298
Chem. Res. Toxicol. 2010, 23, 298–308
The Cu(II)/Aβ/Human Serum Albumin Model of Control Mechanism for Copper-Related Amyloid Neurotoxicity Małgorzata Ro´zga† and Wojciech Bal*,†,‡ Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Pawin´skiego 5a, 02-106 Warsaw, Poland, and Central Institute for Labour Protection, National Research Institute, Czerniakowska 16, 00-701 Warsaw, Poland ReceiVed October 1, 2009
Alzheimer’s disease (AD) is the most common neurodegenerative disorder in the elderly population, above 65 years of age. Multiple lines of evidence confirm the central role of 40-42 residue Aβ peptides in the pathogenesis of AD, but exact mechanisms of Aβ toxicity remain unclear. Recently, evidence has accumulated in favor of small oligomers of the Aβ42 peptide as major toxic species. Metal ions, copper(II) in particular, have been implicated in molecular mechanisms of Aβ neurotoxicity, including oxidative damage of lipid membranes. While monomeric Aβ peptides are not neurotoxic, the deep understanding of their chemical properties is prerequisite for significant progress in Alzheimer research. Monomeric Aβ40 and Aβ42 form a specific mononuclear complex with Cu(II), recruiting donor atoms within their common 16 amino acid N-terminal sequence. The formation of this complex, the exact structure of which is debated, correlates with increased Aβ toxicity. Human serum albumin (HSA) is a versatile carrier protein present, among others, in blood and cerebrospinal fluid. It binds one Cu(II) ion with a high, picomolar affinity and one Aβ molecule with a moderate, micromolar affinity. In this perspective, we present a model of interactions, which make HSA a likely guardian against Cu/Aβ toxicity in extracellular brain compartments. Contents 1. Introduction 2. Basic Facts about Aβ Peptides 3. Oligomerization of Aβ Peptides 3.1. Aβ Oligomers as the Main Toxic Species in AD 3.2. Oligomers in Synthetic Aβ Peptides 4. The Impact of Cu(II) and Zn(II) on Aβ Peptides 4.1. Cu(II) and Zn(II) Ions and Aβ Aggregation 4.2. Cu(II) and Zn(II) Interactions with Aβ Monomer 4.3. The Roles of Zn(II) and Cu(II) Ions in Aβ Toxicity 5. HSA and Its Relation to Cu(II) and Aβ Peptides 6. The Cu(II)/Aβ/HSA Model of Control of Copper-Related Amyloid Neurotoxicity 7. Concluding Note
298 298 299 299 299 300 300 300 301 302 302
boom generation reaches the critical age of 65 (3). The disease usually begins slowly, and its common early symptoms such as confusion, short-term memory disturbances, and mood swings often pass unrecognized. As AD inevitably progresses, symptoms become more severe, and the total loss of executive functions is eventually observed. This frightening perspective boosts vigorous research in all aspects of AD, including molecular mechanisms underlying the disease. The reactivity of amyloid β (Aβ) peptides is a prominent area of these efforts. In this perspective, we would like to propose a model of interactions in the Cu(II)/Aβ/human serum albumin (HSA) triangle that could constitute a control system preventing copperdependent as well as copper-free neurotoxicity of Aβ peptides.
2. Basic Facts about Aβ Peptides 304
1. Introduction Alzheimer’s disease (AD) is the most common neurodegenerative disorder in the elderly population, above 65 years of age. It accounts for 50-60% of all cases of dementia. According to current estimates, more than 30 million individuals worldwide are affected by AD. This number is going to raise rapidly due to a significant increase in life expectancy in both developed and developing countries. It is predicted that by the year 2040 as many as 81 million people will have developed AD worldwide (1, 2). In the United States, for example, AD is the only major cause of death currently on the rise. Moreover, it is expected to explode in the coming years, as the postwar baby * To whom correspondence should be addressed. Tel: +48-22-5922346. Fax: +48-22-6584636. E-mail
[email protected]. † Polish Academy of Sciences. ‡ National Research Institute.
The presence of intracellular neurofibrillary tangles and extracellular amyloid plaques (senile plaques) in the brains of patients is the most characteristic postmortem feature of AD. The tangles consist of hyperphosporylated τ protein, described in detail elsewhere (4), and the plaques are composed predominantly of insoluble fibrils, formed by aggregated Aβ peptides (5, 6). These peptides are proteolytic products of β-amyloid precursor protein (APP), a transmembrane, receptor-like protein that is expressed widely in neuronal as well as non-neuronal cells. The physiological function of the native APP remains unknown (7). The release of Aβ from APP requires the action of two aspartyl proteases, referred to as β- and γ-secretases, acting in a sequence according to a regulated intramembrane proteolysis mechanism. APP is first trimmed within its extracellular domain by β-secretase. The resulting membraneassociated C-terminal fragment is then cleaved by the γ-secretase complex. The free Aβ peptide is thereby released into extracellular space (8-11). In an alternative pathway, present
10.1021/tx900358j 2010 American Chemical Society Published on Web 12/01/2009
PerspectiVe Scheme 1. Sequences of Full Length and Truncated Aβ Peptides Used in AD Research
in neuronal cells, Aβ peptides can also be released intracellularly (12). The γ-secretase cleavage site is slightly variable. As a result, Aβ peptides differ in length, ranging from 38 to 42 amino acids. The two most abundant forms, 40 and 42 amino acids long, are named Aβ40 and Aβ42, respectively (Scheme 1). Aβ40 predominates in biological fluids, whereas Aβ42, which accounts for less than 10% of the total Aβ secreted, is the major constituent of senile plaques (13). Although multiple lines of evidence confirm the central role of Aβ peptides in the pathogenesis of AD, the exact mechanism of Aβ toxicity remains unclear. Rather strikingly, it has become clear that Aβ peptides are continuously produced throughout life and circulate in extracellular fluids, including cerebrospinal fluid (CSF) and blood plasma. This finding disproved the early concept that the mere presence of Aβ peptides would suffice to initiate neuronal dysfunction and degeneration observed in AD patients (14-16). The existence of other factors triggering Aβ toxicity thus became obvious. Self-association has long since been considered to be such a factor. Extended by two C-terminal residues, Aβ42 is by far more prone to self-assembly into ordered structures than Aβ40 (17-19). This molecular property is also demonstrated in vivo. For example, in transgenic BRIAβ40 mice, producing 2-3 times more Aβ40 than the Tg2576 control mice, no Aβ deposits were formed by 24 months of age. In contrast, analogous BRI-Aβ42 mice developed parenchymal and cerebrovascular Aβ deposits as early as 3 months of age, despite a ∼5-10-fold lower transgene expression than BRI-Aβ40 mice (20). In humans, most point mutations in APP and γ-secretase complex components (presenilins PS1 and PS2), linked to early onset familial forms of AD (FAD), result in an increased production of Aβ42, rather than Aβ40 (9, 21, 22), thus suggesting enhanced pathogenicity of Aβ42 as compared to Aβ40. On the basis of the coincidence of higher toxicity and more efficient deposition of Aβ42, the formation of plaques containing mature amyloid fibrils, the end point of Aβ aggregation, was initially considered to be the most probable source of AD neuropathogenesis, according to the amyloid cascade hypothesis (23, 24). It was assumed that in healthy brains, while present, Aβ existed only as soluble, nontoxic monomers (14). This hypothesis failed, however, to explain the fact that many elderly humans who showed no ante-mortem symptoms of dementia had substantial amounts of amyloid plaques in limbic areas of their brains demonstrated by postmortem examinations (25-27). In diagnosed AD patients, little correlation was observed between the plaque density and the degree of cognitive function impairment. Also, multiple experimental animal studies demonstrated that the development of structural and functional neuronal deficits responsible for memory impairment and cognitive decline preceded the formation of extracellular plaques (28). For example, in transgenic mice expressing a FAD mutant of APP, a significant deficit of synaptic transmission, detected already at 1-4 months of age, was not accompanied by amyloid deposition (29). Other studies confirmed the independence of these phenomena (30, 31). Also, the aforementioned BRI-Aβ42 transgenic mice had no obvious neuronal cell loss despite the presence of a large number of Aβ deposits (20). These findings testified against the causative role of amyloid deposits in AD
Chem. Res. Toxicol., Vol. 23, No. 2, 2010 299
pathogenesis and suggested a role of plaque-independent Aβ toxicity in the neurodegeneration. Indeed, other experiments correlated the extent of synaptic deficits and cognitive impairment with levels of nondeposited Aβ rather than with a plaque load (32-35). In accord with this line of evidence, the association of soluble Aβ forms with neuronal membranes is widely considered as a major route of Aβ neurotoxicity (36).
3. Oligomerization of Aβ Peptides 3.1. Aβ Oligomers as the Main Toxic Species in AD. In fact, insoluble amyloid is not the sole product of Aβ selfaggregation. Apart from fibrils, the peptide easily converts into diverse oligomeric states, described briefly below (the term “oligomer” is used in the literature rather liberally with respect to Aβ peptides, covering all water-soluble nonfibrillar assemblies that resist sedimentation upon high-speed centrifugation). Their diffusible nature enables them to flow into the synaptic clefts. Therefore, Aβ oligomers seem to be more reliable candidates for inducing the earliest symptoms of AD, such as neural transmission impairment, than amyloid deposits (37, 38). The presence of sodium dodecyl sulfate (SDS)-stable low molecular weight (low-n) Aβ oligomers was reported in human AD brains (for the sake of simplicity, in this perspective, we apply the “low-n” term equally to all types of Aβ oligomers of low molecular weight, without distinguishing between synthetically and naturally derived species). Western blot analyses detected the dimers and trimers of Aβ both in samples of amyloid plaque core (39-41) and in buffer-soluble fractions of human cerebral cortex (33). The presence of stable dimers of Aβ was also demonstrated in human CSF (42, 43). SDSstable dimers and trimers (at 8 and 12 kDa, respectively, were also abundantly (10-20% of total of immunoprecipitated Aβ) present in conditioned media of Chinese hamster ovary (CHO) cells expressing the human APP gene (44, 45). Aqueous buffer-soluble, SDS-stable Aβ42 hexamers (27 kDa), nonamers (40 kDa), and dodecamers (56 kDa) were detected by Western blots in soluble extracellular-enriched extracts from the brains of Tg2576 mice expressing human APP protein. The 56 kDa species appeared in brain extracts at 6 months of age, coincident with the development of the first memory deficit symptoms in these mice (46). However, the presence of these higher-n Aβ assemblies has yet to be reported in human CSF or soluble extracts of human brain (47). The potential of various Aβ forms to trigger AD was investigated by monitoring their relative ability to block hippocampal long-term potentiation (LTP), which is believed to be the earliest AD manifestation (48, 49). The Aβ fractions were microinjected into lateral ventricles of rat brains. Inhibition of LTP in these experiments was ascribed solely to oligomers and not to monomers or fibrils (50, 51). Similar effects were seen in cultured neurons, grown in the presence of subnanomolar concentrations of natural Aβ oligomers. The trimer, separated by size exclusion chromatography (SEC), appeared to be the most potent LTP blocker (52, 53). 3.2. Oligomers in Synthetic Aβ Peptides. Aβ dimers and trimers can be generated in vitro from the synthetic Aβ42 upon its prolonged incubation at 37 °C. However, depending on experimental conditions, a plethora of other neurotoxic oligomeric states can be generated in a test tube, from both Aβ42 and Aβ40 (54, 55). Protofibrils (PFs) are soluble, metastable structures formed in Aβ40 solutions via low-n Aβ oligomers. These curvilinear structures of 4-11 nm diameter and