Simple is good: yeast models of neurodegeneration

Authors

  • Sandra Tenreiro,

    1. Cell and Molecular Neuroscience Unit, Instituto de Medicina Molecular, Lisboa, Portugal
    Search for more papers by this author
  • Tiago Fleming Outeiro

    1. Cell and Molecular Neuroscience Unit, Instituto de Medicina Molecular, Lisboa, Portugal
    2. Instituto de Fisiologia, Faculdade de Medicina da Universidade de Lisboa, Lisboa, Portugal
    Search for more papers by this author

  • Editor: Bruno Dumas

Correspondence: Tiago Fleming Outeiro, Cellular and Molecular Neuroscience Unit, Instituto de Medicina Molecular, Av. Prof. Egas Moniz, 1649-028 Lisboa, Portugal. Tel.: +351 217 999 438; fax: +351 217 999 436; e-mail: touteiro@gmail.com

Abstract

The budding yeast, Saccharomyces cerevisiae, is the best-studied eukaryotic cell, at both genetic and physiological levels. As a eukaryote, yeast shares highly conserved molecular and cellular mechanisms with human cells. Thus, this simple fungus is an invaluable model to study the fundamental molecular mechanisms involved in several human diseases. In the particular case of neurodegenerative disorders, yeast models have been able to recapitulate several important features of complex and devastating disorders, such as Huntington's and Parkinson's diseases. Once validated, these models have also been used to accelerate the identification of both novel therapeutic targets and compounds with therapeutic potential. Here, we review the recent contributions of this simple, but powerful model organism toward our understanding of neurodegeneration.

Introduction

The yeast Saccharomyces cerevisiae, also known as baker's or budding yeast, is the most extensively studied eukaryotic organism. Yeast models have been instrumental for our current understanding of conserved cellular mechanisms such as cell division, DNA replication, metabolism, protein folding and intracellular transport (Fields & Johnston, 2005). With a cell cycle that involves haploid and diploid forms, the study of lethal mutations in heterozygous diploids and recessive mutations in haploids is extremely simplified in yeast. Furthermore, classical genetic manipulations are facilitated by mating the haploid strains and sporulating diploid strains, while molecular genetics is aided by the high transformation efficiency of these cells and by the presence of a very efficient homologous recombination pathway, which renders it relatively easy to insert, delete or mutate any genomic sequence up to the chromosome level (Sugiyama et al., 2009).

As a result of the advantages described above, S. cerevisiae was the first eukaryote organism to be fully sequenced in 1996 (Goffeau et al., 1996) and had an important role as a test tube to enhance and accelerate the human genome project. After 14 years of postgenomic research, about 80% of the nearly 6000 proteins predicted to be encoded in the yeast genome have been characterized functionally (Suter et al., 2006; Christie et al., 2009). A very active and collaborative yeast research community built a series of biological resources such as the yeast gene deletion strains (YGDS), where every protein-coding gene in the genome was deleted (Winzeler et al., 1999); the green fluorescent protein (GFP)-tagged collection of strains, where each gene was tagged with GFP (Huh et al., 2003); as well as a collection engineered for protein overexpression (Jones et al., 2008). An overview of some of these valuable biological tools is presented in Table 1. Additionally, high-throughput data obtained from functional genomics approaches, such as transcriptomics, proteomics, metabolomics, interactomics (protein–protein interactions) and locasomics (protein subcellular localization), are well organized and permanently actualized in public databases (Table 2). Notably, information about predicted orthologs in humans is also organized and available for each yeast gene (Table 2).

Table 1.   Example of the yeast collections available for genetic screens
CollectionTypeSource/reference
Yeast haploid and diploid deletion strainsStrains with individual genes replaced by the KanMX4 cassetteWinzeler et al. (1999)
Yeast-GFP clone collectionGenome insertionsHuh et al. (2003)
Yeast GST-fusion CollectionVector libraryMartzen et al. (1999)
Yeast-TAP fusion libraryGenome insertionsGhaemmaghami et al. (2003)
Yeast Tet-promoters – yTHCPromoter-shutoff strains for essential yeast genesMnaimneh et al. (2004)
TRIPLESYeast transposon insertion libraryKumar et al. (2002)
Yeast promoter library constructed in a vector containing two reporter genes (EGFP and lacZ)Vector libraryBell et al. (1999)
Table 2.   Yeast databases available on the internet
DatabaseTypeWebsite
Saccharomyces genome database (SGD)Wide-range informationhttp://www.yeastgenome.org/
MIPS comprehensive yeast genome database (CYGD)Wide-range informationhttp://mips.gsf.de/genre/proj/yeast/
European S. cerevisiae archives for functional analysis (EUROSCARF)Strain and plasmid collectionshttp://web.uni-frankfurt.de/fb15/mikro/euroscarf/
Yeast microarray global viewer (yGMV)Microarrays datahttp://transcriptome.ens.fr/ymgv/
Database yeast search for transcriptional regulators and consensus tracking (YEASTRACT)Transcription regulatory associationshttp://www.yeastract.com/
Profiling of phenotypic characteristics in yeast (PROPHECY)Phenotypes of deletion strainshttp://prophecy.lundberg.gu.se/
Mitochondrial proteome (MitoP)Mitochondria-related genes, proteins and diseaseshttp://www.mitop.de:8080/mitop2/
Eukaryotic orthology (YOGY)Orthologous proteins from eukaryotic orgranismshttp://www.bahlerlab.info/YOGY/
Princeton protein orthology database (P-POD)Orthologous proteinshttp://ppod.princeton.edu/
Yeast protein localization database (YPL.db)Subcellular localization of yeast proteinshttp://ypl.uni-graz.at/pages/home.html
BioGRIDGeneral repository for interaction datasetshttp://www.thebiogrid.org/

Taken together, the powerful available genetic resources and the accumulated knowledge of the simple yeast cell have been used to gain an insight into many human diseases, including neurodegenerative diseases (Fig. 1). The recent developments in the context of Parkinson's (PD) and Huntington's disease (HD), two neurodegenerative disorders associated with protein misfolding and aggregation, will be discussed in detail below.

Figure 1.

 Yeast as a model for neurodegeneration. The yeast Saccharomyces cerevisiae is being used as a eukaryotic model organism, in combination with other cell and animal models, to gain an insight into the fundamental molecular mechanisms involved in neurodegeneration.

Neurodegenerative disorders as protein misfolding pathologies

Neurodegenerative disorders, such as PD, HD and Alzheimer's diseases (AD), belong to the wide superfamily of pathologies known as protein misfolding disorders. The common hallmark of these disorders is the folding of particular proteins into an abnormal three-dimensional conformation, which makes these proteins more prone to aggregate and form amyloid-like β-sheet structures. The composition of the aggregates, as well as their localization in organs and tissues, is specific of each disease. For example, intracellular α-synuclein (α-syn) and huntingtin (Htt) aggregates characterize PD and HD brains, respectively, and extracellular aggregates of tau protein and amyloid-β peptide are characteristic of AD.

Amyloid-like aggregates are dynamic structures, where small soluble species can attach to or detach from the largest protein inclusions relatively easily (Kim et al., 2002). As a result of this dynamism, protein inclusions have variable solubility, stability and size, and they have been suggested to play a physiological role as a natural storage and source of peptide hormones (Maji et al., 2009). In neurodegenerative disorders, however, the aggregation of abnormally folded proteins results in the loss of the normal function of the protein, the gain of a cytotoxic function or both, depending on the disorder. The normal function of the protein that aggregates is often unknown, making it very difficult to determine which effects constitute a loss or a gain of function. Furthermore, there is still intense debate regarding the nature of the aggregates and the peptides that are toxic in neurodegenerative disorders and other protein misfolding pathologies. Large, insoluble protein inclusions inside or outside cells were initially thought to be toxic. However, current evidence indicates that they are rather cytoprotective (Arrasate et al., 2004; Bodner et al., 2006), and that the smaller, more soluble protein dimers and oligomers are the ones that exert toxicity (Outeiro et al., 2007; Chen et al., 2009; Polyakova et al., 2009). Regardless of the exact mechanisms involved in each neurodegenerative disease, it is clear that alterations in normal protein homeostasis are central to this group of pathologies. Therefore, the study of other protein misfolding disorders and the basic pathways involved in protein production and processing can provide important clues for the understanding and cure of neurodegenerative disorders.

Humanized yeast models of misfolding diseases

The basic mechanisms and pathways underlying neurodegenerative diseases, such as mitochondrial dysfunction, transcriptional dysregulation, trafficking defects and proteasomal impairment, are highly conserved between yeast and human species. Thus, the fundamental molecular events involved in these pathologic processes can be studied in simple organisms such as yeast (Gitler, 2008). Two different strategies can be envisaged when modeling a human disease in yeast. If the gene implicated in the disease has a yeast homolog, it is possible to study its function directly. If, on the other hand, the gene underlying the disease is absent in yeast, it can still be modeled via the heterologous expression of the human gene in yeast cells. These two approaches have been used successfully to perform a functional analysis of yeast homologs of human disease genes or to characterize the phenotypes caused by expressing human disease genes in yeast (Smith & Snyder, 2006). The first approach was successfully used, for example, in studies of the yeast YHF1 and SOD1 genes. These genes are the homologs of the human genes involved in Friedreich's ataxia (Puccio & Koenig, 2000) and amyotrophic lateral sclerosis (Leitch et al., 2009), respectively, and these studies strongly contributed toward the clarification of the pathogenic mechanisms underlying these disorders. As an example of the second approach, human genes underlying the histopathological features of various neurodegenerative disorders were expressed in yeast, leading to the discovery of central molecular aspects of HD, PD and AD (Greenfield et al., 1999; Krobitsch & Lindquist, 2000; Muchowski et al., 2000, 2002; Willingham et al., 2003), as we will show below.

PD and other synucleinopathies

PD is one of the most common progressive neurodegenerative disorders, affecting about 2% of people over 65 years old. Clinical manifestations of PD consist of severe motor defects produced by resting muscle tremor, muscle rigidity, bradykinesia and postural instability (Maetzler et al., 2009). PD is characterized by the loss of dopaminergic neurons from substantia nigra pars compacta (Irizarry et al., 1998). Lewy bodies, the pathological hallmark of the disease, are cellular inclusions that can be visualized by histological analysis and are mainly constituted by the presynaptic protein α-syn (Spillantini et al., 1997). Insoluble aggregates of α-syn are also found in other neurodegenerative diseases commonly known as synucleinopathies, such as dementia with Lewy bodies and multiple system atrophy (Spillantini et al., 1997, 1998). α-syn was also the first gene to be associated with familial cases of PD, with several missense mutations (A53T, A30P, E46K) (Polymeropoulos et al., 1997; Kruger et al., 1998; Zarranz et al., 2004). In addition, duplication or triplication of the wild-type (WT) locus was shown to cause PD (Singleton et al., 2003; Chartier-Harlin et al., 2004; Ibanez et al., 2004). In the mean time, other genes were genetically linked to PD, such as parkin, UCH-L1, DJ-1, PINK1, LRRK2 and ATP13A2 (Gasser, 2001), which play important roles in mitochondrial function, the ubiquitin–proteasome system, the autophagy-lysosomal pathway and membrane trafficking (Hatano et al., 2009). Although the genetic cases represent only approximately 10% of the PD cases, the understanding of the molecular mechanisms underpinning the genetic forms of the disease has already provided insights into the pathogenesis of the sporadic forms as well (Hatano et al., 2009). Therefore, intense research is focused on investigating the genes and proteins involved in PD, both in terms of their normal physiological role as well as their contribution to disease.

Yeast as a model of PD

Several molecular aspects of PD have been modeled extensively in yeast, even though yeast lacks obvious orthologs for some PD-associated genes. α-syn is one of the proteins that lacks a yeast ortholog. Nevertheless, due to the central role it is known to play in the disease, it has been studied extensively in yeast. Since the initial description of the first yeast model of PD (Outeiro & Lindquist, 2003), about two dozen publications, reporting different achievements obtained in yeast, have been published (Dixon et al., 2005; Flower et al., 2005, 2007; Zabrocki et al., 2005, 2008; Cooper et al., 2006; Sharma et al., 2006; Volles & Lansbury, 2007; Gitler et al., 2008, 2009; Soper et al., 2008; Vamvaca et al., 2009; Yeger-Lotem et al., 2009; Franssens et al., 2010; Su et al., 2010).

Several features of PD can be reproduced in yeast, namely, heterologous expression of α-syn in yeast inhibits growth (Fig. 2a) and promotes cell death in a concentration-dependent manner (Outeiro & Lindquist, 2003). These observations are similar to those made in other models such as rat primary mesencephalic cultures, Drosophila, Caenorhabditis elegans and mouse models, which display cell dysfunction and death when α-syn is overexpressed (Feany & Bender, 2000; Masliah et al., 2000; Zhou et al., 2000; Lakso et al., 2003). Moreover, increased expression of α-syn is known to cause PD (Singleton et al., 2003; Chartier-Harlin et al., 2004; Ibanez et al., 2004), further validating the results obtained in these model organisms.

Figure 2.

 Inclusion formation and toxicity in yeast cells expressing α-syn and Htt mutants. (a) Spotting assays of yeast cells transformed with either α-syn or the Htt fragments 25Q and 72Q encoding genes, under the regulation of the GAL promoter (inducible by galactose). The toxicity associated with α-syn or Htt 72Q expression in yeast is visible by comparing the cell growth with the respective controls (empty vector or Htt 25Q, respectively). (b) Fluorescence microscopy of yeast cells expressing α-syn fused with GFP or an Htt fragment with 42Q fused to CFP, showing intracellular protein inclusions.

Other features associated with PD, such as the formation of α-syn intracellular foci (Fig. 2b) (Outeiro & Lindquist, 2003; Dixon et al., 2005; Zabrocki et al., 2005; Gitler et al., 2008), oxidative stress (Flower et al., 2005), interaction with lipid rafts (Zabrocki et al., 2008), trafficking defects (Cooper et al., 2006; Gitler et al., 2008; Soper et al., 2008; Su et al., 2010) and apoptosis (Flower et al., 2005), have been accurately recapitulated in yeast.

Furthermore, it was observed recently that α-syn is part of a diverse and highly conserved interaction network including proteins with very diverse functions (namely kinases, phosphatases, deubiquitinating enzymes and metal transporters). This network was identified in genetic screens in yeast and was further validated in rat primary neuronal cultures and in nematode models (Gitler et al., 2009; Yeger-Lotem et al., 2009). This conserved network is evident in the observation that in yeast, α-syn is subjected to several post-translational modifications observed in mammalian cells and in PD patients' brains, namely phosphorylation at S129 and acetylation of the N-terminal (Anderson et al., 2006; Zabrocki et al., 2008; Chen et al., 2009).

Since the validation of yeast as a cell model to study α-syn pathobiology, its genetic resources have enabled several screens, in which modifiers of α-syn toxicity were identified (Fig. 3). Deletion of genes involved in lipid metabolism and vesicle-mediated transport increased yeast susceptibility toward α-syn expression (Willingham et al., 2003), reinforcing the idea that α-syn association with lipids is relevant for its cytotoxicity. Additionally, another study identified a group of highly conserved genes involved in vesicle-mediated transport between the endoplasmic reticulum (ER) and Golgi as the largest and most effective class of modifiers antagonizing the cellular toxicity resulting from the accumulation of α-syn (Cooper et al., 2006), in agreement with other reports of α-syn disruption of vesicular traffic (Gosavi et al., 2002). Findings from the yeast screens were extended and validated in neuronal cell models of PD. As an example, coexpression of α-syn and Rab1, the mammalian ortholog of the yeast Ypt1 protein (known to be one of the strongest yeast α-syn toxicity suppressors), was sufficient to suppress the loss of α-syn-induced dopamine neurons (Cooper et al., 2006; Gitler et al., 2008). Yeast deletion mutants were also used to identify genes that alter the localization and promote α-syn inclusion formation. This study resulted in the identification of genes involved in endocytosis and vacuolar degradation (Zabrocki et al., 2008).

Figure 3.

 Yeast genomewide screens for the identification of modifiers of protein-associated toxicity. (a) One strategy takes advantage of a collection of yeast expression plasmids, each harboring a unique full-length yeast ORF that is used to transform cells containing genome insertions of the gene of interest (GOI) under the regulation of a GAL-inducible promoter. (b) In another method, the YGDS collection is transformed with a plasmid containing the GOI under the regulation of a GAL-inducible promoter. The genes that suppress (suppressors) or aggravate (enhancers) the toxicity induced by the GOI are identified by their ability to increase or decrease the yeast growth rate in galactose-containing media (expression is ‘ON’). In media containing glucose, the expression of the GOI is ‘OFF’.

Interestingly, genetic suppressors capable of preventing the accumulation α-syn-induced reactive oxygen species (ROS) show specificity for WT α-syn toxicity and fail to protect cells from the toxicity induced by α-syn mutants A30P or A53T, suggesting that WT α-syn leads to toxicity through mechanisms that differ from those involved in the toxicity induced by the two familial mutations tested (Liang et al., 2008).

Yeast was also used for drug screens that resulted in the identification of several therapeutic candidates that rescue α-syn toxicity (Griffioen et al., 2006; Fleming et al., 2008; Su et al., 2010). Namely, two flavonoids, quercetin and (−)-epigallocatechin-3-gallate (EGCG), were found to counteract α-syn cytotoxicity, reinforcing the role of oxidative stress as a determinant in α-syn-instigated cellular degeneration (Griffioen et al., 2006). The successful identification of small molecules that rescue α-syn toxicity by stimulating Rab1 function and/or increasing Rab1 levels was also achieved (Fleming et al., 2008). Similarly, a class of small molecules was shown to reduce α-syn toxicity in yeast, with the concomitant reduction in the formation of cytoplasmic α-syn foci, the re-establishment of ER-to-Golgi trafficking and the amelioration of α-syn-mediated damage to the mitochondria. In addition, the same small molecules were shown to counteract the toxicity of α-syn in nematodes and in primary rat neuronal midbrain cultures (Su et al., 2010).

Interestingly, a library of head-to-tail cyclic peptides, which are potent natural bioactive compounds, was recently tested in PD yeast. Two related cyclic peptide constructs that specifically reduced α-syn toxicity in yeast were also found to prevent dopaminergic neuronal loss in worms (Kritzer et al., 2009).

HD

HD is a neurodegenerative disorder characterized by specific movement alterations (chorea), personality changes and cognitive decline (Martin & Gusella, 1986). There is no effective therapy to cure HD or to control its physical and neurologic symptoms. At the molecular level, HD is caused by a mutation in the IT15 gene encoding for Htt, a very large protein of unknown function. This mutation involves an increase in the number of CAG codons (encoding for the amino acid glutamine) in exon 1 beyond a critical length in the polyglutamine (polyQ) region of the Htt protein. This region is polymorphic in the general population, with the number of CAG repeats ranging from four to 35. Expansions of 36–39 CAG repeats increase the risk of developing HD, and expansions of 40 or more CAG repeats are fully penetrant (Rubinsztein et al., 1996). The length of the CAG repeat correlates directly with the severity of the disease and inversely with the age of onset (Duyao et al., 1993), with expansions of 70 CAG repeats or longer inevitably leading to juvenile-onset HD (Huntington's Disease Collaborative Research Group, 1993; Telenius et al., 1993). The expansion of the polyQ region beyond the critical length leads to the misfolding, aggregation and toxicity of mutant Htt in various tissues of HD patients and mouse models of HD (Davies et al., 1997; DiFiglia et al., 1997).

Yeast as a model of HD

Although there is no Htt homolog in the yeast genome, heterologous expression of mutant human Htt exon 1 in yeast reproduces many of the cellular and molecular features of HD pathology in patients. For example, yeast cells expressing mutant fragments of Htt show polyQ length-dependent aggregation and toxicity (Fig. 2a) (Krobitsch & Lindquist, 2000; Hughes et al., 2001; Meriin et al., 2002; Sokolov et al., 2006). Furthermore, yeast models of HD also display several types of cellular dysfunction that are also observed in HD patients and higher eukaryote models of the disorder. These include endocytosis impairment (Meriin et al., 2003, 2007), transcriptional dysregulation (Hughes et al., 2001), apoptosis hallmarks (Sokolov et al., 2006), increased levels of ROS (Giorgini et al., 2005; Solans et al., 2006) and mitochondrial dysfunction (Sokolov et al., 2006; Solans et al., 2006). Very recently, a metabolic approach in yeast HD model enabled the identification of four metabolites (alanine, glutamine, glycerol and valine), whose concentrations changed significantly. This result was consistent in HD transgenic models and in HD patients. Thus, these might be promising biomarkers to evaluate new HD therapies and at the same time provide important insights into the mechanisms associated with mutant Htt toxicity (Joyner et al., 2010).

The yeast HD models recapitulate many features of the toxic gain of function of mutant Htt (Krobitsch & Lindquist, 2000; Muchowski et al., 2000; Hughes et al., 2001; Meriin et al., 2002). Several studies showed that, while the length of the polyQ region determines the aggregation and toxicity of mutant Htt, the sequences flanking the polyQ region also play a key role in the sequence of events leading to Htt toxicity (Dehay & Bertolotti, 2006; Duennwald et al., 2006). In particular, the presence of the proline-rich region situated in the C-terminus of the polyQ stretch is protective in yeast. The role that other regions of Htt play in its aggregation and toxicity has been poorly studied in yeast because it is an extremely large protein and most studies focused on studying the effects induced by exon 1. However, observations made in other experimental models suggest that there are sites of the protein with particular relevance for HD pathology, such as the first 17 amino acids or the many cleavage sites found after the proline-rich region of Htt (Luo et al., 2005; Schilling et al., 2006; Rockabrand et al., 2007; Imarisio et al., 2008; Warby et al., 2009). Therefore, it would be very interesting to use yeast HD models to elucidate the molecular mechanisms that determine the relevance of these sites for Htt aggregation and toxicity.

Genetic screens in yeast models of HD identified several modulators of mutant Htt aggregation and toxicity, which could potentially serve as therapeutic targets. These studies showed, for example, that the formation of mutant Htt inclusions in yeast can be modulated by the expression of chaperones, such as members of the Hsp40 and Hsp70 families, or the chaperonin TRiC (Krobitsch & Lindquist, 2000; Muchowski et al., 2000; Willingham et al., 2003; Tam et al., 2006). These observations were consistent with those made in fly and mouse models, where Hsp70 and its partner proteins potently modulate the aggregation and/or the toxicity of mutant polyQ proteins (Warrick et al., 1999; Cummings et al., 2001; Wacker et al., 2009).

Genetic screens also showed that aggregation-prone proteins, such as yeast prions or proteins containing Q/N-rich regions that may mediate prion-like aggregation, are enhancers of mutant Htt cytotoxicity (Meriin et al., 2002; Giorgini et al., 2005; Duennwald et al., 2006).

The kynurenine pathway, involved in tryptophan degradation, was also found to be activated by the expression of mutant Htt, consistent with observations made in HD patients and in animal models of HD (Schwarcz, 2004; Giorgini et al., 2005). Importantly, this provided the first evidence for a direct relationship between the higher levels of two metabolites of this pathway, 3-hydroxykynurenine and quinolinic acid, and an increased toxicity of mutant Htt.

Proteins involved in other basic cellular functions such as vesicular transport, active microtubule-mediated transport, vacuolar degradation and transcription were also found to be putative therapeutic targets in yeast genomic screenings for their role in the aggregation and toxicity of mutant Htt (Meriin et al., 2002; Muchowski et al., 2002; Giorgini et al., 2005).

Finally, yeast models of HD have also been used to identify new active compounds with therapeutic potential, such as the green tea compound EGCG. EGCG was able to improve growth and to reduce the number of yeast cells containing Htt inclusions, as well as reduce neurodegeneration and improve motor impairment in a fly model of HD (Ehrnhoefer et al., 2006). A primary yeast-based high-throughput screen of small molecules led to the identification of the ones with potential therapeutic interest for the treatment of HD, which were tested later for biological activity in vitro in brain slices derived from HD transgenic mice. This approach resulted in the identification of a potent compound, C2-8, that has long-term inhibitory effects on polyQ aggregation in neurons and suppressed neurodegeneration in a Drosophila HD model (Zhang et al., 2005). The preclinical pharmacology and efficacy of C2-8 was further tested in a transgenic mouse model of HD, where it was demonstrated to be nontoxic, orally bioavailable and to have favorable brain pharmacokinetics. Although it reduced the size of mutant Htt aggregates, it was modestly neuroprotective as measured by behavior and neuropathology parameters (Chopra et al., 2007). These results are promising and could lead to the development of more potent analogs based on the C2-8 structural scaffold. Furthermore, they also validate the use of mutant Htt aggregation models as a valid phenotypic screening tool for compounds with potential therapeutic value for HD, and reinforce yeast as a reliable model of HD.

Concluding remarks

The major limitation in the field of neurodegenerative disorders is the lack of preventive or regenerative therapies. This is mostly due to the general lack of knowledge of the fundamental mechanisms underlying cell dysfunction and neuronal cell death. The high degree of conservation in the major biological processes between yeast and human cells makes it possible to model the basic molecular mechanisms involved in neurodegenerative diseases such as PD and HD in this simple organism. Powerful genetic and postgenomic tools, such as collections of strains and plasmids, enabled the identification of novel modifiers of aggregation and toxicity, which are central hallmarks of these disorders. Importantly, the results obtained in yeast were confirmed in other in vitro and in vivo models, such as neuronal cells, worms or flies, validating novel putative targets for therapeutic intervention. The yeast models are also being used for drug discovery efforts. Together with the genetic tools available, these approaches are expected to lead to the identification of novel drug targets for therapeutic intervention. As with other model organisms, yeast models have obvious limitations. Nevertheless, the use of this simple model organism is becoming widespread, especially in combination with other model systems. It is actually expected that the simple yeast models might enable the development of better animal models for PD and HD, because none of the currently available models fully recapitulates the spectrum of alterations observed in each disorder. Thus, the combination of yeast and other models is expected to lead to advances both in the basic knowledge and in the development of novel avenues for therapeutic interventions in neurodegenerative diseases.

Acknowledgements

We thank Federico Herrera, Zrinka Marijanovic and Leonor Miller-Fleming for their critical reading of the manuscript and insightful comments. T.F.O. is funded by a Marie Curie International Reintegration Grant from the European Commission (Neurofold), an EMBO Installation Grant, and FCT Grant PTDC/SAU-NEU/105215/2008. S.T. is supported by FCT (SFRH/BPD/35767/2007).

Ancillary