A new era in research on Friedreich’s ataxia, a rare, fatal neurodegenerative disease, began in 1996 when, 133 years after the disease was first characterized, researchers showed it is caused by mutations in the gene now known as FXN. While this finding was an important advance, it also presented researchers with the daunting task of determining why neurons with FXN mutations were dying in Friedreich’s ataxia patients. To meet this challenge, a team at the University of Utah turned to an unlikely source: the baker’s yeast Saccharomyces cerevisiae. A mere 15 months later, using genome editing, growth assays, and biochemical techniques, the Utah team demonstrated that FXN mutations cause fatal mitochondrial damage. This finding identified an important therapeutic target, and clinical trials have recently demonstrated that two drugs targeting mitochondrial function improve symptoms in Friedreich’s ataxia patients.
This serves as one of the earliest examples in which yeast, a single-celled fungus with a one-week lifespan, provided actionable insights into a complex, age-related disorder of the billions of interconnected neurons of the human brain. But remarkably, such stories are common in neurodegenerative disease research, and not just for rare diseases like Friedreich’s ataxia, but for more common diseases as well, including Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis (ALS). Yeast have long been used to understand the fundamental cellular, molecular, and genetic features common to all forms of eukaryotic life and how disrupting these features leads to disease. After S. cerevisiae became the first eukaryotic organism to have its genome fully sequenced in 1996, researchers determined that 60 percent of yeast genes have a human homolog. This means that humans and yeast share not only a common set of cellular structures and functions, but also a set of genetic instructions that encodes them. Neurodegenerative diseases are likely caused by genetic and environmental factors that impair fundamental biological processes conserved across eukaryotic life, and yeast can serve as a valuable model for understanding the roles of these factors in disease and developing interventions to counter their deleterious effects.
Humans and yeast share not only a common set of cellular structures and functions, but also a set of genetic instructions that encodes them.
Yeast provide neurodegenerative disease researchers with unique advantages compared to other model systems. First, yeast cells grow faster and can be maintained at a fraction of the cost of other eukaryotic organisms. This allows researchers to complete experiments faster and at greater scale than in other systems. Moreover, yeast growth provides a simple and reliable indication of cellular health. In the context of neurodegenerative disease research, this provides a metric to decipher the environmental and genetic factors that cause neuronal death and to screen therapeutics intended to treat or prevent these diseases. And arguably the greatest advantage of yeast is the ease with which it can be genetically modified. Yeast, more than any other eukaryotic organism, readily incorporates foreign DNA into its own genome by the process of transformation. This enables researchers to add and remove native yeast genes, incorporate DNA from other species into yeast, and control gene expression, among many other applications.
These advantages have powered important discoveries in neurodegenerative disease research over the past four decades (see boxes). Early yeast models of neurodegenerative diseases helped researchers characterize the then-unknown functions of disease-linked genes, as in Friedreich’s ataxia. Throughout the 2000s, researchers used large yeast strain collections to systematically search for genes that modify disease-associated pathologies, leading to new insights into the mechanisms of numerous neurodegenerative diseases and revealing new therapeutic targets. More-recent efforts pair the massive scale of modern yeast genetics with powerful artificial intelligence (AI) algorithms to disentangle the complex, multifactorial causes of neurodegeneration. The many insights into neurodegenerative diseases revealed by yeast are remarkable examples of how combining simple model organisms with cutting-edge technology can provide unique insights into complex biological systems.
MODELING NEURODEGENERATIVE DISEASES WITH YEAST
Neurodegenerative diseases kill cells by disrupting basic biological processes shared by species as diverse as humans and yeast. The conservation of cellular structures and functions across eukaryotic life allows researchers to study the genetic and molecular underpinnings of neurodegenerative diseases in various model organisms. Yeast cells are a particularly valuable system, as the fungi grow rapidly, are inexpensive to maintain, and can be genetically modified more readily than any other eukaryote. By introducing mutations linked to certain brain diseases, researchers have made yeast models that lend an unmatched speed and scale to research on neurodegeneration. Here are some examples of disease mechanisms that researchers have begun to untangle using yeast.
When proteins are damaged or misfolded, they often stick together, forming clumps known as protein aggregates. If not removed by the cell, protein aggregates increase in size and number, impairing a variety of cellular functions and ultimately killing cells.
Yeast research helped establish the predominant role of protein aggregation in numerous neurodegenerative diseases, including Alzheimer’s disease, Parkinson’s disease, ALS, dementia, and Huntington’s disease. Yeast have also become an important platform for identifying and screening therapies capable of preventing or reversing protein aggregation.
Cells move molecules into and out of the nucleus for many purposes, such as exporting mRNAs following transcription and importing newly-synthesized nuclear proteins.
Yeast research helped establish that nucleocytoplasmic transport is impaired in ALS and dementia. A Biogen clinical trial is testing a new drug that improves this process.
Mitochondria generate the majority of a cell’s energy and regulate its metabolism. Because neurons consume large amounts of energy, they are uniquely vulnerable to defects in mitochondrial function.
Yeast research established that the mutations that cause Friedreich’s ataxia and Mohr-Tranebjarg syndrome kill cells by damaging mitochondria and that aberrant activity of several proteins, such as the ALS-linked protein SOD1, impairs mitochondrial function, pointing to new therapeutic targets for these diseases.
Cells shuttle membrane-bound vesicles to the plasma membrane to insert transmembrane proteins and secrete vesicular contents. This vesicular trafficking is especially critical for the health and function of neurons, which use this process to release neurotransmitters during synaptic transmission.
Protein quality control pathways
Damaged and misfolded proteins are removed from the cell through the action of two protein quality control pathways, the ubiquitin-proteasome system and autophagy. Age-associated declines in the activity of both pathways cause damaged and misfolded proteins to aggregate and ultimately lead to neurodegeneration.
Many of the components and mechanisms of the ubiquitin-proteasome system and autophagy were first discovered in yeast, and yeast have helped researchers understand how mutations in many disease-linked genes cause impairments in both pathways.
Yeast’s history in neurodegeneration research
Yeast first came to prominence in the study of neurodegenerative diseases in the 1990s. At that time, advances in DNA sequencing technology enabled large-scale sequencing efforts that identified mutations that cause rare, familial neurodegenerative diseases such as Friedreich’s ataxia, Huntington’s disease, and Niemann-Pick disease, as well as inherited forms of Alzheimer’s disease, Parkinson’s disease, and ALS. The mutations that cause these diseases occur in distinct genes whose functions were, at that time, largely unknown. Thus, after researchers identified the mutations that cause these diseases, they then faced the considerable challenge of determining both the normal functions of the associated genes and mechanisms by which the mutations cause cell death.
Yeast models created through genome engineering methods unavailable in other model systems became critical tools in this endeavor. For disease-linked genes with a yeast homolog, researchers could remove the native yeast gene and replace it with wildtype or mutant forms of its human counterpart. This approach became an important platform for identifying mutations that cause disease by ablating a gene’s normal functions, such as the FXN mutations that impair mitochondrial function in Friedreich’s ataxia and the NPC1 mutations that impair lipid metabolism in Niemann-Pick disease.
Other neurodegenerative disease–causing mutations occur in human genes with no yeast homolog. Nevertheless, researchers found that when such genes are expressed in yeast, they often function as they do in human cells. These “humanized yeast” models supported neurodegenerative disease research in the 1990s and early 2000s, particularly for studying a phenomenon termed “toxic gain of function,” in which a mutation converts a gene’s product into a form that negatively affects cellular viability. By building humanized yeast strains expressing various mutant forms of disease-linked genes, researchers discovered that many mutations exert a toxic gain of function by causing the associated protein to misfold and aggregate into inclusions that disrupt a number of essential cellular functions. Researchers used humanized yeast strains expressing mutant, aggregating forms of human proteins linked to Alzheimer’s disease, ALS, and Huntington’s disease to understand how protein aggregates kill cells, identifying impairments in vesicle trafficking, protein degradation, and mitochondrial function.
Prevalence: Approximately 1 in 300 in the US. Parkinson’s disease is the 16th leading cause of death in the country, afflicting approximately 1 million individuals.
Symptoms: Tremors of the extremities, jaw, or head; difficulty initiating voluntary movements; loss of coordination; anxiety; and depression
Pathology: Parkinson’s disease results from the death of dopamine-producing neurons of the substantia nigra. Neuron loss is thought to be caused by impairments in multiple cellular processes that result from the misfolding and aggregation of the α-synuclein protein into structures termed Lewy bodies.
Insights from yeast: Mutant forms of α-synuclein that cause Parkinson’s disease are toxic when expressed in yeast, creating a platform for large-scale screens for genes and chemical compounds that protect cells. By systematically testing the growth of α-synuclein–expressing strains, researchers determined that the protein kills cells by overwhelming protein quality control pathways and impairing vesicular trafficking. With these mechanisms established, a team at MIT screened a library of 190,000 chemicals to search for compounds capable of preventing the cellular defects caused by α-synuclein expression. Several compounds that reduced the levels of unsaturated membrane lipids protected cells from α-synuclein’s toxic effects. Yumanity Therapeutics, a biotechnology company focused on developing neurodegenerative disease treatments using yeast models, is preparing to test one of these compounds, YTX-3379, in Parkinson’s disease patients.
New disease mechanisms and drug targets
In the 2000s, yeast researchers completed several milestone projects that would influence virtually all areas of biomedical research, including the study of neurodegenerative diseases. In 2002, researchers at Stanford University completed the Yeast Deletion Project, a collection of yeast strains, each with one of the yeast genome’s approximately 6,000 genes removed from its genome. The following year, not long after the completion of the Human Genome Project, the laboratories of Erin O’Shea and Jonathan Weissman at the University of California, San Francisco, released the Yeast GFP Collection, a set of 4,000 unique yeast strains, each with a single gene tagged with green fluorescent protein (GFP). In 2005, the laboratories of Brenda Andrews and Charles Boone at the University of Toronto released the Yeast ORF Collection, a set of 5,000 strains overexpressing individual yeast genes.
Researchers engineered the strains in these collections to express human disease genes to build screening platforms that rapidly accelerated the pace of discovery in neurodegenerative disease research. The deletion and overexpression strain collections enabled researchers to search for genetic modifiers of the toxic effects of neurodegenerative disease–linked genes. In particular, researchers searched for “enhancers,” genes whose deletion increased cell death in humanized yeast strains, and “suppressors,” genes whose overexpression protected cells. Pairing screens of the deletion and overexpression collections led to especially powerful insights into the mechanisms that cause neurodegeneration and, in many cases, led to the discovery of therapeutic approaches that are now being tested in humans, while the GFP collection enabled researchers to determine the precise subcellular localization of each protein in the yeast proteome. This helped to identify disease genes that disrupt a protein’s localization, such as by sequestering it in a toxic protein aggregate, and to visualize protein aggregation in humanized yeast strains, providing a rapid and reliable indicator of the effectiveness of candidate therapeutics that target protein aggregation.
Prevalence: Approximately 1 in 50 in the US; approximately 1 in 10 for individuals over the age of 65. Alzheimer’s disease is the sixth leading cause of death in the country, afflicting some 6 million individuals.
Symptoms: Memory loss; impaired thinking, communication, and reasoning; depression; and hallucinations
Pathology: Alzheimer’s disease is characterized by the death of neurons in the hippocampus and cerebral cortex. Neuron loss is thought to be caused by a combination of extracellular plaques composed of aggregated amyloid-β protein and intracellular neurofibrillary tangles composed of aggregated tau protein. Normally, the amyloid-β protein facilitates trafficking of synaptic vesicles and helps neurons form and maintain physical connections; in Alzheimer’s disease, aberrant processing of the amyloid precursor protein leads to high levels of an aggregation-prone form of amyloid-β (Aβ42). Amyloid-β aggregates kill cells by causing defects in synaptic transmission and endocytic sorting, the process by which the membrane-bound cargoes from endocytosis are directed to the appropriate cellular compartment.
Insights from yeast: Saccharomyces cerevisiae does not have an amyloid-β homolog, does not release synaptic vesicles, and does not form cell-to-cell contacts. Nevertheless, researchers were able to recapitulate aspects of amyloid-β biology in yeast by genetically engineering the fungus not only to produce Aβ42, but to target the protein to the yeast mating pheromone secretion pathway, which functions similarly to the human pathway used for secreting neurotransmitters contained in synaptic vesicles. The secretion-targeted Aβ42 formed aggregates, impaired endocytic vesicle trafficking, and caused cell death, recapitulating the major hallmarks of Aβ42 toxicity in Alzheimer’s. Researchers then systematically overexpressed each gene in the yeast genome to search for modifiers of Aβ42 toxicity, identifying homologs of the human PICALM endocytosis gene as potent suppressors of Aβ42 toxicity. A 2020 study used the same approach to understand how the APOE4 allele increases the risk of Alzheimer’s disease, and also found that PICALM potently suppressed the toxicity of APOE4 in yeast by correcting defects in endocytosis. Preclinical studies are now testing pharmacological and gene therapy approaches that increase PICALM activity in mouse models of Alzheimer’s disease.
Understanding neurodegenerative disease networks
Since the turn of the 21st century, technological advances have led to the emergence of entire new fields of study in biomedical research. The development of approaches capable of measuring the full complement of distinct classes of biological molecules in cells, tissues, and whole organisms led to the emergence of various omics disciplines, including genomics, proteomics, and metabolomics, among others. A particular focus of these disciplines is understanding how interactions among genes, proteins, or metabolites create functional networks through which the properties of cells, tissues, and organisms arise. This holistic, network-based approach to scientific inquiry represents a departure from the traditional reductionist approach that considers individual elements of biological networks in isolation.
The advent of new gene editing tools has advanced the use of yeast for studying the human brain.
As these fields matured over the past two decades, researchers began to explore how aberrant interactions among biological networks contribute to neurodegenerative diseases, and they once again leveraged the power of the yeast model to generate new, systems-level insights into disease. For example, in a 2020 study, researchers used a set of 30,000 yeast strains to build a network of 5,000 disease-linked proteins, revealing 30,374 interactions, many of which were not detected by traditional experimental approaches. Moreover, the resulting neurodegenerative disease network, termed NeuroNet, predicted that many more proteins than previously appreciated are likely aggregated in the neurons of neurodegenerative disease patients, a finding that the authors subsequently validated in brain tissue from Alzheimer’s disease patients. This reinforces the growing sentiment that there are common, conserved cellular processes that may underlie different diseases. These and related results suggest that neurodegenerative diseases perturb the normal functions of protein interaction networks in complex, multifaceted ways, which may explain why previous efforts to treat neurodegenerative diseases by focusing on individual proteins met with little success. However, the results also suggest that NeuroNet is a hierarchical network and that targeting upstream events that promote the protein aggregation in disease may help prevent the aggregation of multiple proteins and ultimately yield treatments that could be effective in multiple neurodegenerative diseases.
Prevalence: Approximately 1 in 10,000 in the US; around 30,000 people in the country have Huntington’s disease.
Symptoms: Uncontrollable, involuntary muscle movements; memory loss; anxiety; depression; and personality changes
Pathology: Huntington’s disease is caused by abnormal, dominantly inherited repeat expansions in the HTT gene that create long stretches of the amino acid glutamine in the huntingtin protein. Individuals with 36 or more repeats will typically be afflicted by the disease, which results in the accumulation of misfolded, aggregated huntingtin protein in the cytoplasm of neurons in multiple brain regions and subsequent death of many of these cells.
Insights from yeast: Engineering yeast to produce disease-causing mutant huntingtin recapitulates the aggregation of the protein and cellular toxicity observed in human Huntington’s disease patients. Such studies have provided valuable insights into how the polyglutamine tracts encoded in mutant HTT kill human neurons. In one study, researchers systematically tested which proteins in the yeast proteome physically interacted with huntingtin aggregates. The resulting set of proteins were unrelated in function, but each contained the same type of polyglutamine tract found in the mutant huntingtin protein. Further analysis showed that the overproduction of these polyglutamine-containing proteins led to their recruitment to huntingtin aggregates, where they acted as “molecular shields” that prevented toxicity caused by huntingtin aggregates binding other proteins and depleting them from their normal location in the cell. This result established that aggregates can impair the function of multiple cellular pathways by sequestering proteins needed by cells to perform their normal functions, an insight that likely extends to protein aggregates found in other neurodegenerative diseases.
Other efforts to understand neurodegenerative disease networks have leveraged the large array of publicly available datasets on interactions between genes and proteins in yeast. Owing to both the advantages of the yeast model system and the relatively small number of genes in the yeast genome, yeast has the most comprehensively characterized networks of genetic and physical protein interactions. In a 2017 study, researchers used information from yeast networks to populate corresponding areas in human networks, which are relatively poorly characterized, to understand hidden connections between Parkinson’s disease genes. In this approach, human genes and proteins were connected to α-synuclein if their corresponding yeast homolog was previously shown to interact with this protein. The resulting “humanized network,” called TransposeNet, delineated a previously undetected defect in translation by identifying interactions between α-synuclein and two translation factors, EIF4G and ATXN2.
Alongside the application of various omics technologies to yeast neurodegeneration models, the advent of new gene editing tools has advanced the use of yeast for studying the human brain. Despite the many advances brought by CRISPR-Cas9, yeast remain unparalleled in the ease, speed, and scale with which they can be genetically modified, and new technologies allow researchers to address once-unanswerable questions in biomedical research. For example, yeast are also being used to understand genetic interactions, a phenomenon in which the effect of one mutation depends on the presence or absence of another mutation. Genetic interactions contribute to both the risk for and severity of neurodegenerative diseases, but are challenging to detect in humans. A team at the University of Toronto developed a system, termed the Synthetic Genetic Array, capable of measuring all pairwise genetic interactions in yeast. The resulting network, built using more than 23 million strains and containing approximately 800,000 unique genetic interactions revealed important principles about how genetic interactions shape cellular viability. For example, the yeast genome was already known to contain approximately 1,000 essential genes that cannot be removed from the genome without killing cells; however, the synthetic genetic array methodology identified an additional 10,000 pairs of nonessential genes whose pairwise deletion from the genome is lethal. Moreover, researchers have used the synthetic genetic array and related methods to show that a gene’s essentiality can vary between environments. Methods such as the synthetic genetic array may one day help establish the cause of noninherited forms of neurodegenerative diseases, which likely result from complex interactions between many individual genetic differences and an individual’s life history.
From their early role in understanding the function of disease genes to more recent systems-level insights into disease mechanisms and treatments, yeast have an important place in the history of neurodegenerative disease research. Some of the most highly studied disease mechanisms and therapeutic targets were first identified using yeast, and the fungus continues to provide needed insights into neurodegeneration, showing that model organisms are an important tool in our efforts to understand, treat, and prevent these devastating disorders.
Amyotrophic Lateral Sclerosis (ALS)
Prevalence: Approximately 1 per 50,000 in the US; there are approximately 16,000 ALS patients in the country.
Symptoms: Muscle weakness/stiffness; muscle atrophy; loss of voluntary motor control; and loss of motor reflexes. Approximately 10–20 percent of ALS patients also develop dementia.
Pathology: ALS results from the death of motor neurons in the brain and spinal cord that control voluntary muscle movements. Many factors are hypothesized to contribute to motor neuron death, including RNA binding protein aggregation, impaired RNA processing, reactive oxygen species, and impaired axonal transport.
Insights from yeast: A University of Pennsylvania team used yeast to predict ALS-causing mutations, noting that many occur in RNA-binding proteins, specifically in segments known as prion-like domains. Applying a prion-like domain–finding algorithm to the human genome, the researchers identified 213 RNA binding proteins with prion-like domains. They then built yeast strains expressing each RNA binding protein tagged with green fluorescent protein (GFP) and searched for strains harboring GFP-positive protein aggregates. Targeted sequencing of the aggregate-forming genes in human ALS patients revealed that mutations in the TAF15, hnRNPA2B1, and hnRNPA1 genes all cause rare familial forms of ALS and that motor neurons in these patients contain aggregates of the associated mutant protein.
Meanwhile, a team at Stanford University created yeast strains expressing toxic peptides derived from abnormal expansions of the human C9orf72 gene, which are the most common genetic cause of ALS and dementia. The researchers systematically searched for modifiers of C9orf72 peptides’ toxic effects on yeast growth in a collection of 11,000 unique strains, identifying genes encoding proteins that transport molecules into and out of the nucleus. This result generalized to other species including humans, and two years later researchers identified the same nuclear transport defects in neurons from ALS patients without C9orf72 mutations, suggesting that the original yeast study may have identified a mechanism common to many forms of ALS and dementia. Based on these and subsequent studies, Biogen developed an experimental drug that corrects the nuclear transport defects seen in ALS neurons, and an ongoing clinical trial is testing its efficacy in ALS patients.