Eukaryon

Abstract

Parkinson’s disease (PD) is an incurable neurodegenerative disorder that affects four million people worldwide. Pathogenesis of PD includes aggregation of the protein α-synuclein into structures called Lewy bodies within dying midbrain dopaminergic neurons. Cell death is linked to three properties of α-synuclein: ability to misfold and aggregate, accumulate, and bind lipids. α-synuclein is a highly modified protein. A major hypothesis in the field is that some of these modifications contribute to these toxic properties of α-synuclein. Specifically, phosphorylation and nitration on α-synuclein have been linked to increasing these pathological properties. However, the link between toxicity and post-translational modifications is still not well understood. I used two yeast models to further elucidate the connection between the pathological characteristics of α-synuclein and its phosphorylation and nitration status. We report three significant findings here. 1) α-synuclein is phosphorylated in yeasts. 2) Phosphorylation regulated α-synuclein lipid binding and toxicity. 3) Nitration status affected membrane binding in budding yeast and is toxic in fission yeast.  This thesis provided multiple lines of evidence that post-translational modifications affect the properties of α-synuclein.

Introduction

The diversity of life

Life is so diverse, yet all organisms share similarities at the molecular level. For example, organisms conserve specific biochemical and metabolic pathways, such as autophagy, which occurs when a cell consumes portions of itself in order to survive (Creighton, 1994).  However, the true diversity lies within the molecular workhorses of the cell, proteins.

Made up of as few as 20 amino acids linked by peptide (amide) bonds, proteins make up only 20% of the cell, but perform most of its functions. Proteins can serve as enzymes, store and transport molecules, guide the flow of electrons in photosynthesis, act as communicators for cells via hormones, function as antibodies for the immune system, and act as channels for molecules to pass across cell membranes (Creighton, 1994). The correct structure of the protein mediates its specific function in the cell.   

An important biochemical dogma governs proteins: structure dictates function. Though mistakes in genetic replication are rare, mutations can cause three types of changes to a protein’s structure: neutral, beneficial, or toxic (Figure 1).  A neutral outcome may result from a mutation that does not affect the protein’s function.  A beneficial outcome increases optimal function which improves an organism’s chance of survival in a specific environment. Populations can then adapt and thus drive evolution. Finally, a toxic change occurs when a mutation in the genome changes the protein’s function through misfolding, resulting in either a loss of function or a gain of a new, toxic function.  In addition to genetic mutations, environmental factors can also influence the protein by causing spontaneous toxic structures to form. My thesis focuses on a group of related diseases, called neurodegenerative diseases, all linked by protein misfolding.

Neurodegenerative disease

Neurological disorders are a large class of nervous system diseases that cover anything from autism to brain cancer. Neurological disorders are prominent in the central nervous system (CNS) and include a subclass called neurodegenerative diseases. Unlike the peripheral nervous system (PNS), the CNS is unable to regenerate after injury (Purves, 2008).  Because of the inability for CNS neurons to regenerate, neurodegenerative diseases are highly studied.

Neurodegenerative disease diversity stems from the death of neurons in specific areas of the brain, characterized by the misfolding, aggregation, and plaque formation of proteins (Taylor et al., 2002). These abnormalities lead to phenotypic characteristics unique to the pathology of each disease (Sherman and Goldberg, 2001). Another common pathological feature, inflammation of tissues, is associated with many neurodegenerative diseases including Alzheimer’s disease, Parkinson’s disease, Amyotrophic Lateral Sclerosis, and multiple sclerosis (Glass et al., 2010). 

 At the molecular level, common features of neurodegenerative disease-related proteins include misfolding, decreased degradation, and post-translational modifications. Interestingly, many of these proteins are found in all brain tissues, yet they cause death in only specific cell types.  Many unanswered questions stimulate attempts to uncover why this is the case. One hypothesis is that these proteins interact with cell-specific proteins. For example, huntingtin in Huntington’s disease is found throughout the brain, but interactions with the striatal-specific protein, RHES, causes death in striatal cells (Subramaniam et al., 2009; Figure 2).

Once proteins are made in cells, they can undergo a variety of covalent modifications; these are termed post-translational modifications. Cells use post-translational modifications to dynamically regulate proteins, but the imbalance of these modifications is associated with neurodegenerative disease. For example, the increase of the post-translational modification phosphorylation on the protein tau is a pathological feature of Alzheimer’s disease (Grundke-Iqbal et al., 1986; Morrison and Hof, 1997). In Huntington’s disease, the post-ranslational modification sumoylation binds the huntingtin protein and increases toxicity (Steffan et al., 2004). The contributions of post-translation modifications in neurodegenerative disease is not well understood especially, in proteins linked to Parkinson’s disease.

Parkinson’s Disease

Parkinson’s disease (PD) is an incurable and common neurodegenerative disease afflicting four million people worldwide (Dorsey et al., 2007). Defined as a combination of motor deficits, PD symptoms include loss of postural reflexes, bradykinesia, and tremor-at-rest (Olanow and Tatton, 1999).  Although onset is more common in aging adults, with most cases occurring near 60 years old, onset 

Figure 1: The diversity of life.

Mutations in eukaryotic genes only occur once ever 107 replications. Mutations lead to one of three changes to the conformation of a protein. The structural change can either be neutral, beneficial, or toxic. A toxic change occurs in specific proteins that cause fatal neurodegenerative diseases such as Parkinson’s (α-synuclein), Prion disease (Prion), Huntington’s disease (Huntingtin) and Alzheimer’s disease (Amyloid beta-peptide) can occur at any age (Nass et al., 2008). Studies suggest the death of non-dopaminergic neurons accounts for other features of the disease, such as problems in balance, dementia, sleep, and olfaction (Braak et al., 2004).

The loss of motor function results from the degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNc). The basal ganglia, which include the SNc, the striatum, subthalamic nucleus, and the internal and external segments of the globus pallidus, constructs the motor circuitry of the brain most affected in Parkinson’s patients. The dopaminergic neurons of the SNc project onto the neurons of the striatum, releasing the neurotransmitter dopamine (Raju et al., 2006). Dopamine release excites the striatal neurons regulating the ventroanterior (VA) and ventrolateral (VL) nuclei of the thalamus, modulating movement (Figure 3A).  However, in Parkinson’s disease, a reduction of dopamine due to the degeneration of SNc neurons inhibits the VA and VL, reducing the excitation of the cortex and limiting voluntary movement (Figure 3B).  

The mechanism behind the death of dopaminergic neurons is still unknown. These neurons contain inclusions commonly known as Lewy bodies. Lewy bodies consist of the proteins ubiquitin and α–synuclein (Spillantini et al., 1998). Although both proteins are present, α–synuclein is most prominent. The exact cause of cell death is currently unknown, however, most researchers link the misfolding, aggregation, and accumulation of the protein α–synuclein to toxicity (Masliah et al., 2000; Outeiro and Lindquist, 2003; Cooper et al., 2006). Formation of α–synuclein aggregated Lewy bodies are characteristic of both forms of Parkinson’s disease: familial and sporadic.

Parkinson’s disease diversity: familial and sporadic

PD onset is either caused by environmental (sporadic) or genetic (familial) factors  (Goedert, 2001).  The exact causes of sporadic PD remain unclear, but mitochondrial stress (Langston et al., 1983; Nicklas et al., 1987; Tipton and Singer, 1993), oxidative stress (Olanow and Tatton, 1999; Dawson and Dawson, 2003), herbicides, and other deleterious chemicals (Olanow and Tatton, 1999) are all possibilities.  Although devoid of genetic mutations, sporadic PD patient brains still form Lewy bodies. This suggests a contribution of α–synuclein in all forms of PD.

Familial PD only represents about five percent of cases, worldwide. Presently, eleven genes are known to cause PD, with only eight characterized and studied. PD arises from either autosomal dominant mutations in SNCA (α–synuclein; Polymeropoulus et al., 1997), UCHL1 (Leroy et al., 1998), and LRRK2 (Funayama et al., 2007), or autosomal recessive mutations in Parkin (Matusumine et al., 1997), PINK1 (Valente et al., 2004), DJ-1 (Bonifati et al., 2003), and ATP13A2 (Najim al-Din et al., 1994; Hampshire et al., 2001).  Research using these genetic mutations has increased the knowledge in the field about Parkinson’s disease because, unlike sporadic PD, researchers can study the direct link between the mutation and disease onset.

Currently, we know that three point mutations in the α–synuclein gene cause PD. Point mutations arise from a single nucleotide change in the genetic code of a protein; the result can be a change in an amino acid within the protein, which is what occurs in these three relevant PD point mutations. If such a mutation is present in a critical shape-determining region of the protein, a point mutation can be detrimental to its function. The first familial mutant, A53T, was discovered in a Greek family. The mutation substitutes the 53rd amino acid alanine with a threonine (Polymeropoulos et al., 1997). Secondly, in A30P, an alanine to proline substitution was found in a German family at the 30th amino acid (Kruger et al., 1998). Finally, a family from Spain was found with a mutation E46K, where a glutamic acid at the 46th amino acid was altered to a lysine (Zarranz et al., 2004).

An increase in the misfolding of α-synuclein is characterized by all three mutations in vitro and in vivo (Polymeropoulos et al., 1997; Kruger et al., 1998; Conway et al., 1998; Giasson et al., 1999; Conway et al., 2001). Thus, α-synuclein may aggregate in cells for multiple reasons, including point mutations, simple protein over-expression, environmental influences, oxidative stress, or post-translational modifications (Singleton et al., 2003; Figure 4). Further understanding of α-synuclein’s properties related to PD is central to my thesis.

α –Synuclein in Parkinson’s disease

The synuclein family is a small group of proteins found in the pre-synaptic terminals of neurons and consists of three isoforms: α, β, and γ (Clayton and George, 1998). Although not fully understood, α-synuclein’s physiological role is implicated in multiple functions. α-Synuclein binds and transports fatty acids, and may regulate neurotransmitter release at synaptic terminals (Jensen et al., 1998; Sharon et al., 2001).  Most recently, α-synuclein’s role in SNARE-

Figure 2: Tissue specific death in neurodegenerative disease

Neurodegenerative disease are characterized by the loss of cells in a tissue-specific manner. This horizontal slice shows areas susceptible to tissue specific death in common neurodegenerative diseases. Alzheimer’s disease in the hippocampus, Amytrophic lateral sclerosis in the motor cortex,  and Huntington’s disease in the motor cortex. Tissue specific loss occurs in the substantia nigra caused by the death of dopaminergic neurons in Parkinson’s disease.

Figure 3: Properties of α-Synuclein (A) Diagram of α-synuclein demonstrating the location of serine-129 and serine-87, two primary sites of phosphorylation in Lewy Bodies.  In vitro experiments demonstrated that serine phosphorylation enhances aggregation of α-synuclein.  However, there is conflicting evidence as to the toxic nature of serine phosphorylation.  (B)  Diagram of α-synuclein demonstrating location of the E46K familial mutation.  The mutation occurs when a glutamic acid (E) is mutated to lysine (K) due to a genetic mutation in the α-synuclein gene.  The E46K mutant was shown to enhance membrane interaction and aggregation of α-synuclein in vitro.  (C)  Diagram of α-synuclein demonstrating the location of alanine-76 within the NAC domain of the protein.  Alanine-76 was predicted to govern the hydrophobic properties of the protein, and the A76E and A76R mutations decreased aggregation of α-synuclein in vitro-complex formation, proteins that help release vesicles in the cell, suggests disease onset correlates with the loss of SNARE-complex formation (Burre et al., 2010). Thus, α-synuclein may be a crucial protein to a neuron’s regulation and function in the brain.

α-Synuclein is a highly soluble, natively unfolded, 140 amino acid-long protein (Weinreb et al., 1996).  As a monomer, α-synuclein is unfolded, found in the cytoplasm, and binds lipid membranes. Binding occurs when the unstructured N-terminus of α-synuclein forms an α -helix (Clayton and George, 1998; Perrin et al., 2000; Kim et al., 2006). Increasing evidence suggests that α-synuclein oligomers, intermediate aggregates, are toxic because they perforate the membrane of the cell and mitochondria (Volles and Lansbury, 2002; Zakharov et al., 2007; Kim et al., 2006). Disruption of membranes by oligomeric species may increase mitochondrial dysfunction relating to toxicity in sporadic PD (Banerjee et al., 2010). When the second amino acid aspartate in the N-terminus was deleted or substituted with an alanine (D2A), reduction of toxicity and increased cytoplasmic diffusion were observed in a yeast model (Vamvaca et al., 2009). Thus, the ability to bind membranes,

especially as an oligomer, may be a pathological property of α-synuclein. Due to its flexible nature, α-synuclein tends to aggregate; resulting in the formation of Lewy bodies suggesting aggregation is a key pathological feature of PD (Caughey et al., 2003). The amino acids 71-82 make up the non-ß amyloid component region (NAC) of α-synuclein (Giasson et al., 2001).

The importance of the NAC domain’s role in α -synuclein’s ability to aggregate is illustrated by the structure of β-synuclein. β-synuclein differs structurally from α-synuclein in that it lacks the NAC region and ultimately does not aggregate (Biere et al., 2000).  α-Synuclein first oligomerizes, and then forms fibrils. Pre-oligomeric species appear to have a higher affinity for lipids compared to monomeric or aggregated α-synuclein (Volles and Lansbury, 2002; Ding et al., 2002). Due to the link between oligomeric formation and toxicity, protective mechanisms may cause Lewy body formation (Volles and Lansbury, 2002). Thus, Lewy body formation could prove to be a protective mechanism.

Finally, many different post-translational modifications are found on α-synuclein molecules when they are aggregated into Lewy bodies. These covalent modifications include: lipidation, glycosylation, ubiquitination, and nitration (Sharon et al., 2001; Shimura et al., 2001; Okochi et al., 2000; Giasson et al., 2001; Hodara et al., 2004). My thesis seeks to examine the specific roles phosphorylation and nitration play in the pathological properties of α-synuclein.

 Figure 4: Parkinson’s Disease Hypothesis

The current hypothesis in the field about dopaminergic cell death in Parkinson’s disease involves the misfolding of α-synuclein. Either by mutation or sporadically, α-synuclein misfolds and is able to self-aggregate. However, during the formation of lewy bodies, α-synuclein forms protofibrils, oligomeric species that can perforate membranes. Currently it is unknown whether the fibrils or protofibril form of α-synuclein is the toxic species. Aggregation of α-synuclein can be influenced by several factors; I am most interested in post-translational modifications.

Phosphorylation

Phosphorylation is the addition of a phosphate group, by a group of enzymes called kinases, to one of three amino acids: tyrosine, threonine, or serine. Negatively charged phosphate groups induce electrostatic forces with close ionizable amino acid side chains. These interactions rotate bonds and change the protein’s conformation. In human patients, α-synuclein is heavily phosphorylated in Lewy bodies (Fujiwara et al., 2002). α-Synuclein is phosphorylated at five residues: serine-87, serine-129, tyrosine-125, tyrosine-133, and tyrosine-136 (Fujiwara et al., 2002; Anderson et al., 2006; Paleougou et al., 2010). Currently, the degree of toxicity caused by post-translational modifications is unknown. Chen and Feany (2005), observed phosphorylation-dependent toxicity in a Drosophila model with an increase in α-synuclein aggregation.  However, in a rat model, serine-129 phosphorylation had a protective effect. (Gorbatyuk et al., 2008; Chen et al., 2009). Additionally, another lab found no difference between the phosphorylation-mimic and -deficient mutants (Mcfarland et al., 2009). Interestingly, all three of these studies used different expression vectors. Another study showed that phosphorylated α-synuclein upregulates tyrosine hydroxylase, the precursor enzyme to dopamine (Wu et al., 2011). This study contradicts previous research suggesting phosphorylation at serine-129 inhibits tyrosine hydroxylase (Lou et al., 2010). The contribution of α-synuclein phosphorylation to toxicity and Parkinson’s disease is still not well understood.

Of all the five potential phosphorylation sites, serine-87 is the only site that falls within the non-ß-amyloid component region (NAC). Thus, phosphorylation of serine-87 may be important for α-synuclein’s ability to aggregate (Paleologou et al., 2010). Membrane binding and phopshorylation is not a well-studied property of α-synuclein. One study does suggest phosphorylation of serine-87 decreases α-synuclein’s affinity for lipids; however, a connection between serine-129 phosphorylation and membrane binding has not been identified (Palelogou et al., 2010). Other residues within α-synuclein can potentially be phosphorylated, such as three tyrosine residues at positions 125, 133, and 136.  According to Chen et al. (2009), phosphorylation of tyrosine-125 inhibits the toxic properties of serine-129 phosphorylation, however, tyrosine-125 phosphorylation decreases as flies age, causing toxicity. 

Previous research in my lab observed no change in toxicity from either phosphorylation-deficient or -mimic mutants (Fiske Thesis, 2010).  In addition, Michael Fiske did observe that phosphorylation is important for aggregation and membrane binding in our fission and budding yeast models, respectively. Thus, the role phosphorylation plays in α-synuclein membrane binding and toxicity is still unclear and may be organism- or expression-dependent. Describing these properties in yeast using a high expression vector system and comparing them to previous research by Fiske will help enhance the phosphorylation field. The first goal of my study was to analyze phosphorylation and its effect on α-synuclein toxic properties using a new expression vector and to compare those results to previous results with a low expression vector system in a budding yeast model (Fiske thesis, 2010).

Nitration

Nitration is the addition of a nitrogen oxide group onto tyrosine residues and is mediated by reactive nitrogen species like nitrogen dioxide.  Possible nitration sites on α-synuclein include tyrosine-39, -125, -133, and -136. Three of the four tyrosines are found within the asparagine and glutamine rich carboxy-terminus of α-synuclein, which has been implicated in protein-protein interactions (Clayton and George, 1998). As mentioned before, oxidative stress may contribute to α-synuclein toxicity and may induce nitration. When exposed to oxidative and nitrative agents, α-synuclein forms oligomers stabilized by dityrosine cross-linking (Souza et al., 2000).  Furthermore, tyrosine-39 and tyrosine-125, when substituted with a phenylalanine in Y39F and Y125F nitration-deficient mutants, did not form aggregates and decreased toxicity. This is because they could not form dityrosine cross-linkages (Norris et al., 2003).  Additionally, cysteine is an amino acid that is easily oxidized and mimics the nitration state. Thus, when tyrosine was substituted by cysteine (Y39C, Y125C, Y133C, and Y136C), Y39C and Y125C enhanced aggregation, toxicity, and Y39C decreased membrane binding in vitro and in vivo (Zhou et al., 2002; Zhou et al., 2008).  No additional aggregation or toxicity was observed by Y133C or Y136C, suggesting a less important role for these residues in α-synuclein’s pathological properties. Inducing nitration through oxidative stress causes pathology and symptoms similar to that of PD in a mouse model, as well (Yu et al., 2010).

Most evidence in the field suggests a pathological role of nitrated tyrosine-39 and tyrosine-125. However, there are only limited studies suggesting this relationship in vivo, and there is a lack of research observing the role of nitration with membrane binding.   Since nitration of α-synuclein has yet to be studied in yeast, this will help provide additional insight into nitration of α-synuclein in vivo. The second goal of my thesis was to evaluate the role of nitration and the importance of four individual tyrosine residues in a budding and fission yeast model.

Yeast model

There are multiple models used to study Parkinson’s disease, such as transgenic mice, Drosophila, and C. Elegans. Each model recapitulates different aspects of PD. Overexpression of WT α-synuclein and A53T mutant leads to the formation of Lewy body-like inclusions and PD-like symptoms in mice (Giasson et al., 2001; Mashliah et al., 2001). Drosophila also models PD successfully; in flies, when A53T or A30P is expressed, age-dependent PD occurs selectively to dopaminergic neurons (Feany and Bender, 2000). 

Figure 5: Experimental Hypothesis

(A) Diagram of α-synuclein demonstrating the location of serine-129 and serine-87, two primary sites of phosphorylation in Lewy bodies. We hypothesized that phosphorylation will regulate toxicity, increase aggregation and regulate membrane binding in yeast  (B)  Diagram of α-synuclein demonstrating the location of tyrosine-39, tyrosine-125, tyrosine-133, and tyrosine-136, four sites nitrated in Lewy bodies. We hypothesized that nitration will increase aggregation, decrease membrane binding, and increase toxicity in our yeast models. 

Initially, yeast would seem to be an unsuitable organism to model PD. However, it models many of α-synuclein’s physiological aspects such as degradation by the proteasome and autophagy pathways, and was the first model to show that α-synuclein binds to lipid droplets in Lewy bodies (Outeiro and Lindquist, 2003; Zabrocki et al., 2005; Sharma et al., 2006).  Several labs use yeast to model PD because yeast are relatively inexpensive, have a short lifespan, both budding and fission yeast genomes are mapped, deletion strains are available, and humans and yeast share many homologous genes for protein synthesis, folding and degradation (Outiero and Lindquist, 2003; Willingham et al., 2003; Zabrocki et al. 2005; Dixon et al. 2005; Sharma et al. 2006). For years, budding yeast have served as effective models for neurodegenerative diseases, such as prion disease, ALS, and Huntington’s disease (Ma and Lindquist, 2002; Krobitsch and Lindquist, 2000; Corson et al., 1998).   In our lab and for my thesis, I used two yeast models as tools: Saccharomyces cerevisiae (budding yeast) and Schizosaccharomyces pombe (fission yeast). Our lab helped pioneer a fission yeast model for studying PD and neurodegenerative diseases (Brandis et al., 2006). Interestingly, our fission yeast models a key pathological feature of PD, aggregation and toxicity of α-synuclein due to WT, A53T, A30P, and increased toxicity in E46K (Brandis et al., 2006; Fiske Thesis, 2010).

The yeast genome does not contain an α-synuclein homologue. Currently, many techniques are used to express α-synuclein in yeast. The gene can either be incorporated into the yeast genome (viral vector) or a plasmid vector in the cytoplasm of the yeast (Outeiro and Lindquist, 2003). Outeiro and Lindquist (2003) showed that when one copy of the gene was present, no toxicity resulted. However, when two copies or a high copy 2-micron vector was used, α-synuclein dependent toxicity was observed (Outeiro and Lindquist, 2003). Although often thought as aggregated α-synuclein in the cytoplasm of the yeast, further investigation suggests that many of these are not true aggregates, but accumulation of vesicular structures from intracellular trafficking defects (Gitler et al., 2008; Soper et al., 2008; Dixon et al., 2005; Flower et al., 2005). Our lab used the same 2-micron vector to express α-synuclein in budding and fission yeast (Sharma et al., 2006; Brandis et al., 2006). Additionally, adding stresses, such as disrupting the proteasome or oxidative stress, increases α-synuclein-dependent toxicity in our budding yeast model (Sharma et al., 2006). 

Thus, using both strains as tools, we are able to model the three prominent pathological features of PD, lipid binding, aggregation, and toxicity. For my thesis I used these two yeast models to observe the relationship between post-translation modifications and how they alter the properties of α-synuclein.