked action potentials with tetrodotoxin, low doses of NMDA failed to drive up ISF Aβ, indicating that release of the peptide into the fluid depends on electrical activity. In contrast, Aβ levels still dropped below baseline under tetrodotoxin and high NMDA, suggesting the agonist attenuated peptide release independently of electrical activity. Verges and colleagues then looked to biochemical pathways that might explain the effect. They found that calcium influx and the extracellular regulated kinase (ERK) signaling pathway were essential for reducing ISF Aβ. ERK is known to promote non-amyloidogenic processing of Aβ precursor protein by boosting α-secretase (Kojro et al., 2006) and attenuating γ-secretase activity (Kim et al., 2006). In fact, the researchers found that the NMDA receptor antagonist CPP not only elevated ISF Aβ, but it reduced activation of ERK cascade members ERK2 and MEK1/2, and it suppressed α-secretase. All told, the work suggests that, by promoting action potentials and driving intracellular signaling cascades, NMDA receptor activation can have opposite effects on Aβ production.

“I think that this [research] emphasizes the diversity in the ‘valence’ of the changes in Aβ generation that accompany signaling via different pathways,” wrote Sam Gandy, Mount Sinai Alzheimer’s Disease Research Center, New York, in an e-mail to ARF. Gandy was not involved in this work. Cirrito agrees, and has data extending the observation to other receptors as well. He told ARF that other neurotransmitters, including serotonin, work in a similar fashion.

How does this finding relate to Alzheimer’s disease, and to the use of memantine, an NMDA receptor antagonist that is approved for moderate to severe AD. “There is evidence that memantine can increase Aβ production in the short term, but we don’t know what goes on chronically,” said Cirrito. “The brain habituates and you can get paradoxical effects later on,” he added. Work in mice, for example, suggests that memantine protects against Aβ accumulation (see Scholtzova et al., 2008; Alley et al., 2010; Martinez-Coria et al., 2010). “Given this new data, I'm not sure that one can predict what memantine will do to human brain Aβ,” said Gandy. “You might just have to do the experiment.” In fact, Gandy suggested that memantine might be a good drug to try in prevention trials, using CSF measures to determine how it affects brain Aβ. Recent meta-analysis of clinical trial data suggests that the NMDA antagonist does little to help people with mild AD (see ARF related news story), but as a preventive measure it may be beneficial, suggested Gandy.

Cirrito and colleagues think that ERK activation might be an alternative way to achieve suppression of Aβ production, though they caution that cellular context dictates ERK downstream effects, and that not all activators may have the same outcome. Activating ERK through NMDA receptors is also not the best approach, they posit, since that could have multiple effects, including cytotoxic ones. “It is possible that activation of ERK via other means could be used to depress Aβ levels therapeutically,” they write.—Tom Fagan


  1. Fine work.

    View all comments by Takaomi Saido

Make a Comment

To make a comment you must login or register.


News Citations

  1. Research Brief: Meta-Analysis Finds Memantine Forgettable for Mild AD

Paper Citations

  1. . The neuropeptide PACAP promotes the alpha-secretase pathway for processing the Alzheimer amyloid precursor protein. FASEB J. 2006 Mar;20(3):512-4. PubMed.
  2. . ERK1/2 is an endogenous negative regulator of the gamma-secretase activity. FASEB J. 2006 Jan;20(1):157-9. PubMed.
  3. . Memantine leads to behavioral improvement and amyloid reduction in Alzheimer's-disease-model transgenic mice shown as by micromagnetic resonance imaging. J Neurosci Res. 2008 Sep;86(12):2784-91. PubMed.
  4. . Memantine lowers amyloid-beta peptide levels in neuronal cultures and in APP/PS1 transgenic mice. J Neurosci Res. 2010 Jan;88(1):143-54. PubMed.
  5. . Memantine improves cognition and reduces Alzheimer's-like neuropathology in transgenic mice. Am J Pathol. 2010 Feb;176(2):870-80. PubMed.

Further Reading

Primary Papers

  1. . Opposing synaptic regulation of amyloid-β metabolism by NMDA receptors in vivo. J Neurosci. 2011 Aug 3;31(31):11328-37. PubMed.