Modeling Neurodegenerative Disorders in Drosophila melanogaster

Drosophila melanogaster provides a powerful genetic model system in which to investigate the molecular mechanisms underlying neurodegenerative diseases. In this review, we discuss recent progress in Drosophila modeling Alzheimer’s Disease, Parkinson’s Disease, Amyotrophic Lateral Sclerosis (ALS), Huntington’s Disease, Ataxia Telangiectasia, and neurodegeneration related to mitochondrial dysfunction or traumatic brain injury. We close by discussing recent progress using Drosophila models of neural regeneration and how these are likely to provide critical insights into future treatments for neurodegenerative disorders.


Introduction: Drosophila as a Model System for Studies of Human Disease
The common fruitfly Drosophila melanogaster has been used as a genetic model system for more than 100 years. Because Drosophila are inexpensive to maintain and reproduce rapidly, an enormous repertoire of genetic technologies has been created over the past century [1]. Drosophila possess extensive homology with humans at the genetic level making them a useful model for investigation of the cellular and molecular processes underlying development and disease [2]. Over the past three decades, a variety of human diseases also have been modeled in Drosophila, including many affecting the nervous system [2][3][4]. Human diseases for which there are Drosophila models are curated by the Bloomington Drosophila Stock Center (https://bdsc.indiana.edu/stocks/hd/index.html). The Drosophila studies are facilitated by the fact that the Drosophila nervous system is complex and possesses many features of our own nervous system including: eyes, olfactory organs, gustatory organs, auditory organs, a ventral nerve cord (spinal cord analog), peripheral sensory neurons for proprioception and pain, and a brain [5] and the generation of huge collections of mutants that impact neural development [6][7][8][9][10][11][12][13][14][15][16][17]. Moreover, multiple rigorous assays to score neurodegeneration can be used in Drosophila, providing reliable measurements for the effects of the disease process. Such assays include examination of eye morphology and retinal structures by light microscopy, vacuolization of the central brain using histological staining, lifespan analysis, locomotor performance measurements using a climbing assay as well as assessment of neuromuscular junction morphology to determine potential synaptic abnormalities [18,19]. Immunohistochemical techniques can be used to label specific subtypes of brain cells such as dopaminergic neurons by using an anti-Tyrosine hydroxylase antibody, or to examine the accumulation of deposits such as amyloid plaques, which are a hallmark of the neuropathology accompanying Alzheimer's Disease using Thioflavin S labeling [18,20].
In this review, we focus on several major neurodegenerative diseases being modeled in Drosophila (Figure 1). The diseases discussed here include adult-onset diseases such as Alzheimer's Disease, Parkinson's Disease, Dementia with Lewy Bodies, Amyotrophic Lateral Sclerosis (ALS), Frontotemporal Dementia (FTD) and Huntington's Disease. We also review existing models of Ataxia Telangiectasia, which is a childhood-onset multiorgan disorder, characterized by progressive neurodegeneration as well as emerging models of neurodegenerative diseases with mutations in mitochondrial genes or the Drosophila brain tumor gene. We also discuss Traumatic Brain Injury (TBI) model that is being used to model Chronic Traumatic Encephalopathy (CTE). We conclude with some likely future directions of the Drosophila neurodegeneration field that include use of this powerful model to investigate neural regeneration and how these studies may lead to clinically relevant therapeutics. revealed by histology and hematoxylin and eosin (H&E) staining that shows the appearance of holes in the brain neuropil of 60,114 mutants (ND23 mutants) but not in heterozygous controls (60114/+). Image copyright and permission to use the image were obtained from [21]. (B) Rough eye phenotype (B and B' for magnified image) observed in a Drosophila model of Ataxia Telangiectasia using scanning electron microscopy. Image copyright and permission to use the image were obtained from [22]. (C) Loss of dopaminergic neurons in a Drosophila model of Parkinson's Disease is revealed by immunohistochemistry using an anti-Tyrosine Hydroxylase antibody. Image copyright and permission to use the image were obtained from [23]. (D) Neurodegeneration in photoreceptors (labeled R1-R7) of ommatidia in a Drosophila model of Amyotrophic Lateral Sclerosis (right image) is revealed using Transmission Electron Micrographs. Image copyright and permission to use the image were obtained from [24]. (E). Progressive spreading of Red Fluorescent Protein (RFP)-labeled Huntingtin within the brain is revealed by immunohistochemistry in a Drosophila model of Huntington's Disease. Image copyright and permission to use the image were obtained from [25].

Alzheimer's Disease
Alzheimer's Disease (AD) is a form of dementia that usually manifests after the age of 65. It is characterized by memory loss, mood and behavior changes, the presence of specific protein aggregates in the brain, and a reduction in brain regions including the hippocampus and temporal lobes [26][27][28]. AD is one of the leading causes of death in the US; based on data from the 2010 census, 4.7 million individuals over the age of 65 had AD, and an estimated 13.8 million people in the US will have AD by 2050 [29,30]. Although there are no known cures, treatment options include diet and lifestyle interventions, as well as medications that alleviate the symptoms and/or progression of the disease [31,32].
The etiology of AD is the subject of ongoing research, and different hypotheses have been proposed to account for the variety in associated risk factors and physiological changes. The most prominent of these is the amyloid hypothesis, which holds that the buildup of characteristic extracellular amyloid-beta (Aβ) aggregates causes pathology, notably neurodegeneration. Pathogenic, extracellular Aβ42 is produced by sequential cleavage of the intramembrane amyloid precursor protein (APP) (in Drosophila, amyloid precursor protein-like (APPL)) by BACE1 (β-site APP cleaving enzyme-1), and γ-secretase. In contrast, non-pathogenic cleavage of APP is carried out by α-secretase [33][34][35]. Another characteristic protein, Tau also is implicated in the amyloid hypothesis. Under normal conditions, Tau binds to microtubules; however, when Tau is hyperphosphorylated, it detaches and forms intracellular aggregates, destabilizing the microtubules and thus decreasing neurotransmission. The cause of Tau hyperphosphorylation is not known, but some data suggest that amyloid pathology contributes, or that they share some mutual causation through a mechanism such as innate immunity. Indeed, activation of the innate immune system and chronic inflammation have been implicated in multiple neurodegenerative disorders [36]. The role of Drosophila brain immunity in the contexts of injury and neurodegeneration is reviewed in [37]. Other hypotheses focus on observations related to Tau tangles, cholinergic dysfunction, mitochondrial dysfunction and oxidative stress, calcium homeostasis, vascular dysfunction, glia-mediated inflammation, metal ion toxicity, and poor lymphatic clearance. These proposed mechanisms are all interrelated, often by the role of Aβ aggregates, and all may contribute to the development of AD [35].
Drosophila homologs of AD-associated genes have provided insights into the human genes implicated in the development of AD as well as the pathways that contribute to the disease. The gene Draper (in humans, MEGF10), plays a role in the glial engulfment of Aβ, reducing neurotoxicity in a Drosophila model of AD [38]. In another study of 87 Drosophila genes, each with a human homolog identified in GWAS as an AD-associated genomic locus, nine were found to strongly affect the toxicity of Tau: CD2AP (cindr), SNRPN (SmB), PTPRD (Lar), XYLT1 (oxt), FERMT2 (Fit 1, Fit 2), CELF1 (aret), MAST4 (CG6498), ITGAM (scb), and ITGA9 (scb) [39]. The proteins encoded by CD2AP and FERMT2 both function with integrins in cell adhesion and signaling; ITGAM and ITGA9 produce α-subunits for integrin receptors; and PTPRD and XYLT1 also function in cell adhesion [39,[53][54][55][56].
Environmental factors, including diet, lifestyle, and chemical exposure, are known to contribute to AD in humans [62,63]. The metals iron, copper, and zinc; each found in the human diet; have been studied in Drosophila models of AD, using a variety of methods to increase and decrease exposure. Copper and zinc supplementation exacerbate Aβ42 toxicity, while leveraging chelators, expression of detoxifying proteins, and expression of transport proteins ameliorate toxicity [50,51]. Iron chelators, when overexpressed, can also attenuate Aβ42 toxicity and when expression of these chelators is reduced, toxicity increases [49,52]. Reflecting a different aspect of lifestyle and environment, a unique study using a Tau model of AD found that disruption of the circadian rhythm with dim light exposure increased neurodegeneration [48]. Lifespan [46] Pan-neuronal and muscle-specific expression of Aβ42, exposure to exogenous Aβ42, and treatment with anti-Aβ42 antibody (6E10) Larva 3rd Instar Larvae: Electrophysiology for synaptic transmission, FM1-43 dye imaging for neurotransmitter release, Thioflavin S staining for amyloid deposits [47] Pan-neuronal and photoreceptor-specific expression of two human Tau variants, manipulation of light exposure Adult Lifespan, histology, and light microscopy to quantify neurodegeneration, climbing assay, immunohistochemistry for pTau accumulation, light microscopy for eye morphology [48] Pan-neuronal expression of human APP and BACE1 separately and in combination, treatment with a γ-secretase inhibitor Adult Lifespan, fluorescence microscopy for defects in the whole-brain and mushroom body structure, immunostaining with anti-Aβ (6E10), Thioflavin S, and X-34 for amyloid deposition, climbing assay, conditioned courtship suppression assay [57] Expression of human BACE1 and late-onset induction of human APP Adult Lifespan, climbing assay, immunostaining with anti-Aβ (6E10) for amyloid deposition, fluorescence microscopy for defects in the whole-brain and mushroom body structure, conditioned courtship suppression assay [58] Combined models of Drosophila orthologs and overexpression of human

Lewy Body Dementias: Parkinson's Disease and Dementia with Lewy Bodies
Lewy Body Dementias (LBDs) are characterized by aggregates of α-synuclein (α-syn) within cells of the brain. These aggregates are called Lewy bodies. There are two types of LBD: Parkinson's disease (PD) and Dementia with Lewy Bodies (DLB). Overexpression of α-syn in Drosophila could be considered to model PD, DLB, or both. However, publications to date refer to α-syn overexpression in Drosophila as PD models. We maintain that convention here and therefore do not discuss DLB further. PD is a neurodegenerative disease that affects individuals over the age of 45 at a rate of 572/100,000 in North America. The number of people with PD in the United States is expected to reach 930,000 in 2020 [65]. The characteristic symptoms of PD include tremor and posture instability, which are caused by a loss of midbrain dopaminergic (DA) neurons that deliver dopamine to the basal ganglia [66]. Other brain structures also are affected, such as the cerebral cortex, post-commissural putamen, and olfactory tubercle, leading to diverse symptoms [66]. There are no known cures for the condition, but medications that target dopamine receptors, such as levodopa and dopamine agonists, are effective at treating the symptoms [67]. Other non-pharmacological treatments, such as deep brain stimulation and exercise therapy, may also be effective [68].
Among the molecular mechanisms contributing to PD pathology are neuroinflammation, defects in α-syn proteostasis, mitochondrial dysfunction, oxidative stress, perturbation of calcium homeostasis and defects in axonal transport [67]. Because the etiology of PD is multifactorial, a variety of models have been generated in Drosophila to reflect known contributing factors ( Table 2). Orthologous genes, transgenic constructs carrying human genes, and environmental factors all have been investigated. The Parkinson's disease-related genes that have exploitable homologs in Drosophila include PARK2, PINK1, LRRK2, DJ-1, UCH-L1, HtrA2, GBA, and Tau [69][70][71][72][73][74][75][76][77][78][79][80]. α-Syn and Pael-R do not have Drosophila homologs, and are studied using transgenic models [79,81]; human transgenes also have been introduced for other genes including LRRK2 and Tau [79,82,83]. Moreover, environmental stressors, such as the widely used pesticides rotenone and paraquat, have been tested on Drosophila [84][85][86]. Overexpression of Pael-R in DA neurons Adult TH immunostaining [79] Toxin exposure Rotenone Adult TH immunostaining, climbing assay [84] Paraquat Adult TH immunostaining, lifespan, climbing assay, jumping assay, Dopamine levels [85,86] Functions of the Drosophila orthologs of genes associated with PD can be investigated using mutant flies or by tissue and/or cell-specific overexpression or knock down using binary expression systems [88]. The PARK2 gene codes for the protein Parkin, which targets abnormal proteins for degradation. The PD-related proteins Pael-R and α-Syn are among those surveilled by PARK2 [79]. Importantly, Drosophila brains, as with human brains, have dopaminergic (DA) neurons. Thus, the consequences of particular mutations and gene overexpression can be examined in DA neurons. In Drosophila, PINK1 protein is important for mitochondrial function, and PINK1 mutants have fewer DA neurons and exhibit both olfactory dysfunction and motor deficits [69,75,76,80]. Drosophila LRRK2 loss-of-function mutations also reduce DA neuron numbers and cause impaired locomotor activity [71]. DA neuron-specific knockdown of the Drosophila ortholog of UCH-L1, dUCH, leads to Parkinson's disease-like phenotype illustrated by the loss of DA neurons, while its overexpression leads to caspase-dependent cell death in eye imaginal discs, aberrant patterning of the pupal retina and a rough eye phenotype in the adult [77]. The HtrA2 protein has protease activity and is involved in apoptosis. Knockdown of HtrA2 function in Drosophila DA neurons and photoreceptor cells decreases lifespan, motor ability, and ommatidia number [72]. The gene GBA encodes β-glucocerebrosidase, which is necessary for preventing accumulation of glucosylceramides. GBA mutations in Drosophila result in DA cell death, motor deficits, and decreased lifespan [73]. The protein Tau can form neurotoxic inclusions that are implicated in both human PD and AD [70]. Overexpression of Drosophila Tau in mushroom body neurons results in learning and memory defects [87]. Tau toxicity in Drosophila is increased by both over-and under-expression of LRRK2, and is characterized by loss of tyrosine hydroxylase (TH)-immunoreactive neurons [89].
Because flies lack orthologs of human α-Syn or Pael-R, overexpression of human cDNAs has been used to investigate their functions in Drosophila. Overexpression of variants of human α-Syn in Drosophila, leads to locomotor defects, Lewy body formation in the brain and retinal degeneration [81]. Co-expression of the human proteins α-Syn and Tau, in a Drosophila model, leads to the formation of inclusions of ubiquitinylated proteins that disrupt functions of the cytoskeleton, causing neurodegeneration [83]. Overexpression of wild type and mutant human LRRK2 in Drosophila leads to degeneration in photoreceptor cells and neurons along with symptoms including motor deficit and decreased lifespan [82].
The pesticides rotenone and paraquat have been linked to the development of PD in humans, and Drosophila models of PD have been used to investigate the mechanisms by which this occurs [84][85][86]90]. Rotenone inhibits mitochondrial Complex I, causing oxidative stress. In flies, rotenone causes dose-dependent symptoms including motor deficits and selective DA neuron loss. Moreover, the PD medication L-dopa can be used to treat the motor deficits but not DA neuron loss [84]. Paraquat exposure in Drosophila also induces oxidative stress and DA neuron loss [85]. These changes are similar to those observed in post-mortem samples of paraquat-exposed PD patients [91,92].
Recent studies in Drosophila have demonstrated that paraquat exposure also leads to deregulated innate immune responses [90]. It remains unclear whether deregulation of the innate immune response is a primary contributor to neurodegeneration following paraquat exposure; however, activation of the innate response has been linked to neurodegeneration in other contexts (e.g., [93]).

Amyotrophic Lateral Sclerosis and Frontotemporal Dementia
ALS is characterized by the progressive degeneration of motor neurons. ALS is a relatively rare, but rapidly progressing, neurodegenerative disease that usually leads to death within 5 years of diagnosis. Familial ALS (FALS) accounts for~10% of ALS cases, while sporadic ALS (SALS) accounts for the remaining~90% [94]. A variety of genes have been implicated in ALS. Seven of these genes have been used to generate Drosophila ALS models (Table 3). These are: C9ORF72, TDP-43, FUS, VAPB, UBQLN2, VCP, and SOD-1. Overexpression, reduced expression, and expression of mutant versions of these genes have been used productively in Drosophila. A variety of assays also have been employed including measurement of lifespan, assessment of locomotion; examination of neuromuscular junction (NMJ) phenotypes; quantification of retinal degeneration and sensory neuron dendritic branching.
Frontotemporal dementia (FTD) is a group of disorders characterized by degeneration of the frontal and temporal lobes of the brain. It often has an early onset. Genes that contribute to FTD include: C9ORF72, FUS, VCP, TDP-43, MAPT/tau, CHMP2B, PGRN, TBK1, and TMEM106B (reviewed in [95]), thus there is overlap with other neurodegenerative diseases including: ALS (C9ORF72, FUS, VCP and TDP-43), AD (tau) and PD (tau). C9ORF72, FUS, VCP, and TDP-43 studies in Drosophila are included in Table 3, while tau studies are listed in Tables 1 and 2. Overexpression of Human Transgenes C9ORF72 Pan-neuronal expression of RNA-only constructs expressing (G 4 C 2 ) 106 repeats with both intronic (nucleus) and polyadenylated (cytoplasm) sense and antisense transcripts Pan-neuronal expression of UAS-RNA sense polyA constructs expressing 800-1000 and >1000 (G 4 C 2 ) repeats Adult Lifespan, negative geotaxis, light microscopy for eye morphology [104] Eye and pan-neuronal expression of UAS-(G 4 C 2 ) 3 and UAS-( G 4 C 2 ) 30   Lifespan, locomotion, number of motorneurons, neuronal SOD-1 accumulation, glial stress response [123] The most common contributing factor to ALS is a particular repeat expansion in the gene C9ORF72 that contains hundreds or thousands of the intronic hexanucleotide repeat (G 4 C 2 ) n [124,125]. Hexanucleotide repeat expansion (HRE) has been found in more than 5% of SALS patients and 39% of white American and European FALS patients, although it may be less common in other ethnic groups [126]. Repeat RNA can be neurotoxic. However, repeat-associated non-AUG (RAN) translation from these RNAs also can give rise to dipeptide repeat (DPR) proteins, which can be neurotoxic [127,128].
Several strategies have been used in Drosophila to introduce precise G 4 C 2 repeats and to investigate the potential neurotoxicity mechanisms (Table 3). In one study, as few as 30 repeats were found to be sufficient to cause neurodegeneration [105]. A subsequent study compared a variety of RNA-only expression methods, which was accomplished by interspersing stop codons that prevented dipeptide repeat (DPR) protein production. In this study, the RNA carrying the HRE did not result in toxicity, and the DPR proteins encoded by the hexanucleotide repeats were thought to mediate neurotoxicity [104]. Consistent with this, when the effects have been compared of expressing RNA that would code for different dipeptide combinations without using the G 4 C 2 motif, only arginine-containing DPR proteins were neurotoxic [107]. The Drosophila studies contrast with results in zebrafish where both DPR proteins and clusters of the mutant RNA are neurotoxic [129,130].
Both protein and RNA aggregates typically are observed in motor neurons of ALS patients post-mortem. Furthermore, these aggregates commonly contain both ubiquitin and TDP-43, thereby uniting multiple ALS genes in a common, proteostasis-defective, program. TDP-43 encodes the transactive response (TAR) DNA-binding protein, which can bind to both DNA and RNA. Mutations in TDP-43 account for~4% of FALS cases. TDP-43 protein is normally found in the nucleus, but localizes the cytoplasm in 90% of ALS patient samples. Indeed, cytoplasmic aggregates of TDP-43 are found iñ 90% of SALS brain and spinal cord specimens, making these aggregates one of the most reliable ALS diagnostics [94]. TDP-43 is a heterogeneous nuclear ribonucleoprotein (hnRNP) with demonstrated roles in transcription, mRNA splicing, and mRNA transport. It recognizes a UG-repeat sequence in target RNAs. Multiple mechanisms for TDP-43 toxicity have been proposed. One possibility is that TDP-43 functions in part by suppressing the incorporation of cryptic exons into mRNAs; without normal TDP-43 activity, aberrant transcripts are made that encode aberrant proteins. These RNAs and proteins overwhelm the degradation machinery, forming neurotoxic aggregates.
The Drosophila ALS models provide unique and powerful tools for understanding the etiology of ALS. Sophisticated genetic analyses that are not feasible in other systems have permitted the identification of both cell-autonomous and non-autonomous mechanisms of neurotoxicity [99] and facilitated analysis of protein vs. RNA contributions to neurotoxicity [104,107]. In addition, application of advanced genetic techniques has permitted the identification of interacting loci for known ALS genes [131][132][133][134][135]. These genetic interactions, in turn, have provided insights into the molecular mechanisms underlying neurodegeneration in ALS patients and provide a platform for the assaying of potential ALS therapeutics [136].

Huntington's Disease
Huntington's disease (HD), as with ALS, is caused by repeat expansion mutations. It involves a trinucleotide repeat that results in a polyglutamine (polyQ) segment of 36 or more units in the Huntingtin (Htt) protein [137,138]. It is usually diagnosed between the ages of 30 and 50, and symptoms include progressive problems of coordination, learning ability, decision making, and mood, although it can begin earlier in life with different signs and symptoms [139]. It is most common in European, North American, and Australian populations at 5.7 cases per 100,000 people, as compared to 0.4 cases per 100,000 people in Asian populations [140]. Life expectancy after diagnosis is only 17-20 years and no known treatments can slow the disease's progression, although some can address symptoms such as chorea [139,140].
Because the Drosophila Huntingtin (dHtt) does not have expanded polyQ in its amino terminus [141], most models of HD in Drosophila introduce the mutant human gene transgenically, and focus on large polyQ domains as versus studying the entire protein ( Table 4). Much of the variation in Drosophila models of HD is based on which segments of the Htt protein are expressed. In some studies, expression of fragments of the gene such as exon one only or the first three exons was used, while in others, large segments such as a 12-exon fragment or the sequence encoding the entire protein were used [142][143][144][145]. To introduce compounds of interest, inhibitors of polyQ aggregation such as QBP1 (polyglutamine binding peptide) and bivalent polyQ peptides have been transgenically expressed, genes such NMNAT (nicotinamide mononucleotide adenylyltransferase) have been overexpressed, HDAC (histone deacetylase) inhibitors have been administered through food, and polyQ aggregation inhibitors have been delivered using nanoparticles [144,[146][147][148]. Of course, alongside overexpression of genes such as NMNAT, which leads to reduction in mutant Htt aggregation by promoting autophagic clearance, loss-of-function mutations and conditional expression (after symptoms appear) have also proven useful for investigating pathology and treatment [148]. Treatment with HDAC inhibitors was shown to halt polyglutamine-induced toxicity and improve lethality. Moreover, assays including survival, photoreceptor quantification, circadian rhythmicity, and motor performance are conducive to screening for treatments or deficiency mutations [142,145]. While not a direct assay for neuropathology, circadian rhythm changes are strongly correlated with neurodegeneration in humans as well as animal models [149,150].  immunostaining for brain size [148] Expression of an mRFP-tagged N-terminal fragment of human Q15 or Q138 peptides containing exons 1-12

Adult
Immunofluorescence for spreading of Huntingtin aggregates in the brain [25] Expression of several

Ataxia Telangiectasia (A-T)
Mutations in the human Ataxia Telangiectasia Mutated (ATM) gene lead to a variety of pathologies, including increased risk of certain cancers, increased risk of infections, problems with motor control, and neurodegeneration [154]. ATM encodes an atypical protein kinase involved in the repair of double strand DNA breaks. The Drosophila homolog of ATM has several names, including dATM and tefu (telomere fusion) and was used to model the disease (Table 5). Similar to human patients, Drosophila carrying dATM/tefu mutations exhibit neurodegeneration. Furthermore, analysis of dATM/tefu function in Drosophila has provided critical insights into the mechanisms underlying neurodegeneration in A-T patients. Specifically, loss of kinase activity in glial cells was shown to lead to activation of the innate immune response and the death of both glia and neurons [155,156]. Furthermore, because activation of the innate immune response has been linked to neurodegeneration in multiple contexts [90,93,157,158], the Drosophila studies support the idea of a shared neurodegenerative mechanism underlying multiple disorders, including A-T and AD. More recently, the Drosophila model of A-T has been used to screen 2400 compounds for possible A-T therapeutics. These studies identified 10 lead compounds, including one that provided additional insights into the cellular mechanisms underlying A-T pathologies [159].

Mitochondrial Gene Mutations and Neurodegeneration
Mitochondrial dysfunction has long been associated with neurodegenerative diseases (reviewed in [160]). AD, PD, and HD, among others, are all known to be potentiated by defects in mitochondrial dynamics. This is perhaps not surprising when one considers that neurons have extremely high energy requirements [161]. More recently, forward genetic screens in Drosophila have been used to identify genes that are so critical to mitochondrial function in neurons that their mutation leads to neurodegeneration, even in the absence of other predisposing mutations (Table 6). We describe two examples here.
In 2017, the Bellen laboratory reported that mutations in Nardilysin (dNRD1) were defective in the folding of the mitochondrial enzyme α-ketoglutarate dehydrogenase, which is a rate-limiting enzyme for the Krebs Cycle [162]. The same group had previously shown that mutations in dNRD1 led to neurodegeneration in the Drosophila retina [163]. Upon further investigation, mutations in dNRD1 were found to lead to increased levels of α-ketoglutarate which impaired autophagy via an increase in mTORC1 activity [162]. The work was important both because it offered a molecular mechanism for the neurodegeneration observed in the mutants and a potential therapeutic target -mTORC1 -for neurodegenerative disorders caused or enhanced by mutations in dNRD1. Indeed, rapamycin was shown to alleviate the neurodegeneration caused by loss of either dNRD1 or OGDH mutations [162], providing a clear proof of principle for the approach. We note, however, that because mTORC1 exerts pleiotropic effects, it may not be an ideal target for neurodegeneration therapeutics.
In 2018, the Ganetzky laboratory reported on the identification of a new allele of the nuclear-encoded mitochondrial Complex I enzyme, ND23 [21]. Drosophila ND23 mutations cause progressive neurodegeneration and early death. Another Complex 1 protein, NDUFS8, previously had been implicated in a human disorder called Leigh Syndrome [164]. Leigh Syndrome manifestations include early, progressive neurodegeneration with loss of cognitive and motor function. A confusing aspect of Leigh Syndrome has been the variation in phenotype among patients carrying the identical mutant alleles of NDUFS8. Loewen and Ganetzky now offer both a Drosophila model for Leigh Syndrome and an explanation for Leigh Syndrome phenotypic variability. In particular, they found that the mitochondrial genotype modifies the severity of the neurodegeneration in ND23 mutants and identified a mutation in the mitochondrially encoded ATPase 6 as a strong candidate enhancer of ND23 mutations [21]. Like ND23 and NDUFS8, ATPase 6 is a component of Complex I. Another Complex I mutation in the Drosophila gene ND75, homolog of human NDUFS1, has been shown to contribute to neurodegeneration [165]. The work underscores the utility of Drosophila for understanding neurodegenerative disorders and provides a model for the testing of potential therapeutics.

The Brain Tumor Gene and Neurodegeneration
A novel genetic model for neurodegeneration recently was described in Drosophila [166] (Table 7). This model consists of a new mutation in the TRIM-NHL (tripartite motif-NCL-1, HT2A, and LIN-41) protein encoded by the Drosophila brain tumor (brat) gene. Mutations in human TRIM proteins have been associated with neuropathologies, including AD [167] and axonal neuropathy [168]. However, the new mutation, brat cheesehead (brat chs ), is unusual in leading to both brain tumors and progressive neurodegeneration. While deletions of TRIM3 frequently are found in primary human gliomas [169], the simultaneous presence of neurodegeneration and overgrowth is rare and raises the intriguing question of whether one leads to the other. Epidemiological studies have identified positive associations between PD and an increased risk of malignant brain tumors [170,171], while genetic studies have shown that mutations in the human E3-ubiquitin ligase-coding gene PARK2 are associated with several malignancies in addition to early onset PD [172]. Thus, it seems likely that brat chs flies may serve as an excellent model for identification of still unknown mechanisms underlying neurodegenerative diseases and for the testing of potential therapeutics.

Drosophila Traumatic Brain Injury and Neurodegeneration
In 2013, the first Drosophila model of closed-head TBI was published by the Wassarman and Ganetzky laboratories [173] (Table 8). Similar to humans, TBI in Drosophila leads to temporary incapacitation, ataxia, innate immune response activation, neurodegeneration and death [173]. The neurodegeneration is analogous to the CTE observed in human TBI patients. Over the intervening seven years, much has been learned about the factors influencing in TBI outcomes in Drosophila. These factors include age and diet as well as genetic background [174,175]. Being able to study the mechanisms underlying neurodegeneration in controlled genetic backgrounds is extremely powerful and already providing insights into both genetic and environmental variables that can contribute to neurodegeneration or to neuroprotection.
The standard TBI protocol in Drosophila involves four impacts spaced at 5-minute intervals. A standard outcome measure is the percent of injured flies that die within the first 24 h following the injury. A survey of more than 200 "wild type" Drosophila strains derived from a single wild type population [176], revealed that post-TBI mortality is influenced tremendously by genetic background with some strains exhibiting as little as 10% mortality and others exhibiting 60% mortality [175]. Moreover, using mortality as a measure, TBI outcomes were found to be worse in older adults than in younger adults [174].
Notably, restricting food intake after TBI was shown to have beneficial effects, paralleling TBI outcomes in humans, where increased hyperglycemia (e.g., as seen in patients with diabetes) is linked to significantly increased risk of death after TBI [175]. These results suggest that the secondary injuries leading to organismal death are similar in Drosophila and humans and that further studies in Drosophila are likely to provide additional new information that will help us understand the complex consequences of TBI [175].
Gene expression studies have permitted the identification of genes that are up-or down-regulated following Drosophila TBI. The up-regulated genes include components of the Drosophila innate immune system [174], some of which previously have been linked to neurodegeneration in Drosophila [93,157]. This raises the possibility that pharmacological control of innate immunity programs in human patients could reduce secondary injuries and therefore prevent adverse TBI outcomes.
In recent years, the utility of this model has become sufficiently clear that other laboratories are now using it [177][178][179][180][181][182]. Because of the parallels between Drosophila and human responses to TBI, there are multiple future applications for this model. These include evaluating the utility of various drugs in treating TBI in the clinic [183]. Future applications of this work will include testing the efficacy of these same pharmaceutical agents in preventing genetically induced neurodegeneration. In addition, because TBI patients often require surgery, not only for the head injury but also for other injuries sustained in parallel, the Drosophila model should also prove useful for the safety of individual anesthetics for TBI patients [184]. Table 8. Drosophila models of neurodegenerative traumatic brain injury.

Traumatic Brain injury Drosophila Model Developmental Stage Assay Used for Neuropathology References
Injury from the High-Impact Trauma device Adult Lifespan, histological staining for vacuole quantification [173] Stab injury to the brain through the right eye Adult Lifespan, climbing assay, mobility assay [183]

Looking Ahead
In summary, Drosophila are a powerful model in which to gain insights into human neurodegenerative disorders. Studies in Drosophila have made major contributions to our understanding of neurodegenerative disorders. For instance, the importance of mitochondria to PD was discovered in flies [69,76,80], as was the toxicity of arginine-containing dipeptide repeats in ALS [107]. Candidates from human GWAS can also be readily validated in Drosophila, as shown in [39]. The availability of Drosophila models for diseases such as AD, PD, and HD provides opportunities for the discovery of molecular mechanisms that affect disease progression and tools for the identification of therapeutics. More recently, Drosophila have emerged as a model in which to study neural regeneration. A variety of models have been created, including several in which to investigate axon or dendrite regrowth after injury as well as an adult brain model for the simultaneous analysis of degeneration and regeneration after brain injury (Table 9). An important direction for future research will be to apply what we are learning about neural regeneration to the neurodegenerative disease models to test whether we can slow or reverse specific types of neurodegeneration. For instance, one could imagine using CRISPR/Cas9 to correct a genetic defect in a subset of cells within the brain and then activating those cells to replace neurons and/or glia that were lost or damaged. Using this type of approach, one might first pre-empt Lewy body formation in neural stem cells, then coax those same stem cells to replace lost dopaminergic neurons in a PD model. These types of experiments could be done in Drosophila, paving the way for future studies in mammalian models and human patients. Because of the shared developmental genetics of the Drosophila and human nervous systems, it seems likely that Drosophila will prove as fruitful for modeling neural regeneration as it has for modeling neural degeneration. Visualization of regenerating proximal stump using GFP reporters [185][186][187] In vivo laser axotomy Larva Sensory neuron axons Visualization of degenerating distal stump using GFP reporters Visualization of axon regrowth using GFP reporters [185,[188][189][190] In vivo laser dendriotomy Larva Sensory neuron dendrites n.d.
Visualization of axon regrowth using GFP reporters [192] In vivo axon pruning and remodeling Pupa Mushroom body of the brain axon pruning and remodeling

Immunostaining of fixed samples
Immunostaining of fixed samples [198] Ex vivo axotomy Adult Brain sLN-v neurons Visualization of degenerating distal stump using GFP reporters Visualization of axon regrowth using GFP reporters [199] Olfactory neuron axotomy Adult Antennal olfactory neuron axons Visualization of degenerating distal stump using GFP reporters n.d. [196,200,201] In vivo axotomy Adult Wing sensory neuron axons Visualization of degenerating distal stump using GFP reporters Visualization of axon regrowth using GFP reporters [202][203][204] Traumatic Brain Injury (TBI) Adult Various brain regions histology n.d. [173,183] Penetrating Traumatic Brain Injury (PTBI) Adult Various brain regions TUNEL assays, Cell proliferation, lineage tracing [183,[205][206][207] Acknowledgments: Due to space limitations: we could not be comprehensive in describing all existing Drosophila neurodegenerative disease models. We therefore apologize for omissions. We would like to thank the reviewers for their meticulous reding of the initial version of the manuscript. This work was supported by NIH T32 GM007133 (KLC); NIH NS102698 (GBF); start-up funds from the University of Alabama (SC), the Merrymac/McKinley Foundation (SC) and the Women in Science and Engineering Leadership Institute (WISELI) (GBF).