The researchers discovered the clustering phenomenon using microscopic flames of their own, that is, quantum dot-labeled Aβ oligomers. Quantum dots are small, intense fluorophores that allow the tracking of single particles, a technique Triller uses to observe how molecules diffuse across cell surfaces. He teamed up with Klein to adapt the strategy to studying the movement of Aβ peptides in neuronal membranes. Joint first authors Marianne Renner and Pascale Lacor and colleagues treated cultured hippocampal neurons with synthetic soluble Aβ oligomers (otherwise known as Aβ-derived diffusible ligands or ADDLs; see Chromy et al., 2003) that they first labeled with quantum dot (QD) fluorophores, or biotin—which they detected with QD-coated streptavidin.

The researchers found that, within an hour of application to neuronal cultures, the synthetic Aβ oligomers (mostly 60 kDa species) stick to the cell membrane and cluster near excitatory synapses. It appears the peptides are drawn into the clusters as they laterally diffuse through the cell membrane, not simply taken up from the culture medium, since the researchers rinsed excess Aβ from the cells shortly after applying the oligomer mixture. In fact, single particle tracking (SPT) confirmed that QD-labeled oligomers initially diffuse rapidly along the cell surface, but then quickly slow down, especially around synapses, where the fluorescence due to clustering grows most intense. The tracking suggests that the oligomers are initially free to flutter about the cell membrane, but then get stuck as they form large clusters around excitatory synapses.

What is it about synapses that attracts Aβ oligomers? Other investigators have previously reported that Aβ oligomers can bind to synaptic proteins such as the NMDA receptor subunit NR1 (see ARF related news story on De Felice et al., 2007). Renner and colleagues now report that NR1 and other synaptic proteins including NR2, GluR1, and the metabotropic glutamate receptor mGluR5, immunoprecipitated with Aβ oligomers added to hippocampal neurons. But whereas antibodies to mGluR5 reduced the amount of Aβ oligomers that bound to cultured neurons when applied together, mGluR1 and mGluR2 antibodies did not. Aβ binding fell by half in mGluR5+/- cells and almost totally disappeared in mGluR5-/- cells, suggesting the metabotropic receptor is—quite literally—a major sticking point for Aβ.

If Aβ oligomers cluster in the presence of mGluR5, what happens to the metabotropic receptor in the presence of Aβ? Renner and colleagues noticed that more of mGluR5 becomes detergent insoluble when Aβ oligomers are added, hinting that the receptor might be doing a little clustering of its own. The scientists tested this idea by labeling the receptor with a yellow fluorescent protein to trace its tracks. In the presence of Aβ, mGluR5 fluorescence became brighter at synapses. Diffusion of extrasynaptic mGluR5, determined by QD single particle tracking, dramatically slowed on addition of Aβ oligomers, and the surface area explored by freely diffusing receptors fell by 90 percent. The authors conclude that membrane-bound Aβ oligomers “by direct or indirect interactions, induce the dynamic redistribution of mGluR5 receptors to synapses.” In contrast, Aβ had no effect on distribution of other receptors, such as the GABA or AMPA glutamate receptors. The work provides a rationale for previous studies from Klein’s lab showing that synthetic Aβ oligomers bind to discrete puncta on neurons and activate Arc, which mediates mGluR5 signaling (see Lacor et al., 2004).

Triller told ARF that he believes that Aβ oligomers are acting like an extracellular scaffold that then recruits other proteins. “If you have a protein like Aβ that sticks or binds to a receptor and it stops diffusing, it may encounter other Aβ molecules and they then form microclusters, which get bigger and recruit more and more receptors,” he said.

One consequence of such clustering could be induction of long-term depression, which is induced by recruitment of mGluR5 to synapses and has been linked to Aβ toxicity (see ARF related news story and ARF news story on Li et al., 2009). To explore this idea, Renner and colleagues examined what happens to ionotropic glutamate receptors when Aβ oligomers are added to neurons. They saw loss of NR1 NMDA receptor subunits, which is in keeping with an LTD effect. In addition, more neurons exhibit spontaneous influx of calcium when treated with Aβ oligomers, which could be due to activation of mGluR5. Interestingly, simply cross-linking mGluR5 receptors in the absence of Aβ ablated NR1 subunits and induced calcium influx. The results suggest that Aβ-induced clustering activates mGluR5, leading to downstream effects. Supporting this idea, the researchers found that the mGluR5 antagonist SIB1757 protects neurons against mGluR5 crosslinking and the sequelae of Aβ clustering, including calcium influx and NMDA receptor loss.

“This is a very high-tech paper and a beautiful use of quantum dots,” Gunnar Gouras, Weill Cornell Medical Center, New York, told ARF. He also thought the link between mGluR5 recruitment in response to Aβ and LTD was interesting (see full comment below), but cautioned that other receptor dysfunction might be involved as well. Triller agreed that the clustering could be more widespread. “I think other proteins are likely involved because mGluR5 binds Homer, and Homer may be bound to other elements. So there is probably a molecular complex,” he told ARF. “In that complex, Aβ oligomers could be bound to prion proteins and other receptors.”

Whether the reported interactions may be an early event in AD pathology remains to be seen. Triller told ARF that one next step would be to repeat the experiments in organotypic slices and eventually in vivo to see if the findings hold up in physiological settings. As Gouras wrote in his comment, much of the focus nowadays is on intracellular Aβ, and the relevance of added extracellular Aβ is not clear. The other issue that researchers are grappling with is what form of Aβ is the most important physiologically (see ARF related news story). Different sizes and species of oligomers have been studied, and the 60 kDa oligomers made in vitro may or may not be the most relevant to AD. Dimers isolated from patient tissue are known to be toxic, for example (see ARF related news story on Shankar et al., 2008). Interestingly, those dimers also seem to work in a metabotropic glutamate receptor-dependent manner (see ARF related news story on Li et al., 2009). Triller suggested that mGluR5 might be worth examining as a potential therapeutic target for AD. Some mGluR5 antagonists are already being tested for other conditions, such as fragile X mental retardation (see ClinicalTrials.gov) and Parkinson disease (see Rylander at al., 2010).—Tom Fagan.

Reference:
Renner M, Lacor PN, Velasco PT, Xu J, Contractor A, Klein WL, Triller A. Deleterious effects of amyloid beta oligomers acting as an extracellular scaffold for mGluR5. Neuron 2010 June 10; 66:739-754. Abstract