**6. Alpha-synuclein and membranes**

#### **6.1 Lipid-binding domains**

A characteristic feature of the α-synuclein amino-acid sequence is the set of seven degenerate 11-residue repeating motifs. These are reminescent of the amphipathic αhelical domains of the exchangeable apolipoproteins, which mediate a variety of lipid and protein interactions (Davidson et al., 1998, George et al., 1995). Amphipathicity corresponds to the segregation of polar and nonpolar residues to the two opposite faces of the α-helix, a distribution that is well suited for membrane binding (Drin & Antonny, 2009).

Depending upon the distribution of residues to the polar and nonpolar faces of the helices, Segrest et al. divided the apolipoprotein α-helices into different classes: class A helices bind lipids and are characterized by a clustering of basic residues at the polar/nonpolar interface and acidic residues at the center of the polar face, while class G helices are implicated in protein interactions and are characterized by a random radial distribution of charged residues to the polar face of the helix (Segrest et al., 1992). Davidson et al. subjected the entire α-synuclein sequence to helical wheel analysis and identified five potential amphipathic α-helices that encompass all of the 11-mer repeats and some of the adjacent amino acids. The first four of these five theoretical helices share the defining properties of class A2 lipid-binding helices, and they are distinguished by clustered basic residues at the polar-apolar interface and positioned ±100° from the center of the nonpolar face, with a preponderence of Lys over Arg, and the presence of Glu residues on the polar face. The αhelix on the fifth 11-mer repeat resembles a class G helix, and it is thus a candidate for protein-protein interactions (Davidson et al., 1998).

There is a notable feature that can be used to distinguish between putative amphipathic αsynuclein helices and those in the exchangeable apolipoproteins: the Thr residues at the center of the nonpolar faces of helices 2-4 (Davidson et al., 1998). Although polar, these can reside on the nonpolar face of the helix due to its relatively long aliphatic side chain (Segrest et al., 1992). Thr are conserved among the α-synuclein sequences from canary, human and rat, suggesting that they indeed have an important function. Another unique aspect of the αsynuclein 11-mer repeat sequences is the absence of Pro, which in exchangeable apolipoproteins introduces helix-breaking hairpin turns. In contrast, α-synuclein helices 1-4 appear to be punctuated by nonpolar residues that are predicted to disrupt the amphipathicity of a helix (Davidson et al., 1998).

#### **6.2 Lipid and membrane selectivity**

92 Etiology and Pathophysiology of Parkinson's Disease

Tyr, we created Tyr to Ala mutants to examine the importance of these Tyr residues in fibril formation of α-synuclein *in vitro*. This was completely inhibited in the timescale over which measurements were made (70 hours) when the three C-terminal Tyr were replaced with Ala. In addition, substitution of Tyr133 by Ala also inhibitted fibril formation, whereas the individual Y125A and Y136A mutants showed limited inhibition. Replacement of Tyr39 by Ala also resulted in substantial inhibition of fibril formation. Structural analysis showed that the Y133A α-synuclein mutant has a substantially different conformation, as it is rich in αhelical secondary structure, as compared with wild-type α-synuclein and its other mutants. However, no formation of any tertiary structure was seen, as judged from the near-UV circular-dichroism spectra. These observations suggest that the long-range intramolecular interactions between the N-terminal and C-terminal of α-synuclein are crucial for the

A characteristic feature of the α-synuclein amino-acid sequence is the set of seven degenerate 11-residue repeating motifs. These are reminescent of the amphipathic αhelical domains of the exchangeable apolipoproteins, which mediate a variety of lipid and protein interactions (Davidson et al., 1998, George et al., 1995). Amphipathicity corresponds to the segregation of polar and nonpolar residues to the two opposite faces of the α-helix, a distribution that is well suited for membrane binding (Drin & Antonny,

Depending upon the distribution of residues to the polar and nonpolar faces of the helices, Segrest et al. divided the apolipoprotein α-helices into different classes: class A helices bind lipids and are characterized by a clustering of basic residues at the polar/nonpolar interface and acidic residues at the center of the polar face, while class G helices are implicated in protein interactions and are characterized by a random radial distribution of charged residues to the polar face of the helix (Segrest et al., 1992). Davidson et al. subjected the entire α-synuclein sequence to helical wheel analysis and identified five potential amphipathic α-helices that encompass all of the 11-mer repeats and some of the adjacent amino acids. The first four of these five theoretical helices share the defining properties of class A2 lipid-binding helices, and they are distinguished by clustered basic residues at the polar-apolar interface and positioned ±100° from the center of the nonpolar face, with a preponderence of Lys over Arg, and the presence of Glu residues on the polar face. The αhelix on the fifth 11-mer repeat resembles a class G helix, and it is thus a candidate for

There is a notable feature that can be used to distinguish between putative amphipathic αsynuclein helices and those in the exchangeable apolipoproteins: the Thr residues at the center of the nonpolar faces of helices 2-4 (Davidson et al., 1998). Although polar, these can reside on the nonpolar face of the helix due to its relatively long aliphatic side chain (Segrest et al., 1992). Thr are conserved among the α-synuclein sequences from canary, human and rat, suggesting that they indeed have an important function. Another unique aspect of the αsynuclein 11-mer repeat sequences is the absence of Pro, which in exchangeable apolipoproteins introduces helix-breaking hairpin turns. In contrast, α-synuclein helices 1-4 appear to be punctuated by nonpolar residues that are predicted to disrupt the

process of fibril formation (Ulrih et al., 2008).

**6. Alpha-synuclein and membranes** 

protein-protein interactions (Davidson et al., 1998).

amphipathicity of a helix (Davidson et al., 1998).

**6.1 Lipid-binding domains** 

2009).

Data that have documented the tendency of α-synuclein to colocalize with synaptic vesicles *in vivo* (Maroteaux et al., 1988) and the presence of the 11-residue repeated domains in a pattern similar to that found in the apolipoproteins (George et al., 1995) sparked a series of studies to determine the α-synuclein lipid-binding ability. Alpha-synuclein interactions with membranes have been found to be one of the most contentious areas regarding this protein (Fink, 2006), as there have been numerous reports on sometimes completely contradicting results, and as there might be major differences between the situation *in vivo* and *in vitro*. Also, membranes have been reported to both accelerate (Lee et al., 2002) and inhibit (Narayanan & Scarlata, 2001; Zhu & Fink, 2003) α-synuclein fibril formation, so this probably reflects the varying conditions used in the different studies (Zhu & Fink, 2003). All three of these α-synuclein mutations occur within the N-terminus, which is responsible for its membrane binding, hence suggesting an effect on membrane interactions (Fortin et al., 2010). The A30P α-synuclein mutation, and to a lesser extent that of A53T, disrupts the helical structure of α-synuclein (Bussell & Eliezer, 2001), although it does not significantly

affect the structure of membrane-associated α-synuclein (Bussell & Eliezer, 2004). The E46K α-synuclein mutant binds to negatively charged vesicles with a higher protein/lipid ratio than does wild-type α-synuclein (Choi et al., 2004), while A30P affects the localization, and presumably the membrane binding, of α-synuclein *in vivo* (Fortin et al., 2010).

#### **6.2.1 Membrane interactions** *in vitro*

It is generally accepted that α-synuclein preferentially interacts with small unilamellar vesicles (SUVs) containing negatively charged head groups (Davidson et al., 1998; Jo et al., 2000) or interfacial packing defects (Kamp & Beyer, 2006; Nuscher et al., 2004), and that upon SUV binding, α-synuclein undergoes a conformational transition from an intrinsically disordered state to an α-helical structure (Davidson et al., 1998; Jo et al., 2000; Nuscher et al., 2004). Various combinations of charged and uncharged lipids have been used in these studies. These negatively charged acidic phospholipids include phosphatidylglycerol (PG), phosphatidic acid (PA), phosphatidylserine (PS) and phosphatidylinositol (PI), while the uncharged, neutral lipids commonly used include phosphatidylcholine (PC) and phosphatidylethanolamine (PE) (Valenzuela, 2007).

The interactions of α-synuclein with membranes have been shown to affect the properties of both the protein and the membranes, and both electrostatic and hydrophobic interactions are important in the protein-bilayer association (Zhu et al., 2003). There are several factors that are believed to have central roles in modulation of the binding equilibrium of αsynuclein to membranes, including chemical properties of the membranes (Davidson et al., 1998; Jo et al., 2000), ionic strength of the solution (Davidson et al., 1998; Zhu et al., 2003), vesicle size, or more precisely, the curvature of the phospholipid surface (Davidson et al., 1998; Narayanan & Scarlata, 2001; Rhoades et al., 2006), and mass ratio of α-synuclein to the lipids (Zhu & Fink, 2003). Here, an overview of some of the more important findings regarding the lipid specificities of α-synuclein are given.

Davidson et al. were the first to demonstrate that α-synuclein binds only to acidic phospholipids and preferentially to vesicles with smaller diameters. Circular dichroism spectroscopy was used to determine the effects of this lipid binding on the secondary structure of α-synuclein. In buffer solution, α-synuclein is mainly unstructured, with less than 3% of the structure as α-helix. Incubation of α-synuclein with vesicles made of a

Alpha-Synuclein Interactions with Membranes 95

POPC. This initial binding did not induce changes in the secondary structure of α-synuclein. This study also supported the role of the α-synuclein C-terminus in membrane binding, by showing that lowering the pH of folded α-synuclein, which reduces the negative charge of α-synuclein, greatly increases the binding affinity without altering the secondary structure

Fluorescence correlation spectroscopy was used as a tool for rapid and quantitative analysis of the lipid binding of α-synuclein. Some studies have confirmed the importance of the negatively charged lipids (PA and PS) for α-synuclein binding to LUVs with 120 nm diameter, when no pre-incubation of α-synuclein and the vesicles was used. Alphasynuclein has a significantly higher affinity for vesicles that contain some POPA, over those that contain an equivalent amount of POPS. The reason for this could be the polar POPA head-group, which is smaller in size compared to that of POPS and might therefore be able to pack more closely together in a lipid bilayer, producing a higher charge density. Alphasynuclein shows slightly higher binding affinity to POPE compared to POPC (Rhoades et al., 2006). Combined with other data in the literature (Davidson et al., 1998; Jo et al., 2000; Nuscher et al., 2004), these suggest that each molecule of α-synuclein can bind to a lipid bilayer patch composed of ≤85 acidic lipid molecules, corresponding in the case of POPS to a weight ratio of bound protein to total lipid of approximately 1:5. Interestingly, at higher αsynuclein concentrations, the amount of bound α-synuclein decreases, suggesting a destabilization of the membrane. This study also confirmed the importance of electrostatic

interactions for the binding between α-synuclein and the lipids (Rhoades et al., 2006). The binding of α-synuclein to SUVs has been monitored by measuring the changes in intrinsic fluorescence emanating from the four Tyr residues in α-synuclein. These data have suggested that α-synuclein binds to both negatively charged and electrically neutral SUVs, although slightly weaker for the latter. This binding to electrically neutral vesicles is presumably due to electrostatic interactions between the negatively charged C-terminal region of α-synuclein and the positively charged choline. Binding to different types of vesicles was also detected in high ionic strength solutions. These data indicate that for αsynuclein binding to lipids, not only are electrostatic interactions important, but also hydrophobic interactions. The influence of α-synuclein binding to the membrane has also been examined. Due to the differences in the excitation spectra and polarisation of the Laurdan dye after incubation of dipalmitoyl PA (DPPA)/dipalmitoyl PC (DPPC) and dipalmitoyl PG (DPPG)/DPPC SUVs with α-synuclein, this study concluded that αsynuclein is inserted deep into the membrane and does not remain bound only on the surface. A lack of significant penetration of α-synuclein into the DPPC vesicle bilayer was observed. Random coil–helix structure transition was most notable when SUVs composed of DPPG or dipalmitoyl PS (DPPS) or their mixtures with DPPC or dipalmitoyl PE (DPPE) were used. The amount of helix induced was smaller for DPPA/DPPC. SUVs made of DPPC only do not trigger the formation of the α-synuclein α-helix structure; presumably αsynuclein binds to the surface of these vesicles due to electrostatic interactions, but does not

The small diameter of the SUVs leads to curvature stress in the bilayer, which results in a rather broad phase transition that is centered at ~4-5 °C below the chain-melting phasetransition temperature (Tm), and thus vesicles made of DPPC undergo melting transition at 36 °C rather than at 41 °C (Gaber & Sheridan, 1982). Using isothermal titration calorimetry, differential scanning calorimetry (Nuscher et al., 2004), spin label electron paramagnetic

(Narayanan & Scarlata, 2001).

induce the helical structure (Zhu et al., 2003).

mixture of acidic and neutral phospholipids is accompanied by a large increase in its αhelical content. Alpha-synuclein does not bind to SUVs or multilamellar vesicles composed of PC only, or of a mixture of PC and PE. Here, α-synuclein binds to SUVs containing at least 30% to 50% acidic lipids, per vesicle weight. Comparisons of the ability of α-synuclein to bind to vesicles of different sizes, as SUVs with 20 nm to 25 nm diameters and large unilamellar vesicles (LUV) with 125 ±30 nm diameter, have shown that α-synuclein does not bind to vesicles that contained neutral 1-palmitoyl 2-oleoyl PC (POPC) alone. Also, binding to LUVs composed of POPC/1-palmitoyl 2-oleoyl PA (POPA) was less than to SUVs of the same composition (Davidson et al., 1998).

Binding to negatively charged SUVs was confirmed in another study where they incubated α-synuclein together with vesicles and then fractionated the solution with gel filtration chromatography. Alpha-synuclein eluted together with the SUV fraction when incubated with either POPC/POPA or POPC/1-palmitoyl 2-oleoyl PS (POPS) phospholipids, while its binding to vesicles composed of POPC alone was not detected (Perrin et al., 2000). Althought in nervous tissue PA comprises approximately 1% to 3% of the total phospholipids, while PS is more abundant (12% to 22% of the total phospholipids) (Sastry et al., 1985), it is difficult to relate these values directly to the local composition of specific membranes inside the cell, since these lipids are not distributed evenly in the cell and are generated and metabolized rapidly (Perrin et al., 2000).

Using thin-layer chromatography overlay, it has been shown that α-synuclein binds to the brain or commercially available lipids PE, PI and lyso-PE. These interactions were much weaker with POPS and brain PS, and absent with POPC, POPA, sphingomyelin and cholesterol (Jo et al., 2000). Surprisingly, in contrast with a previous report (Davidson et al., 1998), α-synuclein does not bind to PA, which was attributed to the properties of the thinlayer chromatography overlay method. Replacing PC with PE in acidic lipid vesicles greatly increased the binding of α-synuclein (Jo et al., 2000). Althought both PE and PC are neutral phospholipids that have similar electrostatic properties, they differ in their head-group orientation, lipid bilayer packing, and hydrogen-bonding capacity (Hauser et al., 1981, as cited in Jo et al., 2000). When the neutral head-groups are tightly packed, PE forms a lipid monolayer with high negative curvature (Bazzi et al., 1992). It is believed that α-synuclein can relieve this negative curvature strain, and hence stabilize the PE/acidic lipid vesicles. This study also showed that both SUVs and multilamellar vesicles composed of POPC/POPS induce α-helical secondary structures, which suggests that the vesicle size does not impact on the α-synuclein secondary structure. With the neutral charged PE in the presence of acidic phospholipids (PA and PI), this significant increases the α-synuclein αhelicity. However, it should be emphasized that the changes in α-synuclein secondary structure are much lower in the presence of neutral PC in combination with negatively charged lipids (Jo et al., 2000).

In contrast to previous studies that used prolonged incubation times and mechanical separation of the products, and which found that α-synuclein only bound to SUVs composed of acidic phospholipids (Davidson et al., 1998; Jo et al., 2000; Perrin et al., 2000), in another study, the association between α-synuclein and lipids was viewed immediately after the addition of lipids to α-synuclein. Data were then obtained either by monitoring the change in intrinsic fluorescence emanating from four α-synuclein Tyr residues, or by adding the Laurdan fluorescent probe to the vesicles. Surprisingly, α-synuclein bound with similar affinities to LUVs composed of the negatively charged POPS and the electrically neutral

mixture of acidic and neutral phospholipids is accompanied by a large increase in its αhelical content. Alpha-synuclein does not bind to SUVs or multilamellar vesicles composed of PC only, or of a mixture of PC and PE. Here, α-synuclein binds to SUVs containing at least 30% to 50% acidic lipids, per vesicle weight. Comparisons of the ability of α-synuclein to bind to vesicles of different sizes, as SUVs with 20 nm to 25 nm diameters and large unilamellar vesicles (LUV) with 125 ±30 nm diameter, have shown that α-synuclein does not bind to vesicles that contained neutral 1-palmitoyl 2-oleoyl PC (POPC) alone. Also, binding to LUVs composed of POPC/1-palmitoyl 2-oleoyl PA (POPA) was less than to SUVs of the

Binding to negatively charged SUVs was confirmed in another study where they incubated α-synuclein together with vesicles and then fractionated the solution with gel filtration chromatography. Alpha-synuclein eluted together with the SUV fraction when incubated with either POPC/POPA or POPC/1-palmitoyl 2-oleoyl PS (POPS) phospholipids, while its binding to vesicles composed of POPC alone was not detected (Perrin et al., 2000). Althought in nervous tissue PA comprises approximately 1% to 3% of the total phospholipids, while PS is more abundant (12% to 22% of the total phospholipids) (Sastry et al., 1985), it is difficult to relate these values directly to the local composition of specific membranes inside the cell, since these lipids are not distributed evenly in the cell and are

Using thin-layer chromatography overlay, it has been shown that α-synuclein binds to the brain or commercially available lipids PE, PI and lyso-PE. These interactions were much weaker with POPS and brain PS, and absent with POPC, POPA, sphingomyelin and cholesterol (Jo et al., 2000). Surprisingly, in contrast with a previous report (Davidson et al., 1998), α-synuclein does not bind to PA, which was attributed to the properties of the thinlayer chromatography overlay method. Replacing PC with PE in acidic lipid vesicles greatly increased the binding of α-synuclein (Jo et al., 2000). Althought both PE and PC are neutral phospholipids that have similar electrostatic properties, they differ in their head-group orientation, lipid bilayer packing, and hydrogen-bonding capacity (Hauser et al., 1981, as cited in Jo et al., 2000). When the neutral head-groups are tightly packed, PE forms a lipid monolayer with high negative curvature (Bazzi et al., 1992). It is believed that α-synuclein can relieve this negative curvature strain, and hence stabilize the PE/acidic lipid vesicles. This study also showed that both SUVs and multilamellar vesicles composed of POPC/POPS induce α-helical secondary structures, which suggests that the vesicle size does not impact on the α-synuclein secondary structure. With the neutral charged PE in the presence of acidic phospholipids (PA and PI), this significant increases the α-synuclein αhelicity. However, it should be emphasized that the changes in α-synuclein secondary structure are much lower in the presence of neutral PC in combination with negatively

In contrast to previous studies that used prolonged incubation times and mechanical separation of the products, and which found that α-synuclein only bound to SUVs composed of acidic phospholipids (Davidson et al., 1998; Jo et al., 2000; Perrin et al., 2000), in another study, the association between α-synuclein and lipids was viewed immediately after the addition of lipids to α-synuclein. Data were then obtained either by monitoring the change in intrinsic fluorescence emanating from four α-synuclein Tyr residues, or by adding the Laurdan fluorescent probe to the vesicles. Surprisingly, α-synuclein bound with similar affinities to LUVs composed of the negatively charged POPS and the electrically neutral

same composition (Davidson et al., 1998).

charged lipids (Jo et al., 2000).

generated and metabolized rapidly (Perrin et al., 2000).

POPC. This initial binding did not induce changes in the secondary structure of α-synuclein. This study also supported the role of the α-synuclein C-terminus in membrane binding, by showing that lowering the pH of folded α-synuclein, which reduces the negative charge of α-synuclein, greatly increases the binding affinity without altering the secondary structure (Narayanan & Scarlata, 2001).

Fluorescence correlation spectroscopy was used as a tool for rapid and quantitative analysis of the lipid binding of α-synuclein. Some studies have confirmed the importance of the negatively charged lipids (PA and PS) for α-synuclein binding to LUVs with 120 nm diameter, when no pre-incubation of α-synuclein and the vesicles was used. Alphasynuclein has a significantly higher affinity for vesicles that contain some POPA, over those that contain an equivalent amount of POPS. The reason for this could be the polar POPA head-group, which is smaller in size compared to that of POPS and might therefore be able to pack more closely together in a lipid bilayer, producing a higher charge density. Alphasynuclein shows slightly higher binding affinity to POPE compared to POPC (Rhoades et al., 2006). Combined with other data in the literature (Davidson et al., 1998; Jo et al., 2000; Nuscher et al., 2004), these suggest that each molecule of α-synuclein can bind to a lipid bilayer patch composed of ≤85 acidic lipid molecules, corresponding in the case of POPS to a weight ratio of bound protein to total lipid of approximately 1:5. Interestingly, at higher αsynuclein concentrations, the amount of bound α-synuclein decreases, suggesting a destabilization of the membrane. This study also confirmed the importance of electrostatic interactions for the binding between α-synuclein and the lipids (Rhoades et al., 2006).

The binding of α-synuclein to SUVs has been monitored by measuring the changes in intrinsic fluorescence emanating from the four Tyr residues in α-synuclein. These data have suggested that α-synuclein binds to both negatively charged and electrically neutral SUVs, although slightly weaker for the latter. This binding to electrically neutral vesicles is presumably due to electrostatic interactions between the negatively charged C-terminal region of α-synuclein and the positively charged choline. Binding to different types of vesicles was also detected in high ionic strength solutions. These data indicate that for αsynuclein binding to lipids, not only are electrostatic interactions important, but also hydrophobic interactions. The influence of α-synuclein binding to the membrane has also been examined. Due to the differences in the excitation spectra and polarisation of the Laurdan dye after incubation of dipalmitoyl PA (DPPA)/dipalmitoyl PC (DPPC) and dipalmitoyl PG (DPPG)/DPPC SUVs with α-synuclein, this study concluded that αsynuclein is inserted deep into the membrane and does not remain bound only on the surface. A lack of significant penetration of α-synuclein into the DPPC vesicle bilayer was observed. Random coil–helix structure transition was most notable when SUVs composed of DPPG or dipalmitoyl PS (DPPS) or their mixtures with DPPC or dipalmitoyl PE (DPPE) were used. The amount of helix induced was smaller for DPPA/DPPC. SUVs made of DPPC only do not trigger the formation of the α-synuclein α-helix structure; presumably αsynuclein binds to the surface of these vesicles due to electrostatic interactions, but does not induce the helical structure (Zhu et al., 2003).

The small diameter of the SUVs leads to curvature stress in the bilayer, which results in a rather broad phase transition that is centered at ~4-5 °C below the chain-melting phasetransition temperature (Tm), and thus vesicles made of DPPC undergo melting transition at 36 °C rather than at 41 °C (Gaber & Sheridan, 1982). Using isothermal titration calorimetry, differential scanning calorimetry (Nuscher et al., 2004), spin label electron paramagnetic

Alpha-Synuclein Interactions with Membranes 97

might be a fatty-acid-binding protein (Sharon et al., 2001). In contrast, a later NMR study excluded high-affinity binding of fatty-acid molecules to specific α-synuclein sites (Lucke et al., 2006). Exposure of living mesencephalic neurons to polyunsaturated fatty acids (PUFAs) increased the α-synuclein oligomer levels, whereas saturated fatty acids decreased these. Here, α-synuclein interacts with the free PUFAs to form the first soluble oligomers, which then aggregate into insoluble high-molecular-weight complexes (Sharon et al., 2003a). Indeed, elevated PUFA levels have been detected in the soluble fractions of PD and Lewy bodies dementia brains. The levels of saturated and monounsaturated fatty acids did not change in these PD brains or in cells overexpressing α-synuclein, which indicated that αsynuclein is involved specifically in the maintenance of PUFA levels (Sharon et al., 2003b). Using binding assays, it has been demonstrated that α-synuclein binds saturably and with high affinity to detergent-resistant membranes, to lipid rafts, in permeabilized HeLa cells, and in the presence of synaptosomal membranes from transgenic mice expressing human αsynuclein. The A53T α-synuclein mutation has no detectable effects on this binding, while the A30P mutation disrupts the association, which supports the specificity of the interaction (Fortin et al., 2004). It should also be mentioned that both of these mutations do not generally affect the interactions of α-synuclein with artificial membranes (Perrin et al., 2000), probably because these membranes fail to reproduce the full characteristics of lipid rafts (Fortin et al., 2004). In contrast, the A30P mutation distrupts α-synuclein association with native membranes, such as those of axonal transport vesicles, lipid droplets produced in HeLa cells by the administration of oleic acid, and yeast (Cole et al., 2002; Jensen et al., 1998; Outeiro et al., 2003). Alterations in the electrophoretic mobility of α-synuclein upon membrane binding have confirmed its binding to lipid rafts, with this interaction resistant to digestion of the rafts with proteinase K, which suggests that the lipids, rather than proteins, are required. This assumption is also supported by high affinity binding of α-synuclein to artificial membranes that do indeed mimic lipid rafts. Cholesterol does not appear to be required for the binding, but rather for maintenance of raft integrity; sphingolipid also

appears not to be crucial for these interactions (Kubo et al., 2005).

2005).

Similar to previous reports (Davidson et al., 1998; Perrin et al., 2000), Kubo et al. reported that α-synuclein binding requires acidic phospholipids, with a preference for PS. Synthetic PS with defined acyl chains did not support this binding when used individually, with the combination of 18:1 PS and PS with polyunsaturated acyl chains required both to bind to and to shift the electrophoretic mobility of α-synuclein. The addition of 18:1 PC to 20:4 PS, or conversely, the addition of 20:4 PC to 18:1 PS, also promoted α-synuclein binding. The requirement for both monounsaturated and polyunsaturated acyl chains suggests that the interaction of α-synuclein requires membranes with two distinct phases: lipid rafts in a liquid-ordered phase, and the rest of the cell membrane in a liquid-disordered phase. Alphasynuclein binds with higher affinity to artificial membranes with the PS head-group on the polyunsaturated fatty acyl chain rather than on the oleoyl side chain, apparently reflecting an interaction of α-synuclein with both the acyl chain and the head-group (Kubo et al.,

In contrast to artificial membranes, the interactions of α-synuclein with biological membranes are highly dynamic and they show rapid dissociation. Thus, rather than electrostatic interactions, Kim et al. suggested the involment of hydrophobic interactions. Furthermore, the interactions of α-synuclein with cellular membranes occured only in the presence of nonprotein and nonlipid cytosolic components, which distinguished it from the

resonance (EPR), and fluorescence spectroscopy (Kamp & Beyer, 2006), it has been shown that α-synuclein affects the lipid packing in neutral SUVs. Here α-synuclein induces chain ordering below the Tm, but not in the liquid crystalline state of zwitterionic vesicle membranes. Binding of α-synuclein leads to an increase in the temperature and cooperitivity of the phase transition, which was attributed to defect healing in the curved vesicle membranes. Binding to the vesicles also induces coil-helix transitions of α-synuclein (Kamp & Beyer, 2006; Nuscher et al., 2004). SUVs made of POPC/POPG at a molar ratio of 1:1 and 2:1 cause α-helix formation in the structure of α-synuclein, and this is more pronounced at the 1:1 ratio. A helix structure is not observed in LUVs of the same composition. This again hightlights the importance of the negative charge and size of lipid vesicles for α-synuclein α-helix formation. A more important finding is the formation of the helical structure by binding to SUVs of neutrally charged DPPC under the Tm and not above that temperature (Nuscher et al., 2004).

Recently, Bartels et al. used circular dichroism spectroscopy and isotermal titration calorimetry to investigate peptide fragments from different domains of the full-length αsynuclein protein. They showed that membrane recognition of the N-terminus is essential for the cooperative formation of helical domains in the protein. This suggests that the membrane-induced helical folding of the first 25 residues of α-synuclein might be driven simultaneously by electrostatic attraction and by changes in lipid ordering (Bartels et al., 2010).

#### **6.2.2 Membrane interactions** *in vivo*

Compared with the α-synuclein–lipid interaction *in vitro*, the interaction of α-synuclein with membranes in cells is less well understood. Cole et al. investigated α-synuclein interactions with intracellular lipid stores in cultured cells treated with high concentrations of fatty acids. Here, α-synuclein accumulated on phospholipid monolayers surrounding triglyceride-rich lipid droplets and protected the stored triglycerides from hydrolysis. Chemical cross-linking experiments led to the suggestion that dimers or trimers of αsynuclein were associated with the droplet surface (Cole et al., 2002).

Alpha-synuclein can be imported into cells (Sung et al., 2001) and can be secreted from cells, althought it lacks a conventional signal sequence for secretion (Lee et al., 2005). Lee et al. reported that a portion of α-synuclein is stored in the lumen of vesicles in the cytoplasm, and that the α-synuclein in vesicles might be secreted through an unconventional exocytosis pathway. This study also demonstrated that intravesicular α-synuclein is more prone to aggregation than cytosolic α-synuclein, and that aggregated forms of α-synuclein are also secreted from cells (Lee et al., 2005). They thus used a series of deletion mutants and recombinant peptides to determine the amino-acid sequence motifs of α-synuclein that were required for its membrane translocation. The N-terminal region and the NAC peptide were shown to be necessary for translocation, althought the NAC was less effective than the Nterminal region. This thus suggested that the 11-amino acid repeat sequences bind to the lipid bilayer and that this binding interaction is crucial for α-synuclein translocation. Cellular uptake of α-synuclein was not significantlly affected by treatment with inhibitors of endocytosis, suggesting that this occurs via a mechanism distinct from normal endocytosis (Ahn et al., 2006).

Sharon et al. showed that free fatty acids have specific roles in the formation and maintenance of the soluble α-synuclein oligomers, and they suggested that α-synuclein

resonance (EPR), and fluorescence spectroscopy (Kamp & Beyer, 2006), it has been shown that α-synuclein affects the lipid packing in neutral SUVs. Here α-synuclein induces chain ordering below the Tm, but not in the liquid crystalline state of zwitterionic vesicle membranes. Binding of α-synuclein leads to an increase in the temperature and cooperitivity of the phase transition, which was attributed to defect healing in the curved vesicle membranes. Binding to the vesicles also induces coil-helix transitions of α-synuclein (Kamp & Beyer, 2006; Nuscher et al., 2004). SUVs made of POPC/POPG at a molar ratio of 1:1 and 2:1 cause α-helix formation in the structure of α-synuclein, and this is more pronounced at the 1:1 ratio. A helix structure is not observed in LUVs of the same composition. This again hightlights the importance of the negative charge and size of lipid vesicles for α-synuclein α-helix formation. A more important finding is the formation of the helical structure by binding to SUVs of neutrally charged DPPC under the Tm and not above

Recently, Bartels et al. used circular dichroism spectroscopy and isotermal titration calorimetry to investigate peptide fragments from different domains of the full-length αsynuclein protein. They showed that membrane recognition of the N-terminus is essential for the cooperative formation of helical domains in the protein. This suggests that the membrane-induced helical folding of the first 25 residues of α-synuclein might be driven simultaneously by electrostatic attraction and by changes in lipid ordering (Bartels et al.,

Compared with the α-synuclein–lipid interaction *in vitro*, the interaction of α-synuclein with membranes in cells is less well understood. Cole et al. investigated α-synuclein interactions with intracellular lipid stores in cultured cells treated with high concentrations of fatty acids. Here, α-synuclein accumulated on phospholipid monolayers surrounding triglyceride-rich lipid droplets and protected the stored triglycerides from hydrolysis. Chemical cross-linking experiments led to the suggestion that dimers or trimers of α-

Alpha-synuclein can be imported into cells (Sung et al., 2001) and can be secreted from cells, althought it lacks a conventional signal sequence for secretion (Lee et al., 2005). Lee et al. reported that a portion of α-synuclein is stored in the lumen of vesicles in the cytoplasm, and that the α-synuclein in vesicles might be secreted through an unconventional exocytosis pathway. This study also demonstrated that intravesicular α-synuclein is more prone to aggregation than cytosolic α-synuclein, and that aggregated forms of α-synuclein are also secreted from cells (Lee et al., 2005). They thus used a series of deletion mutants and recombinant peptides to determine the amino-acid sequence motifs of α-synuclein that were required for its membrane translocation. The N-terminal region and the NAC peptide were shown to be necessary for translocation, althought the NAC was less effective than the Nterminal region. This thus suggested that the 11-amino acid repeat sequences bind to the lipid bilayer and that this binding interaction is crucial for α-synuclein translocation. Cellular uptake of α-synuclein was not significantlly affected by treatment with inhibitors of endocytosis, suggesting that this occurs via a mechanism distinct from normal endocytosis

Sharon et al. showed that free fatty acids have specific roles in the formation and maintenance of the soluble α-synuclein oligomers, and they suggested that α-synuclein

synuclein were associated with the droplet surface (Cole et al., 2002).

that temperature (Nuscher et al., 2004).

**6.2.2 Membrane interactions** *in vivo* 

2010).

(Ahn et al., 2006).

might be a fatty-acid-binding protein (Sharon et al., 2001). In contrast, a later NMR study excluded high-affinity binding of fatty-acid molecules to specific α-synuclein sites (Lucke et al., 2006). Exposure of living mesencephalic neurons to polyunsaturated fatty acids (PUFAs) increased the α-synuclein oligomer levels, whereas saturated fatty acids decreased these. Here, α-synuclein interacts with the free PUFAs to form the first soluble oligomers, which then aggregate into insoluble high-molecular-weight complexes (Sharon et al., 2003a). Indeed, elevated PUFA levels have been detected in the soluble fractions of PD and Lewy bodies dementia brains. The levels of saturated and monounsaturated fatty acids did not change in these PD brains or in cells overexpressing α-synuclein, which indicated that αsynuclein is involved specifically in the maintenance of PUFA levels (Sharon et al., 2003b).

Using binding assays, it has been demonstrated that α-synuclein binds saturably and with high affinity to detergent-resistant membranes, to lipid rafts, in permeabilized HeLa cells, and in the presence of synaptosomal membranes from transgenic mice expressing human αsynuclein. The A53T α-synuclein mutation has no detectable effects on this binding, while the A30P mutation disrupts the association, which supports the specificity of the interaction (Fortin et al., 2004). It should also be mentioned that both of these mutations do not generally affect the interactions of α-synuclein with artificial membranes (Perrin et al., 2000), probably because these membranes fail to reproduce the full characteristics of lipid rafts (Fortin et al., 2004). In contrast, the A30P mutation distrupts α-synuclein association with native membranes, such as those of axonal transport vesicles, lipid droplets produced in HeLa cells by the administration of oleic acid, and yeast (Cole et al., 2002; Jensen et al., 1998; Outeiro et al., 2003). Alterations in the electrophoretic mobility of α-synuclein upon membrane binding have confirmed its binding to lipid rafts, with this interaction resistant to digestion of the rafts with proteinase K, which suggests that the lipids, rather than proteins, are required. This assumption is also supported by high affinity binding of α-synuclein to artificial membranes that do indeed mimic lipid rafts. Cholesterol does not appear to be required for the binding, but rather for maintenance of raft integrity; sphingolipid also appears not to be crucial for these interactions (Kubo et al., 2005).

Similar to previous reports (Davidson et al., 1998; Perrin et al., 2000), Kubo et al. reported that α-synuclein binding requires acidic phospholipids, with a preference for PS. Synthetic PS with defined acyl chains did not support this binding when used individually, with the combination of 18:1 PS and PS with polyunsaturated acyl chains required both to bind to and to shift the electrophoretic mobility of α-synuclein. The addition of 18:1 PC to 20:4 PS, or conversely, the addition of 20:4 PC to 18:1 PS, also promoted α-synuclein binding. The requirement for both monounsaturated and polyunsaturated acyl chains suggests that the interaction of α-synuclein requires membranes with two distinct phases: lipid rafts in a liquid-ordered phase, and the rest of the cell membrane in a liquid-disordered phase. Alphasynuclein binds with higher affinity to artificial membranes with the PS head-group on the polyunsaturated fatty acyl chain rather than on the oleoyl side chain, apparently reflecting an interaction of α-synuclein with both the acyl chain and the head-group (Kubo et al., 2005).

In contrast to artificial membranes, the interactions of α-synuclein with biological membranes are highly dynamic and they show rapid dissociation. Thus, rather than electrostatic interactions, Kim et al. suggested the involment of hydrophobic interactions. Furthermore, the interactions of α-synuclein with cellular membranes occured only in the presence of nonprotein and nonlipid cytosolic components, which distinguished it from the

Alpha-Synuclein Interactions with Membranes 99

an elongated helical structure that is devoid of any significant tertiary packing (Jao et al., 2004), or they have suggested a broken helical structure (Bortolus et al., 2008; Drescher et al., 2008). A number of recent studies have highlighted the ongoing debate regarding the physiologically relevant form, as the bent or extended membrane-bound helix (Figure 2).

Fig. 2. Illustrations of the generally proposed α-synuclein structures on (a) micelles (bent helix model) and (b) SUVs (elongated helix model) (Jao et al., 2008; Trexler & Rhoades, 2009).

Several studies have raised the question of the periodicity of the helix that is formed upon binding of α-synuclein to membranes. As indicated above, the N-terminus of α-synuclein contains seven imperfect 11-residue repeats. Using site-directed spin-labeling, it has been shown that repeats 5–7 of α-synuclein are bound to SUVs with an 11/3 periodicity (11 residues to complete three full turns) (Jao et al., 2004)*.* Sodium dodecyl sulphate (SDS) micelle-bound α-synuclein shows the same periodicity, as opposed to the 18/5 periodicity of an ideal α-helix (Bussell et al., 2005). In this ideal 18/5 periodicity, there are 3.6 residues per turn and the rotation per residue is 100°. In the α-synuclein 11/3 periodicity, the number of residues per turn is 3.67 and the rotation per residue is 98.18°. Using theoretical methods, it has indeed been concluded that the periodicity of α-synuclein is 11/3, and that through the positioning of the charged residues, this has implications for α-synuclein membrane binding. These calculations show that the energy penalty for a change in periodicity from the 18/5 to 11/3 on anionic membranes is overcome by the favorable solvation energy (Mihajlovic & Lazaridis, 2008).

Eliezer et al. were the first to use NMR spectroscopy to characterize the conformational properties of α-synuclein when bound to lipid vesicles and lipid-mimetic detergent micelles. They demonstrated that only the first 100 residues of the N-terminal region of α-synuclein bind to both SDS micelles and PA/PC vesicles and fold into an amphipathic helix, while the

Ulmer et al. have described the structure and dynamics of α-synuclein in the micelle-bound form according to solution NMR spectroscopy. In binding to SDS micelles or SDS micelles with dodecylPC (DPC), α-synuclein forms two curved α-helices (Figure 3), helix-N (Val3-Val37) and helix-C (Lys45-Thr92). These helices are connected by a well-ordered, extended linker in an unexpected anti-parallel arrangement, which is followed by another short extended region (Gly93-Lys97) that overlaps with a chaperone-mediated autophagy recognition motif and a

**7.2 Analysis of α-synuclein structure by nuclear magnetic resonance** 

acidic C-terminal region of α-synuclein remains unstructured (Eliezer et al., 2001).

predominantly unstructured highly mobile tail (Asp98-Ala140) (Ulmer et al., 2005).

**7.1 Helix periodicity** 

spontanous interaction with artificial membranes. Here, addition of a cytosolic preparation to the artificial membranes resulted in similar binding of α-synuclein as for biological membranes (Kim et al., 2006).

Lipid rafts contain a lot of the ganglioside GM1, and it has been suggested that the gangliosides mediate or facilitate the association of α-synuclein with neuronal membranes (Martinez et al., 2007). However, recently Di Pasquale et al. identifed a ganglioside-binding domain in α-synuclein that showed a marked preference for interactions with GM3, which is a minor brain ganglioside for which the expression increases with age; the Lys34 and Tyr39 residues were shown to have critical roles in the GM3 recognition by α-synuclein (Di Pasquale et al., 2010).
