**3. Alpha-synuclein pathology in PD and α-synucleinopathies**

The term α-synucleinopathies is usually employed to define a group of neurodegenerative disorders that show common pathologic proteinaceous accumulation of α-synuclein aggregates. In these diseases, α-synuclein aggregates are deposited in selective vulnerable populations of neuronal and glial cells (Goedert, 1999;Spillantini and Goedert, 2000;Galvin et al., 2001;Trojanowski and Lee, 2003). From a clinical point of view, α-synucleinopathies include symptomatically heterogeneous disorders, among them LB-associated diseased such as PD, DLB, multiple system atrophy (MSA) LB dysphagia as well as neurodegeneration with brain iron accumulation type I, pure autonomic failure, and the LB variant of Alzheimer's disease (Uversky, 2007).

However, since α-synuclein has been found to be the main component of LB and LN in the PD brain (Spillantini et al., 1997), and it has been described that mutations in the gene encoding αsynuclein (SNCA gene) are responsible for the onset of familiar forms of PD (Polymeropoulos et al., 1997), in the last decade, much effort has been devoted to the characterization of the molecular basis of α-synuclein-related neuronal degeneration in PD and DLB.

Missense mutations in the gene encoding α-synuclein, as well as duplication and triplication of the locus containing the SNCA gene, have been identified in familial forms of PD (Tofaris and Spillantini, 2007;Lee and Trojanowski, 2006). However, familial forms of PD linked to missense mutations are extremely rare and have different clinical and histopathological features, although they may provide insights into pathogenic mechanisms leading to LB formation. Fibril formation is accelerated by the A53T mutated α-synuclein with respect to A30P-mutated or wild type form of the protein. However, in fibril-producing conditions, the A30P monomer is consumed more rapidly than the wild type monomer, although less rapidly than the A53T monomer, thus indicating that both mutations trigger the accelerated formation of pre-fibrillar α-synuclein oligomers (Conway et al., 2000). A30Pmutated α-synuclein display a reduced binding to vesicles, thus increasing the neuropathology and the bioavailability of the protein for aberrant interactions (Choi et al., 2004). The E46K mutation significantly increases the binding of α-synuclein to negatively charged liposomes and enhances the rate of filament assembly comparably to the A53T mutation (Greenbaum et al., 2005). The deleterious effects of point mutations and the effect of high expression levels have been investigated "in vivo" in transgenic and viral infected animals. Transgenic mice where the expression of A53T α-synuclein has been driven by the mouse prion protein promoter (mPrP) showed accumulation of α-synuclein in cell bodies and dystrophic neuritis as well as motor impaiment (Neumann et al., 2002). However, these transgenic mice showed an extensive α-synuclein pathology in motor neurons and axons which may account for their behavioural phenotype. Other studies showed that transgenic overexpression of the wild type form of α-synuclein under the guidance of the mPrP was not associated to protein accumulation or behavioural deficits. Conversely, transgenic mice overexpressing the A30P mutant α-synuclein under the guidance of the Thy1 promoter developed age dependent LB-like changes such as proteinase K-resistant protein aggregates,

Finally, recent evidence showed that cells from transgenic mice expressing the truncated form of α-synuclein are more susceptible to environmental conditions and that overexpression of the wild type from of the protein in neuronal progenitor cells affect their fate of differentiation, thus supporting the notion that full length α-synuclein is involved in dopaminergic cell differentiation and survival (Michell et al., 2007;Schneider et al., 2007).

The term α-synucleinopathies is usually employed to define a group of neurodegenerative disorders that show common pathologic proteinaceous accumulation of α-synuclein aggregates. In these diseases, α-synuclein aggregates are deposited in selective vulnerable populations of neuronal and glial cells (Goedert, 1999;Spillantini and Goedert, 2000;Galvin et al., 2001;Trojanowski and Lee, 2003). From a clinical point of view, α-synucleinopathies include symptomatically heterogeneous disorders, among them LB-associated diseased such as PD, DLB, multiple system atrophy (MSA) LB dysphagia as well as neurodegeneration with brain iron accumulation type I, pure autonomic failure, and the LB variant of

However, since α-synuclein has been found to be the main component of LB and LN in the PD brain (Spillantini et al., 1997), and it has been described that mutations in the gene encoding αsynuclein (SNCA gene) are responsible for the onset of familiar forms of PD (Polymeropoulos et al., 1997), in the last decade, much effort has been devoted to the characterization of the

Missense mutations in the gene encoding α-synuclein, as well as duplication and triplication of the locus containing the SNCA gene, have been identified in familial forms of PD (Tofaris and Spillantini, 2007;Lee and Trojanowski, 2006). However, familial forms of PD linked to missense mutations are extremely rare and have different clinical and histopathological features, although they may provide insights into pathogenic mechanisms leading to LB formation. Fibril formation is accelerated by the A53T mutated α-synuclein with respect to A30P-mutated or wild type form of the protein. However, in fibril-producing conditions, the A30P monomer is consumed more rapidly than the wild type monomer, although less rapidly than the A53T monomer, thus indicating that both mutations trigger the accelerated formation of pre-fibrillar α-synuclein oligomers (Conway et al., 2000). A30Pmutated α-synuclein display a reduced binding to vesicles, thus increasing the neuropathology and the bioavailability of the protein for aberrant interactions (Choi et al., 2004). The E46K mutation significantly increases the binding of α-synuclein to negatively charged liposomes and enhances the rate of filament assembly comparably to the A53T mutation (Greenbaum et al., 2005). The deleterious effects of point mutations and the effect of high expression levels have been investigated "in vivo" in transgenic and viral infected animals. Transgenic mice where the expression of A53T α-synuclein has been driven by the mouse prion protein promoter (mPrP) showed accumulation of α-synuclein in cell bodies and dystrophic neuritis as well as motor impaiment (Neumann et al., 2002). However, these transgenic mice showed an extensive α-synuclein pathology in motor neurons and axons which may account for their behavioural phenotype. Other studies showed that transgenic overexpression of the wild type form of α-synuclein under the guidance of the mPrP was not associated to protein accumulation or behavioural deficits. Conversely, transgenic mice overexpressing the A30P mutant α-synuclein under the guidance of the Thy1 promoter developed age dependent LB-like changes such as proteinase K-resistant protein aggregates,

**3. Alpha-synuclein pathology in PD and α-synucleinopathies** 

molecular basis of α-synuclein-related neuronal degeneration in PD and DLB.

Alzheimer's disease (Uversky, 2007).

neuritic pathology and formation of some argyrophilic and thioflavin S-positive α-synuclein inclusions. Remarkably, although mPrP but not Thy1 promoter drives high expression of the transgene in the substantia nigra neurons, the tyrosine hydroxylase (TH)–positive neurons of this area in the mPrP-driven transgenic mice were devoid of α-synclein aggregates, and didn't show loss of striatal dopamine or DAT (Giasson et al., 2002;Lee et al., 2002). In line with these findings another study showed that TH-driven expression of wt or mutant α-synuclein didn't result in the formation of intracellular α-synuclein aggregates.

From these studies, it became evident that adeno- or lentiviral-mediated expression of αsynuclein may represent a powerful tool to induce the selective expression of the protein in dopaminergic neurons of the substantia nigra of adult rodents (Kirik et al., 2002;Klein et al., 2002;Lo et al., 2002). Indeed, overexpression of either wild type or mutant form of the protein led to cellular and axonal pathology associated with loss of nigral neurons, reduction in striatal dopamine levels and motor deficits without the formation of fibrillary inclusions (Lo et al., 2002), thus contradicting evidences from studies on transgenic animals. These observations suggest that dopaminergic neurons are vulnerable to high levels of human α-synuclein and that wild type or mutant form of the protein are both toxic. Nonetheless, it may be taken into account that these discrepancies may be related to the high rate of transgene expression that can be achieved by either technology or by the high toxicity which is induced by gene transfection in viral models. To date, neither transgenic nor viral-mediated rodent models expressing wild type or mutated α-synuclein show fibrillary inclusions. Thus the mechanisms by which wild type human α-synuclein assembles in LB in the substantia nigra of PD patients hasn't been clarified by using these models. Nonetheless, filamentous α-synuclein inclusions have been observed in a *Drosophila* model in association with loss of dopaminergic cells and locomotor defects (Feany and Bender, 2000).

Although studies of genetic mutations in α-synuclein helped in the understanding some of the function and pathogenic properties of α-synuclein, they only account for a very small proportion of PD cases. Indeed, more than 90% of PD cases are sporadic, thus characterized by the accumulation of insoluble fibrils of WT α-synuclein (Spillantini et al., 1997;Spillantini et al., 1998). Therefore, much effort has been made in order to understand what are the alterations which convert wild type α-synuclein to a toxic species. It has been shown that wild type α-synuclein aggregates form fibrils identical to those isolated from disease brains, even though the rate of fibril formation is slower than that of the mutant form (Serpell et al., 2000;Conway et al., 1998). To date, several post-translational modifications of α-synuclein can alter its biophysical properties. Thus, it has been hypothesized that these modifications are implicated in the induction of the fibrillation process (Oueslati et al., 2010). These studies indicate that α-synuclein has numerous potential sites for post-translational modifications such as phosphorylation, tyrosine nitration or protein cleavage. In transfected cells, it is constitutively phosphorylated at serine residues 87 and 129, with the latter being the predominant site (Okochi et al., 2000). Residue 129 in α-synuclein lies a consensus sequence for casein kinase 1, which is also present in β- and γ-synuclein, and casein kinase 1 and 2 have been found to phosphorylate this site. Alpha-synuclein can be phosphorylated also by several G-protein-coupled receptor kinases, events which reduce the ability of the protein to interact with phospholipids and PLD2 (Pronin et al., 2000). Tyrosine kinase 72syk can phosphorylate the tyrosine residues in the carboxy-terminus of the protein both *in vitro* and in CHO cells (Negro et al., 2002). The specific phosphorylation of serine 129 in the Cterminal region can decrease the ability of this portion to prevent fibril formation (Fujiwara

Targeting α-Synuclein-Related

aggregates (Trostchansky et al., 2006) .

synuclein cannot be excluded at present.

lipidic and protein molecules.

**4. The** α**-synuclein synaptic proteome** 

Synaptic Pathology: Novel Clues for Parkinson's Disease Therapy 145

found in the brains of patients with PD and recombinant α-synuclein forms multimers "in vitro" upon exposure to vesicles polyunsaturated fatty acids (PUFA) acyl groups (Perrin et al., 2001). Moreover, exposure of mesencephalic neurons to PUFA increases α-synuclein oligomerisation "in vivo" (Sharon et al., 2003). These oligomers precede the formation of insoluble fibrillar aggregates and can bind and permeabilize membrane bilayers through electrostatic and hydrophobic interactions (Volles et al., 2001;Zhu et al., 2003). Conversely, more recent studies have shown that association of α-synuclein with biological membranes can also protect the protein from oxidation and nitrosylation, decreasing the formation of

Finally, it has been found that overexpression of wild type and mutant α-synuclein in cells isolated from transgenic mice disrupted the vesicular pH and increased cytosolic catechol species, which in turn could trigger oxidative damage (Mosharov et al., 2006). Dopamine, on the other hand, has been found to stabilize oligomeric intermediates (Conway et al., 2001;Mazzulli et al., 2006) which can further disrupt the integrity of synaptic vesicles, thus initiating a vicious circle that eventually leads to aggregation and cell death. Dopamine overload can exacerbate the formation of insoluble α-synuclein-positive inclusions and cell death induced by nutrient starvation (Bellucci et al., 2008). Overexpression of wild type, A53T and A30P α-synuclein in human dopaminergic neurons led to 2-2.5-fold increase in apoptosis (Xu et al., 2002;Zhou et al., 2002). Thus, from these findings, it seems that dysregulation of dopamine homeostasis may underlie the vulnerability of dopaminergic neurons in PD. However, the possibility that post-translational modifications during disease process may at least partly be a secondary phenomena resulting from aggregation of α-

Several lines of evidence indicate that α-synuclein is critically involved in the regulation of synaptic functions (Fortin et al., 2010;Bellani et al., 2010;Sousa et al., 2009;Tofaris and Spillantini, 2007). In particular, it has been shown that α-synuclein is loosely associated with the distal pool of synaptic vesicles (Kahle et al., 2000;Lee et al., 2008;Zhang et al., 2008) and with the lipid rafts of the plasma membrane (Fortin et al., 2004), thus indicating that the protein is able to associate with membranes, as confirmed by other studies (Chandra et al., 2003;Davidson et al., 1998;Eliezer et al., 2001;Jo et al., 2000). Nonetheless, it seems that the distribution of α-synuclein at synapses is very dynamic and dependent upon neuronal stimulation, with rapid exchanges taking place among neighbouring synapses (Fortin et al., 2005). The critical importance of α-synuclein in the modulation of synaptic functions is confirmed by numerous studies shading light upon the α-synuclein synaptic proteome. Numerous observations indicate that α-synuclein is crucially involved in the regulation of synaptic functions as it interacts with- and modulates key synaptic components including

Yet, in 2003 Sharon and colleagues (2003) elegantly showed that α-synuclein accumulation can influence cellular and brain phospholipids levels, including certain PUFA, thus affecting membrane fluidity and synaptic membrane trafficking. Later on, Jo and coauthors (2004) found that wild type α-synuclein does not penetrate to the fatty acyl chains and in isolated synaptosomes, while the A53T mutated form of the protein binds to synaptosomal membranes, increases lipid headgroup packing, induces subtle changes in the lipid interfacial space and decreases the fluidity of the fatty acyl chains. However, a recent

et al., 2002). Other post translational modifications, such as tyrosine nitration of α-synuclein, which have been reported in LB, seem to promote fibril formation (Giasson et al., 2000;Hodara et al., 2004). Indeed, it has been shown that the specific nitration of the tyrosine at position 39 in the N-terminal region of α-synuclein decreases its ability to bind to lipid vesicles and shows the efficiency of both the proteasome and calpain I to degrade the protein, thus fostering its accumulation. Thus, collectively it seems that modifications of specific regions of α-synuclein may differentially affect its tendency toward aggregation. In particular, modifications in the C-terminal part of α-synuclein are likely. This is confirmed by the fact that modifications in this region, such as oxidation, nitration and phosphorylation, may influence the propensity of α-synuclein to aggregate "in vivo" in a similar way to truncation (Giasson et al., 2000;Fujiwara et al., 2002;Hashimoto et al., 1999). Similarly, polyamines, dopamine or other positively charged molecules can interact with the C-terminal portion of α-synuclein and promote its aggregation (Antony et al., 2003;Fernandez et al., 2004;Goers et al., 2003;Norris et al., 2005).

Remarkably, it has been reported that LB are rich of C-terminally-truncated α-synuclein (Campbell et al., 2001;Baba et al., 1998) which derives from proteolitic cleavage operated by calpain I (Mishizen-Eberz et al., 2005;Dufty et al., 2007). Several studies demonstrated that C-terminally truncated α-synuclein has an high tendency towards aggregation both "in vitro" and "in vivo" (Crowther et al., 1998;Tofaris et al., 2003;Tofaris et al., 2006;Bellucci et al., 2011) thus indicating that these α-synuclein species could play a role in inducing LB formation. On this line, other findings (Murray et al., 2003) indicate that, while the middle region of α-synuclein forms the core of filaments, the negative charges in the carboxyterminal part of α-synuclein counteract its aggregation, thus implying that a lack or biochemical modifications of this region may increase the propensity of the protein to aggregate.

It is noteworthy that abnormal protein degradation is another process which has been involved in the formation of LB. Indeed, accumulation of misfolded proteins can overwhelm the ubiquitin-proteasome system, leading to aberrant degradation (Bence et al., 2001;Venkatraman et al., 2004) and binding of α-synuclein filaments and soluble oligomers to the proteasome results in marked inhibition of chymotrypsin-like hydrolytic activity (Lindersson et al., 2004). Monomeric WT α-synuclein can be directly degraded by the 20S proteasome in a ubiquitin-independent manner (Tofaris et al., 2003) a process that is slowed down by nitrosylation of monomeric α-synuclein (Hodara et al., 2004) and that can lead to generation of incompletely degraded, C-terminally truncated α-synuclein species (Liu et al., 2005). Furthermore, it has been described that in LB diseases a modified form of α-synuclein of 22-24 kDa is the substrate of predominantly mono- or di-unbiquitination (Tofaris et al., 2003). Thus, ubiquitin-dependent degradation is probably not involved in the physiological degradation of α-synuclein but rather represents a disease specific pathway activated as a last attempt to unfold and/or degrade misfolded proteins either through the 26S proteasome via poly-ubiquitination, or the lysosome by mono-ubiquitination.

This notwithstandings, it has been showed that the direct expression of the chaperone Hsp70 prevents dopaminergic neuronal loss associated with α-synuclein toxicity in Drosophila (Auluck et al., 2002) like the overexpression of the co-chaperone carboxyl-terminus of Hsp70 interacting protein (CHIP) decreases the levels and the formations of intracellular α-synuclein inclusions via the proteasome and lysosome systems (Shin et al., 2005).

Interaction with lipid membranes could also play a role in the conversion of wild type αsynuclein to a pathogenic protein, as detergent stable oligomers of α-synuclein have been

et al., 2002). Other post translational modifications, such as tyrosine nitration of α-synuclein, which have been reported in LB, seem to promote fibril formation (Giasson et al., 2000;Hodara et al., 2004). Indeed, it has been shown that the specific nitration of the tyrosine at position 39 in the N-terminal region of α-synuclein decreases its ability to bind to lipid vesicles and shows the efficiency of both the proteasome and calpain I to degrade the protein, thus fostering its accumulation. Thus, collectively it seems that modifications of specific regions of α-synuclein may differentially affect its tendency toward aggregation. In particular, modifications in the C-terminal part of α-synuclein are likely. This is confirmed by the fact that modifications in this region, such as oxidation, nitration and phosphorylation, may influence the propensity of α-synuclein to aggregate "in vivo" in a similar way to truncation (Giasson et al., 2000;Fujiwara et al., 2002;Hashimoto et al., 1999). Similarly, polyamines, dopamine or other positively charged molecules can interact with the C-terminal portion of α-synuclein and promote its aggregation (Antony et al.,

Remarkably, it has been reported that LB are rich of C-terminally-truncated α-synuclein (Campbell et al., 2001;Baba et al., 1998) which derives from proteolitic cleavage operated by calpain I (Mishizen-Eberz et al., 2005;Dufty et al., 2007). Several studies demonstrated that C-terminally truncated α-synuclein has an high tendency towards aggregation both "in vitro" and "in vivo" (Crowther et al., 1998;Tofaris et al., 2003;Tofaris et al., 2006;Bellucci et al., 2011) thus indicating that these α-synuclein species could play a role in inducing LB formation. On this line, other findings (Murray et al., 2003) indicate that, while the middle region of α-synuclein forms the core of filaments, the negative charges in the carboxyterminal part of α-synuclein counteract its aggregation, thus implying that a lack or biochemical modifications of this region may increase the propensity of the protein to

It is noteworthy that abnormal protein degradation is another process which has been involved in the formation of LB. Indeed, accumulation of misfolded proteins can overwhelm the ubiquitin-proteasome system, leading to aberrant degradation (Bence et al., 2001;Venkatraman et al., 2004) and binding of α-synuclein filaments and soluble oligomers to the proteasome results in marked inhibition of chymotrypsin-like hydrolytic activity (Lindersson et al., 2004). Monomeric WT α-synuclein can be directly degraded by the 20S proteasome in a ubiquitin-independent manner (Tofaris et al., 2003) a process that is slowed down by nitrosylation of monomeric α-synuclein (Hodara et al., 2004) and that can lead to generation of incompletely degraded, C-terminally truncated α-synuclein species (Liu et al., 2005). Furthermore, it has been described that in LB diseases a modified form of α-synuclein of 22-24 kDa is the substrate of predominantly mono- or di-unbiquitination (Tofaris et al., 2003). Thus, ubiquitin-dependent degradation is probably not involved in the physiological degradation of α-synuclein but rather represents a disease specific pathway activated as a last attempt to unfold and/or degrade misfolded proteins either through the 26S

proteasome via poly-ubiquitination, or the lysosome by mono-ubiquitination.

inclusions via the proteasome and lysosome systems (Shin et al., 2005).

This notwithstandings, it has been showed that the direct expression of the chaperone Hsp70 prevents dopaminergic neuronal loss associated with α-synuclein toxicity in Drosophila (Auluck et al., 2002) like the overexpression of the co-chaperone carboxyl-terminus of Hsp70 interacting protein (CHIP) decreases the levels and the formations of intracellular α-synuclein

Interaction with lipid membranes could also play a role in the conversion of wild type αsynuclein to a pathogenic protein, as detergent stable oligomers of α-synuclein have been

2003;Fernandez et al., 2004;Goers et al., 2003;Norris et al., 2005).

aggregate.

found in the brains of patients with PD and recombinant α-synuclein forms multimers "in vitro" upon exposure to vesicles polyunsaturated fatty acids (PUFA) acyl groups (Perrin et al., 2001). Moreover, exposure of mesencephalic neurons to PUFA increases α-synuclein oligomerisation "in vivo" (Sharon et al., 2003). These oligomers precede the formation of insoluble fibrillar aggregates and can bind and permeabilize membrane bilayers through electrostatic and hydrophobic interactions (Volles et al., 2001;Zhu et al., 2003). Conversely, more recent studies have shown that association of α-synuclein with biological membranes can also protect the protein from oxidation and nitrosylation, decreasing the formation of aggregates (Trostchansky et al., 2006) .

Finally, it has been found that overexpression of wild type and mutant α-synuclein in cells isolated from transgenic mice disrupted the vesicular pH and increased cytosolic catechol species, which in turn could trigger oxidative damage (Mosharov et al., 2006). Dopamine, on the other hand, has been found to stabilize oligomeric intermediates (Conway et al., 2001;Mazzulli et al., 2006) which can further disrupt the integrity of synaptic vesicles, thus initiating a vicious circle that eventually leads to aggregation and cell death. Dopamine overload can exacerbate the formation of insoluble α-synuclein-positive inclusions and cell death induced by nutrient starvation (Bellucci et al., 2008). Overexpression of wild type, A53T and A30P α-synuclein in human dopaminergic neurons led to 2-2.5-fold increase in apoptosis (Xu et al., 2002;Zhou et al., 2002). Thus, from these findings, it seems that dysregulation of dopamine homeostasis may underlie the vulnerability of dopaminergic neurons in PD. However, the possibility that post-translational modifications during disease process may at least partly be a secondary phenomena resulting from aggregation of αsynuclein cannot be excluded at present.
