15 March 2007. Two new publications add to growing evidence that toxic Aβ oligomers play havoc with neurons by interfering with N-methyl-D-aspartate (NMDA)-type glutamate receptors (NMDARs) and Ca2+ flow, but they suggest different toxic mechanisms. The findings emphasize the pleiotropic effects of this toxic peptide.
Bernardo Sabatini and colleagues at Harvard Medical School, Boston, and University College Dublin, Ireland, report that dimers and trimers of Aβ decrease NMDAR-mediated Ca2+ influx, and propose that this promotes long-term depression and loss of dendritic spines. William Klein and colleagues at Northwestern University, Chicago, Illinois, and the Federal University of Rio de Janeiro, Brazil, report that Aβ oligomers cause an increase in NMDAR-dependent Ca2+ influx and a subsequent induction of reactive oxygen species (ROS). The two groups also find that NMDA antagonists have different effects. While Klein’s group reports that the NMDAR channel blocker memantine and the competitive glutamate inhibitor APV both protect against Aβ-induced ROS, Sabatini and colleagues found that the competitive inhibitor CPP can mimic the effect of Aβ oligomers, leading to spine loss. The effects of NMDA receptor antagonists are intriguing, given that memantine is approved for the treatment of AD in many countries and that its physiological effects are not completely understood.
Synapse loss is the best correlate of clinical deterioration in Alzheimer disease patients, and many research groups have tried to reconcile those losses with the toxic properties of various amyloid-β (Aβ) species (see Alvarez, 2007, a new review on the plasticity of dendritic spines). Sabatini and colleagues examined how synapses react to Aβ oligomers that are naturally secreted into culture medium by APP transfected cells—a technique pioneered by Dominic Walsh, now at UCD, when he worked in coauthor Dennis Selkoe’s lab at Harvard. In yesterday’s Journal of Neuroscience, first author Ganesh Shankar and colleagues report that while Aβ monomers have no effect on spine density in organotypic hippocampal slices prepared from rat brain, low molecular weight oligomers (primarily dimers and trimers) cause a progressive loss of spines, resulting in about 75 percent loss over 15 days. The oligomers also reduce spontaneous mini-excitatory postsynaptic currents, indicating that the loss of spines has physiological relevance. Some of this work was reported at the Eibsee conference in Germany last year (see ARF related news story).
Shankar and colleagues also found that acute treatment of the hippocampal slices with Aβ oligomers had no effect. This suggests that spine loss is a slow process. It is not irreversible, though, as spine density came back to normal after removing the oligomers or adding an Aβ antibody (6E10). To test if this sluggish toxicity is mediated through receptor signaling, the researchers tried blocking the effect with both the nicotinic acetylcholine receptor blocker α-Bungarotoxin and the NMDAR antagonist CPP. While the α-Bungarotoxin had no effect, high doses of CPP (20 μM) prevented spine loss, suggesting NMDARs may be involved. Since NMDAR signaling is mediated by Ca2+ influx, Shankar and colleagues measured micro Ca2+ transients at spine heads using two-photon laser scanning microscopy and a calcium-sensitive fluorophore. They found that Aβ oligomers indeed reduced the frequency of Ca2+ transients. The scientists also determined that expression of an inactive form of the protein cofilin prevented oligomer-induced spine loss. So did FK506, a drug that prevents activation of calcineurin, a calmodulin-dependent protein phosphatase. Both cofilin and calcineurin have been implicated in the reduction of dendritic spines and NMDAR-mediated long-term depression (LTD) via restructuring of the synaptic cytoskeleton. All told, the data suggest that Aβ oligomers reduce spine density by partially blocking NMDARs, leading to LTD. This theory is supported by the observation that adding enough CPP to achieve 50 percent inhibition of the NMDARs (200-400 nM) mimicked the effect of Aβ oligomers. Previous work from Greg Cole’s lab at the University of California, Los Angeles, has also linked cofilin, which is involved in remodeling of the cytoskeleton, with synapse loss in AD (see ARF related news story).
Klein’s group tested the effects of Aβ using synthetic preparations that were allowed to oligomerize. In a recent paper they described how these oligomers, or ADDLS (Aβ-derived diffusible ligands), bind to NMDA receptors and reduce spine density (see Lacor et al., 2007 and ARF related news story). In an in press article in the February 16 Journal of Biological Chemistry, they report that those same oligomers can promote ROS, also through binding to NMDARs. Mitochondria-produced ROS are a common theme in many neurodegenerative diseases including Alzheimer’s (see ARF related news story).
First author Fernanda De Felice and colleagues used an ROS-sensitive fluorescent probe to determine if Aβ oligomers have any effect on hippocampal cultures. They found a robust increase in ROS as early as 3 to 4 hours after exposing neurons to oligomers, and much of the ROS-induced fluorescence occurred in the dendritic arbors. Given previous links between Aβ and glutamate receptors, the researchers tested if blocking the NMDARs might prevent the toxic effect. They found that both an antibody to the extracellular domain of the NR1 subunit of the NMDA receptor and NMDAR antagonists, including the channel blockers memantine and MK-801 and the competitive inhibitor APV, blocked Aβ-induced ROS.
De Felice and colleagues found that Aβ oligomers bound to the NMDA receptors, since immunoprecipitation of the NR1 receptor subunit from neuronal extracts also pulled down Aβ. Interestingly, while the NR1 antibody and the competitive inhibitor APV prevented oligomer binding to neurons, the channel blockers did not, suggesting that the oligomers might bind near the glutamate site, rather than on it.
Turning to downstream effects, the researchers measured the effect of Aβ oligomers on intracellular Ca2+ levels. They found that the oligomers elicited a rapid increase in Ca2+, which could be prevented by memantine or the NR1 antibody.
At first blush these two reports may seem to be at odds with each other. On the one hand Aβ oligomers seem to slowly inhibit NMDAR-mediated Ca2+ influx, causing LTD and decreased spine loss, while on the other, they appear to quickly cause an inrush of Ca2+ and an increase in toxic ROS. Perhaps these different effects are not mutually exclusive. The oligomers studied by Shankar and colleagues are predominantly dimers and trimers, whereas those formed by Klein’s group are much larger, with only those greater than 50 kDa binding with high affinity to synaptosomes. These are reminiscent of the 56 kDa Aβ dodecamer, Aβ*56, that was recently reported by Karen Ashe’s group as being the best correlate for memory loss in mouse models of AD (see ARF related news story). But smaller species may well play an important role at various stages of pathology.
One thing both papers agree on is that the toxic effects respond to agents that sequester Aβ or interfere with its aggregation. Klein and colleagues were able to prevent the toxic effects of their Aβ preparations by adding a monoclonal antibody, NU1, which recognizes Aβ species from postmortem Alzheimer disease tissue. Shankar and colleagues could similarly block synapse loss by using an antibody (6E10) or scyllo-inositol (AZD-103), a small molecule being developed as a potential treatment for AD. This inositol prevents Aβ aggregation and was recently shown to rescue synaptic function in mouse hippocampal tissue (see ARF related news story).
The value of small molecule inhibitors of Aβ aggregation was also addressed by a paper in the February 6 Journal of Biological Chemistry from Charlie Glabe and colleagues at the University of California, Irvine. First author Mihaela Necula and colleagues have used small-molecule inhibitors of aggregation to address a fundamental question in Aβ biology, namely, do monomers have to form oligomers on the way to becoming fibrils, or can the two assembly pathways coexist independently? Their findings indicate that fibrils can form independently.
Necula and colleagues used the A11 oligomer-specific antibody previously developed in the Glabe lab (see ARF related news story) to monitor Aβ aggregation. From a screen of 29 molecules reported to bind to amyloid or inhibit aggregation, the researchers found three classes of compounds. One prevented oligomerization but not fibrillization, a second did the opposite, and the third prevented both processes. The data indicate that oligomers and fibrils can occur separately. These compounds may not only be useful for studying Aβ aggregation, but they may also serve as lead compounds for drug development. The most potent compound tested, azure C, was effective in the 100 nM range.—Tom Fagan.
Shankar GM, Bloodgood BL, Townsend M, Walsh DM, Selkoe DJ, Sabatini BL Natural oligomers of the Alzheimer amyloid beta-protein induce reversible synapse loss by modulating an NMDAR-dependent signaling pathway. J. Neurosci. Abstract
De Felice FG, Velasco PT, Lambert MP, Viola K, Fernandez SJ, Ferreira ST, Klein WL. Abeta oligomers induce neuronal oxidative stress through an NMDA receptor-dependent mechanism that is blocked by the Alzheimer’s drug memantine. J. Biol. Chem. 2007, February 16. Abstract
Necula M, Kayed R, Milton S, Glabe CG. Small-molecule inhibitors of aggregation indicate that amyloid beta oligomerization and fibrillization pathways are independent and distinct. J. Biol. Chem. 2007, February 6. Abstract