Oligomers are widely believed to be the most neurotoxic Aβ species, but how they exert their harmful effects remains mysterious. Previously, researchers led by Stephen Strittmatter at Yale University, New Haven, Connecticut, reported that at least some toxicity is mediated by cellular prion protein. Now, in the July 22 Nature Neuroscience, Strittmatter and colleagues delineate the downstream signaling pathway. Using cell cultures, they show that the interaction of Aβ and prion protein activates Fyn kinase, which then modifies synaptic signaling through NMDA glutamate receptors. The authors present evidence that this mechanism underlies excitotoxicity and dendritic spine loss. Intriguingly, tau was previously shown to direct Fyn to synapses, so the new findings may help further tie together the roles of Aβ and tau in AD.

“I was impressed by the breadth of the study. [The authors] go a long way to validate this pathway,” said Michael Wolfe at Brigham and Women’s Hospital, Boston. The next step will be to explore the in-vivo relevance of this phenomenon, Wolfe suggested.

Strittmatter and colleagues first discovered that Aβ oligomers strongly bind membrane-anchored cellular prion protein (PrPC), and that genetically deleting PrPC protected cell cultures from the toxic effects of Aβ exposure (see ARF related news story). They followed up by crossing APP/PS1 AD mice with prion protein knockout animals (Prnp-/-). Lack of the prion protected the offspring against axon and synapse loss, memory impairment, and early death (see Gimbel et al., 2010). However, other groups reported different results. Researchers led by Gianluigi Forloni at the Mario Negri Institute for Pharmacological Research, Milan, Italy, injected both wild-type and Prnp-/- mice with synthetic Aβ oligomers and found that the animals developed identical cognitive deficits (see ARF related news story). By contrast, Michael Rowan and colleagues at Trinity College Dublin, Ireland, showed that patient-derived Aβ oligomers disrupted synaptic plasticity in rodents through prion protein (see ARF related news story). Beyond the controversy, one of the big unanswered questions in this research was how PrPC might mediate toxicity.

To look for that mechanism, first author Ji-Won Um applied Aβ oligomers for 15 minutes to cortical neuron cultures prepared from wild-type mice. In some experiments, Um and colleagues used 3 μM synthetic Aβ oligomers (prepared by the ADDL method of Bill Klein), and in other soluble species isolated from AD patients. Patient-derived Aβ bound to purified, immobilized PrPC, and could be detected by an antibody specific for synthetic oligomers, NU-4 (see Lambert et al., 2007), the authors report. In cell culture, both types of Aβ activated Fyn, a Src family kinase. Treating the neurons with an antibody to prion protein, or using neurons from Prnp-/- mice, blocked activation, demonstrating an essential role for PrPC in switching on Fyn. Because PrPC is anchored to the extracellular side of the membrane, and Fyn is intracellular, it is unlikely the two proteins directly interact, the authors note. Nonetheless, more Fyn co-immunoprecipitated with prion protein after Aβ treatment, suggesting the proteins may bind through a membrane-spanning intermediary. Other research has fingered the protein caveolin 1 as a possible link between Fyn and PrPC (see ARF related news story).

The researchers then looked for downstream targets of Fyn. Since PrPC was enriched in postsynaptic densities, the authors examined postsynaptic glutamate receptors. NMDA-type glutamate receptors play a crucial role in synaptic plasticity and memory formation, and are phosphorylated by Src family kinases. Phosphorylation of two NMDA receptor subunits increased fivefold after 20 minutes of Aβ treatment. In addition, NMDA receptors on the cell surface tripled, and calcium influx into the cell increased. This caused excitotoxicity and reduced cell health, the authors report. None of these changes occurred in cultures made from Prnp-/- or Fyn knockout animals, demonstrating a key role of this pathway.

Intriguingly, longer-term exposure to Aβ produced the opposite effect. After one to three hours, phosphorylation of NMDA receptor subunits dropped, receptors internalized, and intracellular calcium levels crashed. This may be due to the activation of STEP tyrosine phosphatase, which opposes Fyn and can also be activated by Aβ, the authors note. In support of this, they saw increased STEP activation as Fyn activity returned to baseline. As with the short-term effects, these longer-term changes did not occur in Prnp-/- or Fyn knockout neurons. The authors wondered whether these increases or decreases in calcium flux might lead to synapse loss, one of the hallmarks of AD. They exposed the neuronal cultures to Aβ for five hours, over which time, 10 to 15 percent of dendritic spines disappeared. By contrast, cultures made from Fyn or prion protein knockouts, or treated with an NMDA receptor antagonist, maintained their spines.

The caveat to these experiments, Forloni pointed out in an e-mail to Alzforum, is that “All the in-vitro data are produced with short exposure to oligomers. We do not know how long the mechanism remains.” For this reason, “The relevance of these results remains unclear,” Forloni wrote (see full comment below).

Since PrPC-mediated Aβ toxicity can lead to both increased and decreased calcium flux through NMDA receptors, what might happen in vivo during continuous exposure to Aβ? To begin to address this, the authors monitored seizure activity in 10-month-old APP/PS1 mice for 72 hours. They found that four out of 10 had at least one seizure during this time, indicating profound changes in synaptic activity. But when they crossed these AD mice with prion protein knockouts, none of the 12 offspring examined suffered any seizures, despite having elevated levels of Aβ in the brain. These mice also had normal lifespans, in contrast to the early death of many APP/PS1 animals. The results hint that the prion protein-Fyn-NMDA receptor pathway may be involved in epileptic-like signaling, but do not yet prove it, Wolfe noted. Epileptic seizures are common in some AD mouse models, and people with AD are at increased risk for seizures (see ARF related news story on Palop et al., 2007).

Other groups have linked Fyn kinase to Aβ toxicity (see Chin et al., 2004; Chin et al., 2005). Researchers led by Jürgen Götz at the University of Sydney, Australia, showed that tau targets Fyn to the NMDA receptor, where the kinase mediates Aβ’s excitotoxic effects (see ARF related news story). “In our study we did not investigate how Aβ might interact with cell surface receptors,” Götz wrote to Alzforum (see full comment below). The new data suggest this occurs through the prion protein pathway, he wrote. He added that the suppression of seizures that Strittmatter and colleagues see in their Prnp-/- AD mice is very similar to the seizure suppression Götz saw when he crossed AD mouse strains to tau knockout mice, suggesting the same mechanism is at work.—Madolyn Bowman Rogers


  1. Of all Src kinases, Fyn is assuming a critical role in mediating Aβ toxicity. In this Nature Neuroscience paper, Stephen Strittmatter and colleagues show that oligomeric forms of Aβ bind to PrPC, a protein on the extracellular side of the membrane, and thereby activate (by an as-yet unknown mechanism) Fyn, which is localized on the intracellular side. The findings do not rule out that Aβ also binds to other receptors and/or the lipid membrane in exerting, in part, its toxic effects.

    This study by Um et al. is very carefully done, employing APP transgenic and PrP knockout mice and several cellular systems. We previously showed that tau is critical in targeting Fyn to the dendritic spine, thereby mediating Aβ toxicity (Ittner et al., 2010). In our study we did not investigate how Aβ might interact with cell surface receptors and/or the plasma membrane. Um et al. identify PrP as the protein with which Aβ interacts to cause Fyn activation. They showed further that in human AD brain, the Aβ species interacting with PrPC is present at critical levels. They further analyzed NMDAR activation upon exposure to Aβ and found phosphorylation at T1472 that was absent on a Fyn knockout background.

    When the authors crossed APP/PSEN mutant mice onto a Prnp-null background, they found that this prevented spontaneous seizures (recorded with EEG) and fully extended lifespan. These data are very similar to what we (Ittner et al., 2010) and the Mucke lab (Roberson et al., 2007) reported for the susceptibility to pentylenetetrazol-induced seizures and lifespan when crossing our APP mutant strains onto a MAPT knockout background. or with mice expressing a truncated form of tau. Also, the survival curve of the APP/PSEN mice (Um et al.) is very similar to that of the J20 and the APP23 mice (Ittner/Roberson), pointing to shared pathomechanisms in the investigated mouse strains.

    One of the most interesting aspects of the current study, as also pointed out by the authors, is that toxicity occurs during a short time window, after which the Fyn counterplayer STEP is activated. In our study, we found that disrupting the NMDAR complex with a small peptide for eight weeks protected APP23 mice permanently from seizure susceptibility, memory impairment, and reduced lifespan, indicating that there are mechanisms in place (possibly by restructuring the NMDAR) that protect the mice from Aβ’s toxicity even after the small peptide is removed. It will be interesting to further examine the Aβ-PrPC-Fyn-tau pathway, and to understand better the modes of acute and chronic Aβ toxicity.


    . Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer's disease mouse models. Cell. 2010 Aug 6;142(3):387-97. Epub 2010 Jul 22 PubMed.

    . Reducing endogenous tau ameliorates amyloid beta-induced deficits in an Alzheimer's disease mouse model. Science. 2007 May 4;316(5825):750-4. PubMed.

    View all comments by Jürgen Götz
  2. This paper by the Strittmatter group describes a significant number of data; however, the relevance of these results remains unclear. All the in-vitro data are produced with short exposure to oligomers (15-20 minutes, sometimes hours). We do not know how long the mechanism remains. When the observation is prolonged in primary cultures to test toxicity, a negligible effect was found at two hours (10 percent of LDH increase and no changes in MTT; Fig. S8 and 5h and 5i) and no effect at 72 hours (S9). Also, the measure of spines gave a modest effect after five hours of exposure to oligomers (Fig. 6b). In addition, the mechanism responsible for the relationship between Aβo/PrP/Fyn and NMDA receptors is elusive.

    While the number of experiments with different conditions and the continuous referral to supplemental information does not help the reader, this is an interesting scientific issue.

    In-vivo experiments show that the epileptic status of the APPSwe/PSenΔE9 mice was strongly attenuated when PrPC was nullified. This result may indicate a potential interaction between the expressed transgene and PrPC, but epileptic discharge is strictly associated with this specific transgenic mouse. Other APP single, double, and triple Tg mice, as well as most AD subjects, do not show this condition. In our lab we have conducted a specific investigation into this aspect, comparing three different Tg mice (PDAPP, TASTPM, and APP/PS2/TAU) in the framework of EC-IMI-Pharmacog project, with at least one week of continuous EEG recordings. We have not seen evidence of epilepsy in any animals.

    In conclusion, the results presented in this paper confirmed the high-affinity association between Aβo and PrPC. Downstream, and in the specific conditions used, Fyn signaling might be activated. However, as mentioned by the authors in the text, and in contrast with the picture of Fig. S10, the physical interaction of PrPC with Fyn is unlikely since PrPC is attached extracellularly to the membrane with a GPI anchor, while Fyn is inside the cell. Furthermore, the timing and the quantitative effects associated with the Aβo/PrPC/Fyn mechanism, in terms of synaptic dysfunction, interaction with NMDA receptors, and toxicity, leave doubts as to the relevance of the phenomenon.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Keystone: Partners in Crime—Do Aβ and Prion Protein Pummel Plasticity?
  2. Model Shows Oligomers Impair Memory, Questions Role of Prion Protein
  3. Patient-Derived Aβ Needs Prion Protein to Harm Synapses
  4. Normal Prion Protein Signals with Fyn
  5. Do "Silent" Seizures Cause Network Dysfunction in AD?
  6. Honolulu: The Missing Link? Tau Mediates Aβ Toxicity at Synapse

Paper Citations

  1. . Memory impairment in transgenic Alzheimer mice requires cellular prion protein. J Neurosci. 2010 May 5;30(18):6367-74. PubMed.
  2. . Monoclonal antibodies that target pathological assemblies of Abeta. J Neurochem. 2007 Jan;100(1):23-35. PubMed.
  3. . 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.
  4. . Fyn kinase modulates synaptotoxicity, but not aberrant sprouting, in human amyloid precursor protein transgenic mice. J Neurosci. 2004 May 12;24(19):4692-7. PubMed.
  5. . Fyn kinase induces synaptic and cognitive impairments in a transgenic mouse model of Alzheimer's disease. J Neurosci. 2005 Oct 19;25(42):9694-703. PubMed.

External Citations

  1. APP/PS1 AD mice

Further Reading

Primary Papers

  1. . Alzheimer amyloid-β oligomer bound to postsynaptic prion protein activates Fyn to impair neurons. Nat Neurosci. 2012 Sep;15(9):1227-35. Epub 2012 Jul 22 PubMed.