Notwithstanding some redeeming features (ARF related New Orleans story) soluble Aβ oligomers have an overwhelmingly bad reputation as suspected mediators of synaptic dysfunction in Alzheimer’s. A number of accusing fingers have pointed to the havoc they wreak in vitro and in vivo (see, for example, ARF related news story and ARF related story), and at the 33rd Annual Meeting of the Society for Neuroscience, several groups incriminated these oligomers more deeply. Here are selected highlights:

Sylvain Lesne, working with Karen Hsiao Ashe and colleagues at the University of Minnesota in Minneapolis, further analyzed Aβ oligomers in an attempt to find out which are the precise species that impair memory in young Tg2576 mice, even before these mice develop plaques (Kotilinek et al., 2002). Using a detergent, the researchers extracted soluble proteins from these APP-overproducing mice throughout life, from embryonic stages to the elderly age of around 20 months, and then identified Aβ oligomers using immunoblots (SfN abstract 772.1).

The scientists were surprised to find trimers present throughout all ages tested. Their amounts did not change with advancing age, suggesting they are not behind the memory problems. Multimers of those trimers, namely hexamers and dodecamers, appeared with a time course paralleling the gradual loss of memory. By contrast, tetramers began appearing at 13 months, when the memory loss was already in full swing, and their emergence did not worsen it. This leads the Ashe group to suspect that trimers are the fundamental form of Aβ oligomers in their mice, and that multimeric forms of them are at the root of the mice’s cognitive problems.

Dennis Selkoe of Brigham and Women’s Hospital Boston, described how his group has further characterized the soluble Aβ oligomers that are made intracellularly and then secreted by Chinese hamster ovary (CHO) cells overexpressing human APP (667.7). This group considers these secreted Aβ oligomers more “natural” than synthetic oligomers, and it appears that they may be structurally different and more potent, too. Selkoe presented new data indicating that of several proteases known to degrade Aβ monomers (neprilysin, IDE, and plasmin were tested), only plasmin readily breaks down oligomers, as well. Selkoe also showed data on two synthetic small molecules made by pharmaceutical companies that can inhibit the formation of Aβ oligomers in medium and in cells.

He then cited prior work by Dominic Walsh, Igor Klyubin, and others showing that injecting low nanomolar concentrations of these oligomers from conditioned medium inhibits hippocampal LTP in live rats (see ARF related news story). In that paper, the scientists had shown that γ-secretase inhibitors could prevent this oligomer-induced LTP block. In New Orleans, Selkoe presented new data indicating that anti-Aβ antibodies, as well, can neutralize this LTP block, suggesting that this might be a third way (besides microglial plaque clearance and passive, peripheral sink effects) by which AD immunotherapy might eventually become useful (see ARF related New Orleans story).

Recently, this group has developed a size exclusion chromatography protocol to isolate Aβ oligomers from this conditioned medium and shown that four different Aβ antibodies recognize oligomers collected in this procedure. This allowed them to test if these oligomers had biological activity affecting learned behavior. To do that, Selkoe and Walsh collaborated with James Cleary, Ashe, and colleagues in Minneapolis. The researchers chose a rat behavior test called the Alternating Lever Cyclic Ratio (ALCR) lever-pressing procedure, in which rats are trained to keep switching between two levers, pressing them until they receive a food pellet. Rats can make perseveration errors, i.e., they don’t alternate levers properly any more, and switching errors, i.e., they switch to the wrong lever. The scientists consider this test more sensitive than previous instruments, and other researchers including Cleary have used it before to test Aβ-induced learning deficits in rats and the ability of NSAIDs to protect against them (see O’Hare et al., 1999; Richardson et al., 2002). Cleary microinjected into rat lateral brain ventricles the conditioned medium, Aβ oligomers, or Aβ monomers isolated by chromatography. Conditioned medium and oligomers, but not monomers, caused the rats to commit both kinds of errors (see also 772.11). The doses used were physiologically comparable to those seen in AD brain.

Taken together, this leads Selkoe to conclude that a biochemically isolated, defined species of Aβ oligomer—in the absence of monomer and fibrils—can disrupt hippocampal LTP and a learned behavior. Many prior studies had wrestled with the technical difficulty of using mixtures of Aβ species, leaving doubt over which component may have had which effect. Despite this new data, many questions remain about the synaptotoxic mechanisms of Aβ oligomers, and how they could be targeted therapeutically, Selkoe added.

Qinwen Wang, Mike Rowan, and Roger Anwyl of Trinity College in Dublin, Ireland, in collaboration with Selkoe and Walsh (who is based both at University College, Dublin, and Brigham and Women’s Hospital Boston), looked closely at how Aβ oligomers affect LTP (904.10). They performed electrophysiology on rat brain slices harboring the perforant pathway from the entorhinal cortex to dentate granule cell synapses in the hippocampus, a nerve projection that shows damage early on in AD. Both synthetic Aβ oligomers and oligomers secreted by APP-overexpressing CHO cells indeed inhibited the induction of LTP that normally follows high-frequency stimulation, but the natural oligomers were much more potent, Wang et al. found.

Since phosphorylation of excitatory neurotransmitter receptors might play into the mechanism underlying this phenomenon, the scientists next asked which kinases might affect it. They applied the Aβ oligomers in the presence of various kinase inhibitors, and reported that two different inhibitors of c-Jun N-terminal kinase (JNK) counteract the LTP inhibition caused by Aβ oligomers. The Cdk5 inhibitors butyrolactone and roscovitine did, as well, as did a p38MAP kinase inhibitor, but a p42/44 MAP kinase inhibitor did not. All of these kinases have been previously implicated in Alzheimer’s in various ways, but their precise role, if any, on learning and memory remains poorly understood (see, for example, Liu et al., 2001; Fisher et al., 2002, Savage et al., 2002; Sun et al., 2003.

Wang et al. presented one more experiment addressing the question of which transmitter receptor might be at play here. They suggest that the metabotropic glutamate receptor (mGluR) might be one, as two antagonists against it prevent the Aβ oligomer effect. By contrast, an angatonist of the α7 nicotinic acetylcholine receptor (α7nAChR) did not (but see alsoARF related New Orleans story). Together, this suggests that Aβ oligomers inhibit LTP through the mGluR5 receptor and that the kinases JNK, Cdk5, and p38 MAP kinase might be involved. (Cdk5 is also suspected of phosphorylating the NR2A subunit of the NMDA receptor complex, see ARF related news story). Again, this update on oligomer news can’t be comprehensive; other noteworthy presentations on Aβ oligomers included 20.13, 133.7, 841.2, 525.22, 772.4, 772.5, 772.9, 841.21, and 876.5. As always, additions and corrections are welcome. You can view abstracts mentioned in this story at the SfN/ScholarOne website.—Gabrielle Strobel.


Make a Comment

To make a comment you must login or register.

Comments on this content

  1. I would like to correct a brief comment on our SfN abstract. Although we wrote that trimers do not seem to play a role in the cognitive decline observed in Tg2576, our current results (presented in the oral presentation during the SFN meeting) demonstrate the opposite. Because 4-5 month-old Tg2576 mice do not display spatial reference memory deficits as compared to their non-transgenic littermates, we believed that trimers (the only Low Molecular Weight Abeta oligomers detected at that age) weren’t involved. However we understood our error when we looked at another mouse line, called Tg5469, overexpressing human wild-type APP at identical levels as APP(Swe) in Tg2576. The former mouse displayed enhanced memory abilities compared to non-transgenic littermates and produces Abeta at considerable lower levels than in Tg2576.
    Thus, we believe that the presence of trimers decreases spatial reference memory in 4-5 month-old Tg2576 to levels similar to the ones of non-transgenic littermates. These data plus the characterization of Abeta oligomerization in Tg2576 performed both in vivo and in vitro indicate that trimeric assemblies of Abeta may be the building block for High Molecular Weight Abeta oligomers and the source of memory impairment observed in these transgenic mice.

Comments on Primary Papers for this Article

No Available Comments on Primary Papers for this Article


News Citations

  1. New Orleans: Aβ Oligomers and Memory: Now They Are Good…
  2. Oligomers in AD: Too Much of a Bad Thing?
  3. Amyloid Oligomer Antibody—One Size Fits All?
  4. Earliest Amyloid Aggregates Fingered As Culprits, Disrupt Synapse Function in Rats
  5. New Orleans: Immunotherapy—The Game Is Still in Town
  6. New Orleans: Kelly Dineley Reports from Satellite Social
  7. Tau Kinase Mediates Stroke Damage

Paper Citations

  1. . Reversible memory loss in a mouse transgenic model of Alzheimer's disease. J Neurosci. 2002 Aug 1;22(15):6331-5. PubMed.
  2. . Delayed behavioral effects following intrahippocampal injection of aggregated A beta (1-42). Brain Res. 1999 Jan 2;815(1):1-10. PubMed.
  3. . Behavioural and histopathological analyses of ibuprofen treatment on the effect of aggregated Abeta(1-42) injections in the rat. Brain Res. 2002 Nov 1;954(1):1-10. PubMed.
  4. . Regulation of cyclin-dependent kinase 5 and casein kinase 1 by metabotropic glutamate receptors. Proc Natl Acad Sci U S A. 2001 Sep 25;98(20):11062-8. PubMed.
  5. . Cyclin-dependent kinase 5 is required for associative learning. J Neurosci. 2002 May 1;22(9):3700-7. PubMed.
  6. . Activation of c-Jun N-terminal kinase and p38 in an Alzheimer's disease model is associated with amyloid deposition. J Neurosci. 2002 May 1;22(9):3376-85. PubMed.
  7. . P38 MAP kinase is activated at early stages in Alzheimer's disease brain. Exp Neurol. 2003 Oct;183(2):394-405. PubMed.

External Citations

  1. SfN/ScholarOne

Further Reading


  1. Alzheimer's Medal Goes to Pioneer in Diffusible Aβ
  2. Playing "Gotcha" with the Phantom: α-synuclein Oligos Spotted in Vivo