Low-n oligomers as therapeutic targets of Alzheimer’s disease

Authors


Address correspondence and reprint requests to Dr K. Ono, Department of Neurology & Neurobiology of Aging, Kanazawa University Graduate School of Medical Science, Kanazawa 920-8640, Japan. E-mail: onoken@med.kanazawa-u.ac.jp

Abstract

J. Neurochem. (2011) 117, 19–28.

Abstract

The pathogenesis of Alzheimer’s disease involves the progressive accumulation of amyloid β-protein (Aβ). Recent studies using synthetic Aβ peptides, a cell culture model, Aβ precursor protein transgenic mice models suggest that pre-fibrillar forms of Aβ are more deleterious than extracellular fibril forms. Recent findings obtained using synthetic Aβ peptides and human samples indicated that low-n oligomers (from dimers to octamers) may be proximate toxins for neuron and synapse. Here, we review the recent studies on the soluble oligomers, especially low-n oligomers in Alzheimer’s disease.

Abbreviations used

amyloid β-peptide

AD

Alzheimer’s disease

ADDL

Aβ-derived diffusible ligand

APP

Aβ precursor protein

CQ

clioquinol

FAD

familiar AD

LTP

long term potentiation

PBT

Prana Biotechnology

PICUP

photo-induced cross-linking of unmodified proteins

SDS

sodium dodecyl sulfate

In Alzheimer’s disease (AD), prominent protein misfolding leads to aggregation and, eventually, formation of a fibrillar structure rich in β-sheets (Conway et al. 2000; Klein et al. 2001; Kirkitadze et al. 2002; Meier et al. 2006). AD is the most common form of late-life dementia and is the focus of this review. Current estimates of AD incidence are > 24 million worldwide, a number that is expected to double every 20 years, reaching 81 million in 2040 (Blennow et al. 2006). AD is a slowly progressive disorder with insidious onset and progressive impairment of episodic memory and executive function coupled with aphasia, apraxia, and agnosia (Blennow et al. 2006).

From genetic evidence, biochemical data and animal models, amyloid β-protein (Aβ) has been suggested to be responsible for the pathogenesis of AD (Hardy and Selkoe 2002; Walsh and Selkoe 2007). Aβ is derived from the Aβ precursor protein (APP) by the action of two aspartyl proteases referred to as β- and γ-secretases (Haass et al. 1992; Seubert et al. 1992; Shoji et al. 1992; Hardy and Selkoe 2002) (Fig. 1). Although two forms of Aβ, comprising 40 and 42 amino acid residues are produced, the relative amount of Aβ42 is particularly critical for AD progression because this longer form of Aβ is more prone to aggregate than shorter Aβ40 peptide (Burdick et al. 1992; Jarrett et al. 1993; Walsh and Selkoe 2007). Aβ molecules tend to aggregate to form oligomers, protofibrils, and β-amyloid fibrils, which have been suggested to cause neuronal dysfunction in the brains of AD patients (Fig. 1) (Ono et al. 2006; Yamin et al. 2008). These Aβ aggregates may cause neuronal injury directly by acting on synapses, or indirectly by activating microglia and astrocytes (Hardy and Selkoe 2002; Haass and Selkoe 2007; Walsh and Selkoe 2007) and therefore pharmacological interventions have been developed to target the sequential events originating from Aβ synthesis (Ono et al. 2006; Yamin et al. 2008).

Figure 1.

 β-Amyloid precursor protein (APP) processing and aggregation. Aβ is produced by the sequential cleavage of APP. β-secretase cleavage produces the Aβ N-terminus, after which γ-secretase releases the Aβ C-terminus from APP. The unfolded Aβ monomer may fold or aggregate to produce toxic aggregates. In vitro and in vivo studies suggest that Aβ proceeds along a linear pathway comprising many populated monomer conformational states, a population of partially folded states, a more restricted distribution of oligomers, protofibrils, and fibrils.

Substantial evidence supports the hypothesis that Aβ oligomers play a seminal role in AD (Kirkitadze et al. 2002; Walsh et al. 2002; Wang et al. 2002; Klein et al. 2004; McLaurin et al. 2006; Cheng et al. 2007; Haass and Selkoe 2007; Walsh and Selkoe 2007; Shankar et al. 2008; Roychaudhuri et al. 2009). Aβ oligomers inhibit hippocampal long term potentiation (LTP) in mice and rats injected intracerebrally with Aβ oligomers (Walsh et al. 2002; Wang et al. 2002; Klyubin et al. 2004; Townsend et al. 2006; Shankar et al. 2008). Recently, formal structure–cytotoxicity studies of pure Aβ40 oligomer populations produced the first determinations of specific activity of dimers, trimers, and tetramers (Ono et al. 2009). Dimers, trimers, and tetramers all were significantly more toxic than monomers.

In this article, recent developments in the research on low-n oligomers (from dimers to octamers) of Aβ, including our findings are reviewed.

Aβ aggregation is required for toxicity

Amyloid β-protein is a normal byproduct of cellular/neuronal metabolism and can be detected as a circulating peptide in the plasma and CSF of healthy humans (Haass et al. 1992; Seubert et al. 1992; Vigo-Pelfrey et al. 1993; Ida et al. 1996; Walsh et al. 2000). Thus, the presence of Aβ itself does not cause neurodegeneration; rather neuronal cell death appears to ensue as a result of the ordered self-aggregation of Aβ molecules (Pike et al. 1991; Busciglio et al. 1992; Geula et al. 1998). Within the amyloid plaques that characterize AD, synthetic Aβ can form mature fibrils in vitro similar to those present in human brain (Castano et al. 1986; Kirschner et al. 1987).

Early studies clearly demonstrated that although aggregation of Aβ was essential for toxicity, the characterization of Aβ aggregates that formed in vitro was limited, and it was assumed that since mature fibrils were detectable, these assemblies might mediate the observed toxicity. However, this could not explain the relatively weak correlation between the severity of dementia and the density of fibrillar amyloid plaques in AD brain (Katzman 1986; Terry et al. 1991; Dickson et al. 1995). Evidence for the involvement of soluble, non-fibrillar Aβ in AD has accumulated through three distinct experimental approaches that utilize (i) in vitro system using synthetic Aβ peptides; (ii) in vivo system using APP transgenic mice; and (iii) human CSF and postmortem brain.

In vitro evidence that low-n oligomers of Aβ are potent neurotoxins

Neurotoxicity studies have shown that Aβ aggregates are potent neurotoxins (Oda et al. 1995; Lambert et al. 1998; Hartley et al. 1999; Hirakura et al. 2000). Oda et al. (1995) found slowly sedimenting Aβ complexes formed in the presence of clusterin, which included aggregates > 200 kDa that resist dissociation by low concentrations of sodium dodecyl sulfate (SDS) by incubating Aβ with human serum clusterin (Oda et al. 1995). These Aβ aggregates enhanced mitochondrial toxicity in neuron like PC12 cells (Oda et al. 1995). Lambert et al. (1998) characterized them and found that these ∼ 53 kDa small diffusible Aβ oligomers, that is Aβ-derived diffusible ligands (ADDLs), killed mature neurons in organotypic central nervous system cultures at nanomolar concentrations. At cell surfaces, ADDLs bound to trypsin-sensitive sites and surface-derived tryptic peptides blocked binding and afforded neuroprotection (Lambert et al. 1998). Later, it was reported that the toxicity of ADDL, can be greater than that of the mature fibrils (Dahlgren et al. 2002). Hartley et al. (1999) reported that both low molecular weight Aβ and protofibrils reproducibly induced toxicity in mixed brain cultures in a time- and concentration-dependent manner.

Recent studies of Aβ oligomers put bigger focus on smaller soluble oligomers, as well as seeking to correlate oligomer size and biological activity. Townsend et al. (2006) found that Aβ trimers fully inhibit LTP, whereas dimers and tetramers have an intermediate potency. Dimers and trimers from this medium also have been found to cause progressive loss of synapses in organotypic rat hippocampal slices (Shankar et al. 2007). However, little is known about the structure-neurotoxicity relationship of these low-n oligomers because of their biochemical instability. Very recently, we succeed in extracting low-n Aβ oligomers, stabilized structurally by using photo-induced cross-linking of unmodified proteins (PICUP) (Fig. 2) and then determined their conformations and neurotoxic activities. Oligomer order-specific effects were consistently observed. We found that the size of each oligomer increased with oligomer order using both electron microscopy and atomic force microscopy (Ono et al. 2009). We also found that Aβ monomers are largely unstructured, but oligomers exhibit order-dependent increases in β-sheet content by circular dichroism spectroscopy. These increases correlated with the abilities of the oligomers to nucleate fibril assembly as well as cytotoxicity (Ono et al. 2009). In the cytotoxicity assay of differentiated PC12 cells, dimers, trimers, and tetramers were all significantly more toxic than monomers. Notably, we found a non-linear dependence of cytotoxicity on oligomer order. Oligomers of higher order were disproportionately more toxic (tetramers >> trimers >> dimers >> monomers). We also observed these relationships in assays of fibril nucleation activity (Ono et al. 2009). However, the biochemical difference between secreted oligomers, stabilized oligomers by PICUP, and oligomers in the human brain is still unclear.

Figure 2.

 Distribution of purified Aβ oligomers. Aβ samples were subjected to PICUP and SDS–PAGE. Individual gel bands were stained with Coomassie Blue, excised, and then extracted under alkaline conditions. The extracts then were reconstituted in F12K medium and analyzed by SDS–PAGE and silver staining. Lane 1, cross-linked Aβ subjected to the entire protocol; Lane 2, monomer band; Lane 3, dimer band; Lane 4, trimer band; Lane 5, tetramer band. This research was originally published in Proc. Natl Acad. USA (Ono et al. 2009).

Very recently, in both the Aβ40 and Aβ42 systems, we found that the English (H6R) and Tottori (D7N) substitutions accelerated the kinetics of secondary structure change from statistical coil→α/β→β and produced oligomer size distributions skewed to higher order relative to their wild-type homologues by PICUP (Fig. 3) (Ono et al. 2010). This skewing was reflected in increases in average oligomer size, as measured using electron microscopy and atomic force microscopy. Stabilization of peptide oligomers using PICUP allowed detailed study of their properties. Each substitution produced an oligomer that displayed substantial β-strand (H6R) or α/β (D7N) structure, in contrast to the predominately statistical coil structure of wild-type Aβ oligomers (Ono et al. 2010). We found that mutant oligomers functioned as fibril seeds, and with efficiencies significantly higher than those of their wild-type homologues by thioflavin T assay. Importantly, the mutant forms of both native and chemically-stabilized oligomers were significantly more toxic in assays of cell physiology and death (Ono et al. 2010). The results show that the English and Tottori mutations alter Aβ aggregation at its earliest stages, monomer folding and oligomerization, and produce oligomers that are more toxic to cultured neuronal cells than are wild-type oligomers.

Figure 3.

 Aβ oligomerization of wild-type, H6R and D7N. PICUP, followed by SDS–PAGE and silver staining, was used to study oligomerization of 25 μM Aβ40 (a, b, c) or Aβ42 (d, e, f) of wild-type (a, d), H6R (b, e) or D7N (c, f). Lanes 1, un-cross-linked Aβ; Lanes 2, Aβ cross-linked at the beginning of incubation at 37°C; Lanes 3, Aβ cross-linked after 2 h; Lanes 4, Aβ cross-linked after 4 h; Lanes 5, Aβ cross-linked after 6 h. Oligomer order is indicated by white numerals over the respective gel bands. This research was originally published in J. Biol. Chem. (Ono et al. 2010).

Although we have examined the toxicity of low-n oligomers of Aβ by using PICUP, a future study comparing their toxicity with high-n oligomers of Aβ in the same experimental model is essential.

In vivo evidence that low-n oligomers play key roles in AD pathogenesis

After the reports that amyloid plaque density in the human brain does not correlate with severity of dementia (Katzman 1986; Terry et al. 1991; Dickson et al. 1995), studies showed that memory impairment and changes in neuron form and function preceded the amyloid deposition in APP transgenic mice (Chapman et al. 1999; Hsia et al. 1999; Moechars et al. 1999; Mucke et al. 2000; Dineley et al. 2002; Westerman et al. 2002; Wu et al. 2004). Hsia et al. (1999) reported that over-expression of familiar AD (FAD) (717V→F)-mutant human APP in neurons of transgenic mice lowers the density of pre-synaptic terminals and neurons before these mice develop amyloid plaques (Hsia et al. 1999). Electrophysiological recordings from the hippocampus revealed prominent deficits in synaptic transmission, which also preceded amyloid deposition by several months (Hsia et al. 1999). Similarly, Mucke et al. (2000) reported that the Aβ is synaptotoxic even in the absence of plaques and that high levels of Aβ are insufficient to induce plaque formation using wild-type and FAD-mutant APP mice (Mucke et al. 2000). Westerman et al. (2002) and Dineley et al. (2002) reported that memory loss and insoluble Aβ aggregates were not connected using APP transgenic mice (Tg2576) and transgenic mice co-expressing mutant presenilin-1 and APP, respectively, suggesting that earlier Aβ aggregates disrupting cognition (Dineley et al. 2002; Westerman et al. 2002). Wu et al. (2004) evaluated the cytological basis for these earlier changes, and showed that substantial dendritic pathology is evident in 90-day-old transgenic mice that over-express a mutation associated with FAD (717V→F) for a spatially defined subset of granule cells well before amyloid pathology occurs (Wu et al. 2004).

Using the APP transgenic mouse model Tg2576, SDS–polyacrylamide gel electrophoresis (PAGE) and gel infiltration, Lesne et al. (2006) found nonamers and dodecamers that appeared at 6 months, the age when these mice first show changes in performance on the Morris-water maze (Lesne et al. 2006). The nonamer and dodecamer levels correlated with the impairment of spatial memory. Injection of the dodecamer purified by immunoaffinity chromatography into the ventricle of normal pre-trained wild-type rats caused a dramatic fall-off in spatial memory performance. This demonstrated that a 56-kDa soluble, brain-derived form of Aβ (Aβ*56) can directly mediate brain dysfunction. However, the Tg2576 mice show impaired performance in a hippocampal-dependent contextual fear conditioning assay, decreased spine density in the dentate gyrus, and impairment of LTP at ages long before the first apparent detection of Aβ dodecamer (Dineley et al. 2002; Jacobsen et al. 2006; Lesne et al. 2006). Thus, the appearance of dodecamer in Tg2576 mice does not correlate with changes in other forms of memory, nor does it correlate with changes in synaptic form and function.

Chinese hamster ovarian cells that express mutant (V717F) human APP (7PA2 cells) can produce and secrete low-n oligomers including mainly dimers and trimers and occasionally tetramers (Podlisny et al. 1995; Walsh et al. 2002). Walsh et al. (2002) reported that microinjection of small volumes of these low-n oligomers into the lateral ventricle of the brain of an anesthetized rat inhibited hippocampal LTP in vivo (Walsh et al. 2002). In addition, they showed the blockage of LTP was specially mediated by low-n oligomers, not by monomers or any larger aggregates (Walsh et al. 2005). They also reported that Aβ immunization antibody neutralizes these Aβ oligomers that disrupt synaptic plasticity in vivo (Klyubin et al. 2005). They furthermore developed a new method to simultaneously assess the ability of soluble Aβ dimers to impair plasticity at synapses and to affect resting and activity-dependent local blood flow in the rat hippocampus in vivo (Hu et al. 2008). Intracerebroventricular injection of soluble synthetic Aβ dimers that were cross-linked covalently using AβS26C rapidly inhibited LTP, but failed to affect vascular function measured using laser-Doppler flowmetry, indicating that the inhibition of synaptic plasticity can be caused by low-n oligomers of Aβ independently of deleterious effects on cerebrovascular dynamics (Hu et al. 2008). Very recently, they reported that intracerebroventricular injection of naturally secreted low-n oligomers of Aβ after early avoidance training for recall of light foot shock caused a significant impairment in memory consolidation with significantly fewer synapses in the dentate gyrus later in vivo, suggesting that low-n oligomers of Aβ target specific temporal facets of consolidation-associated synaptic remodeling whereby loss of functional synapses results in impaired consolidation (Freir et al. 2010).

Although low-n oligomers of Aβ are presumed to cause synaptic cognitive dysfunction of AD, their contribution to other pathological features of AD remains unclear. To determine the interaction of Aβ oligomers with tau pathology, Oddo et al. (2006) used the 3×Tg-AD mice, which develop a progressive accumulation of Aβ plaques and tangles and cognitive impairments. They presented evidence for the formation of SDS-resistant low-n oligomers of Aβ such as dimers, trimers, and pentamers appears to first occur intraneuronally. Then, they showed that a single intrahippocampal injection of a specific oligomer antibody is sufficient to clear Aβ pathology, and more importantly, tau pathology (Oddo et al. 2006). Recently, Tomiyama et al. (2010) generated APP transgenic mice expressing the E693Δ mutation, which causes synaptic dysfunction by enhanced formation of low-n order oligomers such as dimers, trimers, and tetramers without fibril formation. This mouse showed accumulation of intraneuronal Aβ oligomers as well as abnormal tau phosphorylation from 8 months, microglial activation from 12 months, astrocyte activation from 18 months, and neuronal loss at 24 months (Tomiyama et al. 2010). These findings suggest that low-n oligomers of Aβ cause not only synaptic alteration but also other features of AD pathology.

Evidence of the presence of low-n oligomers in human samples

In previous studies, the amyloid plaque number in human brain samples was not highly correlated with the severity of dementia (Katzman 1986; Terry et al. 1991; Dickson et al. 1995). Recently, the soluble Aβ levels were reported to be correlated with the extent of synaptic loss and severity of cognitive impairment (Lue et al. 1999; McLean et al. 1999; Wang et al. 1999), and the soluble Aβ did not pellet after ultracentrifugation indicating that they are not in a fibrillar state in nature (Funato et al. 1998; Morishima-Kawashima and Ihara 1998; Stenh et al. 2005).

Although the primary sequence of Aβ found in human brain has been studied extensively, little is known about the assembly forms of cerebral Aβ. Kuo et al. (1990) isolated non-fibrillar forms of Aβ from both AD and control human brain. They reported that both control and AD brain contained an abundance of low-n oligomers, but the possibility remained that these oligomers are Aβ species bound to other proteins.

McLean et al. (1999) revealed the presence of monomers, dimers and trimers of Aβ in the supernatant of specimen extracted from the frontal cortex and putamen of AD brain (McLean et al. 1999). Similarly, Funato et al. (1999) reported the accumulation of SDS-stable dimers in the hippocampus CA1 in many AD cases with or without neurofibrillary tangles. These SDS-stable low-n oligomers that have also been reported in human CSF (Vigo-Pelfrey et al. 1993) could be the highly stable non-covalently associated dimers of Aβ1–40 and trimers of either Aβ6–42 or Aβ1–35. These findings and the detection of the soluble low-n oligomers in amyloid plaques (Roher et al. 1996; Enya et al. 1999) suggest that low-n oligomers of Aβ develop into insoluble amyloid deposits and could lead to the neurodegeneration in AD pathogenesis.

Human CSF from both healthy older individuals and AD patients contained clearly detectable dimers of Aβ, and completely disrupted synaptic plasticity in a manner similar to animal cell-derived low-n oligomers of Aβin vivo (Klyubin et al. 2008). Moreover, systemic passive immunization against Aβ by monoclonal antibodies 4G8 and 6E10 fully prevented the inhibition of LTP by both human and animal cell-derived Aβ dimers. Aβ monomer isolated from human CSF did not affect LTP. These results strongly support a strategy of therapeutic targeting soluble low-n oligomers of Aβ in early AD (Klyubin et al. 2008).

Selkoe’s group showed soluble Aβ oligomers extracted directly from the cerebral cortex of subjects with AD, potently inhibited LTP, enhanced long-term depression and reduced dendritic spine density in normal rodent hippocampus (Shankar et al. 2008). These oligomers also disrupted the memory of learning behavior in normal rats. Insoluble amyloid plaque cores from AD cortex did not impair LTP unless they were first solubilized to release Aβ dimers, suggesting that plaque cores are largely inactive but sequester Aβ dimers that are synaptotoxic. The Aβ dimers extracted from human AD brain facilitated electrically evoked long-term depression by disrupting neuronal glutamate uptake in the CA1 region of hippocampus (Li et al. 2009).

Using 43 brains from the Newcastle cohort of the population-representative Medical Research Council Cognitive Function and Ageing Study, Walsh’s group examined the relationship between biochemically distinct forms of Aβ and the presence of AD-type dementia (Mc Donald et al. 2010). While Aβ monomers were specifically detected in soluble fraction of AD brain, dimers were also detected in soluble fraction as well as formic acid fraction of AD brain, especially, frontal and temporal cortices (Mc Donald et al. 2010). These dimers detected in both soluble fraction and formic acid fraction were very strongly associated with AD samples but not with non-AD samples. These data suggest that the soluble low-n oligomer is a major factor correlated with AD-pathology and is likely to intimately involved in the pathogenesis of cognitive failure.

Selkoe’s group measured Aβ oligomer species in plasma and postmortem samples from AD samples and control subjects using a new ELISA specific for Aβ oligomers, especially dimers (Xia et al. 2009). Their plasma assays showed that relative Aβ oligomer levels were closely associated with relative Aβ monomer levels (Xia et al. 2009). Moreover, both oligomer and monomer levels of Aβ were higher in brain tissue of AD cases (Xia et al. 2009).

In a recent study on the effect of CSF from 33 patients with AD and 33 age-matched, non-demented controls on oligomerization of Aβ40 and Aβ42 using the technique of photo-induced cross-linking of unmodified proteins, CSF obtained from both groups inhibited low-n order oligomerization of both Aβ40 and Aβ42 (Ikeda et al. 2010). This inhibitory effect was significantly weaker in AD patients than in non-demented controls (Ikeda et al. 2010). Our results indicate that AD patients have a CSF environment favorable for the formation of low-n oligomers of Aβ.

Inhibition of Aβ oligomerization as a preventive and therapeutic approach

Natural removal of the Aβ monomer by the action of degrative enzymes in the brain, or stabilizing of the Aβ monomer by binding with small molecules should prevent oligomerization. Although many inhibitors of in vitro Aβ aggregation have been identified (for a review, see Ono et al. 2006; Yamin et al. 2008), few molecules capable of disrupting pre-formed oligomers have reached the clinical trial stage.

Bush (2002) showed that chelators inhibit the production of hydrogen peroxide by Aβin vitro. These compounds can also reverse the aggregation of Aβin vitro and in vivo (based on analysis of postmortem human brain specimens) (Bush 2002). A particularly important chelator is clioquinol (CQ), an analogue of 8-hydroxy-quinoline. CQ has moderate affinity for Cu2+ and Zn2+ and has been found to inhibit metal-induced Aβ aggregation and reactive oxygen species generation in vitro (Cherny et al. 2001). Oral administration of CQ to Tg2576 mice reduced the brain Aβ burden by ∼ 50% after 9 weeks and was accompanied by a modest increase in cerebral Cu2+ and Zn2+ levels (Cherny et al. 2001). In a pilot phase IIa trial of CQ Prana Biotechnology (PBT1), AD patients showed evidence of delayed cognitive deterioration and significantly lowered plasma Aβ levels (Ritchie et al. 2003). However, phase II/III trials were halted because of impurities in the formulation (Tschape and Hartmann 2006). Adlard et al. (2008) characterized a second-generation 8-hydroxy-quinoline analogue, PBT2, which also targets metal-induced Aβ aggregation and has greater blood–brain barrier permeability. Compared to CQ, PBT2 significantly decreased soluble interstitial brain Aβ levels in Tg mice within hours and improved cognitive performance to levels that were equal to or better than those of non-transgenic littermates. Non-transgenic mice were unaffected by PBT2 administration. A recent 12-week phase IIa double-blind, randomised, placebo-controlled trial of PBT2 was conducted on patients with mild-to-moderate AD (Lannfelt et al. 2008). In this trial, participants received oral PBT2 (50 mg or 250 mg) or placebo. Patients treated with 250 mg PBT2 showed significant diminution in frontal lobe functional deficits, as measured by the Category Fluency Test and Trail Making Part B, and significant decreases in CSF Aβ42 levels (Lannfelt et al. 2008). Importantly, no serious side-effects were noted in patients that received the PBT2 formulation (Lannfelt et al. 2008). Future trials on a larger AD patient pool will be necessary to determine the clinical utility of PBT2.

Tramiprosate (3-amino-1-propanesulfonic acid; AlzhemedTM, Neurochem, Inc., Laval, Canada) has been shown to bind preferentially to pre-fibrillar (monomeric and oligomeric) forms of Aβ, preventing their conversion into higher-order aggregates and fibrils (Wright 2006; Gervais et al. 2007). Studies using primary neuronal cultures from rat brain showed that tramiprosate decreased Aβ42-induced cell death by 38% (Gervais et al. 2007). Administration of tramiprosate to Tg mice that develop cerebral amyloidosis resulted in significant reductions in plasma Aβ (60%), soluble Aβ40 (30%), insoluble Aβ40 (31%), soluble Aβ42 (25%), and insoluble Aβ42 (22%) (Gervais et al. 2007). Tramiprosate crosses the blood–brain barrier and causes minimal toxicity in mice (Gervais et al. 2007). For these reasons, clinical trials of tramiprosate were undertaken. Unfortunately, the latest results from a phase III trial in patients with mild-to-moderate AD showed no significant benefits relative to placebo-treated controls (Yamin et al. 2008). Neurochem, the manufacturer of AlzhemedTM, has abandoned the study and apparently now is marketing this product as a dietary supplement (Yamin et al. 2008).

French and Danish epidemiological studies that moderate wine consumption may protect against the development of AD (Letenneur et al. 1993; Dorozynski 1997; Orgogozo et al. 1997; Truelsen et al. 2002). Using thioflavin T assay, SDS–polyacrylamide gel electrophoresis, surface plasmon resonance, electron microscopy, and atomic force microscopy, we previously found that wine-related polyphenols such as myricetin and tannic acid inhibit Aβ fibrils formation as well as destabilizing preformed Aβ fibrils by in vitro (Ono et al. 2003, 2004a; Hirohata et al. 2007). Destabilized Aβ fibrils may be less toxic than intact Aβ fibrils (Ono et al. 2003). Similarly, we also reported potent polyphenolic antioxidants such as rosmarinic acid and curcumin have similar anti-Aβ activities in vitro (Ono et al. 2004b). Using a combination of thioflavin T assay, PICUP, size exclusion chromatography, circular dichroism spectroscopy and electron microscopy, we recently showed that a commercially available polyphenol-containing grape seed extract, MaegaNatural, block early stages of Aβ assembly, that is, secondary structure transition, oligomerization and protofibril formation in vitro (Ono et al. 2008). Moreover, we showed that phenolic compounds such as myricetin, rosmarinic acid and MegaNatural inhibited both Aβ oligomerization and Aβ burden as well as attenuating cognitive deterioration in vivo (Wang et al. 2008; Hamaguchi et al. 2009). Using the THMT mouse model of tauopathy, we also showed that MegaNatural treatment significantly prevented tau neuropathology, that is the aggregation of phosphorylated tau into neurofibrillary tangles in vivo (Wang et al. 2010).

Curcumin has been suggested to be a preventive or therapeutic agent (for a review, see Hamaguchi et al. 2010). Cole’s group showed that curcumin decreases the amounts of soluble and insoluble Aβ as well as lessening plaque burden in many affected brain regions using Tg2576 mice (Lim et al. 2001). Later, they reported that curcumin inhibits the formation of Aβ oligomers and fibrils, binds plaques, and reduces plaque burden in vivo (Yang et al. 2005). They also reported that a cocktail containing both docosahexaenoic acid and curcumin could limit cognitive deficits and tau pathology in human tau mice beginning treatment at 14 months which is well after tau pathology and cognitive deficits had developed (Frautschy et al. 2010). In a randomized study on 34 patients with probable or possible AD using curcumin at doses of 4, 1, or 0 g/day, there was no significant difference in the changes in Mini-Mental State Examination scores for 6 months although curcumin appeared to cause no side effects (Baum et al. 2008). Although 24-week phase II clinical trials of curcumin C3 complex (curcumin, demethoxycurcumin, and bisdemethoxycurcumin) at doses of up to 4 g/day in persons with mild-to-moderate AD are ongoing, their data do not provide evidence for a large effect of curcumin on cognitive function and biochemical measures of plasma and CSF until July 2008 (Ringman et al. 2005, 2008).

Conclusions

The research on Aβ aggregation, especially oligomerization, has advanced intensively, and the therapies directed at preventing the formation of toxic Aβ aggregates will soon reach the clinical stage. Unlike current therapies that are limited to the treatment of the symptoms of AD, these new therapies may have the potential to delay or even halt further progression of AD. However, in order to succeed in these exciting clinical trials, it is essential to further develop our understanding of the aggregates and mechanisms that underlie the neurotoxicity of Aβ. There is growing evidence that soluble pre-fibrillar aggregates, especially, low-n oligomers of Aβ are proximate neurotoxins and that strategies directed at preventing the formation of low-n Aβ oligomers should prove effective. There are also projects to fully characterize the soluble, low-n oligomers actually present in human brain. Further clarification of the toxic soluble forms of brain Aβ will help us to develop more effective and safe therapeutics as well as novel diagnostic assays.

Acknowledgements

The authors thank Drs. T. Ikeda, A. Morinaga and T. Hamaguchi (Kanazawa University) for assistance in preparing manuscript.

This work was supported by Grant-in-Aid for Young Scientists (B) (K.O.) from the Japanese Ministry of Education, Culture, Sports, Science and Technology, Japan, a the Japan Human Science Foundation (K.O.), a Pergolide Fellowship from EliLilly Japan (K.O.), the Mochida Memorial Foundation for Medical and Pharmaceutical Research (K.O.), the Novartis Foundation for Gerontological Research (K.O.), Alumni Association of the Department of Medicine at Showa University (K.O.), a grant for Knowledge Cluster Initiative [High-Tech Sensing and Knowledge Handling Technology (Brain Technology)] from the Ministry of Education, Culture, Sports, Science and Technology, Japan (M.Y.), and a grant to the Amyloidosis Research Committee from the Ministry of Health, Labour, and Welfare, Japan (K.O. and M.Y.).

Ancillary