Too much Aβ weakens neural synapses by favoring long-term depression, but how? A report published January 18 in Nature Neuroscience suggests that the peptide does so by hijacking the neuron’s own mechanisms of synaptic plasticity. Scientists led by José Esteban from the Spanish Research Council (CSIC), César Venero from the National Distance Education University in Madrid (UNED), and Shira Knafo of the University of the Basque Country find that Aβ helps recruit more than the usual amount of PTEN—a protein that previously has been implicated in long-term depression—to the postsynaptic membrane. This, they report, tips the balance away from long-term potentiation and toward depression. The findings give new insight into how Aβ might mess with cognition. “We knew that plasticity was out of balance in Alzheimer’s disease, but we didn’t know why,” Esteban told Alzforum. “This is one mechanism that explains how it is impaired.”

“The exact mechanism of Aβ-induced depression of synaptic transmission is still mysterious,” said Daniel Choquet of the Université de Bordeaux, France. This study “furthers our understanding of this phenomenon, and opens a new way to preventing Aβ-induced synaptic toxicity.” (See full comment below.)

PTEN Flocks to Synapses.

Treatment with synthetic Aβ oligomers concentrates PTEN (bright green) in the spine heads of primary hippocampal neurons. [Courtesy of Knafo et al., Nature Neuroscience.]

PTEN—short for phosphatase and tensin homolog deleted on chromosome 10—is a lipid phosphatase best known for its role in tumor suppression throughout the body. The enzyme is also abundantly expressed in the adult brain. A decade ago it was first implicated in memory and synaptic long-term depression. LTD is a process that opposes long-term potentiation (LTP), preventing the latter from making synapses too strong (Wang et al., 2006). 

A few years back, Esteban and colleagues picked up on these early reports. They observed that upon LTD, PTEN collects in the postsynaptic membrane, where it counteracts signaling of phosphatidylinositol-(3,4,5)-trisphosphate (PIP3) by dephosphorylating it to PIP2 (Jurado et al., 2010; Arendt et al., 2014). This reduces the activation of some downstream signaling molecules such as AKT, and activates others, such as GSK3BETA. Through further signaling steps that remain unclear, this removed AMPA receptors from the synapse and weakened the neuron’s response to glutamate (Arendt et al., 2010). 

PTENs’ relocation to the postsynaptic membrane required it to bind to the scaffolding protein PSD-95, which anchors proteins to synapses via its 90-amino acid PDZ domain. PDZ domains are common to many proteins and are needed for protein-protein interactions (for a review, see Lee and Zheng, 2010). 

In the present study, first author Knafo wondered if this process contributed to the pathological depression driven by Aβ. To find out, she infused either the PTEN inhibitor VO-OHpic or vehicle into the brain ventricles of four-month-old PS/APP mice or their wild-type littermates over three to four weeks. She then tested the animals’ spatial memories using novel-object-location and contextual-fear-conditioning tasks. While untreated PS-APP mice were impaired on both, treated transgenic mice explored a displaced object as long as wild types, and froze just as much in an environment where they had received foot shocks.

Experiments in hippocampal slices also supported the idea that PTEN was required for the depressive effects of Aβ. Whether slices were bathed in synthetic Aβ or in Aβ secreted by cultured neurons, both LTP and basal synaptic transmission dropped. In contrast, pre-incubating the slices with VO-OHpic protected them from these effects. PTEN and Aβ appeared to work in the same pathway, as LTD was enhanced in hippocampal slices from mice that overexpressed PTEN but synthetic Aβ oligomers increased it no further.

How might Aβ influence PTEN? The mechanism is unclear, but the researchers found that synthetic Aβ added to primary hippocampal neurons concentrated PTEN from dispersed locations into spine heads (see image above). This would be similar to the synapse-directed movement of PTEN the group previously reported in LTD. Like in the previous study, the scientists again report that this relocation depends on PTEN’s interaction with PDZ-binding domains, though it is unclear which specific proteins bind PTEN. Hippocampal slices from mice that produced PTEN lacking the last five amino acids of the PDZ-binding domain maintained normal LTP in spite of Aβ treatment. Together, the data suggest to the authors that Aβ somehow facilitates binding between PTEN and its PDZ-domain containing partners, which draws more of the protein to the synapse, causing more LTD than usual.

Might these findings translate into a therapeutic strategy? PTEN catalytic activity helps prevent cancer, hence blocking it would be dangerous, said Esteban. However, it may be safe to target only the protein’s binding of PDZ domains. Knafo and colleagues designed a peptide matching the last eight amino acids of the PDZ-binding region from rat and mouse PTEN. A pulldown assay revealed that it competed with PTEN for binding to PDZ-containing proteins. Incubated with hippocampal slices from mice, the peptide blocked LTD and warded off ill effects of synthetic Aβ on LTP. Infusing the peptide into the brains of PS/APP mice for three to four weeks rescued deficits in the novel-object-location and contextual-fear-conditioning tasks, the scientists report.

While this provides a preclinical proof of concept, peptides are large, unstable, and do not cross the blood-brain barrier, Esteban said. In an effort to find something suited to the clinic, his group is looking for a small molecule that interrupts PTEN-PDZ-interactions and can be given intravenously. The scientists are also trying to figure out how Aβ relocates PTEN.

This study is interesting and thorough, with clear experimental outcomes, wrote Gary Landreth, Case Western Reserve University, Cleveland, Ohio. However, he added, “The story, while provocative, remains incomplete so long as the PDZ-containing effector(s) that mediate the synaptic effects of the therapeutic peptide remain unknown.” (See full comment below.)—Gwyneth Dickey Zakaib


  1. This is an interesting and thorough study, whose experimental outcomes are pretty clear.

    One question one could ask is why chronic treatment with the PTEN inhibitor is necessary to improve the behavioral outcomes. There is no change in pAKT in the chronic treatment paradigm and if the effects of PTEN are modifying local signaling at the synapse, then acute treatment should be sufficient to measure a treatment benefit.

    In the future, the authors may want to measure the PTEN substrate PIP3 in any of their preparations to verify drug action at this level.  

    The story, while provocative, remains incomplete so long as the PDZ-containing effector(s) that mediate the synaptic effects of the therapeutic peptide remain unknown. However, the authors report that their PTEN-PDZ peptide interacted with four of 10 PDZ domains, hence it may be a challenge to obtain this level of mechanistic detail. 

  2. The exact mechanism of Aβ-induced depression of synaptic transmission is still mysterious. As it is possibly at the base of memory loss in Alzheimer’s disease, it is particularly important to understand its molecular mechanism.

    Esteban et al. make an important step forward in demonstrating that inhibition of PTEN rescued normal synaptic function and cognition in cellular and animal models of Alzheimer’s disease. By finding that the Aβ induced synaptic toxicity and cognitive dysfunction depends crucially on the recruitment of PTEN to synapses through a specific protein interaction domain (PDZ-binding domain), they advance our understanding of this phenomenon and open a new way to prevent Aβ-induced synaptic toxicity through inhibition of PTEN-PDZ interactions. Of particular interest is the finding that peptides that mimic the PTEN-DZ ligand protect against Aβ-induced synaptic toxicity.

    Indeed, the authors found that it is sufficient to specifically target PDZ-dependent interactions of PTEN—preserving its catalytic activity, thus most of PTEN’s normal functions—to disengage the cascade of events triggered by Aβ, therefore preventing synaptic and cognitive dysfunction.

Make a Comment

To make a comment you must login or register.


Research Models Citations

  1. PS/APP

Paper Citations

  1. . The PTEN phosphatase is essential for long-term depression of hippocampal synapses. Neuromolecular Med. 2006;8(3):329-36. PubMed.
  2. . PTEN is recruited to the postsynaptic terminal for NMDA receptor-dependent long-term depression. EMBO J. 2010 Aug 18;29(16):2827-40. PubMed.
  3. . PTEN counteracts PIP3 upregulation in spines during NMDA-receptor-dependent long-term depression. J Cell Sci. 2014 Dec 15;127(Pt 24):5253-60. Epub 2014 Oct 21 PubMed.
  4. . PDZ domains and their binding partners: structure, specificity, and modification. Cell Commun Signal. 2010 May 28;8:8. PubMed.

Further Reading


  1. . Activation of PI3-kinase is required for AMPA receptor insertion during LTP of mEPSCs in cultured hippocampal neurons. Neuron. 2003 May 22;38(4):611-24. PubMed.
  2. . LTP inhibits LTD in the hippocampus via regulation of GSK3beta. Neuron. 2007 Mar 1;53(5):703-17. PubMed.

Primary Papers

  1. . PTEN recruitment controls synaptic and cognitive function in Alzheimer's models. Nat Neurosci. 2016 Mar;19(3):443-53. Epub 2016 Jan 18 PubMed.