N E W S A N D V I E W S

760 VOLUME 12 | NUMBER 7 | JULY 2006 NATURE MEDICINE

Aβ star: a light onto synaptic dysfunction?Seong-Hun Kim, Ya-Ping Tang & Sangram S Sisodia

Learning and memory is impaired in a mouse model of Alzheimer disease. An Aβ 12-mer called Aβ* may be responsible for these cognitive deficits.

Seong-Hun Kim and Sangram S. Sisodia are in

the Department of Neurobiology, Pharmacology &

Physiology, and Ya-Ping Tang is in the Department

of Psychiatry, University of Chicago, Chicago, Illinois

60637, USA.

E-mail: ssisodia@bsd.uchicago.edu

Amyloid-β (Aβ) peptides are widely believed to have a causative role in the pathogenesis of Alzheimer disease. Aβ assembles into a variety of higher-order structures that include dimers, oligomers, protofibrils and fibrils, and evidence has emerged to support a role for soluble oligomeric and protofibril-lar forms in dysregulation of synaptic func-tion1–3. To these latter efforts, Karen Ashe and colleagues now provide evidence that a ∼56 kDa Aβ oligomer, termed Aβ*56 (“Aβ star”), isolated from the brain of a well-characterized mouse model of Alzheimer disease, can induce cognitive deficits when infused into the brains of naive rats4.

The Tg2576 mouse model of Alzheimer disease expresses human amyloid precursor protein (APP) containing a mutation associ-ated with a familial form of Alzheimer dis-ease. Tg2576 mice develop memory deficits at 6 months of age, well before the presence of detectable Aβ deposits4. In view of earlier evidence that oligomeric Aβ injected into rat brain induces cognitive deficits2, the authors sought to identify Aβ species in the brains of Tg2576 mice that showed the strongest cor-relation with memory impairments. A vari-ety of Aβ-related derivatives were identified, and among these, a ∼56 kDa species termed Aβ*56 first appeared at 6 months of age. The levels of this Aβ oligomer in individual mice inversely correlated with their performance in a memory-retention test.

Of course, correlations are a weak substi-tute for causation, and a definitive test of the proposal that Aβ*56 is solely responsible for the cognitive deficits in young Tg2576 mice still remained. To tackle this formidable challenge, the authors attempted a ‘transmis-sion’ experiment that involved purification of Aβ*56 from detergent-solubilized Tg2576 brain extracts and subsequent injection of these preparations into the brains of young

rats. Although infusion of Aβ*56 2 hours before testing the rats in a spatial learn-ing task had no effect on the acquisition of spatial information, rats that received a sec-ond dose of Aβ*56 24 hours later exhibited impairments in memory retention. Notably, the effects of the oligomer on memory reten-tion was a transient phenomenon, as Aβ*56-treated rats performed as well as control rats when tested 10 days later. Nevertheless, Aβ*56, prepared from the brains of cogni-tively impaired mice, can transfer memory deficits to a heterologous host.

As is the case for many exciting advances in science, the present studies raise several intriguing issues and challenges that will surely be the subject of future inquiry. Biochemists will certainly want to know the true identity of Aβ*56. This species is recognized by the Aβ-specific 6E10 and 4G8 antibodies, and the oligomer-specific A11 antibody5. Aβ*56 does not disaggregate in response to urea, a strong denaturing agent. But increasing con-centrations of the hydrogen-bond disrupting agent, hexafluoroisopropanol (HFIP), cause the oligomer to dissociate into highly stable intermediate ‘trimeric’ species and, to a lesser extent, monomeric forms.

Analysis of Aβ*56 shows that Aβ is the major ‘core’ component. Collectively, these data would argue that Aβ*56 is a highly sta-ble dodecameric assembly of Aβ. But some questions remain. First, is Aβ*56 composed primarily of Aβ42, or a mixture of shorter Aβ peptides ‘seeded’ by Aβ42? Second, how ‘pure’ is the Aβ*56 preparation used in the infusions? In this regard, low levels of contaminants would be difficult to visual-ize by current methods, and estimations of the native molecular weight of Aβ*56 are imperfect.

Although it is clear that intraventricular infusion of Aβ*56 in rat brain causes memory deficits, there is little information about fate of the oligomers once delivered, and the specific-ity of the Aβ*56-mediated effects on neuronal dysfunction. How stable is Aβ*56? Does this oligomer mature into insoluble fibrillar aggre-gates or resolve into lower-molecular-weight oligomers, species that have been reported to induce synaptic dysfunction1–3? In this regard, does coadministration of the aforementioned A11 antibody, which selectively recognizes Aβ oligomers and attenuates Aβ oligomer-induced neurotoxicity5, also neutralize the Aβ*56-induced memory deficits? Finally, the

Presynaptic axon Postsynaptic dendrite

Glutamate

Aβmonomer Trimer

Hexamer

Aβ*56

No Aβ +Aβ

Nanomer

Glutamate receptor

Internalization?

Ca2+ influx

?

Figure 1 Aβ*56, most likely a tetramer of Aβ trimers, is the prime candidate suspected of causing early memory deficits in a mouse model of Alzheimer disease4. The mechanism(s) by which Aβ*56 induces memory loss are not clear, but it could disrupt glutamatergic neurotransmission and thereby impair synaptic plasticity, possibly by enhancing glutamate receptor internalization, as reported by Snyder et al 6.

Kim

Cae

sar

©20

06 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

urem

edic

ine

N E W S A N D V I E W S

NATURE MEDICINE VOLUME 12 | NUMBER 7 | JULY 2006 761

authors infused a large amount of Aβ*56 into the brain and used vehicle as the control. In view of the inordinately high levels of infused Aβ*56, it would seem reasonable to examine the effects of other ‘amyloidogenic’ polypep-tides unrelated to Aβ, or for that matter, solu-ble Aβ oligomers, such as the stable trimers, on the behavioral parameters that were tested. At a fundamental level, it will be crucial to iden-

tify the cell-surface target(s) and downstream signaling pathways that impact glutamater-gic neurotransmission6 and Aβ*56-induced memory loss (Fig. 1).

Finally, it will be crucial to establish the presence of Aβ*56 in humans and validate the proposal put forth that Aβ*56 could serve as a valuable biomarker for preclini-cal diagnosis, and ultimately, provide a new

therapeutic target to “...abort the disease before permanent structural changes have developed”4.

1. Tanzi, R.E. Nat. Neurosci. 8, 977–979 (2005).2. Cleary, J.P. et al. Nat. Neurosci. 8, 79–84 (2005).3. Walsh, D.M. et al. Nature 416, 535–539 (2002).4. Lesne, S. et al. Nature 440, 352–357 (2006).5. Kayed, R. et al. Science 300, 486–489 (2003).6. Snyder, E.M. et al. Nat. Neurosci. 8, 1051–1058

(2005).

The author’s perspectiveFourteen years ago, I decided to study how amyloid-β (Aβ) alters brain function as others were investigating how it changes brain structure. I have often had doubts about this choice because it has been an arduous road. When I began tackling this problem, there were no models to study the effects of endogenous Aβ on memory. It took such a long time to create a useful transgenic mouse model that I did not publish any papers the first three years I was at the University of Minnesota, which nearly cost me tenure because scientists at UCSF and Columbia wrote opposition letters, stating quite correctly that I would not receive tenure at their institutions. I was only promoted because the dean overrode the Promotions and Tenure Committee’s negative recommendation. After we created the Tg2576 mouse model in 1996, it took my students and collaborators six years to determine the relationship between memory function, brain pathology and Aβ levels over the two-year lifespan of the mouse. During this time, I ran out of money to support the large aging mouse colony I needed for these experiments, and was struggling to create a lab infrastructure that could handle the volume of mouse breeding and behavioral testing. Fortunately, I had a wonderful group of lab technicians who devised a team method for generating the mice as well as the behavioral data that forms the foundation of our work. In addition, I was given two endowed chairs to support the work financially. The Aβ* hypothesis was forged in 2001, but to prove it, I needed a method for testing acute cognitive effects of minute quantities of Aβ oligomers. Luckily, two collaborators turned their labs over to solving this problem. We also lacked the expertise to find Aβ* and were stuck until Sylvain Lesné joined us in 2002. Although Sylvain’s paper underwent the most rigorous set of reviews I have ever encountered, the paper that was published is much better than the first version we submitted, a clear testament to superb reviewers.

Karen H. Ashe, University of Minnesota Medical School

Highly cited clinical and epidemiological papers on Alzheimer disease published in 2003a

Reference Times cited

Shumaker, S.A. et al. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women—The Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA 289, 2651–2662

497

Reisberg, B. et al. Memantine in moderate-to-severe Alzheimer’s disease. N. Engl. J. Med. 348, 1333–1341 336

Nicoll, J.A. et al. Neuropathology of human Alzheimer disease after immunization with amyloid-β peptide: a case report. Nat. Med. 9, 448–452 240

Hebert, L.E. et al. Alzheimer disease in the US population: prevalence estimates using the 2000 census. Arch. Neurol. 60, 1119–1122 185

Hock, C. et al. Antibodies against β-amyloid slow cognitive decline in Alzheimer’s disease. Neuron 38, 547–554 182

Aisen, P.S. et al. Effects of rofecoxib or naproxen vs. placebo on Alzheimer disease progression: a randomized controlled trial. JAMA 289, 2819–2826

167

Orgogozo, J.-M. et al. Subacute meningoencephalitis in a subset of patients with AD after Aβ42 immunization. Neurology 61, 46–54 125

aNumber of citations as of 13 June 2006. Table includes all clinical trials (including phase 1 trials), epidemiological studies, case reports and biomarker studies that have the terms ‘Alzheimer’ or ‘Alzheimer’s’ in their title, abstract or keywords, and that have been cited at least 125 times. Data source: Scopus

Cou

rtes

y of

Nat

ure

Bio

tech

nolo

gy

©20

06 N

atur

e P

ublis

hing

Gro

up

http

://w

ww

.nat

ure.

com

/nat

urem

edic

ine