A Yeast-Based Model for Hereditary Motor and Sensory Neuropathies: A Simple System for Complex, Heterogeneous Diseases

Charcot–Marie–Tooth (CMT) disease encompasses a group of rare disorders that are characterized by similar clinical manifestations and a high genetic heterogeneity. Such excessive diversity presents many problems. Firstly, it makes a proper genetic diagnosis much more difficult and, even when using the most advanced tools, does not guarantee that the cause of the disease will be revealed. Secondly, the molecular mechanisms underlying the observed symptoms are extremely diverse and are probably different for most of the disease subtypes. Finally, there is no possibility of finding one efficient cure for all, or even the majority of CMT diseases. Every subtype of CMT needs an individual approach backed up by its own research field. Thus, it is little surprise that our knowledge of CMT disease as a whole is selective and therapeutic approaches are limited. There is an urgent need to develop new CMT models to fill the gaps. In this review, we discuss the advantages and disadvantages of yeast as a model system in which to study CMT diseases. We show how this single-cell organism may be used to discriminate between pathogenic variants, to uncover the mechanism of pathogenesis, and to discover new therapies for CMT disease.


Introduction
The peripheral neuropathies, also known as polyneuropathies, are a large group of disorders affecting the three types of peripheral nerves: motor, sensory, and autonomic. The clinical presentations of all neuropathies overlap, and the primary causes are numerous and varied. Infectious, immunemediated, metabolic, toxic, vascular, genetic, and idiopathic forms can all be distinguished from one another. Hereditary neuropathies include Charcot-Marie-Tooth disease (CMT), also known as hereditary motor sensory neuropathy (HMSN); the hereditary motor neuropathies (HMN); the hereditary sensory and autonomic neuropathies (HSAN), also known as hereditary sensory neuropathy (HSN); and small fiber neuropathies (SFN). It is notable that the lines between one class and the next are relative and considered "blurred" or "fluid". The CMT disease can be classified into several types/subtypes, depending on the mode of inheritance (dominant, recessive), the pattern of the injury (axonal, demyelinating) and the genes involved (more than 100 different genes have been identified so far). Despite this high heterogeneity, the clinical presentation allows for the "classical" CMT phenotype to be distinguished. Typically, the disease begins between the first and second decade of can be established by comparing the frequency of rare variants in the patient group compared to the healthy population [10,11]. There are also various computational and experimental methods used to deduce the pathogenicity of rare variants. However, computational methods have the limited predictive potential [12][13][14][15][16][17], and the experimental evaluation of variant function in human cells is difficult, due to inefficient allele replacement methods and the presence of paralogs with overlapping functions. This makes tests in "humanized" model organisms, such as yeast, an attractive alternative.

Therapeutic Approaches for Charcot-Marie-Tooth Disorder
The vast majority of experimental therapies are dedicated to the most common CMT subtypes, i.e., CMT1A (PMP22 gene duplication), CMTX1 (GJB1 gene point mutations), CMT1B (MPZ gene mutations) and CMT2A (MFN2 gene mutations), which, together, make up more than 90% of genetically confirmed CMT cases [18]. However, even in these cases, the proposed therapies are not universal. They usually target a single disease mechanism, and are thus dedicated to a specific set of mutations in the associated gene. Gene therapy has also been attempted as a treatment for CMT disease caused by less common mutations, including IGHMBP2 and SH3TC2 [19][20][21]. As the treatment strategies for different types of CMT were described in detail elsewhere [22], below, we summarize the most important information about possible therapies for CMT disorder.
CMT1A, the most common CMT disease subtype, is the result of elevated expression of PMP22. Down-regulation of PMP22 is, therefore, the main therapeutic strategy targeting this disease subtype. Unfortunately, the first attempts to down-regulate PMP22 gene expression in a mouse model (ascorbic acid) and in a transgenic rat model (progesterone receptor antagonists) could not be translated into human clinical trials [23][24][25]. The discovery that a combination of three medications already on the market (baclofen, sorbitol and naltrexone-PXT3003) was able to ameliorate the long-term phenotypical manifestation of peripheral neuropathy in a CMT1A rat model was a significant step forward (reviewed in [26]). Recently, encouraging news in relation to the Phase III trial of PXT3003 was reported [27]. Another promising pre-clinical therapeutic for patients carrying the CMT1A mutation is ADX71441 (a positive allosteric modulator of GABA B receptors) [28], which was approved for phase I clinical trials for other diseases [29]. In addition, the use of antisense oligonucleotides (ASOs) appears to be a good strategy for the suppression of PMP22 mRNA levels [30]. Nevertheless, it is notable that a variety of missense (amino acid substitutions), nonsense (premature termination due to stop codons), and frameshift mutations have also been described in PMP22, some of them causing more severe phenotypes. These mutations result in a different, not yet fully elucidated, pathogenic mechanism [31] that will require different therapeutic strategies.
The second most common form of CMT disease, CMTX1, is caused by mutations within the GJB1 gene, and seems to be an optimal candidate for gene replacement therapy, due to the small size of the GJB1 gene and the loss-of-function observed for the majority of GJB1 mutations [32,33]. The treatment strategy for the next most common type of CMT, CMT1B, which is caused by mutations in MPZ, is mainly focused on relieving the effects of accumulated mutant forms of the protein in the endoplasmic reticulum (ER). However, this mechanism is not observed in all MPZ mutations, and some may manifest in different manners [31]. It was shown that curcumin was able to release MPZ mutant proteins from the ER to the cytoplasm and cause a significant decrease in apoptosis in HeLa cells [34]. Curcumin also demonstrated positive results in a CMT1B mouse model [35,36]. Sephin1 (a selective inhibitor of a holophosphatase), which attenuates stress resulting from misfolded proteins, also had a beneficial effect [37]. Finally, for CMT2A disease caused by mutations in the MFN2 gene (encoding Mitofusin 2), abnormal mitochondrial trafficking has been reported for at least some mutations. In recent studies, mitofusin agonists have been shown to normalize mitochondrial pathology in the sciatic nerves of MFN2 Thr105Met mice [38].
Despite numerous attempts, no single effective therapeutic for CMT disorders has been registered on the market to date. The vast majority of CMT subtypes and mutations in "common" CMT genes with no "classical" (or an unknown) pathogenic mechanism have not been the subject of research for therapeutic approaches [39]. This state is a result of major barriers to research-including diverse molecular mechanisms for most CMT subtypes (even for different mutations in the same gene)-a lack of knowledge relating to the pathophysiology of specific variants, a deficiency of good disease models, and problems with translating results from animal models into humans. In this context, new models with which to identify new drugs may help to fill the gaps in available CMT therapies.

Studies of CMT in Yeast-Based Models for Human Genes with Yeast Orthologs
The simplicity of yeast is both its principle advantage and disadvantage as a model system. On the one hand, yeast provides an easy, cheap, and rapid platform for conducting research. On the other hand, are we truly able to use it to study processes in very complex systems such as neurons? Can yeast really help to solve the problems that neurogenetics faces: the unknown pathogenicity of rare sequence variants; unclear disease mechanisms; and a lack of effective therapies for patients? It has been shown that yeast can help in all of these cases. In this section, we will present the commonalities between a single-celled yeast and very sophisticated neurons, and how we can exploit these.
Although yeast and humans are separated by a billion years of evolution, a pairwise comparison of genes between these two species reveals more than 2041 groups of orthologs, representing 2386 yeast genes and 3673 human genes [40]. Moreover, there are more than 1000 functional complementation pairs, where a gene from one species can functionally replace (complement) its ortholog in the other [41]. This clearly indicates a significant conservation of function between such distant species, which opens a great number of research possibilities to explore. In the past 30 years, more than 400 yeast genes have been used to study their human counterparts [42]. Whether a human gene will complement mutations of its yeast ortholog cannot be confidently predicted based on the sequence alone [43]. Data related to experimental cross-species complementation are collected and are easy to extract from an open database (for details see [41]). The broad spectrum of possibilities of how to create and utilize yeast as a human model for diseases has been described elsewhere [42,44]. Here, we would like to highlight efficient neuropathy yeast-based models.
The CMT disease consists of a group of disorders displaying high genetic heterogeneity. If we looked at the list of genes associated with neuropathies (Tables 1 and 2), with a particular focus on the function of proteins that they encode, we notice that these proteins occur in a diverse range of cellular pathways and processes. This ranges from the most common of processes, for example, translation (aminoacyl-transfer RNA (tRNA) synthetases), to highly specialized processes such as myelin sheath formation. It is not possible to study the formation of myelin using yeast, but it may be a good model to describe pathogenic mechanisms in more basic processes.

VCP CDC48
A member of the AAA ATPase family of proteins; plays a role in protein degradation, intracellular membrane fusion, DNA repair and replication, regulation of the cell cycle, and activation of the NF-kappa B pathway Wild type VCP partially suppressed the temperature sensitivity growth of cdc48-3 but not the cold sensitivity growth of cdc48-1 and null mutation [61] YARS TYS1 Tyrosyl-tRNA synthetase [62] Genes Possessing Orthologs in Yeast S. cerevisiae

Human Gene
Yeast Gene Protein Function

ABCA1 YOL075C
A member of the superfamily of ATP-binding cassette (ABC) transporters

AIMP1 ARC1
A multifunctional polypeptide with both cytokine and tRNA-binding activities Catalytic component of the active enzyme, which catalyzes the hydrolysis of ATP coupled with the exchange of sodium and potassium ions across the plasma membrane

C12ORF65 RSO55
A mitochondrial matrix protein that appears to contribute to peptide chain termination in the mitochondrial translation machinery

CCT5 CCT5
A molecular chaperone that is a member of the chaperonin containing TCP1 complex (CCT), also known as the TCP1 ring complex (TRiC)

CHCHD10 MIX17
A mitochondrial protein that is enriched at cristae junctions in the intermembrane space; it may play a role in cristae morphology maintenance or oxidative phosphorylation Enzyme that catalyzes the phosphoribosylation of ribose 5-phosphate to 5-phosphoribosyl-1-pyrophosphate, which is necessary for purine metabolism and nucleotide biosynthesis Member of the septin family involved in cytokinesis and cell cycle control Contains a DNA/RNA helicase domain at its C-terminal end which suggests that it may be involved in both DNA and RNA processing

SIGMAR1 ERG2
Receptor protein that interacts with a variety of psychotomimetic drugs, including cocaine and amphetamines

SLC25A19 TPC1
Mitochondrial transporter mediating uptake of thiamine pyrophosphate into mitochondria

SPTLC1 LCB1
The long chain base subunit 1 of serine palmitoyltransferase

SURF1 SHY1
Localized to the inner mitochondrial membrane, involved in the biogenesis of the cytochrome c oxidase complex

TDP1 TDP1
Is involved in repairing stalled topoisomerase I-DNA complexes by catalyzing the hydrolysis of the phosphodiester bond between the tyrosine residue of topoisomerase I and the 3-prime phosphate of DNA

UBA1 UBA1
Catalyzes the first step in ubiquitin conjugation to mark cellular proteins for degradation

Process Gene Protein Function
Lipid metabolism

HADHA
The alpha subunit of the mitochondrial trifunctional protein, which catalyzes the last three steps of mitochondrial beta-oxidation of long chain fatty acids

HADHB
The beta subunit of the mitochondrial trifunctional protein, which catalyzes the last three steps of mitochondrial beta-oxidation of long chain fatty acids HEXA Beta-hexosaminidase subunit alpha: involved in degradation of GM2 gangliosides, and other molecules containing terminal N-acetyl hexosamines

BAG3
Co-chaperone for HSP70 and HSC70 chaperone proteins: acts as a nucleotide-exchange factor (NEF) promoting the release of substrate

DCAF8
Interacts with the Cul4-Ddb1 E3 ubiquitin-protein ligase complex; may function as a substrate receptor

DNAJC3
Acts as a co-chaperone of BiP, a major endoplasmic reticulum-localized member of the HSP70 family of molecular chaperones that promote normal protein folding

Process Gene Protein Function
Signaling

GJB3
Gap junction beta-3 protein: a member of the gap junction protein family

GNB4
Guanine nucleotide-binding protein (G-protein) subunit beta 4; G proteins are involved as a modulator or transducer transmembrane signaling

NDRG1
Belongs to the alpha/beta hydrolase superfamily: a cytoplasmic protein involved in stress responses, hormone responses, cell growth, and differentiation; is necessary for p53-mediated caspase activation and apoptosis NGF Nerve Growth Factor: nerve growth stimulating activity

NTRK1
High affinity nerve growth factor receptor tyrosine kinase: involved in the development and the maturation of the central and peripheral nervous systems

HOXD10
Transcription factor which is part of a developmental regulatory system

IFRD1
Protein related to interferon-gamma: this protein may function as a transcriptional co-activator/repressor that controls the growth and differentiation of specific cell types during embryonic development and tissue regeneration

LAS1L
Involved in the biogenesis of the 60S ribosomal subunit LITAF Plays a role in endosomal protein trafficking and in targeting proteins for lysosomal degradation

MED25
Component of the transcriptional co-activator complex termed the Mediator complex, involved in the regulated transcription of nearly all RNA polymerase II-dependent genes

MORC2
Essential for epigenetic silencing by the HUSH (human silencing hub) complex

PRDM12
A transcriptional regulator of sensory neuronal specification that plays a critical role in pain perception

RBM7
RNA-binding subunit of the trimeric nuclear exosome targeting (NEXT) complex, a complex that functions as an RNA exosome cofactor that directs a subset of non-coding short-lived RNAs for exosomal degradation

SOX10
Transcription factor involved in developing and mature glia TARDBP RNA-binding protein that is involved in various steps of RNA biogenesis and processing

TRIP4
Transcription co-activator, which associates with transcriptional coactivators, nuclear receptors and basal transcription factors   Looking at Tables 1 and 2, of the more than 170 genes involved in various neuropathies, 60 have orthologs in yeast cells (Table 1 and Figure 1).  Some of these genes are so well conserved that even human genes may, at least in part, complement a lack of native yeast orthologs (Table 1). This gives a wide range of possibilities to model and study the neuropathy-associated mutations in yeast cells, which has advantages over other more complex models in terms of its low cost, growth rate, and genetic tractability. However, this opportunity is usually not fully exploited. In most cases, yeast is used only sporadically and to investigate one narrow aspect of a disease-associated mutation. Meanwhile, similar to other rare disorders, yeast-based neuropathy models may, at least partially, solve three of the major problems faced by researchers (Figure 2).
Some of these genes are so well conserved that even human genes may, at least in part, complement a lack of native yeast orthologs (Table 1). This gives a wide range of possibilities to model and study the neuropathy-associated mutations in yeast cells, which has advantages over other more complex models in terms of its low cost, growth rate, and genetic tractability. However, this opportunity is usually not fully exploited. In most cases, yeast is used only sporadically and to investigate one narrow aspect of a disease-associated mutation. Meanwhile, similar to other rare disorders, yeast-based neuropathy models may, at least partially, solve three of the major problems faced by researchers (Figure 2). ) and affects the cell physiology, leading to the selected phenotypes which may be monitored. The system presented may be used to test unknown sequence variants to improve diagnosis, or for screening drug candidates and investigating the molecular mechanisms of pathogenicity to develop future experimental therapies.
The first problem is the significance of rare sequence variants found in patients. As mentioned above, in the current era of NGS, several rare alleles are often identified in individual patients, but their impact on human health is usually poorly understood. This is one of the reasons for prolonged and incomplete diagnoses. There is an urgent need to develop a fast, cheap and reproducible system to test the pathogenicity of identified sequence variants. Yeast has been presented as a good platform for the study of aminoacylation activity, and, thus, the possible pathogenicity of human aminoacyltransfer RNA (tRNA) synthetases (aaRSs) encoded by ARS genes [68]. AaRSs are key enzymes that catalyze the first reaction in protein biosynthesis, charging tRNAs with their cognate amino acids. To date, mutations in six ARS genes have been associated with CMT disorders (GARS, YARS, AARS, MARS, HARS, and WARS, encoding glycyl-tyrosyl-alanyl-, methionyl-, histidyl-, tryptophanyl-RS, respectively) [53,[69][70][71][72][73]. It has been shown that human orthologs can complement the lethality of the deletion of the alanyl-, glycyl-, histidyl-, tyrosyl-RS genes (ALA1, GRS1, HTS1, TYS1, respectively) in yeasts and restore the growth of these cells (see Table 1). The other two human ARS genes associated with CMT (MARS and WARS) possess yeast orthologs (MES1 and WRS1, respectively), which allows researchers to model the mutations found in patients in the corresponding yeast genes. Testing newly identified rare variants in yeast not only allows the identification of loss-of-function (functional null; hypomorphic) alleles but can also reveal gain-of-function (i.e., hypermorphic) alleles [45,73]. Even though a human full-length wild-type WARS failed to complement a wrs1 deficiency in yeast cells, a where it is expressed. The resulting protein interacts with yeast cellular components (proteins, RNAs, lipids, etc.) and affects the cell physiology, leading to the selected phenotypes which may be monitored. The system presented may be used to test unknown sequence variants to improve diagnosis, or for screening drug candidates and investigating the molecular mechanisms of pathogenicity to develop future experimental therapies.
The first problem is the significance of rare sequence variants found in patients. As mentioned above, in the current era of NGS, several rare alleles are often identified in individual patients, but their impact on human health is usually poorly understood. This is one of the reasons for prolonged and incomplete diagnoses. There is an urgent need to develop a fast, cheap and reproducible system to test the pathogenicity of identified sequence variants. Yeast has been presented as a good platform for the study of aminoacylation activity, and, thus, the possible pathogenicity of human aminoacyltransfer RNA (tRNA) synthetases (aaRSs) encoded by ARS genes [68]. AaRSs are key enzymes that catalyze the first reaction in protein biosynthesis, charging tRNAs with their cognate amino acids. To date, mutations in six ARS genes have been associated with CMT disorders (GARS, YARS, AARS, MARS, HARS, and WARS, encoding glycyl-tyrosyl-alanyl-, methionyl-, histidyl-, tryptophanyl-RS, respectively) [53,[69][70][71][72][73]. It has been shown that human orthologs can complement the lethality of the deletion of the alanyl-, glycyl-, histidyl-, tyrosyl-RS genes (ALA1, GRS1, HTS1, TYS1, respectively) in yeasts and restore the growth of these cells (see Table 1). The other two human ARS genes associated with CMT (MARS and WARS) possess yeast orthologs (MES1 and WRS1, respectively), which allows researchers to model the mutations found in patients in the corresponding yeast genes. Testing newly identified rare variants in yeast not only allows the identification of loss-of-function (functional null; hypomorphic) alleles but can also reveal gain-of-function (i.e., hypermorphic) alleles [45,73].
Even though a human full-length wild-type WARS failed to complement a wrs1 deficiency in yeast cells, a mutant WARS incorporating a H257R substitution could partially complement the lack of WRS1, which implies that H257R substitution may change the structure of human tryptophanyl-RS [73].
Another example of using a yeast-based system to investigate rare sequence variants is the study of mutations in the POLG gene, which encodes the catalytic subunit of mtDNA polymerase γ. Pathological mutations in this gene are usually associated with severe mitochondrial disorders, but may also manifest as an isolated neuropathy [74]. Yeast is a suitable model organism for the study of alleles resulting in severe oxygen and/or respiration impairment and mitochondrial dysfunction, due to its ability to survive without oxidative phosphorylation. Comparable phenotypes (the harmful effects observed in yeast reproduce the severity of the phenotypes in humans), the possibility to study variants in heteroallelic states, and easy and fast analysis make budding yeast an excellent model for testing mutations in POLG. In POLG-associated diseases, yeast has enabled researchers to distinguish pathogenic mutations from other single-nucleotide polymorphisms; to show that some polymorphisms may act as phenotypic modifiers; and has demonstrated that certain mutations are not the only cause of a pathology, highlighting the need for further genetic analysis [75][76][77][78][79][80][81].
The second problem is a deficiency or complete lack of knowledge related to the molecular mechanisms underlying the pathogenicity of mutations. Information about the cellular processes that trigger pathogenic changes and ultimately lead to the symptoms observed in patients may provide clues for the production of effective therapeutics. It may also reveal the possible toxicity of the drugs used and which drugs should be avoided when treating specific patients. It may also enable recommendations to be made for lifestyle changes to improve the quality of life and slow the disease's progression. An example of using yeast to learn about the mechanism of pathogenicity of a specific mutation is the study of the human gene MFN2 encoding Mitofusin 2. Mutations in MFN2 result in the most frequent axonal form of CMT (CMT2A) [82]. Mitofusin 2 represents a key player in mitochondrial fusion, trafficking, turnover and the formation of contacts with other organelles [83]. Yeast mitofusin, Fzo1, was used to study the consequences of CMT2A mutations associated with the human MFN2 gene. It was shown that one mutation in particular (causing a substitution analogous to the I213T substitution in Mfn2) is highly deleterious for protein function and stability. Other mutations had variable effects, causing either no phenotype, or a subtle alteration of mitochondrial morphology [84]. The study of Mfn2 function is also of importance because mitofusins belong to the group of proteins important for the formation, regulation, and function of endoplasmic reticulum (ER) and mitochondrial membrane contact sites (MCSs), called mitochondria-associated membranes (MAMs). MCSs are structures where membranes of different organelles are close and connected by a proteinaceous tether but do not fuse. The genes affecting homeostasis in MAMs are over-represented in the group of genes causing several hereditary neurodegenerative disorders, such as Alzheimer's disease [85][86][87][88] and Chorea-acanthocythosis [89]. This is because MCSs are involved in various processes, including mitochondrial dynamics (fusion, fission), lipids metabolism, autophagy, cell survival, energy metabolism, calcium homeostasis, and protein folding [90][91][92]. In the group of genes causing CMT disorders, besides Mfn2, there are other proteins with and without orthologs in yeast, which take part in processes at MAMs, such as VAPB, Opa1, or GDAP1 [93]. Studies on the function of one MCSs component may help uncover the role of MCSs in the development of several other neurodegenerative diseases. More details about the role of MAMs in neurodegeneration can be found in other reviews [85][86][87][88].
The third and final challenge, facing not only neuropathies but all rare diseases, is a lack of therapies. This problem was more specifically described in the Section 3. Yeast may also serve as a rapid and cheap platform with which to screen potentially active compounds, and for a detailed analysis of the cellular effects of drugs (see Section 6).

Studies of CMT for Human Genes Lacking Orthologs in the Yeast Model
The most simple yeast-based neuropathy models are based on the homology between human and yeast genes. However, even in the absence of clear orthologs, yeast may still be used to study the three problematic areas for rare diseases identified in Section 4.
The strategy that allows for this is based on the belief that yeast and the cells of higher organisms are built and function according to the same principles. Thus, human proteins may still be able to modulate essential cellular pathways in yeast. This rule has been shown in the case of yeast-based studies of neurodegenerative disorders including Alzheimer's disease, Parkinson's disease, and Huntington's disease. Here, the heterologous expression of pathological variants of aβ-peptide, α-synuclein or poly-Q repeats of different lengths in the yeast allowed researchers to dissect the molecular mechanisms underlying the pathology of mutant proteins, identify potential drug targets and select active compounds by screening the available drug libraries [94][95][96][97][98]. Similarly, in the case of genes whose mutations lead to CMT diseases and which have no orthologs (Table 2), their expression in yeast may still affect cells and limit their growth.
This strategy, testing the activity of a human protein without an ortholog in yeast, has already been used to study neuropathy associated with mutations in the GDAP1 gene encoding ganglyoside induced differentiation associated protein 1. Although the production of this protein does not affect the growth of wild-type yeast cells, it has manifestations at the molecular level, as shown in two different studies [99,100]. In both studies, the effect of heterologous GDAP1 expression in the wild-type and different yeast mutants was tested. Based on the observation that increased expression of GDAP1 in COS-7 cells results in mitochondrial fragmentation, which is probably due to interference with the mechanism of mitochondrial fission, the influence on yeast mitochondria morphology was investigated [99,100]. The expression of GDAP1 either did not affect yeast mitochondria morphology [99], or caused increased fragmentation of the mitochondria, depending on the specific experimental procedures [100], and also changed mtDNA maintenance [100]. GDAP1 expression in the fis1∆ mutant, defective for mitochondrial fission and for G2/M progression during the cell cycle, did not eliminate the mitochondrial fission defect, but reversed the cell cycle delay phenotype of this mutant [99]. Although the results obtained did not define the molecular function of the GDAP1 protein, they indicate the pathway for further research. Moreover, finding a clear phenotype allowed for further testing of the effect of pathogenic GDAP1 gene missense variants. Since all the investigated variants did not reverse the cell cycle delay phenotype of a fis1∆ mutant, it suggests that this phenotype can be used to study new GDAP1 variants and distinguish pathogenic from non-pathogenic alleles identified in patients during the diagnostic process. However, it is not possible to differentiate between variants in terms of the strength of the clinical symptoms induced. A more functional system with which to evaluate the pathogenicity of GDAP1 gene mutations was reported in our previous study [100]. In this case, the csg2∆ mutant was used with the deletion of a gene required for mannosylation of inositolphosphorylceramide. The sensitivities of the csg2∆ strain to stress conditions, the presence of tunicamycin and high concentrations of calcium ions (Ca 2+ ) were all suppressed by the expression of the wild-type GDAP1 allele, but not by GDAP1 variants encoding proteins that lost the ability to correctly localize to the mitochondria. Different GDAP1 variants (point mutants) exhibited differing abilities to suppress these phenotypes. This system, in addition to enabling testing of the pathogenicity of GDAP1 alleles isolated from patients, should also allow for their classification in terms of the severity of the resulting clinical phenotypes.
These two working systems show that GDAP1 functionality can be assessed in yeast cells even though yeast has no functional orthologs. This is due to the fact that they participate in conserved cellular processes (Table 2) and, consequently, there are functional orthologs of their partner proteins ( Figure 2). Based on this principle, it is possible to build additional yeast-based models for other genes that are involved in CMT disease, but do not have yeast orthologs. An example of this is the human gene MTMR2 coding for Myotubularin 2-related protein, a member of the myotubularin family of phosphoinositide lipid phosphatases. Myotubularin 2 functionally interacts with the phosphoinositide 5-phosphatase protein FIG4, amongst other proteins in Schwann cells and in neurons [101]. Mutations in the FIG4 gene, similar to in MTRM2, are described as resulting in CMT disease. The human FIG4 has a yeast ortholog, making it is possible to build a model to directly study FIG4, and to indirectly test MTMR2 function in a yeast model.
Another possible use of the preservation of biochemical pathways is associated with the CMT subtype, caused by mutations in the MPZ gene. Some mutant MPZ proteins are retained in the ER, where their accumulation triggers an induction of the unfolded protein response (UPR). Although the downregulation of UPR has already been shown to have a positive effect in CMT1B, the problem associated with distinguishing between pathogenic and non-pathogenic variants and the subsequent search for mechanisms of pathogenicity remains. In a case of mutations in the MPZ gene, a yeast model can be used because of the conservation of biochemical pathways.
As a result of UPR induction in mammalian cells, the eIF2 kinases PERK and GCN2 are activated and phosphorylate the translation initiation factor 2 (eIF2a). This leads to the synthesis of transcription factor ATF4 and increased production of the CHOP protein. The induction of this response has been shown to cause demyelination in CMT1B. Depletion of CHOP or its subsequent target, Gadd34, improves myelination [102]. The mammalian signaling pathway leading to eIF2a phosphorylation is homologous to the well-studied general control response in yeast, in which phosphorylation of eIF2a activates genes involved in amino acid biosynthesis. Thus, mammalian cells use a conserved pathway to regulate gene expression in response to various stresses [103]. Therefore, it is possible to use yeast to study the pathogenicity of mutations by monitoring the effects caused by the presence of MPZ protein variants at the molecular level.
Finally, with models such as those described above, we can study the pathogenicity of variants and mechanisms of pathogenesis and use them to find therapies by searching libraries of small molecules. This experimental approach could and should be far more widely applied to the investigation of CMT diseases.

Repositioning of Drugs in Hereditary Neuropathies
In the case of common disorders, effective medicines can be found and tested in clinical trials involving thousands of patients, but for rare and ultra-rare diseases, the classical approaches to drug discovery are very difficult to follow. This is due to the small number of patients and the lack of economic justification for pharmaceutical companies to engage large resources in research that will not be profitable for them.
This problem applies to CMT disease, which can be classified as a rare or even ultra-rare disease. In such cases, one favorable solution is a drug repurposing strategy. Drug repurposing (also known as repositioning, reprofiling, rediscovering or redirecting) may be defined as developing new uses for a drug beyond its original intended use or initially approved use. The best example of drug repurposing is that of chlorpromazine. In 1950, chlorpromazine, synthesized as a potential antimalarial drug, was administered to patients before surgery. Due to its unexpected sedative effects, a weak anti-malarial drug has become a powerful medicine used in both psychiatry (acute mania) and neurology (chorea, epilepsy, muscle spasms, etc.) [104]. Drug repositioning radically reduces the cost of clinical trials, prevents the withdrawal of a drug from the market due to low interest, and usually allows well-known and cheap pharmaceuticals to be selected as the preferred option. This latter feature is especially important when developing therapies for rare disorders, as the alternatives (e.g., gene therapy or cell therapy) are extremely expensive.
Yeast models have great potential to be used as a platform for the screening of drugs libraries to obtain preliminary results. Such models benefit from a fast turnaround, low cost, and easy testing. Yeast models have been successfully used to search for active compounds against mitochondrial diseases [105][106][107], central nervous system diseases [108], and copper-deficiency disorders [109]. Thus, it seems reasonable that, for peripheral neuropathies, yeast models may also be applied and allow for drug repositioning. Nearly a third of human genes involved in the pathogenesis of CMT have yeast orthologs, making it relatively easy to create yeast-based models. For the other cases, there is the possibility of finding a convenient, easy to quantify phenotype (e.g., a toxic effect associated with the expression of gene variants), which may be used for drug screening. The other option is the use of phenologs, which are the phenotype-level equivalent of gene orthologs. Two sets of deeply conserved genes between extremely evolutionary divergent organisms, i.e., yeast and humans, may be manifested in the different molecular contexts as two dissimilar phenotypes. For example, cell-wall maintenance in yeast and vascular growth in human are phenologous processes: they share the same genes conserved between yeast and human. The failure of the equivalent genes will lead to disturbed angiogenesis in humans and reduced cell-wall maintenance in yeast. Using this logic, thiobendazole-marketed as an antifungal drug-would also act as an angiogenesis inhibitor in humans [110,111]. Similarly, phenologous processes may also be used for drug selection in CMT disease.
The repurposing strategy may be key to finding therapies, especially for very rare subtypes of CMT. To date, this approach has been poorly used for CMT diseases. Only a few CMT genes and their mutations (AARS, ATP7A, GARS, HARS, HINT1) have been modeled in yeast so far, and any have not been used for drug repurposing, despite the systems being ready to test collections of highly bioavailable drugs of known toxicity in models to confirm their action and to select the effective dose.

Outlook
Rapidly developing technologies provide ever-improving diagnostics, pathology monitoring and therapies. However, they also reveal the limitations of current approaches and "black holes" in our knowledge. Returning to more basic, simple models may be a key for further research regarding different, and especially very rare, neuropathies. Included in this work are examples that clearly indicate that yeast models have a great potential to serve as neuropathic models. These models are currently underused. Yeast models are a promising tool for the determination of the pathogenicity of newly discovered rare sequence variants and their influence on the progression of a disease. In addition, they may also provide an excellent platform for studying the molecular and cellular background of pathogenesis and offer a cheap, easy, and fast system for the high-throughput screening of genetic and chemical suppressors that would give rise to potential therapeutics. Yeast models have advantages over other models in terms of cost, ease of growth and use, and the facilities necessary to implement them. They could be used as a convenient tool in almost all laboratories, and we hope that they will continue to be developed to support the study of more hereditary neuropathies.