Live Discussion: Now You See Them, Now You Don't: The Amyloid Channel Hypothesis
||Ever since amyloid-β (Aβ) was identified as the major component in amyloid plaques, scientists have been trying to decipher just exactly what the little peptide does. Most researchers would agree that Aβ is toxic to cells, but some are still skeptical. Even among the majority, agreement soon ends. Exactly what form of Aβ is toxic and how? These remain crucial, controversial, and unresolved issues. Are dimers, trimers, dodecamers, or even larger conglomerates of Aβ the principal toxic species? Do oligomers cause oxidative stress, interfere with signal transduction, scupper synaptic transmission, aggravate the immune system, or in some other way damage cells? Is it intracellular or extracellular Aβ that is the culprit, or is it both?
One theory that has received some attention is commonly referred to as the "channel hypothesis." This theory suggests that oligomeric forms of Aβ are hollow structures that allow small ions to pass through. Being hydrophobic, these oligomers can infiltrate the lipid bilayers that surround cells and organelles, and because they are big enough to span these bilayers, they can effectively punch a channel through the membranes, allowing small ions to stream through. For neurons, which spend much of their energy segregating ions across the cell membrane, this spells disaster. For mitochondria, also sensitive to perturbations in ionic strength, this can mean the initiation of an apoptotic cascade that equally spells trouble for the host cell.
- Amyloid peptides from Alzheimer's disease (Aβ), Parkinson's, Huntington's, Prion diseases and Type II diabetes can all form ion channels in planar lipid bilayers.
- These channels share common properties: heterodispersity, irreversibility, non-selectivity, long open times, blockade by zinc, inhibition by Congo red, enhancement by "aging" or acidic pH. These properties would lead to cells becoming "leaky," running down ionic gradients, dysregulating calcium, and using up energy supplies.
- Aβ channels induce ionic currents in neurons, induce increased mitochondrial permeability, are cytotoxic to fibroblasts. (This toxicity is blocked by channel blockers such as zinc).
- Aβ channels inhibit LTP, and Aβ variants which do not form channels, do not inhibit LTP.
- Several amyloid peptides form ring structures, and mutations which cause disease enhance the likelihood of ring structures.
- Channel formation in plasma and/or intracellular membranes could explain the pathophysiology, cellular dysfunction and cytotoxicity observed in amyloid diseases.
Bruce Kagan led this live discussion on 16 June 2005. Readers are invited to submit additional comments by using our Comments form at the bottom of the page.
View Transcript of Live Discussion — Posted 22 August 2006
By Bruce L. Kagan
In this live discussion, I will argue that amyloid-β forms membrane channels, and that this is the form in which the peptide exerts its toxicity in Alzheimer disease. I invite colleagues and readers who agree, but especially those who don't, to join me in a discussion of the merits and weaknesses of this hypothesis. Together, we can identify what are the stumbling blocks to faster progress at this point in time, and what experiments, collaborations, or infusion of new technologies could help the field get past them.
But first, a look at history and background: In the 19th century, the German physician Rudolf Virchow was the first to describe "amyloid" as an amorphous, starch-like deposit in tissues. The deposits stained with iodine, indicating the presence of carbohydrate. Investigators in the next century concluded that amyloid deposits were mostly made of protein, which stained with the dye Congo red in a characteristic fashion, making that dye the standard first-pass test for the presence of amyloid for many years. It also became clear that amyloid deposits were composed of fibrils with a characteristic structure (See Sipe & Cohen, 2000, for review). Fibrils are 80-100 angstroms wide and often very long. Many different proteins can form amyloid deposits. The count is up to 21 by now (see Table 1) and they are associated with a wide variety of human illnesses (Merlini and Westermark, 2004). Besides amyloid fibrils, amyloid deposits also contain myriad nonspecific components, such as glycosaminoglycans, proteoglycans, and the pentraxin serum amyloid P. Their structural role is uncertain.
Although amyloid-forming proteins share no amino acid sequence homology, they all adopt a characteristic β-sheet structure within amyloid fibrils. The β-sheet conformation is essential to the formation of amyloid. However, its role in determining pathology has been less clear, in part because amyloid deposits occur in apparently healthy tissues as well as being frequently associated with disease. Indeed, many studies suggested that amyloid deposits were associated with disease and even played an etiologic role (e.g., see Hardy and Alsop, 1991). But the issue kept bogging down over the irregular and unpersuasive correlation of clinical illness with amyloid burden.
More recent studies have suggested that smaller, soluble aggregates of many proteins may be responsible for cellular dysfunction and death. I'll call these early aggregates "oligomers" here, but keep in mind that this term might refer to species containing as few as two or as many as thousands of molecules. More precise estimates may come from future studies. The important distinction is that these oligomers are soluble, at least for a time, whereas fibrils precipitate out of solution.
Quite a few common, costly, and devastating diseases are amyloidoses. These include Alzheimer disease (AD), type II diabetes mellitus (DM), and the spongiform encephalopathies or prion diseases including scrapie, "mad cow" disease, and new variant Creutzfeld-Jakob disease (see Table 1). The aging of the population in the developed world has caused an epidemic of these age-related diseases, making the search for an understanding of disease pathogenesis more critical than ever. A great deal of research has strongly implicated amyloid proteins in the pathogenesis of these illnesses. Even so, the molecular mechanisms by which these misfolded proteins cause dysfunction and/or toxicity remain elusive. (See Merlini and Berlotti, 2003, for review).
In the space below, I will review the evidence that amyloid oligomers, rather than fibrils, are the pathogenic species. I will keep this section brief, as many investigators already share this view. I will then review current evidence on the more fluid question of how these oligomers cause cellular dysfunction and/or toxicity. This is where I put forth my channel hypothesis. I will include reports suggesting that protein misfolding and aggregation are linked to the pathophysiology of a number of other, nonamyloid diseases. Yet further hints come from reports that proteins unrelated to disease can sometimes form amyloid oligomers that then become toxic. The links among these various phenomena will be explored.
The early findings that amyloid burden in patients did not correlate well with clinical severity of illness was troubling to those who thought amyloid played a causative role in amyloidoses. However, the discovery of familial forms of amyloidoses linked to mutations in amyloid proteins gave new impetus to the notion that these proteins "caused" the illness (for example, Samaia et al., 1997; Pepys et al., 1993). Further research showed that amyloid peptides could damage isolated cell systems, and that these peptides were toxic to the cell types damaged in diseases. In this way, the Aβ42 peptide of AD (Yankner et al., 1990) and then PrP106-126 (Forloni et al., 1994) were shown to be toxic to neurons, and islet amyloid polypeptide (IAPP) was shown to be toxic to β cells from pancreatic islets of Langerhans (Lorenzo, et al., 1994). Amyloid fibrils themselves usually lacked significant toxicity, as did monomers.
In early experiments, however, cytotoxicity was poorly reproducible, especially for the Aβ peptides. This took years to sort out but eventually proved to be due to differences in the aggregation status of Aβ under different experimental conditions (Pike et al., 1993). The aggregation of amyloid peptides turned out to be a highly complex and variable process (Harper and Lansbury, 1997). The initial phases of protein aggregation from monomers into small oligomers progressed quite slowly, but once a critical mass was reached, aggregation into large species sped up. What's more, adding preformed "seeds" to a solution of peptide in effect "nucleated" the growth of fibrils. (See also Alzforum interview with Peter Lansbury.)
In exploring in-vitro aggregation further, we soon discovered that while promoting some amyloid peptide aggregation (by acidic pH, high concentration, "aging," or seeding) was necessary to achieve cytotoxicity, too much aggregation actually led to a decline in toxicity (Hirakura et al., 1998). Some intermediate state of aggregation appeared to be where amyloid peptides exert the most potent toxicity.
At the same time, the notion of changes wrought by the amyloid peptide broadened. While cytotoxicity was clearly relevant to diseases in which cells died, animal models of AD created with transgenic mice expressing amyloid precursor protein (APP) or Aβ developed amyloid deposits, learning and memory deficits, and synaptic loss, but no frank neuronal loss. Some criticized the models as inadequate, or attributed the lack of nerve loss to differences between mice and men. Yet others began to think that perhaps in the early stages of AD, dysfunctions similar to those seen in the mice preceded neuronal loss. Investigations of early stages of transgenic mouse models showed that memory deficits actually preceded amyloid deposits and synaptic loss. Further studies showed that Aβ peptides could inhibit long-term potentiation, a model for memory (Chen et al., 2000). But only Aβ oligomers did that, not monomers or fibrils (Walsh et al., 2002). Furthermore, fibroblasts died from exposure to Aβ that was freshly prepared (and thus less highly aggregated) and appeared "globular" by atomic force microscopy. That toxicity could be blocked by antibodies or the channel-blocking zinc ion, but not antioxidants (Zhu et al., 2002). Thus, oligomers were implicated in cellular dysfunction relevant to clinical symptoms and to cytotoxicity. This idea is supported by the "arctic" mutation of Aβ, which enhances oligomer formation, and which shows increased cytotoxicity compared to native Aβ peptide (Lashuel et al., 2003).
A similar story emerged in Huntington disease (HD), where early deficits in learning and memory could be linked to the presence of aggregates of the disease-related protein huntingtin, which contains a long polyglutamine (polyQ) tract. These deficits appear well before any frank neuronal loss in the brains of affected transgenic mice. Electrophysiologic abnormalities accompanied these deficits and could be detected even prior to the appearance of huntingtin aggregates, suggesting that the large visible aggregates may not be responsible for the early cellular effects of polyQ huntingtin (Cepeda et al., 2003). It has also been shown that polyQ aggregates targeted to the nucleus can be directly cytotoxic (Yang et al., 2002).
Oligomers have also been implicated in the cytotoxicity of α-synuclein (AS) and IAPP. Mutations in AS that promote "protofibril" (oligomer) formation also enhance cytotoxicity (Lashuel et al., 2002), while toxic oligomers of IAPP were shown by light scattering to contain between 25 and 6,000 IAPP molecules (Janson et al., 1999). Once again, fibrils and monomers were shown to lack toxicity.
That all these amyloids might have structures in common was suggested by an antibody to the "soluble oligomer" conformation, which recognized several different amyloid peptides, but only in a precise conformation (Kayed et al., 2003). This antibody could block cytotoxicity of a range of amyloid peptides, strongly suggesting that oligomers were the toxic species, and that the toxic mechanism (or at least a part of it) was common to all the amyloids.
Biological dogma has maintained that the three-dimensional structure and function of proteins is dictated by their primary amino acid sequence. Our knowledge of amyloid diseases has led to the recognition that natively folded proteins may unfold at times and adapt non-native conformations. Proteins may refold to their native conformation (with or without the assistance of chaperones), or they may be targeted for degradation via the ubiquitin-proteasome system (UPS). Alternately, some partially unfolded or misfolded proteins may aggregate. This aggregation may be driven thermodynamically by new hydrogen bonding possibilities (Fernandez and Berg, 2003) or by the hydrophobic effect—the need to shield hydrophobic parts of proteins from the aqueous environment. Proteins that form amyloid appear to have a propensity to misfold into β-sheet structures. These β-sheets tend to self-aggregate, forming intermolecular bonds that aggregate the proteins or peptides.
It is still not well-understood why these complexes are not handled by the cell's usual defense mechanisms, that is, refolding to native structures through chaperone assistance or targeting for degradation through the UPS. It may be that these processes are at work, but are merely overwhelmed by the amount of misfolded protein they must handle. This might be how disease begins. Alternately, it is possible that the nature of these β-sheet aggregates leads them to attack or insert into cellular membranes. This would render them inaccessible to chaperones and to the UPS. Another possibility is that once seeding or nucleation has occurred, the fibrillization process proceeds so rapidly that the aggregates become too large to be handled by the cell's defenses. It has been suggested that fibrillization, with or without vacuolization, is actually a cellular defense mechanism to isolate the misfolded protein and keep it separate from the cell's vital machinery. Support for this idea comes from experiments showing that amyloid fibrils have little, if any, toxicity to cells, while smaller aggregates appear to possess considerable cytotoxicity. It has also been reported that inclusion bodies containing huntingtin appear to protect neurons from dysfunction and death (Arrasate,et al, 2004).
This initial misfolding of a protein may be driven by mutations, metal interactions, and environmental conditions such as temperature, acidic pH, oxidation, proteolysis, or a simple increase in concentration. All these phenomena can potentially increase the presence of misfolded conformations and increase the probability of aggregation. The presence of membranes may also play a key role in this process as lipid bilayers have been shown to influence both the conformation of amyloid peptides and the propensity of amyloid peptides to aggregate (McLaurin et al., 1998), and proteins that possess hydrogen bond defects are more likely to interact with lipid bilayers (Fernandez and Berg, 2003). Thus, the intracellular compartment in which an amyloid protein finds itself may significantly affect its ultimate conformational fate.
Among the many different amyloid peptides, there is no primary sequence homology, but they all share the capacity to assume structures with relatively high β-sheet content. Sometimes this β-sheet content is unpredictable, as in the prion protein where sequences which would usually be predicted to be α helical adopt β-sheet conformation when synthesized (De Gioia et al., 1994).
While the earliest stages of amyloid peptide aggregation cannot be easily observed with current imaging techniques, the results of recent electron and atomic force microscopic studies have shown the early presence of spherical or globular aggregates. The smallest particle appears to aggregate into chains or annular rings referred to as "protofibrils." These structures appear to be the precursors of fibrils that exist in mature amyloid deposits. The protofibrillar structures possess significant toxicity compared to fibrils themselves. It is unclear whether the annular structures or rings are more toxic than their linear brethren, but it is of interest that pathogenic mutations of β amyloid and α-synuclein lead to increased formation of these annular structures (Lashuel et al., 2002).
Cellular Mechanisms of Oligomeric Toxicity
Recently, it has been reported that amyloid disease proteins and peptides are not unique in their ability to form amyloid. Aggregation and amyloid fibril formation has been achieved for a number of non-disease-associated proteins (Fandrich et al., 2003). It has been proposed that under appropriate conditions, all proteins might be capable of forming amyloid. Beyond the structural similarity of these non-disease amyloid proteins to true amyloid, there are functional similarities as well. For example, HypF is able to aggregate, and the aggregates can kill cells and permeabilize liposomes (Relini et al., 2001). Fibrils lack these properties, but aggregation of HypF is required for them, thus implicating smaller oligomers as the cytotoxic, membrane permeabilizing species. Prior to cytotoxicity, HypF oligomers, but not fibrils, raise intracellular Ca2+ levels and levels of reactive oxygen species (Bucciantini et al., 2004). These intracellular effects were reversible, and they were prevented with the omission of Ca2+ from the media or the addition of reducing agents. These physiologic effects are strikingly similar to those observed with disease-related amyloid peptides (Schubert et al., 1995; Kawahara et al., 2000).
These results are also consistent with the findings that antibodies raised against soluble prefibrillar oligomers of various amyloid peptides recognize that state in other amyloid peptides and can prevent cytotoxicity (Kayed et al., 2003). This suggests that amyloid peptides share an oligomeric conformation critical to cytotoxicity and independent of amino acid sequence.
Thus, a wide variety of structural, biochemical, and physiologic approaches suggest that prefibrillar amyloid peptide oligomers act to damage and/or kill cells, especially neurons, and that they appear to share a common mechanism of action. As we discuss below, a growing body of biophysical evidence implicates channel formation by amyloid peptides in cellular membranes as the molecular mechanism of this damage.
The Channel Hypothesis
A large body of evidence has focused on the ability of amyloid peptides to interact with membranes. A series of provocative studies has shown that many (if not all) amyloid peptides can aggregate into β-sheet oligomers capable of spontaneously inserting into lipid membranes. The peptides form relatively permanent ion-permeable channel structures across the cell membrane. These channels are reported to have properties that would damage or kill most cells. The channels are reported to be (a) large, (b) nonselective, (c) heterogeneous, (d) voltage-independent, (e) irreversible, (f) inhibited by agents that prevent aggregation such as Congo red, and (g) blocked by zinc ion. These channels would likely cause a leakage pathway in plasma, endoplasmic reticulum, and mitochondrial, lysosomal, or other membranes. These leaks could damage cells by (a) disrupting membrane potentials and ion gradients, (b) causing loss of vital intracellular ions such as K+ and Mg2+, (c) allowing influx of toxic ions such as Ca2+, (d) running down energy stores by forcing ion pumps to work harder, (e) disrupting mitochondrial membrane potential and initiating apoptosis by allowing cytochrome c to leak out of mitochondria, and (f) allowing toxic enzymes and other factors to leak out of lysosomes and peroxisomes. In the sections that follow, we review the evidence for channel formation by the amyloid peptides which have been most studied by these techniques.
The "channel hypothesis" was first proposed by Arispe et al. (1993a) after they reported that Aβ1-40 could form cation-selective, calcium permeable channels of various conductances in planar lipid bilayer membranes (BLMs). The channels were formed by Aβ1-42 as well, and were large, voltage-independent, and blocked by tromethamine (tris+) and aluminum (Arispe 1993b) The largest channels observed (4nS) could potentially change the interior [Na] of a cell by as much as 10 μM per second. They proposed that ionic leaking of Na+, K+, and Ca+2 could disrupt membrane potential and ionic regulation within a few seconds.
While these findings were not immediately confirmed by other laboratories due to problems with irregular aggregation of Aβ (for example, Mirzabekov et al., 1994), eventually a series of studies found Aβ peptides capable of forming channels in BLMs (Hirakura et al., 1999; Kourie et al., 2001), liposomes (Lin et al., 1999), neurons and oocytes (Fraser et al., 1997), and fibroblasts (Zhu et al., 2000). The aggregation state of Aβ is critical not only to its cytotoxicity properties (Pike et al., 1993; Hirakura et al., 1998), but to its channel forming abilities. Indeed, monomers and fibrils of Aβ, which are nontoxic, fail to form ion channels (Hirakura et al., 1999). Oligomeric species of Aβ form a variety of channel entities which can be distinguished by their single-channel conductance, ionic selectivity, kinetics, and other channel properties (Kourie et al., 2002). It was also found that conditions (aging, acidic pH, etc.) which favored the aggregation of monomers into oligomers led to an increase in channel activity (Hirakura et al., 1999). Exposure to organic solvents, which monomerized Aβ, led to loss of channel activity. However, this activity could be recovered by allowing the peptide to "age" (aggregate) in aqueous solution.
Although Aβ1-40 and 1-42 are the primary Aβ peptides found in vivo, other Aβ fragments have been studied with interest. Aβ25-35, a cytotoxic peptide not found in vivo, is a voltage-dependent, nonselective channel former (Mirzabekov et al., 1994). Studies using variants of Aβ25-35 showed that channel formation was necessary for cytotoxicity, but not sufficient, that is, all cytotoxic species formed channels but there were two channel-forming variants of Aβ25-35 which did not kill cells (Lin, 1996, PhD Thesis). Channel activity could be enhanced by lipids carrying a net negative surface charge, and this effect could be countered by high salt concentrations. Cholesterol, which stiffens membranes, decreased Aβ25-35 channel activity (Lin and Kagan, 2002). Aβ25-35 variants could not form channels if they were not at least 10 residues long, indicating a minimum bilayer spanning length of about 30 angstroms, a result consistent with the β-sheet span lengths of the known channels made by Staphylococcal α toxin and anthrax toxin (Song et al., 1996; Petosa et al., 1997).
In vivo, Aβ1-40 and Aβ1-42 can induce currents in rat cortical neurons (Weiss et al., 1994; Furukawa et al., 1994), HNT cells (Sanderson et al., 1997) and GnRH secreting neurons (Kawahara et al., 1997). The channels observed in vivo seem indistinguishable in their properties from those observed in vitro.
Aβ1-40/Aβ1-42 can also kill fibroblasts in a manner inhibited by antibodies, tromethamine, or zinc, but not by antioxidants, suggesting that channel formation is the mechanism of cytotoxicity. The freshly prepared Aβ used in these studies appeared "globular," consistent with an early stage of aggregation (Bhatia et al., 2000; Zhu et al., 2000). It has also been shown that the cholesterol content of plasma membranes affects a cell's vulnerability to Aβ1-40/Aβ1-42 (Arispe and Doh, 2002), suggesting that the membrane plays a critical role in Aβ cytotoxicity. More recently it has been reported that Aβ can directly induce cytochrome c release from mitochondria. This could occur through a decrease of mitochondrial membrane potential or even through Aβ channel-mediated release of cytochrome c (Kim et al., 2002).
The prion protein (PrP106-126) has at least two distinct tertiary conformations, PrPc and PrPsc, the latter of which results in a transmissible neurodegenerative disease known as a spongiform encephalopathy. Prion diseases include scrapie in sheep and "mad cow" disease as, well as Creutzfeld-Jakob, Gerstmann-Straussler-Scheinker syndrome (GSS), and fatal familial insomnia in humans. These illnesses may be sporadic, infectious, or hereditary. The familial versions are associated with mutations in the prion protein (DeArmond and Prusiner, 2003 for review). PrP106-126 deposits in the brains of afflicted organisms in the form of amyloid fibrils. A critical step in the conformational transition from PrPc to PrPsc is the conversion of α helical and random coil regions of PrP to β-sheet (Pan et al., 1993). One region predicted to be α helical, PrP106-126, actually forms β-sheets when synthesized and self-aggregates into amyloid fibrils (Gasset et al., 1992). The β-sheet-rich form of PrP106-126 binds to membranes, unfolds, and ultimately disrupts the bilayer (Kazlauskaite et al., 2003). Forloni et al. (1993) demonstrated that PrP106-126 was toxic to neurons in culture. Lin et al. (1997) reported that PrP106-126 could form ion-permeable channels in planar lipid bilayer membranes at neurotoxic concentrations. PrP106-126 channels were irreversibly associated with the membranes, demonstrated a multiplicity of single-channel conductances (10-400 pS in 0.1MNaCl), and had relatively long lifetimes (seconds to minutes). Ionic selectivity of the channels was meager, with significant permeability being shown to Na+, K+, Cl-, and Ca2+ (P(Na+)/P(Cl-) = 2.5). Channel activity could be enhanced dramatically by "aging" of the peptide in aqueous solution, a procedure which promotes aggregation and increases neurotoxicity. Incubation of PrP106-126 at acidic pH also enhanced channel activity by nearly 100 times and shifted the distribution of observed single-channel conductances to higher conductance levels. It has also been reported that acidic pH converts α helical PrP106-126 to β-sheet conformation (De Gioia et al., 1994). Kourie and Culverson (2000) characterized three distinct channel types formed by PrP106-126. These included: (a) a dithiodipyridin sensitive channel of 40pS with slow kinetic behavior; (b) a giant channel, 900-1500pS, exhibiting five separate subconductance states; and (c) a triethylammonium (TEA) sensitive channel of 140pS with rapid kinetics.
Bahadi et al. (2003) have reported that PrP82-146, a peptide that forms amyloid fibrils found in the brains of patients with Gerstmann-Straussler-Scheinker syndrome, can also form ion channels. Scrambling the amino acid sequence of the 106-126 region of this longer peptide abolishes channel activity, whereas scrambling the 127-146 region has no effect on channel activity, thus implicating the 106-126 region as key in channel forming ability. The electrophysiologic properties of PrP82-145 are very similar to those of PrP106-126. Channel activity could be decreased by the antibiotic rifampicin, which had previously been shown to decrease aggregation and toxicity of Aβ peptides.
Amyloid deposits are observed in most, but not all, prion diseases. Intriguingly, in one prion disease where amyloid fibrils are not found, the mutant prion protein adapts a transmembrane conformation (Hegde et al., 1998). It is tempting to speculate that this transmembrane protein may be causing ionic leakage across the cell membrane.
Lysosomotropic agents have been reported to inhibit PrPsc accumulation in neuroblastoma cells (Doh-Ura et al., 2000). One of these agents, quinacrine, has been reported to block PrP106-126 channels (Farrelly et al., 2003). Quinacrine is also able to repair the impaired functioning of N-type calcium channels in prion-infected neurons (Sandberg et al., 2004). Thus, it seems likely that channel blockers such as quinacrine may be useful as potential therapeutic agents in prion-related diseases. Indeed, there is at least one report of quinacrine improving the clinical status of four patients with Creutzfeld-Jakob disease (Nakajima et al., 2004). Other acridine derivatives and tricyclic compounds may have even better antiprion efficiency (Korth et al., 2001). Congo red can inhibit channel formation, block PrP106-126 cytotoxicity, and inhibit the development of scrapie (Hirakura et al., 2001; Ingrosso et al., 1995). It remains to be seen whether the antichannel blocking or antiaggregation activities of quinacrine are key to its antiprion effects.
Islet amyloid polypeptide (IAPP, amylin) is a 37-residue amyloidogenic hormone which is cosecreted with insulin from β cells in the islets of Langerhans in the pancreas. Amyloid deposits comprising IAPP are found in the islets of patients with type II diabetes and are positively correlated with β cell loss and clinical insulin requirements (Westermark and Wilander, 1978; Butler et al., 2003). IAPP is cytotoxic to β cells in culture (Lorenzo et al., 1994).
Although IAPP is α helical in aqueous solution, exposure to lipid membranes induced a transition to β-sheet structure (McLean and Balasubramanian, 1992).
Human IAPP formed ion-permeable channels in planar lipid membranes at cytotoxic concentration (Mirzabekov et al., 1996), but rat IAPP, which differs from human IAPP at five amino acid positions and is nonamyloidogenic and nontoxic, did not form channels. Human IAPP channels inserted into membranes irrespective of voltage, however; once inserted, channels rapidly opened at negative voltages and rapidly inactivated at positive voltages (voltages being relative to the IAPP containing side). Inactivation faded gradually over a time course of several minutes. Open IAPP channels were ohmic and exhibited a single-channel conductance of 7.5pS in .01M KCl. Channels were permanently associated with the membrane and showed lifetimes of seconds to minutes depending on voltages. Increasing concentrations of net negatively charged lipids in the membrane lead to an increase in IAPP channel activity. Increasing salt concentrations in the aqueous solution decreased channel activity.
Large fibrils of IAPP have been reported to be nontoxic, whereas smaller aggregates possess cytotoxicity (Janson et al., 1999). These aggregates, but not fibrils, could disrupt planar lipid bilayers. Light scattering showed these oligomers to range in size from 25 to 6,000 IAPP molecules.
Anguiano et al. (2002) showed that liposomes could be permeabilized by IAPP in a graded fashion, allowing Ca2+ to cross the membrane, while not allowing fura-2 (MW 832) or FITC-Dextran (MW 4400) to escape. IAPP has also been reported to disrupt Ca2+ homeostasis in cells in a manner similar to Aβ and prion-related peptides (Kawahara et al., 2000).
Hirakura et al. (2000a) showed that Congo red incubation with IAPP, Aβ, or PrP106-126, prior to membrane exposure, could inhibit channel formation. They also reported that Zn2+ could reversibly block these channels. The concurrence of channel-forming properties, physiologic effects, and cytotoxicity strongly suggests a common mechanism of action for these three amyloid peptides.
Serum amyloid A (SAA) refers to a group of related apolipoproteins. During states of infection or inflammation, the acute phase isoforms of SAA can increase their levels in serum by as much as three orders of magnitude. Fibrils comprising the N-terminal 76 residues of SAA are found as amyloid deposits in various organs such as spleen, kidney, and liver. Patients with chronic infections such as tuberculosis or inflammatory diseases such as rheumatoid arthritis are particularly at risk. Patients with cancer, arteriosclerosis, and Alzheimer disease have also been reported to have SAA fibril deposits (Sipe, 2000).
One acute-phase isoform, SAAp, has been reported to form ion-permeable channels in planar lipid bilayer membranes at physiologically relevant concentrations (Hirakura et al., 2002). A wide variety of single-channel conductances (10-1000pS) were observed, consistent with a peptide aggregated into multiple oligomeric states. SAA channels were permeable to most physiologic ions including Na+, K+, Ca2+ and Cl-, exhibiting only a weak preference for cations over anions. Channel formation could be inhibited by preincubation of SAA with Congo red, but adding Congo red after channel formation had no effect. Channels could be reversibly blocked by 100uM Zn2+.
The related acute-phase isoform SAA1 was reported to lyse bacterial cells when expressed in E. coli, whereas expression of the constitutive isoform SAA4 did not. SAA1 and SAA4 differ in their sequences at approximately 50 percent of residues. SAA1 has a greater concentration of hydrophobic residues in the N-terminal region. The resemblance of these results to the properties of channel-forming toxins such as colicins (Schein et al., 1978), yeast killer toxins (Kagan, 1983), defensins (Kagan et al., 1990), and protegrins (Sokolov et al., 1999) led to the conjecture that SAA might play a role in host defense against microbes.
Electron microscopy revealed that murine SAA 2.2 could exist as an annular hexamer with a central "pore-like" region (Wang et al., 2002). Although membranes were not present, the observed pore diameter of 25 angstroms was consistent with the physiologic findings of Hirakura et al., 2002.
Parkinson disease (PD) is a progressive neurodegenerative disorder characterized by tremor, rigidity, and bradykinesia. The hallmark lesion of PD is the Lewy body, an inclusion body in dopaminergic neurons consisting largely of NAC (nonamyloid component, residues 66-95 of α-synuclein). Incorrectly named, NAC is actually a fibril-forming amyloid peptide. Mutations in α-synuclein, a synaptic protein associated with vesicles but of unknown function, can lead to familial PD and implicates α-synuclein in the pathophysiology of the illness (Baptista et al., 2004). NAC is also found in AD amyloid deposits, which suggests a link between the two amyloid diseases. This notion is supported by the clinical overlap in these illnesses. AD patients are commonly found to have motor abnormalities. PD patients frequently have cognitive problems such as dementia and depression. Intermediate syndromes such as dementia with Lewy bodies also implicate α-synuclein in damage to neurons outside the dopaminergic system (McKeith et al., 2004).
α-synuclein has been reported to permeabilize liposomes in a graded manner to substances of increasing size. This "sieving" action is characteristic of "pore-like" transport systems (Volles and Lansbury, 2002). Pathogenic mutations in α-synuclein such as A30P and A53T accelerated the formation of oligomers (protofibrils) capable of permeabilizing activity (Volles and Lansbury, 2002). Electron microscopy revealed that α-synuclein formed annular, pore-like oligomers, and that the PD-related mutations enhanced the formation of these structures. The pathogenic "arctic" mutation of Aβ also showed a similar enhancing effect on these annular structures (Lashuel et al., 2002). Electrophysiologic studies of NAC have confirmed the formation of ion-permeable channels in lipid bilayers (Azimova et al., 2003). These channels have properties strikingly similar to those of other amyloidogenic peptides. Single-channel conductances are heterogeneous. Ionic selectivity is weak. Channels are irreversible and have extended lifetimes. Channel formation is inhibited by Congo red and channels are blocked by Zn2+.
β-2-microglobulin (β2M) forms amyloid deposits in bones and joints of patients on hemodialysis or peritoneal dialysis, a syndrome referred to as "dialysis-associated amyloidosis." This 99-residue peptide belongs to the MHC Class I complex which is involved in the presentation of foreign antigens to lymphocytes. β2M levels can rise 100-fold during states of renal failure (Drueke, 1998). Renal transplantation can lower β2M levels and clinical symptoms. β2M's physiologic effects include the induction of Ca2+ efflux from calvariae, bone resorption, and increasing collagenase production (Moe and Sprague, 1992; Brinckerhoff et al., 1989; Peterson and Kang, 1994). β2M is found in amyloid deposits as full-length native protein. In contrast to other amyloid protein "misfolding" diseases, the misfolding here appears to be solely a function of increased protein concentration in the serum, rather than mutation or proteolysis.
Channel formation by β2M was reported by Hirakura and Kagan (2001). A multiplicity of single-channel conductances were observed, ranging from 0.5-120pS, with 90pS being the most commonly observed size in 0.1M KCl. Channel lifetimes were typically extended and ionic selectivity was poor. β2M associated irreversibly with the membrane. Incubation of β2M with Congo red inhibited formation of channels. Zn2+ could reversibly block inserted channels. Channel formation could be accelerated by acidic pH, compatible with the idea that the acidosis/uremic state of renal failure could enhance the generation of pathologic oligomers of β2M.
AL or light-chain amyloidosis is characterized by fibrillar deposits of the variable domain of immunoglobulin light chains. Fibril assembly is dependent on the environmental conditions: For example, the process may be different on surfaces versus in solution (Zhu et al., 2000). AL deposits are frequently found on surfaces such as arterial walls and basement membranes. A recent study of AL aggregation using atomic force microscopy found that AL protein of the variable domain SMA could form annular aggregates similar to those seen with α-synuclein. The SMA annular aggregates were significantly larger, however. Acidic pH was critical to the formation of these aggregates. It was suggested that these annular species might form pores in membranes (Zhu et al., 2004).
Although triplet repeat diseases are not classic "amyloidoses," their pathology seems to involve protein misfolding and the accumulation of toxic protein aggregates. Huntington disease (HD) is the most common and best known of these hereditary illnesses which are caused by an expansion of the codon CAG which codes for glutamine (single amino acid code: Q). In Huntington disease, polyQ tracts longer than 37 residues cause disease, although this number varies amongst the different triplet repeat illnesses. Several different proteins are involved in the various triplet repeat diseases, but all are affected by an extended polyQ tract with a minimum threshold length for causing clinical symptoms. In Huntington disease, the best studied, amyloid-like neuronal aggregates of huntingtin, the polyQ expanded protein, correlates with disease progression in transgenic mice (see Li and Li, 2004, for review). Toxicity of huntingtin is proportional to polyQ tract length. Age of onset of illness is inversely proportional to repeat length, too. However, a PG12 cell line shows vulnerability to apoptosis without visible aggregates of huntingtin (polyQ repeat length = 150). Thus, visible deposit or aggregates may not be necessary to cause dysfunction or cell death. Indeed, some transgenic mice show electrophysiologic abnormalities in striatum before aggregates are visible (Cepeda et al., 2003.)
Channel formation by polyQ was reported almost simultaneously by Hirakura et al. (2000) and Monoi et al. (2000). The former group reported channels that were long-lived, nonselective, and heterogeneous in single-channel conductance size, ranging from 19-220pS in 0.1M KCl. Channel formation was increased by acidic pH. Unlike classic amyloid peptides, Congo red preincubation did not inhibit channel activity nor was Zn2+ able to block polyQ channels. These distinct properties indicate that polyQ aggregates are clearly different structurally from classic amyloids.
Monoi et al. (2000) reported that polyQ 40 could form cation-selective, long-lived channels with a single-channel conductance of 17pS in 1m CsCl. PolyQ 29 could not form channels, consistent with the 37-residue cutoff for clinical illness. They also proposed a structural model, the μ helix, for polyQ channels, a model which just spans the bilayer hydrophobic core at a length of 37 residues, again in agreement with the clinical data.
Further evidence of possible channel formation by polyQ in mitochondria was reported by Panov et al. (2002), who showed that Huntington disease mitochondria had decreased membrane potential and depolarized at lower Ca2+ levels than control mitochondria. Brain mitochondria from transgenic mice expressing huntingtin with a pathogenic polyQ tract exhibited a similar dysfunction. Electron microscopy revealed mutant huntingtin localized to mitochondrial membranes. Most strikingly, the mitochondrial defects could be reproduced by a fusion protein with a long polyQ repeat. These results suggest that polyQ tracts are toxic to mitochondria and likely act via a channel-forming mechanism, depolarizing mitochondria and leaving them more vulnerable to apoptosis or metabolic insults. These data are consistent with a report that Aβ peptide can directly induce cytochrome c release in isolated mouse brain mitochondria by directly inducing a permeability increase in the mitochondrial membrane (Kim et al., 2002; Rodrigues et al., 2000).
The recent discovery that nondisease-related amyloid proteins exist has led to a re-examination of the nature of amyloid structure. It has been reported that a wide variety of nondisease proteins can form amyloid under the "appropriate" conditions, suggesting that the amyloid-β-sheet conformation is more universal than previously thought (Stefani and Dobson, 2003). It is more curious that the aggregation process of at least one such nondisease-related amyloid protein, HypF, can lead to formation of oligomeric structures and permeabilization of lipid membranes without forming amyloid fibrils (Relini et al., 2004). These oligomers are cytotoxic, as well. These data strongly suggest that it is the oligomer β-sheet conformation itself that leads to membrane insertion, channel formation, cellular dysfunction, and eventually, cytotoxicity. It has been proposed that these are latent properties of all polypeptide chains.
Human calcitonin is (hCt) is a 32-residue peptide hormone involved in the regulation of calcium and phosphorous metabolism. It is produced by C cells of the thyroid gland and is found in fibrillar amyloid deposits in patients with medullary carcinoma of the thyroid. The peptide, and fragments as small as four residues, can also form amyloid fibrils when incubated in aqueous solution. Calcitonin has pharmacological use in humans as a treatment for Paget disease and osteoporosis. The tendency to form fibrils limits its utility and has favored the use of salmon calcitonin (sCt), which is less fibrillogenic. Residues 16-21 of hCt form antiparallel β-sheet in methanol/water, and peptides consisting of residues 15-19 are highly amyloidogenic (Reches et al., 2002). Fibril formation is accompanied by a conformational charge from α helix to β-sheet in this central region or a change from random coil to β-sheet in the C-terminal region (Kawahara et al., 2000). Several calcitonins, including human, salmon, eel, and porcine, can from ion-permeable channels in lipid bilayer membranes (Stipani et al., 2001). These channels were long-lived, voltage-dependent, nonselective between cations and anions, permeable to Ca2+, and enhanced by the presence of negatively charged phospholipids. Although hCt acts through a human receptor, it is possible that the channel-forming activity of hCt could have other physiologic effects on signal transduction, particularly in light of its ability to allow Ca2+ across membranes. Further investigation is needed to investigate the potential physiological role of these channels.
Lysozyme is an enzyme possessing antimicrobial activity—it cleaves bonds in the outer wall of gram-positive bacteria. Lysozyme also forms toxic amyloid deposits in humans. Mutations or partial protein unfolding of lysozyme leads to aggregation, oligomer formation, and fibrillization. Partial unfolding of lysozyme also leads to the development of a nonenzymatic, broad-spectrum antimicrobial activity and a membrane-permeabilizing activity (Ibrahim et al., 2001). These activities were localized to a helix-loop-helix peptide at the upper lip of the active site cleft (87-114 of hen lysozyme). Similar peptides from human and chicken lysozyme possessed these activities, as well. These results suggest that the unfolding of lysozyme leads to fundamental shifts in protein function and activity. The aggregation of lysozyme monomers into oligomers appears to create a membrane-penetrating, antimicrobial complex which can aggregate into annular rings as seen by imaging (Malisauskas et al., 2003).
Anguiano M, Nowak RJ, Lansbury PT, Jr. 2002. Protofibrillar islet amyloid polypeptide permeabilizes synthetic vesicles by a pore-like mechanism that may be relevant to type II diabetes. Biochemistry 41(38):11338-11343. Abstract
Arispe N, Pollard HB, Rojas E. 1993a. Giant multilevel cation channels formed by Alzheimer disease amyloid-beta-protein [A beta P-(1-40)] in bilayer membranes. Proc Natl Acad Sci U S A 90(22):10573-10577. Abstract
Arispe N, Rojas E, Pollard HB. 1993b. Alzheimer disease amyloid-beta protein forms calcium channels in bilayer membranes: blockade by tromethamine and aluminum. Proc Natl Acad Sci U S A 90(2):567-571. Abstract
Arispe N, Doh M. 2002. Plasma membrane cholesterol controls the cytotoxicity of Alzheimer's disease AbetaP (1-40) and (1-42) peptides. FASEB J 16(12):1526-1536. Abstract
Arrasate, M., Mitra, S., Schweitzer, ES, Segal, MR, Finkbeiner, S. 2004. Inclusion body formation reduces levels of mutant huntingtin and the risk of neuronal death. Nature 431 (7010):747-8. Abstract
Azimova RK, Kagan BL. 2003. Ion channels formed by a fragment of alpha-synucleain (NAC) in lipd membranes. Biophs J 84(2):53a.
Bahadi R, Farrelly PV, Kenna BL, Kourie JI, Tagliavini F, Forloni G, Salmona M. Channels formed with a mutant prion protein PrP(82-146) homologous to a 7-kDa fragment in diseased brain of GSS patients. Am J Physiol Cell Physiol. 2003 Oct;285(4):C862-72. Epub 2003 Jun 18.
Baptista MJ, Cookson MR, Miller DW. 2004. Parkin and alpha-synuclein: opponent actions in the pathogenesis of Parkinson's disease. Neuroscientist 10(1):63-72. Abstract
Bhatia R, Lin H, Lal R. 2000. Fresh and globular amyloid-beta protein (1-42) induces rapid cellular degeneration: evidence for AbetaP channel-mediated cellular toxicity. FASEB J 14(9):1233-1243. Abstract
Brinckerhoff CE, Mitchell TI, Karmilowicz MJ, Kluve-Beckerman B, Benson MD. 1989. Autocrine induction of collagenase by serum amyloid A-like and beta 2-microglobulin-like proteins [see comments]. Science 243(4891):655-657. Abstract
Bucciarelli LG, Wendt T, Rong L, Lalla E, Hofmann MA, Goova MT, Taguchi A, Yan SF, Yan SD, Stern DM, Schmidt AM. 2002. RAGE is a multiligand receptor of the immunoglobulin superfamily: implications for homeostasis and chronic disease. Cell Mol Life Sci 59(7):1117-1128. Abstract
Bucciantini M, Calloni G, Chiti F, Formigli L, Nosi D, Dobson CM, Stefani M. 2004. Prefibrillar amyloid protein aggregates share common features of cytotoxicity. J Biol Chem 279(30):31374-31382. Abstract
Butler AE, Janson J, Soeller WC, Butler PC. 2003. Increased beta-cell apoptosis prevents adaptive increase in beta-cell mass in mouse model of type 2 diabetes: evidence for role of islet amyloid formation rather than direct action of amyloid. Diabetes 52(9):2304-2314. Abstract
Butterfield DA, Bush AI. 2004. Alzheimer's amyloid-beta-peptide (1-42): involvement of methionine residue 35 in the oxidative stress and neurotoxicity properties of this peptide. Neurobiol Aging 25(5):563-568. Abstract
Cepeda C, Hurst RS, Calvert CR, Hernandez-Echeagaray E, Nguyen OK, Jocoy E, Christian LJ, Ariano MA, Levine MS. 2003. Transient and progressive electrophysiological alterations in the corticostriatal pathway in a mouse model of Huntington's disease. J Neurosci 23(3):961-969. Abstract
Chen QS, Kagan BL, Hirakura Y, Xie CW. 2000. Impairment of hippocampal long-term potentiation by Alzheimer amyloid-beta-peptides. Journal of Neuroscience Research 60(1):65-72. Abstract
DeArmond SJ, Prusiner SB. 2003. Perspectives on prion biology, prion disease pathogenesis, and pharmacologic approaches to treatment. Clin Lab Med 23(1):1-41. Abstract
De Gioia L, Selvaggini C, Ghibaudi E, Diomede L, Bugiani O, Forloni G, Tagliavini F, Salmona M. 1994. Conformational polymorphism of the amyloidogenic and neurotoxic peptide homologous to residues 106-126 of the prion protein. Journal of Biological Chemistry 269(11):7859-7862. Abstract
Doh-Ura K, Iwaki T, Caughey B. 2000. Lysosomotropic agents and cysteine protease inhibitors inhibit scrapie-associated prion protein accumulation. J Virol 74(10):4894-4897. Abstract
Drüeke TB. 1998. Dialysis-related amyloidosis. Nephrol Dial Transplant 13 Suppl 1(2):58-64. Abstract
Fandrich M, Forge V, Buder K, Kittler M, Dobson CM, Diekmann S. 2003. Myoglobin forms amyloid fibrils by association of unfolded polypeptide segments. Proc Natl Acad Sci U S A 100(26):15463-15468. Abstract
Farrelly PV, Kenna BL, Laohachai KL, Bahadi R, Salmona M, Forloni G, Kourie JI. 2003. Quinacrine blocks PrP (106-126)-formed channels. J Neurosci Res 74(6):934-941. Abstract
Fernandez A, Berry RS. 2003. Proteins with H-bond packing defects are highly interactive with lipid bilayers: Implications for amyloidogenesis. Proc Natl Acad Sci U S A 100(5):2391-2396. Abstract
Forloni G, Chiesa R, Smiroldo S, Verga L, Salmona M, Tagliavini F, Angeretti N. 1993. Apoptosis mediated neurotoxicity induced by chronic application of beta amyloid fragment 25-35. Neuroreport 4(5):523-526. Abstract
Fraser SP, Suh YH, Djamgoz MB. 1997. Ionic effects of the Alzheimer's disease beta-amyloid precursor protein and its metabolic fragments. Trends Neurosci 20(2):67-72. Abstract
Furukawa K, Abe Y, Akaike N. 1994. Amyloid-beta protein-induced irreversible current in rat cortical neurones. Neuroreport 5(16):2016-2018. Abstract
Gasset M, Baldwin MA, Lloyd DH, Gabriel JM, Holtzman DM, Cohen F, Fletterick R, Prusiner SB. 1992. Predicted alpha-helical regions of the prion protein when synthesized as peptides form amyloid. Proc Natl Acad Sci U S A. 1992 Nov 15;89(22):10940-4.
Hardy J, Allsop D. 1991. Amyloid deposition as the central event in the aetiology of Alzheimer's disease. Trends Pharmacol Sci 12(10):383-388. Abstract
Harper JD, Lansbury PT, Jr. 1997. Models of amyloid seeding in Alzheimer's disease and scrapie: mechanistic truths and physiological consequences of the time-dependent solubility of amyloid proteins. Annu Rev Biochem 66(6):385-407. Abstract
Hegde RS, Mastrianni JA, Scott MR, DeFea KA, Tremblay P, Torchia M, DeArmond SJ, Prusiner SB, Lingappa VR. 1998. A transmembrane form of the prion protein in neurodegenerative disease. Science 279(5352):827-834. Abstract
Hirakura Y, Satoh Y, Hirashima N, Suzuki T, Kagan BL, Kirino Y. 1998. Membrane perturbation by the neurotoxic Alzheimer amyloid fragment beta 25-35 requires aggregation and beta-sheet formation. Biochem Mol Biol Int 46(4):787-794. Abstract
Hirakura Y, Lin MC, Kagan BL. 1999. Alzheimer amyloid abeta1-42 channels: effects of solvent, pH, and Congo Red. Journal of Neuroscience Research 57(4):458-466. Abstract
Hirakura Y, Yiu WW, Yamamoto A, Kagan BL. 2000a. Amyloid peptide channels: blockade by zinc and inhibition by Congo red (amyloid channel block). Amyloid 7(3):194-199. Abstract
Hirakura Y, Azimov R, Azimova R, Kagan BL. 2000b. Polyglutamine-Induced Ion Channels: A Possible Mechanism for the Neurotoxicity of Huntington and Other CAG Repeat Diseases. Journal of Neuroscience Research 60:490-494. Abstract
Hirakura Y, Kagan BL. 2001. Pore formation by beta-2-microglobulin: a mechanism for the pathogenesis of dialysis associated amyloidosis. Amyloid 8(2):94-100. Abstract
Hirakura Y, Carreras I, Sipe JD, Kagan BL. 2002. Channel formation by serum amyloid A: a potential mechanism for amyloid pathogenesis and host defense. Amyloid 9(1):13-23. Abstract
Ibrahim HR, Thomas U, Pellegrini A. 2001. A helix-loop-helix peptide at the upper lip of the active site cleft of lysozyme confers potent antimicrobial activity with membrane permeabilization action. J Biol Chem 276(47):43767-43774. Abstract
Ingrosso L, Ladogana A, Pocchiari M. 1995. Congo red prolongs the incubation period in scrapie-infected hamsters. J Virol 69(1):506-508. Abstract
Janson J, Ashley RH, Harrison D, McIntyre S, Butler PC. 1999. The mechanism of islet amyloid polypeptide toxicity is membrane disruption by intermediate-sized toxic amyloid particles. Diabetes 48(3):491-498. Abstract
Kagan BL. 1983. Mode of action of yeast killer toxins: channel formation in lipid bilayer membranes. Nature 302(5910):709-711. Abstract
Kagan BL, Selsted ME, Ganz T, Lehrer RI. 1990. Antimicrobial defensin peptides form voltage-dependent ion-permeable channels in planar lipid bilayer membranes. Proc Natl Acad Sci U S A. 1990 Jan;87(1):210-4. Abstract
Kagan BL, Hirakura Y, Azimov R, Azimova R. The channel hypothesis of Huntington's disease. Brain Res Bull. 2001 Oct-Nov 1;56(3-4):281-4. Review. Abstract
Kawahara, M., N. Arispe, Y. Kuroda, and E. Rojas. 1997. Alzheimer's disease amyloid-beta-protein forms Zn(2+)-sensitive, cation-selective channels across excised membrane patches from hypothalamic neurons. Biophys. J. 73: 67-75. Abstract
Kawahara M, Kuroda Y, Arispe N, Rojas E. 2000. Alzheimer's beta-amyloid, human islet amylin, and prion protein fragment evoke intracellular free calcium elevations by a common mechanism in a hypothalamic GnRH neuronal cell line. J Biol Chem 275(19):14077-14083. Abstract
Kayed R, Head E, Thompson JL, McIntire TM, Milton SC, Cotman CW, Glabe CG. 2003. Common structure of soluble amyloid oligomers implies common mechanism of pathogenesis. Science 300(5618):486-489. Abstract
Kazlauskaite J, Sanghera N, Sylvester I, Venien-Bryan C, Pinheiro TJ. 2003. Structural changes of the prion protein in lipid membranes leading to aggregation and fibrillization. Biochemistry 42(11):3295-3304. Abstract
Kim HS, Lee JH, Lee JP, Kim EM, Chang KA, Park CH, Jeong SJ, Wittendorp MC, Seo JH, Choi SH, Suh YH. 2002. Amyloid-beta peptide induces cytochrome C release from isolated mitochondria. Neuroreport 13(15):1989-1993. Abstract
Korth C, May BC, Cohen FE, Prusiner SB. 2001. Acridine and phenothiazine derivatives as pharmacotherapeutics for prion disease. Proc Natl Acad Sci U S A 98(17):9836-9841. Abstract
Kourie JI. Characterization of a C-type natriuretic peptide (CNP-39)-formed cation-selective channel from platypus (Ornithorhynchus anatinus) venom. J Physiol. 1999 Jul 15;518 ( Pt 2):359-69.
Kourie JI, Culverson A. 2000. Prion peptide fragment PrP[106-126] forms distinct cation channel types. Journal of Neuroscience Research 62(1):120-133. Abstract
Kourie JI, Hanna EA, Henry CL (2001a): Properties and modulation of alpha human atrial natriuretic peptide (alpha-hANP)-formed ion channels. Can J Physiol Pharmacol. 2001 Aug;79(8):654-64.
Kourie JI, Henry CL, Farrelly P. 2001b. Diversity of amyloid-beta protein fragment [1-40]-formed channels. Cell Mol Neurobiol. 2001 Jun;21(3):255-84.
Kourie JI, Culverson AL, Farrelly PV, Henry CL, Laohachai KN. 2002. Heterogeneous amyloid-formed ion channels as a common cytotoxic mechanism: implications for therapeutic strategies against amyloidosis. Cell Biochem Biophys 36(2-3):191-207. Abstract
Lashuel HA, Petre BM, Wall J, Simon M, Nowak RJ, Walz T, Lansbury PT, Jr. 2002a. Alpha-synuclein, especially the Parkinson's disease-associated mutants, forms pore-like annular and tubular protofibrils. J Mol Biol 322(5):1089-1102. Abstract
Lashuel HA, Hartley D, Petre BM, Walz T, Lansbury PT, Jr. 2002b. Neurodegenerative disease: amyloid pores from pathogenic mutations. Nature 418(6895):291. Abstract
Lashuel HA, Hartley DM, Petre BM, Wall JS, Simon MN, Walz T, Lansbury PT, Jr. 2003. Mixtures of wild-type and a pathogenic (E22G) form of Abeta40 in vitro accumulate protofibrils, including amyloid pores. J Mol Biol 332(4):795-808. Abstract
Li SH, Li XJ. 2004. Huntingtin-protein interactions and the pathogenesis of Huntington's disease. Trends Genet 20(3):146-154. Abstract
Lin H, Zhu YJ, Lal R. 1999. Amyloid-beta protein (1-40) forms calcium-permeable, Zn2+-sensitive channel in reconstituted lipid vesicles. Biochemistry 38(34):11189-11196. Abstract
Lin MC. 1996. Channel formation by amyloldogenic neurotoxic and neurodegenerative disease related peptides. Ph.D. dissertation, Division of Neuroscience, UCLA.
Lin MC, Mirzabekov T, Kagan BL. 1997. Channel formation by a neurotoxic prion protein fragment. Journal of Biological Chemistry 272(1):44-47. Abstract
Lin MC, Kagan BL. 2002. Electrophysiologic properties of channels induced by Abeta25-35 in planar lipid bilayers. Peptides 23(7):1215-1228. Abstract
Lorenzo A, Razzaboni B, Weir GC, Yankner BA. 1994. Pancreatic islet cell toxicity of amylin associated with type-2 diabetes mellitus. Nature 368(6473):756-760. Abstract
Malisauskas M, Zamotin V, Jass J, Noppe W, Dobson CM, Morozova-Roche LA. 2003. Amyloid protofilaments from the calcium-binding protein equine lysozyme: formation of ring and linear structures depends on pH and metal ion concentration. J Mol Biol 330(4):879-890. Abstract
Manunta M, Kunz B, Sandmeier E, Christen P, Schindler H. 2000. Reported channel formation by prion protein fragment 106-126 in planar lipid bilayers cannot be reproduced [letter]. FEBS Lett 474(2-3):255-256. Abstract
McCarthy RE, 3rd, Kasper EK. 1998. A review of the amyloidoses that infiltrate the heart. Clin Cardiol 21(8):547-552. Abstract
McKeith IG, Mosimann UP. 2004. Dementia with Lewy bodies and Parkinson's disease. Parkinsonism Relat Disord 10 Suppl 1:S15-18. Abstract
McLaurin J, Franklin T, Fraser PE, Chakrabartty A. 1998. Structural transitions associated with the interaction of Alzheimer beta-amyloid peptides with gangliosides. J Biol Chem 273(8):4506-4515. Abstract
McLean LR, Balasubramaniam A. 1992. Promotion of beta-structure by interaction of diabetes associated polypeptide (amylin) with phosphatidylcholine. Biochim Biophys Acta 1122(3):317-320. Abstract
Merlini G, Bellotti V. 2003. Molecular mechanisms of amyloidosis. N Engl J Med 349(6):583-596. Abstract
Merlini G, Westermark P. 2004. The systemic amyloidoses: clearer understanding of the molecular mechanisms offers hope for more effective therapies. J Intern Med 255(2):159-178. Abstract
Mirzabekov T, Lin MC, Yuan WL, Marshall PJ, Carman M, Tomaselli K, Lieberburg I, Kagan BL. 1994. Channel formation in planar lipid bilayers by a neurotoxic fragment of the beta-amyloid peptide. Biochemical and Biophysical Research Communications 202(2):1142-1148. Abstract
Mirzabekov TA, Lin MC, Kagan BL. 1996. Pore formation by the cytotoxic islet amyloid peptide amylin. Journal of Biological Chemistry 271(4):1988-1992. Abstract
Moe SM, Sprague SM. 1992. Beta 2-microglobulin induces calcium efflux from cultured neonatal mouse calvariae. Am J Physiol 263(3 Pt 2):F540-545. Abstract
Monoi H, Futaki S, Kugimiya S, Minakata H, Yoshihara K. 2000. Poly-L-glutamine forms cation channels: relevance to the pathogenesis of the polyglutamine diseases. Biophysical Journal 78(6):2892-2899. Abstract
Nakajima M, Yamada T, Kusuhara T, Furukawa H, Takahashi M, Yamauchi A, Kataoka Y. 2004. Results of quinacrine administration to patients with Creutzfeldt-Jakob disease. Dement Geriatr Cogn Disord 17(3):158-163. Abstract
Pan KM, Baldwin M, Nguyen J, Gasset M, Serban A, Groth D, Mehlhorn I, Huang Z, Fletterick RJ, Cohen FE, et al. 1993. Conversion of alpha-helices into beta-sheets features in the formation of the scrapie prion proteins. Proc Natl Acad Sci U S A 90(23):10962-10966. Abstract
Panov AV, Gutekunst CA, Leavitt BR, Hayden MR, Burke JR, Strittmatter WJ, Greenamyre JT. 2002. Early mitochondrial calcium defects in Huntington's disease are a direct effect of polyglutamines. Nat Neurosci 5(8):731-736. Abstract
Pepys MB, Hawkins PN, Booth DR, Vigushin DM, Tennent GA, Soutar AK, Totty N, Nguyen O, Blake CC, Terry CJ, et al. 1993. Human lysozyme gene mutations cause hereditary systemic amyloidosis. Nature 362(6420):553-557. Abstract
Petersen J, Kang MS. 1994. In vivo effect of beta 2-microglobulin on bone resorption. Am J Kidney Dis 23(5):726-730. Abstract
Petosa C, Collier RJ, Klimpel KR, Leppla SH, Liddington RC. 1997. Crystal structure of the anthrax toxin protective antigen. Nature 385(6619):833-838. Abstract
Pike CJ, Burdick D, Walencewicz AJ, Glabe CG, Cotman CW. 1993. Neurodegeneration induced by beta-amyloid peptides in vitro: the role of peptide assembly state. J Neurosci 13(4):1676-1687. Abstract
Qi JS, Qiao JT. 2001. Amyloid-beta-protein fragment 31-35 forms ion channels in membrane patches excised from rat hippocampal neurons. Neuroscience 105(4):845-852. Abstract
Reches M, Porat Y, Gazit E. 2002. Amyloid fibril formation by pentapeptide and tetrapeptide fragments of human calcitonin. J Biol Chem 277(38):35475-35480. Abstract
Relini A, Torrassa S, Rolandi R, Gliozzi A, Rosano C, Canale C, Bolognesi M, Plakoutsi G, Bucciantini M, Chiti F, Stefani M. 2004. Monitoring the process of HypF fibrillization and liposome permeabilization by protofibrils. J Mol Biol 338(5):943-957. Abstract
Rodrigues CM, Sola S, Brites D. 2002. Bilirubin induces apoptosis via the mitochondrial pathway in developing rat brain neurons. Hepatology 35(5):1186-1195. Abstract
Samaia HB, Mari JJ, Vallada HP, Moura RP, Simpson AJ, Brentani RR. 1997. A prion-linked psychiatric disorder. Nature 390(6657):241. Abstract
Sandberg MK, Wallen P, Wikstrom MA, Kristensson K. 2004. Scrapie-infected GT1-1 cells show impaired function of voltage-gated N-type calcium channels (Ca(v) 2.2) which is ameliorated by quinacrine treatment. Neurobiol Dis 15(1):143-151. Abstract
Sanderson KL, Butler L, Ingram VM. 1997. Aggregates of a beta-amyloid peptide are required to induce calcium currents in neuron-like human teratocarcinoma cells: relation to Alzheimer's disease. Brain Research 744(1):7-14. Abstract
Sipe JD, Cohen AS. 2000. Review: history of the amyloid fibril. J Struct Biol 130(2-3):88-98. Abstract
Schein SJ, Kagan BL, Finkelstein A. 1978. Colicin K acts by forming voltage-dependent channels in phospholipid bilayer membranes. Nature 276(5684):159-163. Abstract
Schubert D, Behl C, Lesley R, Brack A, Dargusch R, Sagara Y, Kimura H. 1995. Amyloid peptides are toxic via a common oxidative mechanism. Proc Natl Acad Sci U S A 92(6):1989-1993. Abstract
Simmons, M. A., & Schneider, C. R. (1993). Amyloid-beta peptides act directly on single neurons. Neuroscience Letters, 150(2), 133-136. Abstract
Sipe JD. 2000. Serum amyloid A: from fibril to function. Current status. Amyloid 7(1):10-12. Abstract
Sokolov Y, Mirzabekov T, Martin DW, Lehrer RI, Kagan BL. 1999. Membrane channel formation by antimicrobial protegrins. Biochimica et Biophysica Acta 1420(1-2):23-29. Abstract
Song L, Hobaugh MR, Shustak C, Cheley S, Bayley H, Gouaux JE. 1996. Structure of staphylococcal alpha-hemolysin, a heptameric transmembrane pore. Science 274(5294):1859-1866. Abstract
Stefani M, Dobson CM. 2003. Protein aggregation and aggregate toxicity: new insights into protein folding, misfolding diseases and biological evolution. J Mol Med. 2003 Nov;81(11):678-99. Epub 2003 Aug 27. Review. Abstract
Stipani V, Gallucci E, Micelli S, Picciarelli V, Benz R. 2001. Channel formation by salmon and human calcitonin in black lipid membranes. Biophys J 81(6):3332-3338. Abstract
Volles MJ, Lansbury PT, Jr. 2002. Vesicle permeabilization by protofibrillar alpha-synuclein is sensitive to Parkinson's disease-linked mutations and occurs by a pore-like mechanism. Biochemistry 41(14):4595-4602. Abstract
Walsh DM, Klyubin I, Fadeeva JV, Cullen WK, Anwyl R, Wolfe MS, Rowan MJ, Selkoe DJ. 2002. Naturally secreted oligomers of amyloid-beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature 416(6880):483-484. Abstract
Wang L, Lashuel HA, Walz T, Colon W. 2002. Murine apolipoprotein serum amyloid A in solution forms a hexamer containing a central channel. Proc Natl Acad Sci U S A 99(25):15947-15952. Abstract
Weiss JH, Pike CJ, Cotman CW. 1994. Ca2+ channel blockers attenuate beta-amyloid peptide toxicity to cortical neurons in culture. J Neurochem 62(1):372-375. Abstract
Westermark P, Wilander E. 1978. The influence of amyloid deposits on the islet volume in maturity onset diabetes mellitus. Diabetologia 15(5):417-421. Abstract
Yang AJ, Knauer M, Burdick DA, Glabe C. 1995. Intracellular A beta 1-42 aggregates stimulate the accumulation of stable, insoluble amyloidogenic fragments of the amyloid precursor protein in transfected cells. Journal of Biological Chemistry 270(24):14786-14792. Abstract
Yang W, Dunlap JR, Andrews RB, Wetzel R. 2002. Aggregated polyglutamine peptides delivered to nuclei are toxic to mammalian cells. Hum Mol Genet 11(23):2905-2917. Abstract
Yankner BA, Duffy LK, Kirschner DA. 1990. Neurotrophic and neurotoxic effects of amyloid-beta protein: reversal by tachykinin neuropeptides. Science 250(4978):279-282. Abstract
Zhu M, Souillac PO, Ionescu-Zanetti C, Carter SA, Fink AL. 2002. Surface-catalyzed amyloid fibril formation. J Biol Chem 277(52):50914-50922. Abstract
Zhu M, Han S, Zhou F, Carter SA, Fink AL. 2004. Annular oligomeric amyloid intermediates observed by in situ atomic force microscopy. J Biol Chem 279(23):24452-24459. Abstract
Zhu YJ, Lin H, Lal R. 2000. Fresh and nonfibrillar amyloid-beta protein(1-40) induces rapid cellular degeneration in aged human fibroblasts: evidence for AbetaP-channel-mediated cellular toxicity. FASEB J 14(9):1244-1254. Abstract
Live Discussion led by Bruce Kagan on 2 May 2005.
Participants: Tom Fagan, Alzforum; Bruce Kagan, UCLA Neuropsychiatric Institute; David Corbin, New York; Wonmuk Hwang, Texas A&M University; Nico Stanculescu, Alzforum; Charlie Glabe University of California, Irvine; Larry Nault, Matrix Integrated Linkage LLC; Ratnesh Lal, University of California at Santa Barbara; Dave Teplow, University of California, Los Angeles; Ben Albensi University of Manitoba; Rose Xiao, Memory Pharmaceutical; Yuri Sokolov, University of California, Irvine; Ryan Lee, W3C/MIT CSAIL.
Note: The transcript has been edited for clarity and accuracy.
Hi, Bruce! Hi, Charlie!
Okay, if you didn't guess by the microphone icon, I'm Tom Fagan and I'll be moderating today when necessary. Bruce is going to start us off with a brief introduction and then let the fun begin. Take it away, Bruce!
In 1993, Arispe et al.,1993 reported that Aβ formed ion channels, and proposed that channels could cause cellular pathology and toxicity. Abundant evidence now supports this view, including the facts that virtually all amyloids form channels and that Aβ can inhibit LTP, or long-term potentiation (see ARF related news story), depolarize neurons, allow cytochrome c efflux from mitochondria (see Kim et al., 2002), and kill cells. Channel formation readily explains most amyloid pathophysiology, including memory disturbance, calcium dysfunction, membrane depolarizations, increased reactive oxygen species (ROS), sensitivity to toxins, and apoptosis.
Bruce, do these channels have a hypothetical geometry. A pipe or a duct? Bounded by what?
Larry, Arispe and Guy proposed several possible models for the pore. Recently, Arispe has published evidence that specific regions of the peptide seem to line the pore (see Arispe, 2004).
Bruce, how many channels does it take to kill a cell?
Tom, it depends on the size of the cell. The Aβ channels are large, electrically, and would create a significant leak. A single pore could reduce the sodium (Na+) concentration by about 10 micromolar per second.
Bruce, how does that compare to normal leakage?
Tom, it again depends on the cell. For a neuron, which must maintain a tight membrane for signaling, a single pore would pose a significant but not lethal leak.
How specific are these channels? I understand that they have a distribution of sizes, which likely cause them to be nonspecific.
Won, the channels have been described as permeable to a variety of cations including sodium and calcium. There is little selectivity amongst cations.
This fits in with what Charlie reported recently, right Charlie?
We see fluorescent dyes that are about 600 Daltons (Da) leaking; it is not ion-specific (see ARF related news story).
Charlie, 600 Da is pretty big; do you mean such molecules are getting through these pores?
Yes, going both ways; I'm not sure of the mechanism. Large proteins like lactate dehydrogenase (LDH) do not leak detectably.
Charlie, we see non-electrolyte leakage of sugars with a diameter up to about 10-12 angstroms. Of course, we should temper this with the observation that we see many species of channels and they may all have different sizes and selectivities.
Bruce, so are these channels in a state of flux? Can their size change dynamically with time, and might there be large and small coexisting in the same cell?
Tom, Charlie sees a very different result from ours, but Joseph Kourie has described at least three different Aβ channels with varying kinetics, selectivity lifetime, and so on (see Kourie et al., 2002).
Everyone, any ideas on how these channels can be stoppered? Would this be a therapeutic strategy?
We haven't seen any known channel blockers that inhibit the permeabilization; but that doesn't mean that there aren't any.
Tom, the only blockers we have so far are very nonspecific, such as zinc (Zn2+) or tromethamine ions (e.g., Tris buffer). The development of more specific blockers could prove useful as a therapeutic strategy. Zn2+ does block the toxicity of Aβ on fibroblasts (see Zhu et al., 2000).
About Charlie's reply: How about the oligomer-specific antibodies that you published? Don't they block permeabilization?
Won, not known, but that is a reasonable hypothesis. We find that you need to start with oligomers first. I'm not sure that they insert into the hydrocarbon region in a traditional fashion, either.
Calcium uptake has been blocked specifically by anti-amyloid antibody and zinc in both liposomes and cells (Rhee et al., 1998; Lin et al., 1999; Lin et al., 2001; Bhatia et al., 2000).
So how does the channel hypothesis fit with the slow progression of the disease? Given the degree of leakage, it would seem that channels would be pretty lethal and that the disease should progress more rapidly, or am I being too simplistic?
Maybe a small leak just contributes to chronic stress as the cell has to pump more to keep up with the leak.
Tom, Dennis Selkoe's group has shown that oligomers can impair memory in a reversible fashion. Since we know Aβ channels can inhibit long-term potentiation (LTP), this might be the first step in the pathology (see ARF related news story and ARF news story; also Wang et al., 2002).
Bruce, but in the LTP experiments, basal transmission is okay, which seems odd, given the degree of leakage you could get. I wonder if there is something modulating the channels?
Tom, the LTP experiments require very small doses (less than toxic) of Aβ. I suspect that they are altering membrane potential in a subtle way to inhibit LTP, but not enough to derail normal transmission.
About channel formation: Is it known whether they are formed in solution, then incorporated into the cell membrane, or is it more likely that channels are formed through interaction with the membrane or the substrate (in case of in vitro experiments)?
Evidence is published suggesting that both preformed oligomers can insert in the membrane or monomers can insert and oligomerize to form ion channels.
Won, there is clear evidence that the presence of lipids or membranes affects the folding of Aβ and other amyloids, tending to promote β-sheet and oligomerization (see, e.g., ARF related news story on lipid rafts and ARF news story on the effects of lipid anchors on prion amyloid formation).
Bruce, that is consistent with the finding that fibrils form in partially denaturing conditions—possibly the hydrophobic environment of the lipid tails enhances the condition for fibril formation.
We find that monomers are pretty inert. It isn't necessarily easy to make pure monomers.
Charlie, we find that monomers are inert also, but that with time in aqueous solution, oligomers form and insert into the membrane.
As mentioned above, there is published work showing that monomers can induce cell toxicity as well as calcium uptake in both cells and liposomes.
Ratnesh, how were they sure they were monomers? They may start out that way, but things change pretty rapidly even in the absence of membranes.
Ratnesh, how can you be certain that the peptide remains as a monomer throughout the experiment?
I don't think they remain monomers throughout the experiments. Instead, they do oligomerize. We have imaged those monomers in real time using atomic force microscopy (AFM). We have shown that for an extended period of time, in physiological buffer and in physiologically relevant concentration, they do not form oligomers. Nevertheless, as I mentioned above, monomers will oligomerize once in the association with lipids in the bilayer/liposomes.
Ratnesh, so that would be consistent with all the recent work suggesting toxicity of oligomers?
Yes, that is all we have published in the last 7 years.
Ratnesh, I think that Aβ peptides are produced throughout the lifespan.
Membrane-induced conversion from inert to an active form will be consistent with the expected conformational change in many amyloids associated with misfolding diseases. Won, I agree with you and have made that point in our earlier publications. Aβ peptides are produced throughout the lifespan and in many, if not all cell types.
All, so is it just the induction of LTP that is affected or later stages, as well?
Ben, I believe it is just the induction, but I could be mistaken. Dave, do you know?
All, how many monomers does it take to form the simplest channel—in theory even, if there is no experimental data?
Tom, a model presented by Durrel et al.,1994 suggests that it can be anywhere from tetramer to octamer and more (perhaps up to 12).
Ratnesh, is a tetramer big enough to form a pore and span the membrane?
Tom, Aβ42 tends to form hexamers, and those would be excellent candidates for the smallest channels.
It isn't clear that it has to span the bilayer in the traditional sense. How do you explain the fact that polyQ permeabilizes membranes in the same fashion?
Charlie, I don't know. Any ideas?
Maybe the same reason that highly charged, polar peptides like HIV tat and Antennapedia can cross the bilayer with large passenger molecules attached to them? (See review on cell penetrating peptides).
Charlie, are you suggesting we abandon "the channel" hypothesis and call it something else?
I don't know; this is something we are still trying to sort out.
There is a model by Mobley et al. suggesting that monomers may or may not span the whole membrane. If they do not, it can still form an ion channel (like porins or others); otherwise, one would need an additional four monomers from the other membrane leaflet (and in that case it would be eight monomers total), or six or more, depending upon whether they form tetrameric, pentameric, hexameric, or higher order oligomeric channels.
Charlie, the polyQ story is striking in that there is a minimum length for pathology, yet this minimum varies from illness to illness. For Aβ peptides, we have shown that the minimum length to obtain channels is nine residues, which would span the hydrophobic core of the bilayer in β-sheet form.
How about the pure dimensional argument? The hydrophobic region of a lipid bilayer is about 3 nm, and if the channel diameter is about 2 nm, the oligomer channel must have about 36 nm^2. Given the size of Aβ, we can guess how many monomers are needed to span a bilayer like a channel. But as suggested above, even if the oligomers occupy only one leaflet, it is still possible for there to be a hole, through fluctuation of the lipids. In this case, the hole will be highly dynamic rather than statically open.
Won, the pore diameter of 2 nm is quite big for any channel. The outer diameter of the oligomeric channel diameter would be consistent with what we had published earlier (see Lin et al., 2001). The inner (pore) diameter will be considerably smaller (~1-1.5 nanometer).
Ratnesh, I was referring to Hilal, Lansbury and colleagues' work on torroids—their electron microscopy (EM) images suggest that the inner diameter is about 2 to 3 nm (see, e.g., Rochet et al., 2004).
Won, Lashuel's work is done on annular pores; peptides were never associated with any membrane (either before or after). For their actual channel conformation, they refer to our work (Lin et al., 2001).
If peptides that do span the bilayers are extended β structures, you would expect that they would have to form H bonded β barrels, which typically have ~20 strands.
Charlie's point is well-taken. This would predict very large channels indeed.
Caution to all—there is a distinct possibility that Hilal's work, while relevant in in vitro systems, is not representative of what occurs in vivo.
Charlie, I was just going to ask about molecular models that would have a pore. If the pore idea is correct, then wouldn't it be possible to estimate how many monomers would be needed, and how would that fit with current ideas on trimers, dodecamers, etc., that seem most toxic?
Uli Aebi tried to visualize pores or channels by AFM using preparations and conditions that caused dye leakage, and he did not observe "donuts." Instead, he saw defects in the membrane radiating from the oligomers (see Green et al., 2004).
Charlie, there is a serious difference between Aebi's work and other channel work. In his work, amyloid was added to membranes that were preformed and adsorbed on a substrate, a common practice for AFM work. On the contrary, when you reconstitute in liposome or bilayer and then image with AFM, you do see channels (Lin et al., 2001).
In relation to Tom's question, what is the shortest length of a model peptide tried that forms channels?
There is a report of an Aβ31-35 fragment forming channels. I don't know what to make of this unless there is one peptide in each leaflet of the bilayer (see also Le and Qiao on Aβ31-35 effects on LTP).
Charlie, Bruce, have either of you examined the effect of different lipids on leakage? Sorry, I'm not remembering the data off the top of my head.
Yes, not much specificity. Yuri can jump in on this if I have misspoken.
Tom, cholesterol inhibits. Negatively charged lipids are required for activity. Not much other specificity.
There are some publications from the Arispe group on the effect of cholesterol on amyloid insertion and channel activity (see Arispe and Doh, 2002).
Charlie and Yuri, does membrane fluidity affect your results?
Bruce, we did not study the effect of membrane fluidity. But recently we found very strong lipid-dependence of amyloid-induced conductance in bilayers.
So what about other channels besides Aβ ones? What have we learned from those? Do they clarify anything or just make things more confusing?
I am struck by how similar the channels are that we see from a variety of unrelated peptides. Their physiologic properties are nearly identical. They are all "leakage" channels.
I second that. All amyloids are pretty similar.
Must close out. Interesting hypothesis. Clearance of building channel blocks may be the key to defeat of progression. Building channels over time and diving then through the bilayer is a process that might be interrupted with agents like zinc or by clearing essential oligomer building blocks.
There is published work on many other amyloid proteins showing channel-like activity (Kagan's lab, Arispe's lab, Kourie's lab) and we have structural data from AFM study to confirm channel-like structures.
Ratnesh, what do you mean by "channel-like."
Tom, as the AFM work is purely structural and we do not do electrical recording simultaneously (I mean not on the same structures because we don't have an appropriate recording system), we take a conservative view that it is channel-like. However, in parallel studies on the same batch of specimen, electrical recording shows channel activity.
Tom, in the case of our work and that of Arispe and Kourie, these are clearly channels by all the standard criteria. Charlie's work clearly shows permeability without "channels."
Ratnesh, I thought you were referring to a channel-like structure rather than activity.
Tom, we do both: image the three-dimensional structure and in parallel do electrical recording. We are hoping to do both together someday soon.
Great, Ratnesh; keep us posted!
Thanks, Tom. We will be in touch.
Tom, one thing we have learned from complement and toxin studies is that "holes" that one sees in EM do not always correspond to the electrical pore pathways.
All, we are nearing the end of our hour, but we can keep chatting as long as we like. But before folks start rushing off, I just want to thank you all for coming, and Bruce for agreeing to host this chat.
Is there any evidence that the channel-like structure must have torroidal morphology? How about spheroids? And thanks to all for useful conversations.
So before we go, what are the crucial experiments that need to be done to advance our understanding of what is going on? Anyone, everyone?
I think there are two crucial experiments to be done. First, specific channel blockers need to be developed and tested in cell and animal models. Second, channels (or permeability changes) need to be found in animal models of disease.
Bruce, are there any meta studies that could be done examining patients taking channel blockers for other diseases that would answer your question?
Dave, that's a great idea! However, for the easiest ones—the calcium channel blockers—we have already tried them without success.
Bruce, tried them in vitro?
Tom, right, we tried them in vitro against our Aβ and other channels.
All, we talked about channel blockers, but are there other strategies to stop the leakage of these "channels"? Also, are there synthetic channels that could be made to mimic the natural ones, and how might that be useful?
Antibodies do block, and one can use peptides to block, as well (a practice useful in blocking gap junction ion channel activity).
Tom, yes, one could devise treatments that made the bilayer less susceptible to channel insertion. Arispe did this recently with robust effects on toxicity. He stiffened the cell membrane with cholesterol (see Arispe and Doh, 2002).
Jim Hall and I found that an increase of bilayer thickness inhibits the effect of amyloid-β in bilayers.
Yuri, what's the latest on what ions and other molecules can go through your Aβ-treated membranes?
Bruce, looks like everything we tested. HEPES, Tris, TEA....
Yuri, you do this by using longer chain lipids?
Tom, looks like with longer chains we have only preliminary data. But definitely the effect of amyloid-β is completely blocked by the saturation of bilayer with decane. We are not going to use decane as a treatment, but it is good enough to study the mechanism.
Yuri, yes, for sure, anything that sheds some light on what's going on....
Tom, there are other ways to increase membrane thickness, such as changes in dietary lipids.
Bruce, that's where I was headed. But you said that you tried different lipids and saw no specific effects, right?
Tom, yes we did try many lipids and didn't see major effects except of charge and stiffness, but we did not try to alter membrane thickness (one needs lipids with long tails). We should go back and do that. Good idea!
Or you could use shorter ones to see if the problem is any worse.
Right. I know these manipulations to change lipid composition can be done in animals, but I don't know if they have ever been tried in humans. An interesting prospect! It is interesting to note that increased membrane thickness would inhibit standard channels as well as Yuri's permeability increase.
Bruce, have you also considered the systemic effects in Alzheimer's of impaired calcium transport by fibroblasts (see Peterson et al., 1985)?
David, I'm not familiar with this paper. Do you think it might be evidence of Aβ channels in the fibroblasts?
At least in in vitro studies, we have shown Aβ-channel-mediated toxicity in fibroblasts (Zhu et al., 2000).
Bruce, I do not know. I was just looking at some of the early research on cell membrane abnormality in Alzheimer's to see what others have observed, such as decreased microviscosity.
David, I'm going to look up that paper. Thanks for bringing it to my attention. Goodbye to all.
Okay, I think we should probably wrap up. Thanks, Bruce, most interesting. Thanks to everyone for joining in; I hope it was useful.
Thanks, Bruce, and thanks, Tom!
Nico and Tom, thanks so much. I enjoyed this immensely!
Splendid! We will be forwarding the transcript for review!