By Erene Mina. As evidence mounts against oligomers as culprits of Aβ neurotoxicity (see ARF related New Orleans story), their only defense hinges on the possibility that they may have some normal physiological function—even a slightly neurotrophic effect—on cognition which has yet to be characterized. Sally Frautschy’s group at the VA Medical Center in Sepulveda, California, presented data at the 33rd Annual Meeting of the Society for Neuroscience suggesting that Aβ oligomers can be acutely beneficial for spatial memory performance in rats, but that chronic exposure eventually leads to cognitive decline that parallels Alzheimer’s pathogenesis (SfN abstract 240.11).

Presenting author Marni Harris-White and colleagues infused rats with synthetic, SDS-stable low-molecular weight Aβ oligomers (monomers, dimers, trimers) using a mini osmotic pump that continually delivered the Aβ into a cerebral ventricle. Throughout the 3-month infusion, the researchers tested spatial memory at 6, 9, and 12 weeks in the Morris water maze. The rats spent 3 days training to swim to a visible platform, and then the researchers tested acquisition using a hidden platform. To test for memory retention, the rats attempted the maze again following a 24-hour delay. The researchers reported a steady improvement in memory acquisition and retention starting at 6 weeks and carrying through until 9 weeks post-infusion. However, by 12 weeks post-infusion, the animals began to show a mild decline in acquisition coupled with more pronounced deficits in spatial memory retention, characterized by the rats spending less time in the target quadrant. This apparent worsening of memory retention was correlated with a reduction in crucial post-synaptic proteins, such as PSD-95, in the cortex and hippocampus.

In summary, soluble, low-molecular-weight β-amyloid oligomers seem to support memory acquisition and retention early on. However, their contribution to excitatory neurotransmission is short-lived, as this initially beneficial synaptic stimulation eventually sentences the brain to cognitive deterioration. What remains to be elucidated is how an arguably toxic species of Aβ can be even remotely beneficial or have trophic activity in neurons, proving that this is no open-and-shut case. You can view abstracts mentioned in this story at the SfN/ScholarOne website.—Erene Mina is a graduate student at the University of California, Irvine.


  1. This curious effect of β-peptide oligomers, in first enhancing, then impairing cognition in rats, may have an explanation, as follows: Such oligomers are toxic in Alzheimer's disease because they cause synaptic lipid peroxidation, especially in brain regions like hippocampus, which have higher membrane cholesterol levels, that favor peptide-membrane interaction.

    Low-level lipid peroxidation favors cell differentiation, a physiological effect of fatty acid-derived aldehyde release. Thus, β-peptide oligomers could conceivably enhance maturation in the brain of the rat, especially the younger rat (whose brain is still developing postnatally). Eventually, of course, increased peroxidation (from 4-hydroxy-nonenal) would become toxic by inactivating ion-motive membrane pumps, glucose transporters, and glial glutamate transporters; synaptic degeneration and neuronal apoptosis could then follow.

    Similar impaired learning, in weaned rats raised on 20 percent safflower oil was noted in 1976, by Harman et al. Vitamin E supplemention prevented such learning impairment in a second batch of rat pups, suggesting that Harman's oil was typical steam-refined commercial oil, which is about 30 percent deficient in vitamin E.

    More recently, Greg Cole at UCLA has noted aggravation of AD pathology in FAD transgenic mice given safflower oil, which, if steam-refined, would confirm the hypothesis that common refined food oils are the direct cause of sporadic Alzheimer's disease.

    Ironically, the oil-induced primary lipid peroxidation presumably lays down the basis for the much later secondary peroxidation arising from β-peptide oligomers, since β-peptide accumulation would be favored by hydroxy-nonenal-induced inhibition of synaptic α-secretase.

    The good news is that, whereas fat-induced aqueous oxidation promotes cellular proliferation, the differentiating and apoptotic effects of refined oil-induced lipid peroxidation strongly oppose cancer development, so the take-home message is: Eat your oils, but check that the vitamin E level is = to or > 0.6 mg/ gm of polyunsaturated fatty acid. QED!


    . Free radical theory of aging: effect of dietary fat on central nervous system function. J Am Geriatr Soc. 1976 Jul;24(7):301-7. PubMed.

  2. Amyloid-β Oligomers: Were They Ever Bad?

    I read with interest the ARF news story on the 33rd Society for Neuroscience Annual Meeting 2003
    presentation by Harris/White et al. (abstract
    ). This is because, in our recent study, we showed that amyloid-
    β restored long-term potentiation (LTP, a cellular model for synaptic
    plasticity that underlies learning and memory) in rat hippocampal slices
    (Koudinov et al., 2002; also see ARF
    hypothesis page

    In fact there are several studies reporting on the essential role for
    Aβ in synaptic function and plasticity (see
    and the discussion
    of the cited article).

    With regard to the past reports on the neurotoxicity of Aβ oligomers
    (also called ADDLs) (Walsh
    et al. 2002
    ; Lambert
    et al. 1998
    ; Kayed
    et al. 2003
    ; also see ARF interview
    with Vincent Marchesi
    ), it is important to realize that these studies
    missed critical experimental consideration, the association of Aβ with lipoproteins. This association potently inhibits the neurotoxicity
    of Aβ (Koldamova
    et al. 2001
    ; Farhangrazi
    et al. 1997
    ; Cedazo-Minguez
    et al. 2001
    ) as well as serves to maintain Aβ solubility in
    the body fluids (Koudinov et al. 1998; 1999).

    While there is no question about a certain role for amyloid-β in
    Alzheimer's disease, the protein association with brain tissue lipoproteins
    creates an intriguing possibility of mistaken identity of lipid-bound soluble
    monomeric ApoAβ as plaque or oligomeric Aβ in a contemporary AD research
    (SFN 2003 abstract

  3. I thank Robert Peers for his provocative idea on the possible role of lipid peroxidation on improved cognition. This is also in agreement with our observed increase in F2 isoprostanes as an early endpoint that coincides with improved memory. However, I would like to add a more general comment on the implications of our observations. AD is not a disease with sudden onset, but takes decades to develop. Even MCI patients have extensive pathology without severe dementia. Therefore, I was frustrated, but not that surprised with the initial beneficial effects of amyloid oligomers on memory, which may involve two known effects of Aβ. The first relates to the largely overlooked normal physiological function of Aβ, which has been shown by several groups to be neurotrophic including Cotman's group. Factors potentiating memory by increasing excitability may make cells prone to excitotoxicity through calcium destabilization or energy depletion, leading to the second process: cortical toxicity. So why doesn't initial toxicity worsen memory? Well, because toxicity is initially cortical. In the in-vivo infusion model, sustained elevations in oligomers cause a preferential vulnerability of the entorhinal cortex to amyloid toxicity. Depending on dose and size of oligomer, we see either neuron loss or postsynaptic loss in AD vulnerable regions. By four weeks post-infusion, concomitant to improved learning in water maze, there is either neuron loss in layer II of the entorhinal cortex or loss of postsynaptic proteins. Entorhinal cortex lesion does not affect water maze learning (Bannerman et al., 2001), and causes sprouting of the perforant path, providing neurotrophic factor support for hippocampus, so it is no surprise that this toxicity doesn't yet reverse the initial beneficial effect of amyloid in water maze acquisition until 12 weeks post-infusion. More toxic forms of amyloid can accelerate this process. Amyloid is much like stress, which helps with acute adaptation to environment, but kills you when prolonged (Hans Selye, 1956). The adaptive and compensatory responses to chronic elevations in amyloid could mitigate early-stage illness (MCI or mild AD). I suspect that these early changes set the stage for moderate or late-stage illness, at which time accumulation of the high MW oligomer may be required to explain disease progression. The most frustrating part of these observations is that these data seem in contradiction to the observations of two premier Alzheimer's labs, which showed the acute detrimental effect of injected Aβ oligomer on LTP (Selkoe) and on behavior (Cleary and Ashe). To address this conflict, more research is necessary to determine whether the acute injection effects on LTP and memory are transient or sustained. It has been suggested that the improvement in memory with our chronic infusion model is due to the presence of monomer. This is partly true because all data so far presented were from an Aβ preparation with considerable monomer. However, I have new data showing that infusion of high MW oligomers in the absence of monomer caused the same early improvement in memory, although improvements were more transient and deficits happened sooner. Stage-dependent regional expression of defined oligomer classes in the AD brain (as Ashe has done in Tg2576 mouse) should also help sort some of this out.


    . The role of the entorhinal cortex in two forms of spatial learning and memory. Exp Brain Res. 2001 Dec;141(3):281-303. PubMed.

  4. In response to the commentary by Robert Peers, I'd like to reply that part of his interpretation is plausible and part of it seems less likely to me. When Sally Frautschy and Marni Harris-White infused that particular preparation of Aβ oligomers into older adult animals, they saw an initial response with enhanced cognitive function that was eventually followed by cognitive deficits and decline. While the explanation for initial trophic effects is unclear, developmental maturation endpoints in nine-month-old adults would not be the first explanation to consider. That preparation was not IDE-treated or otherwise shown to contain pure oligomers, so it may have had a trophic effect from low-dose monomer that activates a7 nicotinic acetylcholine receptors and the ERK MAPK cascade as described by David Sweatt's group (Dineley et al., 2001; see also ARF New Orleans story). Trophic effects of low-dose monomer have been suggested in many different systems beginning with Carl Cotman's initial report on cultured neurons. It is also plausible that the injury from pump implantation contributed to an initial glial response that included an upregulation of neuroprotective trophic factors. Finally, it is possible that toxicity involves co-factors such as metals that only become freely available later in the response. Whatever the mechanism of trophic activity in the peptide preparation, it is very dose- and preparation-dependent.

    Frautschy's group also has evidence of extensive toxicity with higher doses of similar preparations and with other oligomer preparations. For example, in collaboration with Caleb Finch's group, they have previously presented data at SFN showing that infused oligomer preparations made with ApoJ can selectively destroy AD-vulnerable CA1 and entorhinal cortex layer II neurons in rats, and similar neurotoxicity at one month has been recently found in collaboration with Jeffrey Craft in Linda van Eldik's group using cold oligomer preparations infused into mice.

    In relationship to effects of safflower oil, I would agree with Robert Peers that there is an important pro-oxidant effect that synergizes with amyloid toxicity. However, to date, our unpublished data suggest that DHA depletion is at least as critical an endpoint as oxidative damage per se in inducing dendritic pathology. The animal model results are, in general, consistent with epidemiology, which to my knowledge consistently identifies low fish and, more specifically, low DHA intake as a risk factor, but has yet to clearly implicate high n-6 PUFA, safflower, or any other oil.

    Finally, there have been a number of recent presentations and papers consistent with oxidative damage as a significant factor in AD pathogenesis. If this is true, one can assume that like atherosclerosis, AD will not only be influenced by dietary PUFA, but a host of other pro- and antioxidant and inflammatory pathways regulated by the diet, most obviously including folate, polyphenolic antioxidants, vitamin E, vitamin C, and lipoate. In the next few years, animal model studies on lipids and all of the above factors should make it possible to design an appropriate and synergistic dietary prevention strategy very likely capable of substantially reducing AD risk. Unfortunately, clinical trials will probably first test the efficacy of each factor one by one on Alzheimer's patients, where the impact is likely to be small at best, just as it was with vitamin E. We can only hope that single interventions will be potent enough by themselves to make a real difference in patients.


    . Beta-amyloid activates the mitogen-activated protein kinase cascade via hippocampal alpha7 nicotinic acetylcholine receptors: In vitro and in vivo mechanisms related to Alzheimer's disease. J Neurosci. 2001 Jun 15;21(12):4125-33. PubMed.

  5. In relation to the possible role of Abeta peptides in memory and learning processes, I would like to mention that the results that we obtained (presented in NSF meeting 2003 Echeverria V, Ducatenzeiler A and Cuello C) analyzing the influence of the Ab-induced ERK2 hyperactivation on the CRE-regulated gene expression,,using PC12 cells as a model, clearly suggested that at least in PC12 cells under stimulated conditions with increased levels of calcium and cAMP, soluble oligomeric Abeta at low concentration stimulates ERK-dependent activation of the CRE-regulated gene expression a signaling pathway involved in the memory and learning processes. At the contrary, higher levels of fibrillar Abeta abolished it.
    It is likely then that A beta could have a normal role in neural plasticity involved in memory formation and a pathological role when it abnormaly accumulates intra and excellularly in the brain

Make a Comment

To make a comment you must login or register.


News Citations

  1. New Orleans: Aβ Oligomers and Memory: …Now They Are Bad

External Citations

  1. SfN/ScholarOne

Further Reading

No Available Further Reading