20 February 2003. In today’s Neuron, scientists at Brigham and Women’s Hospital in Boston, Massachusetts, report that they have devised a biochemical strategy to detect the earliest aggregates of the protein α-synuclein in cultured neurons, mouse brain, and human brain. Ronit Sharon, Dennis Selkoe, and colleagues, discovered that first extracting lipids from the preparation was the trick that made possible the visualization of small, oligomeric forms of this protein. Such oligomers are still soluble, but are considered the precursors to protofibrils and then to large, insoluble deposits. A growing number of scientists believe oligomers may play a role in the early pathogenesis of several neurodegenerative diseases-α-synuclein in Parkinson’s and Lewy body dementia, tau in Alzheimer’s, huntingtin in Huntington’s, among others. Other scientists, still unconvinced, are saying "Show me the goods!" because such oligomers have been difficult to pin down in biological and disease material.
Beyond demonstrating the presence of α-synuclein "oligos" in vivo, Sharon and coworkers report that fatty acids control their formation. These data strengthen an evolving line of thought, namely, that interactions of the natively unfolded α-synuclein with lipids, either in membranes or in the cytoplasm, induce their folding, and that age-related changes in the cell’s lipid makeup could favor the formation of pathogenic aggregates. This could become part of the explanation for why people without strong predisposing mutations develop these diseases.
Much prior work on oligomeric assemblies of a protein implicated in neurodegeneration has been done on Aβ (Lambert et al., 1998; Walsh et al., 2002; Dahlgen et al., 2002). Yet this study focused on α-synuclein, a 17 kD soluble protein in the cytosol that, when mutated, causes rare forms of Parkinson’s and is the major constituent of Lewy bodies, the pathologic hallmark of Parkinson’s. Drosophila and mouse models of α-synuclein recapitulate some aspects of PD-like neurodegeneration, but its normal physiological role is still unclear.
The present study builds on work by, among others, Julia George at the University of Illinois in Urbana, who has shown that soluble α-synuclein associates with membranes and phospholipid vesicles in ways that alter its folding and that change with the addition of fatty acids (Perrin et al., 2001). In their own prior work, Selkoe’s team had noted that α-synuclein had homologies with certain fatty acid-binding proteins (FABP) which transport fatty acids across membranes and have received attention in studies of diabetes and other disorders of lipid metabolism. They spotted a pool of cytosolic α-synuclein with a molecular weight above that of its monomer, but it was tied up with lipids (Sharon et al., 2001). How to characterize this pool, and why were the lipids in the mix?
This is where today’s paper comes in. The study includes too many experiments to summarize here, but briefly, the researchers first extracted lipids from whole mouse brains. This simple step revealed previously unidentifiable α-synuclein immunoreactive bands in a ladder pattern that follows the step-wise increase in molecular weight from monomers to hexamers. To rule out that the oligos might have formed as an artifact of the detection procedure, Sharon et al. used size exclusion chromatography to separate out components of the relevant centrifugation fraction from homogenates of mouse brain and cultured mesencephalic neuronal cells that were prepared in physiological, nondenaturing buffer. They saw dimers and larger oligomers in this preparation, too.
Then, Sharon et al. added different fatty acids to cultured mesencephalic neurons and discovered that oligomer formation shot up in the presence of polyunsaturated fatty acids (PUFAs), such as α-linolenic acid, but dropped in the presence of saturated ones, such as stearic acid. The degree of oligomer formation correlated with the number of unsaturated double bonds. Further characterizing this process, Sharon and colleagues showed that the fatty acid-triggered oligomerization changed over time in cultured cells and transgenic mice in ways suggesting that oligomers which are initially soluble later convert to insoluble forms. They also showed some data suggesting that fatty acids act on α-synuclein directly.
Moreover, Sharon et al. show evidence suggesting that oligomers accumulate faster in cells stably expressing a human mutant α-synuclein gene, and that they accumulate in aging mice. Finally, they checked cerebral cortex samples of five brains each of Parkinson’s, Lewy body dementia, and aging controls. They found more dimers and larger oligomers in the disease brains than in control brains.
The authors speculate that monomeric α-synuclein interacts in the cytoplasm with fatty acids as part of its normal function. In the process, they propose, longer and more unsaturated fatty acids mediate the formation of soluble oligomers. Small amounts of those may be normal. However, time or prolonged exposure to such fatty acids may lead to their aggregation into insoluble assemblies, perhaps as the neuron’s protective response to avoid the accumulation of oligos to toxic levels. In support of this hypothesis, the authors note unpublished data suggesting that the accumulation of α-synuclein oligos in diseased brains correlates with the level of cytoplasmic PUFAs.
A separate study by Sreeganga Chandra, Thomas Sudhof, and others at University of Texas Southwestern Medical Center in Dallas, is currently available in manuscript form in the online Journal of Biological Chemistry. In it, the scientists analyze the folding process that ensues when initially unstructured α-synuclein interacts with phospholipid membranes. They describe the resulting conformation as consisting of two α-helical regions interrupted by a break, and point out structural peculiarities that distinguish α-synuclein from its β- and γ cousins. The authors suggest that this folded form of α-synuclein might function to coat synaptic vesicles.-Gabrielle Strobel.
Sharon R, Bar-Joseph I, Frosch MP, Walsh DM, Hamilton JA, Selkoe DJ. The formation of highly soluble oligomers of alpha-synuclein is regulated by fatty acids and enhanced in Parkinson's disease. Neuron 2003 February 20; (37):583-595. Abstract
Chandra S, Chen X, Rizo J, Jahn R, Sudhof TC. A broken alpha -helix in folded alpha -synuclein.J Biol Chem 2003 Feb 13. Abstract
Q&A with Dennis Selkoe-Posted 20 February 2003.
Q: To you, are these data the "smoking gun" that oligomer proponents have been waiting for?
A: Not necessarily, depending on how one defines "smoking gun." While we believe we provide compelling evidence that entirely soluble oligomers of α-synuclein exist in intact neuronal cells and in the normal brain, we have not yet shown that such lipid-dependent soluble oligomers are directly toxic to the neurons that contain them. Indeed, our data suggest that certain amounts of such oligomers occur normally, and that they may gradually transition to higher MW assemblies that can become insoluble. So, our paper does not provide a "smoking gun" as regards oligomer neurotoxicity, although it does identify the existence of soluble α-synuclein oligomers in vivo.
Q: What do you know about the toxicity of these oligos you’ve seen in vivo?
A: This is a very important question. We know nothing yet, but we would like to model their potential cytotoxicity. In the case of naturally generated, soluble Aβ oligomers released by certain cultured cells, Dominic Walsh and coworkers were able to apply these extracellularly in the living rat brain and demonstrate inhibitory effects on synaptic plasticity. But in the case of the lipid-dependent α-synuclein oligomers we describe here, one would wish to deliver them intraneuronally and assess any resultant toxicity, compared to control (nonoligomer) preparations. It may ultimately be possible to do so, but this will be technically challenging.
Q: How about Aβ? Presumably, oligomers of it exist in vivo, as well, prior to deposition. You’ve even published that Aβ accumulation begins intracellularly (Walsh et al., 2000). Has someone in the group tried delipidation to reveal them?
A: This is a good suggestion, and we need to do so. We focused on delipidation and the effects of fatty acids on the oligomerization of α-synuclein, in particular, because α-synuclein was known to interact with, and be affected by, lipids.
Q: There has been a lot of attention on lipid rafts recently. You found cytoplasmic interactions. Is this part of where the mechanisms for AD and PD/DLB diverge?
A: Much of the interest in lipid rafts in AD relates to the question of whether they represent a principal site for Aβ generation from AβPP and then for Aβ storage/accumulation. I don’t think our work on the effects of free fatty acids on α-synuclein oligomerization is directly related to this question.
Q: Do you know anything about the structure of these oligos? Are they α-helical or β-sheets? Where does the transition happen?
A: We do not. These lipid-dependent "low-n" oligomers are presumably present in relatively small amounts in the neuronal cytoplasm normally, and it is unclear if the initial oligomerization process requires a major conformational transition in the α-synuclein monomer, which has been reported to be "natively unfolded."
Q: Could it be that age-related changes in membrane lipid composition alter the binding and structure of α-synuclein in a way that promotes oligomerization? If yes, how so?
A: We currently favor the hypothesis that α-synuclein can normally interact with both free fatty acids and fatty acids within phospholipid membranes in the neuronal cytoplasm. In this sense, age-related changes in membrane lipid composition and/or in free FA levels in the cytoplasm could well influence α-synuclein oligomerization over time. As you noted, we observed an age-dependent increase in the levels of soluble and insoluble α-synuclein oligomers in mouse brains, but we cannot yet make a link to age-related membrane alterations.
Q: PUFAs are considered the healthier fatty acids. It seems counterintuitive that they promote oligomerization, whereas the "bad" saturated fats inhibit it. Some epidemiological evidence suggests that diets high in PUFAs protect against neurodegeneration. How do you square that?
A: It is too soon to say anything about the implications of our findings for dietary factors in PD and other synucleinopathies. The "bad" and "good" you refer to relate to risk for atherosclerotic disease, and this may not translate easily to risk for neurodegeneration. I think we now require dietary studies in mice and other animals on the effects of consuming diets rich in saturated FAs vs. PUFAs to determine whether exogenous fatty acid consumption has any effect on regulating the levels and progression of these lipid-dependent soluble α-synuclein oligomers in brain.
Q: Where do cholesterol and cholesterol esters fit into the bigger picture?
A: I don’t know yet.
Q: Neurodegeneration is sometimes called a problem of misfolded proteins. Are these "normal" cellular α-synuclein oligomers properly folded or misfolded?
A: If, as we believe, they occur normally in healthy cultured neuronal cells and in normal mouse brains, then they would presumably not be "misfolded," per se. But their accumulation over time and their transition to higher MW assemblies may involve changes in folding, so that some of them become "misfolded," or at least represent too much of a "properly folded" thing.
Q: Your findings, I presume, have nothing to do with α-synuclein’s presynaptic localization, but happen in a pool in the soma?
A: I see no immediate link, other than that the normally occurring α-synuclein oligomers are likely to be present in neuronal processes, not only in the cell bodies.