- <div style="background-image:url(/live/image/gid/32/width/1600/height/300/crop/1/41839_V14Cover_Lynch_Artwork.2.rev.1520229233.png)"/>
Combinatorial Impact of α-Synuclein Post-Translational Modifications in Yeast
Lake Forest College
Lake Forest, Illinois 60045
Parkinson’s Disease is a devastating neurodegenerative disorder in which loss of voluntary movement occurs due to accumulation of misfolded α-synuclein in the midbrain. In diseased individuals, α-synuclein is covalently modified by multiple chemical groups. The pathological contributions of some modifications are well-studied (phosphorylation and nitration), but those of more recently identified ones (sumoylation, acetylation, and glycation) remain scant. Furthermore, how these modifications in combination influence α-synuclein’s properties is unknown. We used a budding yeast model of Parkinson’s Disease to investigate the effects of these modifications on α-synuclein’s localization, expression, and toxicity. We found that: 1) SUMOylation and acetylation are protective, while phosphorylation and glycation are detrimental; 2) SUMOylation and phosphorylation, as well as acetylation and glycation, counteract each other; 3) and glycation alters the toxicity of several α-synuclein mutants that cause familial Parkinson’s Disease. This study suggests that post-translational modifications are a viable target for manipulating α-synuclein’s pathological properties.
ANYONE TOUCHED BY NEURODEGENERATIVE DISEASES
About fifteen years ago, my grandfather was diagnosed with an unspecified neurodegenerative condition. Doctors never determined with certainty whether he had Alzheimer’s Disease, Parkinson’s Disease, or a combination of the two. Regardless of the diagnosis, I remember the terrible emotional toll that the disease took on our family. I was too young to remember my grandfather when he was healthy. I do, however, know that the end of his life saw him bed-ridden, unable to take care of himself. Such a story is too common in today’s society. The devastating grip of neurodegeneration spares no one. I believe that the future holds promise in the fight against neurodegenerative diseases. Through scientific research and public outreach, we can as a society, join the fight against these illnesses and work to end the suffering and costs associated with them.
This work would not have been possible without the great support I have received from my parents. They have been by my side through thick and thin and have always been willing to give me sound advice. I would also like to acknowledge the wonderful help and mentorship from Dr. DebBurman. The solid training that his lab has provided, along with the numerous opportunities to present my research at national and international conferences, have served as constant reminders that science is an ever-changing frontier in which it is hard to hit a dead end. I am also grateful for the mentorship and advice from all of the members on my senior thesis committee: Dr. Jean-Marie Maddux, Dr. Michael Kash, and Dr. Erica Schultz. Thank you to all of the members of Dr. DebBurman’s lab, especially Rosemary Thomas ’19, Alex Roman ’16, and Estella Tcaturian ’21, who worked closely with me on this project. Their energy, friendship, and enthusiasm has kept me going and has always put a fresh perspective on the research.
LIST OF ABBREVIATIONS
A30P: Aggressive early-onset mutant of α-synuclein; alanine 30 replaced by proline
A53E: Aggressive early-onset mutant of α-synuclein; alanine 53 replaced by glutamate
A53T: Aggressive early-onset mutant of α-synuclein; alanine 53 replaced by threonine
DLB: dementia with Lewy Bodies
E46K: Aggressive early-onset mutant of α-synuclein; glutamate 46 replaced by lysine
G51D: Aggressive early-onset mutant of α-synuclein; glycine 51 replaced by aspartate
GFP: Green Fluorescent Protein
H50Q: Aggressive early-onset mutant of α-synuclein; histidine 50 replaced by glutamine
K6A/K10A: double acetylation-blocking and glycation-blocking mutation (K6A/K10A)
K6Q/K10Q: double acetylation-mimicking mutation (K6Q/K10Q)
K6R/K10R: double acetylation-blocking mutation
K96R/K102R: mutant blocking SUMOylation
MSA: Multiple System Atrophy
NDD: neurodegenerative diseases
PD: Parkinson’s Disease
PTM: post-translational modification
S87A/S129A – SUMO: Double phosphorylation-blocking and SUMOylation blocking mutation
S87D/S129D – SUMO: Double phosphorylation-mimicking and SUMOylation blocking mutation
WT: Wild-type (unmutated version)
Part 1: Proteins in Health and Disease
According to a recently-developed hypothesis in molecular biology, RNA initially supported all life on Earth (Alberts, 1998). It was the first molecule capable of both storing genetic information and expressing it to execute the complicated functions of cells (Alberts, 1998). Across evolutionary time, these vital capabilities passed to two more-stable molecules: DNA and proteins, respectively. Thus, it is reasonable to assume that DNA and proteins hold the key to disease. The central dogma of biology states that cells transcribe the genetic code (composed of DNA) to RNA, serving as a template for later protein translation (Figure 1a). DNA stores genetic information, while proteins are the functional components of cells, resulting in all of life’s observable characteristics. By adopting diverse conformations, proteins act as molecular machines and contribute to the phenotypic diversity of life (Alberts, 1998). In fact, it is not an exaggeration to say that the great versatility of proteins creates the division of life into five kingdoms (Figure 1b). All human organs are made up of proteins. While we marvel at the unique qualities of the brain, its dependence on proteins makes it indistinguishable from other organs submicroscopically.
The structure of a protein determines its function. The amino acid sequence (primary structure) dictates local interactions (secondary structure), which come together to give a stable low-energy protein (tertiary structure). A disruption at any one of these levels of organization leads to a non-functional protein that folds incorrectly. The conformation of each protein is suited to its role in the organism. For example, hemoglobin is made of four subunits, each accommodating a heme group and iron, allowing the protein to carry oxygen (Marengo-Rowe, 2006). A disruption of this structure leads to a loss of function and ultimately disease (sickle-cell anemia). Variations
in protein folding can result from many mechanisms, but the most common ones are mutation and post-translational modification. These changes are not always detrimental. While some contribute to disease, others are essential and may contribute to more efficient protein functioning (Figure 2). In a mutation, a change in the DNA leads to a change in the amino acid sequence of the protein. In a post-translational modification, once protein synthesis is complete, a molecule or functional group is covalently attached, changing the protein’s structure and function. These mechanisms have been implicated in many evolutionary pathways but also in many diseases.
Neurodegenerative diseases are closely associated with abnormal proteins. These devastating illnesses are characterized by a misfolded protein coming out of solution and aggregating in the brain. The area of the brain affected and the type of protein that misfolds together determine the symptoms characterizing a particular disease. For example, in Alzheimer’s Disease, cholinergic neurons accumulate aggregates of beta-amyloid and tau (Mufson et al., 2008). As these aggregates lead to the degeneration of the hippocampus and cerebral cortex, it is not surprising that memory loss and personality change become the primary symptoms (Sabuncu et al., 2011). In several different types of spinocerebellar ataxias, aggregates in the cerebellum or brainstem lead to decreased coordination, balance problems, and speech impediments (Seidel et al., 2017; Bird, 2016). Another large family of neurodegenerative diseases is the synucleinopathies. This category of diseases can strike anywhere in the brain (and can afflict either neurons or glia). One common feature of all forms is the pathology associated with the protein α-synuclein. α-Synuclein misfolds and creates abnormal cytoplasmic aggregates (McCann et al., 2014). These neurodegenerative diseases, along with some other representative examples, are showcased in Figure 3a.
In many neurodegenerative diseases, levels of post-translational modifications are altered from the expected distribution in the healthy population. As stated above, the effects of these modifications vary on a case-by-case basis, and they are not always detrimental. Phosphorylation (the attachment of phosphate groups to serine, tyrosine, or threonine amino acids) is one of the most extensively studied post-translational modifications. In Huntington’s disease, phosphorylation of huntingtin is protective and leads to faster degradation of the aberrant protein (Warby et al., 2005). The same modification can be harmful in other cases. For example, in spinocerebellar ataxia type 1, phosphorylation of ataxin-1 is associated with increased aggregate stability and exacerbated pathology (Park et al., 2013). The fact that modifications of key proteins have been implicated in neurodegenerative diseases has raised many questions. What is the role of these modifications? Do they occur before or after the protein misfolds and aggregates? How do they change the structure and function of the implicated protein? Can they serve as targets for treatment? Molecular biology and neuroscience are currently far from definitively answering these questions, but researchers are making steady progress. Given the promising results seen so far in post-translational modification research, this thesis aims to contribute to knowledge about how post-translational modifications affect the progression of one devastating neurodegenerative illness – Parkinson’s Disease.
Part 2: Parkinson’s Disease on the Macroscopic and Microscopic Levels
Parkinson’s Disease (PD) belongs to the broader family of synucleinopathies. The main examples of these illnesses (and the areas of the brain that they affect) are presented in Figure 3b. Synucleinopathies are diseases in which the protein α-synuclein misfolds and comes out of solution. The neurons (or glia) that are afflicted usually die, leading to a loss of critical processes in the brain. Dementia with Lewy Bodies (DLB), Multiple System Atrophy (MSA), and PD are the primary members of the synucleinopathy family (Barker and Williams-Gray, 2016). In all of these cases, α-synuclein forms insoluble cytoplasmic aggregates, and as brain cells die, key cognitive or motor functions deteriorate. In Dementia with Lewy bodies, α-synuclein pathology affects the brainstem, limbic system, and cortex (Garcia-Espacia et al., 2017). The aggregation of this protein in structures called Lewy bodies leads to cognitive disturbances, motor problems, and emotional instability. Another neurodegenerative disease, Multiple System Atrophy, is unusual in that the brain’s insulating agents (myelin-producing oligodendrocytes) degenerate and exhibit α-synuclein pathology (Wakabayashi and Takahashi, 2006). As a result of the death of these glial cells (and the subsequent loss of midbrain neurons), patients experience autonomic disturbances and coordination/motor problems. PD also falls under the synucleinopathy umbrella. Here, the death of neurons producing dopamine in the substantia nigra leads to difficulties in initiating voluntary movements (Mazzoni, Shabbott, and Cortes, 2012). The different synucleinopathy categories, and the areas of the brain that they affect, are highlighted in Figure 3b. Given this context of neurodegeneration, it is worth exploring how in PD, the symptoms of sufferers directly arise from the loss of dopaminergic neurons.
PD is a hypokinetic disorder, meaning that a patient’s movement abilities diminish as the disease progresses (Fahn, 2008). A significant portion of the aging population (2-3%) is afflicted with this disease (Tysens et al., 2017). Symptoms have a gradual onset, eventually affecting the entire scope of movement. Presentation is usually unilateral, and early signs include a resting tremor, stooped posture, and shuffling gait (Rizek et al., 2016). The overall slowing down of movement prevalent in PD is known as bradykinesia (Fahn, 2008). Eventually, the patient experiences difficulty initiating voluntary movements as well as balance and posture deterioration. Movements become rigid and fragmented. As the disease progresses, the symptoms of neurodegeneration spread through the entire body (Hughes, 1994). Even though motor symptoms are the main hallmarks of PD, there are non-motor manifestations of the disease as well. A vast majority of PD patients exhibit speech problems, such as a softening of the voice or difficulty producing the lip movements necessary for proper talking (Ramig et al., 2008). These symptoms are due to the motor impairment characterizing PD. Other characteristic features of the disease include masklike facial expressions, sleep disturbances, and digestive problems (Fahn, 2008). While many PD patients do not experience cognitive decline, dementia affects up to half of sufferers with later stages of the disease (Goldman et al., 2018). By itself, PD is not fatal. Instead, patients often die of secondary complications, such as pneumonia, bronchitis, heart disease, or accidents due to restricted motion (Iwasaki et al., 1990). Figure 4a summarizes the main symptoms of PD. The devastating symptoms of the disease highlight the urgency of elucidating its molecular mechanisms, with the goal of uncovering proper targets for treatment.
Upon autopsy of PD patients, one finds macroscopic evidence of cell death. The dark bands that make up the substantia nigra, a structure in the midbrain, disappear (Petrucelli, 2008). This disappearance indicates the loss of dopaminergic neurons. Dopamine synthesis produces the pigment neuromelanin as a byproduct. In this way, the loss of the dark neuromelanin bands of the substantia nigra correlates with the death of dopaminergic neurons (Federow et al., 2005). Zooming in on the microscopic level, it becomes evident that the dying dopaminergic neurons are plagued by cytoplasmic inclusions known as Lewy bodies (Petrucelli, 2008). These Lewy bodies are composed mainly of misfolded variants of the protein α-synuclein (Spillantini et al., 1997). The hallmarks of PD, from the gross appearance of the diseased brain, to the structure of Lewy bodies, and to the fibrillization of α-synuclein are demonstrated in Figure 4b. Research has not reached a consensus on whether or not Lewy bodies are harmful. Some labs subscribe to the hypothesis that these aggregates are responsible for cell death, while others postulate that the aggregates are mechanisms of localizing aberrant proteins and that pathology would be worse without them (Petrucelli, 2008).
α-Synuclein is a small protein of 140 amino acids, whose function in the brain remains unelucidated (Allendoerfer, 2008). It has an irregular structure that is prone to both membrane binding and aggregation in fibrils (Lasheul et al., 2013). These two properties are thought to endow α-synuclein with the ability to aggregate into intracellular foci as it misfolds. The structure of α-synuclein can best be described as intrinsically disordered and divisible into three domains. The N domain (amino acids 1-59) mediates interactions with lipids and membranes; the M domain (amino acids 60-94) facilitates aggregation; and the C domain (amino acids 95-140) is responsible for keeping the protein in solution (Snead et al., 2014 and Vamvaca et al., 2010). These three domains come together to give α-synuclein its overall function. Figure 4c describes the linear structure of α-synuclein and focuses on membrane-binding and aggregation, which together give the protein its pathological propensities. Researchers have postulated that the protein is important in vesicle transport and may have a chaperone function, helping other proteins fold properly (Petrucelli, 2008). More recently, the function of α-synuclein has been studied from the perspective of neurotransmitter release. An influential paper showed that α-synuclein helps SNARE complexes assemble, allowing for vesicles carrying neurotransmitters to fuse efficiently with the membranes of neurons (Burré et al., 2010). This puts α-synuclein at the forefront of neurotransmission and perhaps explains its prevalence throughout the brain. No matter what function is ascribed to α-synuclein, the protein’s involvement in PD remains undisputed. Understanding the function of α-synuclein and how it is altered in the protein’s diseased state may be the key to unlocking the mysteries of PD and understanding how treatments for this form of neurodegeneration might work.
α-Synuclein pathology is central to all types of PD. In general, PD can be subdivided into familial or sporadic versions. The familial variant accounts for 10% of cases and is autosomal-dominant, as mutant α-synuclein is involved (Ross et al., 2008). Six α-synuclein mutations, A30P, E46K, H50Q, G51D, A53T, and A53E have been identified, and all give early-onset, aggressive versions of the disease (Ross et al., 2008). Each familial mutant is toxic in its own way. A classic paper tracked the aggregation kinetics of the A30P and A53T α-synuclein mutants, finding aggregation to occur at different rates (Narhi et al., 1999). This suggests that different mechanisms operate behind the pathogenesis of each mutant.
In contrast, 90% of PD patients have sporadic versions of the disease, in which α-synuclein misfolding and pathology start in the absence of a genetic cause (del Tredici et al., 2016). Although the causes of sporadic PD remain unelucidated, researchers have pinpointed environmental toxins, oxidative stress, and the inheritance of certain non-disease-causing versions of genes as potential culprits (Chai et al., 2013). In both iterations of the disease, dopaminergic neurons die and α-synuclein misfolds in Lewy bodies (intracellular cytoplasmic aggregates).
Delving into the properties of this protein and understanding what factors control its structure and function could be a useful approach for both familial and sporadic PD. At its core, PD is a protein-misfolding disease. It is an axiom that protein structure dictates function. When α-synuclein misfolds in both types of PD, its structure changes, so it can no longer function properly. Determining what causes the shift from the proper to the toxic structure remains a priority in both basic and translational research on α-synuclein.
the combinatorial impact of post-translational modifications of α-synuclein. Several questions are immediately apparent from this approach. Are there particular combinations of modifications that drastically increase α-synuclein’s toxicity? Can other combinations relieve that toxicity? Are there certain modifications that counteract each other?
The DebBurman PD Research Lab has invested heavily in exploring post-translational modifications of α-synuclein. The lab first demonstrated that α-synuclein is indeed nitrated in yeast and that nitration is detrimental to the growth of fission yeast (Solvang, Senior Thesis, 2011). Later work expanded to phosphorylation, showing that disrupting this modification alters the localization of α-synuclein (Fiske et al., 2011). More recently, the lab focused on SUMOylation. Removing SUMOylation led to increased toxicity and aggregation of α-synuclein (Thomas, Senior Thesis, 2018). The promising nature of these discoveries has led the lab to expand its investigations of the post-translational modifications of α-synuclein and to consider such modifications in combination with each other.
This thesis will focus on four α-synuclein modifications – SUMOylation, phosphorylation, acetylation, and glycation. To my knowledge, the combined effect of α-synuclein SUMOylation and phosphorylation has not been investigated. In addition, evidence for the roles of acetylation and glycation is based on very narrow studies. The proposed effects of these modifications are based on sparse evidence in few model systems. The interplay between acetylation and glycation (as well as acetylation, glycation, phosphorylation, and SUMOylation taken together) has not been investigated in any model system. Finally, the impact of α-synuclein acetylation and/or glycation on the six familial mutations has not been investigated. As alluded earlier, each mutant exhibits a unique toxicity pattern (Narhi et al., 1999). Through the experiments performed in this thesis, I hope to begin filling some of these gaps in the literature. I hope that the work of the DebBurman lab will inspire other labs to perform further experiments with these modifications and ultimately, to translate this research into therapies that help PD patients overcome their pathology, or at least, manage their symptoms.
Part 4: Modeling Parkinson’s Disease
The brief review of the various α-synuclein modifications alludes to a central theme in the experimental sciences – researchers need a suitable model system in which to perform experiments. Common models for PD include rodents (such as mice and rats), fruit flies, earthworms, and neuron culture lines (Allendoerfer et al., 2008). Some labs have even gone as far as using primates to show that toxins can induce parkinsonism and to search for biomarkers of PD (Lin et al., 2015). Each model system is valuable in its own right and brings out different aspects of the disease. For example, mice are valuable for the quantification of behavioral deficits and dopamine levels, while flies allow for quick genetic manipulations in a relatively high-order organism.
The DebBurman PD Research Lab uses an unconventional but equally powerful model organism – yeast cells. Baker’s yeast (Saccharomyces cerevisiae) have many advantages that make them a useful tool in investigations of disease at the protein level. Yeast are among the simplest available eukaryotic organisms. Their quick turnover rates and the fact that they are relatively inexpensive make yeast a good platform for experiments with multiple replications (Allendoerfer et al., 2008). Yeast readily accept foreign DNA, so they can make proteins that are not part of their pre-existing proteome (Orr-Weaver et al., 1983). Yeast make, fold, and degrade proteins just like humans (Allendoerfer et al., 2008). As shown many times in previous research, PD is a protein misfolding disorder on the molecular level, making yeast a good platform for studying this disease.
By tracking the way α-synuclein with various modifications affects yeast cells, we can draw basic conclusions that might be applicable to the neurons of PD patients. Budding yeast have been a powerful model system, yielding important insights into PD pathology. For example, a classic 2003 paper (which was among the first to establish yeast as a main-stream model of PD) showed that α-synuclein can aggregate in lipid droplets and can induce abnormal vesicle trafficking (Outeiro and Lindquist, 2003). Also, the same lab found that α-synuclein inhibits ER/Golgi transport by preventing the proper docking of vesicles (Gitler et al., 2008). Yeast have been crucial to our understanding of α-synuclein in many other ways. They have been used to show that α-synuclein interacts with synphilin-1 to induce inclusion formation (Engelender et al., 1999). A different paper showed that the A30P and A53T α-synuclein familial mutants trigger apoptosis in yeast, which was surprisingly reversible by heat shock (Flower et al., 2005). Another key contribution of the yeast model was that α-synuclein (in its A30P variant) impairs the proteasome system and curtails cells’ ability to both synthesize proteins and withstand aging (Chen et al., 2005). Yeast were critical in determining the functions of the domains of α-synuclein. For example, yeast biology showed that the N-domain is critical for membrane binding and that it adopts an α-helical conformation (Vamvaca et al., 2009).
The DebBurman PD Research lab has shown that α-synuclein pathology is connected to proteasome dysfunction and oxidative stress (Sharma et al., 2006). Past members have also shown that budding yeast and fission yeast (Schizosaccharomyces pombe) capture different aspects of PD pathology (Brandis et al., 2006). While budding yeast demonstrate membrane binding of WT α-synuclein, fission yeast feature prominent α-synuclein aggregation (Brandis et al., 2006). Both of these properties actively contribute to the diseased state of neurons. Our elucidations of the roles of nitration, phosphorylation, and SUMOylation have also come exclusively from yeast (Solvang, Senior Thesis, 2011, Fiske et al., 2011, and Thomas, Senior Thesis, 2018). These experiments, and many others like them, have validated yeast as a useful PD model, despite its unconventionality.
Part 5: About This Study
Gap in Knowledge
The combined effects of α-synuclein SUMOylation and phosphorylation warrant further research. Also, the role of α-synuclein acetylation and glycation in PD pathogenesis has been insufficiently studied. Only one paper in the literature has been dedicated to each, offering limited perspectives in only a few model organisms. By manipulating acetylation and glycation (either on the α-synuclein level or the yeast genome level), I hope to add to the emerging body of findings on these two modifications. In addition, the interplay between acetylation and glycation with SUMOylation and phosphorylation has never been studied. This combinatorial analysis is important, as patients are rarely influenced by solely one modification of α-synuclein. Finally, the six familial mutations of α-synuclein have generated plentiful research, but no labs have investigated them in the context of acetylation or glycation.
My hypotheses are based on limited evidence from studies in relatively non-diverse model systems. I hypothesized that each of the modifications would have either a protective or detrimental effect and that in combination, the modifications would influence each other’s effects. Specifically, I predicted that: 1) SUMOylation is protective (Krumova et al., 2011), 2) phosphorylation is harmful (Chen et al., 2005), 3) SUMOylation and phosphorylation counteract each other, 4) acetylation is protective (Miranda et al., 2017), 5) glycation is harmful (de Oliveira et al., 2017), and 6) when placed in combination, acetylation or glycation would counteract SUMOylation and phosphorylation in terms of toxicity, localization, and expression levels. I also predicted that post-translational modifications would influence the toxicity patterns of the six familial mutants. The direction of this effect is difficult to predict, as the literature is largely silent on the impact of acetylation and glycation in tandem with the familial mutations of α-synuclein. Figure 6 breaks down the aims and predictions of the study, while Figure 7 gives more information about our model organism of choice (yeast).
Breakdown of Experiments
Chapter 1: Combined Effects of SUMOylation and Phosphorylation
Aims: My first goal was to combine artificial α-synuclein mutations blocking SUMOylation (K96R/K102R) with mutations that block phosphorylation (S87A/S129A) or mimic it (S87D/S129D) in budding yeast.
Main findings: I found that while blocking SUMOylation is highly toxic, blocking SUMOylation and phosphorylation together reverses this effect. Thus, phosphorylation serves as a “brake” that opposes the beneficial effects of SUMOylation. This reversal of toxicity corresponds with a reduction in α-synuclein foci in yeast. I concluded that SUMOylation is protective, while phosphorylation is harmful, and the two modifications counteract each other.
Chapter 2: Effects of Acetylation and Glycation
Aims: I next investigated the effects of blocking acetylation (K6R, K10R, and K6R/K10R mutations), mimicking acetylation (K6Q, K10A, and K6Q/K10Q mutations), or blocking both acetylation and glycation (K6A, K10A, and K6A/K10A mutations) on α-synuclein.
Main findings: I found that mimicking α-synuclein acetylation has a slightly protective effect on yeast growth and correlates with a diffuse α-synuclein cellular localization. Blocking both acetylation and glycation has a similar protective effect and much lower expression levels, supporting my initial hypothesis that acetylation is protective while glycation is harmful.
Chapter 3: Acetylation in Combination with SUMOylation and Phosphorylation
Aims: My next goal was to combine acetylation with SUMOylation and phosphorylation. I promoted a hyper-acetylating environment in yeast cells, using that context to study the effects of α-synuclein SUMOylation and phosphorylation.
Main findings: I found that increasing overall acetylation leads to better yeast growth in all conditions. More specifically, hyper-acetylation was sufficient to make up for the toxicity and aggregation associated with deficient SUMOylation. Also, hyper-acetylation led to increased toxicity when phosphorylation was blocked. As this was not observed in WT yeast, I concluded that acetylation and phosphorylation may work together to mediate the phenotype of α-synuclein.
Chapter 4: Glycation in Combination with SUMOylation and Phosphorylation
Aims: My fourth goal was to combine glycation with SUMOylation and phosphorylation. In analogy to Chapter 3, I used yeast promoting a hypo-glycation environment in combination with modifying SUMOylation and phosphorylation at the α-synuclein level. Two different glycation-altering strains were used to provide more support for the role of this post-translational modification.
Main findings: I found that decreasing overall glycation leads to better yeast growth in all conditions. Glycation is thus detrimental to yeast growth in general. However, the loss of glycation reverses the beneficial effect of removing α-synuclein phosphorylation (toxicity and foci formation both increase). This indicates that glycation and phosphorylation together control the toxicity and localization of α-synuclein.
Chapter 5: SUMOylation in Combination with Acetylation and Glycation
Aims: The next part of the project was a reversal of the previous two chapters. Here, I placed α-synuclein manipulated for acetylation and glycation in yeast strains that globally increased or decreased SUMOylation.
Main findings: I found that in the strain deficient for SUMOylation, there were no major profile changes in the acetylation/glycation mutants’ properties (toxicity, localization, and expression). The same could be said of the strain promoting excess SUMOylation. This shows that the interplay we have described previously could have strain-dependent effects that need to be elucidated further.
Chapter 6: Familial Mutations with Modified Glycation
Aims: As a final addition to these investigations, I transformed the six familial mutants of α-synuclein in the glycation-altering strain and assessed changes in toxicity.
Main findings: I found that there may be a mutant-specific effect of glycation on the toxicity of the various α-synuclein mutants. A53E resisted the protective nature of global hypo-glycation.
I joined the DebBurman lab in the summer of 2016, beginning work on the SUMOylation project with Rosemary Thomas ’18. In 2017, I started work on the combinatorial impact of SUMOylation and phosphorylation. In the summer of 2018, I characterized the acetylation/glycation mutants. With Estella Tcaturian ’21, I investigated the modifications in triple combinations. I finished data acquisition over the 2018-19 academic year.
MATERIALS AND METHODS
The α-synuclein manipulations and functional experiments described here are derived from Outeiro and Lindquist (2003), Sharma et al. (2006), and Thomas, Senior Thesis, 2018. Without these preceding studies, this work would not have been possible.
Strategies of Manipulation
In this thesis, I aimed to evaluate the effects of SUMOylation, phosphorylation, acetylation, and glycation. I also investigated the combinatorial impact of these α-synuclein modifications. To achieve these goals, I broke up my work as described below.
Chapter 1: I first created mutants that specifically manipulate SUMOylation and phosphorylation. I removed SUMOylation at one site (through the K96R mutation) and then at two sites simultaneously (K96R/K102R). I next investigated the effects of blocking SUMOylation through the aforementioned double mutation and at the same time either blocking phosphorylation (S87A/S129A – SUMO) or mimicking it (S87D/S129D – SUMO).
Chapter 2: My next goal was to engineer mutants of α-synuclein that either mimic or block acetylation and glycation. Three mutants block acetylation only – K6R, K10R, and K6R/K10R. Three mutants mimic acetylation only – K6Q, K10Q, and K6Q/K10Q. The last three mutants block both acetylation and glycation – K6A, K10A, and K6A/K10A. There is no amino acid substitution that models constitutive glycation, making it infeasible to use this approach to create a set of mutants mimicking that modification.
Chapter 3: For the rest of the thesis, I altered my approach. I combined α-synuclein modifications with alterations at the yeast level. I obtained a yeast strain modified to provide an environment of hyper-acetylation. In this strain, I expressed either WT α-synuclein, versions of the protein modified to block SUMOylation, or versions modified to block SUMOylation along with mimicking or blocking phosphorylation. This allowed me to study the effects of these modifications in tandem.
Chapter 4: As a next step, I used strains of yeast modified to promote a hypo-glycating environment. Once again, I expressed either WT α-synuclein, versions of the protein modified to block SUMOylation, or versions mutated to block SUMOylation along with mimicking or blocking phosphorylation.
Chapter 5: My next idea was to reverse the logic presented in Chapters 3 and 4. I used strains of yeast modified for excess or deficient SUMOylation. In these strains, I expressed α-synuclein modified for acetylation or glycation, as described in Chapter 2. This reversal allowed me to investigate the combinatorial impact of these modifications from a different angle.
Chapter 6: I finally transformed the six familial mutants of α-synuclein (A30P, E46K, A53T, H50Q, G51D, and A53E) in the glycation-altering strain glo1∆ from Chapter 4. That is, I evaluated the effects of these mutants under conditions promoting hypo-glycation.
Table 1 summarizes the different types of α-synuclein modifications (and the strains in which they were placed).
The model organism of choice in these experiments was budding yeast (Saccharomyces cerevisiae). In Chapters 1 and 2, unmodified yeast (of the BY4741 strain) were used. In Chapter 3, a modified strain (sirt2D) provided a hyper-acetylation environment. Sirtuin-2 is an enzyme that removes acetyl groups. When this enzyme is no longer present, hyper-acetylation may result (de Oliveira et al., 2017). In Chapter 4, two different modified strains, glo1D and tpi1D, were used. Both provide deficient glycation by different mechanisms. Knocking out glo1 removes the enzyme that attaches sugars to α-synuclein, while knocking out tpi1 prevents sugars from being isomerized into an attachable form (Miranda et al., 2017). All three knockout strains were obtained from Open Biosystems (Dharmacon). For Chapter 5, I used temperature-sensitive SUMOylation-altering strains. The smt3ts strain is engineered so that it is deficient in SUMOylation at 30 °C (while it is normal at 25 °C). Similarly, the ulp1ts strain is engineered to produce hyper-SUMOylation at 30 °C (while it is normal at 25 °C). Finally, in Chapter 6, I transformed the six familial mutants in the glycation-altering strain glo1∆. The ease with which the genomes of yeast can be manipulated makes these organisms a good platform for studying the combinatorial impact of post-translational modifications at many levels.
Creation of Mutants
PCR-based site-directed mutagenesis was used to create the mutations mimicking or blocking SUMOylation, phosphorylation, acetylation, or glycation. In this technique, wild-type α-synuclein DNA obtained from E. coli is put through a methylation reaction. Then, the desired mutation is amplified by PCR using mutation-carrying primers. The GENEART Site-Directed Mutagenesis Protocol was followed, using a mutagenesis kit developed by Invitrogen Life Technologies (catalog number A13282). Once the desired mutant was created, the DNA was transformed into DH5α E. coli for further amplification and removal of the methylated template. After plasmid isolation, the DNA was sequenced to confirm that the mutation was successfully made. Finally, the DNA was transformed into the appropriate yeast strain, so that it could be used for functional analyses. Figure 8 summarizes this process, while Table 2 lists the primers used.
α-Synuclein Expression in Yeast
All α-synuclein DNA is in the pYES2.1 vector from Invitrogen. This vector puts the gene of interest under the control of a galactose-inducible promoter. Thus, yeast only express α-synuclein when galactose is present in the growth media. If glucose is present, α-synuclein is not produced. For this reason, yeast grown in glucose are used to look at the effects of individual strains (or as loading controls). The effects of α-synuclein can only be evaluated in galactose media. The pYES2.1 vector also allows for selection of only those yeast colonies that contain the gene of interest. BY4741 yeast cannot make their own uracil, an essential component of RNA. The yeast are grown in media that contain no exogenous sources of this nucleotide. Thus, if a colony is to survive, it must have accepted the pYES2.1 vector, which gives the cells a copy of the URA3 gene, allowing them to produce uracil.
The three experimental assays described below were chosen to correspond to key pathological properties of α-synuclein neurodegeneration. In PD, α-synuclein causes toxicity to neurons, loses its normal cellular localization, and is overexpressed. These three characteristics can be evaluated in yeast via spotting, microscopy, and Western blotting, respectively, as shown in Figure 7 of the introduction.
- Serial Dilution Spotting – Toxicity Analysis
To analyze how toxic a particular manipulation is, I performed serial dilution spotting. In this procedure, I begin by growing cell cultures in 5 mL of SC-URA Glucose overnight at the appropriate temperature (either 30 or 25 °C, depending on the strain). Note that SC-URA is a type of selective, non-α-synuclein-inducing media. I then wash the cells and by counting with a hemocytometer, adjust the cell density for each sample to 2 X 106 cells per mL. A 1 mL suspension at the new density is then made. In a 96-well microtiter plate, the suspension is diluted either 5-fold or 10-fold, depending on the experiment. 5-fold dilutions utilize 6 lanes in the microtiter plate, while 10-fold dilutions accommodate only 4. Using a multi-channel pipette, the diluted cells are spotted onto glucose and galactose plates (in triplicate for each). The spottings on glucose are used as a loading control, to make sure that the colonies have the same cell density. The spottings on galactose are used to identify the effects of the α-synuclein manipulation under investigation. Each experiment is performed for a minimum of 5 repetitions, each in triplicate.
2a. Fluorescence Microscopy – Localization Analysis
All variants of α-synuclein in the lab are tagged with enhanced Green Fluorescent Protein to allow for localization analysis. Cultures are grown overnight at 30 °C (except for the temperature-sensitive strains, which must be grown at 25 °C) in SC-URA Glucose media (repressing α-synuclein expression). The cells are then transferred to galactose media to induce α-synuclein expression. The cells are observed under a Nikon TE2000-U Fluorescent Microscope 6, 12, 18, and 24 hours after α-synuclein expression has begun (using 1000X magnification). In each condition, pictures of over 1,000 cells per time-point are taken in fields selected without conscious bias. The viewing field on the microscope is changed three times before each image is taken, and the plane is selected without regard to observed cellular phenotype. At least five trials for each α-synuclein variant are performed.
2b. Microscopy Quantification and Statistics
Each cell in an image is grouped into one of five categories depending on how α-synuclein is distributed. The protein could be binding the membrane, diffuse throughout the cytoplasm, clumping in foci (intracellular aggregates), binding and clumping simultaneously, or binding and diffuse simultaneously. These phenotypic categories are shown in Figure 9. In each condition at each time-point, the number of cells falling in each category is counted. Then, a chi-square test is run to determine whether the phenotypic distribution of mutant cells in each category differs from the wild-type or whether a particular strain differs from BY4741.
- Western Blotting – Expression Analysis
To measure how much protein is made in each condition, I used Western blotting. In this technique, cells are cultured, washed, and adjusted to a density of 1.5 X 107 cells/mL. The yeast cells are lysed with 100 mmol sodium azide, electrophoresis sample buffer, and glass beads. The extracted proteins are then run on a Tris-Glycine gel from Invitrogen (catalog number XP00100BOX). Along with the samples, SeeBlue ladder is run on the gel to ensure adequate comparison of molecular size and to gauge the quality of the transfer. The contents of the gels are then transferred to membranes, which are probed with anti-α-synuclein primary antibodies (and the mouse anti-alkaline phosphatase secondary antibody). I used PGK (phosphoglycerokinase) as a loading control in all cases. Mouse anti-α-synuclein and anti-PGK antibodies from Molecular Probes were used (at a1:200 dilution). Reagents for the Western blot came from an Invitrogen Western Breeze Kit (catalog number WB7103).
In chronological order, the samples were made by the following students: Nijee Sharma (GFP, WT α-synuclein, A30P, and A53T); Michael Fiske (E46K), Wase Tembo (H50Q, G51D, and A53E); Alex Roman (SUMOylation blocking mutants); Rosemary Thomas and Yoan Ganev (SUMOylation and phosphorylation-manipulating mutants); Yoan Ganev, Chisomo Mwale, Ariane Balaram, and Estella Tcaturian (KR, KQ, and KA sets of acetylation/glycation manipulating mutants).
COMBINED EFFECTS OF SUMOYLATION AND PHOSPHORYLATION
*Experiments performed in collaboration with Rosemary Thomas
My first goal was to determine whether or not α-synuclein SUMOylation and phosphorylation counteract each other. A few papers suggest that SUMOylation is protective (Krumova et al., 2011), while phosphorylation is harmful (Chen et al., 2005). Our lab has contributed to these findings as well. See Appendix A1, in which we demonstrated that SUMOylation protects against toxicity (Thomas, Senior Thesis, 2018), while phosphorylation drives harmful phenotypes (Fiske et al., 2011). The question of whether SUMOylation and phosphorylation together control the effects of α-synuclein remains open. This combinatorial approach was intriguing, as in patients, modifications often appear in tandem. I used the following samples for these experiments: vector control (no α-synuclein), WT α-synuclein, K96R/K102R α-synuclein (blocking SUMOylation), K96R/K102R/S129A (blocking SUMOylation and phosphorylation), S87A/K96R/K102R/S129A (blocking SUMOylation and phosphorylation), K96R/K102R/S129D (blocking SUMOylation and mimicking phosphorylation), S87D/K96R/K102R/S129D (blocking SUMOylation and mimicking phosphorylation), and finally GFP (no α-synuclein control). I performed experiments that assessed the toxicity (spotting), localization (microscopy), and of these α-synuclein variants in yeast. I predicted that there would be a combinatorial effect of α-synuclein SUMOylation and phosphorylation. For these studies, I collaborated with Rosemary Thomas ’19.
Phosphorylation Counteracts SUMOylation’s Effects on α-Synuclein Toxicity
I first tested whether the aforementioned modifications increase the toxicity of α-synuclein or curtail it. The death of dopaminergic neurons as a result of α-synuclein toxicity is a hallmark of PD pathology. To capture this aspect of the disease, any PD model must incorporate α-synuclein toxicity as a characteristic. I used serial dilution spotting to this end, systematically monitoring yeast growth levels on glucose and galactose plates. In the yeast model, galactose induces α-synuclein expression, so the glucose plates serve as a loading control.
Figure 10 depicts representative results (5-fold dilution spotting). Growth in the glucose plates is even, meaning comparisons in galactose are possible. Upon induction, WT α-synuclein is slightly toxic relative to the vector and the GFP controls. Less growth across the lane indicates decreased colony formation. Toxicity increases with the SUMOylation blocker. A striking result occurs with the double SUMOylation and phosphorylation blocker. A rescue of toxicity relative to the SUMOylation blocker occurs. One can counteract the negative effects of blocking SUMOylation by also blocking phosphorylation. Conversely, toxicity increases with the phosphorylation mimic on a blocked-SUMOylation background. These trends were all moderate in appearance. Taken together, our spotting results support the hypothesis that SUMOylation and phosphorylation counteract each other.
SUMOylation Controls α-Synuclein Aggregation
Next, I asked whether α-synuclein modified for SUMOylation and phosphorylation exhibits differential localization. In PD, α-synuclein aggregates in Lewy bodies. The protein no longer holds its normal position in the cell, creating abnormal cytoplasmic deposits. I tagged α-synuclein with eGFP (an enhanced variant of green fluorescent protein), allowing me to track its position in the cell over time. I was particularly interested in five phenotypes: membrane binding, formation of foci (which may resemble aggregates in Lewy bodies), cytoplasmic diffusion, membrane binding combined with foci, and membrane binding combined with diffusion. I observed the SUMOylation/phosphorylation α-synuclein mutants over a course of 6, 12, 18, and 24 hours under the microscope, as shown in Figure 11. I found that WT α-synuclein localizes to the membrane by 18 hours. When SUMOylation is blocked, stark foci form, and even by 24 hours, almost no membrane binding occurs. In contrast, membrane binding returns by 24 hours when both SUMOylation and phosphorylation are blocked. Removing phosphorylation at the same time as SUMOylation reverts the phenotype back to the WT pattern. Finally, a slightly diffuse phenotype results with the SUMOylation blocker and phosphorylation mimic. Figure 12 quantifies the results for SUMOylation, providing statistical backup for the conclusions. The differences described above were statistically significant by a chi-squared analysis (p < 0.005 for all distributions relative to WT α-synuclein). My data suggest that both SUMOylation and phosphorylation impact the localization of α-synuclein and are responsible in tandem for the mis-localization of the protein in PD.
α-Synuclein SUMOylation and Phosphorylation Do Not Heavily Impact Expression
Finally, I asked whether SUMOylation and phosphorylation impact the expression levels of α-synuclein. In PD, α-synuclein is overexpressed as a result of excess production, deficient degradation, or both. Yeast also exhibit α-synuclein overexpression, detectable via Western blotting. I ran a Western blot of the SUMOylation/phosphorylation altering samples to determine the modifications’ effects on α-synuclein expression levels. Figure 13 summarizes the results of these experiments. I found that cells express WT α-synuclein less than the SUMOylation-blocking variant. Blocking SUMOylation together with a phosphorylation blocker/mimic leads to a similar expression increase. These effects are very slight. Overall, these expression data suggest that post-translational modifications of α-synuclein affect the amount of protein present in yeast cells.
EFFECTS OF ACETYLATION AND GLYCATION
My second goal was to evaluate the effects of α-synuclein acetylation and glycation. Only two papers have delved into the functions of these α-synuclein modifications (de Oliveira et al., 2017 & Miranda et al., 2017). To my knowledge, acetylation and glycation have not previously been studied in yeast. To address this gap, I created three sets of α-synuclein mutants. The K-R set (K6R, K10R, and K6R/K10R) blocked acetylation alone. The K-Q set (K6Q, K10Q, and K6Q/K10Q) mimicked acetylation alone. Finally, the K-A set (K6A, K10A, and K6A/K10A) blocked both acetylation and glycation. No mutant mimics glycation or blocks it alone. I predicted that acetylation of α-synuclein is protective, while glycation is harmful. In analogy to Chapter 1, I determined the toxicity, localization, and expression levels of α-synuclein modified for acetylation and glycation using spotting, microscopy, and Western blotting.
Acetylation Attenuates α-Synuclein Toxicity while Glycation Promotes It
I first asked whether acetylation and glycation alter the toxicity of α-synuclein. As in Chapter 1, I performed serial dilution spotting, using GFP and WT α-synuclein samples as controls. I then spotted three sets of α-synuclein N-terminal lysine mutants. Figure 14 summarizes representative results of these 10-fold dilution experiments. I found that blocking acetylation alone is about as toxic to yeast as WT α-synuclein. This effect was relatively weak. Mimicking acetylation has a stronger effect, eliminating α-synuclein’s toxicity. When I blocked both acetylation and glycation, toxicity once again disappeared. This effect must be due to the removal of glycation. Considered in tandem, these data support my initial hypothesis that acetylation is protective while glycation is harmful.
Acetylation and Glycation Alter α-Synuclein Localization
I next investigated whether the three sets of acetylation- and glycation-manipulating mutants influence the localization of α-synuclein. I performed fluorescence microscopy, using WT α-synuclein as a control. Figure 15 shows representative microscopy images. I found that while WT α-synuclein binds to the membrane by 18 hours, foci persist in the K-R mutant set (blocking acetylation), even by 24 hours. In contrast, the K-Q set (mimicking acetylation) exhibits two dominant phenotypes by 24 hours. K6Q binds the membrane like WT, while K10Q and K6Q/K10Q both become cytoplasmically diffuse. The K-A set (blocking both acetylation and glycation) mostly binds the membrane but exhibits some cytoplasmic diffusion as well.
My statistical analysis supported these qualitative descriptions (Figure 16). I found that the 24-hour time-point most relevantly captured the phenotypes resulting from each manipulation. By 24 hours, 91% of the WT α-synuclein samples bound the membrane, 4% formed foci, and only 3.6% were diffuse throughout the cytoplasm. The remaining 1.4% exhibited combinatorial phenotypes. For the K6R/K10R samples at 24 hours, the phenotype was more mixed. 75.4% of cells exhibited membrane binding, 3.5% formed foci, 20.6% were diffuse, and the remaining 0.5% exhibited combinatorial phenotypes. For the K6Q/K10Q samples at 24 hours, I observed marked cytoplasmic diffusion, capturing a dominant 93.6% of the total phenotypic distribution. Of the same sample, 2.1% exhibited membrane binding, 1.8% formed foci, and the remaining 2.5% exhibited combinatorial phenotypes. Finally, the distribution of the K6A/K10A mutant was as follows: 87.6% membrane binding, 3.6% formation of foci, 5.5% cytoplasmic diffusion, and 3.3% mixed phenotype. These distributions differed significantly from the expected WT α-synuclein distribution, according to a chi-squared analysis for goodness of fit: X2 (6, N = 16) = 260.145, p < .0005. I concluded that both mimicking acetylation and blocking acetylation and glycation together leads to a more diffuse phenotype.
α-Synuclein Expression Reduced with Blocked Acetylation and Glycation
Finally, I asked whether acetylation and glycation affect the expression levels of α-synuclein. I used a Western blot to answer this question. Figure 17 shows a representative example. Surprisingly, both the acetylation blocking mutants and the acetylation mimicking mutants had higher expression of α-synuclein than the WT control. Only the mutants blocking both acetylation and glycation (K-A set) showed markedly reduced expression of the protein.
GLYCATION IN COMBINATION WITH SUMOYLATION AND PHOSPHORYLATION
My fourth goal was to study how glycation affects the phenotypes of the SUMOylation and phosphorylation-manipulating mutants. As in Chapter 3, I manipulated SUMOylation and phosphorylation on the genetic level (via α-synuclein mutations), while I manipulated glycation on the yeast strain level. I achieved hypo-glycation in two separate ways – via a knockout of glo1 or via a knockout of tpi1, essential glycating enzymes. After obtaining these yeast strains, I transformed the GFP, WT α-synuclein, K96R/K102R (blocking SUMOylation), S129A-SUMO, S87A/S129A-SUMO (blocking SUMOylation and phosphorylation), S129D-SUMO, and S87D/S129D-SUMO (blocking SUMOylation and mimicking phosphorylation) mutants in them. I then performed serial dilution spotting, fluorescence microscopy, and Western blotting as in Chapter 3. I predicted that glycation would influence the effects of SUMOylation and phosphorylation. Specifically, removing glycation should promote healthy phenotypes in the SUMOylation and phosphorylation manipulating mutants.
Hypo-Glycation Protects Against α-Synuclein Toxicity
I first asked how an environment promoting hypo-glycation affects the toxicity of the SUMOylation and phosphorylation-altering α-synuclein mutants. To answer this question, I spotted these mutants in both the glo1∆ and tpi1∆ strains. Figure 23 shows representative results of these experiments. In Part A of the figure, I have provided a reproduction of the profile in BY4741. Parts B and C, respectively, show representative samples of my work in glo1∆ and tpi1∆.
In glo1∆, I found that, in general, the mutants preserved the pattern that they exhibited in BY4741 (compare Figure 23a to 23b). Blocking SUMOylation presents toxicity, while blocking phosphorylation and SUMOylation together reverses this trend. Blocking SUMOylation while mimicking phosphorylation leads to growth similar to WT α-synuclein. I noted that in glo1∆, the SUMOylation and phosphorylation double blocker appeared to grow exaggeratedly better compared to BY4741. In tpi1∆, the hypo-glycating environment completely obliterates the differences between each of the post-translational modification manipulations (compare Figure 23a to 23c). All of the lanes grow at about the same level as the GFP control.
I next performed a follow-up spotting, comparing the SUMOylation and phosphorylation mutants in either glo1∆ or tpi1∆ to their BY4741 counterparts to better capture strain effects. Figure 24 (a and b) shows representative samples of this work. When I compared glo1∆ to BY4741, I found that overall, glo1∆ yeast grow more robustly. I also found that upon induction, the GFP control, WT α-synuclein, and the SUMOylation blocker all grow better in glo1∆ than BY4741. This trend was not as apparent when I blocked SUMOylation and either blocked or mimicked phosphorylation. In contrast, when I compared the growth of the same mutants in tpi1∆ to BY4741, I found that all samples (regardless of modification) grew better in tpi1∆ than in BY4741. Thus, the beneficial effect of hypo-glycation overpowers the effects of α-synuclein SUMOylation and phosphorylation.
Hypo-Glycation Promotes α-Synuclein Membrane Localization
The second question I asked for this chapter was whether hypo-glycation via glo1∆ or tpi1∆ influences the cellular localization of α-synuclein modified for SUMOylation and phosphorylation. As in the previous chapters, I used fluorescence microscopy. Figure 25 shows representative images of all samples in glo1∆ at all time-points. I found that when glycation was globally blocked, the SUMOylation-blocking mutants (K96R and K96R/K102R) both bound the membrane by 24 hours. This is in stark contrast to the findings in BY4741, where the SUMOylation blockers formed distinct foci. In addition, both the phosphorylation/SUMOylation blockers and the phosphorylation mimics that also block SUMOylation ended up at the membrane by 24 hours. Overall, when glycation is globally blocked, membrane binding prevails, even when that is not the case in BY4741. My statistical analysis supports these conclusions (Figure 26). The majority of samples bound the membrane. By 24 hours, 89.6% of WT cells, 95.9% of SUMOylation blocking cells, 95.6% of SUMOylation and phosphorylation blocking cells, and 86.2% of SUMOylation blocking and phosphorylation mimicking cells exhibited membrane binding. These results were significantly different from the expected BY4741 distribution, according to a chi-squared test of goodness of fit: X2 (3, N = 8) = 10.890, p = .0123.
In tpi1∆, I found a similar pattern as in glo1∆. Figure 27 shows representative images. All of the mutants (blocking SUMOylation and blocking SUMOylation while blocking/mimicking phosphorylation) bind the membrane more than their BY4741 counterparts. This conclusion once again is supported by statistics (Figure 28). By 24 hours, 89.6% of the WT cells, 97.7% of the cells blocking SUMOylation, 89.9% of the cells blocking both SUMOylation and phosphorylation, and 71.8% of the cells blocking SUMOylation and mimicking phosphorylation exhibited membrane binding. This distribution in tpi1∆ was significantly different from that in BY4741: X2 (3, N = 8) = 11.236, p = .0105. Taken together, the data in glo1∆ and tpi1∆ suggest that when glycation is blocked on the yeast level, α-synuclein is more efficiently cleared from the cytoplasm of the cells. Like hyper-acetylation, hypo-glycation promotes membrane binding.
Hypo-Glycation Promotes α-Synuclein Overexpression Upon Modified Phosphorylation
Next, I asked whether hypo-glycation affects α-synuclein expression levels. I used a Western blot to answer that question. Figure 29 recapitulates the main findings. In both glo1∆ and tpi1∆, WT α-synuclein exhibited similar expression levels to α-synuclein blocking SUMOylation. Interestingly, manipulating phosphorylation (mimic or blocker) elicited overexpression in both glo1∆ and tpi1∆.
Additive Effects on Toxicity of Beneficial Manipulations
As a follow-up to the data reported in Chapters 2-4, I was interested in whether or not the most beneficial manipulations of α-synuclein (and yeast) together would have an additive protective effect. I transformed the K6A, K10A, and K6A/K10A mutants, which had a protective effect in BY4741 (especially K10A) in the sirt2∆ and glo1∆ strains. Figure 30 shows a representative result. I found that compared to BY4741, the K6A, K10A, and K6A/K10A α-synuclein mutants in sirt2∆ and glo1∆ exhibited more robust growth. In addition, within each strain, the K-A mutant sets exhibited more growth than the respective WT α-synuclein control. This trend was very faint.These trends show that protective α-synuclein and yeast-level manipulations can additively protect against α-synuclein toxicity.
SUMOYLATION IN COMBINATION WITH ACETYLATION AND GLYCATION
My fifth goal was to reverse the paradigm established in Chapters 3 and 4. I now manipulated acetylation and glycation on the α-synuclein level, while I used yeast-level manipulations for SUMOylation. Specifically, I transformed the K-R (acetylation-blocking), K-Q (acetylation-mimicking), and K-A (acetylation and glycation blocking) mutant sets in two SUMOylation altering strains. The smt3ts strain exhibits decreased SUMOylation when grown at 30 °C, while the ulp1ts strain exhibits excess SUMOylation at that temperature. Both strains exhibit normal SUMOylation levels at 25 °C (control temperature). I once again performed spotting, microscopy, and Western blotting using these samples. The goal of this undertaking was to investigate whether global SUMOylation influences the effects of α-synuclein acetylation and glycation. I predicted that excess SUMOylation would shift the K-R, K-Q, and K-A mutants towards protective phenotypes, especially for the mutants that were already protective in BY4741.
Altered SUMOylation Does Not Impact Acetylation/Glycation Mutant Toxicity
I first began with spotting experiments to determine the effects on toxicity of globally manipulating SUMOylation in the acetylation/glycation mutants. In the smt3ts strain, I found that at 25 °C (normal SUMOylation), the K-A mutant set (blocking both acetylation and glycation) grows better, especially K10A. The K-Q mutant set (mimicking acetylation) grows better as well. Figure 31a shows representative samples of these experiments. My findings parallel those in BY4741. Given the agreement between BY4741 and smt3ts, I moved on to investigate the effects of the temperature-induced SUMOylation deficit. When the temperature was increased to 30 °C (leading to deficient SUMOylation in smt3ts), there was no appreciable effect on the overall growth of the yeast cells (glucose plates). When α-synuclein expression was induced, overall toxicity increased. This shows that SUMOylation is essential to keeping α-synuclein non-toxic to yeast. Interestingly, the acetylation/glycation mutants did not exhibit alterations in their profile. The K-Q and K-A sets remained protective. This suggests that although global SUMOylation affects α-synuclein, it does not impact acetylation and glycation. I made similar conclusions from the ulp1ts strain. Figure 31b shows representative images of spotting experiments in that strain. There was no overall increase in toxicity for this strain when α-synuclein expression was induced. This suggests that excess SUMOylation does not hurt yeast cells expressing the protein. As with the other strain, the growth profiles of the K-R, K-Q, and K-A mutants was preserved in ulp1ts. Global SUMOylation has a minimal impact on the toxicity effects of acetylation and glycation α-synuclein mutants.
Altered SUMOylation Favors Cytoplasmic Diffusion
I next asked whether α-synuclein localization changes in response to global SUMOylation manipulations. To answer this question, I observed the smt3ts yeast under the microscope for α-synuclein localization. Representative images are shown in Figure 32. Most of the profiles were similar to the localization patterns exhibited in BY4741 for each of the mutants. The most notable exceptions were K10Q and K6Q/K10Q. While they were almost exclusively diffuse in BY4741, some cells trended towards the membrane in smt3ts when SUMOylation was deficient. These qualitative observations are supported by my statistical analysis. By 24 hours, 93.4% of WT cells trended to the membrane, 88.2% of K6R/K10R exhibited a membrane-bound phenotype, and 96.2% of K6A/K10A cells presented with α-synuclein binding the membrane. In contrast, 25% of K6Q/K10Q cells had a membrane-binding phenotype, with 16.7% forming foci, and 50% becoming cytoplasmically diffuse. The remaining cells exhibited mixed phenotypes. While qualitatively the distributions seemed to follow the trend set by BY4741, the chi-squared goodness of fit test showed that in fact, there was a statistically-significant difference between the BY4741 and smt3ts localization trends for the acetylation/glycation mutants: X2 (3, N = 8) = 252.8, p < .0005 (Figure 33).
I performed the same type of experiments in ulp1ts, and the results are summarized in Figure 34. As in smt3ts, most of the patterns I observed were similar to those in BY4741, except for minor deviations. For example, K6R/K10R retained some cytoplasmic diffusion by 24 hours, while K6Q/K10Q lost its diffusion and bound the membrane. My quantifications agreed with these observations. Overall, membrane binding predominated, but to a lesser extent than in smt3ts. 81.1% of our WT samples, 64.6% of the K6R/K10R mutant, 11.1% of the K6Q/K10Q mutant, and 65.8% of the K6A/K10A mutant bound the membrane. Unusually. in the K6Q/K10Q mutant, the dominant phenotype was aggregation (20.8% of samples). The distribution of these samples was different from the expected WT distribution: X2 (3, N = 8) = 46.621, p < .0005 (Figure 35). Taken together, my localization data suggest that the effects of SUMOylation predominate over those of acetylation and glycation. SUMOylation may influence acetylation (since the K-R and K-Q mutants are affected in both strains, respectively).
Altered SUMOylation Decreases α-Synuclein Expression
Finally, I asked whether hyper- and hypo-SUMOylation levels influence expression levels of α-synuclein. I observed two interesting trends in my Western blots. First, I found that overall, there was a much higher expression of α-synuclein in smt3ts than in ulp1ts. Second, the stark reduction in expression of the K6A/K10A mutants disappeared in smt3ts, but it remained intact in ulp1ts. One explanation for this is that cells need SUMOylation (which does not occur in smt3ts) to execute the protective effects of removing acetylation and glycation by reducing α-synuclein expression. These results are summarized in Figure 36.
FAMILIAL MUTATIONS WITH ALTERED MODIFICATION ENVIRONMENTS
My sixth and final goal was to investigate whether altered post-translational modification environments influence the toxicity patterns of the six α-synuclein familial mutants. Each familial mutant has a unique toxicity mechanism. This makes it likely that acetylation and glycation would differentially impact the familial mutants of α-synuclein. To investigate this question, I transformed the A30P, E46K, H50Q, G51D, A53E, and A53T mutants in the sirt2∆ and glo1∆ strains and began initial characterizations. I exclusively focused on the spotting assay. I predicted that there may be mutant-specific effects of hyper-acetylation or hypo-glycation on the growth of mutant α-synuclein. For this pilot examination, I only report data from the glo1∆ strain.
Growth Profile of α-Synuclein Familial Mutants in BY4741
To understand the results in the knockout strain better, I first provide a summary of the trends for the familial mutants of α-synuclein in BY4741. The DebBurman lab has extensively studied these six familial mutants. Sharma et al. (2006) described the properties of A30P and A53T, and Fiske et al. (2011), described the properties of E46K. More recently Maiwase Tembo’s thesis (2015) characterized H50Q, G51D, and A53E. Figure 37a recapitulates these phenotypes in the control strain BY4741. First, WT α-synuclein is toxic relative to control yeast that express GFP. Second, the mutants E46K, A53T, H50Q, and A53E are just as toxic as WT α-synuclein. However, depending on expression levels, A53T and H50Q can be significantly more toxic (Tembo, Senior Thesis, 2015). Third, A30P and G51D mutants are surprisingly protective, growing better than WT α-synuclein. Other yeast labs corroborate this protection (Outeiro and Lindquist, 2011, Fares et al., 2014). This result is counterintuitive, as these mutants cause aggressive early-onset Parkinson’s just like the others.
Glycation Does Not Strongly Affect Familial Mutant Toxicity
I next asked whether or not hypo-glycation from the glo1 knockout has an effect on the growth profiles of familial mutants. Figure 37b shows a representative result of these spottings. I have already established that in glo1∆, WT α-synuclein grows better than in BY4741. Relative to this new WT baseline, A30P and G51D were once again protective, showing that the glo1∆ had a minimal effect on the growth of these mutants. The profiles of the other mutants were also similar to those in BY4741. It is interesting to note that A53E remained toxic, despite the hypo-glycating environment. Perhaps the A53E mutant is resistant to the beneficial effects of removing glycation. This result would hint at a mutant-specific effect of the post-translational modifications.
As a follow up to this investigation, I am currently performing trials of the same familial mutants in sirt2∆. While the results are not yet conclusive, I hope to uncover other mutant-specific effects of altered post-translational modification environments. I am also in the process of performing experiments in which the same mutants are spotted on the same plate in BY4741, sirt2∆, and glo1∆. While instructive, these studies will inherently be affected by the strain effects of sirt2∆ and glo1∆, as reported in Chapters 3 and 4.
Parkinson’s Disease is a devastating neurodegenerative disorder characterized by tremors and difficulty initiating movement. The pathology of the disease involves loss of dopaminergic neurons in the substantia nigra. These compromised cells exhibit Lewy bodies, aggregates of the protein α-synuclein. Multiple hypotheses attempt to explain the molecular underpinnings of Parkinson’s Disease. Post-translational modifications present a promising mechanism that explains the patterns of α-synuclein toxicity, localization, and expression.
In this study, I hypothesized that SUMOylation and acetylation promote a healthy function of α-synuclein, while phosphorylation and glycation serve a detrimental role. I posed all of my questions in the powerful budding yeast model. In general, my hypotheses were confirmed. I found that: 1) some modifications are protective, 2) some modifications are harmful, 3) modifications can counteract each other’s effects, and 4) modifications affect familial mutants in a selective way. I discuss these findings in detail below. I also evaluate the strengths and weaknesses of the study, and I end by considering future directions based on the findings of this project.
Some Modifications are Protective
SUMOylation is Protective
For the first question I asked in this thesis, I hypothesized that SUMOylation is protective, guarding against the toxicity of α-synuclein. In general, my work supported this hypothesis. The first line of evidence that SUMOylation is protective came from my toxicity analysis. When I blocked SUMOylation via an artificial mutation, α-synuclein became more toxic – yeast grew less in serial dilution spotting. This shows that SUMOylation is essential for colony formation in yeast expressing α-synuclein. A second line of evidence (localization) is that without SUMOylation, α-synuclein has a greater propensity to form foci and aggregate. My microscopy findings point to abnormal α-synuclein localization in the absence of SUMOylation. These results suggest that SUMOylation is essential for keeping cells expressing α-synuclein relatively healthy. My expression level analysis did not show a correlation to increased expression, as would have been expected. Still, these data build a convincing case for the beneficial effects of SUMOylation.
The protective nature of SUMOylation is consistent with the limited literature on the topic. For example, one team of researchers showed that in vitro, SUMOylated α-synuclein exhibited slower aggregation kinetics than its non-modified counterparts (Krumova et al., 2011). Another study showed that SUMOylation helps sort α-synuclein into extracellular vesicles, implying that the modification is crucial for the efficient transport of the protein (Kunadt et al., 2015). Another in vitro paper confirmed that homogenously-SUMOylated α-synuclein (at lysines 96 and 102) exhibited less aggregation (Abeywardana et al., 2015). Interestingly, SUMOylation of α-synuclein has also emerged as a protective modification through some recent drug studies. These papers came out after I began work on this portion of the thesis. For example, one paper found that methamphetamine, an illicit drug, causes α-synuclein aggregation by decreasing SUMOylation levels (Zhu et al., 2018). Another study used PC12 cells treated with rotenone to model neurodegeneration. These effects were reversed by rifampicin, a drug that increased α-synuclein SUMOylation levels (Lin et al., 2017). My data adds to this body of findings by providing among the first in vivo evidence that loss of SUMOylation is harmful (previous studies deal with hyper-SUMOylation or naturally-occurring SUMOylation). Taken together, all of these lines of evidence suggest that SUMOylation may be a potential target against α-synuclein toxicity. As postulated in Krumova et al., 2011, SUMOylation may exert protective effects by increasing the solubility of the C-terminal of α-synuclein, lessening the propensity of the protein to aggregate.
Acetylation is Protective
I also investigated the role of α-synuclein acetylation. Based on very limited studies, I hypothesized that acetylation is protective. My work expanded the available evidence that this modification counters the adverse phenotypic effects of α-synuclein, confirming my hypothesis. In these studies, I showed both at the α-synuclein and yeast-strain levels that acetylation is beneficial. The first line of evidence was that when I mimicked α-synuclein acetylation via an artificial mutation, yeast cells grew better, and α-synuclein became cytoplasmically diffuse. These phenotypes indicate protection from the toxicity of α-synuclein in the yeast model (Jones, Senior Thesis, 2018; Thomas, Senior Thesis, 2018). From these trends, I concluded that acetylation attenuates the harmful effects of α-synuclein. Blocking acetylation of α-synuclein did not appreciably increase the protein’s toxicity, but it did lead to more aggregation, a traditionally harmful phenotype.
A second line of evidence came from my analysis of the sirt2∆ strain. These yeast cells are deficient in a deacetylase, promoting a hyper-acetylating environment. In general, sirt2∆ yeast grew better than their BY4741 counterparts. Even WT α-synuclein lost its toxicity in that strain. Also, the sirt2 knockout promoted more membrane binding, a healthy phenotype, as assessed by microscopy. Considered together, my data point in the direction that α-synuclein acetylation is protective.
Recent findings, generated at the same time as this thesis in other labs, support this conclusion. One study found that N-terminal acetylation of α-synuclein protects against artificially-induced oligomerization (Lima et al., 2018). This directly implicates acetylation as a protective mechanism against α-synuclein aggregation. An in vitro study suggested that acetylation is crucial for the ability of α-synuclein to recognize its binding partners (Toal et al., 2017). Yet another study demonstrated that acetylation promotes the helical conformation of α-synuclein’s N-terminus, thus protecting against aggregation (Iyer et al., 2016). These papers suggest mechanisms by which the acetylation mimic I used could protect yeast cells. Perhaps there is a decrease in the aggregation of α-synuclein, leading to lower toxicity levels and more membrane binding. Another possibility is that acetylation increases the solubility of α-synuclein by making the protein overall more hydrophilic. Acetyl groups are small and highly polar, making this likely especially if multiple sites are acetylated at once. Regardless of the case, my data suggest that acetylation is protective and that promoting it can curb the disease-causing properties of α-synuclein.
Some Modifications are Harmful
Phosphorylation is Harmful
Some α-synuclein modifications push the protein towards more harmful phenotypes, promoting aggregation and neurodegeneration. I hypothesized that phosphorylation falls under this category. My data gave extensive support for this hypothesis. The first line of evidence came from mutations blocking phosphorylation and SUMOylation. When I removed phosphorylation via an artificial mutation, yeast cells grew better. In addition, the cells exhibited a membrane-bound localization resembling the WT α-synuclein control. This rescue of toxicity and reversal of aggregation in the absence of phosphorylation suggests that phosphorylation is detrimental. Mimicking phosphorylation by way of another artificial mutation had the opposite effect. Yeast grew less, and α-synuclein failed to bind the membrane, remaining cytoplasmically diffuse. Once again, this alteration of normal localization points to a negative effect of phosphorylation.
The literature has been relatively split on the role of phosphorylation, leaving the topic open to debate. Some papers have suggested that phosphorylation is protective and that it can curtail α-synuclein toxicity and aggregation in yeast (Shahpasandzadeh et al., 2014). In a worm study, researchers showed that S129 phosphorylation of α-synuclein decreased neurodegeneration by promoting α-synuclein membrane binding (Kuwahara et al., 2012). These findings directly contradict the results from this thesis. Still, other labs have come out with the opposite finding, providing support for my results. One influential paper found that α-synuclein phosphorylation is harmful, inducing Parkinsonian phenotypes in flies (Chen et al., 2005). A more recent in vitro paper found that synthetically S129-phosphorylated α-synuclein exhibits denaturation (as measured by circular dichroism) and faster aggregation compared to WT (Ma et al., 2016). These findings are in line with my conclusion that phosphorylation is harmful.
It is difficult to explain the conflicts in the literature regarding the nature of phosphorylation. One possibility is that even the most effective phosphorylation mimics or blockers do not encompass all of the phosphorylated sites on α-synuclein. Thus, the results of each experiment are different depending on which sites are manipulated and to what extent. Another possibility is that while all studies have focused on serine phosphorylation, tyrosine phosphorylation has been almost completely ignored. It may serve as a confound for these analyses. Regardless of the case, this thesis falls under the camp supporting a harmful role of phosphorylation.
Glycation is Harmful
I also investigated the role of α-synuclein glycation. I hypothesized that glycation is detrimental. My data supported this hypothesis, but the logic for my conclusions was slightly less robust than those for the other modifications. The first line of evidence came from α-synuclein mutations that remove acetylation and glycation at the same time. I found that yeast grow better in the absence of both acetylation and glycation. Since I had already established that acetylation is protective, this improvement in growth must be due to the loss of glycation. I found that removing glycation leads to a more diffuse phenotype, often associated with protective properties. Crucially, the mutant blocking both acetylation and glycation exhibited a drastic reduction in α-synuclein expression. These data suggest that yeast cells benefit when α-synuclein cannot be glycated at the key modification sites.
The second line of evidence came from two glycation-blocking strains, which both exhibited better growth in glucose. This shows that in general, yeast grow better under reduced glycation conditions. The effect carried over when α-synuclein was expressed, so I concluded that glycation may increase α-synuclein’s toxicity. Interestingly, knocking out tpi1 reduced the toxicity of α-synuclein so much that the effects of the modification-manipulating mutations were obliterated. This shows that glycation has a particularly strong effect on α-synuclein. One downside to these investigations is that I did not have any way of investigating hyper-glycation. This makes my data on glycation less convincing than those on acetylation.
The literature on α-synuclein glycation is much less developed than that on acetylation. Some papers suggest that advanced glycation end products (AGEs) can make α-synuclein prone to aggregation (Munch et al., 2000). A more recent publication brings out the hypothesis that anti-diabetes drugs may be helpful against Parkinson’s, as they inadvertently lower glycation of α-synuclein (Koenig et al., 2018). In the future, more research will either confirm or refute the detrimental role of α-synuclein glycation. Glycation is so ubiquitous and complex that mechanisms for its toxicity have not yet been elucidated.
Modifications Counteract Each Other’s Effects
Phosphorylation Counteracts SUMOylation
A major focus of my thesis was on the combinatorial impact of the post-translational modifications of α-synuclein. I first studied the combined impact of SUMOylation and phosphorylation. To my knowledge, this was the first study to investigate the effects of modifying SUMOylation and phosphorylation both at the level of α-synuclein. I hypothesized that SUMOylation and phosphorylation would cooperate to contribute to the properties of α-synuclein. As I expected, the two modifications countered each other, confirming my hypothesis. The first line of evidence was that the two modifications influenced the toxicity of α-synuclein in opposite directions. I found that while removing SUMOylation is toxic, removing phosphorylation counteracts this toxicity and favors a return of α-synuclein to the membrane. Also, this combination leads to the lowest overall expression of the protein. My data suggest that the different modifications of α-synuclein work together to influence the protein’s distribution in the cell. This conclusion is interesting, as in patients, modifications often occur in combinations. For example, one recent study found that erythrocyte-derived α-synuclein was affected by phosphorylation, nitration, and glycation at the same time, and these modifications served as biomarkers for PD (Miranda et al., 2017). Finding support for other combinations of modifications working in tandem, such as SUMOylation and phosphorylation in my case, may be a step forward in furthering scientists’ understanding of Parkinson’s Disease.
Glycation Counteracts Acetylation
I also examined the interdependence of acetylation and glycation. I hypothesized that just as I found for SUMOylation and phosphorylation, there would be a counteraction between acetylation and glycation. Removing acetylation was slightly harmful but removing glycation on top of that counteracted the former manipulation’s effects. This shows that acetylation and glycation work in opposite directions relative to each other. Interestingly, these two modifications occur on shared lysine residues (mainly K6 and K10 for both). Thus, the balance between these two modifications may affect the state of α-synuclein and its propensity to become pathogenic. As emphasized before, it is crucial to study the modifications of α-synuclein in tandem, as they rarely occur separately in patients. The literature is silent on the combined impact of α-synuclein acetylation and glycation.
Other Ways in Which Modifications Influence Each Other’s Effects
In Chapters 3-5, I attempted to uncover interdependences between different combinations of SUMOylation, phosphorylation, acetylation, and glycation. I did this by first studying the SUMOylation/phosphorylation-altering mutants in the acetylation (Chapter 3) and glycation (Chapter 4) altering strains. For the fifth chapter, I reversed the paradigm and studied the acetylation/glycation mutants in SUMOylation altering strains. I found that some interesting interdependences emerged by combining the modifications in these ways. For example, global hyper-acetylation was only protective when α-synuclein phosphorylation was intact. The implications of this result are interesting. While removing phosphorylation and promoting acetylation are by themselves beneficial, when done at the same time, the effect is detrimental. This combinatorial impact is important to keep in mind for future targets of α-synuclein modifications. The individual effects of one modification are not necessarily additive.
I also found that when glycation is blocked at the yeast level, the protective effect of removing phosphorylation is enhanced. Thus, targeting two harmful modifications at the same time reduces the toxicity of α-synuclein and may be a useful strategy in the fight against the protein’s pathogenicity. Investigating the acetylation/glycation mutants in the SUMOylation-altering strains did not yield as many fruitful insights. This may be due to the fact that SUMOylation is such a key modification to many cellular pathways that interrupting it (whether by up- or downregulation) has massive effects that mask those of α-synuclein’s other manipulations.
Modifications Affect Familial Mutants in a Selective Way
In Chapter 6, I found that while most familial mutants exhibited growth similar to BY4741 patterns in hypo-glycating conditions, A53E was an exception. Unlike the other mutants, it did not benefit from the protective nature of hypo-glycation in yeast. These results indicate that glycation differentially impacts the toxicity of the familial mutants. Perhaps in patients with the A53E mutation, glycation is even more harmful than it is in those diagnosed with sporadic PD. The literature is largely silent on the combined impact of α-synuclein glycation and familial mutations. Perhaps predictive structural biology could show why α-synuclein carrying the A53E mutation does not benefit from hypo-glycation. One safe prediction is that the protein adopts a more stable non-aggregatable conformation. I am currently in the process of expanding this pilot investigation. My next steps are to spot the familial mutants in sirt2∆ and ask whether hyper-acetylation differentially impacts the familial mutants. It may also be instructive to spot the mutants in BY4741, sirt2∆, and glo1∆ on the same plate to make direct growth comparisons.
Explanation of Microscopy Phenotypes
I observed three main phenotypes as I was tracking α-synuclein throughout time – membrane binding, clumping, and cytoplasmic diffusion. While there are also combinatorial phenotypes, they are so rare that they do not contribute much to the statistical analysis. While it is difficult to definitively assign characteristics to phenotypes, some trends undeniably emerge. WT α-synuclein binds the cell membrane by 24 hours. This phenotype is the default in budding yeast, but it is pathological (along with aggregation) in humans. Perhaps this is consistent with the fact that even WT α-synuclein is toxic to yeast. Often, α-synuclein clumps up in distinct foci or puncta. These vary in size and number by manipulation. One possibility is that these puncta are analogous to Lewy bodies. This cannot be confirmed using just fluorescence microscopy as it does not allow us to determine whether or not the protein is fibrillized in a beta-sheet conformation. Nevertheless, the formation of foci usually correlates with less growth in the spotting assay. Other times, α-synuclein is diffuse throughout the cytoplasm. Usually, diffusion correlates with better growth (especially in the A30P and G51D familial mutants as in Jones, Senior Thesis, 2018). This is paradoxical, as the manipulations that end up diffuse (and growing better) are highly aggressive in humans. One possibility is that the intracellular transport of α-synuclein breaks down completely as a result of the protein’s increased pathogenicity. In yeast, this breakdown may prevent the protein from reaching organelles where it causes toxicity. These conclusions show that while microscopy phenotypes can be instructive, it is important to consider them in the context of toxicity and expression data to draw conclusions.
Strengths of the Study
One of the major strengths of this study is that many replications were performed for each experiment. The yeast model allows for a quick experimental turnover, allowing me to repeat each experiment up to five times (sometimes even more). This large number of replications increases my confidence in the data and gives me the opportunity to report only the trends that appear in multiple trials. Another strength of the study is that yeast have an easily manipulatable genome and can accept plasmids that carry extra-genomic genes. This allows me to induce yeast to produce α-synuclein and also to perform global protein knockouts. Such flexibility makes the study of the combinatorial impact of post-translational modifications possible. Additionally, in my experiments, all microscopy data are backed up by statistical analyses. This step is rarely taken in other yeast labs, but it allows me to increase my confidence in the trends I report. From a molecular biology perspective, this study provides basic data on the functional consequences of combinations of post-translational modifications, a novel perspective in Parkinson’s research. Yeast make, fold, and degrade proteins like humans do. This study gives adequate preliminary data on the functions of the modifications in an in vivo model.
Limitations of the Study
While this research provides a good overview of the preliminary functions of the chosen modifications, there are multiple limitations to the scope of the conclusions. For example, dopamine release, neurotransmitter balance, and locomotor behavior cannot be evaluated in a single-celled organism such as yeast. All of these are crucial to the study of Parkinson’s Disease and would create a better connection of our manipulations to the symptoms of the illness. There are some methodological weaknesses in the study as well. For example, all of the toxicity data are based on qualitative evaluations of growth. There is no efficient way to quantify spotting results. I cannot back my toxicity claims with statistics, and it is difficult to make accurate predictions about growth rates. Thus, I cannot determine whether one toxic sample is more or less toxic than another, unless there is an extreme phenotype that makes the answer obvious. One could perform survival assays or optical density analyses to obtain quantitative toxicity results. In my microscopy analyses, while I could track and quantify changes in phenotype, I could not be sure whether these changes reflect a transition towards the characteristics of Lewy bodies. For example, even though I describe some phenotypes as “aggregated” using my analyses, it is impossible to tell whether or not the α-synuclein is fibrillized as it is in Lewy bodies. Biochemical assays (like proteinase K digestion and the thioflavin T assay) or electron microscopy would provide more evidence on the nature of the modified α-synuclein. Yet another criticism is that my Western blots exhibit lots of degradation. Also crucially, the position of the α-synuclein band in the Western blot does not change when modifications are altered. For example, the SUMO protein is heavy, contributing 15 to 17 kDa to the molecular weight of its substituent (Sarge & Sarge, 2008). However, in my Western blots, WT α-synuclein and the variants that block SUMOylation appear at the same molecular size. There are two possible explanations for this discrepancy. The SUMO protein may be getting cleaved off when the lysates are prepared and all proteins are denatured using SDS. Another possibility is that other lysine amino acids on α-synuclein are getting SUMOylated to a large extent, making the loss of two SUMOylation sites insignificant to the molecular weight (but still crucial to the function of α-synuclein). These criticisms show that for more convincing arguments, it is important to consider data from different model organisms and varied experimental assays.
Sources of Unexpected Results
The results of this project mostly support my hypotheses, although some of the trends are not as strong as I initially hoped. The most unexpected result occurred in the smt3ts and ulp1ts strains, at 30 °C in glucose. Under these conditions, the smt3ts strain promotes cell-wide deficient SUMOylation, while the ulp1ts strain promotes cell-wide hyper-SUMOylation. In previous theses, under the same conditions, smt3ts grows much less than BY4741, while ulp1ts grows much better than BY4741. These trends were not observed in this thesis, with both strains growing equally as well as BY4741 when α-synuclein expression is not induced. There are several possible explanations of the differences in these reported trends. For instance, the experiments reported in this project were performed during the winter, and it may have been hard for the incubators to reach the full temperature (30 °C) required to activate the enzyme mutations. Thus, the effects of the temperature-sensitive mutations may have been weaker. Alternatively, the temperature-sensitive mutant enzymes may have been damaged during the transformation process, weakening our manipulation. This is highly unlikely, but it still represents a possible scenario.
The research presented in this thesis can take many directions in the future. For example, the KQ, KR, and KA mutant sets could be studied in other knockout strains to glean out other interdependences between modifications. It may be interesting to study these mutants in the nitration-altering strains cox5a and cox5b. These strains mimic and block nitration, respectively, and it may be instructive to investigate how the acetylation and glycation mutants behave in them.
Another possibility is to repeat all of the experiments described in this thesis in fission yeast. Our lab was among the first to start the fission yeast model of α-synuclein (Brandis et al., 2006). Other students have found that fission yeast often highlight different aspects of PD pathology than budding yeast. For example, while in budding yeast, WT α-synuclein tends to bind to the membrane, in fission yeast, it forms aggregates in the cytoplasm. Given this key difference in the baseline condition, it may be the case that different phenotypes will emerge with the mutants modifying acetylation and glycation.
Another study in yeast that would be interesting to perform would be to investigate the effects of acetylation and glycation on other members of the family, β-synuclein and γ-synuclein. These two proteins are closely related to α-synuclein, and their role in neurodegeneration has only recently been acknowledged (Taschenberger et al., 2013). Whether these additional members of the synuclein family are affected by these modifications (and whether the functions of the modifications are preserved) remains to be discovered. If SUMOylation, phosphorylation, acetylation, and glycation have similar effects in β-synuclein and γ-synuclein as they do in α-synuclein, this would strengthen the notion that post-translational modifications are pivotal mediators of neurodegenerative mechanisms.
Finally, expanding away from yeast models, it may be interesting to investigate SUMOylation, phosphorylation, acetylation, and glycation in a mouse model of PD (such as the 6-OHDA model). This would allow us to verify our conclusions from yeast in a eukaryotic multicellular organism and would also allow us to ask questions related to pathology, locomotor behavior, and dopamine levels. Worm models and fly models would also yield valuable behavioral data, in addition to providing easier genetic manipulation than mice. Induced pluripotent stem cell lines lines could be used to track amounts of post-translational modifications over time. Regardless of the model system, this thesis, as shown in Figure 38, poses the question of whether a balance of post-translational modifications drives α-synuclein from a healthy to a diseased conformation.
In this thesis, I investigated the individual and combinatorial effects of four post-translational modifications on α-synuclein. I found that SUMOylation and acetylation are both protective, while phosphorylation and glycation are both harmful. SUMOylation and phosphorylation counteract each other, as do acetylation and glycation. In addition, acetylation and phosphorylation (as well as glycation and phosphorylation) may have additive effects that do not necessarily conform to the individual effects of these modifications.
These findings were all derived from experiments in budding yeast, a powerful model for investigations of α-synuclein’s pathogenic properties. Yeast divide quickly, making it possible to perform many replications of the experiments in a short time. Just like humans, yeast make, fold, and degrade proteins, making them a suitable platform for studying neurodegenerative diseases at the protein processing level. Finally, yeast tend to magnify phenotypes that are not readily apparent in other model systems. It is yeast models that first demonstrated that α-synuclein can both bind membranes and aggregate, leading to the later conclusion that both phenotypes may be pathogenic.
This thesis presented several ideas that had not been investigated before. For example, acetylation and glycation of α-synuclein had not been studied in yeast before these experiments. Studying combinations of modifications (both at the α-synuclein level and the strain level) is a novel experimental undertaking that has yet to be repeated in other labs (potentially using models other than yeast).
Returning to the big picture, it is important to study the molecular basics of Parkinson’s Disease. Ultimately, the functions and phenotypes of all living cells are determined by the proteins expressed in them. Thus, targeting cellular dysfunction is equivalent to targeting dysfunctional proteins. The post-translational modifications of α-synuclein provide a novel treatment target for the disease. By investigating post-translational effects in a combinatorial fashion, we can pinpoint the groups of modifications that can be simultaneously manipulated to yield the most protective effects or avoid the most harmful interactions.
Studying Parkinson’s Disease is an important endeavor with the potential to help many people. The PD population is very large – over 60,000 people are diagnosed in the United States each year, with the worldwide figure skyrocketing to 6-7 million new patients annually. Solving the problem of PD is an urgent issue that affects the global community. By performing basic experiments that can be translated into new therapies, molecular biology labs can contribute their individual pieces to the puzzle that is PD research.
Abeywardana, T., and Pratt, M. R. (2015). Extent of inhibition of α-synuclein aggregation in vitro by SUMOylation is conjugation site-and SUMO isoform-selective. Biochemistry, 54, 959-961.
Alberts, B. (1998). The cell as a collection of protein machines: preparing the next generation of molecular biologists. Cell, 92, 291-294.
Alberts, B. (2002). The RNA World and the Origins of Life. In Molecular Biology of the Cell, 4th Edn. (New York: Garland Science).
Allendoerfer, A. L., Su., and L. J., Lindquist, S. (2008). Yeast cells as a discovery platform for Parkinson’s disease. In Parkinson’s Disease: Molecular and Therapeutic Insights from Model Systems, Przedborski, S., ed. pp. 505-536. (Cambridge, UK: Academic Press).
Barker, R. A., and Williams‐Gray, C. H. (2016). The spectrum of clinical features seen with alpha synuclein pathology. Neuropathology and Applied Neurobiology. 42, 6-19.
Beyer, K., and Ariza, A. (2013). Alpha-synuclein posttranslational modification and alternative splicing as a trigger for neurodegeneration. Molecular Neurobiology. 47, 509-524.
Bird, T. D. (2016). Hereditary ataxia overview. GeneReviews. Seattle VA Medical Center. Online: Retrieved from: https://www.ncbi.nlm.nih.gov/books/NBK1138/
Brandis, K. A., Holmes, I. F., England, S. J., Sharma, N., Kukreja, L., & DebBurman, S. K. (2006). α-synuclein fission yeast model. Journal of Molecular Neuroscience, 28, 179-191.
Burré, J., Sharma, M., Tsetsenis, T., Buchman, V., Etherton, M., and Südhof, T. C. (2010). α-Synuclein promotes SNARE-complex assembly in vivo and in vitro. Science, 1195227.
Chai, C., and Lim, K. L. (2013). Genetic insights into sporadic Parkinson’s disease pathogenesis. Current Genomics, 14, 486-501.
Chen, L. I., and Feany, M. B. (2005). α-Synuclein phosphorylation controls neurotoxicity and inclusion formation in a Drosophila model of Parkinson disease. Nature Neuroscience, 8, 657.
Chen, Q., Thorpe, J., and Keller, J. N. (2005). α-synuclein alters proteasome function, protein synthesis, and stationary phase viability. Journal of Biological Chemistry, 280, 30009-30017.
de Oliveira, R. M., Miranda, H. V., Francelle, L., Pinho, R., Szegö, É. M., Martinho, R., … and Fonseca, L. (2017). The mechanism of sirtuin 2–mediated exacerbation of alpha-synuclein toxicity in models of Parkinson disease. PLoS Biology, 15, e2000374.
Del Tredici, K., and Braak, H. (2016). Sporadic Parkinson’s disease: development and distribution of α‐synuclein pathology. Neuropathology and Applied Neurobiology, 42, 33-50.
Ebrahimi-Fakhari, D., McLean, P. J., and Unni, V. K. (2012). Alpha-synuclein’s degradation in vivo: opening a new (cranial) window on the roles of degradation pathways in Parkinson disease. Autophagy, 8, 281-283.
Engelender, S., Kaminsky, Z., Guo, X., Sharp, A. H., Amaravi, R. K., Kleiderlein, J. J., … and Dawson, V. L. (1999). Synphilin-1 associates with α-synuclein and promotes the formation of cytosolic inclusions. Nature Genetics, 22, 110.
Fahn, S. (2008). Clinical aspects of Parkinson’s disease. In Parkinson’s Disease: Molecular and Therapeutic Insights from Model Systems, Przedborski, S., ed. pp. 505-536. (Cambridge, UK: Academic Press).
Fares, M. B., Ait-Bouziad, N., Dikiy, I., Mbefo, M. K., Jovičić, A., Kiely, A., … and Lashuel, H. A. (2014). The novel Parkinson’s disease linked mutation G51D attenuates in vitro aggregation and membrane binding of α-synuclein, and enhances its secretion and nuclear localization in cells. Human Molecular Genetics, 23, 4491-4509.
Fiske, M., Valtierra, S., Solvang, K., Zorniak, M., White, M., Herrera, S., … and DebBurman, S. (2011). Contribution of alanine-76 and serine phosphorylation in α-synuclein membrane association and aggregation in yeasts. Parkinson’s Disease, 2011.
Fedorow, H., Tribl, F., Halliday, G., Gerlach, M., Riederer, P., and Double, K. L. (2005). Neuromelanin in human dopamine neurons: comparison with peripheral melanins and relevance to Parkinson’s disease. Progress in Neurobiology, 75, 109-124.
Flower, T. R., Chesnokova, L. S., Froelich, C. A., Dixon, C., and Witt, S. N. (2005). Heat shock prevents alpha-synuclein-induced apoptosis in a yeast model of Parkinson’s disease. Journal of Molecular Biology, 351, 1081-1100.
Garcia-Esparcia, P., López-González, I., Grau-Rivera, O., García-Garrido, M. F., Konetti, A., Llorens, F., … and Gelpi, E. (2017). Dementia with Lewy bodies: molecular pathology in the frontal cortex in typical and rapidly progressive forms. Frontiers in Neurology, 8, 89.
Gitler, A. D., Bevis, B. J., Shorter, J., Strathearn, K. E., Hamamichi, S., Su, L. J., … and Barlowe, C. (2008). The Parkinson’s disease protein α-synuclein disrupts cellular Rab homeostasis. Proceedings of the National Academy of Sciences, 105, 145-150.
Goldman, J. G., Vernaleo, B. A., Camicioli, R., Dahodwala, N., Dobkin, R. D., Ellis, T., … and Golden, R. (2018). Cognitive impairment in Parkinson’s disease: a report from a multidisciplinary symposium on unmet needs and future directions to maintain cognitive health. Nautre - npj Parkinson’s Disease, 4, 19.
Hughes, R. C. (1994). Parkinson’s Disease and its Management. BMJ: British Medical Journal, 308, 281.
Iwasaki, S., Narabayashi, Y., Hamaguchi, K., Iwasaki, A., and Takakusagi, M. (1990). Cause of death among patients with Parkinson’s disease: a rare mortality due to cerebral haemorrhage. Journal of Neurology, 237, 77-79.
Iyer, A., Roeters, S. J., Schilderink, N., Hommersom, B., Heeren, R. M., Woutersen, S., … and Subramaniam, V. (2016). The impact of N-terminal acetylation of α-synuclein on phospholipid membrane binding and fibril structure. Journal of Biological Chemistry, 291, 21110-21122.
Jones, P. (2018) Differential impact of multiple Parkinson’s Disease associated genes on the toxicity of alpha-synuclein in a yeast model. Senior Thesis (Lake Forest, IL: Lake Forest College).
Krumova, P., Meulmeester, E., Garrido, M., Tirard, M., Hsiao, H. H., Bossis, G., … & Bähr, M. (2011). Sumoylation inhibits α-synuclein aggregation and toxicity. The Journal of Cell Biology. 194, 49-60.
Koenig, A., Vicente Miranda, H., and Outeiro, T. F. (2018). Alpha-Synuclein Glycation and the Action of Anti-Diabetic Agents in Parkinson’s Disease. Journal of Parkinson’s Disease, 8, 33-43.
Kunadt, M., Eckermann, K., Stuendl, A., Gong, J., Russo, B., Strauss, K., … and Krumova, P. (2015). Extracellular vesicle sorting of α-Synuclein is regulated by sumoylation. Acta Neuropathologica, 129, 695-713.
Kuwahara, T., Tonegawa, R., Ito, G., Mitani, S., and Iwatsubo, T. (2012). Phosphorylation of α-synuclein protein at Ser-129 reduces neuronal dysfunction by lowering its membrane binding property in Caenorhabditis elegans. Journal of Biological Chemistry, 287, 7098-7109.
Lashuel, H. A., Overk, C. R., Oueslati, A., and Masliah, E. (2013). The many faces of α-synuclein: from structure and toxicity to therapeutic target. Nature Reviews Neuroscience, 14, 38.
Lima, V. D. A., do Nascimento, L. A., Eliezer, D., and Follmer, C. (2018). Role of Parkinson’s Disease-Linked Mutations and N-Terminal Acetylation on the Oligomerization of α-Synuclein Induced by 3, 4-Dihydroxyphenylacetaldehyde. ACS Chemical Neuroscience, 10, 690-703.
Lin, X., Shi, M., Masilamoni, J. G., Dator, R., Movius, J., Aro, P., … and Zhang, J. (2015). Proteomic profiling in MPTP monkey model for early Parkinson disease biomarker discovery. Biochimica et Biophysica Acta (BBA)-Proteins and Proteomics, 1854, 779-787.
Lin, D., Jing, X., Chen, Y., Liang, Y., Lei, M., Peng, S., … and Yang, L. (2017). Rifampicin pre-treatment inhibits the toxicity of rotenone-induced PC12 cells by enhancing sumoylation modification of α-synuclein. Biochemical and Biophysical Research Communications, 485, 23-29.
Ma, M. R., Hu, Z. W., Zhao, Y. F., Chen, Y. X., and Li, Y. M. (2016). Phosphorylation induces distinct alpha-synuclein strain formation. Scientific Reports, 6, 37130.
Marengo-Rowe, A. J. (2006, July). Structure-function relations of human hemoglobins. In Baylor University Medical Center Proceedings (Vol. 19, No. 3, pp. 239-245). Taylor & Francis.
Mazzoni, P., Shabbott, B., and Cortés, J. C. (2012). Motor control abnormalities in Parkinson’s disease. Cold Spring Harbor Perspectives in Medicine, a009282.
McCann, H., Stevens, C. H., Cartwright, H., and Halliday, G. M. (2014). α-Synucleinopathy phenotypes. Parkinsonism & Related Disorders, 20, S62-S67.
Miranda, H. V., Cássio, R., Correia-Guedes, L., Gomes, M. A., Chegão, A., Miranda, E., … and Outeiro, T. F. (2017). Posttranslational modifications of blood-derived alpha-synuclein as biochemical markers for Parkinson’s disease. Scientific Reports, 7, 13713.
Mufson, E. J., Counts, S. E., Perez, S. E., and Ginsberg, S. D. (2008). Cholinergic system during the progression of Alzheimer’s disease: therapeutic implications. Expert Review of Neurotherapeutics, 8, 1703-1718.
Münch, G., Lüth, H. J., Wong, A., Arendt, T., Hirsch, E., Ravid, R. A., and Riederer, P. (2000). Crosslinking of α-synuclein by advanced glycation end products—an early pathophysiological step in Lewy body formation. Journal of Chemical Neuroanatomy, 20, 253-257.
Narhi, L., Wood, S. J., Steavenson, S., Jiang, Y., Wu, G. M., Anafi, D., … and Louis, J. C. (1999). Both familial Parkinson’s disease mutations accelerate α-synuclein aggregation. Journal of Biological Chemistry, 274, 9843-9846.
Orr-Weaver, T. L., Szostak, J. W., and Rothstein, R. J. (1983). Genetic applications of yeast transformation with linear and gapped plasmids. Methods in Enzymology, 101, 228-245.
Outeiro, T. F., Lindquist, S. (2003). Yeast cells provide insight into α-synuclein biology and pathobiology. Science, 302, 1775-1779.
Park J., Al-Ramahi I., Tan Q., Mollema N., Diaz-Garcia J. R., and Gallego-Flores T. (2013). RAS-MAPK-MSK1 pathway modulates ataxin 1 protein levels and toxicity in SCA1. Nature, 498, 325–331.
Park-Sarge, O. K., and Sarge, K. D. (2008). Detection of sumoylated proteins. In The Nucleus (pp. 255-265). Humana Press, Totowa, NJ.
Petrucci, S., Ginevrino, M., and Valente, E. M. (2016). Phenotypic spectrum of alpha-synuclein mutations: New insights from patients and cellular models. Parkinsonism & Related Disorders, 22, S16-S20.
Petrucelli, L. Dickson, D. W. (2008). Neurobiology of Parkinson’s disease. In Parkinson’s Disease: Molecular and Therapeutic Insights from Model Systems, Przedborski, S., ed. pp. 505-536. (Cambridge, UK: Academic Press).
Ramig, L. O., Fox, C., and Sapir, S. (2008). Speech treatment for Parkinson’s disease. Expert Review of Neurotherapeutics, 8, 297-309.
Rizek, P., Kumar, N., and Jog, M. S. (2016). An update on the diagnosis and treatment of Parkinson disease. Canadian Medical Association Journal, 188, 1157-1165.
Ross, O. A., Braithwaite, A. T., Farrer, M. J. (2008). Genetics of Parkinson’s disease. In Parkinson’s Disease: Molecular and Therapeutic Insights from Model Systems, Przedborski, S., ed. pp. 505-536. (Cambridge, UK: Academic Press).
Sabuncu, M. R., Desikan, R. S., Sepulcre, J., Yeo, B. T. T., Liu, H., Schmansky, N. J., … and Fischl, B. (2011). The dynamics of cortical and hippocampal atrophy in Alzheimer disease. Archives of Neurology, 68, 1040-1048.
Schmid, A. W., Fauvet, B., Moniatte, M., and Lashuel, H. A. (2013). Alpha-synuclein post-translational modifications as potential biomarkers for Parkinson disease and other synucleinopathies. Molecular & Cellular Proteomics, 12, 3543-3558.
Seidel, K., Siswanto, S., Fredrich, M., Bouzrou, M., den Dunnen, W. F., Özerden, I., … and Auburger, G. (2017). On the distribution of intranuclear and cytoplasmic aggregates in the brainstem of patients with spinocerebellar ataxia type 2 and 3. Brain Pathology, 27, 345-355.
Shahpasandzadeh, H., Popova, B., Kleinknecht, A., Fraser, P. E., Outeiro, T. F., and Braus, G. H. (2014). Interplay between sumoylation and phosphorylation for protection against α-synuclein inclusions. Journal of Biological Chemistry, 289, 31224-31240.
Sharma, N., Brandis, K.A., Herrera, S.K. et al. (2006). Alpha-synuclein budding yeast model. Toxicity enhanced by impaired proteasome and oxidative stress. Journal of Molecular Neuroscience, 28, 161-178.
Singleton, A. B., Farrer, M., Johnson, J., Singleton, A., Hague, S., Kachergus, J., … and Lincoln, S. (2003). α-Synuclein locus triplication causes Parkinson’s disease. Science, 302, 841-841.
Snead, D., Eliezer, D. (2014). α-synuclein function and dysfunction on cellular membranes. Experimental Neurobiology, 23, 292-313.
Solvang, K. (2011) Evaluating the role of post-translational modifications of alpha-synuclein in yeast models. Senior thesis (Lake Forest, IL: Lake Forest College)
Spillantini, M. G., Schmidt, M. L., Lee, V. M. Y., Trojanowski, J. Q., Jakes, R., and Goedert, M. (1997). α-Synuclein in Lewy bodies. Nature, 388, 839.
Spinelli, K. J., Taylor, J. K., Osterberg, V. R., Churchill, M. J., Pollock, E., Moore, C., … and Unni, V. K. (2014). Presynaptic alpha-synuclein aggregation in a mouse model of Parkinson’s disease. Journal of Neuroscience, 34, 2037-2050.
Taschenberger, G., Toloe, J., Tereshchenko, J., Akerboom, J., Wales, P., Benz, R., … and Zweckstetter, M. (2013). β‐synuclein aggregates and induces neurodegeneration in dopaminergic neurons. Annals of Neurology, 74, 109-118.
Tembo, M. (2015). Characterization of new familial mutants of Parkinson’s disease protein ɑ-synuclein using yeast models. Senior thesis (Lake Forest, IL: Lake Forest College).
Thomas, R. (2018) Impact of Parkinson’s Disease modifications on alpha-synuclein toxicity in a yeast model: SUMOylation and phosphorylation. Senior thesis (Lake Forest, IL: Lake Forest College)
Toal, S., Trexler, A., DeWitt, D., Brown, M., and Rhoades, E. (2017). Determining the Role of N-Terminal Acetylation on α-Synuclein Function. Biophysical Journal, 112, 366a.
Tysnes, O. B., and Storstein, A. (2017). Epidemiology of Parkinson’s disease. Journal of Neural Transmission, 124, 901-905.
Vamvaca, K, Volles, M. J., Lansbury, P. T. (2009) The first N-terminal amino acids of α-Synuclein are essential for α-Helical structure formation in vitro and membrane binding in yeast. Journal of Molecular Biology, 389, 413-424.
Vicente Miranda, H., Szegő, É. M., Oliveira, L., Breda, C., Darendelioglu, E., de Oliveira, R. M., … and Munari, F. (2017). Glycation potentiates α-synuclein-associated neurodegeneration in synucleinopathies. Brain, 140, 1399-1419.
Wakabayashi, K., & Takahashi, H. (2006). Cellular pathology in multiple system atrophy. Neuropathology, 26, 338-345.
Warby S. C., Chan E. Y., Metzler M., Gan L., Singaraja R. R., Crocker S. F., et al. (2005). Huntingtin phosphorylation on serine 421 is significantly reduced in the striatum and by polyglutamine expansion in vivo. Hum. Mol. Genet., 14, 1569–1577.
Yu, Z., Xu, X., Xiang, Z., Zhou, J., Zhang, Z., Hu, C., and He, C. (2010). Nitrated α-synuclein induces the loss of dopaminergic neurons in the substantia nigra of rats. PloS One, 5, e9956.
Zhu, L. N., Qiao, H. H., Chen, L., Sun, L. P., Hui, J. L., Lian, Y. L., … and Qiu, P. M. (2018). SUMOylation of Alpha-Synuclein Influences on Alpha-Synuclein Aggregation Induced by Methamphetamine. Frontiers in Cellular Neuroscience, 12.
- Fungi: https://askopinion.com/what-do-lizards-eat
- Monera: https://biology.tutorvista.com/organism/kingdom-monera.html
- Protista: https://bio.rutgers.edu/~gb102/lab_1/2i2m-protista.html
- Plantae: https://www.studyblue.com/notes/note/n/kingdom-plantae/deck/14437075
- Animalia: https://twitter.com/savethebelugas
- α-Synuclein good shape: http://file.scirp.org/Html/6-8201833_24987.htm
- α-Synuclein bad shape: http://www.rcsb.org/pdb/results/results.do?qrid=551B9603&tabtoshow=Current
- Same protein images also appear in Figures 4, 5, and 38
- Alzheimer’s: https://s-media-cache-ak0.pinimg.com/564x/5c/14/9c/5c149c751c774deb37ce162508a6f837.jpg
- SCA: http://neuropathology-web.org/chapter9/images9/9-13.jpg
- Prion Diseases: http://media.mnn.com/assets/images/2010/01/prions.jpg.560x0_q80_crop-smart.jpg
- Synucleinopathies: http://file.scirp.org/Html/6-8201833%5C18ed12be-b1c8-409a-961d-adf640dc48e9.jpg
- PD: http://neuropathology-web.org/chapter9/chapter9dPD.html
- DLB: https://consultqd.clevelandclinic.org/research-collaborative-launched-to-advance-progress-against-lewy-body-dementia/
- MSA: https://en.wikipedia.org/wiki/Multiple_system_atrophy
- Stick figure: https://www.shutterstock.com/image-vector/male-figure-silhouette-icon-656542672
- PD brain: https://beyondthedish.wordpress.com/2012/12/27/stem-cells-from-your-nose-to-treat-parkinsons-disease/substantia-nigra/
- Lewy body: Spillantini et al., 1997
- Substantia nigra: https://www.researchgate.net/post/Could_anyone_tell_me_the_way_to_count_TH_stained_Domapinergic_neurons_in_Substantia_nigra
Part A1 – SUMOylation is Protective
The work presented in this thesis hinges on the experiments of past thesis students. My mentor in lab, Rosemary Thomas, showed that SUMOylation is protective. Figure A1 shows a representative spotting from her data. She showed that WT α-synuclein is toxic relative to controls, and it becomes more toxic when SUMOylation is blocked. The double SUMOylation blocker (K96R/K102R) has the most powerful effect. Since yeast grow less without SUMOylation, Rosemary concluded that SUMOylation is protective. Localization and expression data support these conclusions. When SUMOylation is blocked, α-synuclein forms foci, and expression levels are slightly higher. This work justified the hypothesis that SUMOylation is protective.
Part A2 – Chapter 1 Replicability
One major objection to work with yeast cells is that the data are often difficult to replicate and that conclusions are based on trends that appear slight. To counter these arguments against the replicability of the experiments, I have provided additional samples of some of the work presented in this thesis, coming from different trials. Figure A2 shows two separate spotting replicates (labeled a and b), corroborating the conclusions drawn in Chapter 1 (reported in Figure 10).
Figure A3 shows a Western blot replicating the results reported in Figure 13 of Chapter 1. These results closely match the conclusions of other trials, showing that the trends I reported are replicable.
Part B – Chapter 2 Replicability
I provide a similar set of separate trials for Chapter 2 as I did for Chapter 1. Figure B1 shows two alternative spotting trials (labeled a and b) for the acetylation/glycation mutants in BY4741. As in Chapter 2 (Figure 14), the K-Q and K-A mutant sets are protective.
Figure B2 shows an alternative Western blot trial. As in Figure 17, K6A/K10A exhibits a striking reduction in expression.
Part C – Chapter 3 Replicability
Figure C1 shows a different spotting trial from the experiment manipulating SUMOylation and phosphorylation in sirt2∆. In agreement with Figure 18, a hyper-acetylating environment does not promote healthy growth when SUMOylation and phosphorylation are blocked at the same time.
Figure C2 shows a replication of my Western blot for this section, in which I found that the mutant blocking SUMOylation and phosphorylation together exhibited the highest expression.
Part D – Chapter 4 Replicability
Figures D1 and D2 show replications of my spottings in glo1∆ and tpi1∆. The trend for both hypo-glycating strains was the same as presented. Removing glycation has a beneficial effect on the growth of yeast even in the presence of α-synuclein.
Figure D3 shows a representative Western blot from both strains. The results are consistent with the report from the main body of the thesis. In glo1∆ and tpi1∆, overexpression of α-synuclein occurs with the SUMOylation/phosphorylation double blocker and with the SUMOylation blocker/phosphorylation mimic.
Part E – Chapter 5 Replicability
In analogy to the appendices for the previous chapters, I show replications of my spottings and Western blots for Chapter 5. The acetylation/glycation mutants in smt3ts and ulp1ts at 30 degrees exhibited the same growth (Figure E1) and expression patterns (Figure E2) as reported in the main body of the text.
Part F – Chapter 6 Replicability
Our results for Chapter 6 are reliable, as separate trials give the same overall trend. See Figure F1. In glo1∆, all of the mutants preserved their profiles as in BY4741, except for A53E, which became slightly toxic.
Eukaryon is published by students at Lake Forest College, who are solely responsible for its content. The views expressed in Eukaryon do not necessarily reflect those of the College.
Articles published within Eukaryon should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.